INCORPORATING AN ETHIC OF CONTEXT AND PLACE IN MECHANISTIC
RESEARCH: A PLACE-BASED WILDFIRE RISK ASSESSMENT IN THE XÁXLI’P
SURVIVAL TERRITORY
by
Aita Bezzola
B.A. & Sc., McGill University, 2012
ADGIS, Selkirk College, 2013

THESIS SUBMITTED IN PARTIAL FULFILLMENT OF
THE REQUIREMENTS FOR THE DEGREE OF
MASTER OF SCIENCE
IN
NATURAL RESOURCES AND ENVIRONMENTAL STUDIES

UNIVERSITY OF NORTHERN BRITISH COLUMBIA
July 2020
© Aita Bezzola, 2020

Abstract
Wildfires pose a significant economic and social challenge to communities
throughout British Columbia. For some Indigenous communities, a large landscape
fire has the potential to change their traditional territory and communities
permanently. To allocate limited resources to the costly effort of wildfire mitigation,
communities need a baseline for the spatial distribution of risk. For the Xáxli’p and
their community forest, the wildfire risk is an urgent concern, locally effected by
forest fuels, human ignition, and wind. Local knowledge of community members
gathered through workshops were used to validate existing forest, wind, and access
data. Using existing data and community data, areas of higher risk and other
landscape considerations were identified and mapped to support planning by the
Xáxli’p Community Forest to create a fire-resilient landscape.

ii

Table of Contents

Abstract

…………………………………………………………………………………..ii

Table of Contents .................................................................................................................. iii
List of Figures ....................................................................................................................... vii
List of Tables ............................................................................................................................ x
Acknowledgement ................................................................................................................ xi
Chapter 1: Introduction to the Xáxli’p Community Forest, the Research
Collaboration and Wildfire .............................................................................. 1
Research Introduction ..................................................................................................... 1
Research Grounded in People and Place ................................................................... 2
Xáxli’p Pursuit of Self-determination. ..................................................................... 5
Collaboration for Localized Solutions ........................................................................ 8
Research Objective. ..................................................................................................... 9
Applying Indigenist Philosophy ............................................................................... 11
Community-Directed Research. ............................................................................. 13
The Position of the Researcher. ............................................................................... 14
Research for Reconciliation ..................................................................................... 15
Cultural Understanding of Wildfire and Ecosystem Health ................................... 15
Anthropogenic Fires ................................................................................................... 16
Fire Regime................................................................................................................... 18
Land Management Effects on Wildfire Ecosystems ............................................... 19
Institutional Culture of Wildfire Suppression. ..................................................... 20
Fire Regime Changes on the Landscape. ............................................................... 21
Fuel Accumulation in Wildfire Suppressed Landscapes. ................................... 22
Facilitating Wildfire Risk Mitigation .......................................................................... 24
Xáxli’p Eco-cultural Restoration................................................................................ 28
The Components of Wildfire Risk and their Mitigation Strategies ...................... 31
iii

Fuel Risk Mitigation. ................................................................................................ 32
Prescribed Burning. .............................................................................................. 33
Thinning and Biomass Removal. ........................................................................ 35
Comparison. .......................................................................................................... 36
Fire Weather Risk Factors. ....................................................................................... 37
Wind. ...................................................................................................................... 38
Climate Change. .................................................................................................... 38
Geography and Wildfire Risk. ................................................................................ 39
Variable Terrain Creates Highly Variable Burn Conditions. .......................... 40
Human Driven Ignition Patterns. ....................................................................... 41
Limitations Created by The Road Network...................................................... 42
Conclusion ...................................................................................................................... 42
Chapter 2: Exploration of Existing Datasets for Wildfire Risk Assessment Guided by
Community ...................................................................................................... 44
Connecting Research Purpose to Data ....................................................................... 44
Developing the Research Question with the XCF .................................................. 44
Xáxli’p Ecosystem Based Conservation Plan Dataset. ......................................... 45
Fuel Classification in Canada .................................................................................... 49
Description of the FBP Fuel Categories. ................................................................ 51
Converting Ecosystem Data to Fuel Categories ........................................................ 53
Combining EBCP and VRI Datasets ......................................................................... 53
Focusing on Locally Created Data to Improve Classification............................... 59
Improving Iterations of Fuel Classification ............................................................. 62
Fire Risk Beyond Fuel ................................................................................................... 63
Assessing Wind as a Risk Factor ............................................................................... 64
Incorporating Access as an Ignition Proxy .............................................................. 67
A Need for More Context ............................................................................................. 70
iv

Chapter 3: Enhancing Wildfire Risk Datasets using Xáxli’p Community Knowledge
of the Local Wildfire Landscape .................................................................... 71
Holding Conversations with Knowledge Holders ................................................... 71
Identifying and Mapping Concerns About Wildfire.............................................. 72
Site Visits and Discussing Solutions ......................................................................... 75
Wildfire Risks Discussed in the Workshop................................................................ 77
Fuel Risks Identified Around High Occupancy Areas .......................................... 78
Transmission of Wildfires from Ignition to Forest via Grasslands. .................. 79
Risks Posed by Dense Forest Fuels. ....................................................................... 82
Other Fuel Concerns. ................................................................................................ 84
Ignition Risks Identified ............................................................................................. 85
Human Ignition Patterns in the Territory. ............................................................ 86
Road Access Associated with Ignition Patterns. .................................................. 87
Lightning Ignitions. .................................................................................................. 89
Wind Patterns............................................................................................................... 89
Risk Mitigation Measures Discussed .......................................................................... 91
Creating Fuel Management Areas ............................................................................ 91
Controlled Burns ......................................................................................................... 94
Maintenance to Reduce Fuels .................................................................................... 94
Emergency Response .................................................................................................. 96
Conclusion ...................................................................................................................... 97
Chapter 4: Integrating Data Types and Exploring Key Learnings ................................ 98
Integrating the Quantitative Categorical Data .......................................................... 99
Combining Fuel and Wind Risk Data ...................................................................... 99
Wildfire Ignition Risk Attributed to Road Access Narrows the Focus Area .... 101
Overlaps in High Fuel, Wind and Ignition Risks ................................................. 102
Refining Risk Indicators Using Community Perspectives ..................................... 104
Community Experience with Access and Ignition Patterns................................ 104
v

Community Fuel Priorities Emphasize the Risk of Grasslands ......................... 105
Community Identified Wind Patterns.................................................................... 108
Data and Research Limitations .................................................................................. 108
Uncertain Accuracy of Fuel Data ............................................................................ 109
Localized Wind Data for Complex Topography .................................................. 110
Ignition Depends on More than Roads .................................................................. 110
Synthesis of Findings .................................................................................................. 111
Community Perspectives Focus Broad Data Assumptions ................................. 112
Growth and Challenges for the Researcher ........................................................... 112
Recommendations for Research and Management ................................................ 115
Future Research Opportunities ............................................................................... 116
Improving Fuel Data. ............................................................................................. 116
Implementation of Wind Monitoring. ................................................................. 117
Assessment of Temporal Variability. ................................................................... 117
Land Management Recommendations .................................................................. 118
Managing Grassland Fuels. ................................................................................... 118
Continued Forest Thinning. .................................................................................. 119
Creation of Fire Management Areas. ................................................................... 120
Community Communication and Monitoring. .................................................. 121
Conclusion .................................................................................................................... 122
Bibliography ........................................................................................................................ 123

vi

List of Figures
Figure 1: Boundaries of the Xáxli’p Survival Territory and Community Forest. .......... 3
Figure 2: Timeline of some key Xáxli’p events towards achieving self-determination
(Diver, 2016). ............................................................................................................................ 6
Figure 3: XCF forest crew and staff spring/summer 2017 (Green, S. 2017). ................... 7
Figure 4: Fire scarring on old Douglas fir tree (Green, S. 2017). .................................... 16
Figure 5: Dense regeneration associated with wildfire suppression (Green, S. 2017).
................................................................................................................................................. 16
Figure 6: History of fires recorded in the Xáxli’p Survival Territory by the Provincial
Government (Ministry of Forests, Lands, Natural Resource Operations and Rural
Development - BC Wildfire Service, 2019). ....................................................................... 19
Figure 7: Wildfire history in Pavilion Ranges eco-section (Ministry of Environment
and Climate Change Strategy - Knowledge Management, 2019; Ministry of Forests,
Lands, Natural Resource Operations and Rural Development - BC Wildfire Service,
2019). ....................................................................................................................................... 22
Figure 8: Open ponderosa pine stands at the south end of the Survival Territory
illustrates an ecosystem characteristic of a low-intensity fire regime that is non-fatal
to mature Douglas fir and ponderosa pine (Green, S. 2017). ......................................... 23
Figure 9: Examples of forests in the Survival Territory with an accumulation of fuels
from closely spaced trees and dead woody debris, increasing wildfire risk (Green
2017). ....................................................................................................................................... 24
Figure 10: Open patches visible in the foreground as well as on the opposite side of
the valley build a fire absorbent landscape in parts of the Xáxli’p Survival Territory
(XCF Forest Crew, 2017). ..................................................................................................... 27
Figure 11: Example of an XCF restoration area with thinned trees and low branches
removed (Green, S. 2017) ..................................................................................................... 30
Figure 12: Edge of restoration area – thick regrowth associated with fire suppression
(back trees) compared to restoration ecosystem (foreground) (Green, S. 2017) .......... 30
Figure 13: Examples of forest thinning undertaken for eco-cultural restoration in the
Survival Territory with debris left on the ground for ecological benefits (Green, S.
2017). ....................................................................................................................................... 36
Figure 14: The complex terrain of the Xáxli’p Survival Territory that supports varied
ecosystems and fire behaviour (XCF Forest Crew, 2017)................................................ 40
vii

Figure 15: The variation of Ecosystem Based Conservation Plan ecosystem
descriptors in the Survival Territory. ................................................................................. 48
Figure 16: VRI and EBCP combined data, A) first iteration Open and Sparse
combined, B) Open and Sparse categories separated...................................................... 56
Figure 17: Projected forest age from the BC VRI dataset (Ministry of Forests, Lands,
Natural Resource Operations and Rural Development - Forest Analysis and
Inventory, 2017)..................................................................................................................... 57
Figure 18: Example of a west-facing picture, 1 of 4 taken at a photo-point (XCF
Forest Crew, 2017). ............................................................................................................... 58
Figure 19: Example of a set of pictures captured at a photo-point in the four cardinal
directions (XCF Forest Crew, 2017). ................................................................................... 60
Figure 20: Progression of the EBCP fuel categories through different iterations of
information and knowledge. ............................................................................................... 62
Figure 21: Iterations of fuel classification from VRI/EBCP classification to straight
EBCP categories. ................................................................................................................... 63
Figure 22: Comparison of predominant winds at the Lillooet weather station (Pacific
Climate Impacts Consortium, 2019). .................................................................................. 65
Figure 23: Modelled windspeed across the Survival Territory for the average
summer fire weather conditions at the Lillooet Airport; A) the modelled raster
output, B) the average wind speed for each EBCP ecosystem unit. .............................. 67
Figure 24: Distribution of fire ignitions relative to access roads. ................................... 69
Figure 25: Quantitative data mapped in earlier research stages and presented for
correction and comment to the focus group. .................................................................... 73
Figure 26: Wall maps used to capture areas identified by focus group participants. . 74
Figure 27: 'Dot-mocracy' colour legend (A) and participants placing 'dot-mocracy'
stickers (B) (Bezzola, 2018) .................................................................................................. 75
Figure 28: Sites visited with participants on the second workshop day to see some
examples of wildfire issues identified on day 1. .............................................................. 76
Figure 29: Areas contributing to wildfire risk identified by the focus group. ............. 77
Figure 30: Grass and shrubland contributing to wildfire risk around the north of the
valley and Kwotlenemo Lake and Chilhil Lake identified by the focus group. .......... 80
Figure 31: View looking south-east over Chilhil lake showing Fountain Valley Road
providing easy access to the lake and fields that have an accumulation of dry grassy
fuels. In the foreground, you can see dead standing trees on the hillside on the west
side of the lake (XCF Forest Crew, 2017). .......................................................................... 82
viii

Figure 32: Forest areas contributing to wildfire risk identified through the centre of
the valley by workshop participants.................................................................................. 83
Figure 33: Fieldtrip viewpoint from Sobotka’s ranch overlooking the valley and the
dead standing trees on the west side of Fountain Valley (Bezzola, 2018). ................... 84
Figure 34: Sign at the northern entrance of Fountain Valley, informing residents and
visitors of wildfire hazard rating, access limitation and fire ban to reduce the risk of
ignition (Bezzola, 2018). ....................................................................................................... 86
Figure 35: Areas identified for high ignition risk associated with road access. .......... 88
Figure 36: Areas identified for particularly high wind by the focus group. ................ 90
Figure 37: Fuel breaks identified by focus group participants. ..................................... 93
Figure 38: Example of forest thinning undertaken by the forest crew for eco-cultural
restoration near Chilhil Lake (Green, S. 2017). ................................................................. 95
Figure 39: Average wind speed in ecosystem units with high-risk fuels (C4, C3, and
M1-75). .................................................................................................................................. 100
Figure 40: Fuel risk categories within 500 metres of roads. .......................................... 102
Figure 41: Average windspeeds for ecosystem units within 500 metres of roads. .... 102
Figure 42: Average wind speeds stratified into high (<10 m/s) and low categories (>10
m/s) in high-risk fuel types (C4, C3, M1-75) and high ignition risk areas (within
500m of roads) to show the integration of the three key risk variables. ..................... 103
Figure 43: Fuel risk categories in the high ignition risk areas identified in the
community workshop compared to the ignition risk of the entire road network
buffer. ................................................................................................................................... 105
Figure 44: The fuel classifications in the areas of fuel concern identified by the
community. .......................................................................................................................... 106
Figure 45: Grassland area near Chilhil lake that could quickly transmit fire to
surrounding dense forests under the right conditions (Green, S. 2017). .................... 107
Figure 46: Discussing photopoints and field data with the XCF forest crew (XCF
Forest Crew, 2017). ............................................................................................................. 114
Figure 47: Areas recommended for targeted fuel management. ................................. 119
Figure 48: Example of natural fire-resistant ecosystems that could be tied into other
eco-cultural restoration activities to create larger fire management areas (Green, S.
2017). ..................................................................................................................................... 120

ix

List of Tables
Table 1: Ecosystem descriptors used in the Ecosystem Based Conservation Plan
dataset..................................................................................................................................... 46
Table 2: Canadian Forest Fire Behaviour Prediction (FBP) system fuel categories
(Taylor, Pike, & Alexander, 2000). ...................................................................................... 52
Table 3: VRI and EBCP - V1 and V2 classification categories. ........................................ 54
Table 4: History of ignitions in Fountain Valley comparing natural and human
ignitions within buffer distances from the road network. .............................................. 68
Table 5: Discussion questions used to guide the conversation with the focus group. 73

x

Acknowledgement
I want to thank the Xáxli’p community and particularly the Xáxli’p
Community Forest employees and board for their time, patience, sharing their
knowledge and allowing me to work with them. I am grateful and humbled by the
generosity shown by the Indigenous community members I have had the privilege
to work with through the years working on my thesis, both related to my thesis
work and as part of my extra-thesis learning. Thank you for sharing and everything
you do; I will pay it forward.
I am also grateful for the autonomy provided and encouraged by my
supervisor Scott Green and committee members Herb Hammond and Agnes
Pawlowska-Mainville to learn, explore and create my own experience. The learning
this allowed (and pushed me towards) has no substitute.

xi

Chapter 1: Introduction to the Xáxli’p Community Forest, the Research
Collaboration and Wildfire
Research Introduction
This research is a partnership with the Xáxli’p Community Forest (XCF),
created by and for the Xaxli’pmec of Fountain Valley in what they refer to as their
“Survival Territory.” The relationships between Indigenous communities and
researchers have a history of imbalance (L. Smith, 1999), with many researchers
misrepresenting these communities and claiming expertise through the “objective”
scientific method that often removes relationships in deference to statistics and
classifications (L. Smith, 1999). Guided by Indigenist philosophy and communitydirected methods, this research explores how western science can support
Indigenous knowledge and self-determination – specifically supporting Xáxli’p ecocultural restoration and wildfire mitigation by providing specific technical expertise
and resources.
Throughout colonial history in Canada, Indigenous peoples have been forced
to make considerable cultural compromise and adaptations when differences arose,
subjecting them to government and institutional authority and practice (King, 2012).
The Xáxli’p Survival Territory was claimed by the crown and sold, their language
and traditional livelihoods were suppressed, and their children were sent to
residential school (Drake-Terry, 1989; T. Smith, 1998). With such legacies, the burden
should now be on researchers to put in the work and reach across this cultural gap
1

to help create solutions that are suitable for the individual culture and unique
setting of the community.
Research Grounded in People and Place
Xáxli’p is an Indigenous community located in the mountains of
southwestern British Columbia in the Pavilion Ranges eco-section (Ministry of
Environment and Climate Change Strategy - Knowledge Management, 2019) 10 km
east of Lillooet looking over the Fraser River from Fountain Valley (Figure 1). They
are a member of the Lillooet Tribal Council and the second-largest community of the
St’át’imc Nation with around 1059 members, about 263 who live on their reserves
(Indigenous and Northern Affairs Canada, 2019).

2

Figure 1: Boundaries of the Xáxli’p Survival Territory and Community Forest.

