Technical ReportPDF Available

Trans Mountain Expansion Project and Oil Spills: Power Analysis on Pacific Salmon Data

Authors:
  • Central Coast Indigenous Resource Alliance
Trans Mountain Expansion Project and Oil
Spills: Power Analysis on Pacific Salmon Data
Prepared for Adams Lake Indian Band
April 9, 2020
CONTACT:
Shannon Gavin
shannon.gavin@mses.c
a
403-241-8668
ADDRESS:
207 Edgebrook Close, NW
Calgary, AB, Canada
T3A 4W5
Photo by Marco Tjokro on Unsplash
TMX Salmon Research Study- Literature Review and Power Analysis
April 2020
Page i
List of Contributors
Dr. Megan Thompson Surface Water Quality
Dr. Sarah Alderman Fish Ecotoxicology
Dr. Kyle Wilson Fisheries Ecologist
Mr. Sebastian Dalgarno Computational Biologist
Dr. Joseph Thorley Computational Biologist
Ms. Shannon Gavin, M.Sc., P.Biol. Terrestrial Wildlife
Dr. Ave Dersch Traditional Resource Use
Mr. Logan Boyer Report Management
TMX Salmon Research Study- Literature Review and Power Analysis
April 2020
Page ii
Executive Summary
Pacific salmon that migrate through Kamloops Lake and into the North Thompson River and Shuswap
Lake complex are at risk from oil spills associated with the Trans Mountain Expansion pipeline project
(TMX). These watersheds and salmon are culturally important to Adams Lake Indian Band (ALIB)
members and the larger Secwépemc Nation (Kwusen Research and Media 2019). Salmon have an
important role in the ecosystem as they provide a nutrient and food source for traditional users, and for
a variety of plants and wildlife. Therefore, reductions in numbers or contamination of salmon can have
consequences for the broader ecosystem. ALIB requested that Management and Solutions in
Environmental Science (MSES) assist the community by conducting a technical study. The focus of the
study was to examine whether existing scientific data would be able to detect potential impacts on salmon
stocks in these watersheds if a spill were to occur associated with the TMX project.
The TMX project involves a twinning of the existing Trans Mountain pipeline with the construction of
Line 2. The TMX will transport diluted bitumen, other conventional crude oils and diluent. Unlike
conventional crude oil, bitumen is too thick to be pumped through pipelines, so it is diluted with either
conventional light crude oil or a mix of natural gas liquids, to give it a thinner consistency. This is called
diluted bitumen. There is less known about the toxicity of diluted bitumen to salmon compared to
conventional crude oil. There is also less known about oil spills in freshwater systems compared to marine
environments. Diluted bitumen has physical characteristics that make it more unpredictable in its
behaviour when spilled into a freshwater ecosystem, relative to other crude oil types. The outcome of
diluted bitumen exposure in salmon can include short term effects (acute effects) and long-term effects
(chronic effects). The magnitude of these effects on salmon depends on the concentration and duration
that fish are exposed to various constituents of the diluted bitumen.
The first step in our study was to review the existing scientific literature to provide background
information for our analysis and to identify gaps in the existing state of knowledge on oil spills and its
impacts to salmon. Critical knowledge gaps identified from this literature review are presented within the
context of assessing impacts to salmon populations due to a pipeline spill. Populations of salmon are
declining in the Fraser Basin and Thompson River Complex and the number of returning adults is highly
variable among sites within and between years. Although there are five species of salmon (Coho, Pink,
Chinook, Sockeye, and Chum) in the Fraser Basin watershed, our study focused on Chinook and Sockeye
salmon populations. These populations provide major fishery resources to the ALIB and may be affected
by the TMX project. The key difference between Chinook salmon and Sockeye salmon is their migratory
behaviour; Chinook salmon rear in streams and spend three years at sea while Sockeye salmon rear in
lakes and spend two years at sea. Given the differences in migratory behaviour of both these species, the
timing and location of any potential oil spill will determine the magnitude of the potential impact to each
species and each life stage (egg, fry, smolt, sub-adult, pre-spawning adult and adult).
The second step in our study was to complete a power analysis to test how precisely we could estimate
the size of a hypothetical impact using currently available fisheries data as a baseline. We considered an
extreme worst-case scenario wherein we assumed a large spatial extent for a spill and high concentrations
of toxic substances. Our analysis considered that this worst-case scenario could occur under three
different river flow conditions: high flow freshet, open water and winter under ice. The relative survival
of salmon following a spill was estimated for each life stage for both species, and a mechanistic population
TMX Salmon Research Study- Literature Review and Power Analysis
April 2020
Page iii
model was used to estimate resulting adult returns. We assumed that the spatial coverage from an oil spill
would extend through the North Thompson, Thompson complex and Kamloops Lake. We did not directly
assess impacts to salmon in the Shuswap Complex because based on our worst-case scenario, it was
assumed that the Shuswap Complex salmon would be less impacted less than the North Thompson
salmon. This does NOT mean that the Shuswap Complex salmon would be unaffected by an oil spill. We
then used an empirical model in a power analysis to determine whether we could detect the impact of an
actual spill from the available adult return data.
Below we provide a brief outline of key findings from our literature review and power analysis study.
Key Findings: Literature Review and Current State of Knowledge
- Knowledge gaps on what size/cohorts of juvenile salmon would use areas in the watershed that
could be vulnerable to an oil spill event.
- Details are needed on the natal origins of juvenile salmon that use areas of the watershed that
would be vulnerable during an oil spill event.
- Knowledge gaps regarding impacts to fish health from a bitumen spill include a lack of data on life
stage-specific and species-specific sensitivities to diluted bitumen. Also, how long-term effects may
influence salmon migration patterns and migratory behaviour.
- There is a lack of information regarding the transfer of contaminants in the food web, which has
implications for the health of the broader ecosystem.
- Consequences of a decline in prey availability can lead to corresponding population declines in
higher-order predators. Food web relationships are quite complex and there is still much more
information needed to understand these interactions.
- A full list of knowledge gaps can be found in the summary table in Section 2.9 of this report.
Key Findings: Modelling Spill Scenario Impacts to Salmon and Power Analysis
- Negative impacts on Chinook and Sockeye salmon survival are likely to be greatest under winter
conditions because the concentration of contaminants in the river water are at their highest under
low stream flow, and because young vulnerable salmon life stages are still overwintering in the
River.
- Based on our worst-case spill scenario, using our mechanistic model, we estimated the number
of returning adults which was most directly impacted by the spill. The results suggest that a worst-
case bitumen spill of 4,000 m3 of diluted bitumen from the TMX Project could cause as much as
a 71% and 73% reduction in the number of Chinook and Sockeye adult salmon, respectively,
returning to the main stem North Thompson. The number of adults returning to the tributaries
could be reduced by between 12% and 53% for Chinook and Sockeye, respectively.
- The power analysis considered whether available baseline adult return data for the North
Thompson River would allow us to detect an oil spill given a reduction of up to 71% and 73% of
returning adults.
- The existing fisheries data is too variable among sites within and between years to reliably detect
changes in the number of returning adults due to an oil spill.
Recommendations
- Our analysis and results do NOT mean that an impact would be small on salmon BUT that the
data on returning adults are not adequate to detect the impact of an oil spill on the number of
returning adults.
TMX Salmon Research Study- Literature Review and Power Analysis
April 2020
Page iv
- The current data on returning adults are not adequate to capture the variability in salmon
abundance at the local scale.
- We suggest the following research is needed that would help inform our understanding of impacts
from oil spills on salmon:
o More extensive and comprehensive monitoring of suspended sediments, water temperature,
and existing contaminants of potential concern.
o Characterize distribution, origin, and abundance of juvenile salmon within the watershed,
particularly for the smallest life history stages most vulnerable to spill impacts.
o Define the timing and locations of juvenile habitat use in the watershed.
o Establish reference values for effects from diluted bitumen exposure for different salmon
species and life stages.
o A full list of recommendations can be found in Table 11 in Section 4.1 of this report.
TMX Salmon Research Study- Literature Review and Power Analysis
April 2020
Page v
TABLE OF CONTENTS
PAGE
LIST OF CONTRIBUTORS .......................................................................................................... I
EXECUTIVE SUMMARY.............................................................................................................. II
ACRONYMS/ABBREVIATIONS ............................................................................................... IX
1.0 INTRODUCTION ............................................................................................... 12
1.1 Report Structure .......................................................................................................................... 13
2.0 PART A: CURRENT STATE OF KNOWLEDGE AND LITERATURE
REVIEW ............................................................................................................... 13
2.1 Study Area of Interest ................................................................................................................. 13
2.1.1 North Thompson River ................................................................................................ 16
2.1.2 Thompson River ............................................................................................................. 19
2.1.3 Kamloops Lake ............................................................................................................... 19
2.1.4 Groundwater .................................................................................................................. 19
2.2 Importance of Salmon in the Ecosystem .................................................................................. 20
2.3 Salmon in the Fraser River Basin ............................................................................................... 20
2.4 Salmon in the Area of Interest ................................................................................................... 22
2.5 Potential Risks of TMX to Salmon ............................................................................................ 22
2.6 Current State of Knowledge on Diluted Bitumen in Freshwaters ...................................... 23
2.6.1 What is diluted bitumen? .............................................................................................. 23
2.6.2 Diluted bitumen composition ...................................................................................... 23
2.6.3 Behaviour of diluted bitumen in freshwaters ............................................................ 24
2.6.4 Freshwater Oil Spill Case Studies ............................................................................... 26
2.7 Current State of Knowledge on Toxicity of Diluted Bitumen to Fish ................................ 27
2.7.1 Studying diluted bitumen toxicity in the laboratory ................................................. 27
2.7.2 Lethality of diluted bitumen exposure ........................................................................ 28
2.7.3 Sublethal effects of diluted bitumen exposure .......................................................... 31
2.7.4 Diluted bitumen toxicity in relation to Pacific salmon............................................. 33
2.7.4.1 Species and life stage sensitivities to diluted bitumen ............................................................................. 33
2.7.4.2 Anadromous life history ............................................................................................................................... 33
2.7.4.3 Behavioural responses to crude oil exposure .......................................................................................... 34
2.7.4.4 Contaminant uptake and clearance ............................................................................................................ 34
2.8 Current State of Knowledge on Impacts to Terrestrial Wildlife that Rely on
Salmon ............................................................................................................................................ 35
2.8.1 Changes in Food Availability ........................................................................................ 35
2.8.2 Changes in Food Quality .............................................................................................. 36
2.8.3 Reduced Habitat Quality............................................................................................... 37
2.8.4 Monitoring Impacts to Wildlife Following Oil Spills ................................................. 37
2.9 Summary of Knowledge Gaps .................................................................................................... 38
3.0 PART B: QUANTIFYING AND DETECTING IMPACTS TO
SALMON .............................................................................................................. 40
3.1 Defining Hypothetical Spill Scenarios and Impacts to Salmon .............................................. 40
TMX Salmon Research Study- Literature Review and Power Analysis
April 2020
Page vi
3.1.1 Hypothetical Spill Considerations ............................................................................... 41
3.1.1.1 Example of the TMX pipeline project spill risk assessment................................................................... 41
3.1.1.2 Spill Conditions Used for this Analysis ...................................................................................................... 44
3.1.2 Calculation of contaminant concentrations ............................................................... 46
3.2 Estimating Impacts on Salmon in the North Thompson Watershed................................... 51
3.2.1 Estimating Baseline Salmon Survival in the Area of Interest ................................... 51
3.2.1.1 Life Cycles ........................................................................................................................................................... 51
North Thompson Chinook ...................................................................................................................................... 51
North Thompson Sockeye ....................................................................................................................................... 53
3.2.2 Estimating Salmon Survival Under Hypothetical Spill Scenarios ............................ 54
3.2.3 Risks of Hypothetical Spills to Salmon in the Shuswap Complex .......................... 55
3.2.4 Estimating the Spatial Coverage of Impacts to Salmon ............................................ 56
3.2.5 Site-level impacts within the North Thompson Watershed .................................. 60
3.2.6 Results: What are the Impacts on Adult Salmon Returns? ..................................... 61
3.3 Power Analysis Estimating Impacts on North Thompson Salmon ..................................... 64
3.3.1 Power Analysis Model ................................................................................................... 64
3.3.2 Long-term Trends in the Fraser Basin ........................................................................ 67
3.3.3 Precision of Estimates ................................................................................................... 68
3.3.4 Results: With the available data, how precisely can we detect oil spill
impacts to salmon returns? .......................................................................................... 69
4.0 DISCUSSION AND CONCLUSIONS .............................................................. 70
4.1 Monitoring and Future Research Recommendations ............................................................. 71
5.0 LITERATURE CITED ......................................................................................... 73
APPENDIX A……………………………………………………………………………...…A1
List of Figures
Figure 1. Map illustrating the Fraser River Basin with our Area of Interest highlighted in blue
(Thompson watersheds). The existing and proposed TMX pipeline route overlays
our study boundaries. .................................................................................................................. 14
Figure 2. The Area of Interest for this study includes the North and South Thompson River
watersheds, and the mainstem Thompson River downstream to Kamloops Lake.
The TMX Line 2 route and facilities are also shown. Water Survey of Canada
(WSC) flow monitoring stations on the North Thompson River Mainstem are
shown. ............................................................................................................................................ 15
Figure 3. The North Thompson River daily discharge at Birch Island (08LB047) from 1960 to
2015. The grey ribbon indicates the minimum and maximum discharge, the
black line the median discharge and the blue line the 2015 discharge. ............................. 17
Figure 4. The North Thompson River daily discharge at McLure (08LB064) from 1960 to 2018.
The grey ribbon indicates the minimum and maximum discharge, the black line
the median discharge and the blue line the 2018 discharge. ............................................... 18
Figure 5. Illustration of the general life cycle of salmon including six life stages eggs, fry, smolt,
sub-adult, pre-spawning adult and spawning adult. .............................................................. 21
Figure 6. Time of onset and relative importance of weathering processes over time after an oil
spill into water. The onset and magnitude of effect will vary with temperature
TMX Salmon Research Study- Literature Review and Power Analysis
April 2020
Page vii
and for different oils (note the time scale, which emphasizes the early onset of
most processes) (figure and caption taken from Lee et al. 2015, Figure 2.4, p.77). ......... 25
Figure 7. Location of hypothetical spills from the proposed TMX pipeline initially considered for
the assessment of impacts to Chinook and Sockeye salmon. Hypothetical spill
location 1 represented the worst-case scenario of a release of 4,000 m3 of
diluted bitumen. Hypothetical spill location 2 represents the spill location
assessed in the TMX qualitative ecological risk assessment (QERA) (TMP ULC
2013, Volume 7), which involved a release of 1,400 m3 of diluted bitumen. .................... 43
Figure 8. Sockeye salmon populations could use the vulnerable areas of the North Thompson
watershed in different ways. To address this complexity in their lifecycle, we
grouped Sockeye salmon based on their use of the watershed (mainstem of the
river compared to two grouping types for tributaries. .......................................................... 58
Figure 9. Chinook salmon populations could use the vulnerable areas of the North Thompson
watershed in different ways. To address this complexity in their lifecycle, we
grouped Chinook salmon based on their use of the watershed (mainstem of the
river compared to two grouping types for tributaries. .......................................................... 59
Figure 10. The assumed spill coverage for Chinook by scenario, life history parameter and
grouping by watershed origin. Life cycle parameters on the x axis represent
transitions between life stages which includes pre-spawning adult to spawning
adult (P2A), fry to smolt (F2S), sublethal impacts to future fecundity for exposed
cohorts (F, number of eggs a female will lay) and egg to fry (E2F) (Table A1 in
Appendix A). .................................................................................................................................. 61
Figure 11. The assumed spill coverage for Sockeye by scenario, life history parameter and
grouping by watershed origin. Life cycle parameters on the x axis represent
transitions between life stages which includes pre-spawning adult to spawning
adult (P2A), fry to smolt (F2S), sublethal impacts to future fecundity for exposed
cohorts (F, number of eggs a female will lay) and egg to fry (E2F) (Table A1 in
Appendix A). .................................................................................................................................. 61
Figure 12. The total impact as the % reduction in returning adult abundance for Chinook by
scenario and grouping (given the assumed life-history parameters, impacts and
coverage). ...................................................................................................................................... 63
Figure 13. The total impact as the % reduction in returning adult abundance for Sockeye by
scenario and grouping (given the assumed life-history parameters, impacts and
coverage). ...................................................................................................................................... 64
Figure 14. The site-level spawner counts (plotted on a logarithmic scale) by spawner year and
conservation unit for Chinook salmon. ..................................................................................... 66
Figure 15. The site-level spawner counts (plotted on a logarithmic scale) by spawn year and
conservation unit for Sockeye salmon. ..................................................................................... 67
Figure 16. The estimated Chinook returning adults per recruit by year with annual variation in
the Fraser Basin. ........................................................................................................................... 68
Figure 17. The estimated Sockeye returning adults per recruit by year with annual variation in
the Fraser Basin. ........................................................................................................................... 68
Figure 18. The estimated impact of a diluted bitumen spill by year (with 95% confidence
intervals). ....................................................................................................................................... 70
TMX Salmon Research Study- Literature Review and Power Analysis
April 2020
Page viii
List of Tables
Table 1. Summary of published studies reporting lethal effects of diluted bitumen exposure in
fish. Dilbit refers to diluted bitumen. ........................................................................................ 29
Table 2. Summary of published studies reporting sublethal effects of diluted bitumen in fish.
