ENVIRONMENT, CLIMATE AND TREE GROWTH RELATIONSHIPS AT THE
WESTERN CANADIAN ARCTIC TREELINE
by
Sean P. Sweeney
B.Sc. University of Montana, 2006

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

UNIVERSITY OF NORTHERN BRITISH COLUMBIA
May 2011
©Sean P. Sweeney, 2011

1*1

Library and Archives
Canada

Bibliotheque et
Archives Canada

Published Heritage
Branch

Direction du
Patrimoine de I'edition

395 Wellington Street
Ottawa ON K1A 0N4
Canada

395, rue Wellington
Ottawa ON K1A 0N4
Canada
Your file Votre reference
ISBN: 978-0-494-75170-1
Our file Notre reference
ISBN: 978-0-494-75170-1

NOTICE:

AVIS:

The author has granted a nonexclusive license allowing Library and
Archives Canada to reproduce,
publish, archive, preserve, conserve,
communicate to the public by
telecommunication or on the Internet,
loan, distribute and sell theses
worldwide, for commercial or noncommercial purposes, in microform,
paper, electronic and/or any other
formats.

L'auteur a accorde une licence non exclusive
permettant a la Bibliotheque et Archives
Canada de reproduire, publier, archiver,
sauvegarder, conserver, transmettre au public
par telecommunication ou par I'lnternet, preter,
distribuer et vendre des theses partout dans le
monde, a des fins commerciales ou autres, sur
support microforme, papier, electronique et/ou
autres formats.

The author retains copyright
ownership and moral rights in this
thesis. Neither the thesis nor
substantial extracts from it may be
printed or otherwise reproduced
without the author's permission.

L'auteur conserve la propriete du droit d'auteur
et des droits moraux qui protege cette these. Ni
la these ni des extraits substantiels de celle-ci
ne doivent etre im primes ou autrement
reproduits sans son autorisation.

In compliance with the Canadian
Privacy Act some supporting forms
may have been removed from this
thesis.

Conformement a la loi canadienne sur la
protection de la vie privee, quelques
formulaires secondaires ont ete enleves de
cette these.

While these forms may be included
in the document page count, their
removal does not represent any loss
of content from the thesis.

Bien que ces formulaires aient inclus dans
la pagination, il n'y aura aucun contenu
manquant.

1*1

Canada

ABSTRACT
The latitudinal forest-tundra ecotone is an area that is experiencing substantial changes with
respect to tree growth and climate change. We examined the response of radial tree growth to
climate in adjacent regions of northern Yukon and Northwest Territories, Canada, across
environmental and spatial gradients using dendrochrono logical methods. Principal
components analysis was used to derive the primary modes of variation in the tree-ring
records, which were subsequently attributed to environmental and climatic features. We
found that slope gradient (small spatial scales) and ecoregional classification (larger spatial
scales) played substantial roles in determining the response of tree growth to climate. Climate
correlations were found for current and previous years to growth, many of which challenge
currently held assumptions regarding the dominant climatic determinants of tree growth at
high latitudes. These findings indicate that Arctic forest environments are highly complex,
and that expected changes in the biosphere will occur at various rates, times and places.

11

TABLE OF CONTENTS
Abstract

ii

Table of Contents

hi

List of Tables

vi

List of Figures

vii

Dedication

vii

Acknowledgements

ix

Chapter 1: Introduction
1.1 Climate change in the Arctic
1.1.1 Temperature
1.1.2 Precipitation
1.1.3 Topography
1.1.4 Soil moisture balance
1.2 Climate change and sub-Arctic forests
1.2.1 Treeline and the forest-tundra ecotone
1.2.2 White spruce & black spruce
1.2.3 Climate-growth relationships of spruce
1.2.3.1 Temperature
1.2.3.2 Precipitation
1.2.4 Environment of the study region in northwest Canada
1.3 Research objectives
1.4 Literature cited

1
2
3
4
5
6
7
7
10
11
11
13
14
15
18

in

Chapter 2: The effect of environmental variation on the radial growth response of spruce at
the western Canadian sub-Arctic treeline
31
2.1 Introduction
32
2.2 Methods and materials
35
2.2.1 Field sites
35
2.2.1.1 Field site locations
36
2.2.2 Ecoregional classification and descriptions
36
2.2.2.1 Environment of Yukon
36
2.2.2.2 Environment of NWT
37
2.2.3 Sampling strategy
40
2.2.4 Chronology development
42
2.2.5 Data analysis
43
2.2.5.1 Chronology analysis
43
2.2.5.2 Multivariate analysis
43
2.3 Results
44
2.3.1 Chronologies
44
2.3.2 Multivariate analysis
45
2.3.2.1 Yukon principal components analysis
46
2.3.2.2 NWT principal components analysis
46
2.4 Discussion
47
2.4.1 Yukon
48
2.4.2 NWT
50
2.4.3 Conclusion
52
2.5 Literature Cited
54
Appendix A. Yukon site data
70
Appendix B. NWT site data
71
Appendix C. Soil, vegetation, and ground cover data
72
Appendix D. Chronology summary data
73
Appendix E. Tree growth indices derived from PCA
74

IV

Chapter 3: Scale-dependant climate and tree growth relationships at the western Canadian
subarctic treeline
75
3.1 Introduction
76
3.1.1 Climate-growth relationships of spruce
78
3.1.1.1 Temperature
78
3.1.1.2 Precipitation
80
3.2 Methods and materials
81
3.2.1 Field sites
81
3.2.1.1 Site Locations
82
3.2.2 Ecoregional classification and descriptions
82
3.2.3 Sampling strategy
82
3.2.4 Chronology development
83
3.2.5 Data analysis
83
3.2.5.1 Climate data
83
3.2.5.2 Climate-growth analysis
84
3.3 Results
84
3.3.1 Climate data
85
3.3.2 Climate-growth relationships
86
3.3.2.1 Yukon
86
3.3.2.2 NWT
87
3.4 Discussion
88
3.4.1 Climate and tree growth in Yukon
88
3.4.2 Climate and tree growth in the NWT
92
3.4.3 Conclusion
96
3.5 Literature cited
100
Chapter 4: Summary
4.1 Introduction
4.2 Key results and implications
4.3 Research limits and future research
4.4 Conclusions

114
114
114
118
121

v

LIST OF TABLES
Table 2.1 Site descriptions and statistics

59

Table 2.2 Site and chronology correlations

60

Table 2.3 Principal components analysis results of entire study area

61

Table 2.4 Summary of regional principal components analysis

62

Table 2.5 Yukon PCA loadings and site characteristics

63

Table 2.6 PCA groups and ecological characteristics

63

Table 3.1 Climate station monthly mean temperature correlations

108

Table 3.2 Seasonal climate & PCA correlations

109

vi

LIST OF FIGURES
Figure 2.1 Location of study area in northwest Canada

64

Figure 2.2 Site locations in Yukon and Northwest Territories

65

Figure 2.3 Ecoregions and PCA groups

66

Figure 2.4 Mean regional temperatures

67

Figure 2.5 Mean regional annual precipitation

68

Figure 2.6 Typical Yukon white spruce stands

69

Figure 2.7 Typical NWT site conditions

69

Figure 3.1 Current and prior year monthly climate correlations

110

Figure 3.2 Temperature and continentality trends in Yukon study area

111

Figure 3.3 Concurrent river flows and temperature trends in the NWT

112

Figure 3.4 Climate trends along the Dempster Highway, NWT

113

vn

DEDICATION
This thesis is dedicated to the memory of my father, Dan P. Sweeney, who taught me to love
and value the forest, and to seek wisdom and solace amongst the trees.

via

ACKNOWLEDGEMENTS
I would like to thank my supervisor Dr. Scott Green for giving me the opportunity to fulfill a
long-time dream to study and work in the far North. His patience, guidance, and support
guaranteed a productive and meaningful research experience. I would also like to thank my
committee members, Dr. Paul Sanborn and Dr. Aynslie Ogden, for their thoughtful input and
pragmatic advice throughout the course of my studies. Many thanks to those who helped me
in the field and lab: Joel Hanthorn, Nicole Winstanley, Robin Chang, Alyson Watt, and Kate
Snow. Also, thanks to Hardy Griesbauer for advice on statistical analyses, Dr. Stephen Dery
for advice on climate data, and Ben Sweeney for objective comments and review of this
document. Lastly, special thanks to the Hanthorn family of Ft. McPherson, NT, for their
gracious accommodation, logistical support, and genuine interest in my activities.

This study was respectfully conducted on the traditional lands of the Tetlit and Gwichya
Gwich'in First Nations with their consent.

This project was funded by a grant from the International Polar Year (IPY) Canada.

A Scientists and Explorers license was granted for research conducted in the Yukon, pursuant
to the provisions of the Yukon Scientists and Explorers Act of 1958. A Scientific Research
License was granted by the Aurora Research Institute for research conducted in the
Northwest Territories.

IX

1. Introduction
Tremendous ecological change is occurring in the circumpolar North, particularly along the
boreal forest's northern boundary, known as the forest-tundra (FT) ecotone. The northern
treeline is a diverse region where the vast boreal forest and the treeless tundra converge in a
delicate balance that is largely controlled by the prevailing climate. In the extreme
environment of the Arctic, slight changes in the climate can significantly disrupt the FT
ecotone by altering natural disturbance patterns, melting permafrost, and altering the
reproductive and growth capacity of trees at the northernmost extent of the boreal forest.
Currently, general observations through remote sensing and ground-based research have
indicated that the FT ecotone is already responding to climatic change in the Arctic,
including movement and densification of the treeline, expansion of shrubs onto formerly
barren tundra, and greater frequency and severity of wildfire and outbreaks of forest insects
and disease.

In an effort to provide a better understanding regarding ongoing and potential changes of the
FT ecotone, this study addressed the issue of radial tree growth at the forest's edge in
northwestern Canada, and how environmental variation can affect the growth response to
climate. This is a key issue as it relates to the carbon cycling capacity of the boreal forest, as
well as its contribution to warming temperatures via feedback mechanisms. This chapter
provides a review of the current state of knowledge regarding the historical and
contemporary context of climate change and tree growth in the Arctic, and identifies the
climatic and environmental features most likely to have significant impacts on the FT
ecotone. The following chapter (2) will explain how landscape variation can affect the
observed patterns of tree growth through varying spatial scales. Chapter 3 will then show
1

how the environmental features identified in Chapter 2 can influence which climate
parameters have the most impact on radial tree growth. Finally, the summary chapter (4) will
highlight the main findings and implications of this study, and will provide an assessment of
the main limitations and potential future work based on this research. As a whole, this thesis
is a contribution towards the efforts of the PPS (Present processes, Past changes, and
Spatiotemporal dynamics) Arctic Canada component of the International Polar Year (IPY)
2007-2008, which primarily focused on the effects and subsequent ecological and social
consequences of climate change on treeline position and structure.

1.1 Climate change in the Arctic
According to the Arctic Climate Impact Assessment (2005), warming temperatures in the
Arctic are expected to exert tremendous changes on vegetation, resulting in taller, denser, and
more abundant plant coverage. Exotic species may begin to find their way into the Arctic via
warming temperatures and increased human activities (Sala et al, 2000). Thawing
permafrost will pose significant adverse effects as well, by destabilising soils, increasing the
paludification of northern forests, significantly altering soil moisture levels, and releasing
carbon dioxide (CO2) into the atmosphere (Lloyd et al, 2003; Zimov et al, 2006;
Vygodskaya et al, 2007). These patterns are substantiated by local knowledge from
indigenous peoples in the Arctic, who frequently refer to unpredictable and changing weather
patterns, warmer and shorter winters, and shifting wind patterns (Kassi, 1993; McDonald et
al, 1997; Krupnik & Jolly, 2002). However, the patterns of ecological change in the Arctic
are and will be characterized by a great degree of spatial complexity, which will only
increase the challenge of predicting and managing for future changes. A better understanding
of the underlying processes behind recent and forthcoming changes, and how these processes
2

operate at different spatial scales, is needed if sufficient measures are to be taken to adapt to
and mitigate for future scenarios of circumpolar change.

1.1.1 Temperature
The multiproxy record (tree rings, lake sediment and ice cores) of the last 400 years
indicates that 20th century temperatures have been the warmest during this period (Overpeck
et al, 1997). Much evidence exists to show the Arctic warmed substantially from 1920 to
1940, and again beginning in the 1970s (Serreze et al, 2000). However, these patterns are
marked by regional differences; while these patterns hold true in the western North American
Arctic, eastern Canada and Greenland have experienced relatively stable, and occasionally
cooling, temperatures in the past century (Hinzman et al, 2005). The greatest degree of
warming has generally occurred during the winter and spring months (Chapman & Walsh,
1993; Serreze et al, 2000; Mbogga et al, 2009), and has been most pronounced in Alaska,
Siberia, and northwestern Canada (Maxwell, 1997; McBeam et al, 2005). In Inuvik,
Northwest Territories (NWT), average annual temperatures have increased approximately
2°C since the 1940s when instrumental measurements first began (Govt, of the NWT, 2008).
Satellite data have shown that warming spring temperatures have been attended by an
increase in net photosynthetic activity in some northern regions (Mynemi et al, 1997).
However, photosynthetic potential is also dependent upon soil moisture and nutrient
availability, so warm temperatures alone may not yield increased photosynthesis in many
scenarios.

Warming temperatures in the Arctic may also increase the length of the growing season by
creating earlier spring snow melt dates and extending the frost free period. Along with
3

temperature, growing season length is a major controlling factor on the stasis of the northern
treeline (MacDonald et al, 2008), as well for many other Arctic plants (Callaghan et al,
2004b). In areas with limited nutrient availability, growing season length may also have a
strong effect on the ability of Arctic plants to respond to warming temperatures (Walker et
al., 2006). Lengthened growing seasons typified by earlier melt dates may also impart
shortages of soil moisture towards the end of the growing season, a condition that may carry
over into the following year (Kirdyanov et al, 2003; Barnett et al, 2005).

1.1.2 Precipitation
Modeled and observed data have indicated the Arctic has generally experienced higher levels
of precipitation in the last century, which will likely continue to rise into the next century
(McBeam et al, 2005; Mbogga et al, 2009). This increase has come predominantly in the
form of winter snowfall (Maxwell, 1997; Kattsov & Walsh, 2000), due to increased
atmospheric vapour capacity resultant of concurrent warming air temperatures (Kattenberg et
al, 1996). This increasing precipitation trend is substantiated by concurrent records of
increased streamflows (Groisman et al, 2005; Trenberth et al, 2007). However, earlier
snowmelt dates and

decreased snow cover have been found throughout the northern

hemisphere, primarily due to increasing spring temperatures (Stone et al, 2002; Barry et al,
2007). Early snowmelt dates can increase spring soil temperature and affect permafrost
dynamics through heightened exposure to solar radiation, which in turn can lengthen the
effective growing season for many plants (Mynemi et al, 1997; Hinzman et al, 2005). In
high mountain environments, increased springtime soil moisture availability, due to earlier
snow melt dates, has been shown to be strongly positively tied to total growing season carbon
uptake as well (Schimel et al, 2002). Longer snow-free periods also reduce the annual
4

surface albedo, leading to increased atmospheric heating via feedback mechanisms (Serreze
et al, 2000; Chapin et al, 2005; Euskirchen et al, 2007). Furthermore, average snow depths
have decreased across Canada and some parts of Europe and western Russia (McBeam et al,
2005), although some opposing examples of increased snow depths and duration have been
found in much of northern Eurasia (Kitaev et al, 2005; Kohler et al, 2006) and some areas
of Yukon (A.E. Ogden, personal communication, March 1, 2011). Decreased snow depths
likely lead to lower soil temperatures that are accompanied by reduced net ecosystem
respiration and CO2 efflux in the winter (though winter efflux would be minimal even with
warmer soils) (Ling & Zhang, 2007; Morgner et al, 2010; Sullivan, 2010). The decreasing
trends in snow cover area and depth, and earlier melt dates, appear to be counterintuitive to
increased precipitation, which has been attributed to warming temperatures. However,
warming temperatures are likely to melt the thinner snowpack at the margins of snow
covered areas, resulting in decreased snow coverage and earlier melt dates in some places
(McBeam et al, 2005). As well, it is possible that more precipitation may periodically fall as
a liquid during the winter, which would potentially diminish snowpack depths. Though these
patterns remain relatively unclear on a global scale, decreases in snow coverage and depths
are predicted for the northern hemisphere, though some areas are likely to experience
opposite increases (Trenberth et al, 2007).