The Xáxli’p traditional territory covers about 31,419 ha from the height of
land on the western side of Fountain Valley to past the height of land east of the
valley (Diver, 2016). Xáxli’p community members started referring to this area as the
“Xáxli’p Survival Territory” after a Xáxli’p Elder coined the phrase at a community
meeting because txhough their traditional territory covers a small area it provided

3

for their livelihood for thousands of years (Diver, 2016). The core of the Survival
Territory is Fountain Valley, where the local topography transitions from the Fraser
River Valley at about 200 m above sea level (asl) to a valley which connects back to
the Fraser River Valley at the south end surrounded by steep and rugged mountains
reaching over 2000 m asl. Xáxli’p means “brow of the hill” reflecting the location of
their main village at the north end of the valley on a terrace above the Fraser River
(Xáxli’p, 2017).
The Xáxli’p connection with this land goes back millennia, with evidence that
the region was occupied 8000 years ago, with nearby archaeological sites of villages
dating back 1,600 – 2,600 years (Diver, 2016). The Survival Territory provided for the
Xáxli’p (and continues to in a reduced scope today); “from a birds-eye view of
looking down in Fountain Valley, it’s like an oasis within an arid area” (Diver, 2016,
p. 31), creating habitat for deer for hunting, berries and plants for food and
medicines, and annual salmon runs in the Fraser River which still brings Xáxli’p
families to their fishing holes in August to catch and dry salmon in the winds off the
Fraser (Diver, 2016; Xáxli’p Community Forest, 2017). A network of historic trails
traces the territory connecting different ecosystems and elevations and are an
essential part of Xáxli’p culture used today by hunters and the community to bring
youth up into the mountains (Diver, 2016; Xáxli’p Community Forest, 2017). The
Traditional Ecological Knowledge (TEK) of Xáxli’p Elders is rooted in the wisdom of
4

many generations of ancestors that lived in the Survival Territory, and their way of
life depends on relationships with the land and passing that on to future
generations.
The last century has brought ranching, recreational development and high
grade-logging in the Survival Territory. However, with the valley’s small size, the
community recognizes that resource extraction cannot be hidden away without
impacting them. Due to the ecological importance of water in the arid climate,
maintaining the local hydrology is a priority for community members (Weinstein,
1995), which led to the XCF gradually ending grazing and clear-cut logging in the
valley by 2008 (Diver, 2016).
Xáxli’p Pursuit of Self-determination.
The Xáxli’p “Mission Statement” to guide their strategic planning is
characterized by several key goals: “to be proud, independent, self-sufficient
Xáxli'pemc and continue to pursue a land settlement with the Canadian
Government; to work with the best interest of Xáxl'ipmec with support, trust and
respect of one another; and for all to be open-minded” (Xáxli’p, 2017). In accordance
with the goals, the Xáxli’p have a long history of resistance against the loss of control
of their Survival Territory and strategically working to regain self-determination
starting with the Declaration of the Lillooet Tribe, logging blockades, and treaty

5

negotiations (Figure 2) through to the recent procurement of a Community Forest
Agreement in 2011.

Figure 2: Timeline of some key Xáxli’p events towards achieving self-determination (Diver, 2016).

The changes on the landscape observed by Xáxli’p community members have
been a significant consequence of the appropriation of Xáxli’p authority under
colonialism. Having a Community Forest Agreement in place affords the Xáxli’p
community greater autonomy towards ensuring that the integral social, cultural and
land relationships can continue for future generations. The Community Forest
Agreement with the provincial government covers 23,265 ha, of the 31,419 ha Xáxli’p
Survival Territory and excludes private land and reserves in the valley bottom, as
well as the eastern edge of the territory (Figure 1). It is managed under the Xáxli’p
Community Forest (XCF), and this arrangement is viewed by the Xáxli’p community
as an interim measure to protect the land until they regain title (Diver, 2016). From
the XCF mission statement, “considering the needs of present and future
generations, the XCF carries out ecologically and culturally sustainable land use for
the benefit of Xaxli’pmec and other beings in our Survival Territory” (Xáxli’p
6

Community Forest, 2013). The XCF team consists of a board of directors that guide
the mandate of the XCF, a forest crew that implements the eco-cultural restoration
work, a forester and office staff that manage the implementation (Figure 3) as well as
the Range Riders who travel around the territory maintaining a physical presence on
the land. The focus of the XCF eco-cultural restoration is to “restore degraded
ecosystems, to restore cultural resources and activities, and to create a sustainable
community economy based on high quality, value-added timber and non-timber
forest products” (Xáxli’p Community Forest, 2017).

Figure 3: XCF forest crew and staff spring/summer 2017 (Green, S. 2017).

7

Collaboration for Localized Solutions
In past land-management work, the Xáxli’p community has emphasized that
creativity and flexibility were necessary for negotiations between cultures; they are
deliberately trying not to emulate existing systems of land management but are
looking to create locally suitable plans that support their cultural values (Diver,
2016; Weinstein, 1995). This approach aligns well with the Indigenist research
methodology (described in the next section), placing importance on the research
team working closely with the XCF team to ensure culturally appropriate, placebased research and solutions.
To create localized, useful information for community-directed research, the
limitations of existing data and processes must be acknowledged, and place-based
solutions such as Traditional Ecological Knowledge (TEK) should be implemented.
TEK is qualitative, observational, long-term and highly localized knowledge that
should not be generalized to other environments or locations (Kimmerer, 2000).
Likewise, for this partnership with the XCF, it is a long-term project that will
generate site-specific information, not assume to be transferrable beyond its borders.
A major concern of the Xáxli’p community is keeping their TEK relevant in
the wake of changes wrought by resource extraction, wildfire suppression and
climate change. The potential erosion of their TEK will have consequences on intergenerational relationships when youth no longer see value in the obsolete
8

knowledge of their Elders that rely on intricate landscape relationships (Weinstein,
1995), creating a barrier to cultural continuance. A large wildfire that could
significantly change the ecosystem poses an existential threat to Xáxli’p traditions
and culture, and in response to this, the XCF team is interested in obtaining a clearer
understanding of the forest fuels and their spatial dispersal in order to manage them
to reduce this risk.
Research Objective.
This research was initiated by and designed closely with the XCF to follow
their direction and priorities that emerged over the research period and to provide
culturally appropriate and localized research. In conversations with Sybil Diver
(Diver, 2016), community member Roger Adolf highlights those priorities:
“We knew that we had to do something up here in Fountain Valley. We really
needed (tree) spacing. Because the other thing that we were afraid of was
forest fires. Because if we didn’t do anything about the land, the land might
be endangered. Spacing was really important – which we have done a lot of…
but at the same time, we always wanted to do our own logging, and protect
the water mainly.” (Diver, 2016, p. 42).
The goal of this research is to provide the XCF team with information and
resources to support their eco-cultural restoration objectives, primarily related to
wildfire risk that threatens their cultural continuance. One definition for wildfire
risk is the probability that a fire will change the net value of the forest resources
(Thompson & Calkin, 2011); for the Xáxli’p community, the ‘net value’ is not limited
9

to infrastructure or focused on loss of future income from the trees, but includes all
relationships with the land that supports their cultural continuance.
As the first stage of research laying the groundwork of what will hopefully be
a long-term partnership between UNBC and the Xáxli’p Community Forest, the
objective of the research is to address the following questions: 1) What is the wildfire
risk in the Xáxli’p Survival Territory? 2) What data and data quality gaps exist to
answer this question? And 3) What can we do with the existing data for XCF land
management until we have more locally suitable data?
The existing datasets and systems for fuel classification used by the British
Columbia government and many other practitioners are not necessarily the best
resources for local use because they are created using extrapolated assumptions and
measurements to service the whole range of ecosystems in Canada with a particular
focus on boreal forests (Forestry Canada, 1992; Perrakis & Eade, 2016). With these
systemic limitations, it is useful to bring in TEK to build suitable solutions for
Indigenous land management where “disagreements between TEK and resource
management policies do not prevent collaboration, but rather indicate places where
national narratives may not fit local environments, making traditional ecological
knowledge and regional science essential to sustainable management” (Ray, Kolden,
& Chapin III, 2012). Following this collaborative understanding, this project seeks to
improve local classifications building on the detailed data collected by the XCF
10

forest crew and Xáxli’p knowledge of their territory to start developing the datasets
required for wildfire modelling.
Applying Indigenist Philosophy
For some, Indigenist research is meant to be undertaken exclusively by
Indigenous people in their own communities (Coghlan & Brydon-Miller, 2014). For
this research, we take a philosophy more aligned with Rix, Wilson, Sheehan and
Tujague (2018) in which the view is that “Indigenist research respects and honors
Indigenous ways of knowing, being, and doing through using methods that are
informed by, resonate with, and are driven and supported by Indigenous peoples”
and allows space for non-Indigenous people to apply this way of thinking (Wilson,
2013). Indigenist philosophy is not about claiming ideas, but about the relationship
with the ideas, how they shape you and you shape them, and the complex
relationships between different ideas and people (Wilson, 2013). For non-Indigenous
researchers like myself, an important starting point is acknowledging that I have a
non-Xáxli’p worldview, and an exploration of the differences (Rix et al., 2018).
Therefore everything I observe and learn about the Xáxli’p Survival Territory and
community is interpreted through my worldview or “epistemological lens”
(Absolon & Willett, 2005), which does not capture the entire story.
Relationality captures the approach of being mindful of the
interconnectedness between everything; people, landscapes, and ideas are all in
11

relation to one another, and we are our relationships (Wilson, 2013). For Xáxli’p
people, their land ethic focuses on relationships and responsibilities, highlighting
the importance of thinking of future generations and the dynamics and learning
between generations (Weinstein, 1995).
“It all falls into the same dish, if you will. If you don’t have trees, you don’t
have life. And it just keeps going right down the chain. If you don’t have trees
you don’t have water, and water is our survival.” - David Adolf (Diver, 2016,
p. 159).
From an institutional perspective, the land is a pool of resources we manage,
rather than us being intimately part of the landscape with complex relationships
tying us to all parts (Leopold, 1949). By seeing it exclusively for its resources, the
exploitation of the land compartmentalizes and simplifies those relationships.
Xáxli’p member Arthur Adolf interviewed by Sybil Diver (2016) highlighted the
limitations that institutional views introduce:
“It’s especially hard for some community leaders to see the vision when
they’ve been trained or educated thorough a system like contemporary
forestry, where the only element that they envision for an economy is
exploiting the resource.” - Arthur Adolf (Diver, 2016, p. 160)
Indigenist researchers must help illuminate a path to success for Indigenous
communities (Wilson, 2013); therefore, it is paramount that this research supports a
self-determined future for the Xáxli’p community. Researchers are accountable for
creating and maintaining healthy relationships (Wilson, 2013), which in this context
means sustaining essential relationships within the landscape (between the fire
12

ecology, wildlife, hydrology, etc.) and with the people in and around it. Context is
vital to Indigenist research and applying theories from other ecosystems assumes
that the relationships are the same for both, which is very unlikely.
Community-Directed Research.
In community-directed projects, outsiders should support and not impose a
process; the aim is for communities to define the priorities and objectives supported
by scientific expertise, innovation, and research (Minkler, 2004). In the history of
many Indigenous communities throughout the world and Canada, there is a colonial
legacy of outside agents (NGO’s, governments, industry, funders, academics,
professionals) dictating research priorities and objectives through various
mechanisms, including funding, expertise and colonial power and control. This has
undermined the authority and wisdom of Indigenous cultures and often failed to
produce meaningful benefits to local communities and ecosystems compared to the
community time and money invested (Castellano, 2004; Government of Canada,
2016; L. Smith, 1999).
With the Indigenist and community-directed approach, the methodology and
outcomes are not fixed; instead, they are adaptive to the emerging processes and
relationships that arise during the research. The focus on community priorities that
guides community-directed research aligns with the trend of measuring research

13

success for their social, economic and environmental impact rather than through
institutional indicators such as peer-reviewed publications (Proctor, 2010).
The Position of the Researcher.
Claiming objectivity is counter to Indigenist methodology where the idea of
‘putting yourself forward’ allows the readers to understand your relationships and
bias, and holds the researcher accountable (Absolon & Willett, 2005). I grew up in
the Rocky Mountain Trench in southeastern BC, between the spine of the Rockies
and the wetlands of the Columbia River. My parents immigrated to Canada from
Switzerland a decade before I was born, which created a family space of exploring
new aspects of culture, without many entrenched rules about ‘this is what culture
is.’ They built a space where curiosity was encouraged, and there was a keen
knowledge that we are only ever understanding part of what people are
communicating with us. In contrast, my background as a Western-trained scientist
has been a barrier in my comprehension of Indigenist methodology and Indigenous
ways of knowing because of my skill honed to classify, measure, compare and
explain – stripping away the relationships and nuance. This thesis is part of my
exploration and learning process so that I can be another strand supporting
reconciliation. It is my sincere hope that I have respected the voice and ethic of the
Xáxli’pmec in this work.

14

Research for Reconciliation
Given the historic power imbalance between Indigenous peoples and
Canada’s institutions, reconciliation must be woven into the research design, paying
particular attention to decolonizing the research process (L. Smith, 1999). As
outlined by the guidelines of the Truth and Reconciliation Commission of Canada
(2015), research in support of reconciliation should be “supporting Aboriginal
peoples’ cultural revitalization and integrating Indigenous knowledge systems, oral
histories, laws, protocols, and connections to the land.” Reconciliation is consistent
with the goals and principles of Indigenist research, which provides a framework to
support the mandate for reconciliation. Non-Indigenous professionals in the natural
resource sector can support reconciliation by creating, supporting and enabling the
capacity within communities for self-determination through ensuring access to land
and the autonomy to manage it with a set of cultural values (Wilson, 2013; Wyatt,
2008). From the beginning, it has been the intent that this research collaborates
closely with the Xáxli’p Community Forest and follows their guidance to produce
outcomes they can apply.
Cultural Understanding of Wildfire and Ecosystem Health
Fire is an important landscape driver throughout the Survival Territory with
wide-spread evidence of fire-scarring on old trees (Figure 4) and remnant stands of
widely spaced old trees that have now filled in with thickets of dense young trees
15

since fires have been actively suppressed (Figure 5). Fire is not isolated from culture;
many Indigenous cultures across North America and the world actively used fire to
manage their landscapes, including the Xáxli’p (Cape York Elders, 2013; Carroll,
Cohn, Paveglio, Drader, & Jakes, 2010; Mistry, Bilbao, & Berardi, 2016; Ray et al.,
2012).

Figure 4: Fire scarring on old Douglas fir tree
(Green, S. 2017).

Figure 5: Dense regeneration associated with
wildfire suppression (Green, S. 2017).

Anthropogenic Fires
With the process of colonization, European perceptions of the wastefulness
and danger of fire increased wildfire suppression efforts and increasingly
discouraged and then criminalized cultural fire until many Indigenous groups had
lost their traditional knowledge of fire management (Kimmerer & Lake, 2001; Pyne,
2008). Without fires on the landscape, forests became increasingly dense, which
supported less frequent but larger and more intense fires (Williams, 2013), as has
been observed in the Xáxli’p Survival Territory. The landscape around Lillooet has a
16

history of anthropogenic fire like many other forests in BC (Boyd, 1999; Gray,
Andrew, Blackwell, Needoba, & Steele, 2002; Lewis, Christianson, & Spinks, 2018)
and has a history of low-intensity cultural burning used by the Xáxli’p to thin the
underbrush and maintain a patchwork landscape of high-quality habitat for wildlife
and community food and medicines (Xáxli’p Community Forest, 2017). Forest
densification and ground-level fuel loading associated with fire suppression have
been observed by the nearby Lytton First Nation (Lewis et al., 2018), and some
community members there continue to use fires to reduce ground fuels with
reduced extent and purpose compared to historic traditional use.
The difference between anthropogenic versus natural fires lies in the seasonal
timing, frequency, location and extent of the fire, resulting in different outcomes on
the ecosystems (Kimmerer & Lake, 2001). The seasonal timing varies by the location,
the type of ecosystem, and what was to be achieved by burning. Similarly, the
frequency of historic burn applications depended on what was the desired fireoutcome; grass forage could be burned every year, berry-picking areas every 3 to 5
years. The spatial extent of the fire was controlled by the timing of fire applications,
natural fuel breaks, and regular burning that reduced fuel (Kimmerer & Lake, 2001).
To help growing conditions and reduce the summer fire hazard in Lytton, fires were
lit in spring and fall at low elevations, which burned themselves out as they
travelled upslope into the forest (Lewis et al., 2018).
17

Fire Regime
Ecosystems that are adapted to a specific fire regime rely on certain intensities
or frequencies of wildfire to endure (Alexander, Cruz, Vaillant, & L. Peterson, 2013;
BC Wildfire Service, 2015). The fire regime is defined by the fire frequency, area
burned, season, intensity and severity (Stephens, North, & Collins, 2015) and is
determined by historical wildfire records, fire evidence in the forest, or stand
composition (Gray et al., 2002; Rogeau, Parisien, & Flannigan, 2016). Fire regimes
vary with forest type, geography, climate, and land management practices and is a
critical influence on forest ecology.
Understanding the local fire regime is important for land management in
order to avoid the oversimplification of applying national or regional narratives
about the role of fire on a different ecosystem (Ray et al., 2012). The landscape
around Lillooet has been shaped by fire, both natural and anthropogenic, for
hundreds of years (Gray et al., 2002; Province of British Columbia, 2011); it has one
of the shortest fire intervals in the province, with high variability of burn intensity
(Gray et al., 2002). Though not much of the Survival Territory has burned in the last
20 years (Figure 6), sites studied nearby had a historic mean fire interval ranging
from 6.6 to 11.6 years over the last 500 years (Gray et al., 2002) leaving most of the
Survival Territory overdue for fire.

18

Figure 6: History of fires recorded in the Xáxli’p Survival Territory by the Provincial Government
(Ministry of Forests, Lands, Natural Resource Operations and Rural Development - BC
Wildfire Service, 2019).

Land Management Effects on Wildfire Ecosystems
Fire suppression, along with high-grade logging, has been impacting Xáxli’p
ecosystems since the mid-1900s, increasing forest densities. This has resulted in a
lower abundance of wildlife, reduction in food and medicine plants, and caused
many areas to become more arid (Xáxli’p Community Forest, 2017) because forest
19

densification increases competition for water and has been shown to result in
stressed forests in water-limited landscapes (Cáceres et al., 2015).
Institutional Culture of Wildfire Suppression.
In southern BC, forestry practices in the early 1900’s focused on the extraction
of Ponderosa Pine (Pinus ponderosa) until it became scarce, then switching to
Douglas fir (Pseudotsuga menziesii). During this time, the practice was to primarily
extract the largest trees, creating a considerable change in the forest structure, with
smaller trees (Klenner, Walton, Arsenault, & Kremsater, 2008). There was also
increased wildfire suppression throughout southern British Columbia to protect
timber resources (Klenner et al., 2008).
Application of institutional forestry practices and a single wildfire narrative
to diverse landscapes with different fire regimes led to a century of dramatic
ecosystem changes across North America. Colonialists brought a culture of firesuppression with them and imposed it in their new environment, curtailing the
traditional use of fire by First Nations people and increasing fire-fighting efforts
(Carroll et al., 2010; Kimmerer & Lake, 2001). Many Indigenous cultures around the
world are still limited in their fire use by land ownership, regulations, and the
ecological legacy of wildfire suppression (Carroll et al., 2010). The imposition of
these European land management practices over the past 150 years have changed
forests across North America, resulting in more homogenized forests that are denser
20

with smaller, less fire-resistant trees, and often less shrub understory (Gray et al.,
2002; Hessburg et al., 2016; Kimmerer & Lake, 2001)
Fire Regime Changes on the Landscape.
Ecosystems with historic frequent, low-intensity wildfires have shown a shift
to less frequent, higher intensity wildfires since the 1800s (Gray et al., 2002;
Williams, 2013). In the Pavilion Ranges eco-section, the area surrounding the Xáxli’p
Survival Territory including Lillooet and Lytton, 94 wildfires have overlapped the
eco-section since 2000 (Figure 7), burning a total of 75,052 ha (Data BC & Ministry of
Forests, Lands, Natural Resource Operations and Rural Development - BC Wildfire
Service, 2019). After a century of fuel accumulation associated with wildfire
suppression and climate change, wildfires in the eco-section over the last 20 years
are larger than before. Low to moderate severity fire regimes like this may have
missed 4 to 10 wildfire cycles in the last 150 years due to wildfire suppression in
North America compared to high-severity regimes that may have only missed 1 to 2
wildfire cycles during the same period (Stephens et al., 2015). The earlier recorded
fires in the Pavilion Ranges eco-section (1919-1999) had a modest average size of 198
ha, but the average wildfire size since 2000 has increased to 2,497 ha. The largest of
these fires burned 192,017 ha in the 2017 Clinton/Cache Creek fire, over six times the
area of the Survival Territory, highlighting the risk of a large wildfire burning
through their entire territory.
21

Figure 7: Wildfire history in Pavilion Ranges eco-section (Ministry of Environment and Climate
Change Strategy - Knowledge Management, 2019; Ministry of Forests, Lands, Natural
Resource Operations and Rural Development - BC Wildfire Service, 2019).