Dilbit refers to diluted bitumen. ................................................................................................ 31
Table 3. Summary of Knowledge Gaps identified during Literature Review. .................................................. 38
Table 4. Spill scenarios considered for a spill of diluted bitumen into the North Thompson River
near Finn Creek Provincial Park (note: this location is upstream of WSC station at
Birch Island). Dilbit refers to diluted bitumen. ........................................................................ 46
Table 5. Summary of available water quality data for case study spills of crude oil and diluted
bitumen from pipelines into freshwaters, focusing on hydrocarbon constituents
(BTEX and PAHs) ........................................................................................................................... 48
Table 6. Estimates of hydrocarbon constituent concentrations (BTEX and PAHs) derived from
reported concentrations following the spill of diluted bitumen into the
Kalamazoo River in 2010. ............................................................................................................ 49
Table 7 Chinook Salmon adult abundance (N), fecundity, egg deposition, fry population
estimates, and egg-to-fry survival in the North Thompson watershed............................... 52
Table 8. Average survival and fecundity (number of eggs) of North Thompson Chinook and
Sockeye salmon across life stages (egg, fry, smolt, pre-spawning adult, spawning
adult) and plausible effects of diluted bitumen spills at highest concentration to
survival and fecundity of exposed cohorts. Note - specific Conservation Units
can vary in their own vital rates, and the sub-adult life stage is merged with pre-
spawning adults. On-route survival between pre-spawning and spawning adult
assumes water temperatures ≤16°C. ........................................................................................ 55
Table 9. The Chinook fry carrying capacity (K), number of returning adults at currently fished
equilibrium (Eq) and total impact as the additional mortality (M) by scenario and
grouping (given the assumed life-history parameters, impacts and coverage). ................ 62
Table 10. The Sockeye fry carrying capacity (K) number of returning adults at currently fished
equilibrium (Eq) and total impact as the mortality (M) by system and scenario
(given the assumed life-history parameters, impacts and coverage). ................................. 63
Table 11. Summary of monitoring and future research recommendations. ................................................... 72
LIST OF APPENDICES
Appendix A: Salmon in the North Thompson Watershed
Appendix B: The Impact of an Oil Spill on North Thompson Salmon 2019 (Attached)
TMX Salmon Research Study- Literature Review and Power Analysis
April 2020
Page ix
Acronyms/Abbreviations
AER
Alberta Energy Regulator
ALIB
Adams Lake Indian Band
ANSCO
Alaska North Slope crude oil
AWB
Access Winter Blend
BC
British Columbia
BTEX
benzene, toluene, ethylene and xylene
CABIN
Canadian Aquatic Biomonitoring Network
CCME
Canadian Council of Ministers of the Environment
Cl-
chloride Ion
CLB
Cold Lake Blend
CLWB
Cold Lake Winter Blend
COPC
contaminants of potential concern
CU
conversion unit
cyp1A
cytochrome P450 1A
dilbit
Diluted bitumen
dpf
days post-fertilization
DFO
Fisheries and Oceans Canada
EC50
the effective concentration at which 50% of the population
displays a sublethal endpoint of interest
EIA
Environmental Impact Assessment
EPA
United States Environmental Protection Agency
Eq
equilibrium
TMX Salmon Research Study- Literature Review and Power Analysis
April 2020
Page x
EROD
Ethoxyresorufin-O-deethylase
ES
early summer
FOSC
Federal On Scene Coordinator
FSC
food, social and ceremonial
g/L
grams per litre
K
carrying capacity
LC50
the concentration of a compound that is lethal for 50% of the
exposed population
M
mortality
m3
cubic metres
m3/s
cubic metres per second
MAH
monocyclic aromatic hydrocarbons
mg/g
milligrams per gram
mg/L
milligrams per litre
MSES
Management and Solutions in Environmental Science
N
abundance
Na+
sodium ion
NTSB
National Transportation Safety Board
NTU
Nephelometric Turbidity Unit
PAC
polycyclic aromatic compounds
PAH
polycyclic aromatic hydrocarbons
QERA
Qualitative Ecological Risk Assessment
TMP
Trans Mountain Pipeline
TMX
Trans Mountain Expansion
TMX Salmon Research Study- Literature Review and Power Analysis
April 2020
Page xi
TPAH
total polycyclic aromatic hydrocarbons
TPH-F
total petroleum hydrocarbons quantified using fluorescence
TSS
total suspended solids
USFWS
U.S. Fish and Wildlife Service
WAF
water-accommodated fraction
WCS
Western Canadian Select
wk
weeks
WSC
Water Survey of Canada
µg/L
micrograms per litre
µS/cm
microsiemens per centimetre
TMX Salmon Research Study- Literature Review and Power Analysis
April 2020
Page 12
1.0 Introduction
Anadromous (migrating up rivers from the sea to spawn) Pacific salmon are ecologically, economically and
socially important to the Fraser Basin. Salmon provide fishery resources to Indigenous peoples,
commercial fishers, and recreational anglers throughout the region. The resilience of Pacific salmon,
including Sockeye and Chinook, has been jointly challenged in recent decades by environmental stressors
such as habitat loss and conversion, climate changes, overfishing, and pollution (Cohen, 2012). Cumulative
effects from these stressors continue to impact freshwater, coastal, and marine ecosystems. As a
consequence of these multiple challenges, Pacific salmon fisheries have declined in recent decades
throughout much of British Columbia (Walters et al. 2019). Spatial impacts from different fishing
communities on salmon populations can vary from one another. First Nations and Indigenous food, social,
and ceremonial (FSC) fisheries often operate locally with “terminal-based” fishing in nearby streams,
rivers, and lakes. For efficiency, commercial fisheries tend to operate at the “mixed-stock” level, fishing
either the open ocean or coastline during larger aggregations of migrating adults. Recreational anglers can
fish at both aggregate- (e.g., trolling from boats on BC coasts) and local-scales (e.g., fly-fishing in streams).
Climate changes affect freshwater and marine habitats in different ways. Pacific decadal oscillations, El
Nino and La Nina events, and other oceanic drivers can alter conditions for marine growth and survival.
Climate changes can also alter precipitation (e.g. rain and snow) and other factors that affect the flow and
temperatures of freshwater habitats. Activities from ongoing and new development projects are a
concern, given the existing stressors to this fishery.
Adams Lake Indian Band (ALIB) have expressed their concerns with the proposed Trans Mountain
Expansion pipeline project (TMX) which has the potential to impact Pacific salmon in the North Thompson
River and Shuswap Lake complex in a variety of ways, including impacts associated with potential oil spills.
These watersheds and salmon are culturally important to ALIB members and the larger Secwépemc
Nation (Kwusen Research and Media 2019). There is an existing Trans Mountain Pipeline (TMP) on the
landscape which was built in 1953 and continues to operate today. The TMX Project will twin this existing
pipeline through the construction of Line 2. Currently the pipeline transports liquid hydrocarbon
products, including crude oil and diluted bitumen. The location of the TMX Project in the North
Thompson River valley, and crossing the Thompson River upstream of Kamloops Lake, will continue to
be a cause for concern to land users and a source of risk to salmon. A paramount concern is a pipeline
leak or spill that results in contamination of these water bodies. The increased capacity of the TMX Project
compared to the existing TMP, along with the uncertainty around the impacts of diluted bitumen spills
into freshwater, escalates concerns already voiced by Indigenous groups.
The goal of this report is to provide a comprehensive review of potential impacts and risks associated
with oil spills to salmon stocks (e.g. Sockeye, Chinook), within the North Thompson River, Thompson
River and Kamloops Lake area of the Fraser Basin. To do this, we have identified the following research
questions.
Research Questions:
1) What is the state of knowledge about impacts of oil and/or diluted bitumen spills on Chinook and
Sockeye salmon in freshwater systems, and specifically in the Thompson River complex?
2) Given specific hypothetical spill conditions, what are the potential impacts of a spill of a diluted
bitumen product in the North Thompson River on Chinook and Sockeye salmon?
TMX Salmon Research Study- Literature Review and Power Analysis
April 2020
Page 13
3) Using the available DFO data on numbers of returning adults, how precisely would we be able to
quantify such an impact on the number of returning Chinook and Sockeye salmon adults?
Technical experts reviewed publicly available literature and empirical data related to freshwater oil spills,
fish toxicity impacts, salmon populations, habitat use, distribution, and movements in the Fraser Basin. A
hypothetical spill scenario was considered, and the potential effects on adult salmon returns was estimated
using a population model. Quantitative population-based power analyses were completed to determine
how much baseline data is required in order to detect a particular effect or level of impact from an oil
spill. Understanding the adequacy of the current baseline data on salmon and its ability to be used in
detecting impacts to these key resources is essential for the management and protection of this
ecosystem.
1.1 Report Structure
This report is organized into two sections. Section 2.0 is a literature review of the current state of
knowledge for salmon in the Thompson River complex and the known adverse effects of diluted bitumen
exposure on fish (Research Question #1). Critical knowledge gaps identified from this literature review
are presented within the context of assessing impacts to the salmon population in the event of a pipeline
spill. Section 3.0 comprises an analysis of a hypothetical spill scenario. Here, an estimation of the nature
and magnitude of the impact of a spill on salmon returns was estimated using a salmon population model
(Research Question #2), and a statistical power analysis was used to determine whether the currently
available baseline salmon data could be reliably used to identify such an impact (Research Question #3).
2.0 Part A: Current State of Knowledge and Literature
Review
2.1 Study Area of Interest
Our study focused on the Thompson watershed within the larger Fraser Basin watershed in British
Columbia (BC), Canada (Figure 1). The TMX pipeline route closely follows the North Thompson River
for about 275 km from south of Albreda, BC to just west of Kamloops, BC, where it crosses the
Thompson River upstream of Kamloops Lake (see Figure 2 below). The pipeline route deviates from the
North Thompson River course briefly at the Clearwater, BC and Barrière, BC areas. The pipeline crosses
the North Thompson River at least three times along this course, and also crosses tributaries of the North
Thompson, including the Albreda, Blue and Clearwater Rivers (TMP ULC 2013, Volume 5C, Table 5.2, p.
5-4 and 5-5). South of Kamloops Lake, the pipeline route leaves the Thompson River mainstem and crosses
through a portion of the South Thompson River Watershed and the Nicola River watershed before
rejoining the Fraser River Mainstem at Hope, BC. The area of interest for this study includes the North
Thompson, South Thompson and Thompson Rivers, as well as, Kamloops Lake.
TMX Salmon Research Study- Literature Review and Power Analysis
April 2020
Page 14
Figure 1. Map illustrating the Fraser River Basin with our Area of Interest highlighted in blue
(Thompson watersheds). The existing and proposed TMX pipeline route overlays our study
boundaries.
TMX Salmon Research Study- Literature Review and Power Analysis
April 2020
Page 15
Figure 2. The Area of Interest for this study includes the North and South Thompson River watersheds,
and the mainstem Thompson River downstream to Kamloops Lake. The TMX Line 2 route and facilities
are also shown. Water Survey of Canada (WSC) flow monitoring stations on the North Thompson River
Mainstem are shown.
TMX Salmon Research Study- Literature Review and Power Analysis
April 2020
Page 16
Our Area of Interest does not include the smaller drainages south of Kamloops or the Nicola River
watersheds for two reasons:
1. The stated interest in terms of potential impacts of a pipeline failure on salmon applies primarily
to Thompson River salmon returns, and specifically in the Adams River and Adams Lake areas.
Although these areas are within the Fraser River watershed, they constitute separate salmon
conservation units/populations.
2. The potential for a pipeline failure and oil release that might occur south of Kamloops to reach
the South Thompson and/or Thompson Rivers is less likely than a failure that might occur along
the North Thompson River or at the Thompson River, as the watercourses and their valleys are
smaller, and in some cases ephemeral, and there are several intervening lakes that can act as
collection basins for spilled oil. It should be noted that impacts to these aquatic ecosystems from
such a spill could be significant, but the potential for rapid transport of a spill downstream is
somewhat moderated.
2.1.1 North Thompson River
The North Thompson River is a free-flowing system with a large contribution of meltwater coming from
glacial or snow melt (Nordin 1993). Shoreline habitat surveys of the lower North Thompson River
completed in the 1990’s characterize the substrate as mainly fine sand gravels, with steep and often
unstable banks (Hickey & Trask 1994). Vegetation, both riparian and instream, was not abundant except
for macrophytes in side-channels and embayments where flows were relatively slow and open water
temperatures might have been higher. High streamflow velocity and significant bedload sand transport and
deposition were also noted in the mainstem (Hickey & Trask 1994). The North Thompson River has a
moderate gradient for a large river, including an elevation change along the TMX route of 397 m over 275
km, with an average slope of approximately 0.14%.
The Water Survey of Canada (WSC) has two recently or currently active water flow monitoring stations
on the North Thompson River: Birch Island (# 08LB047) and McLure (# 08LB064). Several tributaries are
also monitored for stream flow at their mouths. Streamflow in the Thompson River is monitored at
Kamloops (# 08LF023) and downstream of Kamloops Lake at Spences Bridge (#08LF051). At Birch Island,
maximum historical discharge from 1960-2015 was about 950 m3/s, recorded in July, while the median
maximum for the same period was about 475 m3/s and the minimum was about 15 m3/s, recorded during
the winter low flow (December through March). During any given year, localized peak discharges can
occur, likely associated with storms (Figure 3).
Further downstream at McClure, the maximum historical discharge is much higher, at about 2750 m3/s
recorded in June between 1958 and 2018. The median maximum historical discharge at McClure is about
1450 m3/s and the minimum discharge is less than 50 m3/s during the winter low flow. Localized peaks in
discharge are also apparent at this station but are more moderate compared to the maximum late
spring/early summer discharges (Figure 4).
TMX Salmon Research Study- Literature Review and Power Analysis
April 2020
Page 17
Figure 3. The North Thompson River daily discharge at Birch Island (08LB047) from 1960 to 2015. The
grey ribbon indicates the minimum and maximum discharge, the black line the median discharge and
the blue line the 2015 discharge.
TMX Salmon Research Study- Literature Review and Power Analysis
April 2020
Page 18
Figure 4. The North Thompson River daily discharge at McLure (08LB064) from 1960 to 2018. The grey
ribbon indicates the minimum and maximum discharge, the black line the median discharge and the
blue line the 2018 discharge.
An analysis of future changes in land use combined with climate change in the North Thompson River
watershed indicated that mean annual flow is expected to increase by as much as 55% by 2070, and that
peak flows were expected to occur earlier in the spring and be 20-30% higher by 2050 relative to 1981-
2010 peak flows (Carlson et al. 2019). These changes in flow were predicted to occur mainly as a result
of climate change, specifically increased precipitation, and were predicted for the North Thompson River
and its tributaries. The authors point out that the “design flows” of infrastructure that crosses or is in
close proximity to these watercourses, including pipelines, should take into account the potential for
increases in peak flow resulting from climate change.
Available provincially collected water quality data for the North Thompson River are sparse, and the most
recent data are in some cases decades old. At Birch Island (# 0600025) in July 1993, total suspended solids
(TSS) was 11 mg/L and specific conductance was 56 µS/cm. At McLure (site # 0600002), in April 1980,
TSS data are not available but turbidity was low at 2.8 Nephelometric Turbidity Units (NTU) and specific
conductance was 106 µS/cm. Historical average TSS at McLure (1967-1977) is 44 mg/L in June. At North
Kamloops (site # 0600164), much more recently collected samples from February 2020 indicate a specific
conductance of 131 µS/cm, but no TSS or turbidity data are available. (data obtained from
TMX Salmon Research Study- Literature Review and Power Analysis
April 2020
Page 19
https://governmentofbc.maps.arcgis.com/apps/webappviewer/index.html?id=0ecd608e27ec45cd923bdcfe
efba00a7, retrieved March 2020).
2.1.2 Thompson River
The Thompson River begins at the confluence of the South and North Thompson Rivers, and flows
approximately 15km into Kamloops Lake, below which it continues to the Fraser River. At their
confluence, the North Thompson River contributes about 60% and the South Thompson River about 40%
of mean annual flow in the Thompson River (Nordin & Holmes 1992). Shoreline fish habitat surveys of
the Thompson River upstream of Kamloops Lake noted extensive sand substrates in a predominantly
straight channel (Hickey & Trask 1994). The flow is generally smooth, with some back-channel areas.