1.1.3 Topography
Confounding the patterns and behaviour of both precipitation and temperature and their
combined effect on soil moisture is the presence of mountainous topography, which can lead
to more localized climate variations that may not follow the general climate trends in the
Arctic (Pojar, 1996). At a local scale, soil moisture tends to be greater on the lower portion of
5

mountainside hill slopes simply due to the effects of gravity (Carey & Woo, 2001). On a
larger, regional scale in mountainous areas, orographic precipitation results as air rises and
cools along windward slopes, causing the air moisture to condense and form clouds and
precipitation at higher elevations (Bonan, 2008). Strong temperature inversions (a positive
(e.g. reverse) adiabatic lapse rate) in northern Yukon can occur in mountainous areas in calm
conditions (Wahl et al, 1987; Bonan, 2008). Mountain ranges also pose as barriers to oceanderived air masses and storms, creating continental (e.g. more extreme and variable) climates
(Wahl et al, 1987). Such is the case in northern Yukon, where the British-Richardson
Mountains effectively block much of the Arctic Ocean air masses, which are prevalent within
the adjacent Mackenzie delta region of the NWT.

1.1.4 Soil moisture balance
Soil moisture balance is highly dependent on the prevailing local temperature and
precipitation patterns, as well as the topography and geology of the landscape. Changes in
soil moisture availability can impart significant impacts on tree growth in cold climates
(Bunn et al, 2005). For example, decreased soil moisture levels have led to decreased growth
and greater mortality of white spruce in interior Alaska (Juday et al, 2005; McGuire et al,
2010). As a result of declining soil moisture availability, drought stress has been implicated
as an important factor influencing declining tree growth and increased forest disturbance in
northern boreal regions (Zhang et al, 2008; D'Arrigo et al, 2009), although recent evidence
from Alaska suggests that heightened evaporative demands may be the primary cause of
drought-induced growth limitations (Beck et al, 2011). Regardless of the exact mechanisms,
temperature, precipitation, and topography remain the primary sources affecting the growing
environment limitations of the northern treeline.
6

1.2 Climate change and sub-Arctic forests
1.2.1 Treeline and the forest-tundra ecotone
There are many definitions for the northern treeline, and many ways to delineate it. The socalled treeline (or krummholz-line) is commonly cited on large-scale maps as closely
following the 10°C July isotherm, but this does not well characterize tree distributions at
regional and local scales (Koppen, 1936; Halliday & Brown, 1943). The forest-line is
identified where the forest cover exceeds 50% of the land cover (Black & Bliss, 1978). For
the purposes of this study, the area of investigation is defined as the transitional zone
between continuous forest cover and open tundra (shrub or tussock), termed the forest-tundra
(FT) ecotone (Hare & Ritchie, 1972; Scott et al, 1997). The FT ecotone is typified by sparse
individual and aggregated trees, often reflecting the underlying hydrologic, topographic, and
reproductive (i.e. seed vs. vegetative layering) conditions (Arno, 1984; Scott et al, 1997).
Individual trees within the FT ecotone resemble those in the continuous forest, but only reach
about 50% of their height (Scott et al, 1993). At the furthest northern extent, trees assume
krummholz stature before giving way to shrubs and open tundra.

Polar-region treelines are highly sensitive to changes in climate (Hinzman et al, 2005), and
are expected to exhibit some of the earliest and most pronounced visible indications of a
changing climate (Payette et al, 2001; Grace et al, 2002; Juday et al, 2005). Equally, other
studies have indicated the position and density of the treeline can also affect global climate
patterns (Bonan & Sirois, 1992; Foley, 1994; Crawford, 2008; MacDonald et al, 2008). For
instance, greater vegetative density is darker than bare tundra, and thus absorbs more solar
radiation, which in turn can lead to warming of the atmosphere (Chapin et al, 2005). Early
7

responses in tree growth and recruitment, and a subsequent northward advance, densification,
or species alteration of the treeline, are expected to occur in response to changes in various
climate parameters (Rizzo & Wiken, 1992; Suarez et al, 1999; Holtmeier & Broil, 2005),
including increased temperatures and atmospheric CO2, changing precipitation and wind
patterns, and soil nitrification (Rizzo & Wiken, 1992; Lenihan & Neilson, 1995; Hartley et
al, 1999; Grace et al, 2002; Hinzman et al, 2005; Rees, 2007). In addition to climate, other
factors can affect northern treeline dynamics: competition from shrubs that react positively
and relatively more rapidly to warmer temperatures (Hobbie & Chapin, 1998); changes in
snowpack depths and duration (Holtmeier & Broil, 2005); and changing soil temperatures
and moisture levels (Black & Bliss, 1980; Wilmking et al, 2004).

Treeline movement is contingent on the establishment of new seedlings in areas previously
not conducive to successful tree recruitment and growth. The prevailing theory of advancing
treelines suggests that increasing temperatures will lead to more favourable growing
conditions for pioneer seedlings (Hobbie & Chapin, 1998). However, many other factors also
contribute to the success rate of new recruits, such as fire severity (Sirois, 1993) and thinner
post-fire soil organic layers (Greene et al, 2007) in existing forest stands, and below-ground
competition (Hobbie & Chapin, 1998) and growing degree-days thermal sum in both forested
and unforested sites (Sirois, 2000; Meunier et al, 2007). Current findings indicate that
treeline movement has not occurred uniformly across the northern hemisphere in recent
decades, though evidence of a recession of the treeline has not been prevalent (MacDonald et
al, 2008; Harsch et al, 2009). Recent treeline advance has been observed in Quebec
(Gamache & Payette, 2005; Caccianiga & Payette, 2006), northern Alaska (Suarez et al,
1999), the Kluane Range of Yukon (Danby & Hik, 2007), and the Polar Urals of Siberia
8

(Devi et al, 2008). Conversely, other studies have indicated a relatively stable treeline across
Canada, with little to no advancement in recent decades (Szeicz & MacDonald, 1995b;
Masek, 2001). However, the modern extent of the Arctic treeline has remained relatively
stable during both warm and cold periods of the late Holocene (2000-3000 years BP) (Lavoie
& Payette, 1996; Crawford, 2008).

Another component of treeline change is the densification of the FT ecotone by increased
tree and shrub recruitment (Chapin et al, 1995; Tape et al, 2006). Shrubs have been
particularly responsive to warming temperatures throughout the circumpolar region,
evidenced by increased physical size, density, and expansion onto tundra (Sturm et al, 2001;
Tape et al, 2006; Walker et al, 2006; Olthof & Pouliot, 2010). Soil thaw depth is a primary
control of arctic willow growth, and increasing thaw depths will likely yield more fervent
willow growth in high latitude areas (Pajunen, 2009).

Additionally, forest disturbance agents, namely fire and insects and disease, have always
been present in the boreal forest, and are natural components of normal forest ecosystem
processes. However, climate change may predispose the boreal forest to insect outbreaks of
greater scale, intensity, and frequency, due to a low diversity of host tree species and
insufficient cold temperatures to control insect populations (Juday et al, 2005). Instances of
severe outbreaks spruce budworm and spruce bark beetle have been recently documented in
Alaska (Werner, 1996; Juday et al, 2005), while Yukon is experiencing the worst spruce
bark beetle outbreak in Canada's history (Yukon Department of Energy, Mines, and
Resources, 2009). For these cases, warming temperatures are cited as the primary cause.
Severe cases of insect outbreaks reduce the carbon uptake capacity of the forest in the short
9

term, and may eventually cause shifts in tree species compositions and distributions, as new
species better adapted to warmer temperatures and heightened insect pressures migrate to the
affected areas. Wildfire is also innately connected to climate change, as continuing drought
will create conditions in the forest that highly are conducive to large, severe, and frequent
forest fires (Rupp et al, 2000). Fire intensity and frequency are highly important factors that
control the range and reproductive success of black spruce at the northern treeline (Greene et
al, 2007; Kasischke et al, 2007; Lloyd et al, 2007), as well as the successional trajectory of
northern spruce forests (Johnstone & Chapin, 2006).

The implications of changing patterns of tree growth in the FT ecotone will include a
shifting of the timing, location, and severity of disturbance regimes; alteration of the carbon
cycling capacity of the boreal forest; and a change in surface albedo and roughness, which in
turn could potentially further increase a warming trend (Bonan & Sirois, 1992; Foley, 1994;
Callaghan et al, 2005; Chapin et al, 2005; Juday et al, 2005). This is in addition to many
more indirect consequences, including declines in wildlife populations and altered migration
routes (Post & Forchhammer, 2008), changing traditional culture and livelihoods of
indigenous peoples (Ford & Smit, 2004), and loss of biodiversity. Therefore, it is crucial to
understand the potential reactions of the arctic treeline to climatic regime changes (Szeicz &
MacDonald, 1995b).

1.2.2 White spruce & black spruce
The most common tree species in the Canadian FT ecotone are white spruce {Picea glauca
(Moench) Voss) and black spruce {Picea mariana (Mill.) BSP), which were the subjects of
this study. Tamarack {Larix laricina (Du Roi) K. Koch) is the only other gymnosperm
10

present, and occurs sporadically throughout the NWT. Other prevalent species are paper
birch {Betula papyrifera), balsam poplar {Populus balsamifera), and trembling aspen
{Populus tremuloides). White spruce occupies well-drained upland slopes and floodplains,
where the permafrost is deep or non-existent (Yarie, 1983). Mature stands are associated with
well-developed moss layers, which regulate rooting zone temperatures and simultaneously
compete for nutrients (Nienstaedt & Zasada, 1990). In Alaska, trees growing along
floodplains do not correlate well with the climate record, and thus are not suitable sites for
climate-growth studies (Juday et al, 2003). White spruce is the near-exclusive species in the
Yukon FT ecotone, but occurs only intermittently in the NWT, as suitable sites become more
infrequent in the severe climates of northern latitudes (Nienstaedt & Zasada, 1990). Black
spruce typically occupies sites with poorly-drained, wet, and cold soils, and occasionally on
north aspects (Johnson et al, 1995). It is often present in pure dense stands on organic soils
(occasionally with tamarack), and assumes stunted postures with thin crowns (Viereck &
Johnston, 1990). Reproduction is either by asexual vegetative layering, or by stand-replacing
wildfire which produces the heat necessary to open the serotinous cones (Black & Bliss,
1980; Viereck & Johnston, 1990). It is the most common tree in the NWT FT ecotone. In
northern Yukon, it is the primary species within the continuous forest, before giving way
entirely to white spruce at the onset of the FT ecotone and our study area.

1.2.3 Climate-growth relationships of spruce
1.2.3.1 Temperature
Though summer temperature is regarded as the most influential climate component that
positively affects spruce growth at the northern treeline (Bonan & Sirois, 1992; Vaganov et
al, 1999), there is no single relationship that defines climate and tree growth (Crawford,
11

2008). Rather, these relationships vary accordingly to changing biophysical, genetic and
climatic conditions. For example, in central Canada, black spruce at the northern and
southern treelines were found to have pronounced negative correlations with summer
temperatures (Brooks et al, 1998), while sites at Ft. Wainwright, Alaska responded
positively to winter temperatures (Juday et al, 2005). Conversely, black spruce at the
northern treeline in Alaska exhibited positive responses to summer temperatures, while cooccurring white spruce showed an inverse relationship (Lloyd et al, 2005). These
relationships are all derived from the strongest significant correlations between climate and
tree ring parameters; it is possible other climate variables have some varying degrees of
effect as well. Though patterns of climate-tree growth relationships vary, drought stress has
been most frequently cited as a primary limitation to growth in the northern boreal forest
(Jacoby & D'Arrigo, 1997; Barber et al, 2000; Lloyd & Fastie, 2002; ACIA, 2005).

There is mounting evidence that some trees at the northern treeline are experiencing
declining positive correlations with growing-season temperatures. Wilmking et al. (2004)
found that approximately 40% of a large sample (n >1500) of white spruce in the Alaska and
Brooks ranges of Alaska responded negatively to summer temperatures (hereafter negative
responders), while less than 40% of the sample population still responded positively
(hereafter positive responders) in the latter half of the 20th century. In the Mackenzie River
delta, Pisaric et al. (2007) found that only 25% of their sample population of black spruce (n
= 654) showed a positive correlation with summer temperatures, with the remaining portion
beginning to lose its sensitivity in the 1930s. Similar findings have been reported for white
spruce in southwest Alaska (Driscoll et al, 2005), north-central Canada (D'Arrigo et al,
2009), and for larch in the Taymir region of northern Siberia (Jacoby et al, 2000). It appears
12

that the trees with declining positive summer temperature associations are exceeding certain
temperature thresholds beyond which they begin to alter the nature of their climatic growth
response (Wilmking et al, 2005). Specific threshold values are few, but estimates between
11° and 12°C (mean summer temperature) have been suggested for central Yukon and
Alaska (D'Arrigo et al, 2004; Wilmking et al, 2004).

This trend of divergent growth responses to summer temperature has come to be known as
the 'divergence problem' (Briffa et al, 1998; D'Arrigo et al, 2008), and has developed into a
significant area of interest in northern dendroclimatic studies (Lloyd & Bunn, 2007; Wilson
et al, 2007; Esper & Frank, 2009; Loehle, 2009). The direct cause has largely been attributed
to increasing drought stress (Barber et al, 2000; Davi et al, 2003; Wilmking et al, 2004;
Driscoll et al, 2005; Pisaric et al, 2007), though other local or indirect factors, such as
increased plant competition (Hogg & Hurdle, 1995), insect herbivory (Fleming & Volney,
1995), or increased UV-B levels (Callaghan et al, 2004a) could potentially play supporting
roles. It remains unknown why particular trees respond positively or negatively to warming
temperatures, and a general explanation is not likely to be construed.

1.2.3.2 Precipitation
There is a substantial lack of reliability and confidence in measured precipitation data in high
latitude regions due to instrumental measurement errors and an inability to capture and record
localized precipitation events. Significant correlations between precipitation and tree-ring
growth are not commonly found at high latitudes, as temperature generally emerges as the
dominant climatic factor. Where significant positive precipitation effects are found, summer
temperatures are often simultaneously found to exert negative effects. For example, during
13

the latter half of the 20 century in central Alaska, black spruce was found to be positively
correlated with August precipitation, accompanied by negative relationships with July and
August temperatures (Wilmking & Myers-Smith, 2008). Similarly, white spruce negative
responders (to summer temperatures) in the Mackenzie River delta reacted positively to April
precipitation (Pisaric et al, 2007). However, the effect of precipitation on growth is
ultimately determined by soil characteristics, topography, and rainfall distribution (Kljun et
al, 2007).

1.2.4 Environment of the study region in northwest Canada
Northwestern Canada, specifically northern Yukon and the northwestern corner of the
Northwest Territories, was chosen as the study region due to its relative lack of treeline
research, high degree of natural landscape and climate variability, and relatively easy access
via the Dempster Highway. Much of the treeline research in North America has been
conducted in Alaska and eastern Canada, and so our study area will also serve to bridge these
distant regions and provide a more complete coverage of the North American FT ecotone.
Only one limited tree growth/climate study has been conducted in this region of Yukon
(Szeicz & Macdonald, 1994), and it primarily focused on reconstructing historical
temperature records, without the knowledge of the now-identified 'divergence problem'.
Similarly, only a small amount of treeline research have been done in the adjacent NWT
region (e.g. Black & Bliss, 1980; Szeicz & MacDonald, 1996; Pisaric et al, 2007). This
study attempted to build upon these past investigations and provide a more integrated and
comprehensive assessment for this region of Canada.

14

Few long-term climate records exist in this region as well, similar to most high-latitude
regions. As such, regional climate trends are generally based on large-scale climate
oscillations and a few intermittent climate monitoring stations, in addition to empirical
evidence obtained from the landscape. However, this evidence exists on a large scale
throughout the circumpolar region, and it has been firmly established that the Arctic is
undergoing rapid climatic and environmental change, and is widely predicted to continue
experiencing changes with increasing complexity, frequency and scale (ACIA, 2005; IPCC,
2007;Harsch^a/.,2009).

1.3 Research objectives
Climate change is continually presenting new challenges for people and landscapes in the
Arctic. The Arctic is typified by a complex and varied assortment of environments that will
exhibit varied responses to climate change. This is also true within the FT ecotone and along
the circumpolar treeline, where variable plant and tree species, topography, permafrost
dynamics and climates will create multiple scenarios of potential changes. A migration or
densification of the treeline will play a large role in surface albedo feedbacks to the
atmosphere, as well as changing the patterns of snow accumulation, soil temperatures and
permafrost activity. Carbon cycling will also be a critical component of these changes, as the
boreal forest represents one of the earth's great carbon sinks, yet stands to become a source
as the landscape reacts to warming temperatures. Livelihoods of those dependent on the
northern boreal forest will be affected as well, as plant and wildlife resources become more
or less abundant as vegetation shifts.

In response to these serious implications of a rapidly

changing Arctic treeline, much research has been initiated in recent years to address the
potential effects and impacts in response to climate change. Though many important and
15

useful findings have been realized, there remain substantial gaps in our current knowledge
and understanding of the dynamics of a changing northern treeline. This study aims to
address some of these gaps, in an area that has yet to receive much attention on the global
Arctic research front.