Fuel Accumulation in Wildfire Suppressed Landscapes.
Understory fires, where the fire runs low along the ground, correspond to
low wildfire severity and tend to have a short return period. These forest areas that
burn frequently have more open tree spacing and less dense fuels (Figure 8), with
lighter (but highly flammable) grasses and herbs resulting in lower-intensity burns.
Dense coniferous forests with connected tree canopies support severe fires and burn
22

infrequently. Mixed fire regimes like the area around Lillooet fall in between, with
moderate wildfire severity over a patchwork of low and high severity areas where
tree mortality is between 20 to 70 percent (Hessburg et al., 2016).

Figure 8: Open ponderosa pine stands at the south end of the Survival Territory illustrates an
ecosystem characteristic of a low-intensity fire regime that is non-fatal to mature Douglas fir and
ponderosa pine (Green, S. 2017).

With longer wildfire intervals (due to the natural fire regime, or because of
wildfire suppression), fuel accumulation is greater, and there is denser tree spacing,
thicker branches and trunks, and more dead woody debris in the forest litter (Figure
9). This increase in fuel leads to stand-replacing fire regimes where fires burn with
higher severity and burn into the canopy with enough intensity to kill off mature
trees (Gray et al., 2002). When the landscape connectivity of dangerous fuels
increases, the area that burns in a single event may also increase, reducing the
consequent forest patchiness (Gray et al., 2002). The increased wildfire intensity and
23

area burned associated with increased fuel loading can have adverse effects on the
soil, water regime, wildlife habitat, and carbon storage (Stephens et al., 2015; United
Nations University, 2009). Permanent ecosystem changes have been attributed to
high-intensity fires due to increased soil erosion caused by the missing organic
materials resulting in trees being unable to re-establish from the lack of soil
(Stephens et al., 2015).

Figure 9: Examples of forests in the Survival Territory with an accumulation of fuels from closely
spaced trees and dead woody debris, increasing wildfire risk (Green 2017).

Facilitating Wildfire Risk Mitigation
More recent recognition of the adverse effects of wildfire suppression on
ecosystems has led policymakers to encourage land managers to focus on returning
forests to pre-settlement characteristics, which is supported by the re-integration of
wildfire in land management (Kimmerer & Lake, 2001). Differences in landscapes
24

and land-use priorities mean effective mitigation plans are unique to a location in
order to balance the different physical and social factors at play. For the Xáxli’p
community, managing the eco-cultural landscape and community safety is a key
priority over managing for forest productivity. Their eco-cultural restoration plan
prioritizes a landscape network that is healthy, resilient and supports cultural
continuance (Xáxli’p Community Forest, 2017) while using standard western
approaches with thinning and prescribed burns and increasingly more Indigenous
management methods.
Managing the wildfire risk in wildfire suppressed landscapes is necessary to
maintain ecological and cultural values due to potential consequences for ecology,
health, safety and economics (Thompson & Calkin, 2011). Strategic planning
requires an analysis of the risk of wildfire and its effect on the landscape in order to
mitigate these effects (Thompson & Calkin, 2011). Additionally, conflicting land-use
values require land managers to set landscape priorities regarding how they will
manage for wildfire risk. For example, the use of wide swaths of logged and/or
burned areas as fuel breaks around communities is an efficient, cost-effective
method of reducing fuels. However, the public may be opposed to the resulting
visual, ecological or hydrological impacts. Identifying threats on the landscape such
as fuels, ignition points, weather and uncertainty around them is a critical first step
for land managers trying to adjust management practices to mitigate the risk of
25

potentially devastating wildfires (Thompson & Calkin, 2011). Addressing wildfire
risk requires balancing community safety, ecosystem, cultural, timber, and health
values; each is met with different mitigation measures (Ohlson, Berry, Gray,
Blackwell, & Hawkes, 2006).
Landscapes adapted to frequent ecosystem fires with high variability and
patchiness dominated by fire-maintained species are considered fire-absorbent
landscapes (Leverkus et al., 2017) (Figure 10). These patches provide fuel breaks on
the broader landscape because they burn more frequently and, therefore, have
reduced fuel loads and can absorb wildfire without becoming stand-replacing fires
(Leverkus et al., 2017). The presence of fire absorbent areas within a greater
landscape matrix is essential for resilience in withstanding wildfire disturbances and
avoiding large-scale, high-intensity fires (Leverkus et al., 2017).

26

Figure 10: Open patches visible in the foreground as well as on the opposite side of the valley build a
fire absorbent landscape in parts of the Xáxli’p Survival Territory (XCF Forest Crew, 2017).

Variations in topography, weather, forest structure, and fuels can
dramatically affect the behaviour of a wildfire moving across a landscape, resulting
in variable tree and shrub mortality (Gray et al., 2002). With time, this variable
mortality creates a heterogenous ecosystem with pockets of older fire-resistant
forests, open shrubs, and young regeneration (Stephens et al., 2015). Steep, complex
terrain, like what is in the Survival Territory, creates highly localized fire regimes
(Klenner et al., 2008; Parisien, Walker, Little, Simpson, & Wang, 2012). Research on
these localized fire regimes suggests that historically the low-severity fires did not
27

likely spread far, resulting in occasional higher severity fires despite forest and
climate conditions consistent with frequent low-severity burns (Klenner et al., 2008).
With the high risk that large area crown-fires pose to people and resources,
forest managers seek to reduce the area of forest that supports high-severity fire,
breaking it up and increasing heterogeneity and pyrodiversity by creating a
landscape mosaic with both older forest stands (Hessburg et al., 2016) and younger
and/or less dense forests to provide fuel breaks. Landscapes with high connectivity
between dangerous fuels have greater spread and severity of fires. Restoring fireadapted landscapes focuses on increasing heterogeneity with more forest openings
and lower tree densities (Stephens et al., 2015), and focusing on creating more fireabsorbent and climate change resilient forests. To support this, land managers can
prioritize restoration on ecosystems that historically had frequent wildfire activity
such as valley bottoms and south-facing slopes, and use cooler, high elevations to
identify fire refugia where old-growth forest patches can be maintained and created
(Rogeau & Armstrong, 2017).
Xáxli’p Eco-cultural Restoration
What was likely a fire absorbent landscape before institutionalization, the
Survival Territory has become quite homogenous over the past few decades. The
suspension of Xáxli’p fire management has put the ecosystem in a state that will now
take considerable effort, cost, and time to restore historic, pre-suppression fuel
28

conditions. In response to such ecosystem changes, the XCF team developed an
Ecosystem Based Conservation Plan (EBCP) to protect and restore areas necessary
for healthy water, forests, wildlife and Xáxli’p cultural continuity (Xáxli’p
Community Forest, 2019). This landscape-level plan defines a network of connected
areas of protection that supports forest functions and varied land-uses while
sustaining biodiversity (Hammond, 2009) and provides a project framework that
integrates well with creating a fire absorbent landscape in the Xáxli’p Survival
Territory. The strategic execution of the EBCP plan is vital to maximizing the efficacy
of the treatments with limited community resources. A primary mechanism for
rebuilding a fire-absorbent landscape has been restoration treatments undertaken by
the forest crew (Figures 11 and 12) to thin the forest and reduce ladder fuels.

29

Figure 11: Example of an XCF restoration
area with thinned trees and low branches
removed (Green, S. 2017)

Figure 12: Edge of restoration area – thick
regrowth associated with fire suppression (back
trees) compared to restoration ecosystem
(foreground) (Green, S. 2017)

Further steps will likely include larger-scale mitigation measures (such as
large burns) to return the landscape to a state where incremental, multi-scale
measures can be used, such as small patches of annual low-intensity burns to
maintain a more traditional fire absorbent landscape. Other mitigation measures
suggested by outside government reports for Fountain Valley include controlled
burns of forest understory, surface fuel reduction in areas of wider tree spacing,
strategic fuel treatment placement to prevent north-south fire spread, buffers around
the local campground in the middle of Fountain Valley to reduce the risk of runaway campfires, fuel reduction around private properties to protect dwellings, and
annual burning, irrigating or grazing of fields (Province of British Columbia, 2011).

30

The Components of Wildfire Risk and their Mitigation Strategies
Understanding fire behaviour is necessary to implement effective riskmitigation measures. “Assessing wildfire risk requires an understanding of the
likelihood of wildfire interacting with valued resources, and the magnitude of
potentially beneficial and negative effects to resources from fire” (Thompson &
Calkin, 2011). Fire behaviour is defined as “the manner in which fuel ignites, flame
develops, and fire spreads and exhibits other related phenomena as determined by
the interaction of fuels, weather and topography” (Canadian Interagency Forest Fire
Center, 2002). The three commonly recognized types of wildfires are ground fires
burning the organic layer on or below the ground, surface fires burning fuels above
the soil layer, and crown fires that burn in the forest canopy (Scott & Reinhardt,
2001). Fires may burn as ground or surface fires and may fizzle out on their own
because of low temperatures or high moisture/relative humidity, or they may take
off and spread to the forest canopy, starting a crown fire (Alexander et al., 2013).
Crown fires can exhibit extreme fire behaviour with the potential to burn at high
temperatures and move very quickly (Alexander et al., 2013) and as a result can pose
significant risks impacting ecosystems that are not adapted to high-intensity fires
and the nearby human activities (Stephens et al., 2015).
Understanding the dynamics between fuel, wind and topography informs
planning and executing fuel mitigation (BC Wildfire Service, 2015) to get the best
31

results for resource investment. As mentioned previously, efforts to reduce the
threat in high-risk areas may include thinning trees, removing ladder fuels, and
incorporating prescribed burns (BC Wildfire Service, 2015), which may prioritize
Wildland Urban Interface’s (WUI) (Province of British Columbia, 2012).
Fuel Risk Mitigation.
The trio of fuel, weather and topography determine actual fire conditions, but
land managers can only directly change the fuel component, while the risks
introduced by weather and topography can only be managed indirectly. As a result
of decades of wildfire suppression in the Xáxli’p Survival Territory, the forest fuel
loads have increased, and Xáxli’p traditional knowledge used to manage the
landscape with fire has been mostly lost.
Fuels for wildfires include dead and living organic material, including
standing, fallen and decomposing in the duff layer. The structure, chemistry,
volume, state of decomposition of forest fuels is highly variable throughout a
landscape (Gray et al., 2002), but can be roughly classified into categories that
translate to their effect on fire behaviour. Coniferous forests with more small and
medium-sized trees have been found to have greater wildfire severity due to the
increased vertical continuity in the forest structure (ladder fuels), leading to crown
fires (Alvarez, Gracia, Castellnou, & Retana, 2013). The Fire Behaviour Prediction
(FBP) system is a classification system that is widely used in Canada to classify fuels
32

(Forestry Canada, 1992), though other systems exist and take a different approach to
quantifying fuels using measured values. The categories of the FBP system can be
used to identify areas with high-risk fuels to inform mitigation measures.
Research in south-central BC found that fuels had the greatest effect on
wildfire behaviour, followed by weather and ignition sources (Wang et al., 2016),
though fuels have been found to be less significant in other infrequent high and
mixed-severity fire regimes (Alvarez et al., 2013). Methods for reducing the fuel risk
include prescribed burning, forest thinning, and irrigation and grazing of fields (BC
Forest Service, 2018; Province of British Columbia, 2011). The spatial pattern of fuel
reduction treatments on the landscape is also key, creating fuel breaks, targeting
high-risk areas, and reducing the fuel continuity by creating a patchwork of low and
high-risk fuel types (Hessburg et al., 2016; Province of British Columbia, 2011). Fuel
reduction and other mitigation measures are not a one-time fix; they will require
ongoing work to ensure efficacy, and in many cases, require subsidies because of
high resource demands and limited value of the extracted materials (Hessburg et al.,
2016).
Prescribed Burning.
Traditional burning practices were (and to some extent still are) used by
many Indigenous communities throughout the world to modify the environment to
improve survival; it fulfilled many needs, from cleaning up village sites, creating
33

habitat for flora and fauna that could be harvested, to destroying pests (Kimmerer &
Lake, 2001). In wildfire adapted ecosystems, many species thrive with frequent low
burning fires and become scarce with few fires, so the reintroduction of wildfire can
restore ecosystem health as well as act as a risk management strategy. Recognizing
the long term use of wildfire by Indigenous groups in many ecosystems, more
institutional forest managers have come to appreciate the value of wildfire on the
landscape, both economically and for the ecosystem health (Taylor & Alexander,
2006; Williams, 2013) which has encouraged the use of fire in landscape
management.
Contemporary use of prescribed burning includes hazard reduction,
silviculture, wildlife habitat enhancement, range burning, insect and disease control,
conservation of natural ecosystems (Weber & Taylor, 1992). The management of
mixed-severity forest regimes is more complex than low and high severity regimes
because a patchwork and balancing of ecosystems is necessary. The use of fire for
restoration requires frequently burning open forests but ensuring it is controlled and
localized in order to retain adjacent areas of older, denser growth (Hessburg et al.,
2016). To reduce the severity and maximize ecological benefits, applications of
managed fire are restricted to moderate fire weather rather than extreme fire
weather (Hessburg et al., 2016).

34

Thinning and Biomass Removal.
The forest crew’s restoration treatments have focused primarily on reducing
ladder fuels and thinning the forest density in plots around the territory (Figure 13).
The location of this work was guided by the EBCP and dialogue with Elders to
prioritize areas of higher value to the community for berry picking, wildlife,
sensitive ecosystems and sacred areas (Diver, 2016). Thinned trees and branches are
left on the ground to support water retention in these forests stands and support
healthy soils; suitable sized logs are removed for firewood for the community.
Thinning is a more flexible option for land managers because it can be undertaken at
any time while prescribed burns have a limited season of use. Even though thinning
reduces ladder fuels and stand density, it may not reduce crown fires because there
is often an increase in surface fuels if they are not removed from the stand entirely,
or vegetation grows up (Raymond & Peterson, 2005). The removal of trees to reduce
the stand density and the removal of biomass such as dead woody debris and ladder
fuels such as the lower branches of trees is recommended by the BC Forest Service
(BC Forest Service, 2018).

35

Figure 13: Examples of forest thinning undertaken for eco-cultural restoration in the Survival
Territory with debris left on the ground for ecological benefits (Green, S. 2017).

Comparison.
Highly altered landscapes may have too much fuel loading to safely and
effectively consider using prescribed burns and may require initial fuel reduction
before returning fire to the landscape (Hessburg et al., 2016). A combination of
treatments or a phased approach over many years can be used to achieve the lowest
risk and most cost-effective treatments to restore landscapes to a more fire-absorbent
state. Fuel reduction through forest thinning is costly and slow while controlled
burns can be applied to larger areas, but have associated issues with public
perception, air quality and liability (Stephens et al., 2015).
Mechanical thinning alone does not sufficiently mimic wildfire as a
disturbance, but thinning followed by prescribed burning is an effective fuel and
wildfire risk management strategy that reduces more of the overall fuel load and
provides ecological conditions for fire-adapted species (Hessburg et al., 2016;
36

Prichard, L. Peterson, & Jacobson, 2010). Using thinning to reduce young regrowth
and ladder fuels associated with forest densification allows wildfires to burn
through older-aged stands without high mortality (Hessburg et al., 2016). In a
review of studies looking at the effect of fuel treatments on reducing wildfire
severity Kalies and Yocom Kent (2016) reported that “all studies found a positive
effect of at least one treatment. Several studies found that thin + burn treatments had
the greatest positive effects while burning or thinning alone had either less of an
effect or none at all” (Kalies & Yocom Kent, 2016).
Fire Weather Risk Factors.
Under moderate weather conditions, reducing fuel is an effective strategy to
manage the wildfire risk, but in regions with high fire weather severity and during
extreme fire weather conditions, the efficacy is still scrutinized (Moritz, Moody,
Krawchuk, Hughes, & Hall, 2010). Weather conditions that contribute to higher
wildfire risk and can cause fires to spread quickly are hot temperatures, high winds,
and low moisture (BC Wildfire Service, 2015). The Fire Weather Index (FWI) System
is used in Canada using temperature, relative humidity, wind speed, wind direction
and precipitation (Turner & Lawson, 1978). Fire weather varies throughout the year,
creating the conditions for wildfire seasons that can start in April and extend into
September in the Lillooet area.

37

Wind.
Fuels and terrain show a greater effect on wildfire risk in low wind areas, but
severe wind conditions cause fires to become extremely large and unstoppable
(Moritz et al., 2010). Wind has been shown to have a significant impact in the
initiation of crown fires in areas with high wind (Alvarez et al., 2013) with an
increase in the relative importance of wind as a risk factor in these areas; mountain
passes and other terrain features that funnel wind are correlated to high wildfire
danger and the quick spread of fires in California (Moritz et al., 2010). In the
Survival Territory, the narrow valley with steep slopes funnels the north-south wind
of the Fraser Canyon, allowing for the possibility that a wildfire could sweep the
whole valley.
Climate Change.
The concern about higher fuel loads and unhealthy forests resulting from
fires suppression may be compounded by the effects of climate change with longer
wildfire seasons and more extreme fire weather (BC Wildfire Service, 2015). Climate
change is expected to increase fire-weather and ignitions in many areas in BC,
increasing the number of fires (Nitschke & Innes, 2008; Wang et al., 2016). The fire
weather season is also expected to lengthen by nearly 30% by 2070, with most of that
gain in the spring (Nitschke & Innes, 2008).