Substrates were mainly fine particles (clay, mud, sand) (71%) or man-made with a small percentage of
gravel and cobbles (13%). Fine particles also made up the majority of streambank materials along the
Thompson River upstream of Kamloops Lake. Instream vegetation was recorded in this stretch of the
River, but not in great abundance (Hickey & Trask 1994).
2.1.3 Kamloops Lake
Kamloops Lake is a long, narrow water body that is heavily influenced by Thompson River inflows (Nordin
& Holmes 1992). Inflow turbidity peaks during high flows at the freshet period (high flow, early summer),
and the lake acts as a sink for particulate material (Carmack et al. 1979). Documented historical water
quality impacts on the lower Thompson River from discharge points upstream of Kamloops Lake further
indicate the flow-through nature of the Lake and interconnectedness with the Thompson River (Nordin
& Holmes 1992).
Kamloops lake is about 25 km long, 2.1 km wide and has a maximum depth of 143m (Nordin & Holmes
1992, Carmack et al. 1979). The inflow of the Thompson River upstream represents a sizeable fraction of
the lake volume, so that bulk residence times are very short (the annual average is 60 days), and lake levels
fluctuate significantly during periods of high river flow (Nordin & Holmes 1992, Carmack et al. 1979). The
inflow water from the Thompson River enters Kamloops Lake and either moves through the lake mainly
at the surface (in late spring and winter), or at a middle depth according to temperature stratification (in
summer and early autumn), or along the bottom of the lake (during turnover conditions in spring and
autumn)(Carmack et al. 1979).
2.1.4 Groundwater
There are mapped aquifers (i.e. underground layer of water-bearing permeable rock, rock fractures,
gravel, sand, or silt), along the pipeline route in the area of interest. These are generally located at certain
points along the North Thompson River (e.g., Blue River and Clearwater River confluences), and clustered
around the Thompson River and lower South and North Thompson Rivers near Kamloops. There is
limited groundwater data available for the northern North Thompson River valley (TMP ULC 2013,
Volume 5C, Waterline Resources Inc. 2013).
According to the TMX Project Environmental Impact Assessment (EIA) (TMP ULC 2013), many of the
aquifers along the North Thompson River are considered vulnerable to contamination. Aquifers may be
considered vulnerable because they are shallow, porous and/or have high-permeability confining layers,
among other factors (Liggett et al. 2011). This is important when considering how constituents from an
oil spill might be transported through salmon habitats, as shallow aquifers are often connected to surface
water and may serve as a pathway for contamination.
TMX Salmon Research Study- Literature Review and Power Analysis
April 2020
Page 20
2.2 Importance of Salmon in the Ecosystem
Anadromous (migrating up rivers from the sea to spawn) Pacific salmon are ecologically, economically and
socially important to the Fraser Basin. Salmon provide fishery resources to Indigenous peoples,
commercial fishers, and recreational anglers throughout the region. Indigenous people in this region have
fished for Pacific salmon since time immemorial, with archeological evidence of at least 9,000 years of
food, social, and ceremonial fishing practices (Diaz 2019). Pacific salmon provide critical linkages between
marine, coastal, freshwater and terrestrial ecosystems throughout the region by providing large pulses of
biomass to the flora and fauna of the region. They are prey and predators to many aquatic and terrestrial
animals including aquatic invertebrates, fishes, birds, bears, and marine mammals.
For terrestrial wildlife, the availability of salmon as a food source has implications for their overall fitness
in terms of growth rates, litter sizes and reproductive timing and success (Ben-David 1996, Hilderbrand
et al 1999). Higher salmon consumption by coastal bears has been shown to have improved their body
condition, reproductive success and population densities compared to areas without access to salmon
(Hilderbrand et al 1999, Service et al 2018). Population dynamics of other organisms are also influenced
by the availability of salmon. For example, during the breeding season, insectivorous passerines (birds)
have been found in greater densities along salmon streams compared to other streams (Gende et al 2002).
This is thought to be associated with higher invertebrate densities feeding on the salmon carcasses (Gende
et al 2002). Therefore, fluctuations in the availability of salmon to this ecosystem has ramifications for
other species within the overall food web and broader ecosystem.
In addition, salmon have a role in contributing to the recycling of nutrients into the surrounding riparian
and upland habitat via decomposition processes (Reimchen et al 2002). Increased nitrogen and
phosphorus from decomposition support algae and primary production and contributes to soil nutrients
and riparian vegetation growth (Gende et al 2004). Bears and other predator foraging activities often
result in fish carcasses being left within the riparian and upland habitat (Quinn et al 2009). Up to 50% of
the salmon biomass of returning adults may actually end up on the ground as a result of bear foraging
activities (Gende et al 2004) which has benefits for the overall productivity of the ecosystem.
2.3 Salmon in the Fraser River Basin
All five Pacific salmon (e.g., Coho, Pink, Chinook, Sockeye, and Chum) and sea-run trout (e.g., Steelhead
and Coastal Cutthroat Trout) are present in the Fraser River Basin, but this report will focus on two,
Sockeye salmon (Oncorhynchus nerka) and Chinook (Oncorhynchus tshawytscha) salmon, as they provide
major fishery resources to ALIB and may be affected by the TMX project. Sockeye and Chinook salmon
have tremendous life history diversity throughout the Fraser Basin including the North Thompson
watershed, the latter of which is the focus of this report.
The life cycles of Chinook and Sockeye salmon can both be broken into six life stages eggs, fry, smolt, sub-
adult, pre-spawning adult and spawning adult (Figure 5). Eggs, fry, smolt, and sub-adults can all be
considered juveniles (i.e., reproductively immature), but sub-adults reside only in the ocean. Salmon
undertake at least four major migrations throughout their lives three of which must be understood to
assess potential TMX impacts on local salmon populations. First, salmon fry migrate downstream from
natal streams into higher order streams or rivers for growth and rearing as they develop into smolts. For
Sockeye salmon, fry migrate downstream into rearing lakes and, for Chinook salmon, fry migrate
TMX Salmon Research Study- Literature Review and Power Analysis
April 2020
Page 21
downstream into larger streams or rivers. Second, salmon smolts migrate downstream to the estuary and
open Pacific Ocean and develop into sub-adults. Third, salmon sub-adults migrate across the Pacific Ocean
for growth and maturation and return to BC coastal waters as pre-spawning adults (this migration is not
considered relevant to assessing TMX impacts). Last, pre-spawning adults migrate upstream to their natal
streams and lakes as spawning adults.
Figure 5. Illustration of the general life cycle of salmon including six life stages eggs, fry, smolt, sub -
adult, pre-spawning adult and spawning adult.
Hence, both Sockeye and Chinook extensively rely upon interior BC freshwaters as juveniles for growth
and rearing the same freshwaters susceptible to human impacts like TMX. Each of these migration
timings and usage of lakes and rivers can vary by species and natal Sockeye and Chinook populations. In
the open ocean, sub-adult salmon grow and mature over several years before returning to their natal
streams and lakes in the Fraser Basin for spawning, but this can also vary by cohorts. For example, ~90%
of Fraser sockeye adults return at four-years old while 10% return at five years old, and this ratio varies
by population (DFO 2018a).
Sockeye and Chinook salmon also exhibit spatial complexity in their movement across and usage within
Fraser Basin watersheds. Sockeye and Chinook salmon use upper headwater streams, tributary rivers,
large mainstem rivers, and lakes throughout the Fraser Basin. Overall, the Fraser Basin provides an
important spawning and rearing area for sockeye globally, with over 150 natal areas that constitute 24
Conservation Units (CU) - ecologically and genetically unique populations that are defined by the
Department of Fisheries and Oceans (DFO) for fisheries management. These CUs typically vary in
spawning timing, spatial usage within the Basin, and rearing areas. Chinook Salmon in the Fraser Basin
constitute 19 CUs, which group Chinook populations together based on geographic distribution, duration
of freshwater residence as juveniles, oceanic distribution and dispersal, and timing of spawning migration
(Burgner 1991). One-third of Chinook CUs are ocean-type, including upper Adam’s River and Shuswap
(Riddell et al. 2013) with the rest being stream-type. Since the 1950s, DFO monitors adult escapement
(the number of returning spawning adults) for all CUs in the Basin of both Sockeye and Chinook, but
sampling is not always at the scale of local spawning streams and rivers. In other words, sampling may or
TMX Salmon Research Study- Literature Review and Power Analysis
April 2020
Page 22
may not occur at the scale of spawning sites, with the result that there may be a mismatch between the
monitoring data and the ecological dynamics of the salmon populations. This data is used year-to-year to
adjust commercial and recreational fisheries management (catch) targets.
2.4 Salmon in the Area of Interest
Sockeye and Chinook salmon populations vary in how they use and move across the North Thompson
watershed. These spatial usage patterns can be categorized by whether and how much salmon use three
key habitats: (1) tributaries, headwater streams, and lakes outside of the TMX pipeline corridor, (2) the
mainstem North Thompson River that the TMX pipeline follows, and (3) Kamloops Lake, where the North
Thompson and South Thompson Rivers converge and the last catchment area for potential contaminants
prior to higher-gradient downstream flows of the Thompson River. Given that there are dozens of local
spawning and natal Sockeye and Chinook populations within the North Thompson Watershed that vary
in local spawning and natal streams, we focused on reviewing how each of these populations might vary
in terms of how many and which life stages use each of the above three habitats.
Sockeye salmon life histories were represented, for our purposes, in three unique groupings each with
variable life histories (see details in Section 3.2.4). Tributary populations were defined as Sockeye
populations where adults spawn in tributaries of the North Thompson River, but where fry rear in
Kamloops Lake (as part of the North Thompson Conservation Unit). Mainstem populations were defined
as sockeye populations that spawn in the mainstem North Thompson River and that rear in Kamloops
Lake (as part of the North Thompson Conservation Unit). The last grouping were populations of the
Barriere Conservation Unit where adults still pass through the mainstem North Thompson River on their
way to spawning in the Barriere River, and fry rear in Barriere Lake. Juvenile Sockeye (from eggs to fry to
smolts) spend approximately 18-24 months in freshwater (two springs, one winter) most of this time is
spent in their rearing lake, e.g., Kamloops Lake. In total, most Sockeye salmon in the North Thompson
watershed express a four-year life cycle spending ~2 years in freshwater and ~2 years at sea.
Chinook salmon life histories were represented with two general groupings (see details in Section 3.2.4).
First, the mainstem North Thompson River population where adult Chinook spawn in the mainstem river
and fry also rear in the mainstem. The second grouping were the six tributary populations (Louis Creek,
Finn Creek, Barriere River, Clearwater River, Mahood River, and Raft River) where adults pass through
the North Thompson River on their way to spawn in their natal tributaries. Fry from tributary populations
then migrate downstream, and some will rear in the North Thompson River. Juvenile Chinook (from eggs
to fry to smolts) spend approximately 18-24 months in freshwater (two springs, one winter) most of
this time will be spent in their rearing stream or river (e.g., North Thompson River). Most Chinook salmon
in the North Thompson watershed express a five-year life cycle spending ~2 year in freshwater and ~3
years at sea.
2.5 Potential Risks of TMX to Salmon
The TMX may pose a significant risk to the Sockeye and Chinook populations along the Fraser Basin by
crossing much of their freshwater and estuarine habitats. The expansion to the existing pipeline follows
and crosses several major rivers and tributaries of the Fraser Basin including the North Thompson River
and just upstream from large Sockeye rearing lakes like Kamloops Lake. The pipeline may pose localized
and downstream risks to salmon populations that include environmental contamination from oil spills, and
TMX Salmon Research Study- Literature Review and Power Analysis
April 2020
Page 23
habitat deterioration from infrastructure development and runoff. The forms and spatial extent of impacts
from the TMX may vary due to the upstream and downstream connectivity of surface water, groundwater,
nutrients, thermal energy (water temperature), and salmon throughout the Fraser Basin and nearby coastal
waters. These impacts can include spills or disturbances that are localized to a few populations or
watersheds or that broadly affect the entire portfolio of salmon populations throughout the Fraser Basin.
2.6 Current State of Knowledge on Diluted Bitumen in
Freshwaters
2.6.1 What is diluted bitumen?
Bitumen is a “heavy” type of crude oil found in rich deposits throughout the oil sands region of Western
Canada. Bitumen differs from conventional crude oils, which are classified as “light”, because bitumen is
very dense, thick, and sticky. The peanut butter-like consistency of raw bitumen makes it impossible to
pump through pipelines. Therefore, before bitumen extracted from oil sands can be transported by rail
and pipeline to oil refineries, it must first be mixed with lighter petroleum products to make it more liquid
and able to flow. This modified crude oil is called diluted bitumen (or dilbit) and is one of the main products
that is transported in the TMP.
2.6.2 Diluted bitumen composition
Diluted bitumen, like all crude oils, is a complex mixture of thousands of chemicals. The contaminants of
potential concern present in crude oil include aliphatic (non-aromatic) hydrocarbons, naphthenic acids,
and aromatic hydrocarbons. Each of these contaminant groups contains hundreds of unique chemicals.
The toxicity of crude oil to fish is attributed to the aromatic hydrocarbon fraction. It is widely accepted
that aromatic hydrocarbon toxicity in fish and other organisms has many root causes, and the nature of
the toxic response depends on the specific chemical(s) involved, their concentrations, and the length of
exposure. Aromatic hydrocarbons are divided into two groups:
Monocyclic aromatic hydrocarbons (MAH) are the largest component of the volatile (easily
evaporated) organic compounds in crude oil and are defined by having a chemical structure of a
single 6-carbon benzene ring. Benzene, toluene, ethylene, and xylene (BTEX) comprise ~90% of
the MAH in crude oil, and are considered acutely toxic to fish. The BTEX fraction dissipates
quickly (24-96 h) via evaporation, and is a main driver of the acute toxic response to crude oil
exposure (Kennedy, 2015).
Polycyclic aromatic hydrocarbons (PAH; also frequently referred to as polycyclic aromatic
compounds, PAC) are a diverse group of chemicals that contain 2 or more fused benzene rings.
PAH can comprise as much as 10% of the total weight of diluted bitumen. The specific toxicity
of a single PAH to fish is influenced by its molecular weight and solubility (ability to dissolve).
Small molecular weight PAH (2-3 ring) are also a major contributor to the acute toxic response
to crude oil exposure. These and other PAH compounds are also responsible for longer term
chronic effects. Given the important roles that PAH play in crude oil toxicity, the sum total of
measurable PAH (in µg per liter of oil) is typically used to describe crude oil exposure
concentrations.
TMX Salmon Research Study- Literature Review and Power Analysis
April 2020
Page 24
2.6.3 Behaviour of diluted bitumen in freshwaters
The fate of spilled oil products in freshwater ecosystems has not been as thoroughly studied as in marine
environments. This information is particularly scarce for diluted bitumen products. Conventional crude
oils generally do not sink in water, making the recovery after a spill more straightforward. Because diluted
bitumen appears to have a greater propensity to sink to the sediments in aquatic systems, its specific
impacts may be different from those of a conventional crude oil spill and its recovery may be more
complex. A federal government status report issued in April 2018 indicates that the Government of
Canada is funding ongoing studies of diluted bitumen in several federal departments, as is the oil and gas
industry (DFO 2018b). That status report included a brief review of nine spills of oil into freshwater
systems as well as one brackish estuary, of which four involved diluted bitumen products. Thirty-six
laboratory experimental treatments were also reviewed in the report, all of which were conducted by
Canadian or United States federal agencies. These reviews did not involve summary or comparative
analyses that linked experimental findings to real-world observations, and the experimental findings had
highly variable results. Generally, it was found that suspended sediments in water influenced the formation
of oil-particle aggregates, and therefore sinking potential, although this varied with sediment type.
The status report authors indicated that the most important aspect of spill response is timing, and that a
prompt response is the most important determinant of the ultimate fate of diluted bitumen spilled into
aquatic systems (DFO 2018b). Conventional spill response measures were also generally seen as effective
for diluted bitumen spills, although it was acknowledged that more rapid evaporative weathering processes
of diluent and the inherent heavy nature of the bitumen itself means that changes in viscosity and density
of diluted bitumen are more rapid than for conventional crude oils. The window of opportunity for
application of oil recovery and spill countermeasures aimed at the water surface can range from hours to
weeks, depending on the product and the site-specific conditions. In the case of diluted bitumen, this
window is generally shorter, requiring a more rapid response before the properties of the oil change
significantly. A helpful summary of the time of onset and relative importance of weathering mechanisms
for conventional oil spilled into water are also shown in Figure 6 below. It is apparent from this figure that
evaporation (particularly important for diluent), dissolution and dispersion are important short-term
processes.
TMX Salmon Research Study- Literature Review and Power Analysis
April 2020
Page 25
Figure 6. Time of onset and relative importance of weathering processes over time after an oil spill
into water. The onset and magnitude of effect will vary with temperature and for different oils (note
the time scale, which emphasizes the early onset of most processes) (figure and caption taken from
Lee et al. 2015, Figure 2.4, p.77).