The primary objective of this study was to determine the varying effects of climate on tree
growth at the northern latitudinal treeline, and the role that environmental variation has in
shaping the climate/tree growth interrelationship. A secondary objective was to identify
environmental conditions that may be contributing to disparate growth trends inherent to the
'divergence problem'.

Specific questions addressed by this study were:

1) How do climate-tree growth relationships change across an environmental gradient at the
northwestern Canadian treeline? We hypothesized that tree growth would react to climate
differently between and within the two study areas. For this question we assumed that the
adjacent Yukon and NWT study areas were distinct environments that experience different
climates, and would thus represent a viable environmental gradient. Also, we had to assume
that sampling along the Dempster Highway would allow us to identify a reasonable variety of
sites that would reliably represent the range of conditions throughout the study area.

2) What environmental characteristics affect the response of tree growth to climate? For this
question, we tested the hypothesis that tree growth would be dependent on particular
landscape characteristics unique to each study area, and relative to the spatial scale at which
16

the responses were observed. We assumed that by standardizing some of the site selection
criteria (e.g. aspect, topography, slope position, tree species), other important environmental
variables would become evident in our analysis. This required an effort to record as much
pertinent environmental data as possible, given the constraints of time, available equipment
and expertise, and scope of the study.

3) Which climatic variables are most influential on Arctic tree growth? We tested the
hypothesis that summer temperature would not be the primary controlling climate factor over
tree growth at all sites and spatial scales. For this question we assumed that climate variables
could be seasonally defined to provide a more general sense of the conditions in this area.
Also, we assumed that modelled data for this region would accurately reflect the general
trends and variability of the prevailing climate, and would serve as a better (i.e. longer,
complete, and locally representative) record than the available climate station data.

4) Does landscape variation appear to be a contributing factor to the 'divergence problem'?
We hypothesized that climatic factors attributed to disparate tree growth in previous studies
would be evident in our area based on landscape differences. We assumed that disparate tree
growth could be delineated using principal components analysis, rather than physically
categorizing individual tree-ring series. While this would have been a more thorough and
robust method, this question is not the central point of this thesis, and instead we sought to
provide a general assessment and starting point for further, more detailed investigations.

17

1.4 Literature Cited
ACIA. (2005). Arctic Climate Impact Assessment. Cambridge.
Arno, S.F. (1984). Timberline: Mountain and Arctic Forest Frontiers. Seattle: The
Mountaineers.
Barber, V.A., Juday, G.P., & Finney, B.P. (2000). Reduced growth of Alaskan white spruce
in the twentieth century from temperature-induced drought stress. Nature, 405, 668673.
Barnett, T.P., Adam, J.C., & Lettenmaier, D.P. (2005). Potential impacts of a warming
climate on water availability in snow-dominated regions. Nature, 438{7066), 303309.
Barry, R.G., Armstrong, R.L., Callaghan, T.V., Cherry, J., Gearhead, S., Nolin, A., et al.
(2007). Snow. In Global Outlookfor Ice and Snow: United Nations Environment
Programme.
Beck, P.S.A., Juday, G.P., Alix, C, Barber, V.A., Winslow, S.E., Sousa, E.E., et al. (2011).
Changes in forest productivity across Alaska consistent with biome shift. Ecology
Letters, doi: 10.1111/j. 1461-0248.2011.01598.x.
Black, R.A., & Bliss, L.C. (1978). Recovery sequence of Picea-mariana-Vacciniumuliginosum forests after burning near Inuvik, Northwest-Territories. Canadian
Journal of Botany, 56{\1), 2020.
Black, R.A., & Bliss, L.C. (1980). Reproductive ecology of Picea mariana (Mill.) BSP., at
tree line near Inuvik, Northwest Territories, Canada. Ecological Monographs, 50(3),
331-354.
Bonan, G.B., & Sirois, L. (1992). Air temperature, tree growth, and the northern and southern
range limits to Picea mariana. Journal of Vegetation Science, 3, 495-506.
Bonan, G.B. (2008). Ecological Climatology: Concepts and Applications (2nd ed.).
Cambridge: Cambridge University Press.

18

Briffa, K.R., Schweingruber, F.H., Jones, P.D., Osborn, T.J., Shiyatov, S.G., & Vaganov,
E.A. (1998). Reduced sensitivity of recent tree-growth to temperature at high northern
latitudes. Nature, 391, 678-682.
Brooks, J.R., Flanagan, L.B., & Ehleringer, J.R. (1998). Responses of boreal conifers to
climate fluctuations: indications from tree-ring widths and carbon isotope analyses.
Canadian Journal of Forest Research, 25(4), 524-533.
Bunn, A.G., Waggoner, L.A., & Graumlich, L.J. (2005). Topographic mediation of growth in
high elevation foxtail pine (Pinus balfouriana Grev. et Balf.) forests in the Sierra
Nevada, USA. Global Ecology andBiogeography, 14{2), 103-114.
Caccianiga, M., & Payette, S. (2006). Recent advance of white spruce (Picea glauca) in the
coastal tundra of the eastern shore of Hudson Bay (Quebec, Canada). Journal of
Biogeography, 33(12), 2120-2135.
Callaghan, T.V., Bjorn, L.O., Chernov, Y., Chapin, T., Christensen, T.R., Huntley, B., et al.
(2004a). Responses to projected changes in climate and UV-B at the species level.
Ambio, 33(1), 418-435.
Callaghan, T.V., Bjorn, L.O., Chernov, Y., Chapin, T., Christensen, T.R., Huntley, B., et al.
(2004b). Biodiversity, distributions and adaptations of arctic species in the context of
environmental change. Ambio, 33(1), 404-417.
Callaghan, T.V., Bjorn, L.O., Chapin, F.S., III, Chernov, Y., Torben, R.C., Huntley, B., et al.
(2005). Arctic tundra and polar desert ecosystems. In Arctic Climate Impact
Assessment (pp. 243-352). Cambridge: Cambridge University Press.
Carey, S.K., & Woo, M. (2001). Spatial variability of hillslope water balance, Wolf Creek
basin, subarctic Yukon. Hydrological Processes, 75(16), 3113-3132.
Chapin, F.S., III, Shaver, G.R., Giblin, A.E., Nadelhoffer, K.J., & Laundre, J.A. (1995).
Responses of Arctic tundra to experimental and observed changes in climate.
Ecology, 76(3), 694-711.
Chapin, F.S., III, Sturm, M., Serreze, M.C., McFadden, J.P., Key, J.R., Lloyd, A.H., et al.
(2005). Role of land-surface changes in Arctic summer warming. Science, 310(5748),
657-660.

19

Chapman, W.L., & Walsh, J.E. (1993). Recent variations of sea ice and air-temperature in
high-latitudes. Bulletin of the American Meteorological Society, 74(1), 33-47.
Crawford, R.M.M. (2008). Plants at the Margin: Ecological Limits and Climate Change.
Cambridge: Cambridge University Press.
Danby, R.K., & Hik, D.S. (2007). Evidence of recent treeline dynamics in southwest Yukon
from aerial photographs. Arctic, 60(A), 411-420.
D'Arrigo, R.D., Kaufman, R.K., Davi, N., Jacoby, G.C, Laskowski, C , Myneni, R.B., et al.
(2004). Thresholds for warming-induced growth decline at elevational tree line in the
Yukon Territory, Canada. Global Biogeochemical Cycles, 18(3), GB3021,
doi:3010.1029/2004GB002249.
D'Arrigo, R.D., Wilson, R., Liepert, B., & Cherubini, P. (2008). On the 'Divergence Problem'
in Northern Forests: A review of the tree-ring evidence and possible causes. Global
and Planetary Change, (50(3-4), 289-305.
D'Arrigo, R.D., Jacoby, G.C, Buckley, B., Sakulich, J., Frank, D., Wilson, R., et al. (2009).
Tree growth and inferred temperature variability at the North American Arctic
treeline. Global and Planetary Change, 65(1-2), 71-82.
Davi, N.K., Jacoby, G.C, & Wiles, G.C. (2003). Boreal temperature variability inferred from
maximum latewood density and tree-ring width data, Wrangell Mountain region,
Alaska. Quaternary Research, 60(3), 252-262.
Devi, N., Hagedorn, F., Moiseev, P., Bugmann, H., Shiyatov, S., Mazepa, V., et al. (2008).
Expanding forests and changing growth forms of Siberian larch at the Polar Urals
treeline during the 20th century. Global Change Biology, 14(1), 1581-1591.
Driscoll, W.W., Wiles, G.C, D'Arrigo, R.D., & Wilmking, M. (2005). Divergent tree growth
response to recent climatic warming, Lake Clark National Park and Preserve, Alaska.
Geophysical Research Letters, 32(20), L20703, doi:20710.21029/22005GL024258.
Esper, J., & Frank, D. (2009). Divergence pitfalls in tree-ring research. Climatic Change,
94(3-4), 261-266.

20

Euskirchen, E.S., McGuire, A.D., & Chapin, F.S., III. (2007). Energy feedbacks of northern
high-latitude ecosystems to the climate system due to reduced snow cover during 20th
century warming. Global Change Biology, 13(11), 2425-2438.
Fleming, R.A., & Volney, W.J.A. (1995). Effects of climate-change on insect defoliator
population processes in Canada boreal forest - some plausible scenarios. Water Air
and Soil Pollution, 52(1-2), 445-454.
Foley, J.A. (1994). The sensitivity of the terrestrial biosphere to climatic-change - a
simulation of the middle Holocene. Global Biogeochemical Cycles, 8(A), 505-525.
Ford, J.D., & Smit, B. (2004). A framework for assessing the vulnerability of communities in
the Canadian Arctic to risks associated with climate change. Arctic, 57(A), 389-400.
Gamache, I., & Payette, S. (2005). Latitudinal response of subarctic tree lines to recent
climate change in eastern Canada. Journal of Biogeography, 32, 849-862.
Govt, of the NWT. (2008). NWT Climate Change Impacts and Adaptation Report.
Yellowknife: Government of the Northwest Territories, Environment and Natural
Resources.
Grace, J., Berainger, F., & Nagy, L. (2002). Impacts of climate change on the tree line.
Annals of Botany, 90, 537-544.
Greene, D.F., Macdonald, S.E., Haeussler, S., Domenicano, S., Noel, J., Jayen, K., et al.
(2007). The reduction of organic-layer depth by wildfire in the North American
boreal forest and its effect on tree recruitment by seed. Canadian Journal of Forest
Research, 37(6), 1012-1023.
Groisman, P.Y., Knight, R.W., Easterling, D.R., Karl, T.R., Hegerl, G.C, & Razuvaev, V.N.
(2005). Trends in intense precipitation in the climate record. Journal of Climate,
18(9), 1326-1350.
Halliday, W.E.D., & Brown, A.W.A. (1943). The Distribution of Some Important Forest
Trees in Canada. Ecology, 24(3), 353-373.
Hare, F.K., & Ritchie, J.C. (1972). Boreal Biochmates. Geographical Review, 62(3), 332365.

21

Harsch, M.A., Hulme, P.E., McGlone, M.S., & Duncan, R.P. (2009). Are treelines
advancing? A global meta-analysis of treeline response to climate warming. Ecology
Letters, 12(10), 1040-1049.
Hartley, A.E., Neill, C , Melillo, J.M., Crabtree, R., & Bowles, F.P. (1999). Plant
performance and soil nitrogen mineralization in response to simulated climate change
in subarctic dwarf shrub heath. Oikos, 86, 331-343.
Hinzman, L.D., Bettez, N.D., Bolton, W.R., Chapin, F.S., III, Dyurgerov, M.B., Fastie, C.L.,
et al. (2005). Evidence and implications of recent climate change in northern Alaska
and other arctic regions. Climatic Change, 72(3), 251-298.
Hobbie, S.E., & Chapin, F.S., III. (1998). An experimental test of limits to tree establishment
in Arctic tundra. Journal of Ecology, 86, 449-461.
Hogg, E.H., & Hurdle, P. A. (1995). The aspen parkland in western Canada - a dry-climate
analog for the future boreal forest. Water Air and Soil Pollution, 82(1-2), 391-400.
Holtmeier, F.K., & Broil, G. (2005). Sensitivity and response of northern hemisphere
altitudinal and polar treelines to environmental change at landscape and local scales.
Global Ecology and Biogeography, 14, 395-410.
IPCC. (2007). Climate Change 2007: The Physical Science Basis. Cambridge and New York:
Intergovernmental Panel on Climate Change.
Jacoby, G.C, & D'Arrigo, R.D. (1997). Tree rings, carbon dioxide, and climatic change.
Proceedings of the National Academy of Sciences, 94, 8350-8353.
Jacoby, G.C, Lovelius, N.V., Shumilov, O.I., Raspopov, O.M., Karbainov, J.M., & Frank,
D.C. (2000). Long-term temperature trends and tree growth in the Taymir region of
northern Siberia. Quaternary Research, 53(3), 312-318.
Johnson, D., Kershaw, L., MacKinnon, A., & Pojar, J. (1995). Plants of the Western Boreal
Forest & Aspen Parkland. Edmonton: Lone Pine Publishing.
Johnstone, J.F., & Chapin, F.S., III. (2006). Fire interval effects on successional trajectory in
boreal forests of northwest Canada. Ecosystems, 9(2), 268-277.

22

Juday, G.P., Barber, V.A., Duffy, P., Linderholm, H., Rupp, T.S., Sparrow, S., et al. (2005).
Forests, land management, and agriculture. In Arctic Climate Impact Assessment (pp.
781-862). Cambridge: Cambridge University Press.
Juday, G.P., Barber, V.A., Rupp, T.S., Zasada, J.C, & Wilmking, M. (2003). A 200-Year
Perspective of Climate Variability and the Response of White Spruce in Interior
Alaska. In D. Greenland, D. G. Goodin & R. C. Smith (Eds.), Climate Variability and
Ecosystem Response at Long-Term Ecological Research Sites (pp. 226-250). New
York: Oxford University Press.
Kasischke, E.S., Bourgeau-Chavez, L.L., & Johnstone, J.F. (2007). Assessing spatial and
temporal variations in surface soil moisture in fire-disturbed black spruce forests in
Interior Alaska using spaceborne synthetic aperture radar imagery - Implications for
post-fire tree recruitment. Remote Sensing of Environment, 108(1), 42-58.
Kassi, N. (1993). Native perspectives on climate change. In G. Wall (Ed.), Impacts of
Climate Change on Resource Management in the North (pp. 43-49). Waterloo,
Ontario: University of Waterloo.
Kattenberg, A., Giorgi, F., Grassl, H., Meehl, G.A., Mitchell, J.F.B., Stouffer, R.J., et al.
(1996). Climate Models - Projections of Future Climate. In J. T. Houghton, L. G.
Meira Filho, B. A. Callander, N. Harris, A. Kattenberg & K. Maskell (Eds.), Climate
Change 1995, The Science of Climate Change, Contribution of Working Group 1 to
the Second Assessment of the Intergovernmental Panel on Climate Change (pp. 289357). Cambridge: Cambridge University Press.

Kattsov, V.M., & Walsh, J.E. (2000). Twentieth-century trends of arctic precipitation from
observational data and a climate model simulation. Journal of Climate, 13(8), 13621370.
Kirdyanov, A., Hughes, M., Vaganov, E., Schweingruber, F., & Silkin, P. (2003). The
importance of early summer temperature and date of snow melt for tree growth in the
Siberian Subarctic. Trees, 17, 61-69.
Kitaev, L., Forland, E., Razuvaev, V.N., Tveito, O.E., & Krueger, O. (2005). Distribution of
snow cover over Northern Eurasia. Nordic Hydrology, 36(A), 311.
Kljun, N., Black, T.A., Griffis, T.J., Barr, A.G., Gaumont-Guay, D., Morgenstem, K., et al.
(2007). Response of net ecosystem productivity of three boreal forest stands to
drought. Ecosystems, 10(6), 1039-1055.
23

Kohler, J., Brandt, O., Johansson, M., & Callaghan, T. (2006). A long-term Arctic snow
depth record from Abisko, northern Sweden, 1913-2004. Polar Research, 25(2), 91113.
Koppen, W. (1936). Das geographische System der Klimate. In Handbuch der Klimatologie.
Berlin.
Krupnik, I., & Jolly, D. (2002). The Earth is Faster Now: Indigenous Observations of Arctic
Environmental Change. Fairbanks: Arctic Research Consortium of the United States.
Lavoie, C , & Payette, S. (1996). The long-term stability of the boreal forest limit in subarctic
Quebec. Ecology, 77(A), 1226-1233.
Lenihan, J.M., & Neilson, R.P. (1995). Canadian vegetation sensitivity to projected climaticchange at 3 organizational levels. Climatic Change, 30(1), 27-56.
Ling, F., & Zhang, T.J. (2007). Modeled impacts of changes in tundra snow thickness on
ground thermal regime and heat flow to the atmosphere in Northernmost Alaska.
Global and Planetary Change, 57(3-4), 235-246.
Lloyd, A.H., & Fastie, C.L. (2002). Spatial and temporal variability in the growth and
climate response of treeline trees in Alaska. Climatic Change, 52, 481-509.
Lloyd, A.H., Yoshikawa, K., Fastie, C.L., Hinzman, L.D., & Fraver, M. (2003). Effects of
permafrost degradation on woody vegetation at arctic treeline on the Seward
Peninsula, Alaska. Permafrost andPeriglacial Processes, 14(2), 93-101.
Lloyd, A.H., Wilson, A.E., Fastie, C.L., & Landis, R.M. (2005). Population dynamics of
black spruce and white spruce near the arctic tree line in the southern Brooks Range,
Alaska. Canadian Journal of Forest Research, 35(9), 2073-2081.
Lloyd, A.H., Fastie, C.L., & Eisen, H. (2007). Fire and substrate interact to control the
northern range limit of black spruce (Picea mariana) in Alaska. Canadian Journal of
Forest Research, 37, 2480-2493.
Lloyd, A.H., & Bunn, A.G. (2007). Responses of the circumpolar boreal forest to 20th
century climate variability. Environmental Research Letters, 2(A), 1-13.