38

The ecosystem composition changes predicted because of climate change are
expected to impact wildfire risk, though studies have produced conflicting
projections for the province. Some studies suggest that areas currently displaying a
mixed-surface fire to intermittent crown fire regime are expected to become mixed
intermittent crown fires to full crown fire regime (Nitschke & Innes, 2008). Other
studies suggest fuel categories will shift to lower-risk categories as areas expected to
get less precipitation have reduced forest productivity and areas of higher
precipitation have more fuel (Meyn et al., 2013; Wang et al., 2016). These
contradictions are likely due to the complex topography in BC, causing temperatureprecipitation-vegetation relationships to vary with different ecosystems (Meyn et al.,
2013) and confounding generalizations. The shift in ecosystem composition expected
as a result of climate change will likely be slow because of long-lived tree species,
but recruitment after disturbances will provide opportunities for more rapid
ecosystem shifts capitalizing on the change in climate (Wang et al., 2016).
Geography and Wildfire Risk.
The physical and human geography of the Xáxli’p Survival Territory plays a
pivotal role in the local wildfire risk and shape the options for mitigation. The steep
mountainous side hills (Figure 14) and flat valley bottom affect how fires might
move. The concentration of habitation and land-use along the valley bottom

39

concentrates ignition risk and human safety concern while the limited road access
restricts where mitigation can be undertaken on a restricted budget.

Figure 14: The complex terrain of the Xáxli’p Survival Territory that supports varied ecosystems and
fire behaviour (XCF Forest Crew, 2017).

Variable Terrain Creates Highly Variable Burn Conditions.
Topography is linked to both weather and fuel, by providing the surface for
both factors to develop, leading to covariant effects on fire-behaviour. Topography
in mountainous terrain including slope steepness, slope aspect, elevation and
landforms can slow or speed up fires, or funnel and divert the path, creating pockets
of forest with longer wildfire return intervals and creating heterogeneous burn
patterns (Rogeau & Armstrong, 2017).
In the Lillooet area, dry, upland sites have more frequent, lower severity
wildfire regimes, whereas northerly aspects and drainage bottoms have older forests
with high severity fires (Gray et al., 2002). This has also been found in other mixed40

severity fire regimes similar to the Survival Territory, where wildfire frequency is
lower and wildfire intensities higher on north-facing aspects (Gray et al., 2002;
Lecina-Diaz, Alvarez, & Retana, 2014). Higher elevations display more frequent fires
resulting in higher stand densities of older trees lower on the slopes and stand
density decreasing as elevation and snowpack increases (Gray et al., 2002). Steeper
slopes show signs of higher flames, indicating a greater wildfire intensity as slopes
steepen, and increasing tree mortality (Lecina-Diaz et al., 2014). This causes steeper
slopes to have lower tree densities and have species compositions that favour fireresistant species (Gray et al., 2002). The downward slope displays an effect in the
cessation of crown fires, but the effects of slope are greatest at low wind speeds
where the dominant effects of wind were not affecting wildfire spread (Alvarez et
al., 2013).
Human Driven Ignition Patterns.
In the Xáxli’p Survival Territory, the majority of modern recorded fires are
ignited by people (Ministry of Forests, Lands, Natural Resource Operations and
Rural Development - BC Wildfire Service, 2018). The areas around recreation sites,
as well as near private dwellings and along the network of back roads, have been
identified as higher risk features for human ignitions (Province of British Columbia,
2011). Research suggests human ignitions may be reduced by higher policing, and
limiting access to high-risk areas during severe fire weather (BC Forest Service, 2018;
41

Collins, Penman, & Price, 2016; Province of British Columbia, 2011). This can be
supported by education for backcountry users and people living in the wildlandurban interface as well as focusing mitigation on areas where human ignition is
more likely to occur.
Limitations Created by The Road Network.
All the inhabitants in the Survival Territory live near the valley bottom,
coinciding with the areas of high ignition. With only one narrow road providing
access through Fountain Valley, emergency egress in case of a wildfire is a concern
(Province of British Columbia, 2011). In addition to the higher ignition risk near
roads, fuel mitigation efforts are also restricted to the forest area accessible by the
existing road network and must be taken into account in planning (Hessburg et al.,
2016).
Conclusion
In the Xáxli’p Survival Territory, the increased fuels from decades of wildfire
suppression, the steep mountainous terrain with complex wind patterns from the
nearby Fraser River valley and the increasingly extreme fire weather create a
wildfire risk that has become a growing concern for community members. The ecocultural restoration program for the Survival Territory provides a mechanism to
work toward restoring ecosystem resilience while reducing the wildfire risk. The
restoration of the ecosystem is critical to Xáxli’p cultural relevance for future
42

generations, and the return of anthropogenic fire to the landscape has significant
implications for their cultural and ecological health. The cultural implications of this
Indigenist research approach necessitates the incorporation of traditional knowledge
and creating locally adapted solutions.

43

Chapter 2: Exploration of Existing Datasets for Wildfire Risk Assessment Guided
by Community
Connecting Research Purpose to Data
Following the Indigenist and community-based model, this research is
undertaken as a fluid and dynamic process to adapt to the operational needs of the
Xáxli’p Community Forest (XCF) as their objectives develop and information gaps or
needs are identified. The foundation for this thesis rests on the relationships created
between the XCF and the University of Northern British Columbia (UNBC) research
team. Ongoing meetings and conversations with the XCF board and employees
throughout the first two years identified their top concerns and the research
direction and were used to share information and substantiate and verify the data
products produced. At each stage, the data and conversations informed the next
steps, all while trying to uphold the ethic of context and place to honour Xáxli’p
objectives.
Developing the Research Question with the XCF
The partnership started in November 2016 through conversations with Herb
Hammond, a forester and consultant for the XCF for many years, who suggested
that the XCF might have projects that could use external resources. An introductory
letter sent to the XCF board proposed a partnership where UNBC research could
support the work of the XCF employees with technical expertise and resources. In
early 2017, after an introductory meeting with the XCF board and employees to
44

discuss the knowledge of the research team and critical issues for the XCF, the board
approved the research partnership and prioritized wildfire risk as the key issue to
focus initial research on. Throughout the research process, the board acted as an
advisory committee along with the XCF employees to guide the research.
Discussions with Herb identified the existing Ecosystem Based Conservation
Plan (EBCP) ecosystem classifications developed for the XCF as a potential data
source that could be adapted for location-specific fuel-classification. In April, three
days were spent in the Xáxli’p Survival Territory, in the field and the office, building
relationships with the XCF board and employees and visiting a range of ecosystems
and restoration sites the XCF forest crew has treated with various methods to reduce
the forest fuels. The forest crew explained their restoration work along with the
associated forest surveys and EBCP ecosystem data. The XCF board and employees
expressed that the creation of a wildfire risk assessment showing areas of high-risk
would be a useful focus of the research to prioritize fuel treatments and offered the
use of data and fieldwork undertaken up to that point.
Xáxli’p Ecosystem Based Conservation Plan Dataset.
As part of the Xáxli’p Ecosystem Based Conservation Plan, a spatial layer
categorizing the ecosystems in the valley was developed (Silva Forest Foundation,
2011). It was generated from aerial and satellite interpretation, site visits and local
knowledge and is used by the XCF team in their current eco-cultural restoration
45

work. The dataset covers the full Xáxli’p Survival Territory even though the XCF
agreement does not cover the full extent. The footprint of this dataset was used for
all further analysis to support the full Survival Territory and Xáxli’p selfdetermination goals.
This dataset identifies landcover type (ex. alpine, cliff, forest, urban, shrub,
talus slope) and additional details for vegetated areas. All vegetated areas are
classified by dominant vegetation type (forbs, shrubs, trees), ground cover of
vegetation (very sparse, sparse, open, dense), slope (flat, moderate, steep, very
steep), moisture regime (very dry, dry, sub-mesic, mesic, wet) (Table 1). There are
additional modifiers that may also be applied (ex. complex terrain, exposed rock,
wetland, snow-dominated ecosystem).
Table 1: Ecosystem descriptors used in the Ecosystem Based Conservation Plan dataset
Attribute

Description

Data Code

Forbs

F

Shrubs

S

Trees

T

Groundcover of

Very Sparse: 1-10%

VSP

Vegetation

Sparse: 11-25%

SP

(crown cover in

Open: 26-60%

OP

the forest)

Dense: >60%

DE

Very Dry

VD

Dry

D

Sub-Mesic (slightly dry)

SM

Mesic (average to moist)

M

Wet

W

Flat: 0-20%

F

Moderate: 21-40%

M

Steep: 41-60%

S

Dominant
Vegetation Type

Moisture
Regime

Slope Gradient

46

Very Steep: >60%

VS

Complex terrain

CT

Shallow soil over rock

SR

Exposed rock

ER

Talus/scree

TA

Wetland

WL

Large riparian ecosystem

RZ

Snow dominated

SD

Alkali flats beside water

ALKALI

Alpine shrubs or meadow ecosystems

ALPINE

Industrial clear-cuts

CCUT

Avalanche chute

CHUTE

Cliff

CLIFF

Dry, open ponderosa pine forest

DRY_PY

Dry ravel - steep slope

D_RAV

Very Dry site with sparse vegetation

D_SPAR

Landscape

Earthflow

EARTHFLOW

Descriptors

Exposed rock

ER

Agricultural field

FIELD

Forest

FOR

Meadow

MDW

Sagebrush

SAGE

Other Shrubs

SHRUB

Talus slope/rock scree

TALUS

Settled area

URB

Utility or road clearing

UTIL

Modifiers

The different attributes separately observe a single variant or are combined to
provide a generalized sense of the ecosystem conditions at the site. Throughout the
Survival Territory, the dominant vegetation is trees with a mixed range of density
(Figure 15). At the same time, site moisture and slope show similar patterns that vary
with the terrain: where the slope is steeper, the moisture tends to be lower. When the
attributes are combined, each ecosystem type gets a code that describes the site

47

follow the order of vegetation, density, moisture and slope. For example, a site
coded F OP D M is forested with open stand structure, dry, and a moderate slope.

Figure 15: The variation of Ecosystem Based Conservation Plan ecosystem descriptors in the Survival
Territory.

Leveraging the accuracy and local suitability of the EBCP ecosystem data is
vital for the local scale of the project and the work of the XCF team. However, for it
48

to have maximum usefulness and opportunity for assessing and mitigating wildfire
risk modelling, the way it is adapted to fuel risk should still interface or translate to
existing wildfire risk assessments and models to make the most of existing research
and application.
Fuel Classification in Canada
The standard method for provincial and federal wildfire risk assessment uses
the Canadian Forest Fire Danger Rating System (CFFDRS) and associated fuel
classifications to run models such as Burn-P3 (Parisien et al., 2005) to understand
wildfire behaviour on the landscape. The CFFDRS was developed in Canada to
support decision-making by fire management agencies to allocate resources based
on fire danger levels (Taylor & Alexander, 2006). The fire danger rating system
incorporates ignition, weather, topography and fuels and has two primary subsystems, the Canadian Fire Weather Index (FWI) and the Canadian Forest Fire
Behaviour Prediction (FBP) system (Taylor & Alexander, 2006).
The FBP system “provides quantitative estimates of head fire spread rate, fuel
consumption, fire intensity, and fire description”(Forestry Canada, 1992) based on
experimental, wild and prescribed burns across Canada, leaning heavily on boreal
forests. The model uses fuels, weather, topography, foliar moisture content and time
since ignition. It relies on fuel categories characterized on a mixture of dominant tree
species, tree spacing, and stand age, which relates to the fuel loading that these
49

conditions display (Perrakis & Eade, 2016). Canada’s wide range of forest types is a
challenge for the development and implementation of a national system, and the
domination of boreal forests in the CFFDRS and FBP reflects this challenge (Taylor &
Alexander, 2006). Because of the issues that arise when trying to categorize fuel
types outside of the specific ecosystems forming the base of these systems (primarily
boreal and sub-boreal forests) (BC Wildfire Service, 2015), some ecosystems in BC
require adaptations of the FBP categories using local expertise on fire behaviour
(Parisien et al., 2012)
While an in-person assessment by a wildfire professional provides the best
estimation of the fuel category beyond using the FBP categories, over large scales,
this is impractical. Additionally, most areas do not have localized datasets like the
EBCP in the Xáxli’p Survival Territory. To address this issue in British Columbia, the
Vegetation Resource Inventory (VRI) has been used to classify the entire province
into adapted FBP fuel types. VRI is a BC Government forest resource inventory
dataset with detailed site descriptions for forested areas throughout the region based
on photo interpretation and some ground sampling (Ministry of Forests Lands and
Natural Resource Operations, 2018). By using this system, it allows for the unique
ecosystems in the province that are not represented in the national system to get a
fire behaviour characterization (BC Wildfire Service, 2015). For example, an
ecosystem type not represented in the Canadian FBP system may use a category
50

from the other side of the country that has a similar fuel structure to achieve a
realistic fire behaviour when modelled (Parisien et al., 2012). Many wildfire models
in BC use the BC VRI dataset because it provides standardized and continuous
coverage for the province and can easily convert to fuel categories following
standardized methods (Parisien et al., 2005; Perrakis & Eade, 2016). However, the
appropriate application of this is partial to landscapes with uniform and continuous
fuels and simple, homogenous topography with constant and unidirectional wind
(Taylor, Pike, & Alexander, 1996), limiting its suitability in the highly variable terrain
of the Xáxli’p Survival Territory. In addition to this, the spatial and temporal
accuracy of the VRI dataset is variable across the province. Within the Survival
Territory, assessments are from between 1977 and 2012 and all projected values are
only valid up to 2015 (Ministry of Forests, Lands, Natural Resource Operations and
Rural Development - Forest Analysis and Inventory, 2017).
Description of the FBP Fuel Categories.
There are 16 benchmark fuel categories in the FBP system (Table 2),
qualitatively described by their stand structure and composition, surface and ladder
fuels, and forest floor cover and organic layer (Forestry Canada, 1992). “A fuel type
is a fuel complex of sufficient homogeneity and extending over an area of sufficient
size that equilibrium fire behaviour can be maintained over a considerable time
period” (Forestry Canada, 1992). This qualitative description requires practitioners
51

to make subjective decisions about forest stand characteristics such as “wellstocked” or “moderate density” (Perrakis & Eade, 2016).
Stand structure includes the crown closure, species, stand and crown height
of a given forest. Composition, distribution, and height describe the three classes of
fuels - surface, ladder, or litter fuels. Surface and ladder fuels compose the layer that
sits on top of the forest litter, which includes the herb, shrub, and young conifer
understory as well as flaky bark, dead woody debris, low tree crowns. On the forest
floor are the litter fuels consisting of the low ground cover (such as dead grass,
lichen, and moss) and the duff layer.
Table 2: Canadian Forest Fire Behaviour Prediction (FBP) system fuel categories (Taylor, Pike, &
Alexander, 2000).
Coniferous

Mixed wood

C-1

Spruce-lichen woodland

C-2

Boreal spruce

C-3

Mature jack or lodgepole pine

C-4

Immature jack or lodgepole
pine

C-5

Red and white pine

C-6

Conifer plantation

C-7

Ponderosa pine – Douglas fir

Deciduous
D-1

M-3
M-4

Boreal mixed-wood – leafless
Boreal mixed-wood – green
Dead balsam fir mixed-wood –
leafless
Dead balsam fir mixed-wood –
green

Slash
Leafless aspen

Open
O-1

M-1
M-2

Grass

S-1
S-2
S-3

Jack or lodgepole pine slash
White spruce-balsam slash
Coastal cedar-hemlock –
Douglas fir slash

52

When modelled, each fuel type has associated equations for fire spread
(Perrakis & Eade, 2016; Wang et al., 2016); “the C-2 (boreal spruce) and C-4
(immature lodgepole pine) fuel types are highly conducive to spread, whereas the C5 (red and white pine) and C-7 (ponderosa pine) fuel types have slower spread
potential. The deciduous (D-1) and mixed-wood (M-1/2) fuel types have faster fire
spread rates in the spring, before leaf flush than later in the season. Fire spread
potential in the grass (O-1) fuel type also varies seasonally with the curing
percentage ”(Wang et al., 2016).
Converting Ecosystem Data to Fuel Categories
As discussed in Chapter 1, there is a significant relationship between forest
composition and wildfire severity. Ensuring functional fuel categories that are
usable for future research into wildfire modelling requires having forest data with
appropriate metrics and spatial resolution. As a preliminary stage of an ongoing
relationship between the XCF and researchers, a necessary foundation was exploring
the usability of existing datasets and determining whether the accuracy and
precision are appropriate for the scale of the Survival Territory or whether at this
scale and application better forest composition information is required.
Combining EBCP and VRI Datasets
The EBCP and VRI datasets were used together as an initial attempt to
classify forest fuels and get the strengths of each by combining the higher confidence
53

and accuracy of the EBCP forest density data with the critical fuel structure variable
of stand age from the VRI dataset. The spatial features of the EBCP formed the
framework and the attributes of the VRI dataset were merged to this dataset using
the VRI record that covers the largest area of each given EBCP polygon when there
were significant boundary discrepancies. This dataset was classified into ecosystem
categories using density from the EBCP and forest age from VRI (Table 3).
Ecosystems dominated by forbs and shrubs were kept as individual fuel types, while
the forest densities were used for further categorization (dense, sparse & open, and
very sparse). The dense stands were then further classified by stand age from the
VRI attributes into young forests (less than 60 years old), middle-aged forests (6099), and old forests (100+).
Table 3: VRI and EBCP - V1 and V2 classification categories.
Stand Density
(Restoration Ecotype
Dataset)

Stand Age
(VRI Dataset)

FBPS/Severity
Rating (Taylor et

V1

V2

al., 2000)

Dense

60-99 years

C2 / Very High

x

x

Dense

>100 years

C3 / High

x

x

Dense

<60 years

C4 / Very High

x

x

Open

All ages

M1 (75% conifer) /

Combined

Moderate

into one

Sparse

All ages

M1 (25% conifer) /

Open/Sparse

Moderate

category

Very Sparse

All ages

C7 / Low

x

x

n/a

O1 / Very Low

x

x

x
x

Grasslands,
non-forested
vegetation

54

The EBCP/VRI ecosystem categories were classified into fuel risk categories
(Table 3) according to the fuel type descriptions from the Field Guide to the
Canadian FBP System (Taylor et al., 1996) and BC specific work by Perrakis and
Eade (2016). The result was a heterogeneous landscape dominated by the mixed
coniferous, deciduous M1/2 fuel category, with large patches of dense mature C3
fuels (Figure 16, A). The fire behaviour of the M1/2 category has significant variation
depending on the ratio of coniferous to deciduous trees – higher coniferous
composition can support higher risk wildfires. Such a large area categorized as
sparse (M1/2 – 25% coniferous, 75% deciduous) and open (M1/2 – 75% coniferous,
25% deciduous) prompted another iteration to separate the sparse and open
categories into different fuel categories (Table 3) to represent the different levels of
risk better and provide more nuance (Figure 16, B).

55

A)

B)

Figure 16: VRI and EBCP combined data, A) first iteration Open and Sparse combined, B) Open and
Sparse categories separated.