The status report also highlighted operational tools that were still required to enable the findings of
scientific research to be integrated into spill response planning, preparation and operation, including more
robust and comprehensive predictive models, summaries and comparisons between categories of known
diluted bitumen blends (DFO 2018b).
Ongoing scientific research has included several studies that have attempted to predict how sunken
bitumen may interact with the bottom of rivers and streams, in particular gravel beds, and whether it is
likely to become embedded, or to travel downstream with the bedload of a watercourse. The findings
have highlighted the importance of water temperature and stream velocity in determining where sunken
bitumen will accumulate and/or become mobile.
Apart from these general observations, the status report identified several knowledge gaps regarding the
fate and behaviour of diluted bitumen that is spilled into aquatic systems. These include:
- The fate and behaviour of diluted bitumen under low temperature and ice conditions;
- Physical, chemical and environmental processes that most influence diluted bitumen fate and behaviour;
- Natural weathering processes;
- Impacts of degradation and weathering on toxicity;
- The vulnerability of species to diluted bitumen blends;
- Methods to detect, track and monitor product movement when spilled;
- Processes for the formation and breakup of oil-mineral aggregates in the environment; and
TMX Salmon Research Study- Literature Review and Power Analysis
April 2020
Page 26
- Further analysis of hydrocarbon composition present in fresh and weathered diluted bitumen blends
(DFO 2018b, p. 18)
2.6.4 Freshwater Oil Spill Case Studies
The spill of crude oil into the Red Deer River that occurred in Alberta on June 7, 2012 provides helpful
insight in determining the likely concentrations and behaviour of dissolved constituents derived from
spilled oil in rivers and lakes. That spill was caused by high river flows stressing a pipeline at a watercourse
crossing, and spatially comprehensive water quality data were collected for weeks afterward. The pipeline
belonged to Plains Midstream Canada, and it had become exposed at the river crossing. Pipeline valves
were closed within two hours of the initial indications of a spill, and 462.75 m3 of crude oil was released.
The spill entered the Red Deer River and was carried approximately 40 km downstream to the Glennifer
Reservoir (AER 2014). A boom at the river mouth effectively stopped a significant amount of the spilled
crude oil on the water surface from moving downstream and further into the reservoir. However,
dissolved constituents, including light hydrocarbons (BTEX) and PAHs were measured in the reservoir
water after less than two days. These dissolved constituents continued to be detected for the next 9 days
of monitoring in the area and were described as moving downstream in a “pulse” that at times exceeded
water quality guidelines for the protection of aquatic life (Teichreb 2013; AER 2012, 2014). Therefore,
there is evidence that containing the physical oil product in a spill to freshwater does not equate to
containment of potentially harmful dissolved hydrocarbon constituents, which are free to move further
downstream. It is important to note that no sediment quality data are readily available from this spill
monitoring effort.
Another spill of 3,190 m3 of diluted bitumen and other crude oil product in July 2010 from an Enbridge
pipeline near Marshall, Michigan, into a tributary of the Kalamazoo River (Talmadge Creek) occurred
under high flow conditions. The spilled oil traveled about 65 km downstream over two days, and it affected
both watercourses and riparian areas (Dollhopf et al. 2014; Fitzpatrick et al. 2015; FOSC 2016). Water
quality data were collected following the spill for several months. PAH and BTEX concentrations in water
measured in the affected area peaked at some sampling locations in the weeks after the spill, and some
significantly exceeded water quality guidelines, in particular for benzene, toluene, benzo(a)pyrene and
naphthalene (Michigan Department of Community Health 2013). By the fall of 2010 (October
December), the maximum measured concentrations of these constituents in the water was lower and at
times undetectable, although toluene concentrations still exceeded guidelines at some locations. These
constituents were then detected at lower concentrations in the winter of 2011 (January-April), but in the
vast majority of samples collected (>95%, when they were measured). Only toluene concentrations
exceeded guidelines during that time, but the pervasiveness of hydrocarbon detection across sampling
locations indicates a system-wide input, possibly from impacted riparian and shoreline areas during
snowmelt. Concentrations rose again slightly during the period May-August 2011, and fluoranthene in
particular was measured at its highest concentrations during this time (Michigan Department of
Community Health 2013). The variability in constituent concentrations over space and time indicates the
complexity of estimating the magnitude and duration of contaminant exposures for aquatic organisms in
the case of a spill, in particular where spilled oil is stranded in riparian areas and/or a portion of it sinks to
the sediments.
A significant proportion of the spilled oil in the Marshall Spill became submerged in the Kalamazoo River,
including at several impoundments along its length. As of July-August 2012, there was an estimated 681m3
TMX Salmon Research Study- Literature Review and Power Analysis
April 2020
Page 27
of submerged oil present in depositional areas along the entire 38-mile-long reach of the Kalamazoo River
impacted by the spill (Graan & Zelt 2013). Extensive dredging of the river and impoundments was
completed in 2012-13, and sediment traps were used to collect submerged oil that was migrating
downstream. In 2015, the estimate of residual submerged oil in the Kalamazoo River was between 185
and 325 m3 (FOSC 2016). While sediment traps remained in place, it was decided that some of this residual
oil would be left in place, as the potential ecological damage from recovery by dredging outweighed the
requirement to recover the oil (FOSC 2016). Sediment quality as measured by Enbridge in the affected
area of the Kalamazoo River and its impoundments varied, but measures of PAH concentrations taken in
2010 and 2011 indicate concentrations that exceeded CCME interim sediment quality guidelines, and in
some cases probable effects levels, for benzo(a)pyrene, benzo(a)anthracene, phenanthrene and pyrene
(enbridge_sediment_data.csv file, accessed in October 2019 from
https://archive.epa.gov/region5/enbridgespill/data/web/html/index.html). However, it was noted that the
Kalamazoo River and its impoundments have been previously impacted by releases of such constituents,
as well as other potential pollutants, so that assigning cause for these exceedances is difficult. This highlights
the importance of establishing baseline conditions through comprehensive monitoring to allow for
accurate impact assessment, and the importance of continuing to monitor over time.
In terms of the recovery of spilled oil products, including diluted bitumen, from river systems, there are
several helpful case studies that can inform predictions about oil behaviour and recovery success. Many of
these were outlined in the DFO status report (2018b). One case study is of value in considering spills
during winter where an ice cover is present. Recovery crews attempted to access spilled Bakken crude
oil, which penetrated the ice cover and entered the Yellowstone River in Montana in 2015, by drilling
holes through the ice and vacuuming or sopping up the oil. However, under-ice recovery of the spilled oil
was estimated at less than 10% (State and Federal Trustees, State of Montana and U.S. Department of
Interior, 2017). Concentrations of benzene and PAHs were also elevated in the river for several kilometres
downstream, and oil sheens were observed in the river over 90 km downstream. These impacts were
observed both in the days immediately after the January spill, but also in March during river ice breakup
(State and Federal Trustees, State of Montana and U.S. Department of Interior, 2017). The poor recovery
rate and extended period of impact highlights the challenges of oil recovery from rivers during winter
when an ice cover is present.
2.7 Current State of Knowledge on Toxicity of Diluted Bitumen to
Fish
2.7.1 Studying diluted bitumen toxicity in the laboratory
Most studies that investigate the biological effects of crude oil exposure in fish use water-accommodated
or water-soluble fractions (WAF). WAF solutions are generated by mixing defined volumes of crude oil
in water, allowing the mixture to settle so that undissolved oil droplets collect on the surface and the
soluble contaminants can be removed from the lower phase of the mixture. Therefore, only soluble
components of the crude oil (and potentially oil micro-droplets, depending on preparation method) are
present in the exposure water. Fish take up these contaminants primarily at the gill (i.e. fry to adult stages)
or the skin (i.e. eggs), and potentially also via the gut (e.g. consumption of contaminated prey). WAF
prepared from diluted bitumen tends to contain lower concentrations of MAH and PAH relative to
conventional crude oils (Barron et al., 2018; Philibert et al., 2016).
TMX Salmon Research Study- Literature Review and Power Analysis
April 2020
Page 28
Many factors will contribute to the toxicity of diluted bitumen, most notably its concentration and the
exposure duration, as well as environmental co-factors that drive weathering. Weathering refers to the
deterioration of crude oil in the environment. Evaporation of volatile organic compounds (e.g. BTEX)
occurs rapidly (minutes to days) and alters the physical properties of the oil and the relative proportions
of other constituents in the oil. Sunlight (ultraviolet radiation) alters the chemical structure of PAH and
makes them more toxic to fish (i.e. phototoxicity). The use of chemical surfactants and dispersants during
clean-up efforts may also influence diluted bitumen toxicity, as these work to disturb surface slicks by
generating smaller oil droplets that become more broadly distributed in the water column (i.e. dilution
effect) and may be more readily degraded. In general, dispersants can exert their own toxic effects and
also increase the toxicity of crude oil to fish (by increasing the uptake and bioavailability of certain
contaminants in the oil). Of note, dispersants were not applied as part of the spill response tactics in either
the Kalamazoo River or Red Deer River spills.
There are currently 12 peer-reviewed papers presenting data on diluted bitumen toxicity to fish. The
majority of studies have assessed toxicity in fish eggs, and the results are summarized below by lethal and
non-lethal endpoints. In general, younger life stages are considered more sensitive to environmental
contaminants, including crude oil (Kennedy, 2015). In addition, younger life stages are more at risk of
exposure since they cannot escape a contaminated site until they have hatched and are able to swim freely.
Salmonids develop slowly during the winter months and emerge in the spring, therefore, the window for
exposure risk is longer than for other species, like yellow perch, that hatch in less than one month. In
addition, during the weathering process, the density of spilled diluted bitumen increases, and it may sink
(e.g. Kalamazoo), which would complicate/limit any clean-up efforts in river systems containing salmon
eggs and prolong exposure times.
2.7.2 Lethality of diluted bitumen exposure
Death occurs rapidly upon exposure to crude oil, and is influenced by concentration, duration of exposure,
and life stage of fish. For example, 150 mature fish and 1000 juvenile fish died within 48 h of 30,000 liters
of gasoline and 24,000 liters of diesel fuel being released into the Pine River, BC in August 1994 (Hodson
et al., 2011). Six years later, a pipeline rupture in this same river released 475,000 liters of a light crude
oil that spread as a surface oil slick for 20 km downstream. A total of 1600 dead fish were collected
downstream from the spill site; however, total fish kill estimates ranged from 25,000 to 250,000 (Hodson
et al., 2011). Unfortunately, there are no fish kill estimates for the 2010 Kalamazoo River spill, but fish
community and habitat surveys were conducted annually as part of a post-spill monitoring program. During
the 5 years of monitoring, no differences were reported for overall species diversity between impacted
and reference sites. Certain species, like small-mouth bass, experienced a temporary drop in numbers
that was likely linked to poor habitat quality in impacted areas (USFWS 2017).
A standardized test to define a chemical’s acute toxicity is the LC50, which refers to the lethal
concentration at which 50% of the cohort (group) dies in a defined time period. Acute tests last 24 h to
96 h, and chronic LC50 tests are typically 7 days (d). For example, the 96h LC50 for BTEX ranges from
10 μg/L to 72 mg/L depending on life stage and species of fish. During these tests, the exposure solution
is replaced and/or replenished daily to maintain a relatively constant concentration of volatile components.
Table 1 provides details from studies that report fish mortalities resulting from diluted bitumen exposure.
For context, we note that single PAH species were measured at concentrations as high as 86 μg/L (e.g.
fluoranthene) in the Kalamazoo River during the summer months after the Enbridge pipeline failure. For
TMX Salmon Research Study- Literature Review and Power Analysis
April 2020
Page 29
the North Thompson study system, our estimated maximal concentrations of BTEX (279-12,892 ug/L)
and PAH (292-13,480 ug/L sum total of 16 EPA priority PAHs) exceed all exposure concentrations used
in published laboratory studies (see Section 3.0 for methods used to derive these estimates). However,
these high concentrations in real-world spill scenarios are restricted to discrete locations and times, and
therefore the maximal observed effects from published studies are likely outcomes in our hypothetical
spill scenarios for the North Thompson.
Table 1. Summary of published studies reporting lethal effects of diluted bitumen exposure in fish.
Dilbit refers to diluted bitumen.
Species & life
stage
Test
Results
Rainbow trout:
Fry (75-125 d
old; 0.3-0.8 g)1
96h LC50
LC50 = 5.66 g oil/L* unweathered
CLB, most mortalities occurred in
24 h; 100% mortality at 10 g oil/L
(= 35 ug/L PAH);
no mortalities for weathered CLB
at 18 g oil/L
Fathead
minnow:
Hatched eggs1
Acute (96h LC50),
Chronic (24h
exposure, mortality
up to 7d)
Acute LC50 = 0.63 g oil/L and 2.1 g
oil/L, unweathered and weathered
AWB;
Chronic = 0.6 g oil/L and 1.3 g oil/L
unweathered and weathered AWB
Fathead
minnow: Fry (7-
10 d old)2
96h LC50
Lowest LC50 = 20.5 ug/L PAH*
(CLB unweathered)
Highest LC50 = >40 ug/L PAH
(WCS unweathered)
i.e. CLB is more toxic than WCS
LC50 are only slightly lower for
weathered than unweathered oil
Zebrafish:
Eggs to fry (fert.
to 7 d larvae)3
7d LC50
LC50 = 88% dilbit WAF (there is
46 ug/L PAH in 100% WAF)
Ran similar tests on 2 conventional
crudes, and overall 100% dilbit
WAF was less toxic than 100%
WAFs prepared from other oils,
but dilbit WAF contained 3-4 fold
lower total polycyclic aromatic
hydrocarbons (TPAH)
TMX Salmon Research Study- Literature Review and Power Analysis
April 2020
Page 30
Sockeye salmon:
Egg (fert. to
hatch, ~72 d,
then reared 8
months in clean
water)4
Cumulative and
delayed mortality
Cumulative mortality (0-72d
exposure, including failed hatching)
was 10.5% higher than controls for
fish exposed to 100 ug/L TPAH;
mortality was 3.5% higher than
controls during the first 7 d.
Delayed mortality (0-150 d post-
exposure) was 43% higher than
controls for fish exposed to 100
ug/L TPAH (mortality was 36%
higher than controls during the first
60 d post exposure)
Sockeye salmon:
Smolt (~1.5y)5
challenged with a
marine bacterial
pathogen (vibriosis)
Fish exposed for 42 d to 124.5 ug/L
TPAH were more susceptible to
disease than unexposed control
fish: cumulative mortality was
29.5% higher than controls
* Exposure concentrations are inconsistently reported in the literature, and may reflect a loading value (grams of oil per
liter of water or substrate), or a sum total of a subset of 16 to 75 individual PAH measured in the exposure water.
1Robidoux et al., 2018; 2Barron et al., 2018; 3Philibert et al., 2016; 4Alderman et al., 2018; 5Lin et al., 2019
LC50 = lethal concentration for 50% mortality; TPAH = total polycyclic aromatic hydrocarbons; WAF = water-
accommodated fraction
Laboratory experiments present mixed results regarding the acute lethality of diluted bitumen exposures.
Zebrafish exposed from fertilization for 7 d did not reach 50% mortality using a 100% WAF exposure
(>46 μg/L PAH, sum total of 16 EPA priority PAHs; Philibert et al. 2015). Similarly, sockeye exposed
continuously to a diluted bitumen WAF with total PAH ~100 μg/L (sum of 75 PAH) from fertilization to
young fry reached cumulative mortality of only ~8% during the exposure period; however, delayed
mortality surpassed 50% in these sockeye during the 6 months post-exposure (Alderman et al. 2018). In
contrast, all rainbow trout fry exposed to a diluted bitumen WAF prepared at 10 g/L oil load (maximum
PAH = 15.4 ug/L) died within 96 h, with an estimated LC50 of 5.66 g/L oil load (Robidoux et al. 2018).
This latter study exposed a second subset of rainbow trout fry to a WAF prepared from weathered
diluted bitumen, and here LC50 was not reached in 96 h even at oil loadings of 18 g/L, indicating a
considerable reduction in toxicity as the diluted bitumen weathers.
Based on the available data, a 4,000 m3 diluted bitumen spill into the Thompson river system or Kamloops
Lake could result in anywhere from 10% to 100% mortality of salmon present in rivers and in lakes during
the first 24 96 h after the spill (i.e. prior to significant weathering of spilled product, and while spill
response procedures are focused more on containment than clean-up). It is reasonable to assume
mortality on the high end of this range, based on mortality estimates from other crude oil spills into river
systems (e.g. Pine River, BC (Hodson et al., 2011)). Delayed mortality may occur in fish surviving the first
days post-spill (e.g. Alderman et al. 2017), as a result of developmental malformations or other sub-lethal
effects.