24

Loehle, C. (2009). A mathematical analysis of the divergence problem in dendroclimatology.
Climatic Change, 94(3-4), 233-245.
MacDonald, G.M., Kremenetski, K.V., & Beilman, D.W. (2008). Climate change and the
northern Russian treeline zone. Philosophical Transactions of the Royal Society BBiological Sciences, 363(1501), 2285-2299.
Masek, J.G. (2001). Stability of boreal forest stands during recent climate change: evidence
from Landsat satellite imagery. Journal of Biogeography, 25(8), 967-976.
Maxwell, B. (1997). Recent Climate Patterns in the Arctic. In W. Oechel, T. V. Callaghan, T.
Gilmanov, J. I. Holten, B. Maxwell, U. Molau & B. Sveinbjornsson (Eds.), Global
Change and Arctic Terrestrial Systems (pp. 21-46). New York: Springer.
Mbogga, M.S., Hamann, A., & Wang, T.L. (2009). Historical and projected climate data for
natural resource management in western Canada. Agricultural and Forest
Meteorology, 149(5), 881-890.
McBeam, G., Alekseev, G., Chen, D., Forland, E., Fyfe, J., Groisman, P.Y., et al. (2005).
Arctic climate: Past and present. In Arctic Climate Impact Assessment (pp. 21-60).
Cambridge: Cambridge University Press.
McDonald, M., Arragutainaq, L., & Novalinga, Z. (1997). Voices from the Bay: Traditional
Ecological Knowledge oflnuit and Cree in the James Bay Bioregion. Ottawa:
Canadian Arctic Resources Committee and Environmental Committee of the
Municipality of Sanikiluaq.
McGuire, A.D., Ruess, R.W., Lloyd, A.H., Yarie, J., Clein, J.S., & Juday, G.P. (2010).
Vulnerability of white spruce tree growth in interior Alaska in response to climate
variability: dendrochronological, demographic, and experimental perspectives.
Canadian Journal of Forest Research, 40(1), 1197-1209.
Meunier, C , Sirois, L., & Begin, Y. (2007). Climate and Picea mariana seed maturation
relationships: A multi-scale perspective. Ecological Monographs, 77(3), 361-376.
Morgner, E., Elberling, B., Strebel, D., & Cooper, E.J. (2010). The importance of winter in
annual ecosystem respiration in the High Arctic: effects of snow depth in two
vegetation types. Polar Research, 29(1), 58-74.

25

Mynemi, R.B., Keeling, C D . , Tucker, C.J., Asrar, G., & Nemani, R.R. (1997). Increased
plant growth in the northern high latitudes from 1981 to 1991. Nature, 386, 698-702.
Nienstaedt, H., & Zasada, J.C. (1990). White Spruce. In R. M. Burns & B. H. Honkala
(Eds.), Silvics of North America (Vol. 1. Conifers). Washington: U.S. Dept. of
Agriculture, Forest Service.
Olthof, I., & Pouliot, D. (2010). Treeline vegetation composition and change in Canada's
western Subarctic from AVHRR and canopy reflectance modeling. Remote Sensing of
Environment, 114(A), 805-815.
Overpeck, J., Hughen, K., Hardy, D., Bradley, R., Case, R., Douglas, M., et al. (1997). Arctic
environmental change of the last four centuries. Science, 275(5341), 1251-1256.
Pajunen, A.M. (2009). Environmental and biotic determinants of growth and height of arctic
willow shrubs along a latitudinal gradient. Arctic Antarctic and Alpine Research,
41(A), 478-485.
Payette, S., Fortin, M.J., & Gamache, I. (2001). The subarctic forest-tundra: the structure of a
biome in a changing climate. BioScience, 51(9), 709-718.
Pisaric, M.F.J., Carey, S.K., Kokelj, S.V., & Youngblut, D. (2007). Anomalous 20th century
tree growth, Mackenzie delta, Northwest Territories, Canada. Geophysical Research
Letters, 34(5), L05714, doi:05710.01029/02006GL029139.
Pojar, J. (1996). Environment and biogeography of the western boreal forest. Forestry
Chronicle, 72(1), 51-58.
Post, E., & Forchhammer, M.C (2008). Climate change reduces reproductive success of an
Arctic herbivore through trophic mismatch. Philosophical Transactions of the Royal
Society B-Biological Sciences, 363(1501), 2367-2373.
Rees, W.G. (2007). Characterisation of Arctic treelines by LiDAR and multispectral imagery.
Polar Record, 43(221), 345-352.
Rizzo, B., & Wiken, E. (1992). Assessing the sensitivity of Canada's ecosystems to climatic
change. Climatic Change, 21, 37-55.

26

Rupp, T.S., Chapin, F.S., III, & Starfield, A.M. (2000). Response of subarctic vegetation to
transient climatic change on the Seward Peninsula of northwest Alaska. Global
Change Biology, 6(5), 541-555.
Sala, O.E., Chapin, F.S., III, Armesto, J.J., Berlow, E., Bloomfield, J., Dirzo, R., et al.
(2000). Global biodiversity scenarios for the year 2100. Science, 257(5459), 17701774.
Schimel, D., Kittel, T.G.F., Running, S.W., Monson, R., Turnispeed, A., & Anderson, D.
(2002). Carbon sequestration studied in western U.S. mountains. EOS, Transactions,
American Geophysical Union, 53(40), 445-456.
Scott, P.A., Staniforth, R.J., & Hansell, I.C. (1993). Responses of white spruce to tundra
environments: woodland development.Unpublished manuscript.
Scott, P.A., Lavoie, C , MacDonald, G., Sveinbjomsson, B., & Wein, R.W. (1997). Climate
Change and Future Position of Arctic Tree Line. In W. Oechel, T. V. Callaghan, T.
Gilmanov, J. I. Holten, B. Maxwell, U. Molau & B. Sveinbjomsson (Eds.), Global
Change and Arctic Terrestrial Ecosystems (pp. 245-265). New York: Springer-Verlag
Inc.
Serreze, M.C, Walsh, J.E., Chapin, F.S., III, Osterkamp, T., Dyurgerov, M., Romanovsky,
V., et al. (2000). Observational evidence of recent change in the northern highlatitude environment. Climatic Change, 46(1-2), 159-207.
Sirois, L. (1993). Impact of fire on Picea-mariana and Pinus-banksiana seedlings in subArctic lichen woodlands. Journal of Vegetation Science, 4(6), 795-802.
Sirois, L. (2000). Spatiotemporal variation in black spruce cone and seed crops along a boreal
forest - tree line transect. Canadian Journal of Forest Research, 30(6), 900-909.
Stone, R.S., Dutton, E.G., Harris, J.M., & Longenecker, D. (2002). Earlier spring snowmelt
in northern Alaska as an indicator of climate change. Journal of Geophysical
Research-Atmospheres, 107, D10, doi: 10.1029/2000JD000286.
Sturm, M., Racine, C , & Tape, K. (2001). Increasing shrub abundance in the Arctic. Nature,
411, 546-547.

27

Suarez, F., Binkley, D., Kaye, M.W., & Stottlemyer, R. (1999). Expansion of forest stands
into tundra in the Noatak National Preserve, northwest Alaska. Ecoscience, 6(3), 465470.
Sullivan, P.F. (2010). Snow distribution, soil temperature and late winter CO2 efflux from
soils near the Arctic treeline in northwest Alaska. Biogeochemistry, 99(1-3), 65-77.
Szeicz, J.M., & Macdonald, G.M. (1994). Age-dependent tree-ring growth-responses of subArctic white spruce to climate. Canadian Journal of Forest Research, 24(1), 120-132.
Szeicz, J.M., & MacDonald, G.M. (1995). Recent white spruce dynamics in the subarctic
alpine treeline of north-western Canada. Journal of Ecology, 83, 873-885.
Szeicz, J.M., & MacDonald, G.M. (1996). A 930-year ring-width chronology from moisturesensitive white spruce (Picea glauca Moench) in northwestern Canada. Holocene,
6(3), 345-351.
Tape, K., Sturm, M., & Racine, C. (2006). The evidence for shrub expansion in Northern
Alaska and the Pan-Arctic. Global Change Biology, 12(A), 686-702.
Trenberth, K.E., Jones, P., Ambenje, P., Bojariu, R., Easterling, D.R., Klein Tank, A., et al.
(2007). Observations: Surface and atmospheric climate change. In S. Solomon, D.
Qin, M. Manning, Z. Chen, M. Marquis, K. B. Averyt, M. Tignor & H. L. Miller
(Eds.), Climate Change 2007: The Physical Science Basis. Contribution of Working
Group 1 to the Fourth Assessment Report of the Intergovernmental Panel on Climate
Change. Cambridge and New York: Cambridge University Press.
Vaganov, E.A., Hughes, M.K., Kirdyanov, A.V., Schweingruber, F.H., & Silkin, P.P. (1999).
Influence of snowfall and melt timing on tree growth in subarctic Eurasia. Nature,
400(61 A0), 149-151.
Viereck, L.A., & Johnston, W.F. (1990). Black Spruce. In R. M. Bums & B. H. Honkala
(Eds.), Silvics of North America (Vol. 1. Conifers). Washington: U.S. Dept. of
Agriculture, Forest Service.
Vygodskaya, N.N., Groisman, P.Y., Tchebakova, N.M., Kurbatova, J.A., Panfyorov, O.,
Parfenova, E.I., et al. (2007). Ecosystems and climate interactions in the boreal zone
of northern Eurasia. Environmental Research Letters, 2, 045033, doi:
045010.041088/041748-049326/045032/045034/045033.

28

Wahl, H.E., Fraser, D.B., Harvey, R.C, & Maxwell, J.B. (1987). Climate of Yukon. Ottawa:
Environment Canada.
Walker, M.D., Wahren, C.H., Hollister, R.D., Henry, G.H.R., Ahlquist, L.E., Alatalo, J.M., et
al. (2006). Plant community responses to experimental warming across the tundra
biome. Proceedings of the National Academy of Sciences of the United States of
America, 103(5), 1342-1346.
Werner, R.A. (1996). Forest health in boreal ecosystems of Alaska. The Forestry Chronicle,
72(1), 43-46.
Wilmking, M., Juday, G.P., Barber, V.A., & Zald, H.J. (2004). Recent climate warming
forces contrasting growth responses of white spruce at treeline in Alaska through
temperature thresholds. Global Change Biology, 10, 1724-1736.
Wilmking, M., D'Arrigo, R.D., Jacoby, G.C, & Juday, G.P. (2005). Increased temperature
sensitivity and divergent growth trends in circumpolar boreal forests. Geophysical
Research Letters, 32( 15), doi: 10.1029/2005GL023331.
Wilmking, M., & Myers-Smith, I. (2008). Changing climate sensitivity of black spruce
(Picea mariana Mill.) in a peat land-forest landscape in Interior Alaska.
Dendrochronologia, 25(3), 167-175.
Wilson, R., D'Arrigo, R.D., Buckley, B., Bantgen, U., Esper, J., Frank, D., et al. (2007). A
matter of divergence: Tracking recent warming at hemispheric scales using tree ring
fate. Journal of Geophysical Research, 112, doi: 10.1029/2006JD008318.
Yarie, J. (1983). Environmental and successional relationships of the forest communities of
the Porcupine River drainage, interior Alaska. Canadian Journal of Forest Research,
13(5), 721-728.
Yukon Department of Energy, Mines, and Resources. (2009). Spruce Beetle in the Yukon.
Retrieved March 2, 2011, from http://www.emr.gov.yk.ca/forestry/sprucebeetle.html
Zhang, K., Kimball, J.S., Hogg, E.H., Zhao, M.S., Oechel, W.C, Cassano, J.J., et al. (2008).
Satellite-based model detection of recent climate-driven changes in northern highlatitude vegetation productivity. Journal of Geophysical Research-Biogeosciences,
113(G3).

29

Zimov, S.A., Davydov, S.P., Zimova, G.M., Davydova, A.I., Schuur, E.A.G., Dutta, K., et al.
(2006). Permafrost carbon: Stock and decomposability of a globally significant
carbon pool. Geophysical Research Letters, 33, 1-5.

30

Chapter 2: The effect of environmental variation on radial growth patterns of spruce at
the western Canadian sub-Arctic treeline
Abstract
Tree growth responds to the combined effect of climate and environment. At the latitudinal
treeline, it is frequently believed that climate exerts the greatest effect upon variable tree
growth. However, emerging evidence suggests that tree growth in the North is much more
complicated and variable across the landscape. We investigated the role environmental
variation has in determining the annual variation of tree growth in northwestern Canada. Tree
cores were sampled from across a range of sites in Yukon and Northwest Territories, and
principal components analysis was used to derive the main sources of variation in the treering chronologies. We found that slope gradient exerts a strong effect on tree growth at
smaller spatial scales. At a larger scale, ecoregional classification seems to be the main
element determining variable growth. These results highlight the importance of considering
the spatial scale at which observations are made of tree growth in high latitude regions.

31

2.1 Introduction
The circumpolar boreal forest represents about 30% of the worlds forested area (ACIA,
2004), encompassing a wide array of environment and climatic types, from the dry,
mountainous regions of western North America, to the relatively flat muskeg of Siberia. The
northern extent of the boreal forest is of particular interest, as trees there exist at or near their
physiological limits within an arctic environment that is currently experiencing rapid changes
in climate (Hinzman et al, 2005; Ballantyne et al, 2010). In response to arctic climate
change, it is expected that climatically sensitive northern treelines will exhibit some of the
earliest and most pronounced indications of environmental change (Payette et al, 2001). In
order to observe emerging and ongoing changes within the circumpolar treeline, this study
addresses the western Canadian arctic treeline in northern Yukon and Northwest Territories,
in two adjacent but environmentally distinct regions. Specifically, the study regions
encompass the transitional zone between continuous forest cover and open tundra (shrub or
tussock), termed the forest-tundra (FT) ecotone (Hare & Ritchie, 1972; Scott et al, 1997).

The FT ecotone is typified by sparsely spaced individual and aggregated trees, the
distributions of which often reflect the underlying hydro logic, topographic, and reproductive
(i.e. seed vs. vegetative layering) processes (Arno, 1984; Scott et al, 1997). Trees within the
FT ecotone resemble those in the adjacent continuous forest, but are only about half as tall
(Scott et al, 1993).Trees assume krummholz stature at the furthest northern extent of the
forest's range before giving way to shrubs and open tundra.

Tree growth is a response of highly complex, integrated systems. It is a consequence of
climate acting upon a specific environment supporting particular species and compositions.
32

In many cases in the Arctic, the physical environment may be a primary determinant of plant
growth and variation (Chapin et al, 1992). The effects of climate and environment are
evident in the tree-ring patterns of individual trees, and these patterns have long been used to
observe and study historic climate fluctuations. The dominant prevailing climatic conditions
(i.e. temperature and available moisture) of each calendar year of growth, combined with the
effects of microclimate resulting from environment and other biological functions, produce a
chronological record of the integrated effect of the growing environment over the lifespan of
each individual tree. From these records, it is presumed that we can reconstruct past climate
fluctuations accurately beyond the scope of the climate station records. We can also infer
particular monthly or seasonal climate factors that may be most influential on annual tree
growth, an important step when using tree rings as climate proxies to model tree growth
responses to future climate change.

The objective of this study was to examine tree-growth patterns across a range of sites within
and between two unique environments at the western Canadian arctic treeline, in order to
distinguish important environmental traits that influence tree growth at high-latitude sites.
Once indentified, these traits may indicate how landscape variation can potentially create
notable differences in growth limitations (and possibly adaptive traits) across small distances.