From visiting the Xáxli’p Survival Territory, exploring the datasets and
conversations with the XCF board and employees as well as other wildfire experts, it
was determined that the VRI data is too outdated to add accurate information about
fuel loading. Substantial regeneration (and therefore densification) of the forest in
the last 40 years has resulted in many stands being 2-aged. There is a widely spaced
older forest demographic (likely due to frequent historic fires) and a very dense
young forest growing up around this (without fires, young seedlings were not
thinned out). The VRI stand ages do not reflect this pattern because of the

56

assessment year and dataset attributes. Relying on an age classification generated
decades earlier may not provide enough sensitivity for fuel classifications,
particularly for the oldest age category (covering over 80% of the landscape, Figure
17) because the 2-aged nature of the stands changes the fuel structure significantly
for both canopy density and ladder fuels.

Figure 17: Projected forest age from the BC VRI dataset (Ministry of Forests, Lands, Natural
Resource Operations and Rural Development - Forest Analysis and Inventory, 2017).

When presented with this update, the XCF board and employees shared site
survey data that had been undertaken by the forest crew which included photopoints with GPS locations in various locations in the territory (Figure 18). An
57

expansion of this program would allow the use of these photo-points to validate the
ecosystem dataset and get a view on the fuel structure for the ecosystem categories.
With the crew travelling around the territory, taking sample pictures in each
ecosystem type could increase the reliability of the fuel categorization and provide a
photo record of the location. Using these pictures enabled the consultation of outside
fuel categorization experts to improve classification skills and verify existing
classifications.

Figure 18: Example of a west-facing picture, 1 of 4 taken at a photo-point (XCF Forest Crew, 2017).

58

Focusing on Locally Created Data to Improve Classification
The forest cover, density, moisture, and slope profiles of the EBCP ecosystem
classification was used exclusively in all the subsequent fuel classifications. This
provided consistency and locally confident data that could be matched to the fire
behaviour that fits the vegetation type, slope, moisture, and density of each
ecosystem. Forest ecosystems are represented by their vegetation density, moisture
regime, slope gradient. With these attributes, in addition to the shrub and forb
ecosystems, there are 51 ecosystem categories covering areas from 4 ha to 3000 ha.
To prioritize time and resources, ecosystem types that covered less than 200 ha were
not further assessed, resulting in 25 working categories that include 75% of the
territory. Existing photo-samples were mapped to ecosystems, and an inventory of
the usability and dispersal of the existing photo-points was undertaken to identify
gaps. There were 278 existing photo-points sampled in 15 ecosystems; however, they
were highly clustered around restoration areas where the crew had done work,
some in ecosystem types that covered little area in the Survival Territory. With
redundancies and ecosystems not of interest removed, 113 photo-points remained.
With a best-case-scenario target of 10 photo-points per ecosystem type, a
sampling procedure was developed with the XCF forest crew (closely following the
procedure used for previous site surveys) to take additional photo-points in undersampled ecosystems. Photo-points were situated in the middle of a polygon (as
59

geographically determined by GPS enabled tablet and map-viewing service) and
were taken to capture the forest structure from ground to mid-canopy. Pictures were
taken in portrait orientation in the four cardinal directions (north, east, south, west)
with the coordinates, compass direction, date, EBCP category, and samplers written
on a whiteboard in the picture (Figure 19).
N

E

S

W

Figure 19: Example of a set of pictures captured at a photo-point in the four cardinal directions (XCF
Forest Crew, 2017).

There were 15 categories with existing photo-points leaving 18 ecosystem
types needing sampling to reach 10 photo-points. Fortunately, the timing of
sampling fit well with the work schedule and staffing numbers of the forest-crew,
and they were able to collect over 60 additional photo-points. However, the final
samples fell short of the target of 10 samples per EBCP category during the study
period because of the many other demands on their time. By the end of sampling, of
the 25 ecosystems covering most of the valley, 15 ecosystems got 8 or more samples,
9 got 2 – 5, and there was one unsampled category.

60

Categorization into fuel types based on the Canadian Forest Service FBP field
guide (Taylor et al., 1996) started once the samples were collected. By referencing the
manual, the photo-points from each ecosystem category were assessed and assigned
a fuel type. Each category was evaluated a second time on a separate date in reverse
order to reduce learning-curve errors (Figure 20-A). This assessment was then
shared, discussed and re-assessed with supervisor Scott Green to reduce the effects
of personal bias (Figure 20-B). The categories were re-visited upon receiving the last
18 points from the XCF forest crew. The categories were further refined following
consultation with Mark Parisien (fire research scientist providing context for how to
localize FBP) and Dana Hicks (Lillooet fire management specialist), individuals with
a practised eye for fuel categorization. These conversations emphasized how the
localized nature of fire behaviour would greatly benefit from additional localized
fire expertise (Figure 20-C).

61

A

B

C

Figure 20: Progression of the EBCP fuel categories through different iterations of information and
knowledge.

Improving Iterations of Fuel Classification
There were substantial changes to the mapped fuel risk (Figure 21) as more
information was added to the fuel categorization process, and additional expertise
was incorporated. Alpine areas, grasslands, and shrubs stayed consistent throughout
the iterations because the focus was predominantly on the forested areas. However,
in the high-risk categories of C2 and C4, there were considerable shifts where the
last iteration has minimal area classified as high-risk and generally a much more
homogenous landscape of moderately high-risk fuel types.

62

Figure 21: Iterations of fuel classification from VRI/EBCP classification to straight EBCP categories.

Fire Risk Beyond Fuel
Throughout the iterations, the variability within ecosystems and the
homogeneity of fuel risk between ecosystems became apparent. The large area
covered by C3 fuel categories is a management challenge requiring additional data
to separate priority areas for wildfire risk mitigation. In an earlier iteration of fuel
categories, Marc Parisien suggested the inclusion of wind information as a risk factor
that echoed conversations with the XCF board and employees about how windy the
valley gets and how fires from their fishing sites along the Fraser River blow quickly
up the hillside with the wind.
63

The fuel categories produced were first used by the XCF team in November
2017 in a meeting with the Xáxli’p forest manager to select locations for restoration
thinning, focusing on locations within 200m or 500m of roads because of the
prohibitive cost of building roads. This highlighted the need for bringing in road
access as an element of identifying high-risk areas to target. Additionally, because of
the homogenous fuel landscape, the capacity of a small crew to reduce risk in high
fuel areas is overwhelmed. Focusing the logging projects to enhance low-risk areas
to connect them with nearby low-risk areas was proposed as a start to fuel reduction
work.
Assessing Wind as a Risk Factor
There is no regularly recorded wind data for the Xáxli’p Survival Territory;
the closest station with wind data is a BC Wildfire Service operated weather station
over the mountains in Lillooet (Station 1830). Wind data obtained from the weather
station was downloaded (Pacific Climate Impacts Consortium, 2019) and analyzed to
determine the predominant winds in the area in different conditions.
Throughout the day at the Lillooet weather station as temperature rises and
relative humidity (RH) falls, the wind speed increases and shifts to predominantly
westerlies (Pacific Climate Impacts Consortium, 2019). Throughout the year, the
average wind direction is evenly distributed, with a higher frequency of northerly
and southerly winds (Figure 22) and an average wind speed of 5.6 m/s. However,
64

this pattern shifts towards predominately west winds during the hot, dry, summer
conditions when wildfire risk is at its most significant concern (Figure 22). The
average westerly wind speed for hourly recordings with temperatures over 20˚C and
relative humidity less than 50% was 9.9 m/s, and it rises to 10.6 m/s at 25˚C and 30%
RH.

Figure 22: Comparison of predominant winds at the Lillooet weather station (Pacific Climate Impacts
Consortium, 2019).

65

Wind data was modelled using WindNinja V3.3.1 (Forthofer, Butler, &
Wagenbrenner, 2018), a software used for modelling the effects of topography on the
local wind (Parisien et al., 2012) providing an output that identifies areas of high
wind. The model inputs required for WindNinja include a digital elevation model
(DEM) and wind speed and direction. An average westerly wind speed of 9.9 m/s
were the inputs for the model, which was obtained from wind recordings from April
through September for recorded hours over 20˚C and a relative humidity less than
50% to capture summer fire conditions (Pacific Climate Impacts Consortium, 2019).
The conservation of mass model was run on a 25m DEM, using the standard 10m
Canadian weather station height. The output resolution was 25m and showed a
band of high winds at the top of the ridge on the western side of the valley and some
smaller high bands in the upper elevations on the east side of the valley (Figure 23,
A). The valley bottom had relatively low wind. The average WindNinja windspeed
for each EBCP ecosystem unit provides a comparable landscape management unit to
compare across the wildfire risk variables (Figure 23, B). This generalized the wind
patterns more, with fewer of the high wind pockets showing up in the averages.

66

A)

B)

Figure 23: Modelled windspeed across the Survival Territory for the average summer fire weather
conditions at the Lillooet Airport; A) the modelled raster output, B) the average wind speed for each
EBCP ecosystem unit.

Incorporating Access as an Ignition Proxy
The road layer was included in the wildfire risk assessment for the Survival
Territory because of concerns around human wildfire ignitions that are limited to
roadways. Additionally, Herb Hammond and XCF forester Robin Strong focused
logging planning on areas within 500m of the existing road because it could be easily
reached and there is hesitation towards creating new roads for logging (and for
other activities) because of concerns over how it opens access and increases the
ignition risk and pressure on the ecosystem.

67

The road network has a few implications to predicting the wildfire risk: 1)
road access dictates the distribution of ignition in the predominantly human-caused
ignition pattern local to the Survival Territory; 2) the road network limits the ability
to undertake eco-cultural restoration treatments without significant road-building
costs limiting this work to areas close to the roads and; 3) human habitation is
concentrated around these roads.
Table 4: History of ignitions in Fountain Valley comparing natural and human ignitions within
buffer distances from the road network.

LIGHTNING
PERSON
OTHER
TOTAL

200M ROAD BUFFER
in
out
5
31
211
52
12
6
228
89

500M ROAD BUFFER
in
out
7
29
245
18
12
6
264
53

ALL
36
263
18
317

There is only 11,205.9 ha in the Survival Territory that is within 500m of
access roads. Still, the road network acts as a conduit for ignition risk with over 80%
of recorded fires in the area over the last 70 years ignited by people (Table 4), and
93% of those human ignitions occurred within 500m of a road (Ministry of Forests,
Lands, Natural Resource Operations and Rural Development - BC Wildfire Service,
2018). The distribution of human ignitions is primarily along the valley bottom and
near the camping areas near Kwotlenemo and Chilhil Lakes (Figure 24).

68

Figure 24: Distribution of fire ignitions relative to access roads.

The data for the road layer was obtained from the XCF and cross-referenced
with the Digital Road Atlas layer from DataBC (Ministry of Forests, Lands, Natural
Resource Operations and Rural Development - GeoBC, 2017). The road layer was
verified by the XCF forest crew to try to obtain the most up to date and accurate
representation of the roads in the valley, however, the level of road maintenance and
driveability of the roads was not considered for this analysis but would be a future
step to refine the ignition risk posed by road access.
69

A Need for More Context
The level of uncertainty in the quantitative spatial data and processes of
assessing wildfire risk as a desk-top exercise does not produce the level of
confidence needed for a small, resource-limited organization like the XCF to
optimize their planning. Through the iterations of the fuel classification, the fuel
classification became more and more homogenous across the landscape. This
necessitated the integration of the wind and access layers, which incorporated a
broader understanding of the risks but brought assumptions and higher levels of
data uncertainty to the wildfire risk assessment.
Even though the XCF team guided the process, the fuel, wind, and ignition
assessments undertaken were unable to represent the complexity on the land,
emphasizing that basing land management decisions on remote data and models has
essential limitations for understanding the landscape and relationships with and on
the land. Discussions with the XCF team determined that it would be appropriate to
engage the community in small groups to interview Xáxli’p knowledge holders who
have experience on the land and with fire. Given there are people much more
knowledgeable about the landscape, who are living the relationships within the
community and spending time on the land, it is critical to incorporate this
information into the Indigenist approach of this gap analysis.

70

Chapter 3: Enhancing Wildfire Risk Datasets using Xáxli’p Community
Knowledge of the Local Wildfire Landscape
Following a year and a half assessment of forest fuels, wind and ignition
using existing data, classification maps were presented to the Xáxli’p Community
Forest (XCF) board, employees and community members who provided feedback
about specific details and concerns that were not being captured (see Chapter 2).
Incorporating community knowledge holders in the process was proposed to
provide additional context and nuance to the quantitative data.
Holding Conversations with Knowledge Holders
The XCF board and staff organized a focus group of Xáxli’p community
members with recognized knowledge of the Xáxli’p territory to attend a workshop.
The question I hoped to answer was: Where are the areas in the Xáxli’p Community
Forest tenure that local knowledge holders were most concerned about wildfire risk?
This local knowledge and perception of wildfire risk could then be displayed in a
risk layer that could be combined with the existing categorized quantitative data
assessing wildfire risk (wind, fuel and access), providing better context and nuance
to narrow the spatial scope of priority risk areas.
The workshop was held over two days in July 2018, with 14 participants,
including the Xáxli’p Range Riders, XCF forest crew, Elders, board members, and
valley residents selected for their knowledge of the territory and an expressed
interest or concern about wildfire. Some participants were able to attend both days,
71

others only for parts. A mapping exercise on day one and a field trip on day two
provided a space for participants to share and brainstorm with each other, as well as
to provide information to the research team. The primary discussion question was,?”
“Where is the greatest wildfire risk in Fountain Valley with subsequent inquiries
about fuel, wind, access, and important assets to protect in the valley.
Identifying and Mapping Concerns About Wildfire
The first day with the focus group was held at the XCF office and started with
a Xáxli’p prayer and an introduction to the larger project and workshop objectives.
Each participant received a map of the fuel categories, access roads and wind levels
that had been presented to the board and community from earlier research stages
(Figure 25) to provide some context of our findings about wildfire risk to that point.
Participants were asked to point out any areas of disagreement with their knowledge
of the landscape or provide feedback on the maps.

72

Figure 25: Quantitative data mapped in earlier research stages and presented for correction and
comment to the focus group.

After introducing the research question and workshop objectives, the group
conversation was roughly guided by a set of questions (Table 5) relating to
participants' concerns, knowledge and observations of wildfire risk, while also
following the natural direction of the conversation. Large poster-sized aerial
imagery maps that spanned the Survival Territory from north to south in three
sections were used to record areas identified by the participants (Figure 26). Areas
were drawn on the maps by participants, or me when participants only provided
verbal descriptions or gestures.
Table 5: Discussion questions used to guide the conversation with the focus group

1) What do you think matters most when you think about the wildfire risk in
Xáxli’p?
a. Where are you most concerned about a wildfire burning?
2) Are there areas in Fountain Valley that have a lot of combustible fuel?
73

a. What are the areas of highest concern?
b. When you look at the map, is there anything that strikes you as
incorrect?
3) Where is wind the biggest issue for wildfire risk?
a. Does the map match high-wind areas from the Fraser?
b. Is wind consistent throughout the day?
4) Where is there the biggest chance of a wildfire starting?
a. What starts fires in Xáxli’p
b. Which access roads have the highest traffic?
5) What areas are most important to avoid burning?
6) Where would adding fire too the landscape be good?
7) Are there any other fire risk aspects we are not considering?

Figure 26: Wall maps used to capture areas identified by focus group participants.

At the end of the first day, participants were each given three coloured
stickers to participate in a “dot-mocracy” data summary (Diceman, 2010) to help the
group rank the identified areas. Red stickers for the locations they felt were “most
important” regarding wildfire risk, green stickers for locations they felt were an
“easy fix” (low-hanging fruit that could be easily or inexpensively addressed to
reduce wildfire risk), and blue stickers for lower priority areas (Figure 27, A).
Participants were given time to go up to the maps and study the risk areas the group
74

had identified, to discuss with each other and to put their stickers on the maps
(Figure 27, B). Following the workshop, the areas drawn and dots placed during the
workshop were digitized and mapped for the following day.
B)

A)

Figure 27: 'Dot-mocracy' colour legend (A) and participants placing 'dot-mocracy' stickers (B)
(Bezzola, 2018)

Site Visits and Discussing Solutions
On the second workshop day, the group visited a few sites in the valley
identified the day before as either high-risk or that provided a view of high-risk
areas. Sites were chosen by the participants at the beginning of day two based on
interest, accessibility and time constraints. The west side of Kwotlenemo Lake,
Sobotka’s Ranch and the area below the band office were visited over three hours
(Figure 28).

75

Figure 28: Sites visited with participants on the second workshop day to see some examples of wildfire
issues identified on day 1.

Participants were provided with maps of the priority risk areas identified the
previous day (Figure 29). While in the field, participants added some additional
areas and details, but for the most part, participants observed and confirmed what
had been expressed on the first day. Conversations in the field evolved toward
identifying and discussing risk mitigation options that the XCF team and the wider
Xáxli’p community and Fountain Valley residents could undertake. Following the

76

field discussions, all the maps and meeting notes were synthesized into a report and
shared back to participants through the XCF team.

Figure 29: Areas contributing to wildfire risk identified by the focus group.

Wildfire Risks Discussed in the Workshop
Wildfire risk is not a new concern for the Xáxli’p community; they anticipate
and experience wildfires of varying sizes and causes in their Survival Territory every
77

season. Despite the Xaxli’pmec history of lighting fires to manage the landscape for
food and safety, land management practises and wildfire suppression imposed by
the government over the last century have increased fuel loading in the forest and
curtailed the ability for Xaxli’pmec to manage the fuel levels in their Survival
Territory (see Chapter 1 for more details). The workshop attendees were
knowledgeable about wildfire risk, and their concerns over wildfire risk are not a
recent phenom; they expressed that their concerns had been accumulating over
years and referenced conversations or community efforts relating to wildfire risk
undertaken over the last decade. Over the two days, there were recurring themes
that highlighted locations and management practices that participants identified as
contributing to increasing wildfire risk.
Clear patterns in the distribution of identified risk areas were generally
mirrored in the dot-mocracy points (Figure 29). The north area around IR1, the
middle of the valley around Kwotlenemo Lake and Chilhil Lake, and in the south
around Kirby Flats were the primary areas of concern, with the first two coinciding
with areas of high occupancy, while the south end has high recreation but fewer
inhabitants.
Fuel Risks Identified Around High Occupancy Areas
Mirroring the homogeneous fuel pattern identified by the quantitative data
classification, “the whole valley is high-risk” was a point that got brought up
78

frequently, and this initially caused participants some difficulty in identifying
specific areas of high fuel risk. The priority wildfire fuels identified by participants
throughout the valley were grasslands that dry out in the hot summer and/or
accumulate over years if they do not get mowed or grazed, dense forest fuels that
have grown thick because of wildfire suppression and land management practices
and dead trees from pest and disease. Additionally, in the southern end of the
valley, the deep duff layer and deadfall were identified as a priority risk.
The initial fuel assessment using the existing quantitative datasets focused on
dense, mature forest areas of C2, C3, C4 that were presumed to carry the greatest
risk (see Chapter 2). However, the workshop participants expressed more concern
over the grasslands as a fuel type due to their proximity to residential and recreation
areas and their ability to act as a conduit between high ignition areas and forests.
Transmission of Wildfires from Ignition to Forest via Grasslands.
Most grasslands (excluding alpine meadows) are situated along the valley
bottoms through Fountain Valley and along the wide flat terrace between the Fraser
River and the entrance of Fountain Valley. Shrubs dominate the step up from the
river terrace into Fountain Valley, around the band office (IR1), and the southern end
of the valley has open ponderosa pine forest with a grassy understory. These grass
and shrub fuels were identified by participants as being a major concern at the north

79

end of the territory at the fishing grounds, in fields along the highway, as well as
grassy yards on the reserves (Figure 30, Area A).