TMX Salmon Research Study- Literature Review and Power Analysis
April 2020
Page 31
2.7.3 Sublethal effects of diluted bitumen exposure
Sublethal effects of crude oil exposure refer to those biological responses that do not result directly in
mortality, but are likely to impact individual health and performance, including physiology and behaviour,
which in turn could impact population recruitment (new individuals added to a population). Sublethal
effects are expected to occur at lower diluted bitumen concentrations and/or longer exposure times. A
standardized test to define a chemical’s chronic sublethal toxicity is the EC50 – the effective concentration
at which 50% of the population displays a sublethal endpoint of interest (e.g. malformation); however,
EC50 is not always calculated or reported. Table 2 provides details from studies that report various
sublethal effects in fish following diluted bitumen exposure.
Table 2. Summary of published studies reporting sublethal effects of diluted bitumen in fish. Dilbit
refers to diluted bitumen.
Species & Life
stage
Exposure
Endpoint
Results
Fathead minnow:
Egg (fertilization to
hatch)1
CLB and AWB
(unweathered) WAF,
loading rate 1:9 v/v
oil:water (=0.92 g
oil/L), with renewal
Malformation,
hatching success
Malformation: EC50 = 0.6 0.9 ug/L PAC (=
0.83-1.1 mg/L TPH-F).
Hatching: 74-84% success (at 10-32% v/v
WAF dilutions), compared to >90% in
unexposed controls.
Yellow perch:
Egg (fertilization to
hatch, 16d)2
CLB and AWB
as in1
Malformation,
hatching success
Malformations: 100% of embryos have
malformations at 32% WAF (vs 3.8% in
controls), but incidence is not significant at
more dilute concentrations; EC50 = 3.3-3.8
mg/L TPH-F.
Hatching: 2.5% (at 32% v/v AWB WAF),
60% (at 32% v/v CLB WAF), >80% with
more dilute exposures and in unexposed
controls.
Zebrafish:
Egg to fry (fert. to 7
d larvae)3
Dilbit WAF
(unspecified blend),
loading rate = 1:10 v/v
oil:water, with renewal
Malformations,
hatching success,
behavior
Malformations: larvae exposed to 100%
WAF (~47 ug/L PAH) had ~20% higher
occurrence of pericardial and/or yolk sac
edema relative to unexposed controls.
Hatching: 20% fewer hatched at 2 d post-
fertilization (dpf) (onset of hatching in
zebrafish) but caught up by 3 dpf
Behaviour: avoidance behaviour in larvae
exposed to 100% WAF
Japanese medaka:
Egg (fert. to hatch,
17d)4
CLB, as in1
Malformations
Malformation: EC50 = 2.83 ug/L TPAH
Japanese medaka:
Egg (fert. to hatch,
17d)5
AWB, as in1
Blue sac disease
EC50 = 20.5 ug/L TPAH
Sockeye salmon:
Egg (fert. to swim-
up, 72 d)6
CLB WSF,
unweathered, renewed
every 2 wk
Malformation,
Hatching success,
growth
Deformity: 17-20% of embryos with at least
1 deformity at 35 and 100 ug/L TPAH
(occurrence in control fish is 2%).
TMX Salmon Research Study- Literature Review and Power Analysis
April 2020
Page 32
Hatching: onset of hatching delayed 1-3 d
relative to control; time to 50% hatch
delayed 3-4 d relative to control fish; effects
on hatching observed at concentrations as
low as 4 ug/L TPAH. Hatching success was
98% in unexposed control fish, and 92% in
sockeye exposed to 100 ug/L TPAH.
Body composition: more lipids and less
protein in fry exposed to 35 and 100 ug/L
TPAH (reduced yolk consumption for
growth)
Sockeye salmon:
Smolts (1 yr
juveniles, pre-smolt
exposed 1 and 4
wk)7,8
CLB WSF 4 wk;
unweathered,
unreplenished;
initial TPAH given, but
decreases exponentially
with time
Swimming
performance,
histopathology,
EROD; serum
proteome
Reduced swim performance (10% lower
relative to control) and cardiac fibrosis after
4 wk exposure to 67 ug/L TPAH; liver
EROD increased at 1 wk with 3.5 ug/L
TPAH (=lowest concentration)
Serum proteome: increases in non-specific
immune proteins; other changes suggest fish
are more susceptible to exercise-induced
tissue damage (after 4 wk exposure to 67
ug/L TPAH)
Atlantic salmon:
Smolts (1 yr
juveniles, smolts,
exposed 24 d)9,10
CLB WSF,
unweathered,
unreplenished over 24
d
Swimming, muscle
physiology,
histopathology,
EROD
No effect on swimming, gill or kidney
histology (max conc. 67 ug/L TPAH);
Liver EROD takes 2 wk to return to
baseline after 67ug/L TPAH exposure;
Seawater acclimation response not affected
by dilbit
Sockeye salmon:
Parr (~1.5y)11
CLB WAF,
unweathered, 24h or
21d (no renewal,
TPAH reported are
initial); 13.7 124.5
ug/L TPAH
blood parameters,
stress indices,
EROD
Time and concentration-specific increase in
liver EROD activity
Concentration-dependent changes in stress
indicators within 24 h of exposure and
maintained through 21 d at higher WAF
concentrations (increased plasma cortisol,
glucose, and lactate; decrease liver glycogen)
Ionoregulatory disturbances at higher WAF
concentrations and/or longer exposures
(plasma osmolality, Na+, Cl-)
1Alsaadi et al., 2018; 2McDonnell et al., 2019; 3Philibert et al., 2016; 4Madison et al., 2017; 5Madison et al., 2015; 6Alderman et al., 2018;
7Alderman et al., 2017b; 8Alderman et al., 2017a; 9Alderman et al., 2020; 10Avey et al. 2020; 11Lin et al., 2019
EC50 = effective concentration for 50% occurrence in population; EROD = Ethoxyresorufin-O-deethylase; PAC = polycyclic aromatic
compounds; TPAH = total polycyclic aromatic hydrocarbons; TPH-F = total petroleum hydrocarbons quantified using fluorescence
For the study system considered in this report, sublethal exposures may occur kilometers from the spill
site as contaminants are carried downstream, even when a containment barrier is in place (e.g. dissolved
contaminants moved underneath the boom, detailed in Plains Midstream Canada Spill Interim Water
Quality Report June 2012). In addition, flooding of oiled shorelines or snowmelt can redistribute
contaminants into the water column many months after the spill (e.g. Kalamazoo River case study). For
eggs and other young life stages, sublethal concentrations are likely to induce hatching delay (Philibert et
al. 2016; Alderman et al. 2018; Alsaadi et al. 2018; McDonnell et al. 2019) and changes in growth and
energy metabolism (Alderman et al. 2018). For example, pink salmon fry (starting fork length ~30 mm)
TMX Salmon Research Study- Literature Review and Power Analysis
April 2020
Page 33
fed Alaska North Slope Crude Oil (ANSCO, from the Exxon Valdez disaster in the Prince William Sound,
Alaska, 1988) contaminated food for six weeks experienced 45% and 95% growth reductions at 2.8 mg oil
per g food (mg/g) and 34.8 mg/g, respectively (Wang et al., 1993). In addition to growth effects,
developmental abnormalities such as yolk sac and/or pericardial edema (fluid around the heart) can occur
in as much as 40% of the population. Malformations may lead to post-exposure mortalities (Alderman et
al. 2018), and pericardial edema during egg development may be a contributing factor to reduced
swimming performance in later life stages (e.g. Incardona et al. 2015, but see also Alderman et al. 2018).
Sockeye smolts exposed to sublethal concentrations of diluted bitumen WAF for 4 wk experienced a 10%
reduction in aerobic swimming performance (Alderman et al. 2018), which could limit success of
outmigration; however, Atlantic salmon smolts exposed similarly did not experience any decline in
swimming performance (Avey et al 2020).
2.7.4 Diluted bitumen toxicity in relation to Pacific salmon
2.7.4.1 Species and life stage sensitivities to diluted bitumen
There are not enough studies published to assess life stage- and species-specific sensitivities to diluted
bitumen; however, earlier work on ANSCO suggests similar sensitivities among six species of salmonids,
with emergent fry showing the greatest sensitivity across life stages (Moles et al., 1979). Older fish
experience adverse effects when exposed to conventional crude oil, including reduced aerobic swimming
capacity, altered behaviours, and changes in growth and energy metabolism (e.g. reduced body condition;
altered lipid metabolism). Whether or not this suite of adverse effects is consistent with diluted bitumen
exposure in older fish is not known. The majority of studies to date on diluted bitumen toxicity exposed
fish as eggs, with limited effort on post-hatch life stages and no studies on adult fish. Aerobic swimming
capacity was reduced in sockeye salmon smolts (Alderman et al., 2017b) but not Atlantic salmon smolts
(Avey et al. 2020) exposed to diluted bitumen at similar concentrations and durations. Atlantic salmon
smolts exposed to diluted bitumen were able to repeat swimming performance after 24 h rest but showed
a shift from aerobic to anaerobic metabolism in the heart and red muscle after exercise, which could set
limits on endurance exercise capacity (Avey et al. 2020). More research needs to be conducted on post-
hatch fish from species at risk of exposure to fully understand the potential impacts of a diluted bitumen
spill in the aquatic environment.
2.7.4.2 Anadromous life history
The potential for sublethal effects of diluted bitumen exposure to impact the unique ecology of Pacific
salmon, specifically their migrations and transitions between freshwater and seawater environments
(anadromous), represents a major knowledge gap. For example, reduced swimming performance and/or
shifts in metabolic strategy within swimming muscles could decrease a smolt’s ability to complete sustained
exercise necessary for outmigration to seawater or challenge its capacity to escape larger and faster
predatory marine fish (Alderman et al., 2017; Avey et al. 2020). Smoltification (adaptation from living in
freshwater to living in seawater) could be impacted by diluted bitumen exposure as a result of stress,
disruption in hormone pathways, and/or an inability to regulate the body’s salt and water composition. All
of these processes have been shown to be disrupted in fish exposed to conventional crude oil (Kennedy
and Farrell, 2006, 2005), and that are critical for successful seawater transition of smolts. In fact, out-
migrant smolts from six salmonid species tested in seawater were twice as sensitive as those tested in
freshwater to ANSCO (Moles et al., 1979). Results from a recent study in Atlantic salmon smolts suggests
that osmoregulatory capacity is not greatly affected by diluted bitumen exposure (Alderman et al. 2020);
however, similar experiments are needed in other salmonids to verify this result as the Atlantic salmon in
TMX Salmon Research Study- Literature Review and Power Analysis
April 2020
Page 34
this study were generally less sensitive to diluted bitumen than sockeye salmon from a similar study
(Alderman et al., 2017; Avey et al. 2020).
Whether diluted bitumen exposure affects olfactory imprinting (early-life bond to a location) during early
life stages is unknown, but such an effect would carry significant long-term costs to salmon populations by
reducing navigation success during the spawning migration. Olfactory (i.e. smell) imprinting occurs during
transitional periods, such as emergence and smoltification, and ultimately enhances the fish’s sensitivity to
olfactory cues from its early rearing environments. The process of imprinting is regulated by thyroid
hormones (Specker 1988; Lema and Nevitt 2004), an endocrine system that is also important for
development, metabolism, and growth. It is not known if crude oil exposure alters thyroid hormone
signaling, but other endocrine systems are affected by crude oil (Lin et al. 2019). Crude oil exposure also
induces morphological (Solangi and Overstreet, 1982) and functional (Cave and Kajiura, 2018) changes to
the olfactory epithelium, which could further impair the ability of fish to imprint to olfactory cues and/or
use these cues during spawning migrations.
2.7.4.3 Behavioural responses to crude oil exposure
Many behavioural responses have been observed in fish exposed to crude oil, including lethargy (lack of
energy), increased respiration (breathing) rates, altered activity patterns, and impaired feeding and
predatory behaviour. Behavioural effects after diluted bitumen exposure in salmon are unknown, but
young zebrafish exposed to diluted bitumen for 7 d showed an increased tendency to swim at aquarium
edges, suggesting an anxiety-like behavioural response (Philibert et al. 2016). Sub-adult and adult Coho
salmon avoided water containing conventional crude oil (230-530 ug/L) and MAH mixtures (2-5 mg/L).
While this behavioural response can be beneficial by helping to limit exposure times, this avoidance
response could prevent mature adults from reaching their spawning grounds.
2.7.4.4 Contaminant uptake and clearance
Water-soluble contaminants in crude oil can enter the fish via thin, high surface area epithelial surfaces
such as the gills and gut. The liver processes these contaminants by altering their chemical structure
through a series of enzymatic steps (detoxification pathway) that facilitate their excretion via urine. The
detoxification response can be quantified by measuring the activity and expression of a key enzyme
involved in this process, cytochrome P450 type 1a (cyp1a), which is activated by PAH and crude oil
exposure. Thus, quantification of cyp1a gene expression or activity can serve as a biomarker of crude oil
exposure. In Sockeye smolts exposed to diluted bitumen, the cyp1a response is evident after 1 wk
exposure to concentrations as low as 35 ug/L initial TPAH, and after 4 wk exposure to 3.5 ug/L initial
TPAH, which highlights the sensitivity of this biomarker to diluted bitumen exposure (Alderman et al.
2017a). After 4 wk exposure to 67 ug/L initial TPAH, a cyp1a response was also detected in certain tissues
(e.g. kidney, muscle), suggesting that (i) PAH were present at high enough concentrations in tissues other
than the liver to elicit a cellular response, and (ii) that continued exposure to diluted bitumen causes a
bioload of PAH in the body that exceeds the liver’s capacity for detoxification. Cyp1a activation was also
evident in muscle tissue from Atlantic salmon smolts exposed to diluted bitumen concentrations as high
as 67 ug/L initial TPAH for 3 wk, and cyp1a levels did not return to baseline levels until 2 wk recovery in
clean water (Avey et al., 2020). Combined, these studies suggest that the body burden of PAH will vary
depending on the nature of the exposure (concentration and duration) as well as tissue type.
How quickly contaminants are cleared from the body will depend on their rate of uptake from the
environment and the maximal rate at which the liver can process these contaminants, both of which are
TMX Salmon Research Study- Literature Review and Power Analysis
April 2020
Page 35
likely influenced by species, life stage, and environmental factors (e.g. temperature). If uptake surpasses
the body’s capacity for clearance, PAH and other contaminants in crude oil can accumulate in fish tissues.
For example, sea bream (Sparus aurata) exposed to benzo(a)-pyrene, a model carcinogenic PAH, can
accumulate sufficient amounts in muscle tissue to pose a health risk for human consumption (Zena et al.
2015). This has relevance not just to human health, but broader ecosystem health as well given the
importance of salmon in both aquatic and terrestrial food webs.
2.8 Current State of Knowledge on Impacts to Terrestrial Wildlife
that Rely on Salmon
A diluted bitumen spill into the Thompson river system and Kamloops Lake could result in high mortality
salmon. Although current knowledge is limited for life stage and species-specific sensitivities to diluted
bitumen, some research indicates that adult fish experience adverse behavioral, growth and metabolic
effects when exposed to conventional crude oil (Section 2.7.3). Each of these effects can impact salmon
population recruitment and have consequences for species that rely on salmon. There is also concern
associated with the exposure of wildlife to harmful compounds via the ingestion of contaminated prey,
which could lead to bioaccumulation or biomagnification of these compounds in other wildlife species.
Bioaccumulation refers to the accumulation of a toxic chemical in the tissue of an animal that is caused by
the presence of that chemical in its food. Biomagnification is a related process that occurs when the
concentrations of chemicals increases in the tissues of animals along with their position in the food web.
Depending on how quickly these contaminants can be cleared from the body, PAH and other contaminants
in crude oil can accumulate in fish tissues (Section 2.7.4) which could have ramifications for other
terrestrial wildlife.
Although not a focus for this research report, we have briefly discussed the potential consequences of
reduced food availability and quality to the broader terrestrial ecosystem.
2.8.1 Changes in Food Availability
Consequences of a decline in prey availability can lead to corresponding population declines in higher-
order predators. Researchers have noted that declines in salmon abundance from other stressors can lead
to reduced densities in eagle, grizzly bear and other wildlife populations (e.g. kokanee salmon collapse,
Spencer et al. 1991; eagles, salmon and climate change; Rubenstein et al. 2018). Therefore, depending on
the temporal and spatial extent of an oil spill, it could potentially have far reaching effects to the broader
ecosystem.
When food resources are scarce, the overall fitness in individuals can decline. For example, a decline in a
preferred fish species of cormorant, (European shag, Phalacrocorax aristotelis), following an oil spill off the
coast of Spain was thought to contribute to lower cormorant reproductive success (Valendo et al 2005
in Henkel et al 2012). If birth rates decline, this can have population level effects, which is of particular
concern for those species that may be of special management concern or for those species requiring these
resources to prepare for key seasonal cycles (e.g. migration, hibernation).