Results from climate and tree-ring studies in circumpolar regions are often extrapolated from
small, widely spaced study sites to much larger geographical areas or regions, in an attempt
to make broad observations about historic and potential climate conditions in an immense
region (e.g. Briffa et al, 2003; Lloyd & Bunn, 2007). Such a low-density sampling approach
(while logistically understandable) inhibits our ability to address site-specific localized
33

environmental conditions. Consequently, certain environmental characteristics, such as local
topographic variation, soil properties, plant inter- and intra-specific competition, or genetic
variability, must be generalized and, as such, are not fully considered in broadly scoped study
designs and sampling procedures. In turn, a sampling approach that filters out fine-scale
environmental components by necessity will potentially mask the effects of certain
environmental components that could play a significant role in determining the nature of
relationships between climate and tree growth in high-latitude circumpolar regions.

Furthermore, few representative sites exist in some large regions, such as Siberia and west
and central northern Canada (with the exception of the Mackenzie River delta), due in large
part to the general inaccessibility and high costs of conducting research in some regions of
the far North. As a result, northern research areas with the easiest access to sites and
facilities, such as Alaska, eastern Canada and Fennoscandia, are often favoured for northern
dendroclimatic research sites. Consequently, this has led to a general disproportionate
representation of these areas and subsequent biased model results based on a limited range of
sites. Large unrepresented areas are also likely to harbour some of the most extreme
terrestrial environments in the circumpolar north, traits that may inhibit easy access and use
by researchers. This potential scenario of climatic and environmental change, combined with
a general lack of ground-based research in these regions, leads to uncertain assumptions
about the response of arctic tree growth to climate change, and how this response will factor
into the carbon balance of the boreal forest, climate feedback mechanisms, and overall
biodiversity of arctic regions (Bonan & Sirois, 1992; ACIA, 2004; Chapin et al, 2005).
Without a clearer understanding of these patterns the range of variability of the response and
underlying mechanisms of projected changes at the Arctic treeline will remain questionable
34

(Vygodskaya et al, 2007). Therefore, in order to improve projections of future conditions,
both large- and small-scale measures of change should be integrated and assessed
simultaneously.

This study addressed a significant knowledge gap in northern dendroclimatic studies by
assessing the effect of environmental variation on tree growth patterns. Variable tree growth
was examined in two adjacent but distinct environments in order to identify important
landscape elements that may provide more thorough understanding of the effect of climate
change on tree growth. Additionally, few studies of variable tree growth have been conducted
in this region of northwestern Canada, and as such this study will serve to fill a gap in the
coverage of circumpolar dendroclimatic studies. The primary question addressed in this study
is what degree of the variation in tree growth responses to climate can be attributed to
landscape variation in the Arctic? Building upon this, we ask what specific environmental
features can help to define the response of tree growth to climate? And do these effects differ
between tree species, or do they occur independently of species? We hypothesize that
environmental variation will significantly affect the way tree growth responds to climate
within the forest-tundra ecotone in northwestern Canada.

2.2 Methods and materials
2.2.1 Field sites
We selected 33 sites along the northern Dempster Highway, 14 sites in Yukon and 19 sites in
the Northwest Territories (NWT). Sites were selected in the field, focusing on open-grown
stands on southerly slope aspects, while avoiding concave topography that would likely
provide more favourable moisture conditions and limit the climate sensitivity of trees
35

growing in these areas. We chose sites that appeared to be the most moisture-limited in order
to obtain the most complete, climatically sensitive tree ring records (Fritts, 1976), as strong
limitations to growth to due to temperature are prevalent across the entire latitudinal treeline.
Moisture-sensitive sites were located on predominantly dry upland slopes, generally with
rocky soils, that should experience rapid soil moisture movement and runoff, as well as
having limited moisture retention abilities.

2.2.1.1 Field site locations
The Yukon sites spanned the region from the Arctic Circle north to the northern extent of tree
growth, just south of Wright Pass (67° 2'56.81" N, 136°12'24.73" W) in the Richardson
Mountains, and were distributed along a north-south gradient, following the Dempster
Highway along the western edge of the mountains (Figures 2.1, 2.2). Continuing along the
Dempster corridor into the NWT, sites were distributed along an east-west gradient, from the
eastern foothills of the Richardson Mountains and across the Mackenzie delta and Arctic Red
River plain nearly to the town of Inuvik (Figures 2.1, 2.2). Site elevations ranged from 491 m
to 758 m in Yukon, with altitudinal treeline generally observed around 700 m, comprised
entirely of white spruce. Conversely, site elevations in the NWT ranged from 31 m in the
delta to 400 m in the Richardson foothills, where a composition of tamarack, white and black
spruce existed at the altitudinal treeline near 400 m.

2.2.2 Ecoregional classification & descriptions
2.2.2.1 Environment of Yukon
All sites in the Yukon were located in the British-Richardson Mountains ecoregion of the
Taiga Cordillera ecozone (Smith et al, 2004), which represents 5% of the total Yukon area,
36

and is characterized by the largest unglaciated mountain range in Canada (Figure 2.3). Soils
in this ecozone are formed on mountainside colluvium deposits and on the flat pediment
surfaces of the valleys, where much of the soil development is driven by cryoturbation and
physical weathering processes. On slopes, shale bedrock is often exposed or at shallow
depths, and patterned ground surfaces are common. In the valleys, tussock tundra is the
primary feature, with scattered populations of white spruce (Picea glauca) and shrubs. Nearsurface permafrost is continuous in nearly all areas, the active layer depth usually near 0.5m.
Vegetation in this ecozone is predominantly classified as shrub tundra, with trees generally
limited to riparian zones and non-north facing slopes (Yukon Ecoregions Working Group,
2004). Climatic conditions in this region are modified by mountain topography, which
contribute to orographic effects on precipitation and often channels strong winds through the
valleys (Wahl et al, 1987). The following estimated climate data are derived from the
ClimateWNA v.4.52 model for the period 1901-2006, as no suitable representative climate
station exists in this region (Wang et al, 2006). Mean annual temperature for this region is
about -6.4°C January mean temperature is -24.7°C with extreme minimum temperatures
approaching -40°C, whereas July mean temperatures are around 12.9°C, with extreme
maximum temperatures reaching the lower 20s°C (Figure 2.4). Winter temperatures are
milder at higher elevations due to inversions (Wahl et al, 1987). The region receives an
average annual precipitation of 437 mm, with the largest accumulations in the summer
months, and falling as snow from September onward (Figure 2.5).

2.2.2.2 Environment of NWT
In the NWT, most sites were located within the Mackenzie Delta and Arctic Red Plain High
Subarctic ecoregions (Level IV), which together comprise the northernmost portion of the
37

Taiga Plains High Subarctic ecoregion (Level III) (Ecosystem Classification Group, 2007).
Three additional sites were located within the Richardson Plateau High Subarctic ecoregion
(Level IV), a component of the larger Tundra Cordillera High Subarctic ecoregion (Level III)
(Figure 2.3) (Ecosystem Classification Group, 2007).

Encompassing the largest river delta in Canada, the Mackenzie Delta ecoregion is
characterized by level alluvial deposits surrounded by thousands of small lakes and stream
channels. This region was glaciated by the Laurentide ice sheet, which led to varied soil
parent materials. Soil formation is strongly affected by permafrost and the fluvial
environment, exhibiting little to no horizon development and tussock formation. Vegetation
is typified by dense forests of white spruce and black spruce (Picea mariana), with tamarack
(Larix lyallii) and balsam poplar (Populus balsamifera) interspersed in some areas. Shrubs
are prevalent, particularly green alder (Alnus viridis) and dwarf birch (Betula glandulosa). In
higher elevations where permafrost is closer to the soil surface, forests become more open
and stunted, and shrubs assume low-growing statures. The Arctic Red Plain ecoregion, also
subject to Laurentide glaciations, exhibits soils similar to the Mackenzie ecoregion. Black
spruce is the dominant tree species, and typically grows very slowly in open stands (Viereck
& Johnston, 1990). Low shrubs such as crowherry (Empetrum nigrum) and cloudberry
(Rubus chamaemorus) are common understory components, along with lichens and peat
moss. Frequent fires have created large areas of pioneer dwarf birch (Ecological
Classification Group, 2007).

Climate in the Taiga Plains ecoregion is largely dictated by unobstructed weather patterns
originating off the adjacent Beaufort Sea. The following climate data are derived from the
38

ClimateWNA v.4.52 model for the period 1901-2006 (Wang et al, 2006), which is
approximately double the length of the permanent climate station record in Fort McPherson.
Mean annual temperature in for the Fort McPherson area (representing this ecoregion) is
-8.2°C Mean January temperature is -29.4°C with extreme minimums approaching -42°C,
whereas the mean July temperature is 14.4°C with extreme maximums exceeding 23°C
(Figure 2.4). Mean annual precipitation is about 284mm, with the largest amounts falling in
late summer to autumn (Figure 2.5).

Adjacent to the Taiga Plains ecoregion, the Tundra Cordillera/Richardson Plateau ecoregion
is a transition zone from the Mackenzie Delta towards the Yukon-NWT border in the
Richardson Mountains. The lower elevations of this ecoregion, where our sites were located,
were also subject to glaciations, reflected in the prominence of till deposits (Duk-Rodkin et
al., 2004). Permafrost is nearly continuous, excepting some valley slopes and riparian areas,
and consequently Cryosols are the dominant soil type. Black spruce occurs in open, wet
woodlands, with shrubs - predominantly northern Labrador tea (Ledum palustre) and
crowberry - and lichens comprising the understory. White spruce is found along narrow
valley slopes and river terraces, along with abundant green alder, willow (Salix spp.), and
dwarf birch in the understory. The climate in this ecoregion is influenced strongly by weather
coming in from the Arctic Ocean, which is subsequently affected by the mountains. Mean
annual temperature for this ecoregion is -7.5°C Mean January temperature is -28.1°C, and
mean July temperature is 13.6°C (Figure 2.4). Mean annual precipitation is estimated to be
378 mm, most of which falls in the summer and autumn (Figure 2.5). As there is no
permanent climate station in this region, those data are derived from the ClimateWNA v.4.52
model, for the period 1901-2006 (Wang et al, 2006).
39

Dry, forested areas in this ecoregion proved much more difficult to find, as slope gradients
were generally lower and soils appeared to be moister than those found in the adjoining
regions in the NWT and Yukon. Thus, we identified sites that exhibited drier conditions
relative to the rest of the ecoregion, though the site conditions were less dry than desired.
However, we made these concessions with respect to the great variability of the landscape in
this area, realizing that finding perfectly suited conditions is an unreasonable assumption.

2.2.3 Sampling strategy
Tree cores were collected from 15-18 healthy, dominant trees at each site. Inter-tree
competition was not a significant factor, as nearly all the stands were open grown with
relatively low stem densities (Figure 2.6, 2.7). We collected tree increment cores oriented
parallel to the slope contour to minimize the influence of slope stress on ring formation (i.e.
compression and tension wood). Two samples from each tree were obtained from fulldiameter cores, allowing us to select the best quality core samples, as well as providing a
means to visually check for missing or false rings.

White spruce was exclusively sampled in the Yukon, as it was essentially the only spruce
species present within the sample site conditions. Both white and black spruce were present
in the NWT on similar sites, occasionally intermixed but more often occurring separately.
Tamarack and balsam poplar were also present in some stands (Figure 2.7). White spruce
was favoured when possible to maximize cross-site comparisons, but in some areas black
spruce was the only species present.

40

Three unique hill features (hereafter termed Hill sites), approximately 15-20 m in relief of the
flat surroundings, were sampled within the Arctic Red River plain between the towns of Fort
McPherson and Tsiigehtchic. These small mounds featured prominent, discrete white spruce
stands on the south faces, with black spruce covering every other slope as well as the
surrounding flatlands (Figure 2.7). While these features are infrequent on the landscape of the
plain and may not wholly represent the conditions of the entire region, finding habitat that
supported white spruce in the NWT was needed in order to facilitate cross-comparisons with
same species and similar habitat types found in the Yukon study area. The objective was not
to extrapolate findings to entire regions, but rather to observe changes in growth patterns
across a defined gradient, and sampling from the Hill sites provided the best means of doing
so.

An additional chronology from the Dolomite Uplands near Inuvik, NWT was previously
constructed in the mid 1990s (Szeicz & MacDonald, 1996). This data set was obtained from
the International Treering Database (http://www.ncdc.noaa.gov/paleo/treering.html), and was
subsequently incorporated into our analyses to increase our sample range.

Basic site and understory information was collected at each site, including slope aspect and
gradient, elevation, coordinate location, slope position, and basic soil substrate types. Stand
photos were taken in the four cardinal directions from a single point within each site. The
relative abundance of basic vegetation types (high and low shrubs, grass, forbs, lichen/moss,
litter, course and fine woody debris, litter, and bare surfaces) (Ogden, 2008) were estimated
visually from a random point within each site (Appendix C).

41

2.2.4 Chronology development
Tree cores were prepared according to standard dendrochronological methods (Stokes &
Smiley, 1968). Samples were sanded with progressively finer grits (80-400) of sandpaper,
and were visually cross-dated to assign calendar years to each ring (Cook et al, 1990). Ring
widths were measured using a Velmex™ measurement system, accurate to 0.001mm. Dating
accuracy was checked using the standard program COFECHA (Holmes, 1983), and
measurement errors were corrected where identified. Tree cores that did not crossdate well
with each master chronology after visual and statistical inspection were subsequently
omitted.

The R package dplR (Bunn, 2008) was used for chronology detrending. Detrending is an
important step that better isolates the climate signal in the tree-ring series, by removing
biological and other non-climatic growing trends, such as inter- and intra-specific
competitive stresses or periodic insect or disease outbreaks. Early radial growth is often
marked by greater relative ring widths in the first decade or so of a tree's existence, that
gradually become smaller as limited resources are available to add to the yearly increasing
surface area of the bole. Often, this leads to general J-shaped negative exponential growth
patterns. However, typical negative exponential growth trends were not commonly observed
in our tree ring samples. Therefore, rather than using a more traditional negative exponential
or linear detrending method, each tree-ring series was detrended using a cubic smoothing
spline to remove the non-climatic growth trend in each series (Cook & Peters, 1981). We
used a cubic smoothing spline with a wavelength at 67% of the series length and a frequency
response of 50% for all series (Cook et al, 1990). A more adaptable percentage-based spline
rigidity value - as opposed to a fixed number - was chosen due to the disparate length of all
42

the series, as it could adapt similarly to short and long series. It is important to note that no
detrending method can remove all biological growth trends, but a consistent method that can
apply to all tree-ring series is central to creating robust chronologies. Once all series were
detrended, a mean value chronology was calculated for each site by averaging each year's
ring-width index value. Autocorrelation was removed from each series before averaging,
which removed the cumulative growth effects from the previous year's growth within each
tree, which is a standard accepted treatment to tree-ring chronologies. This is done in dplR by
fitting autoregressive models to the data, and selected using Akaike's information criteria
(Venables & Ripley, 2002; Bunn, 2008). The residuals from the model are then used to
construct a prewhitened chronology using Tukey's biweight robust means. This ensures a
relatively pure climatic signal in the chronology, with the biological and cumulative effects
on growth limited in the final chronology.

2.2.5 Data analysis
2.2.5.1 Chronology analysis
Basic statistics were computed for each site chronology with COFECHA, including series
length, mean sensitivity (a relative measure of ring-width variability), and intra-series
correlations (a measure of the strength of the overall population signal). All chronologies
(sites) were then correlated with each other in order to observe trends amongst and between
regions and site types.

2.2.5.2 Multivariate analysis
Principal components analysis (PCA) is a statistical method that reorganizes a dataset so that
independent groups of variables with similar patterns are identified, which can provide a
43

basis to explore the underlying mechanisms that bring about the specific groupings. PCA was
performed on the entire set of chronologies to determine the main modes of growth variation
over a common time period (1929-2007) (Peters et al, 1981; Driscoll et al, 2005). The R
package 'psych' was used to perform this analysis, including varimax rotation (Revelle,
2011). Each region was then subsequently treated separately to obtain a closer examination
of each region's sources of variability. Varimax-rotated principal components (PCs) that
supported the most significant modes of variation were retained for analysis (Griesbauer &
Green, 2010). The PCs were Varimax-rotated to better facilitate the interpretation of the
loadings, and to allow easier comparison to various environmental and climate variables
(Tabachnick & Fidell, 1989). The number of retained components was determined using the
scree test; the number of retained components was found at the inflection point of the scree
plot of eigenvalues (Tabachnick & Fidell, 1989).

2.3 Results
2.3.1 Chronologies
In Yukon, 11 chronologies (all white spruce) were retained for analysis, ranging from 104 to
233 years in length. Intra-site correlations ranged from .523 to .627 with a mean of .584,
indicating a strong population climate signal in all stands. Mean sensitivity ranged from .192
to .277 with a mean of .225 (Table 2.1). Mean sensitivity of the Yukon sites was significantly
positively correlated with elevation (r = .692, p = 0.018), however intra-site correlations did
not significantly correlate with any environmental factor (Table 2.2). Approximately 5% of
the sample cores that did not adequately visually crossdate were omitted from the final
chronologies.