Figure 30: Grass and shrubland contributing to wildfire risk around the north of the valley and
Kwotlenemo Lake and Chilhil Lake identified by the focus group.

In the center of the valley around the Kwotlenemo Lake and Chilhil Lake
(Figure 30, Area B), participants were concerned that the grassy fuels around the
80

houses, campsites and fields would transmit wildfire easily to the dense forest west
of the lakes. This dense forest west of the lakes was the primary concern for forest
fuels in the territory, flagged for the forest density and high amounts of dead
standing trees (Figure 31). Participants identified an increase in grassland fuels
associated with recent decreases in grassland management and less cleaning up
around fishing grounds. Grassy fuels in the fields north of Kwotlenemo Lake and
south of Chilhil Lake (Figure 31) have become an increased concern since the ranch
stopped haying the fields, and horses or cattle are no longer grazing other fields like
they used to. Along the Fraser River are the traditional Xáxli’p fishing grounds
where participants expressed concern over the dead grasses and other debris
building up around the fishing racks (Figure 30, Area A). Traditionally, these areas
were burned to clean them, but this is a practise that no longer occurs regularly.
Recorded wildfires have swept up the hillside that started at the fishing grounds
because of this issue.

81

Figure 31: View looking south-east over Chilhil lake showing Fountain Valley Road providing easy
access to the lake and fields that have an accumulation of dry grassy fuels. In the foreground, you can
see dead standing trees on the hillside on the west side of the lake (XCF Forest Crew, 2017).

Risks Posed by Dense Forest Fuels.
Many workshop participants commented that they had observed the forest
getting increasingly dense throughout the valley over the years and that because old
trails have grown in, it has changed how they access specific locations in the
Survival Territory. A particular concern over forest fuels focused on the center of the
territory, along the valley bottom and a bit up the valley sides (Figure 32). The west
side of Kwotlenemo Lake was the first area identified in the workshop, described
with comments like “it is going to burn soon” (Figure 32). They have observed high
82

tree density with plenty of ladder fuels and dead standing trees in this area, and the
area nearby was identified as a risk for grass fires that could transmit wildfire from
the high probability ignition campsites to the forests with the highest fuel concern
(Figure 32).

Figure 32: Forest areas contributing to wildfire risk identified through the centre of the valley by
workshop participants.

83

Other Fuel Concerns.
Standing and fallen woody debris and duff were concerns throughout the
valley, both during the focus group and the field trip. The dead trees in the forest,
identified by participants as “bug-kill,” contributing to the fuel load were a concern
for some participants, “the bugs are going to get the valley, or the fire is going to get
the valley.” The viewpoint above Sobotka’s ranch visited during the field trip was
chosen partially because from there, the dead trees were visible on the west side of
the valley (Figure 33).

Figure 33: Fieldtrip viewpoint from Sobotka’s ranch overlooking the valley and the dead standing
trees on the west side of Fountain Valley (Bezzola, 2018).

The open grassy ponderosa pine forest at the south end of Fountain Valley
around Kirby Flats was identified as a fuel risk for its deep duff layer. There was
also concern expressed over the fuel loading around the Kwotlenemo Lake and

84

Chilhil Lake because residents are getting older and less able to manage the fuels on
their properties in an area of higher ignition potential.
The role of deciduous trees in a wildfire landscape was not clear to all
participants. Some participants with experience fighting wildfires explained the
lower flammability of deciduous trees when they are leafed out. The understanding
of many was that all dense vegetation would increase the wildfire risk, and they felt
safer after cottonwood removal; however, others were concerned over losing
cottonwoods for both cultural and wildfire risk reasons.
Ignition Risks Identified
According to provincial records, historically, the predominant cause of
wildfires in the Survival Territory has been human activities, with limited
observation of wildfires ignited by lightning by workshop participants, which was
reflected in the quantitative data (see Chapter 1 & 2). The grassland areas align with
land-uses that are associated with higher ignition risk areas: valley bottoms where
people live, roads, campgrounds, fishing grounds. Ignitions from lightning,
backcountry access, and people accessing roads were not identified to point sources
by the focus group but were raised as a concern throughout the valley, with some
regions of higher likelihood identified. Managing the activities of people to limit the
chance of ignition through fire-bans, access restrictions (Figure 34), and patrolling

85

the backcountry on the part of the Range Riders contributes to Xáxli’p wildfire
prevention throughout the summer.

Figure 34: Sign at the northern entrance of Fountain Valley, informing residents and visitors of
wildfire hazard rating, access limitation and fire ban to reduce the risk of ignition (Bezzola, 2018).

Human Ignition Patterns in the Territory.
The portion of the Survival Territory that is in the Fraser Valley has some of
the highest concerns for wildfires starting, with highway traffic, sparks from the
railway tracks, and activity at the fishing grounds which have started wildfires in
the past. Just up-slope at the mouth of Fountain Valley is the main reserve (IR1),
with a higher concentration of houses creating higher safety and cost consequences
if a wildfire were to burn there. This human-wildland interface creates more
complex management conditions to keep ground fuels low relative to areas with no
houses or off-reserve land.
At the campsite, ignitions from campfires are a significant concern because
even with measures in place to control this risk with fire bans and posted signs,
there are often still fires getting lit. Just days before the workshop, an abandoned fire
86

was found by residents despite the fire ban and extreme fire danger rating. Activity
at the fishing grounds, the gathering place, and the campsites is seasonal, with
elevated risks occurring when the areas are most used and when the fuels are driest.
Road Access Associated with Ignition Patterns.
Road access is a concern for fire ignition because it increases the area
accessible to the public, and therefore, the area with a higher chance of human
ignited wildfire. The main areas of ignition concern identified by the focus group
were along Fountain Valley Road, Rough Creek, the bike trails around Kirby Flats,
and Sobotka’s ranch (Figure 35). To the east of Kwotlenemo Lake, there is road
access to the powerline; this concerned the focus group because the high traffic
increases the ignition risk. Participants suggested reducing traffic and posting
information signs during times of higher risk and reducing fuels 10-20 m on either
side of the road to reduce the likely hood that sparks evolve into wildfires.

87

Figure 35: Areas identified for high ignition risk associated with road access.

In the event of a wildfire, access is a limiting factor for both public safety and
wildfire suppression. The main Fountain Valley Road connects Highway 99 at the
north to Highway 12 in the south, providing the only other access into the valley.

88

Lightning Ignitions.
The pattern of lightning in the valley and associated wildfire risk was unclear
to most participants. In conversations about ignition, there were questions from
participants whether lightning was a threat, some observations that there was not
much lightning or observations that wildfire ignited by lightning tend to stay small.
Wind Patterns
During the workshop, there was little detail about the wind patterns because
of its variability throughout the day and territory. Out in the field, it was not a topic
that came up at all, perhaps reflecting that the wind weighed less on the minds of
participants than other issues. The key characteristic of wind in the Survival
Territory identified by participants was the wind tunnel effect created by the
topography, with the wind blowing mostly from the north or south, though down at
the Fraser River, the wind from the south is funnelled up-river. It was identified by a
few participants that it is “windy all the time at the main reserve” with the wind
coming off the Fraser River and over the high ignition risk areas of the
highway/tracks/fishing grounds up to the band office which therefore has a highrisk for quickly pushing a wildfire south into Fountain Valley (Figure 36). The
grassland areas are susceptible to wind-blown wildfires in these areas, transmitting
the fire to surrounding forests, which participants identified as a high concern. The

89

southern end of the valley around Kirby Flats was also identified as a risk for high
wind.

Figure 36: Areas identified for particularly high wind by the focus group.

The variability of wind was highlighted with comments about the wind
switching every hour, the strongest winds in spring and fall, the diurnal patterns of
90

the wind coming downslope at night, and a shift in the wind behaviour at one
participant's house after the 10-Mile burn.
Risk Mitigation Measures Discussed
After highlighting the main areas and causes of wildfire risk, participants
started to suggest options for reducing fuels and managing the community’s safety.
This natural progression of the conversation was not planned, but it was evident that
participants had considered these threats over the years and developed ideas about
potential solutions.
Creating Fuel Management Areas
Fuel management areas (called firebreaks, fireguards, or fuel breaks by some
community members) were raised as a wildfire risk mitigation option multiple times
by participants. Viewing the landscape in the field prompted participants to identify
and discuss potential fuel management locations. Two specific locations were
identified in the valley, both tied into natural low risk landscape features and to
isolate high-risk fuel conditions that could threaten the nearby residences.
In the north end of the valley, participants suggested the creation of a fuel
management area to mitigate wildfire risk posed by the high probability of ignition,
grassy fuels and high winds (Figure 37, Area A). Measures included keeping a wellmowed area along the edge of the field at the mouth of the valley (Figure 37, Field
Fuel Break), as well as a fuel break in the shrub/deciduous area between Fountain
91

Valley Road and Rose Ellen’s house on IR1 west of the Band Office (Figure 37, IR1
Fuel Break). During the field trip, the viewpoint looking over the valley from
Sobotka’s ranch provided an opportunity to identify where a fuel management area
could be tied into existing low fuel-risk features in this area (Figure 37 Area B). From
the more open, rocky bluffs on the west side of the valley, to the field at the valley
bottom up the east side along a creek draw, these features could be further thinned
and maintained, while dense connecting forest could be cleared or aggressively
thinned to provide a fuel break that connects with the transmission line running
parallel with the valley.

92

Figure 37: Fuel breaks identified by focus group participants.

Views on what targeted fuel management areas should look like ranged from
logging most of the valley, to clearing strips across the valley at regular intervals
(1/2-mile-wide breaks, every 5 km), to creating a few breaks that build on natural
low fuel-risk features. There were some conflicting views on the role of logging; on
one side was the suggestion that more of the forest should be logged, and on the
other side that the focus needs to be on restoring the ecosystems, not logging. If the
thinning and creation of fuel management areas can be profitable it can provide

93

some income to the XCF to increase other wildfire mitigation measures and can be
tied into eco-cultural restoration work.
Controlled Burns
Some Xáxli’p community members and the XCF forest crew use fires to
reduce grassy fuels, particularly around houses. The option of increasing the use of
controlled burns, specifically south of Chilhil Lake, between Kwotlenemo Lake and
Chilhil Lake (using the road as a break), below the Xáxli’p band office, and around
the fishing grounds were discussed during the workshop. There was a suggestion to
increase fall burning, which tied into a separate comment about the higher amounts
of dry grass fuels they are expecting next spring because of the wet summer and that
with fall burning this fuel risk could be reduced in the spring. There was a reminder
from the group that with the amount of fuel in the forest, it was not going to be easy
to reintroduce wildfire to large areas in the territory. Participants rely on the XCF
forest crew to decide where to burn, and people feel safer after the burning
undertaken by the crew.
Maintenance to Reduce Fuels
Workshop participants wanted to see more routine fuel management happen
on properties throughout the valley. This included helping older people who can no
longer manage to clean up fuels around the lake, cleaning up around the reserve and
the fishing racks along the Fraser River. They identified grasslands as straight94

forward fuels that could be directly managed with grazing, mowing, haying and
spring burning. Participants mentioned that there had been a study done to identify
how many head of cattle could be run in the valley, and there was further discussion
around bringing back horses to graze or other animals such as pigs or goats for
cleaning the ground.
Participants expressed support for work the XCF forest crew has undertaken
to reduce forest fuels such as tree density thinning, live branch pruning, and slash
management as well as the plans to re-introducing wildfire (Figure 38). The XCF
staff had already scheduled some specific areas identified by the focus group for
high-risk fuels for fuel reduction work, which was another positive reinforcement
for the work of the XCF.

Figure 38: Example of forest thinning undertaken by the forest crew for eco-cultural restoration near
Chilhil Lake (Green, S. 2017).

95

The positive view of ongoing eco-cultural restoration work was tempered by
feedback from a participant with experience doing prescribed burns who indicated
that clearing underbrush would need to happen as frequently as every five years to
keep the new seedlings down and bringing back wildfires will need to be done very
carefully. They advised that the fire regime has shifted so that areas can not just be
lit at the bottom and let go. Additionally, the manual fuel reduction work has
limitations because of limited resources available to the XCF; one participant
estimated that it would take 40 years to thin the valley at the current rate.
Emergency Response
The participants recognized that in a massive wildfire event, there would be
nothing the community or fire-fighters could do to stop it even with extensive
thinning and fuel management areas. In this case, participants felt the priority
should be to focus on “saving” the valley bottom, not on the upper elevations
because of the risk to lives. Logistical concerns about ensuring the safety of residents
included how to evacuate the valley inhabitants without having the correct
addresses of valley residents and poor communications infrastructure.
The desire for the formation of a Xáxli’p wildfire crew to fight wildfires in the
valley was expressed, with concerns over the current capacity and resources
available in the valley to extinguish wildfires. Ensuring access to water throughout
the valley by stationing water throughout the valley and fixing the waterline on the
96

flats was suggested. Additionally, there are improvements needed to the road and
bridge that currently prevent access to the west side of Kwotlenemo Lake with the
fire wagon.
Conclusion
Patterns in the distribution of high-risk areas became apparent when the
focus group mapped out their concerns. The north around the main reserve (IR1),
the middle of the valley around Kwotlenemo Lake and Chilhil Lake, and in the
south around Kirby Flats were the major clusters of identified risks, with the south
generally identified as a lower priority. The first two areas coincide with areas of
high occupancy, while the south end has high recreation but fewer inhabitants.
The significant concern participants expressed for the risk posed by grass and
shrubland contrasts with the greater focus on forest fuels in Chapter 2, this
prompted a re-assessment of the context of the fuel categories and how it fits in with
the eco-cultural landscape. The workshop also focused the priority on the lower
elevations with the higher ignition risk, greater consequences because of habitation
and more detailed information about the landscape.

97

Chapter 4: Integrating Data Types and Exploring Key Learnings
Keeping this research grounded in the relationships between people,
information, ecosystems, and history was vital to maintaining the relevance of
finding and solution that aligned with the Xáxli’p priorities for their Survival
Territory. As the initial phase of a long-term partnership between the University of
Northern British Columbia and the Xáxli’p Community, the objective was to
examine the utility of existing datasets, information, and methods as surrogates for
wildfire risk in usable and suitable outputs for XCF eco-cultural restoration
planning. Through the research process, existing quantitative datasets for ecosystem
structure, weather and road access were adapted to represent the fuel, wind and
ignition conditions. As proxies for risk, we found inherent limitations in this
approach; consulting community members attempted to explore, validate and
mitigate these limitations.
As individual elements, the value of the wildfire risk classification layers (fuel
type and connectivity, average wind conditions throughout the valley, and ignition
source and distribution) for supporting eco-cultural restoration planning remains
unclear. Building on these classifications, the conversations and information
gathered in the community workshop brought together the fuel, wind, and ignition
issues, weaving between different concerns and locations to build a more holistic
and spatially focused view. Combining the quantitative and community data was
98

not an attempt to discredit either, but to contextualize and prioritize the different
risk factors and to identify patterns and shortcomings. Bringing the layers together
provided a more spatially focused view where overlapping wildfire concerns
highlighted potential priority areas for mitigation activities, and differences in risk
perception highlighted locations and systems that would benefit from further
research.
Integrating the Quantitative Categorical Data
As stand-alone risk layers, the fuel, wind, and ignition risk data individually
fail to capture the relationships on the landscape, providing an incomplete
assessment of the overall risk. Identifying critical overlaps in high-risks of the fuel,
wind and human access were examined to identify potential high-priority areas.
Ecosystem Based Conservation Plan data were used for fuel classifications and
provided a unit of aggregation to examine the modelled wind speed, and the
ignition/access areas for each ecosystem patch to spatially refine potential priority
risk areas.
Combining Fuel and Wind Risk Data
Higher wind areas that would stoke the spread of wildfires tended to occur at
higher elevations, while lower winds tended to occur in the valley bottom. When
isolated to high-risk fuels (C4, C3, and M1-75), ecosystems classified in the highest
categories (>15 m/s) made up a very small portion of the landscape, with most areas
99

in the 5-10 and 10-15 m/s wind speed categories (Figure 39). The dispersal of the
high wind (>10m/s) covers about half the high fuel-risk landscape, an area too large
to apply fuel mitigation and mainly at higher elevations, inaccessible areas in the
Survival Territory, while the high fuel-risk ecosystems with lower average wind
speeds (<10 m/s) are nearer in the valley bottom.

Figure 39: Average wind speed in ecosystem units with high-risk fuels (C4, C3, and M1-75).

100

Wildfire Ignition Risk Attributed to Road Access Narrows the Focus Area
Historically, wildfires started by humans have been the dominant ignition in
the Survival Territory with the ignition of wildfires heavily concentrated to within
500 metres of roads (described in Chapter 2). Consequently, human-caused fires
may indicate a primary factor in identifying priority treatment areas. While it is
more likely that wildfires start within 500 metres of a road, there remains a risk of an
ignition occurring anywhere else in the territory, either by human cause or lightning.
Integrating ignition probability as a wildfire risk factor reduces the high-risk
area substantially, providing smaller locations for eco-cultural restoration work to
target. Within 500 metres of a road, high-risk fuels (C4, C3, M1-75) dominate much
of the valley bottom and a large area in the northeast region (Figure 40). Lower risks
fuels are more common near the road at the south and northwest ends of the valley.
Low modelled wind speeds characterize most of the valley bottom (below 10 m/s),
with many areas below 5 m/s. Higher wind speeds expected in the higher elevation
roads in the northeast and steep highway sections in the northwest (Figure 41). The
high-risk fuel areas within the road access buffer highlight areas for fuel reduction
where the chance of ignition is higher and can be accessed without adding more
road infrastructure.

101

Figure 40: Fuel risk categories within 500
metres of roads.

Figure 41: Average windspeeds for ecosystem
units within 500 metres of roads.