TMX Salmon Research Study- Literature Review and Power Analysis
April 2020
Page 36
During the late summer and fall, grizzly bears will focus their energy on obtaining food to increase their
body mass prior to hibernation. The availability of high-quality foods, such as salmon, is essential for
accumulating those fat reserves for hibernation and in supporting successful reproductive outcomes
(during hibernation, embryos will implant on the uterine wall). Higher salmon consumption by coastal
bears has been shown improve their body condition, reproductive success and population densities
compared to areas without access to salmon (Hilderbrand et al. 1999, Service et al. 2019). Although
grizzly bears on coastal BC consume more salmon than their counterparts in the interior of BC (Mowat
and Heard 2006), hotspots of salmon consumption inland have been identified (Adams et al. 2017). Further
research is needed to understand the role of salmon consumption for interior grizzly bear populations
and whether differences in salmon abundance would have any effects on their fitness.
Bald eagles are also a species that regularly feed on salmon, generally carcasses left behind from other
predators, and are opportunistic feeders whose diet includes fish, avian species and carrion. Generally,
eagles migrate from Alaska southward following the progression of salmon spawning activity. Although
several studies have shown bald eagle abundance linked with salmon density (e.g. Hunt et al 1992, Hansen
et al 1984 in Elliott et al 2011), other studies have highlighted the complexity of this relationship and
perhaps regional differences. For example, Elliott et al (2011) found that in the south-coastal area of BC,
the number of chum salmon carcasses peaked well before the peak in eagle numbers. It is clear that these
relationships are quite complex and that there is still much more information needed to understand the
interactions in these food web systems, as well as, how the system would respond and recover from
various natural and anthropogenic disturbances.
2.8.2 Changes in Food Quality
As described in Section 2.6.2, there are several chemical compounds that are found within different types
of oil such as monocyclic aromatic hydrocarbons (e.g. benzenes, toluenes, and xylenes), PAHs, and various
heavy metals (e.g. mercury). Species within the same taxa may show different sensitivities to pollutants
because of life history or physiology differences (Bergeon Burns et al. 2014) making it difficult to generalize
impacts even across a single species. However, research suggests that vertebrates may have the capacity
to metabolize and eliminate PAH residues which would suggest that there is limited potential for
bioaccumulation or biomagnification effects further up in the food chain (Neff 1979, Henkel et al 2012)
related to that compound. Although there has been research on heavy metal toxicity following oil spills,
the exposure route of these organisms to heavy metals is not well understood. For example, chromium
and nickel can bioaccumulate in organisms, and were measured in the tissue of whales within the
Deepwater Horizon spill site at higher concentrations compared to tissue samples of whales outside of
this area (Wise Jr, JP et al 2014). However, it is not clear whether the uptake of these metals was via
exposure through inhalation, dermal absorption, via ingestion of contaminated food, or a combination of
all of these processes. Furthermore, while it is likely that the Horizon oil spill contributed to the higher
concentrations of metals in these whales, there are additional sources in the Gulf of Mexico, such as the
release of industrial waste and exposure to boat paint that contain metals as antifouling agents (Wise Jr JP
et al 2014). Thus, pre-existing stressors can confound the ability to ascribe these effects specifically to oil
spills which can lead to dismissal of concerns surrounding impacts to the health of wildlife, unless there is
adequate baseline data prior to the spill that would capture existing levels of contaminants in the system.
TMX Salmon Research Study- Literature Review and Power Analysis
April 2020
Page 37
2.8.3 Reduced Habitat Quality
The role of salmon in the ecosystem is not solely via a food source for predators but also as a source of
nutrients to the surrounding environment that supports plant growth and health, and in turn, wildlife. For
many of the terrestrial species, the consumption of salmon is only one part of their diet, and the continued
health of plants and invertebrates in the surrounding landscape is essential. If fewer salmon return, there
will be fewer salmon carcasses left by terrestrial scavengers that will contribute to the nutrient cycling
process. Bears also consume fruits and serve as seed dispersal agents via their scat. If salmon were limited,
bear densities may be lower, which could alter seed dispersal patterns and vegetation growth (Gende et
al 2002). Changes in plant diversity and density could then affect densities of omnivorous species, leading
to broadscale reductions in biodiversity in the area.
2.8.4 Monitoring Impacts to Wildlife Following Oil Spills
In 2010, a pipeline operated by Enbridge ruptured in a wetland which flowed into surrounding wetlands,
the Talmadge Creek and the Kalamazoo River (see Section 2.6.4 of this report). The follow up monitoring
after the release of diluted bitumen into the environment, included surveys in the floodplain habitats of
the Talmadge Creek and Kalamazoo River in order to characterize the extent and degree of oiling to
inform clean up techniques. A fish health assessment and fish community assessment were conducted to
evaluate impacts from the spill on fish health and abundance (US Fish and Wildlife Service et al 2015).
Enbridge also established a wildlife response center that cared for and released almost 4,000 animals,
including 3,650 reptiles and 196 birds (NTSB 2012). Long term effects to wildlife are unclear.
Plains Midstream’s 2012 pipeline incident spilled over 460 m3 litres of oil into the Red Deer River and
Glennifer Lake near the town of Sundre, Alberta (see Section 2.6.4 of this report). Although the
deployment of booms helped to stop the migration of visible oil on the surface, dissolved components
continued to move downstream into Glennifer Lake. Impacts to wildlife and vegetation included upstream
soiling/smothering of aquatic organisms and wildlife which the company responded to by employing
deterrent systems in areas with the greatest possibility for wildlife coming into contact with contaminated
soil, plants and water, as well as rescuing any reported oiled wildlife. However, it remains unclear how
regulators and the company addressed potential long-term impacts to wildlife associated with the
transport of dissolved components downstream, other than through water quality monitoring. Few details
are publicly available as to the type and length of wildlife monitoring that was conducted but one news
release indicated that Alberta Environment determined that impacts to wildlife from the spill had been
minimal
(http://www.calgaryherald.com/technology/meet+oily+animals+rescued+from+deer+river+spill/6786070/
story.html).
Follow up efforts for both the Enbridge and Plains Midstream spills appeared to focus on the short-term
effects from a spill, including employing deterrents to minimize continued or future exposure and the
rescue and rehabilitation of any oiled wildlife. Open access to information and transparency is important
to our understanding of how regulators and proponents are monitoring long-term effects of oil spills on
wildlife health, changes to resource availability and quantity, changes to habitat use and the overall
cumulative effects from a spill, combined with other stressors to wildlife.
TMX Salmon Research Study- Literature Review and Power Analysis
April 2020
Page 38
Quantifying impacts on terrestrial vertebrates can be challenging because for some species, there is a lack
of baseline data on population sizes, habitat use, or dispersal. Any impact assessment would need to
establish recovery endpoints that would represent a return of a population or habitat to some pre-spill
condition. Without baseline data, these endpoints will be difficult to characterize and establish. Trying to
track impacts from oil exposure to individuals, to populations and community structures are further
complicated by the cumulative effect of other stressors on the ecosystem. For example, changes to the
abundance and breeding success of seaside sparrows in the Louisiana marshes following the Deepwater
Horizon oil spill were compounded by effects from Hurricane Isaac (Bergeon Burns et al 2014). Long term
and integrated approaches to monitoring are needed to fully understand impacts from oil spills, regardless
of their size and extent.
2.9 Summary of Knowledge Gaps
As a first step in determining the potential impacts of a diluted bitumen spill on Pacific salmon populations
in the Area of Interest, this report compiled data relevant to TMX pipeline spill scenarios, as well as data
on how the North Thompson River and Shuswap complex are used by freshwater life stages of Sockeye
and Chinook salmon. This dataset, along with the current state of knowledge for diluted bitumen toxicity
to fish, underlies the modeling and power analysis presented in Section 3.0. However, the comprehensive
literature review revealed critical knowledge gaps relevant to the objectives of this report. As summarized
in Table 3, these knowledge gaps apply to almost every aspect of predicting how a spill of diluted bitumen
from the TMX pipeline will affect salmon. For example, defining the location and spatial extent of a worst-
case spill scenario is hindered by a limited understanding of how diluted bitumen will behave in a
freshwater system with considerable seasonal variability. Furthermore, the spatial resolution describing
how juvenile salmon utilize the area of interest during their freshwater residency is poor, and research
into the nature and outcomes of diluted bitumen exposure in fish is only in early stages. Filling these
knowledge gaps may improve the capacity to predict, calculate, or mitigate the potential effects of the
TMX project on salmon.
Table 3. Summary of Knowledge Gaps identified during Literature Review.
Knowledge
Category
Sub-category
Knowledge Gap
When and where
will a spill occur?
Pipeline and location
conditions
How will pipeline- and location-specific conditions influence
when and where a spill might occur?
Environmental
conditions
How will future changes in streamflow resulting from climate
and land use change affect the likelihood of a spill from the
pipeline?
What will the
spatial extent of a
spill be?
Diluted bitumen
properties
What will the spatial extent be of physical oil product and
dissolved constituents?
Diluted bitumen
properties and
Environmental
conditions
What are the dominant processes determining diluted bitumen
fate and behaviour in an aquatic ecosystem (degradation,
transformation)?
TMX Salmon Research Study- Literature Review and Power Analysis
April 2020
Page 39
Environmental
conditions
What is the behaviour of diluted bitumen in low temperatures,
or under an ice cover?
Environmental
conditions
How exactly will turbulence, suspended sediments and other
conditions in rivers and lakes influence when and where diluted
bitumen will become submerged or sink?
Where will fish be
at the time of a spill
(life-stage specific)?
Fish ecology
What size-cohorts of juvenile salmon typically use the
vulnerable area of the watershed?
Fish ecology
What are the natal origins of juvenile salmon (e.g., from
tributaries or other populations) inhabiting the vulnerable area
of the watershed?
What will be the
nature of diluted
bitumen exposure
(concentration and
duration)?
Diluted bitumen
properties
How will weathering of diluted bitumen affect its availability and
toxicity to fish?
Diluted bitumen
properties
What are the specific hydrocarbon compositions across various
types of diluted bitumen?
Environmental
conditions
How will conditions in water bodies (e.g. turbulence, sunlight,
temperature) at the time of a spill specifically translate into
concentrations of contaminants of potential concern (e.g.,
hydrocarbons)?
Response effort
How will the spill mitigation response influence the volume
spilled, spatial extent of impact, and environmental persistence
of contaminants?
What are the direct
effects of diluted
bitumen exposure
on salmon?
Diluted bitumen
toxicity
What is the comprehensive and detailed scope of diluted
bitumen toxic effects?
Diluted bitumen
toxicity
What are the relative salmon life stage and species-specific
sensitivities?
Diluted bitumen
toxicity
How long do adverse effects persist in salmon?
Diluted bitumen
toxicity
Do sublethal toxic effects ultimately contribute to population
declines?
What are the
indirect effects of a
pipeline spill and
diluted bitumen
exposure on fish?
Diluted bitumen
toxicity
Will toxic effects in salmon prey items decrease food
availability?
Environmental
conditions
How will a spill and the mitigation response impact salmon
habitat quality?
TMX Salmon Research Study- Literature Review and Power Analysis
April 2020
Page 40
How are adverse
effects in salmon
transferred to
other wildlife?
Food Web Impacts
Food web interactions and relationships are complex - how will
each trophic (food web) relationship respond to perturbations?
Contaminant Exposure
in Higher Order
Predators
What are the exposure routes via the food web and what are
species-specific sensitivities to oil?
3.0 Part B: Quantifying and Detecting Impacts to Salmon
The TMX project poses several plausible risks to salmon in the Area of Interest (North Thompson River,
Thompson River and Kamloops Lake) that may vary in spatial extent and timing across the salmon life
cycle. As outlined in our introduction, this report addresses three key research questions (listed below).
The first research question (greyed out in the list) was addressed under Section 2.0. In Section 3.0, we
now assess the remaining two research questions related to quantifying these potential risks to the number
of returning adult salmon, which is an important metric of fishery resources for the Adams Lake Indian
Band.
1) What is the state of knowledge about impacts of oil and/or diluted bitumen spills on Chinook and
Sockeye salmon in freshwater systems, and specifically in the Thompson River complex?
2) Given specific hypothetical spill conditions, what are the potential impacts of a spill of a diluted
bitumen product in the North Thompson River on Chinook and Sockeye salmon?
3) Using the available DFO data on numbers of returning adults, how precisely would we be able to
quantify such an impact on the number of returning Chinook and Sockeye salmon adults?
We address the second research question by calculating the impacts under a worst-case scenario spill
based on our current knowledge. We address the third research question by estimating the uncertainty
in the recruitment variation for the North Thompson versus the rest of the Fraser basin for individual
years. We also calculate the power (probability of simply detecting an effect) for a worst-case scenario
spill. It is important to note that the considered spill is worst-case in terms of timing, location and spatial
extent of the spill but due to the knowledge gaps identified in the previous section, this scenario may
impact the salmon more or less severely than calculated below. We note that the current impact
assessment framework assumes that if an impact cannot be detected then there is no impact even though
the data may also be consistent with a substantial negative impact. This is a flawed framework for assessing
impacts but has unfortunately been used for several large crude oil spills (e.g., Ward et al. 2017, 2018;
Shelton et al. 2018).
3.1 Defining Hypothetical Spill Scenarios and Impacts to Salmon
Understanding the potential impacts to North Thompson salmon required us to identify how oil spills
might translate into impacts on adult salmon. We began by identifying:
1) the timing, location, and magnitude of potential spills,
2) the fate of diluted bitumen and water in the watershed,
3) the timing, location, and abundance of salmon,
4) the vulnerability and impacts to those salmon from exposure, and
TMX Salmon Research Study- Literature Review and Power Analysis
April 2020
Page 41
5) how those impacts span the salmon life cycle.
To do this, we first reviewed case studies of oil spills and the downstream consequences to fishes
(described in Section 2.6.4). We then constructed a life cycle model of Chinook and Sockeye Salmon
within the North Thompson accounting for spatial variability in productivity (i.e. quantity of fish produced),
habitat usage, and survival within the mainstem and tributaries of the watershed. Next, we reviewed likely
direct impacts to survival (i.e. ability to persist/live) and fecundity (i.e. ability to produce offspring) of
salmon directly exposed to a worst-case diluted bitumen spill scenario. Last, we calculated the loss of
returning adults given their likely reduced survival and fecundity.
3.1.1 Hypothetical Spill Considerations
3.1.1.1 Example of the TMX pipeline project spill risk assessment
The TMX Project involves a twinning of the existing TMP with the addition of Line 2, including new
construction of pipeline segments, reactivation of some deactivated segments and construction of new
pump stations. The current TMP transports conventional crude oil, diluted bitumen and other petroleum
products from Alberta to the west coast of Canada at Burnaby, BC (TMP ULC 2013, Volume 2).
As part of the TMX Project EIA, a qualitative ecological risk assessment (QERA) was completed for
pipeline spills (TMP ULC 2013, Volume 7). The QERA focused on the environmental effects of
hydrocarbons that would be released with diluted bitumen, including BTEX, as well as PAHs and/or PACs,
while metals were not thoroughly considered as contaminants of potential concern (COPC). The
receptors chosen as part of the QERA were generalized across regions, and in BC, these receptors
included fish, fish eggs and larvae of Chinook salmon, Coho salmon, bull trout, Dolly varden, rainbow
trout/steelhead, and cutthroat trout, among other fish species, with a focus on a generic salmonid species
to represent the group.
The specific amount of released oil that was considered as part of the QERA was determined based on
potential outflow volumes expected at specific locations along the pipeline route. Those volumes were
projected based on the assumption of a response to a full-bore rupture comprised of mainline block valves
and check valves near watercourses being fully closed 15 and 10 minutes after the rupture occurs,
respectively. The “shutdown volume” is therefore expected to be constant at all locations considered, at
500 m3, while the “draindown volume” is dependent on the pipeline elevation and relative location of
valves at the location considered. The TMX Project’s outflow modeling was completed for the section of
pipeline that runs through the North Thompson River watershed from the Albreda River valley to
Blackpool, just downstream, of the Clearwater River confluence, and from Black Pines, downstream of
McLure, to Kamloops (Figure 7). It is not clear why this modeling was not completed for the entire length
of pipeline in the watershed. The cumulative outflow volumes (shutdown plus drain down volume)
predicted along this stretch ranged from a minimum of 500 m3 to about 4,000 m3 near the Finn Creek
Pump Station (TMP ULC 2013, Volume 7, Section 3.1.6 and Appendix B).
The spill scenarios that TMP ULC considered in its QERA included consideration of multiple weathering,
and transport mechanisms in aquatic environments (e.g. evaporation, dispersion, dissolution, microbial
degradation). Each of these mechanisms and processes were initially discussed in the QERA in relation to
the conditions influencing fate and transport of oil spilled into inland waters in case studies from t he
northern United States and Canada (TMP ULC 2013, Volume 7, QERA, Section 6.2.2). In aquatic
TMX Salmon Research Study- Literature Review and Power Analysis
April 2020
Page 42
ecosystems, the presence of oil slicks and the dissolution of BTEX substances in water often leads to
acutely toxic conditions for fish and other aquatic organisms.