44

In the NWT, 17 chronologies - 9 of which were black spruce stands - were constructed,
ranging from 80 to 258 years in length. Three north slope black spruce stands at the Hill sites
were combined into one chronology, due to an insufficient number of quality samples in each
stand. Intra-site correlations were similar to those in Yukon, ranging from .466 to .663, with
a mean of .566, again indicating a relatively strong cohesive climate signal. Mean sensitivity
ranged from .178 to .237 with a mean of .214 (Table 2.1). Mean sensitivity in the NWT
chronologies showed only a weak, non-significant positive correlation (r = .257, p = 0.32)
with elevation, but was significantly negatively correlated with longitude (r = -.531, p =
0.034). Intra-site correlation values did not significantly correlate with any other variable
(Table 2.2). Approximately 12% of the sample cores were omitted from the final
chronologies, as they did not adequately crossdate with the master chronology. However,
most of these omitted samples came from the three aforementioned north slope Hill sites,
where extremely poor growing conditions resulted in low quality tree-ring samples. Only 4%
of the total cores were omitted from the rest of the sites, a comparable figure to the
percentage of total cores omitted from Yukon.

2.3.2 Multivariate analysis
Principal components analysis (PCA) was conducted on the entire dataset, and the results
indicated the primary source of growth variation across the entire study area was regionally
divided between Yukon and the Northwest Territories (Table 2.3). Subsequently, separate
PCAs were performed for each region, so that traits underlying the variation could be
assessed in each region.

45

2.3.2.1 Yukon principal components analysis
Principal components analysis performed on the Yukon chronologies extracted two principal
components (PCs) that explained about 77% of the total chronology variance. Individually,
the two PCs explained 38.9% and 38.5% respectively (Table 2.4). The first PC (YTPC1) was
most strongly associated with low slope-gradient sites (between 10% and 33%). The second
PC (YTPC2) was most strongly associated with high slope-gradient sites (38% to 84%)
(Table 2.5). A lone exception was CaribouL, which had a recorded gradient of 10% but was
correlated with the high-gradient group (YTPC2).

2.3.2.2 NWT principal components analysis
The NWT chronologies were represented by three modes of variability which explained
nearly 73% of the total variance. The three PCs represented 38.9%, 17.8%, and 16.2% of the
total chronology variance, respectively (Table 2.4). Occasionally two loadings were very
similar for a site, and selection was then based subjectively on the most logical PC group.
However, it is possible that sites with two similar loadings may represent transitional sites,
exhibiting some characteristics of more than one PC group.

The first PC (NTPC1) correlated most strongly with sites located within the Mackenzie River
delta, including three additional sites (slump sites) on the adjacent plateau (Table 2.6). The
second PC (NTPC2) correlated most strongly with sites near the western extent of forest
cover, in the foothills of the Richardson Mountains (Table 2.6). These were also the sites
with the highest elevations in the NWT, and were comprised entirely of black spruce. The
third PC (NTPC3) correlated most strongly with sites near the Peel River, in the Peel River

46

plateau ecozone that serves as a transition zone between the delta and the mountains (Table
2.6).

2.4 Discussion
The overall PCA results indicated that the Yukon and NWT study regions can reasonably be
characterized as distinctly different environments, which confirms our a priori assumption
that the growing conditions were different between the two areas. Nearly all of the Yukon
sites were responsible for the highest loadings in the first principal component, with the
NWT sites accounting for the remaining components. These components represent distinct
sources of variation that can be attributed to various environmental variables; in this case,
patterns of tree growth in the Yukon study area were distinct from the patterns observed in
the NWT study area. This suggests the presence of a distinct bioclimatic boundary at a small,
regional scale; specifically, the narrow Richardson Mountain range that delineates the
disparate adjacent regions in Yukon and the NWT. The results of the overall PCA also reflect
the sampling constraints in the two regions, dictated by the course of the Dempster Highway.
In Yukon, sampling sites were part of a single ecoregion, running parallel along the western
flank of the Richardson Mountains across a latitudinal gradient, which subsequently led to
more intensive sampling of similar site types. The NWT sites, however, ran across a
longitudinal transect and encompassed

multiple ecoregions and greater regional

environmental diversity along the way, which resulted in a larger scale of observation with
fewer sampling sites within each ecoregion. Had each ecoregion been sampled more heavily
in the NWT, it is likely that more site-level characteristics would have emerged, similar to
the Yukon sites.

47

Additionally, the strong positive correlation between mean sensitivity and elevation in the
Yukon corroborates others' findings (Fritts, 1976; D'Arrigo et al, 2004) and underscores the
effect of specific limiting factors commonly present at elevational treeline sites. Such a
strong relationship highlights the effectiveness of sampling at elevational sites to obtain the
most sensitive climate signals, even at the latitudinal tree line. Conversely, the elevational
trend was not evident in the NWT even with a larger elevational range (> 350 m), likely due
to separate elevational clusters in each region. This seems to suggest that elevation may not
be as important in determining sensitivity when it is considered in a broader context with a
higher number of other environmental factors.

2.4.1 Yukon
In Yukon, growth patterns appeared to separate based on slope angle (Table 2.5). This
finding was somewhat unexpected, as slope angle is not commonly referenced as an
important driver of tree-growth variation. It is a logical finding, however, as slope angle may
be an important contributing factor for water runoff and retention, as well as overall site
productivity (i.e. greater surface area per horizontal hectare) (Fritts, 1976; Barnes et al,
1998).
Another factor affected by slope angle, particularly at high-latitude sites, is incident solar
radiation (Q), which directly affects soil temperature and snow melt timing on some slope
aspects (Pohl et al, 2006). Solar radiation is closely related to temperature variation and
moisture regimes, making it one of the most important drivers of ecological processes in
nearly all ecosystems, particularly in the northern FT ecotone (Smith, 1996; Liepert, 2002).
However, with increasing latitude, the overall value of Q decreases (Chapin et al, 1992).
48

s

Simultaneously, the effective heating capacity of the incident solar radiation is reduced
through the FT ecotone as the dark, radiation-absorbing forest transitions to a more reflective,
snow-covered tundra (Hare & Ritchie, 1972). However, Hare & Ritchie note that mountains
can greatly complicate these general patterns by imposing locally specific microclimates and
weather (i.e. clouds) conditions. This assertion may be particularly relevant in this region
since most of the Yukon sites are located proximate to the Richardson Mountains, as opposed
to the NWT where many of the sites are located within the flat delta.

Solar radiation levels at high latitudes are dominated by annual cycles, with near constant
irradiance during the summer and zero irradiance during the dark winter. Similarly,
irradiance varies widely during the day. For example, at Barrow, Alaska (71°17'26.00" N,
156°47'19.00" W), incoming radiation levels at solar noon were 15 times greater than levels
measured at solar midnight during the summer solstice (24 hours of daylight) (Tieszen,
1972). The range of radiation levels is created as the incident angle of incoming solar
radiation changes with the sun's elevation in the sky. This suggests that slope angle (and to a
lesser degree, slope aspect) would have an effect on total net incident radiation (Kimmins &
Wein, 1986). The effect of slope angle has previously been quantified by the "equivalent
latitude" concept, where for any given slope on the earth, the latitude of a corresponding
horizontal surface that receives the same amount of insolation can be calculated, and thus
provide an index of potential insolation (Lee, 1964). However, in at least one study, slope
angle was found to have a negligible effect on soil temperature in interior Alaska (Bonan,
1991).

49

In turn, solar insolation can have considerable effects upon permafrost, soil temperature and
moisture, and runoff dynamics, particularly at high latitudes (Chapin et al, 1992). Higher
levels of solar insolation tend to warm soils and deepen active layers (Dingman & Koutz,
1974); deeper active layers were observed on steeper slopes in the study area (data not
shown). Steep slopes will also facilitate rapid runoff; coupled with earlier snowmelt (due to
increased solar radiation), trees growing on steeper slopes are likely to experience higher
levels of drought stress than those growing on lower slope angles or flat surfaces. These
factors appear to readily explain the difference of growth patterns between the high- and lowangle slopes observed in this study, where slope angle emerged as an unexpected but
important factor in determining tree growth response to environment and climate. The
dividing point between the two PC groups (between 33% and 38%) may be arbitrary in a
broader context, but it represents an identifiable break point within the context of this study
region.

2.4.2 NWT
The PC groupings in the NWT appear to indicate the presence of two distinct ecoregions as
well as a transition zone between them (Tables 2.4, 2.6), which contrasts to the relatively
homogenous environment in Yukon. The distinct ecoregions in the NWT, in addition to the
smaller number of sampling sites in each ecoregion, likely explains why slope angle and
related solar radiation were not apparent drivers of growth variation within the larger context
of the NWT study area. The largest sampled environment was the Mackenzie delta (NTPC1),
a unique area that harbours specific hydrologic, environmental and climatic conditions
typical of large river deltas (Burn & Kokelj, 2009; Kanigan et al, 2009; Cassano & Cassano,
2010). Tree growth in the delta region was strongly connected between sites regardless of
50

tree species, underlying permafrost conditions or slope conditions, which suggests that the
delta may function as a large climatic "oasis." That is, the growing conditions here may be
relatively uniform throughout and less extreme from those in the surrounding areas, as
indicated by its unique ecoregion classification (Ecological Classification Group, 2007).
Similarly, the foothills group (NTPC2) represents a unique set of growing conditions within
the NWT study range, influenced primarily by mountain-affected weather, higher elevation,
and less available soil moisture (compared to the relative abundance of soil moisture in the
delta). Sites located in the narrow zone between these two regions (NTPC3) represent
another set of growing conditions in the area transitioning from the uplands to the delta,
which is emphasized by the similar proportion of variance to the other PC groups indicated
by the principal components analysis. The differentiation of this transition zone environment
underscores the importance of identifying and addressing bioclimatic transition zones, as
landscape variation in the Arctic is highly spatially complex within and between distinct
regions.

The results from the NWT suggest that multiple climatic envelopes exist in this study area of
the NWT, as the sites of each specific PC grouping can be generally be sorted by the general
ecoregional classifications established by the Ecosystem Classification Group. Indeed,
multiple environments may have been included in some large-scale tree-ring studies in
Alaska that included sites with a large degree of spatial separation (Lloyd & Fastie, 2002;
Wilmking et al, 2004). It is clear that relationships between tree growth and environmental
variation can be uniquely complex at multiple scales, and determining the most suitable scale
for observing these relationships remains a significant, yet important challenge. There

51

remains a substantial need for a more detailed understanding of the underlying mechanisms
of environmental variation that lead to varied patterns of tree growth.

2.4.3 Conclusion
This study highlighted the importance of identifying key context-specific environmental
factors that may impose limits to tree-growth relationships with climate. Arctic ecosystems
are fundamentally distinct from the world's other great ecosystems, and thus require specific
considerations and assumptions that extend beyond more commonly held principles of basic
ecology. Identifying and addressing the key components and linkages of high-latitude
bioclimatic systems is essential when considering potential scenarios of ecologic and climatic
change.

This study also served to reiterate the importance of considering the context in which an area
is studied and sampled. In the Yukon study area, slope gradient exerted a significant
influence on determining tree growth variation on south aspects. While this is not a novel
notion, slope angle regardless is not often considered in tree-ring studies at high latitudes.
However, our findings indicate that it likely plays a stronger role than previously considered
in determining climate/growth interactions, specifically within the context of a latitudinal
gradient along the Richardson Mountain foothills. This factor may or may not have similar
effects in other areas of the sub-arctic treeline, across other climatic or environmental
gradients, or even during different periods of history. Similarly, environmental variation
plays a significant role in the NWT, albeit reflecting different influential sources due to a
much broader range of environmental conditions. Here, groups of sites that were represented
by distinct growth patterns (i.e. PCA groupings) distinguished themselves based primarily on
52

two ecoregions and an associated transition zone between them, reflecting more complex
geologic, climatic, and physiographic conditions. Though viewed in a larger, more complex
context than in the Yukon study area, environmental variation still appeared as a strong
influence, and can be readily addressed with a basic preliminary landscape analysis.

These case studies show how varying degrees of environmental variation can play a
significant role in coupled climate/tree growth interactions, and thus should be fully
considered regardless of the scale or context of future dendroclimatic studies at high
latitudes. Based on these initial results, future research should assess the direct effects
specific components of environmental variation have on the overall and temporal effects of
important climatic components upon tree growth. A broader examination of the role of slope
angle throughout Yukon would also be a warranted, to determine its relative effect in a
broader scope. Different latitudes, environment types (mountainous, plateau, riparian, etc.),
or species compositions may affect the relative influence of slope angle, and as yet this
remains relatively unknown.

53

2.5 Literature Cited

ACIA. (2004). Arctic Climate Impact Assessment. Cambridge.
Arno, S.F. (1984). Timberline: Mountain and Arctic Forest Frontiers. Seattle: The
Mountaineers.
Ballantyne, A.P., Greenwood, D.R., Damste, J.S.S., Csank, A.Z., Eberle, J.J., & Rybczynski,
N. (2010). Significantly warmer Arctic surface temperatures during the Pliocene
indicated by multiple independent proxies. Geology, 38(1), 603-606.
Barnes, B.V., Zak, D.R., Denton, S.R., & Spurr, S.H. (1998). Forest Ecology (4th ed.). New
York: John Wiley & Sons, Inc.
Bonan, G.B. (1991). A biophysical surface-energy budget analysis of soil-temperature in the
boreal forests of interior Alaska. Water Resources Research, 27(5), 767-781.
Bonan, G.B., & Sirois, L. (1992). Air temperature, tree growth, and the northern and southern
range limits to Picea mariana. Journal of Vegetation Science, 3, 495-506.
Briffa, K.R., Osborn, T.J., & Schweingruber, F.H. (2003). Large-scale temperature
inferences from tree rings: a review. Global and Planetary Change, 40, 11-26.
Bunn, A.G. (2008). A dendrochronology program library in R (dplR). Dendrochronologia,
26(2), 115-124.
Bum, C.R., & Kokelj, S.V. (2009). The environment and permafrost of the Mackenzie Delta
area. Permafrost andPeriglacial Processes, 20(2), 83-105.
Cassano, E.N., & Cassano, J.J. (2010). Synoptic forcing of precipitation in the Mackenzie
and Yukon River basins. International Journal of Climatology, 30(5), 658-674.
Chapin, F.S., III, Sturm, M., Serreze, M.C., McFadden, J.P., Key, J.R., Lloyd, A.H., et al.
(2005). Role of land-surface changes in Arctic summer warming. Science, 310(5148),
657-660.

54

Chapin, F.S., III, Jefferies, R.L., Reynolds, J.F., Shaver, G.R., & Svoboda, J. (Eds.). (1992).
Arctic Ecosystems in a Changing Climate. San Diego: Academic Press, Inc.
Cook, E.R., & Peters, K. (1981). The smoothing spline: a new approach to standardizing
forest interior tree-ring width series for dendroclimatic studies. Tree Ring Bulletin,
41, 45-54.
Cook, E.R., Briffa, K.R., Shiyatov, S., Mazepa, A., & Jones, P.D. (1990). Data Analysis. In
E. R. Cook & L. A. Kairiukstis (Eds.), Methods of Dendrochronology: Applications
in the Environmental Sciences (pp. 97-162). Dordrecht, the Netherlands: Kluwer
Academic Publishers.
D'Arrigo, R.D., Kaufman, R.K., Davi, N., Jacoby, G.C, Laskowski, C , Myneni, R.B., et al.
(2004). Thresholds for warming-induced growth decline at elevational tree line in the
Yukon Territory, Canada. Global Biogeochemical Cycles, 18(3), GB3021,
doi:3010.1029/2004GB002249.
Dingman, S.L., & Koutz, F.R. (1974). Relations among vegetation, permafrost, and potential
insolation in central Alaska. Arctic and Alpine Research, 6(1), 37-47.
Driscoll, W.W., Wiles, G.C, D'Arrigo, R.D., & Wilmking, M. (2005). Divergent tree growth
response to recent climatic warming, Lake Clark National Park and Preserve, Alaska.
Geophysical Research Letters, 32(20), L20703, doi:20710.21029/22005GL024258.
Duk-Rodkin, A., Barendregt, R.W., Froese, D.G., Weber, F., Enkin, R., Rod Smith, I., et al.
(2004). Timing and extent of Plio-Pleistocene glaciations in north-western Canada
and east-central Alaska. In J. E. a. P. L. Gibbard (Ed.), Developments in Quaternary
Science: Quaternary Glaciations-Extent and Chronology-Part II: North America (1st
ed., Vol. 2, pp. 313-345). Amsterdam: Elsevier.
Ecosystem Classification Group. (2007). Ecological Regions of the Northwest Territories Taiga Plains. Yellowknife, NT: Department of Environment and Natural Resources,
Government of the Northwest Territories.
Fritts, H.C. (1976). Tree Rings and Climate. Caldwell: The Blackburn Press.
Griesbauer, H.P., & Green, D.S. (2010). Regional and ecological patterns in interior
Douglas-fir climate-growth relationships in British Columbia, Canada. Canadian
Journal of Forest Research, 40(2), 308-321.