Lower wind in the valley bottom where there is a higher ignition risk reduces
the overall risk for this corridor. However, the high ignition risk area simply has a
higher probability of ignition than surrounding areas; it only represents a more
likely scenario. There is a risk of ignition outside of this area, by lightning for
example, that may be less likely but needs to be considered along with other
scenarios for land-use planning and risk management.
Overlaps in High Fuel, Wind and Ignition Risks
Combining all three quantitative datasets highlights potential priority areas
for eco-cultural restoration. In the area 500 metres from roads with high-risk (C4, C3,
102

and M1-75) fuel types, the high wind areas tend to be in the northeast, a few small
pockets mid-slope in Fountain Valley where side roads go up the hillside, and a few
pockets in the northwest up from the Fraser River (Figure 42).

Figure 42: Average wind speeds stratified into high (<10 m/s) and low categories (>10 m/s) in highrisk fuel types (C4, C3, M1-75) and high ignition risk areas (within 500m of roads) to show the
integration of the three key risk variables.

While the northeast access area may have the highest risk of severe wildfire in
the Survival Territory, there is less risk to public safety and community
infrastructure, potentially making it a lower priority for fuel reduction than the

103

classification would suggest. There is uncertainty about the value and utility of the
indicators used to assess the wildfire risk because they do not embody community
interests and values; community input was incorporated to clarify, confirm or revise
the utility of the indicators.
Refining Risk Indicators Using Community Perspectives
In order to capture more nuance and understand the local wildfire risk and
restoration priorities, the priority areas identified by the community workshop were
overlaid on the quantitative categories for fuel, wind, and ignition risk. The
differences and similarities between the community perspectives and the
quantitative datasets were examined to show the priorities of each dataset.
Community Experience with Access and Ignition Patterns
Community workshop participants were confident about which roads
presented a greater ignition risk because of increased use, the type of use or limited
monitoring (Figure 43). The northeast section disappears entirely as a priority risk
area because the focus was on the locations people use most for recreation through
the valley bottom and fishing along the Fraser River. The fuel risk in this high
ignition area is highest around the lakes, on the slopes west of the lakes and higher
up the valley sides near Rough Creek in the south. In the more open forest at the
south and the shrub/grassland areas around the fishing grounds at the north end of

104

the valley have lower risk fuel classifications for these higher ignition risk areas
(Figure 43).

Figure 43: Fuel risk categories in the high ignition risk areas identified in the community workshop
compared to the ignition risk of the entire road network buffer.

Community Fuel Priorities Emphasize the Risk of Grasslands
When restricted to the priority areas identified by community workshop
participants, the critical fuel risk areas emerged along the west side of the lakes and
up the adjacent hillsides (Figure 44) in a pattern similar to the ignition risk (Figure
43). In the quantitative data exploration, the quantitative fuel risk assessment
105

associates hazardous fuel categories with continuous, dense forest stands with
plenty of ladder fuels that support intense crown fires with high tree mortality. In
contrast to this, for the community members that participated in the workshop, the
assessment of fuel type priority reflected more about whether it could support
wildfire spread across the landscape from high ignition areas, with particular
concern regarding grasslands.

Figure 44: The fuel classifications in the areas of fuel concern identified by the community.

106

Grasslands that can quickly transmit wildfire to adjacent forest fuels (Figure
45) are a hazardous fuel area despite not burning as intensely as a dense forest.
Because of this, these grasslands pose a substantial risk of supporting largelandscape wildfire events – the worst-case scenario for the Xáxli’p priority of cultural
continuance. Dense forests and grasslands display distinct wildfire behaviour and
require different management treatments to reduce risk. The prioritization by the
workshop attendees identified grasslands, like those at the north end of the territory,
as priority fuels requiring different fuel treatments (grazing or mowing) than the
high-risk forest fuels (thinning).

Figure 45: Grassland area near Chilhil lake that could quickly transmit fire to surrounding dense
forests under the right conditions (Green, S. 2017).

107

Community Identified Wind Patterns
As was identified in Chapter 3, the information collected from the community
members in the workshop was too limited to identify high-risk areas for wind. The
two areas of high wind identified by the community members are modelled as low
wind with the quantitative data model. The conflicting information and limited
community information highlights the need to have more wind monitoring in the
Survival Territory to provide more systematic and accurate information for decision
making and planning.
Data and Research Limitations
The data used and created in this research could be input into predictive fire
models to evaluate the productivity, patterns, and behaviour of wildfire on the
landscape; however, some important limitations should be considered when using
the data at this stage. The quality of input data determines the quality of the model
outputs; differences in measurements, models, and terms applied by practitioners
propagate errors and inconsistencies and may underpredict the wildfire risk in
many fire models (Alexander et al., 2013). Even with high-quality data and good
models, wildfire predictions reflect probabilities that cannot reconcile stochastic,
random, emergent and unpredictable events. There is a higher probability of ignition
near roads in the Survival Territory, but a random event such as sparks from a
nearby wildfire or a lightning strike in a remote area cannot be predicted.
108

Additionally, efforts to reduce wildfire risk through fuel treatments have varying
degrees of success when modelled (Alexander et al., 2013), and in the event of
extreme fire weather or a severe wildfire, the most drastic fuel reduction treatment
will not guarantee that wildfires will not start or spread.
Uncertain Accuracy of Fuel Data
There are a few factors that could influence the reliability of the data and
methods for fuel classification. The collection of the photo-points introduces
potential misrepresentation because of the uncertain accuracy of the GPS units.
Some photo-points were on the edge of two distinct ecosystems. With the range of
GPS accuracy, this variation could result in point coordinates being in a
neighbouring ecosystem from the ecosystem captured in the pictures or on the
border looking into two different ecosystems. With only ten photo-points to capture
the nature of an ecosystem type, there may be some bias and misrepresentation for
the ecosystem. The dispersal of the photo-points is not randomized; there is some
clustering of points, and the whole range of any given ecosystem was sampled as
best as could be done by the crew. In ecosystems that did not get all ten photopoints, there may be a higher risk of sample bias.
Though there are some accuracy concerns with the sampling, additionally,
the visual review of pictures suggests that considerable natural variability exists in
some ecosystem units, more so than might be seen in other environments (such as
109

boreal forests, where classifications were developed). The fuel categories assume a
homogeneous fuel structure throughout a single fuel type and the variability
indicated from the pictures indicates that some ecosystems may have more variation
than should be grouped to a single fuel category.
Localized Wind Data for Complex Topography
Modelling the natural environment always requires simplifying assumptions
and generalized results in the output. In the case of the WindNinja model for the
Xáxli’p Survival Territory, it is unclear how the wind observation values from the
Lillooet airport might affect the accuracy of the model outputs for the territory.
There is significant terrain variability in this area, and mountains separate the
Lillooet airport from the Survival Territory. With strong diurnal wind patterns
coming up from and along the Fraser River, it is unclear whether the model input
should reflect that, or whether the model is sensitive to that pattern. The information
provided by community members was very localized to areas they spend a great
deal of time (ex. the main reserve, where they live); this small sample size prevents
comparison with the wind model.
Ignition Depends on More than Roads
A better proxy for ignition risk in the Survival Territory needs development.
The road network is a starting framework because it is the limiting factor for where
people go. However, based on community input, not all roads are used equally or
110

for the same activities, with the northeast road network seeing less traffic than the
valley, for example. The stratification of the road network by how often, and for
what purpose, roads are used would then be able to put greater emphasis on areas
that have more traffic and higher risk activities such as campfires. The roads
identified in the community workshop is a good first start but could be further
enhanced with a more systematic review of road conditions and traffic counts.
Synthesis of Findings
As an examination of available indicators of wildfire risk in the Survival
Territory to help prioritize Xáxli’p eco-cultural restoration areas and activities, the
research scope and objectives evolved as new limitations and possibilities came to
light. The initial focus of the research was on developing an accurate
characterization of fuels, which was assumed to be the critical factor in defining
mitigation measures (BC Wildfire Service, 2015; Thompson & Calkin, 2011). Each
iteration of fuel classification resulted in a more homogenous fuel landscape,
making it more challenging to define spatial priorities to support XCF strategic
planning for eco-cultural restoration. The incorporation of wind and road access
added more holistic and locally relevant information on risk. However, the
limitations (primarily questions about relevance and accuracy) in all the datasets left
a great deal of uncertainty and focused the project on a ‘gap analysis’ process from
which further work could build on.
111

Community Perspectives Focus Broad Data Assumptions
Preceding the community workshop, the research was primarily focused on
forest fuels, followed by wind and ignition considerations, but with limited
information about how the community used the land and perceived wildfire risk.
The knowledge and experience shared by the community workshop participants
regarding the critical importance of grassy fuels broadened the primary focus on
forest fuels, particularly in the absence of comprehensive fire modelling. The reevaluation of grassy fuels is confirmed by Ray et al. (2012) in that Traditional
Ecological Knowledge (TEK) provides localized calibration to national narratives
and that disagreements do not invalidate either side. The localized knowledge
refined the potential treatment areas, which were initially less focused and specific
with the broad, general information from the quantitative data classifications. Local
input helped to identify relevant, more targeted areas and activities that could be
addressed by a small forest crew. Even though the knowledge holders frequently
mentioned that “the whole valley is high-risk,” on a secondary level of risk
assessment, they identified the more nuanced relationships on the landscape relative
to the quantitative information.
Growth and Challenges for the Researcher
The process of undertaking a thesis project, cultivating a new perspective
with Indigenist research, and exploring my responsibility for reconciliation provided
112

many opportunities to grow and reflect, and accept the underlying uncertainties
involved. Following a community’s lead and pace in community-led research was a
lesson in becoming comfortable with the uncertainty of not knowing, and not
insisting on controlling, where and when things will occur or evolve.
A crucial learning for myself as a researcher was the challenges and barriers
for Indigenist research, which relies on the ability to build relationships (Figure 46),
that being removed from the research location, both geographically and culturally,
pose. Access to the Survival Territory was available in day trips on paved highways;
however, winter weather, large wildfires, and mudslides all presented challenges
during the research, preventing visits or doubling driving time (and therefore
restricting time spent in the community). Additionally, over half of the planned inperson meetings got rescheduled because of weather or unexpected events in the
community.

113

Figure 46: Discussing photopoints and field data with the XCF forest crew (XCF Forest Crew, 2017).

Not being from the community or having an established relationship was also
a barrier to doing the best possible research and delivering the most useful outputs
to the community because without the relationships, knowing the culture and land,
there was a limit to what I even knew to enquire about or follow up on. During the
workshop with knowledge holders, there were references to spiritual sites without
any detail, which I avoided asking about because I did not want to overstep. This is
an assumption I made, which may have been appropriate or may have prevented a
useful and enlightening conversation for myself, the workshop participants, and the
research. This shortcoming reinforces the importance of the work of Indigenous
researchers in their communities and the importance of building strong partnerships
114

where both sides have the time and resources to feel comfortable and safe to have
open and honest dialogue.
Beyond respecting Xáxli’p values and honouring the Indigenist research
approach, there was a process of learning and understanding my ethical
responsibilities to the community when producing outputs and discussing findings.
There is an ethical responsibility to not just produce isolated outputs, but to explain
the risks and limitations of the datasets and outputs created and to contextualize the
benefits and dangers of applying the maps and conclusions so that there is no
wrongly placed confusion, fear, confidence that could affect land-use and
community decision making. Maps display discrete results which can create a false
sense of certainty. With most of my outputs and results communicated in maps, it
was important to ensure the limitations were well communicated.
Recommendations for Research and Management
As a gap analysis to explore the existing datasets, the gaps and strengths
identified in this research can be used as a foundation for further study in the Xáxli’p
Survival Territory to provide more tailored and nuanced guidance. In the interim,
the following sections outline broad recommendations for land management and
highlight gaps in the quality and type of quantitative data available for fuel, ignition
and wind.

115

Future Research Opportunities
There is significant uncertainty in both the fuel and wind data for the Survival
Territory. Increased data quality for both would enable more in-depth fire modelling
using Burn-P3 (Parisien et al., 2012), allowing for more significant predictive
research into the effects of climate change in the Survival Territory, the effectiveness
of existing and planned restoration treatments and plans for safety and evacuations.
The combined quantitative fuel, wind and ignition risk layers identified a high-risk
area in the northeast, where the most clear-cut logging and plantations are located.
This calls for a closer look at whether plantations increase the wildfire risk
significantly in this landscape (and to what stocking standard) and should be
managed differently going forward, or whether the existing fuel risk categorization
should be adjusted for plantations.
Improving Fuel Data.
Emerging methods for assessing forest fuels (such as airborne laser scanning
or LiDAR) can provide very detailed information about the ground surface, forest
understory, and canopy, such as tree spacing and forest history. It could be used to
model forest fuels to a very fine resolution, with consistent coverage and accuracy of
the territory, to establish tree-spacing objectives that resemble historic fire-resilient
forests and many other uses. Increased photo-sampling of the Survival
Territory to include all ecosystem types and more uniform coverage beyond
116

restoration sites and valley bottoms would also increase the quality of the fuel data
for relatively low cost and provide valuable ground-truthing and insight into the
ecosystem. An important outcome of this research is already underway with the
initiation of the second phase of the research partnership between UNBC and the
XCF to address this need with LiDAR data gathered in 2019 and 2020.
Implementation of Wind Monitoring.
Data from a weather station in the Survival Territory, or more basic ad-hoc
sampling could provide a higher degree of accuracy in the wind model (Forthofer et
al., 2018; Turner & Lawson, 1978), and therefore greater confidence in wildfire
models. More localized wind monitoring and modelling could also support local fire
fighting, increasing efficacy and safety. Any future research projects in the Survival
Territory may also benefit from having local weather data available, helping the XCF
planning and operations. Recognizing that resources are limited in such a small
organization, as a first step, simple hand-held wind measurement units could be
taken out into the field and be used to do spot measurements, which would be better
than having no wind measurements at all.
Assessment of Temporal Variability.
Seasonal variation in many aspects of the wildfire risk in the Survival
Territory, particularly activities that cause ignitions, pose an ongoing challenge in
prioritizing areas and activities in Xáxli’p eco-cultural restoration planning. For
117

example, the high ignition risk identified around the fishing sites is most relevant
during the times these areas are in use for fishing, or when the sites are maintained.
Further refinement of the wildfire risk assessment should take a closer look at
temporal variability to provide more useful management strategies.
Land Management Recommendations
The ongoing eco-cultural restoration approach and fieldwork by the XCF
align with many practices for creating a fire-absorbent landscape and managing the
wildfire risk. Continuation of this work is essential, and a few recommended new
practices and locations of new eco-cultural restoration work are offered below. All
suggestions should be aligned with the EBCP and Traditional Use Study to ensure
they are suitable for the Xáxli’p community and eco-cultural objectives.
Managing Grassland Fuels.
A recurrent grassland management program should be considered for lowelevation grasslands to reduce fuels on reserve land, private property, at fishing
sites, in fields that are no longer being worked and around the camping areas. Some
of these grassland areas may be good target areas to reintroduce prescribed burning
because their fuel loads are more manageable and less risk of escape than the forest.
In addition to prescribed burning, mowing or grazing could be used to reduce these
fuels, but all three will need to occur regularly to be effective.

118

Figure 47: Areas recommended for targeted fuel management.

Continued Forest Thinning.
The forest thinning undertaken through eco-cultural restoration treatments
should continue and new areas can be built into a landscape patchwork, targeting
forest areas where ignition is more likely such as on the west side of the lakes and in
the forests around the campgrounds and the areas within 500 meters of roads
(Figure 47, rust colour). The ability for the landscape to absorb and withstand
wildfire can be enhanced by creating greater patchiness in the forest by breaking up
119

the continuous fuel as well as building connectivity between surrounding
ecosystems that resist the spread of fire such as cliffs, boulder fields and lakes
(Figure 48). The densities and spacing required in these management areas to
sufficiently reduce the fuel risk is beyond the scope of this project; consultation with
the local Lillooet fire crew or other wildfire professionals is recommended.

Figure 48: Example of natural fire-resistant ecosystems that could be tied into other eco-cultural
restoration activities to create larger fire management areas (Green, S. 2017).

Creation of Fire Management Areas.
Workshop participants identified specific management options that should be
considered for wildfire risk mitigation. The features drawn in Figure 47 are only
intended to note the suggested location, not the actual spatial configuration on the
ground. Maintaining a strip of land with reduced fuels through mowing or plowing
along the edge of the fields at the north end of the valley would help prevent the
120

spread of grassfires from the roads, railway and fishing areas up into the valley with
the winds from the Fraser River (Figure 47, A). The creation of fire management
features through the forest near the south end of the Survival Territory was
suggested (Figure 47, C) with a few options of where it would protect houses and tie
together existing open forest types and natural fire-resistant ecosystems. It could be
supplemented with another fire management area farther up the valley (Figure 47 B)
that connects and reinforces low fuel areas on the valley slopes and provides another
break to prevent wildfires from taking off through the whole valley. These potential
fire management areas could be integrated into the Ecosystem Based Conservation
Plan.
Community Communication and Monitoring.
Creating a more fire absorbent landscape, particularly near populated areas,
will increase community safety but should be complemented by the creation of an
evacuation plan in case of an emergency. Though most of the recommendations of
this work are tailored to fit within the landscape level eco-cultural restoration plan, a
FireSmart inventory (British Columbia FireSmart Committee, 2020) to assess the
current risk to infrastructure is recommended to bring greater detail to the fine scale
around reserve land and housing areas. Continuing efforts to reduce human
ignitions through patrols and information is also necessary, with possible benefits to
eliminating or reducing access on side roads during high-risk weather. The presence
121

of the Range Riders and the XCF forest crew out on the land as responsible and
informed community members is an essential component to monitoring activities
throughout the valley and fostering a conversation about wildfire risk. To
complement these efforts, education on fuel types and wildfire risk for the
community through presentations and pamphlets can be delivered in the lead-up to
the wildfire season to proactively inform and prepare the community.
Conclusion
The physiographic, biogeographic, and community data identified some
areas of high-risk in the XCF, but with a high degree of uncertainty. Further research
is recommended to refine the fuel dataset to enable the use of fire models to assess
the efficacy of fuel treatments and plan for wildfire in the Xáxli’p Survival Territory.
Additionally, a significant lack of information on the localized wind patterns is a
prime opportunity for collecting baseline data that could be useful for many
applications.