These processes and patterns were considered in the discussion of specific hypothetical spill scenarios as
part of the QERA. The spill scenario considered along the North Thompson River involved the release of
1,400 m3 of Cold Lake Winter Blend (CLWB) from a full-bore rupture at pipeline route kilometer 766.0
near Darfield. In each case, it was assumed that oil in the ruptured section of pipe, as defined by control
valve locations, would completely drain through the pipe rupture, and that most of this oil would reach a
watercourse. Three different timing scenarios were considered: summer condition (June August), river
in freshet, flow >1,250 m3/s; and spring or fall condition (April June or September-November), flow is
moderate at about 500 m3/s; winter condition (December March) with ice cover on the river and snow
on land, flow at 100 m3/s or less.
For the high-flow summer scenario, oil was expected to travel up to 60 km downstream of the Darfield
spill location, with trace amounts reaching upper Kamloops Lake. Due to turbulence, oil was expected to
be entrained in the water column. The QERA assumes that suspended sediment loads would be low, but
this is unlikely during the freshet. Some oil was expected to adhere to shorelines and some was expected
to sink. Fish mortality was expected to occur within 10 km of the spill site, but as concentrations decline
further downstream, lethal exposures were considered less likely. The QERA indicates that recovery of
fish habitat was expected within 12 months (TMP ULC 2013, Volume 7, Table 6. 22).
For the spring/fall scenario, oil was expected to travel about 25 km downstream. Low turbulence was
expected, so entrainment of oil was expected to be limited. Oil was expected to adhere to shorelines,
especially cobble and gravel. Oil was not expected to reach the confluence of the South and North
Thompson Rivers (TMP ULC 2013, Volume 7, Table 6. 23).
The probable longitudinal extent of oil travel was not provided for the winter scenario the oil was
expected to be absorbed by snow on land, and then to spread out on the ice. Some oil was expected to
enter the River through open water patches and to move downstream. Effects on fish and fish eggs were
expected to be low because very little oil was expected to contact the water. These effects were also
expected to be limited spatially because most of the spilled oil was assumed to be recovered, and recovery
of fish habitat was assumed to occur within 6 months (TMP ULC 2013, Volume 7, Table 6. 21).
TMX Salmon Research Study- Literature Review and Power Analysis
April 2020
Page 43
Figure 7. Location of hypothetical spills from the proposed TMX pipeline initially considered for the
assessment of impacts to Chinook and Sockeye salmon. Hypothetical spill location 1 represented the
worst-case scenario of a release of 4,000 m3 of diluted bitumen. Hypothetical spill location 2
represents the spill location assessed in the TMX qualitative ecological risk assessment (QERA) (TMP
ULC 2013, Volume 7), which involved a release of 1,400 m3 of diluted bitumen.
TMX Salmon Research Study- Literature Review and Power Analysis
April 2020
Page 44
3.1.1.2 Spill Conditions Used for this Analysis
The North Thompson River does not have particularly high sediment load (measured as TSS); however,
it is a glacial-fed River which means that, especially during freshet, it is likely to have peak sediment loads
made up of fine sediments. More recent and comprehensive water quality data are needed to determine
this for certain. The presence of such sediment would increase the potential for diluted bitumen to sink.
Flows in the North Thompson are highly variable seasonally, and compared to the Kalamazoo River (see
discussion of this case study in Section 2.6.4), are many times higher during the freshet. It should be noted
that pipeline ruptures near watercourses have occurred during high flow or freshet conditions due to
erosive forces, especially where depth of cover at watercourse crossings is shallow (Teichreb 2013). The
stream gradient in the North Thompson is also higher than in the Kalamazoo River. Both of these
conditions are likely to contribute positively to turbulence, which is associated with an enhanced potential
for diluted bitumen to sink as it weathers.
The increased turbulence may lead to relatively low initial concentrations of dissolved constituents in the
water column of the River, by inhibiting dissolution. However, over time and if the oil is not recovered,
the concentrations of these constituents will peak. It is also possible that in winter the presence of an ice
cover may reduce the role of volatilization (i.e. evaporation) in removing light hydrocarbons from the
River immediately after a spill.
Aquatic oil spills are dynamic. The concentrations of the toxic constituents of crude oil vary over time
and space, as demonstrated in the case studies described in Section 2.6.4. Based on the water chemistry
data collected during and after these reference spills, this report assumes that a worst-case scenario spill
from the TMX into the North Thompson would generate sufficient concentrations of toxic constituents
to result in the maximal adverse outcomes reported in laboratory studies.
The location of the hypothetical spill will greatly influence the geographic extent and relative severity of
impacts to fish and fish habitat. A spill further upstream in the North Thompson River (e.g., Birch Island)
would enter the River at a location with lower discharge, and therefore less dilution capacity, than at a
location further downstream (e.g., McLure). A spill nearer to the mouth of the North Thompson River
or along the section of the Thompson River would be more likely to reach Kamloops Lake, all else being
equal.
While TMP ULC used a hypothetical spill volume of 1,400 m3 with a spill location near Darfield for its
QERA (TMP ULC 2013, Volume 7) (see Figure 7, Hypothetical Spill Scenario 2), the maximum modeled
cumulative outflow volumes for the section of pipeline along the North Thompson River and at the
Thompson River was about 4,000 m3 (TMP ULC 2013, Volume 7, Section 3.1.6 and Appendix B). This
estimate of outflow volume applies to a location just south of the Project’s Finn Creek Pump Station,
around pipeline kilometre 640, near Finn Creek Provincial Park and just north of a planned watercourse
crossing (approximate location 51°36'06.0"N, 119°54'51.0"W; 51.601667 N, -119.914167 W)(see Figure
7, Hypothetical Spill Scenario 1). This volume was determined under the assumption that detection of a
pipeline rupture would be immediate and that mainline valves would be fully closed within 15 minutes of
the rupture occurring (TMP ULC 2013, Volume 7).
TMX Salmon Research Study- Literature Review and Power Analysis
April 2020
Page 45
The time required for 4,000 m3 of diluted bitumen to drain from the pipeline depends on the type of
pipeline failure that leads to the spill. In the case of the Plains Midstream spill of 462.75 m3 of crude oil
into the Red Deer River, the pipeline operator initiated closure of block valves 15 minutes after the first
indication of a problem, and the pipeline was “offline” at the time (AER 2014), although the meaning of
offline was not provided. The pipeline had a diameter of about 30cm, and failure analysis after the spill
indicated that a “circumferential” (girth) pipeline weld had failed, likely due to instream exposure of the
pipeline and vibration caused by river flows (AER 2014). In contrast, the Enbridge spill of 3,190 m3 diluted
bitumen into wetlands and a tributary of the Kalamazoo River occurred over about 17 hours, during which
the operator attempted to make changes to the pressure in the pipeline (NTSB 2012). The failure
investigation found that the pipeline was ruptured along a longitudinal weld, with an opening of about 2 m
long and 15 cm wide, and that the probable cause of the rupture was corrosion fatigue cracks (NTSB
2012). For this analysis, the period of time required for the hypothetical spill volume to enter the North
Thompson River was estimated to be one hour. Again, this represents a rather extreme worst-case
scenario, since the release duration of such a large volume of diluted bitumen would depend on the size
of the rupture and would slow as pressure in the pipeline is reduced.
We considered three flow scenarios for our assessment of impacts, similar to those considered for various
case studies in the TMX QERA (TMP ULC 2013, Volume 7). These include the following:
A. Under freshet conditions (high flow, early summer), recovery may be difficult/delayed due to
flood conditions. There is a greater potential for diluted bitumen to sink if allowed to weather for
several days. A spill into high flows would maximize the extent of floodplain oiling this can
become a long-term source of oil as it is mobilized in subsequent flood/high flow events.
B. Under winter conditions, the ratio of spilled diluted bitumen to river flow (discharge) will be
much higher, and the difficulties of oil recovery from an ice-covered river will reduce oil recovery
immediately after a spill. In this case, submerged oil or sinking oil may not be a major concern as
much as the large relative volume of product spilled and delayed or ineffective recovery.
C. Under open water conditions (spring, late summer and autumn) a spill will likely result in a
more moderate risk, with more favourable recovery conditions (i.e., not high flow, not under ice)
and a moderate diluted bitumen:discharge ratio.
Conditions in these scenarios are summarised in Table 4 below.
TMX Salmon Research Study- Literature Review and Power Analysis
April 2020
Page 46
Table 4. Spill scenarios considered for a spill of diluted bitumen into the North Thompson River near
Finn Creek Provincial Park (note: this location is upstream of WSC station at Birch Island). Dilbit refers
to diluted bitumen.
Scenario
name
Timing
Flow
condition
(measured at
Birch Island)1
TSS
concentration2
Turbulence?
Oil sinking
potential?
4000 m3 oil:
one-hour
discharge
ratio
Product recovery
rate over time
Spring/early
summer
freshet
May-June (may
occur
periodically
during July and
August resulting
from storms)
high, possibly
flood flows
(median max
discharge 485
m3/s)
~41 mg/L or
higher
high, high
0.0023
~90% over three years
(estimated 49,000 -
86,000 gallons dilbit
residual in the aquatic
environment from
Marshall spill)3
Early spring
or fall
April-May, Sept-
Nov
moderate-low
(discharge 90
m3/s)
~16.4 mg/L
moderate, low
0.0123
~85% over six weeks
(Husky dilbit spill into
North Saskatchewan)4
Winter,
under ice
Dec-Mar
low (median
min discharge
30 m3/s)
~2.3 mg/L
low, low
0.0370
<10% over months,
remaining permanently
in system (Yellowstone
spill in winter)5
1Obtained from https://wateroffice.ec.gc.ca/search/historical_e.html
2 Obtained from
https://governmentofbc.maps.arcgis.com/apps/webappviewer/index.html?id=0ecd608e27ec45cd923bdcfeefba00a7
3 Federal On Scene Coordinator (FOSC), 2016, FOSC desk report for the Enbridge oil spill in Marshall, Michigan, U.S.
Environmental Protection Agency, 241 pp., url: https://www.epa.gov/enbridge-spill-michigan/fosc-desk-report-enbridge-
oil-spill, accessed September 2019.
4 Reported in Derowiz, C. 2019. ‘We’re deeply sorry:’Husky Energy fined $3.8M for leak into North Saskatchewan River.
Toronto Star June 12, 2019.
5 State and Federal Trustees State of Montana and U.S. Department of Interior. 2017. Partial Claim for Past and Future
Assessment Costs January 2015 Yellowstone River Oil Spill.
3.1.2 Calculation of contaminant concentrations
Detailed water quality data available from oil spills that have occurred in the past are scarce, especially for
light hydrocarbons including BTEX and PAHs/PACs, and especially in the hours and days immediately
following a spill. We used the following data sources to obtain water quality data for two oil spill incidents:
The Marshall, Michigan spill of diluted bitumen from an Enbridge pipeline into the Kalamazoo River in 2010
(Michigan Department of Community Health 2013) and the spill of crude oil from a Plains Midstream
pipeline into the Red Deer River in 2012 (AER 2012). This information is summarized in Table 5. After
consideration of the two spills, it was clear that maximum concentrations of BTEX and PAH constituents
in water was highest for the Kalamazoo River spill of diluted bitumen. For this reason, and with the goal
of highlighting a worst-case scenario, our focus was on the Kalamazoo River concentrations.
To estimate the maximum potential BTEX and PAH concentrations in water, we used the spilled oil to
river discharge ratio determined for the Kalamazoo River spill, compared this to the worst-case spill
volume considered of 4,000 m3, and assumed one hour of river discharge in the North Thompson River
at Birch Island under each of the spill scenario flow conditions considered (freshet, under ice,
spring/autumn open water) (summarized in Table 6). In all cases, these maximum estimates for BTEX and
PAH meet or exceed lethal concentrations. It is important to note that these concentrations are unlikely
to occur instantaneously throughout the entire system. Concentrations will fluctuate spatially and
temporally during and after the spill, and so there will be an opportunity for some fish to escape these
TMX Salmon Research Study- Literature Review and Power Analysis
April 2020
Page 47
lethal doses. Again, with the goal of highlighting a worst-case scenario, we considered the impacts that
would occur if these maximum potential concentrations occurred at all locations within the Area of
Interest located downstream of the hypothetical spill location (i.e., mainstem North Thompson River,
Thompson River, and Kamloops Lake) for a period of three months coincident with and following each of
the spill scenarios (freshet, open water, under ice). These conditions represent extreme worst-case
scenarios.
TMX Salmon Research Study- Literature Review and Power Analysis
April 2020
Page 48
Table 5. Summary of available water quality data for case study spills of crude oil and diluted bitumen
from pipelines into freshwaters, focusing on hydrocarbon constituents (BTEX and PAHs)
Incident and
References
BTEX concentration
(water)
Total PAH (water)
Oil or diluted
bitumen:water volume
ratios
Notes
Crude oil spill into
Red Deer River,
2012 (AER 2012,
2014; Teichreb
2013)
maximum recorded in river
>0.025 mg/L toluene and
xylenes, >0.01 mg/L benzene, <
0.005 mg/L ethylbenzene (day
1); max recorded in lake
>0.014 mg/L toluene, ~ 0.01
mg/L xylene, ~0.005 mg/L
benzene (day 2), no detectable
concentrations after 9 days -
downstream pulse
2-Methylnaphthalene, Fluorene,
Naphthalene, Phenanthrene
were measured, maximum
measured in river day after spill
was 0.0065 mg/L (2-
methylnaphthalene), 0.0047
mg/L (naphthalene), and 0.0004
mg/L (phenanthrene) - the latter
exceeded provincial protection
of aquatic life guidelines, and
naphthalene exceeded at times
during the following 9 days
462.75 m3 spilled into Red
Deer River at flows of
about 990 m3/s, ratio for 87
minutes of spill time is
462.75/5167800 =
0.0000895
extent of visible product spill was about
40 km on the Red Deer River, ending at
Glennifer reservoir where booms
contained the surface spill - dissolved
constituents continued to move as a pulse
downstream and into the reservoir, at
times exceeding guidelines, and detectable
in the monitored area for about 9 days.
Dilbit spill into
Kalamazoo River,
2010
(Michigan
Department of
Community Health
2013, Fitzpatrick et
al. 2015, FOSC
2016)
maximum recorded in river
0.046 mg/L toluene,
0.0043mg/L xylenes, 0.049
mg/L benzene, 0.043 mg/L
ethylbenzene (measured July
through September 2010).
many PAHs were measured,
maximum measured July
through September 2010 was
0.00022 mg/L (anthracene),
0.025 mg/L
(benzo(a)anthracene), and 0.038
mg/L (benzo(a)pyrene), 0.0028
mg/L (fluoranthene), 0.0011
mg/L (fluorene), 0.057 mg/L
(naphthalene), 0.0024 mg/L
(phenanthrene), 0.027 mg/L
(pyrene)
3190 m3 spilled into
Talmadge Creek/Kalamazoo
River at flows of about 100
m3/s for 17 hours of spill
time is 3190/6120000=
0.00052
traveled 65 km downstream over two
days, through several impoundments on
the Kalamazoo River
TMX Salmon Research Study- Literature Review and Power Analysis
April 2020
Page 49
Table 6. Estimates of hydrocarbon constituent concentrations (BTEX and PAHs) derived from reported
concentrations following the spill of diluted bitumen into the Kalamazoo River in 2010.