55

Hare, F.K., & Ritchie, J.C. (1972). Boreal Biochmates. Geographical Review, 62(3), 332365.
Hinzman, L.D., Bettez, N.D., Bolton, W.R., Chapin, F.S., III, Dyurgerov, M.B., Fastie, C.L.,
et al. (2005). Evidence and implications of recent climate change in northern Alaska
and other arctic regions. Climatic Change, 72(3), 251-298.
Holmes, R.L. (1983). Computer-assisted quality control in tree-ring dating and
measurement. Tucson: University of Arizona.
Kanigan, J.C.N., Burn, C.R., & Kokelj, S.V. (2009). Ground temperatures in permafrost
south of treeline, Mackenzie delta, Northwest Territories. Permafrost and Periglacial
Processes, 20(2), 127-139.
Kimmins, J.P., & Wein, R.W. (1986). Nature of Taiga Environment: Introduction. In K. Van
Cleve, F. S. Chapin, III, P. W. Flanagan, L. A. Viereck & C T. Dyrness (Eds.),
Forest Ecosystems in the Alaskan Taiga (Vol. 57). New York: Springer-Verlag.
Lee, R. (1964). Potential insolation as a topoclimatic characteristic of drainage basins.
International Association of Scientific Hydrology. Bulletin, 9(1), 27 - 41.
Liepert, B.G. (2002). Observed reductions of surface solar radiation at sites in the United
States and worldwide from 1961 to 1990. Geophysical Research Letters, 29(10).
Lloyd, A.H., & Fastie, C.L. (2002). Spatial and temporal variability in the growth and
climate response of treeline trees in Alaska. Climatic Change, 52, 481-509.
Lloyd, A.H., & Bunn, A.G. (2007). Responses of the circumpolar boreal forest to 20th
century climate variability. Environmental Research Letters, 2(A), 1-13.
Ogden, A. (2008). Yukon Forestry Monitoring Program: Field Manual & Monitoring
Protocol (1 ed.). Whitehorse: Government of Yukon.
Payette, S., Fortin, M.J., & Gamache, I. (2001). The subarctic forest-tundra: the structure of a
biome in a changing climate. BioScience, 51(9), 709-718.
Peters, K., Jacoby, G.C, & Cook, E.R. (1981). Principal Components Analysis of Tree-Ring
Sites. Tree-Ring Bulletin, 41, 1-19.
56

Pohl, S., Marsh, P., & Pletroniro, A. (2006). Spatial-temporal variability in solar radiation
during spring snowmelt. Nordic Hydrology, 37(1), 1-19.
Revelle, W. (2011). psych: Procedures for Personality and Psychological Research (Version
1.0-94). Evanston, Illinois: Northwestern University.
Scott, P.A., Staniforth, R.J., & Hansell, I.C (1993). Responses of white spruce to tundra
environments: woodland development.Unpublished manuscript.
Scott, P.A., Lavoie, C , MacDonald, G., Sveinbjomsson, B., & Wein, R.W. (1997). Climate
Change and Future Position of Arctic Tree Line. In W. Oechel, T. V. Callaghan, T.
Gilmanov, J. I. Holten, B. Maxwell, U. Molau & B. Sveinbjomsson (Eds.), Global
Change and Arctic Terrestrial Ecosystems (pp. 245-265). New York: Springer-Verlag
Inc.
Smith, C.A., Meikle, J.C, & Roots, C.F. (2004). Ecoregions of the Yukon Territory:
Biophysical properties of Yukon landscapes (PARC Technical Bulletin No. 04-01).
Summerland, British Columbia: Agriculture and Agri-Food Canada.
Smith, R.L. (1996). Ecology and Field Biology (5th ed.). Menlo Park, California: AddisonWesley Educational Publishers, Inc.
Stokes, M.A., & Smiley, T.L. (1968). An Introduction to Tree-Ring Dating. Tucson: The
University of Arizona Press.
Szeicz, J.M., & MacDonald, G.M. (1996). A 930-year ring-width chronology from moisturesensitive white spruce (Picea glauca Moench) in northwestern Canada. Holocene,
6(3), 345-351.
Tabachnick, B.G., & Fidell, L.S. (1989). Using Multivariate Statistics (2 ed.). New York:
HarperCollinsP«Z)fo/2era, Inc.
Tieszen, L.L. (1972). The seasonal course of aboveground production and chlorophyll
distribution in a wet arctic tundra at Barrow, Alaska. Arctic and Alpine Research, 4,
307-324.
Venables, W.N., & Ripley, B.D. (2002). Modern Applied Statistics with S (A ed.). New York:
Springer.

57

Viereck, L.A., & Johnston, W.F. (1990). Black Spruce. In R. M. Burns & B. H. Honkala
(Eds.), Silvics of North America (Vol. 1. Conifers). Washington: U.S. Dept. of
Agriculture, Forest Service.
Vygodskaya, N.N., Groisman, P.Y., Tchebakova, N.M., Kurbatova, J.A., Panfyorov, O.,
Parfenova, E.I., et al. (2007). Ecosystems and climate interactions in the boreal zone
of northern Eurasia. Environmental Research Letters, 2, 045033, doi:
045010.041088/041748-049326/045032/045034/045033.
Wahl, H.E., Fraser, D.B., Harvey, R.C, & Maxwell, J.B. (1987). Climate of Yukon. Ottawa:
Environment Canada.
Wang, T., Hamann, A., Spittlehouse, D.L., & Aitken, S.N. (2006). Development of scale-free
climate data for western Canada for use in resource management. International
Journal of Climatology, 26(3), 383-397.
Wilmking, M., Juday, G.P., Barber, V.A., & Zald, H.J. (2004). Recent climate warming
forces contrasting growth responses of white spruce at treeline in Alaska through
temperature thresholds. Global Change Biology, 10, 1724-1736.
Yukon Ecoregions Working Group. (2004). Yukon Taiga Cordillera. In C. A. Smith, J. C.
Meikle & C. F. Roots (Eds.), Ecoregions of the Yukon Territory: Biophysical
properties of Yukon landscapes (pp. 101-106). Summerland, British Columbia:
Agriculture and Agri-Food Canada.

58

Table 2.1 Site descriptions and statistics
Chronology
Intraseries Mean
Region Species Lat. Long. Hev. (m) Slope (%) Aspect Length (years) Correlation Sensitivity Site Position
6675 136 33 758
125
YT
W
46
0 523
0 277
s
us
6682 136 36 684
173
YT
W
17
0 582
0216
Crest W
s
CUSMSLS
6682 136 33 662
209
W
33
S
0 604
023
Hideawayl . YT
LS
Site
Quarry

HideawayU YT
YT
Gut

W

66 82

S

162

0 585

0 249

66 65

13633 689
13634 670

22

W

38

s

183

0 563

0 246

CUSMS

CanbouL

YT

13638 632

233

0611

0 205

LST

YT

66 72

136 38 716

10
10

S

CanbouU

S

139

0 627

0222

CUS

Shed

YT

N

104

0 596

0 192

CUS

YT

6691 -136 36 491
66 92 136 34 504

72

Cutbank

84

S

148

0 575

0218

Bench

YT

66 92

169

0605

S

165

0 549

0206
0 211

CUS
CUSMS

YT
NWT

43
21

S

North

w
w
w
w
w
w

66 72

US MS

3

s

85

0613

0219

C US MS LS

23
35

N

152

s

95

0 557
0661

0218
0 234

US MS
CUS

33
32

S

149

0485

0232

CUSMS

S

156

0663

0 202

CUSMS

Midway
PeclN

B

NWT
NWT

B

w

136 32 541

66 93

136 25 677
67 22 -135 49 368
67 33 135 95 40

CUS

NWT

B

HillAG

NWT

W

67 33 -134 95 75
67 34 134 93 31
67 39 134 22 78

HillA^S

NWT

B

67 39

134 22 62

12

S

251

0 522

0196

LS

Slump 1

NWT

W

67 25

135 27 383

25

S

0 569

0219

CUSMS

Slump3

NWT

W

6726

135 25 369

43

S

258
207

0 563

0213

CUSMS

Slump2

NWT

w
w

s

199
154

US MS

NWT

21
38

0 623

HillBG
HillB S

11

S

HillC_S

NWT

B
B

67 25 -135 26 369
67 39 134 22 85
67 39 134 22 68
67 39

134 20 75

13

S

HiUC G
West

NWT

W
B

134 20 90
135 53 374

38

NWT

67 39
6721

135 49 398

7

133 40 83
67 39 -134 21 78

27

s
s
s
s

20

N

256

PcelG
PeelM

NWT

Foothill

NWT

B

67 22

Dolomite

NWT

W

68 21

Hill N

NWT

B

6

S

0 661

0212
0 187

249
252

05

022

CUSMS
LST

0466

0225

LST

143

0 622

0178

CUSMS

81

0 505

0237

MS

80

0 547

022

US

253

0553
0518

0 205

CUSMS

0214

MSLS

W=white spruce, B=black spruce, S=south aspect, N=north aspect, C=crest, US=upper slope, MS=mid slope,
LS=lower slope, T=toe of slope, Intra-senes Correlation=relative strength of chronology, Mean
Sensitivity=relative variance of chronology

59

Table 2.2 Site and chronology correlations
NT

YT
Sensitivity
Elevation
Latitude
Intra. Corr

Elevation
.692"

Latitude
-0.492
-.606**

Longitude
0.063
0.033

Intra.
Corr.
-0.599*
-.315
-.034

Elevation
.257

Latitude
-.277
-.435*

Longitude
-0.531**
-0.609**

Intra.
Corr
-.446*
.079
-.056

-0.561*
0.015
-correlation significant at p<0.05; *=correlation significant at p<0.1; Intra. Corr. = Intra-series Correlation

60

Table 2.3 Principal components analysis results for entire study area
REGION
NWT
NWT
NWT
NWT
NWT
NWT
NWT
NWT
NWT
NWT
NWT
NWT
NWT
NWT
NWT
NWT
NWT
YT
YT
YT
YT
YT
YT
YT
YT
YT
YT
YT
YT
YT
YT
YT
YT

STAND
Midway
PeelN
PeelG
PeelM
HillA_G
HiUA__S
Slump 1
Slump3
Slump2
HillB_G
HillBS
HillC_S
HillC_G
West
Foothill
Dolomite
Hill_N
SheepN
SheepS
Quarry
CrestW
HideawayL
HideawayU
Gut
CaribouL
CaribouU
GlacierCr
Shed
Cutbank
Bench
RockRiverl
RockRiver2
North

Loadings
Proportion of
Variance
Cumulative
Variance

RC5
0.272
0.189
0.030
0.252
0.253
0.331
0.327
0.247
0.233
0.219
0.198
0.341
0.295
0.147
0.272
0.374
0.182
0.216
0.378
0.814
0.801
0.601
0.720
0.737
0.685
0.795
0.611
0.696
0.465
0.472
0.464
0.511
0.389

RC6
0.072
0.293
0.294
0.360
0.798
0.326
0.491
0.514
0.406
0.800
0.511
0.573
0.780
0.041
0.161
0.647
0.548
0.104
0.272
0.180
0.244
0.258
0.221
0.237
0.230
0.255
0.275
0.285
0.268
0.217
0.073
0.171
0.177

RC3
0.659
0.751
0.767
0.768
0.239
0.239
0.300
0.246
0.283
0.300
0.234
0.326
0.246
0.238
0.201
0.117
0.272
0.482
0.290
0.169
0.268
-0.047
0.006
0.170
0.396
0.185
0.008
0.409
0.335
0.580
0.162
0.189
0.239

RC2
0.212
0.188
0.093
0.289
0.295
0.699
0.322
0.347
0.463
0.189
0.691
0.430
0.224
0.055
0.109
-0.017
0.528
0.008
0.525
0.065
0.188
0.308
0.318
0.334
0.250
0.183
0.254
0.000
0.444
0.291
0.462
0.501
0.166
RC2
3.787

RC1
0.244
0.146
0.036
0.046
0.228
0.067
0.315
0.126
0.344
0.002
0.079
0.008
0.158
0.038
0.074
0.413
0.217
0.583
0.429
-0.038
0.270
0.519
0.266
0.161
0.270
0.251
0.570
0.145
0.294
0.372
0.407
0.419
0.667
RC1
3.088

RC5
7.071

RC6
5.113

RC3
4.392

0.214

0.155

0.214

0.369

RC7
0.297
0.027
0.185
0.030
0.086
0.136
0.437
0.400
0.311
0.154
0.016
0.064
0.177
0.081
0.799
-0.033
0.026
0.242
0.228
0.325
0.124
0.088
0.158
-0.052
0.208
0.132
-0.102
0.198
0.211
0.106
0.175
0.130
0.053
RC7
1.786

RC4
0.281
0.118
0.247
0.004
0.048
0.148
0.107
0.303
0.276
0.017
-0.083
-0.184
-0.050
0.854
0.066
0.226
0.209
-0.268
-0.120
0.115
0.099
0.108
0.232
-0.034
-0.093
0.099
-0.033
0.069
-0.031
0.038
0.197
0.168
0.191
RC4
1.536

0.133

0.115

0.094

0.054

0.047

0.502

0.617

0.711

0.765

0.811

Table 2.4 Summary of regional principal components analysis
Component
Eigenvalues
Variance (%)
CumulativeVariance (%)

1
4.27
38.9
38.9

Yukon
2
4.23
38.5
77.3

Loadings:
Quarry
CrestW
HideawayL
HideawayU
Gut
CaribouL
CaribouU
Shed
Cutbank
Bench
North

.565
.736
.889
.851
.562
.527
.690
.237
.375
.362
.709

.595
.593
.271
.360
.597
.722
.601
.851
.765
.810
.343

Midway
PeelN
PeelG
PeelM
HillA G
HillA_S
Slump 1
Slump3
Slump2
HillB G
HillB S
HillC S
HillC_G
West
Foothill
Dolomite
HillN

1
6.62
38.9
38.9

NWT
2
2.75
16.2
55.1

3
3.02
17.8
72.9

.224
.350
.234
.462
.870
.684
.695
.650
.641
.773
.805
.773
.841
-.039
.292
.658
.752

.591
.227
.367
.170
.175
.330
.504
.583
.571
.110
.057
-.023
.116
.766
.555
.322
.309

.607
.792
.732
.795
.270
.237
.253
.157
.221
.344
.287
.398
.299
.191
.199
.079
.246

Numbers in bold indicate highest loading

62

Table 2.5 Yukon PCA loadings and site characteristics
SITE
HideawayL
HideawayU
CrestW
North
CaribouU
Shed
Bench
Cutbank
CaribouL
Gut
Quarry

PCI
0.889
0.851
0.736
0.709
0.690
0.237
0.363
0.375
0.527
0.562
0.566

PC2
0.271
0.360
0.593
0.343
0.601
0.851
0.810
0.765
0.722
0.597
0.595

SLOPE (%)
33
22
17
21
10
72
43
84
10
38
46

ASPECT
135
135
120
180
84
20
158
142
84
112
166

ELEVATION
662
689
684
677
716
491
541
504
632
670
758

Table 2.6 PCA groups and ecological characteristics
Ecological Description

YTPC1

Ecological Classification
(Level III/IV)
Taiga Cordillera Ecozone /
British-Richardson Mountains
Ecoregion

YTPC2

See above

Sites with slope gradients between
38% and 84%

NTPC1

Taiga Plains High Subarctic /
Mackenzie Delta & Arctic Red
River High Subarctic

Sites within the Mackenzie River
delta region.

NTPC2

Tundra Cordillera High Subarctic /
Richardson Plateau High Subarctic

Westernmost sites in foothills of
Richardson Mountains.

NTPC3

Borderline of two regions above

Sites located between delta and
foothills near Peel River

Region

ID

Yukon

NWT

Sites with slope gradients between
10% and 33%

figure 2.1 Location of study area in northwest Canada
//\

// \
/ /
\
/
;•
\
/
/
)
/ "

/
/

/

'

1

AH
Yukon

<f.
f\

"'fX
/

.

v

'

• i

\

\

\ llo.thwest\
' Tenitoiies \

*N

V

\

X ^
^

\

\

-

"J

1

>'-..

/

/

T—

/

->•-.._
;

/
;'
\

/'
•
•

V-v•

l

N.