122

Bibliography
Absolon, K., & Willett, C. (2005). Chapter 4: Putting ourselves forward: location in
Aboriginal research. In L. A. Brown & S. Strega (Eds.), Research As Resistance:
Critical, Indigenous and Anti-oppressive Approaches (pp. 97–125). Retrieved from
http://ezproxy.auckland.ac.nz/login?url=http://search.ebscohost.com/login.aspx?direct
=true&db=nlebk&AN=145914&site=ehost-live&scope=site&ebv=EB&ppid=pp_97
Alexander, M., Cruz, M., Vaillant, N., & L. Peterson, D. (2013). Crown fire behavior
characteristics and prediction in conifer forests: A state-of-knowledge synthesis.
Boise, Idaho: Joint Fire Science Program.
Alvarez, A., Gracia, M., Castellnou, M., & Retana, J. (2013). Variables That Influence
Changes in Fire Severity and Their Relationship with Changes Between Surface and
Crown Fires in a Wind-Driven Wildfire. Forest Science, 59(2), 139–150.
https://doi.org/10.5849/forsci.10-140
BC Forest Service. (2018). The Home Owners Fire Smart Manual: B.C. Edition. Retrieved
March 8, 2018, from https://www2.gov.bc.ca/assets/gov/public-safety-andemergency-services/emergency-preparedness-responserecovery/embc/preparedbc/homeowner-firesmart.pdf
BC Wildfire Service. (2015). Provincial Strategic Threat Analysis 2015 Wildfire Threat
Analysis Component. Retrieved from http://www2.gov.bc.ca/assets/gov/farmingnatural-resources-and-industry/forestry/wildfire-management/fire-fuelmanagement/bcws_provincial_strategic_threat_analysis_psta_2015_report.pdf
Boyd, R. (1999). Indians, fire, and the land in the Pacific Northwest (1st ed). Corvallis,
Oregon: Oregon State University Press.
123

British Columbia FireSmart Committee. (2020). FireSmart BC. Retrieved April 4, 2020, from
FireSmart BC website: https://firesmartbc.ca/
Cáceres, M. D., Martínez-Vilalta, J., Coll, L., Llorens, P., Casals, P., Poyatos, R., … Brotons,
L. (2015). Coupling a water balance model with forest inventory data to predict
drought stress: The role of forest structural changes vs. climate changes. Agricultural
and Forest Meteorology, 213, 77–90. https://doi.org/10.1016/j.agrformet.2015.06.012
Canadian Interagency Forest Fire Center. (2002). The 2002 Glossary of Forest Fire
Management Terms. Retrieved from
http://www.env.gov.bc.ca/esd/fire_mgmt_gloss_2002.pdf
Cape York Elders. (2013). Fire and the Story of Burning Country (Peter McConchie, Ed.).
Retrieved from https://books.google.ca/books?id=vM2NrgEACAAJ
Carroll, M. S., Cohn, P. J., Paveglio, T. B., Drader, D. R., & Jakes, P. J. (2010). Fire Burners
to Firefighters: The Nez Perce and Fire. Journal of Forestry, 108(2), 71–76.
Castellano, M. B. (2004). Ethics of Aboriginal Research. Journal of Aboriginal Health;
Victoria, 1(1), 98–114.
Coghlan, D., & Brydon-Miller, M. (Eds.). (2014). The Sage encyclopedia of action research.
Los Angeles: SAGE Reference.
Collins, K. M., Penman, T. D., & Price, O. F. (2016). Some Wildfire Ignition Causes Pose
More Risk of Destroying Houses than Others. PLOS ONE, 11(9).
https://doi.org/10.1371/journal.pone.0162083
Diceman, J. (2010). Dotmocracy Handbook. Retrieved from
http://library.uniteddiversity.coop/Decision_Making_and_Democracy/dotmocracy_ha
ndbook.pdf

124

Diver, S. (2016). Community Voices, The Making and Meaning of the Xáxli’p Community
Forest. A Report to the Xáxli’p Community Forest. Retrieved from
http://www.xcfc.ca/media/Community_Voices_Final_Report_2016_Full.pdf
Drake-Terry, J. (1989). The same as yesterday: The Lillooet chronicle the theft of their lands
and resources. Lillooet, B.C: Lillooet Tribal Council.
Forestry Canada. (1992). Development and Structure of the Canadian Forest Fire Behavior
Prediction System (Vol. 3). Retrieved from
https://cfs.nrcan.gc.ca/publications?id=10068
Forthofer, J., Butler, B., & Wagenbrenner, N. (2018). Wind Ninja (Version 3.3.1). Missoula
MT: Rockey Mountain Research Station Fire Sciences Laboratory.
Government of Canada, I. A. P. on R. E. (2016). Interagency Advisory Panel on Research
Ethics. Retrieved April 5, 2018, from http://www.pre.ethics.gc.ca/eng/policypolitique/initiatives/tcps2-eptc2/chapter9-chapitre9/
Gray, R. W., Andrew, B., Blackwell, A., Needoba, A., & Steele, F. (2002). The Effect of
Physiography and Topography on Fire Regimes and Forest Communities [Draft].
Retrieved from Habitat Conservation Trust Fund website:
https://www.for.gov.bc.ca/hfd/library/FIA/2002/FIA2002MR009.pdf
Hammond, H. (2009). Maintaining Whole Systems on Earth’s Crown: Ecosystem-based
Conservation Planning for the Boreal Forest. Silva Forest Foundation.
Hessburg, P. F., Spies, T. A., Perry, D. A., Skinner, C. N., Taylor, A. H., Brown, P. M., …
Riegel, G. (2016). Tamm Review: Management of mixed-severity fire regime forests
in Oregon, Washington, and Northern California. Forest Ecology and Management,
366, 221–250. https://doi.org/10.1016/j.foreco.2016.01.034

125

Indigenous and Northern Affairs Canada. (2019, July). Xaxli’p Registered Population.
Retrieved August 14, 2019, from First Nations Profiles website: http://fnp-ppn.aandcaadnc.gc.ca/fnp/Main/Search/FNRegPopulation.aspx?BAND_NUMBER=592&lang=
eng
Kalies, E. L., & Yocom Kent, L. L. (2016). Tamm Review: Are fuel treatments effective at
achieving ecological and social objectives? A systematic review. Forest Ecology and
Management, 375, 84–95. https://doi.org/10.1016/j.foreco.2016.05.021
Kimmerer, R. W. (2000). Native Knowledge for Native Ecosystems. Journal of Forestry,
98(8), 4–9.
Kimmerer, R. W., & Lake, F. K. (2001). The Role of Indigenous Burning in Land
Management. Journal of Forestry, 99(11), 36–41.
King, T. (2012). The Inconvenient Indian: A Curious Account of Native People in North
America (First edit). Toronto: Doubleday Canada.
Klenner, W., Walton, R., Arsenault, A., & Kremsater, L. (2008). Dry forests in the Southern
Interior of British Columbia: Historic disturbances and implications for restoration
and management. 6th North American Forest Ecology Workshop: From Science to
Sustainability, 256(10), 1711–1722. https://doi.org/10.1016/j.foreco.2008.02.047
Lecina-Diaz, J., Alvarez, A., & Retana, J. (2014). Extreme Fire Severity Patterns in
Topographic, Convective and Wind-Driven Historical Wildfires of Mediterranean
Pine Forests. PLOS ONE, 9(1), e85127. https://doi.org/10.1371/journal.pone.0085127
Leopold, A. (1949). The Land Ethic. In A Sand County Almanac: Sketches Here and There.
United States: Oxford University Press.
Leverkus, S., Fuhlendorf, S. D., Geertsema, M., Elmore, R. D., Engle, D. M., & Baum, K. A.
(2017). A Landscape Disturbance Matrix for Conserving Biodiversity. Journal of
126

Ecosystems and Management, 17(1). Retrieved from http://jemonline.org/index.php/jem/article/view/591
Lewis, M., Christianson, A., & Spinks, M. (2018). Return to Flame: Reasons for Burning in
Lytton First Nation, British Columbia. Journal of Forestry, 116(2), 143–150.
Meyn, A., Schmidtlein, S., Taylor, S. W., Girardin, M. P., Thonicke, K., & Cramer, W. (2013).
Precipitation-driven decrease in wildfires in British Columbia. Regional
Environmental Change, 13(1), 165–177. https://doi.org/10.1007/s10113-012-0319-0
Ministry of Environment and Climate Change Strategy - Knowledge Management. (2019).
Ecosections—Ecoregion Ecosystem Classification of British Columbia [Data
Warehouse]. Retrieved June 19, 2019, from BC Data Catalogue website:
https://catalogue.data.gov.bc.ca/dataset/ecosections-ecoregion-ecosystemclassification-of-british-columbia
Ministry of Forests Lands and Natural Resource Operations. (2018). Forest Inventory—
Province of British Columbia. Retrieved May 16, 2019, from
https://www2.gov.bc.ca/gov/content/industry/forestry/managing-our-forestresources/forest-inventory
Ministry of Forests, Lands, Natural Resource Operations and Rural Development - BC
Wildfire Service. (2018). Fire Incident Locations—Historical [Data Warehouse].
Retrieved from BC Data Catalogue website:
https://catalogue.data.gov.bc.ca/dataset/fire-incident-locations-historical
Ministry of Forests, Lands, Natural Resource Operations and Rural Development - BC
Wildfire Service. (2019). Fire Perimeters—Historical—Data Catalogue [Data
Warehouse]. Retrieved June 19, 2019, from BC Data Catalogue website:
https://catalogue.data.gov.bc.ca/dataset/fire-perimeters-historical
127

Ministry of Forests, Lands, Natural Resource Operations and Rural Development - Forest
Analysis and Inventory. (2017). VRI - Forest Vegetation Composite Polygons and
Layer 1—Data Catalogue [Data Warehouse]. Retrieved from BC Data Catalogue
website: https://catalogue.data.gov.bc.ca/dataset/vri-forest-vegetation-compositepolygons-and-layer-1
Ministry of Forests, Lands, Natural Resource Operations and Rural Development - GeoBC.
(2017). Digital Road Atlas (DRA)—Master Partially-Attributed Roads—Data
Catalogue. Retrieved February 5, 2020, from BC Data Catalogue website:
https://catalogue.data.gov.bc.ca/dataset/digital-road-atlas-dra-master-partiallyattributed-roads
Minkler, M. (2004). Ethical Challenges for the “Outside” Researcher in Community-Based
Participatory Research. Health Education & Behavior, 31(6), 684–697.
https://doi.org/10.1177/1090198104269566
Mistry, J., Bilbao, B. A., & Berardi, A. (2016). Developing a case for integrating Indigenous
fire practices within government fire management policy—Working paper 1.
Retrieved from http://projectcobra.org/wp-content/uploads/Working-Paper-1Indigenous-fire-management_English.pdf
Moritz, M. A., Moody, T. J., Krawchuk, M. A., Hughes, M., & Hall, A. (2010). Spatial
variation in extreme winds predicts large wildfire locations in chaparral ecosystems.
Geophysical Research Letters, 37(4), L04801. https://doi.org/10.1029/2009GL041735
Nitschke, C. R., & Innes, J. L. (2008). Climatic change and fire potential in South-Central
British Columbia, Canada. Global Change Biology, 14(4), 841–855.
https://doi.org/10.1111/j.1365-2486.2007.01517.x

128

Ohlson, D. W., Berry, T. M., Gray, R. W., Blackwell, B. A., & Hawkes, B. C. (2006). Multiattribute evaluation of landscape-level fuel management to reduce wildfire risk.
Forest Policy and Economics, 8(8), 824–837.
https://doi.org/10.1016/j.forpol.2005.01.001
Pacific Climate Impacts Consortium. (2019, July). BC Station Data—Provincial Climate
Data Set. Retrieved July 9, 2019, from PCIC Data Portal version 2.7.0 website:
https://data.pacificclimate.org/portal/pcds/map/
Parisien, M.-A., Kafka, V., Hirsch, K. G., Todd, J. B., Lavoie, S. G., & Maczek, P. D. (2005).
Mapping wildfire susceptibility with the BURN-P3 simulation model. 405. Retrieved
from https://cfs.nrcan.gc.ca/publications?id=25627
Parisien, M.-A., Walker, G. R., Little, J. M., Simpson, B. N., & Wang, X. (2012).
Considerations for modeling burn probability across landscapes with steep
environmental gradients: An example from the Columbia Mountains, Canada.
https://doi.org/10.1007/s11069-012-0495-8
Perrakis, D., & Eade, G. (2016). British Columbia wildfire fuel typing and fuel type layer
description. https://doi.org/10.13140/RG.2.1.2622.7288/1
Prichard, S., L. Peterson, D., & Jacobson, K. (2010). Fuel treatments reduce the severity of
wildfire effects in dry mixed conifer forest, Washington, USA. Canadian Journal of
Forest Research, 40. https://doi.org/10.1139/X10-109
Proctor, W. (2010). Integrated Mission-directed Research. Retrieved from
http://www.publish.csiro.au/book/6353/
Province of British Columbia. (2011). Fountain Valley Fire Management Strategies in Fuel
Managed Areas.

129

Province of British Columbia. (2012). A guide to fuel hazard assessment and abatement in
British Columbia. Retrieved from
http://www.llbc.leg.bc.ca/public/pubdocs/bcdocs2017/605032/bcws hazard
assessment abatement guide.pdf
Pyne, S. J. (2008). Awful Splendour: A Fire History of Canada. Vancouver B.C.: UBC Press.
Ray, L., Kolden, C., & Chapin III, F. (2012). A Case for Developing Place-Based Fire
Management Strategies from Traditional Ecological Knowledge. Ecology and
Society, 17(3). https://doi.org/10.5751/ES-05070-170337
Raymond, C. L., & Peterson, D. L. (2005). Fuel treatments alter the effects of wildfire in a
mixed-evergreen forest, Oregon, USA. Canadian Journal of Forest Research. 35:
2981-2995. Retrieved from https://www.fs.usda.gov/treesearch/pubs/24571
Rix, E. F., Wilson, S., Sheehan, N., & Tujague, N. (2018). Indigenist and Decolonizing
Research Methodology. In P. Liamputtong (Ed.), Handbook of Research Methods in
Health Social Sciences (pp. 1–15). https://doi.org/10.1007/978-981-10-2779-6_69-1
Rogeau, M.-P., & Armstrong, G. W. (2017). Quantifying the effect of elevation and aspect on
fire return intervals in the Canadian Rocky Mountains. Forest Ecology and
Management, 384, 248–261. https://doi.org/10.1016/j.foreco.2016.10.035
Rogeau, M.-P., Parisien, M.-A., & Flannigan, M. D. (2016). Fire History Sampling Strategy
of Fire Intervals Associated with Mixed- to Full-Severity Fires in Southern Alberta,
Canada. Forest Science, 62(6), 613–622. https://doi.org/10.5849/forsci.15-053
Scott, J. H., & Reinhardt, E. D. (2001). Assessing Crown Fire Potential by Linking Models of
Surface and Crown Fire Behavior (p. 59) [Res. Pap. RMRS-RP-29]. Fort Collins,
CO: U.S. Department of Agriculture, Forest Service, Rocky Mountain Research
Station.
130

Silva Forest Foundation. (2011). Restoration Ecotypes. Xaxli’p Community Forest
Corporation.
Smith, L. (1999). Decolonizing Methodologies: Research and Indigenous Peoples. Zed
Books.
Smith, T. (1998). Our Stories Are Written on the Land: A Brief History of the Upper St’at’imc
1800-1940. Lillooet, B.C: Upper St’at’imc Language Culture and Education Society.
Stephens, S. L., North, M. P., & Collins, B. M. (2015). Large Wildfires in Forests: What Can
be Done? Action BioScience. Retrieved from
http://www.actionbioscience.org/environment/large_wildfires_in_forests_what_can_b
e_done.html
Taylor, S. W., & Alexander, M. E. (2006). Science, technology, and human factors in fire
danger rating: The Canadian experience. International Journal of Wildland Fire,
15(1), 121–135. https://doi.org/10.1071/WF05021
Taylor, S. W., Pike, R. G., & Alexander, M. E. (1996). Field Guide to the Canadian forest fire
behaviour prediction (FBP) system (First edition). Retrieved from
https://cfs.nrcan.gc.ca/publications?id=4321
Taylor, S. W., Pike, R. G., & Alexander, M. E. (2000). Retrieved from
http://site.ebrary.com/lib/alltitles/docDetail.action?docID=10125098
Thompson, M. P., & Calkin, D. E. (2011). Uncertainty and risk in wildland fire management:
A review. Journal of Environmental Management, 92(8), 1895–1909.
https://doi.org/10.1016/j.jenvman.2011.03.015
Truth and Reconciliation Commission of Canada. (2015). Honouring the Truth, Reconciling
for the Future: Summary of the Final Report of the Truth and Reconciliation
Commission of Canada [Summary]. Retrieved from
131

http://www.trc.ca/websites/trcinstitution/File/2015/Honouring_the_Truth_Reconciling
_for_the_Future_July_23_2015.pdf
Turner, J. A., & Lawson, B. D. (1978). Weather in the Canadian forest fire danger rating
system. A user guide to national standards and practices. 177. Retrieved from
https://cfs.nrcan.gc.ca/publications?id=1843
United Nations University. (2009). Traditional indigenous fire management techniques
deployed against climate change. Retrieved from
https://www.sciencedaily.com/releases/2009/11/091129125133.htm
Wang, X., Parisien, M.-A., Taylor, S. W., Perrakis, D., Little, J., & Flannigan, M. (2016).
Future burn probability in south-central British Columbia. International Journal of
Wildland Fire, 25. https://doi.org/10.1071/WF15091
Weber, M. G., & Taylor, S. W. (1992). The use of prescribed fire in the management of
Canada’s forested lands. The Forestry Chronicle, 68(3), 324–334.
https://doi.org/10.5558/tfc68324-3
Weinstein, M. (1995). Thinking Like an Inhabitant: Understanding Xáxli’p Environmental
Values. Prepared for Xáxli’p Nation - BC Government Joint Land Use Planning
Workshop.
Williams, J. (2013). Exploring the onset of high-impact mega-fires through a forest land
management prism. Forest Ecology and Management, 294, 4–10.
https://doi.org/10.1016/j.foreco.2012.06.030
Wilson, S. (2013). Using Indigenist research to shape our future. In Decolonizing social work
(M Grey, J, pp. 311–322). Burlington, VT: Ashgate Publishing.

132

Wyatt, S. (2008). First Nations, forest lands, and “aboriginal forestry” in Canada: From
exclusion to comanagement and beyond. Canadian Journal of Forest Research,
38(2), 171–180.
Xáxli’p. (2017). Xáxli’p About Us. Retrieved April 4, 2017, from
http://www.xaxlip.ca/about.html
Xáxli’p Community Forest. (2013). Xáxli’p Community Forest Short and Long Term Goals.
Retrieved from
http://www.xcfc.ca/media/documents/XCFC_Short__Long_Term_Goals_2013-18.pdf
Xáxli’p Community Forest. (2017). Eco-cultural Restoration. Retrieved from Xáxli’p
Community Forest website: http://www.xcfc.ca/pages/management/eco-culturalrestoration.php
Xáxli’p Community Forest. (2019). Ecosystem-based planning. Retrieved from Xáxli’p
Community Forest website: https://www.xcfc.ca/traditional-use-study

133