Scenario
name
Parameter
Unit
(water)
Maximum
reported
(Kalamazoo
River)1
Case study
oil:water
ratio
(Enbridge)
(unitless)
4000 m3 oil:
one-hour
discharge
ratio
(unitless)
Maximum
estimated
concentration
Spring/early
summer
freshet
benzene
mg/L
0.049
0.00052
0.0023
0.216
ethylbenzene
mg/L
0.043
0.189
toluene
mg/L
0.046
0.203
xylene
mg/L
0.043
0.189
acenaphthene
mg/L
0.0004
0.002
acenaphthylene
mg/L
0.00013
0.001
anthracene
mg/L
0.00022
0.001
benzo(a)anthracene
mg/L
0.025
0.110
benzo(a)pyrene
mg/L
0.038
0.167
benzo(b)fluoranthene
mg/L
0.0025
0.011
benzo(g,h,i)perylene
mg/L
0.0022
0.010
benzo(k)fluoranthene
mg/L
0.00054
0.002
chrysene
mg/L
0.028
0.123
dibenzo(a,h)anthracene
mg/L
0.001
0.004
fluoranthene
mg/L
0.0028
0.012
fluorene
mg/L
0.0011
0.005
indeno(1,2,3-cd)pyrene
mg/L
0.00097
0.004
naphthalene
mg/L
0.057
0.251
phenanthrene
mg/L
0.0024
0.011
pyrene
mg/L
0.027
0.119
Early spring
or fall
benzene
mg/L
0.049
0.00052
0.0123
1.163
ethylbenzene
mg/L
0.043
1.021
toluene
mg/L
0.046
1.092
xylene
mg/L
0.043
1.021
acenaphthene
mg/L
0.0004
0.009
acenaphthylene
mg/L
0.00013
0.003
anthracene
mg/L
0.00022
0.005
benzo(a)anthracene
mg/L
0.025
0.594
benzo(a)pyrene
mg/L
0.038
0.902
benzo(b)fluoranthene
mg/L
0.0025
0.059
benzo(g,h,i)perylene
mg/L
0.0022
0.052
benzo(k)fluoranthene
mg/L
0.00054
0.013
chrysene
mg/L
0.028
0.665
TMX Salmon Research Study- Literature Review and Power Analysis
April 2020
Page 50
dibenzo(a,h)anthracene
mg/L
0.001
0.024
fluoranthene
mg/L
0.0028
0.066
fluorene
mg/L
0.0011
0.026
indeno(1,2,3-cd)pyrene
mg/L
0.00097
0.023
naphthalene
mg/L
0.057
1.353
phenanthrene
mg/L
0.0024
0.057
pyrene
mg/L
0.027
0.641
Winter,
under ice
benzene
mg/L
0.049
0.00052
0.0370
3.490
ethylbenzene
mg/L
0.043
3.063
toluene
mg/L
0.046
3.276
xylene
mg/L
0.043
3.063
acenaphthene
mg/L
0.0004
0.028
acenaphthylene
mg/L
0.00013
0.009
anthracene
mg/L
0.00022
0.016
benzo(a)anthracene
mg/L
0.025
1.781
benzo(a)pyrene
mg/L
0.038
2.707
benzo(b)fluoranthene
mg/L
0.0025
0.178
benzo(g,h,i)perylene
mg/L
0.0022
0.157
benzo(k)fluoranthene
mg/L
0.00054
0.038
chrysene
mg/L
0.028
1.994
dibenzo(a,h)anthracene
mg/L
0.001
0.071
fluoranthene
mg/L
0.0028
0.199
fluorene
mg/L
0.0011
0.078
indeno(1,2,3-cd)pyrene
mg/L
0.00097
0.069
naphthalene
mg/L
0.057
4.060
phenanthrene
mg/L
0.0024
0.171
pyrene
mg/L
0.027
1.923
1 Data taken from Michigan Department of Community Health. 2013. Public Health Assessment: Kalamazoo River/Enbridge Spill: Evaluation of
Kalamazoo River surface water and fish after a crude oil release
TMX Salmon Research Study- Literature Review and Power Analysis
April 2020
Page 51
3.2 Estimating Impacts on Salmon in the North Thompson
Watershed
Research Objective: Given specific hypothetical spill conditions, what are the potential impacts of a spill of a
diluted bitumen product in the North Thompson River on Chinook and Sockeye salmon?
Below, we describe the details and parameters for Sockeye and Chinook population dynamics that we
compiled for our analysis. The data on the fecundity (i.e. production of offspring) and survival of each life
stage in the watershed were used to understand how many adults could possibly return each year to the
Area of Interest to spawn. Within the salmon lifecycle, each life stage will use a variety of habitats between
marine and freshwater environments. Life stages can also vary in which habitats they use within these
environments. For example, some salmon populations may spawn in a tributary (outside the vulnerable
Area of Interest) but emerging fry may migrate downstream into the North Thompson River six months
later (into the vulnerable Area of Interest). Other populations may be inside or outside the Area of
Interest the whole time. To address some of these complex components of the salmon life cycle, we
categorized the spatial extent of each life stage into areas of the watershed. For example, for each of our
key species, we grouped life stages that would be found in the mainstem of the watershed compared to
tributaries. We then estimated how much and how long each of the population within those life stages
use that habitat. We used these estimates to create a baseline for returning Sockeye and Chinook adults
under an unimpacted scenario. We then used the literature review on diluted bitumen impacts from
Section 2.7 to estimate how hypothetical oil spills could impact the baseline life cycle for each species and
local population given the spatial extent and timing of that spill compared to where and how many salmon
would be affected. This then allowed us to compare changes in the returning adults under each oil spill
scenario relative to baseline.
We present our process in the sections that follow but further details on life history parameters used in
the analysis can be found within Appendix A and further the detailed technical report in Appendix B or
at http://www.poissonconsulting.ca/f/1714378447.
3.2.1 Estimating Baseline Salmon Survival in the Area of Interest
3.2.1.1 Life Cycles
The life cycles of Chinook and Sockeye salmon can both be broken into six life stages (eggs, fry, smolt,
sub-adult, pre-spawning adult, and spawning adult) which vary in their use of freshwater or marine habitats
(Quinn 2018). As noted in Section 2.3, spawning adults deposit eggs in stream gravels which incubate over
winter before emerging as fry in the spring. In their first spring, emerging fry migrate downstream to their
rearing habitats (streams in the case of Chinook and lakes in the case of Sockeye) t o grow and develop
into smolts. Smolts subsequently migrate downstream to the sea in the following spring. After three years
at sea in the case of Chinook and two years in the case of Sockeye, sub-adults begin returning to the BC
coast as pre-spawning adults and begin the spawning migration as spawning adults upstream to their natal
freshwater streams to deposit their eggs in the stream gravels. Given the number of eggs each female
deposits and the carrying capacity of the streams, the abundance of the remaining life stages is defined by
the survival rates from one stage to the next. These lifecycle parameters are defined in Table A1and a
discussion of the density dependent relationships for the life cycles can be found in Appendix A.
North Thompson Chinook
TMX Salmon Research Study- Literature Review and Power Analysis
April 2020
Page 52
Chinook in the North Thompson typically express a stream-type life cycle. Adult Chinook in the Blue,
Finn, and Raft Rivers migrate upstream in the spring, while the remaining populations in the North
Thompson migrate upstream in the summer. Adult female Chinook average 6,184 eggs in the North
Thompson but not all females expend their eggs before senescence (i.e. deterioration) and mortality (Scott
et al. 1982c). Female spawning success (i.e., percent eggs retained per female) is ~2.2% in the mainstem
North Thompson and ~15% in tributaries (e.g., Raft River and Finn Creek). While Joseph and Lemieux
Creeks had juvenile sampling done in early studies (Scott et al. 1982c), they were not included in our
quantitative analyses because we lacked adult monitoring for these tributaries to contrast findings.
Adult Chinook spawn throughout the North Thompson with the largest habitat in the mainstem North
Thompson; however, there are also productive and high-density tributaries such as Finn Creek (Table 7
below, as well as Table A2 in Appendix A). Eggs incubate over the winter and develop into fry that emerge
in the spring. Egg-fry survival is highest in Finn Creek, Raft River, and Clearwater River (14-69%) with
reduced survival in the North Thompson, Barriere River, Lemieux Creek, and Joseph Creek (Table 7).
Table 7 Chinook Salmon adult abundance (N), fecundity, egg deposition, fry population estimates, and
egg-to-fry survival in the North Thompson watershed.
Site
Adult N
Fecundity
Eggs (Million)
Fry estimate
% Survival
Finn Creek
550-1025
6225
1.8-2.2
810,000
37-69
Raft River
200-405
5837
.5-1.2
150,000
14-28
Clearwater River
3000
6184
9.3
2,900,000
32
Joseph Creek
-
6184
-
20,000
-
Lemieux Creek
15
6184
0.05
1,600
3
Barriere River
100
6184
0.3
7,100
2
North Thompson
2250-4000
6490
7.3-13
1,200,000
9-18
Chinook fry emerge from their natal streams in the late spring and migrate into rearing streams and rivers
(Table A3 and A4 in Appendix A). Some fry migrate downstream of the Area of Interest while others will
rear in the mainstem North Thompson (Table A5 in Appendix A). Peak fry migration tends to occur later
in the season, further downstream in the North Thompson watershed, likely tracking water temperatures
(Scott et al. 1982a). Key tributaries like Finn Creek, Raft Rivers, and Clearwater Rivers tend to produce
relatively large fry cohorts (0.15-2.9 Million) owing to high incubation, survival, or growth in early life
(Scott et al. 1982b; Stewart et al. 1983). Joseph and Lemieux Creeks and Barriere River tend to produce
small fry cohorts due to poor incubation, survival or growing conditions.
Most fry and smolts produced within the North Thompson watershed rear downstream of the Kamloops
Lake area (Stewart et al. 1983). Sampling from the early 1980s estimated ~18% of fry cohorts and ~1% of
smolt cohorts were rearing within the mainstem North Thompson suggesting large downstream migration
from the portion of the watershed likely to be directly impacted by TMX (Scott et al. 1982a; Stewart et
al. 1983). However, this data is from some highly variable sampling in high-flow conditions suggesting the
estimate of ~18% is on the lower range. Most fry rearing within the North Thompson watershed rear
within the mainstem North Thompson and not tributaries (Table A5 in Appendix A).
TMX Salmon Research Study- Literature Review and Power Analysis
April 2020
Page 53
Fry rear for one year in their rearing stream or rivers before developing into smolts and outmigrating
towards the Pacific Ocean. The average fry-smolt survival is ~37% (Bradford 1995). Chinook smolts
migrate from their rearing streams and rivers into the Fraser estuary and reside for 30 days. Smolt
outmigration takes 10 days. In tributaries, smolt outmigration begins in mid-April and ends in early May;
smolts in the mainstem North Thompson begin outmigrating in late April and end in mid July (Table A6 in
Appendix A).
Variation in smolt outmigration timing may be linked to body sizes (Table A7 in Appendix A). For example,
smolts in Lemieux Creek appear to leave earliest at the smallest body sizes suggesting poor fry-smolt
rearing and better foraging conditions in the mainstem North Thompson or Fraser estuary.
Juvenile/sub-adult Chinook spend three years at sea before beginning migrations to their natal stream.
Marine survival for Chinook Salmon, i.e., survival from smolts to pre-spawning adults, can vary from year-
to-year with an average coefficient of variation of ~12.5% (Bradford 1995). Together with on-route
mortality, this suggests that smolt-to-spawning adult survival for Chinook typically ranges 0.62.1%
consistent with observations in Duffy and Beauchamp (2011). Spawning adults begin arriving in their natal
streams from July to August, depending on the population (Table A8 in Appendix A). Spawning occurs a
few days after peak arrival beginning in early August and ends by early October and post-spawn die-off
occurs ~7-14 days later.
North Thompson Sockeye
Sockeye Salmon in the North Thompson watershed have a simpler life history and spatial structure than
Chinook. Spawning habitat is predominately the mainstem North Thompson, Finn Creek, and Raft River
(rearing in Kamloops Lake) and Fennell and Harper Creeks (rearing in Barriere Lake). As such, Sockeye
in the North Thompson watershed are composed of two Early Summer (ES) CUs: (1) North Barriere-ES
(de novo) and (2) Kamloops-ES (composed of Raft River and miscellaneous North Thompson) (Grant et
al. 2011, 2018). Of the Kamloops-ES CU, Sockeye from the Raft River composes ~35-60% of the total
escapement, miscellaneous tributaries ~5%, and the rest coming from the mainstem North Thompson.
The total spawner abundance for miscellaneous tributaries of the North Thompson approximated 7,000
returning adults in 2018 (DFO 2018a). The majority of North Thompson sockeye exhibit a four-year life
cycle (termed a 42 life cycle where lake-rearing Sockeye outmigrate to the sea in their second year). Age
structure of returning adults in the Kamloops-ES are 71% four year olds and 29% five year olds (DFO
2018a). Age structure of returning adults in the North Barriere-ES are 80% four year olds and 20% five
year olds (DFO 2018a). Marine survival for Sockeye salmon, i.e., survival from smolts to pre-spawning
adults, can vary exhibit more year-to-year variability than Chinook an average coefficient of variation of
~78% (Bradford 1995). Together with on-route mortality, this suggests Sockeye smolt-adult survival can
range 1.5-24%. Sockeye adults return to the North Thompson watershed beginning in mid-August and
early-September (Table A9 in Appendix A).
Fry emergence begins in early April with migration to Kamloops Lake (Table A10 in Appendix A). Sockeye
fry from the Fennell and Harper Creek populations rear in Barriere Lake. Sockeye fry rear in Kamloops
Lake for approximately one year before migrating to the Fraser estuary as smolts. Smolt migration begins
~April 10th, peaks ~April 28th and ends by May 28th (Welch et al. 2009). The outmigration lasts 5 days and
sockeye spend ~30 days in the estuary (DFO 2016). Sockeye juveniles and subadults spend ~2 years at
TMX Salmon Research Study- Literature Review and Power Analysis
April 2020
Page 54
sea before migrating to their natal streams for spawning in early summer. On route mortality is ~35-40%
for returning adults (Bradford 1995; Martins et al. 2011).
These baseline vital rates for Chinook and Sockeye were included in the next step of our analysis, where
we examined how these survival rates might change under hypothetical spill scenarios (Sockeye vital rates
are in Table A11 in Appendix A).
3.2.2 Estimating Salmon Survival Under Hypothetical Spill Scenarios
Impacts from the spill scenarios were translated into changes in fish survival across the Chinook and
Sockeye life cycle for the mainstem North Thompson and tributary populations. The following impacts to
survival are based on a literature review (Table 1 in Section 2.7.2) under scenarios of the maximum
concentration of a spill releasing 4,000 m3 in a short time period (1 hr) and the spatial variation in
demography and habitat usage between the mainstem and tributaries. All impacts to fry survival were
assumed to occur after density dependent egg-to-fry survival. Below we describe mortality rates for the
life stages present during these three spill scenario timings based on the current knowledge of toxicity
effects to salmon:
Spills during the Freshet scenario would occur when young salmon are smallest and the most
vulnerable in their development. Freshet spills could affect fry-to-smolt survival for those fry
rearing in the mainstem North Thompson River and Kamloops Lake. Based on our synthesis of
research (Table 1 in Section 2.7.2), assumptions on impacts were that Freshet spills would increase
mortality by 10.5% for the egg-to-fry stage (a lower range of mortality caused by the short
duration) and 29.5% for the fry-to-smolt stage (Alderman et al. 2018; Lin et al. 2019). Sublethal
effects of cohorts exposed to diluted bitumen was included as a 6.5% reduction to adult fecundity
(Heintz et al. 2000).
Spills during the Fall scenario would occur during moderate flow and would impact returning
adults, eggs from that brood year, and rearing fry or smolts from previous brood years for
populations in the mainstem North Thompson River. Fall spills would not affect fry or smolts that
reared outside of this area (~82% of fry cohorts may rear outside the impacted area; Table A5 in
Appendix A) or eggs spawned in the tributaries. Based on our synthesis of research (Table 1 in
Section 2.7.2), assumptions, on impacts were that Fall spills would increase mortality by 36% for
the egg-to-fry stage (a moderate mortality impact from longer duration), 29.5% for the fry-to-
smolt stage (Alderman et al. 2018; Lin et al. 2019), and 15% for returning adults (Heintz et al.
2000). sublethal effects of cohorts exposed to diluted bitumen was included as a 6.5% reduction
to adult fecundity (Heintz et al. 2000).
Spills during the Winter scenario would occur during lowest flow and under ice. Winter spills
would affect all eggs from that brood year and any rearing fry in the mainstem North Thompson
from previous brood years. Winter spills would particularly affect sockeye cohorts rearing in
Kamloops Lake. Based on our synthesis of research (Table 1 in Section 2.7.2), assumptions, on
impacts were that Winter spills would increase mortality by 43% for the egg-to-fry stage (the
highest impact) and 29.5% for the fry-to-smolt stage (Alderman et al. 2018; Lin et al. 2019).
Sublethal effects of cohorts exposed to Winter diluted bitumen spills was included as a 33%
reduction to adult fecundity a higher non-lethal impact than the other two scenarios (Heintz et
al. 2000).
TMX Salmon Research Study- Literature Review and Power Analysis
April 2020
Page 55
The above described mortality estimates from toxicity effects due to exposure to oil on different salmon
life stages (based on literature review outlined in Table 1 in Section 2.7.2), were used to estimate changes
on the average survival and fecundity of Chinook and Sockeye salmon life stages for each hypothetical spill
scenario (Table 8).
Table 8. Average survival and fecundity (number of eggs) of North Thompson Chinook and Sockeye
salmon across life stages (egg, fry, smolt, pre-spawning adult, spawning adult) and plausible effects of
diluted bitumen spills at highest concentration to survival and fecundity of exposed cohorts. Note -
specific Conservation Units can vary in their own vital rates, and the sub-adult life stage is merged
with pre-spawning adults. On-route survival between pre-spawning and spawning adult assumes
water temperatures ≤16°C.
Stage
Chinook
Sockeye
Baseline
survival
(%)
Survival after spill (%)
Baseline
survival
(%)
Survival after spill (%)
Freshet
Fall
Winter
Freshet
Fall
Winter
Eggs-Smolt1,2
7.7
-
-
-
2.1
-
-
-
Eggs-Fry1-4
21.0
18.8
13.4
11.2
9.3
8.3
5.9
5.3
Fry-Smolt1-4
36.7
25.8
25.8
25.8
21.9
15.4
15.4
15.4
Smolt-Adult1,2
1.4
-
-
-
6.2
- </