-

V'

r"' /"v

• -

1

4*V

^

4 / W

\

/

\

~ \

i

/

i

~"~~--i-—_;

'\

•'
t

.*->

*v*'«^._—' ^

V

A

^ >• ^

) >.-~

Hr'MirMt r » -.crr.pKijht-IC-r.M-A Uiu .il Ruvu ,(,•„-> i»J-i
£,-< M-tmli 1-t R"t r- \'. M ( J i " i w h Pfh-i < i *-• llrtli r - k ^ i n ^

r

\
"X v>

^.il'

.'
\

c

',-

;

\
\

\X '

CANADA

64

Figure 2.2 Site locations in Yukon and Northwest Territories
1

1
JDId Crow

\

3

(
\
<

v\-„

Wright Pass.
Dempster Hwy

f

Aklavik

-#*

"'

Inuvik

/ " *
Fort McPbdfson
A^^/Tsiigehtchic

•

Yukon Sites

A

NWT Sites

— — Northern edge of FT ecotone
0
25
50
100
150

-I

200
•••Kilometers

-*fr
r

Forest-tundra (FT) ecotone data courtesy of the Alaska Geobotany Center, Institute of Arctic Biology,
University of Alaska-Fairbanks.

65

Figure 2.3 Ecoregions and PC groups

Ecoregions

0

15

30

60

90

120
iKilometers

j Mackenzie Delta
Arctic Red Plain

4-

I British-Richardson Mountains
I Eagle Plains
1 Peel River Plateau

66

Figure 2.4 Mean regional temperatures
Average Regional Temperatures, 1901-2006
(ClimateWNA v.4.52)
20

10

1 1 1

o
<D
CD
<D

a

-20 -

-30

m
1
-40 - 1

Mean July Temperature
1 Mean Annual Temperature
1 Mean January Temperature
Delta

Plateau

Yukon

Region

61

Flgmre 2„S Regiminal mean amnmnal precipitation!

Regional Mean Annual Precipitation, 1901-2006
(ClimateWNA v. 4.52)
500 -T

400 -

300 -

^

I

I

E

200 -

100 0 "J

1

~I
Delta

1

1

1
Plateau

1

1

1
Yukon

68

Figure 2.6 Typical Yukon white spruce stands

(L) Photo depicts typical open-grown stand with low understory cover in Yukon. (R) White spruce stand on a
south-facing slope in the Yukon.

Figure 2.7 Typical NWT site conditions

(L) Common stand conditions in the NWT, with black spruce interspersed with tamarack and poplar.
Understories typically more dense than those found in Yukon. (R) Example of a Hill site, NWT. Stand of tall
white spruce growing on south face of small hill, surrounded by black spruce and tamarack. Near Tsiigehtchic,
within the Arctic Red River plain.

69

Appendix A.Yukon site data
All coordinates in UTM zone 08 west. W = white spruce; F = flat; C = slope crest; US = upper slope; MS =

Name
Quarry

Year
2009

Species
W

Slope (%)
46

Aspect
166

Elevation
(m)
758

Latitude
(UTM)
7404656

Longitude
(UTM)
441584

CrestW
HideawayL
HideawayU

2009
2009
2009

W

w
w

17
33
22

120
135
135

684
662
689

7412117
7412362
7412362

440362
441569
441569

Gut
CaribouL
CaribouU
GlacierCr
Shed
Cutbank

2009
2009
2009
2009
2009
2009

w
w
w
w
w
w

38
10
10
0
72
84

112
84
84
F
20
142

670
632
716
660
491
504

7392516
7400513
7400513
7392525
7421843
7422812

440867
439117
439117
441333
440368
441308

541
482
482
677

7423430
7421423
7421423
7424335

442221
440379
440379
445477

Bench
2009
43
158
w
RockRiverl
0
F
2009
w
RockRiver2 2009
0
F
w
21
North
2009
180
w
middle slope; LS = lower slope; T = toe of slope; R = riparian

Slope
Position
US
CUS
MSLS
LS
CUS
CUS
MS
LST
CUS
R
CUS
CUS
CUS
MS
R
R
US MS

70

Appendix B. Northwest Territories site data
Name

Year

Species

Slope (%)

Aspect

Elevation
(m)

Latitude
(UTM)

Longitude
(UTM)

Midway
PeelN

2008
2008
2008
2009

B
B

3
23

S
N

368
40

7455907
7468846

478865
459343

W

35

S

75

7468399

502318

PeelG

Slope
Position
CUS
MSLS
US MS

CUS
CUS
PeelM
B
33
S
31
7468886
502981
MS
2009
CUS
32
S
78
7474778
533582
HillAG
2009
W
MS
12
62
7474714
HillAS
B
533624
LS
2009
s
CUS
Slump 1
25
383
7459399
2009
W
488360
MS
s
CUS
Slump3
43
369
7459988
2009
W
489016
MS
s
Slump2
21
369
7459568
2009
488803
US MS
w
s
CUS
HiUB_G
2009
38
85
7474574
533570
MS
w
s
7474564
HillB_S
B
11
68
533525
LST
2009
s
B
13
75
7475033
534331
LST
HillC_S
2009
s
CUS
38
90
7475065
HillC_G
2009
534390
MS
w
s
374
West
B
6
7454846
477002
MS
2009
s
Foothill
B
7
398
7456316
479037
2009
US
s
CUS
Dolomite
W
27
83
7566640
566190
MS
2009
s
7474822
Hill N
B
N
78
N/A
2009
20
MSLS
All coordinates in UTM zone 08 west. W = white spruce; B = black spruce; C = slope crest; US = upper slope;
MS = middle slope; LS = lower slope; T = toe of slope

71

Appendix C. Soil, vegetation, and groundcover data
Mean Active
Layer Depth
(m)

Tall
Shrub

Low
Shrub

Grass

Herb

Moss/
Lichen

Litter

CWD

FWD

Burn

Bare
Soil

Rock

0 54

0

10

5

1

5

0

0

0

0

0

70

CrcstW

5

25

1

0

35

5

5

0

0

0

30

HideawayL

5

35

10

5

60

0

0

0

0

0

0

HideawayU

5

40

0

1

15

0

0

5

0

0

25

Name
Yukon
Quarry

Gut

10

35

0

0

20

5

0

5

0

0

50

CanbouL

70

20

5

0

20

5

0

10

0

5

0

CanbouU

5

20

0

0

30

5

0

0

0

0

70

Shed

5

30

0

1

80

5

0

0

0

0

5

Cutbank

40

60

0

5

30

10

5

10

0

0

0

Bench

80

50

1

0

15

5

0

0

0

0

1

90

40

1

15

60

0

1

5

0

0

0

80

40

10

5

5

60

5

10

0

0

0

15

40

5

1

20

15

1

5

0

5

0

GlacicrCr

RockRiverl
RockRivcr2
North
NWT
Midway

051

PeelN
PeclG

0 67

PeelM
HillAG

0 97

15

20

0

10

40

30

5

10

0

0

0

HillA_S

0 72

30

40

0

0

80

10

5

5

0

0

0

Slumpl

80

25

5

15

30

1

5

5

0

0

0

Slump3

90

80

5

0

75

15

10

10

0

0

0

S!ump2

65

50

1

1

75

5

1

5

0

0

0

HillBG

1 12

5

60

1

5

15

20

5

10

0

5

0

HillBS

07

20

30

1

0

80

1

0

15

0

0

0

HillC_S

0 73

5

70

0

0

70

5

1

5

0

0

0

HillC_G

108

10

40

1

5

25

15

5

15

0

5

0

West

5

30

5

20

40

1

0

5

0

1

0

Foothill

35

60

10

5

20

1

0

5

0

0

0

Dolomite

30

50

5

5

25

10

1

5

0

0

0

15

40

0

0

75

5

5

10

0

0

0

Hill N

0 75

Mean active layer depths were computed from measurements taken 1.5 m upslope of every sampled tree within
each site during mid-August, 2009. Accurate depth probe measurements in the Yukon were not possible due to
extremely rocky soils. Percentage coverage of vegetation and ground cover was estimated visually from a
random point within each site. Class specifications were defined by the Field Manual and Monitoring Protocols,
Yukon Forestry Monitoring Program (Ogden, 2008). Data were not collected from riparian sites or sites
sampled in 2008, which were not included into the analyses of this project.

72

Appendix D. Chronology summary data
Name
Quarry
CrestW
HideawayL
HideawayU
Gut
CaribouL
CaribouU
GlacierCr
Shed
Cutbank
Bench
RockRiverl
RockRiver2
North
Midway
PeelN
PeelG
PeelM
HillA_G
HillAS
Slump 1
Slump3
Slump2
HiUB__G
HillBS
HiUC_S
HillC_G
West
Foothill
Dolomite
Hill N

Region
YT
YT
YT
YT
YT
YT
YT
YT
YT
YT
YT
YT
YT
YT
NWT
NWT
NWT
NWT
NWT
NWT
NWT
NWT
NWT
NWT
NWT
NWT
NWT
NWT
NWT
NWT
NWT

Trees
(n)
12
29
17
15
12
30
27
19
29
16
16
13
19
27
15
15
34
14
28
13
18
16
17
15
14
12
16
15
16
18
21

Chronology
(years)
125
173
209
162
183
233
139
191
104
148
169
187
284
165
85
152
95
149
156
251
258
207
199
154
249
252
143
81
80
253
256

Interseries
Correlation
0.523
0.582
0.604
0.585
0.563
0.611
0.627
0.592
0.596
0.575
0.605
0.583
0.581
0.549
0.613
0.557
0.661
0.485
0.663
0.522
0.569
0.563
0.623
0.661
0.500
0.466
0.622
0.505
0.547
0.553
0.518

Mean
Sensitivity
0.277
0.216
0.230
0.249
0.246
0.205
0.222
0.208
0.192
0.218
0.206
0.199
0.191
0.211
0.219
0.218
0.234
0.232
0.202
0.196
0.219
0.213
0.212
0.187
0.220
0.225
0.178
0.237
0.220
0.205
0.214

Appendix E. Tree growth indices derived from PCA
YTPC1 (shallow)

x
«

2

0
-2
-4
1920

1940

1960

1980

2000

1980

2000

1980

2000

1980

2000

YTPC2 (steep)

1920

1940

1960

NTPC1 (delta)

1920

1940

1960

NTPC2 (uplands)

1920

1940

1960

NTPC3 (transition)

1920

1940

1960

1980

2000

74

Chapter 3: Scale-dependent climate and tree growth relationships at the western
Canadian subarctic treeline
Abstract
Summer temperature has frequently been assumed to be the dominant climate component
controlling tree growth at the latitudinal treeline. However, recent findings have indicated
more complex patterns of tree growth and climate interrelationships. We investigated how
these relationships change through variations in environment and spatial scales in northern
Yukon and Northwest Territories. Climate data were correlated to annual growth series
derived from a principal components analysis performed on a large set of tree-ring
chronologies. We found that climate effects on tree growth significantly changed based on
high and low slope gradients at smaller, local scales. Differences were also found at a larger,
regional scale, where significant climate effects were differentiated by ecoregional
classifications. These results show that climate and tree growth associations are not static
across the landscape or through different spatial scales, and future studies of potential
changes of the circumpolar treeline should acknowledge the inherent complexity of this
region.

75

3.1 Introduction
The location, density, and species composition of circumpolar latitudinal treelines are widely
expected to experience significant responses to emerging climatic changes in Arctic regions
(Holtmeier & Broil, 2005). In order to understand the effects of climate change on
circumpolar treeline structure and processes, it is necessary to identify and assess the most
important climatic parameters that affect tree growth. Effects of climate on tree growth can
be inferred from tree-ring patterns by identifying the climate signal apart from the effects of
non-climatic factors, such as topography, stand density, or disturbance. To obtain a clear
climate signal within the tree-ring record, specific site selection criteria, such as identifying
moisture-limited sites with low inter-tree competition, are employed to identify areas with the
most climatically sensitive tree populations (Fritts, 1976). However, environmental
characteristics not directly accounted for in site selection, such as soil and permafrost
characteristics, variable solar insolation, or slope steepness, can often be overlooked, but may
yet impart a significant influence on the true nature of climate and tree growth interactions.
Regional responses of tree growth may be influenced by specific combinations of climate,
environment, and forest characteristics. The combined effects of these regional responses will
determine the overall trajectory of future changes to the entire circumpolar treeline.
Therefore it is important to identify distinct areas with unique relationships between climate
and tree growth, and the specific environmental characteristics that determine the nature of
these relationships and the boundaries that separate them. The objective of this study was to
identify important climate variables as they occurred across an environmental gradient in
northwestern sub-arctic Canada. A secondary objective was to assess the viability of using
interpolated climate data for high-latitude sites, in lieu of using climate station data from far
away, unrepresentative sites, which may also have short or incomplete historical records.
76

The historic and projected ecological processes and climatic response of the northern treeline
is often inferred from tree-ring proxy records sampled from distant sites across broad and
diverse regions (Luckman & Wilson, 2005; Wilson et al, 2007; MacDonald et al, 2008).
However, this coarse approach to observing northern treeline dynamics may not adequately
consider important relationships between the forest, climate, and environment that are
inherent at different scales, which may skew our assumptions and interpretations of
climatically induced changes at the northern treeline. Broad regional assessments are
important, and for now provide some of the best information available for addressing
circumpolar issues of climatic and environmental change. However, a more detailed
understanding regarding the environmental limits to tree growth and the scale at which these
limits are evident is needed to address future scenarios of climate change and associated
forest responses at high latitudes.

In this study, we attempted to understand the role of environmental variation and scale of
observation when determining treeline growth responses to climate at the western Canadian
forest-tundra (FT) ecotone. To what degree do variable landscape features determine the
dominant climatic factors on tree growth? Can these factors help explain disparate tree
growth trends in the Arctic? We hypothesize that the environmental features identified in
Chapter 2 will support distinct climatic influences upon tree growth at the sub-Arctic treeline.

77

3.1.1 Climate-growth relationships of spruce
3.1.1.1 Temperature
Summer temperature most often correlates (positively) with tree growth at the northern
treeline (Bonan & Sirois, 1992; Vaganov et al, 1999; Kirdyanov et al, 2003). For example,
black spruce at the northern treeline in Alaska exhibited positive responses to summer
temperatures (Lloyd et al, 2005), as did white spruce at the northern treeline in central
Canada (D'Arrigo et al, 2009). Late summer temperature was also identified as having a
significant positive effect on growth of both white and black spruce at the treeline in Nunavut
(MacDonald et al, 1998). Similar observations were made of black spruce in northern
Quebec (Wang et al, 2002) as well as larch in the Siberian subarctic (Kirdyanov et al,
2003). Additionally, white spruce has been shown to lose its summer temperature sensitivity
with advancing age in the Yukon sub-arctic (Szeicz & MacDonald, 1995a). Occasionally
winter temperatures also affect black spruce growth, evidenced by positive correlations with
prior October (e.g. early winter) temperatures in Nunavut (MacDonald et al, 1998), and prior
winter seasonal temperatures at Ft. Wainwright near Fairbanks, Alaska (Juday et al, 2005).
These types of varied growth responses are likely the result of different regional climates and
environments, indicating that no one climatic variable can best explain tree growth across the
entire circumpolar region.

The relationship between summer temperature and tree growth holds true for most latitudinal
treeline sites, but is not evident south of the treeline within the continuous forest. At multiple
sites across Canada from Yukon to Quebec, these trends tend to reverse, as the strong
positive effect of summer temperature diminishes in lieu of stronger effects of spring and
summer precipitation (positive) and early summer temperatures (negative) (Brooks et al,
78

1998; Tardif & Bergeron, 2001; Miyamoto et al, 2010). This reversing trend indicates that
the forest-tundra ecotone is a distinct bioclimatic zone apart from the boreal forest.

There is mounting evidence that some trees at the northern treeline are losing their positive
association with growing-season temperatures. Wilmking et al. (2004) found that
approximately 40% of a large sample (n >1500) of white spruce in the Alaska and Brooks
Ranges of Alaska responded negatively to summer temperatures (hereafter negative
responders), while less than 40% of the sample population still responded positively
(hereafter positive responders) in the latter half of the 20th century. In the Mackenzie River
delta, Pisaric et al (2007) found that only 25% of their sample population of black spruce (n
= 654) showed a positive correlation with summer temperatures, with the remaining portion
beginning to lose its sensitivity in the 1930s. Similar findings have been reported for white
spruce in southwest Alaska (Driscoll et al, 2005), north-central Canada (D'Arrigo et al,
2009), and for larch in the Taymir region of northern Siberia (Jacoby et al, 2000). It appears
that the trees with declining positive summer temperature associations are exceeding a
certain temperature threshold at which they become unduly stressed in response to increasing
Arctic temperatures. Specific threshold values are few, but estimates between 11° and 12°C
(mean summer temperature) have been suggested for central Yukon and Alaska (D'Arrigo et
al, 2004; Wilmking et al, 2004).

These emerging patterns of contrasting growth responses to summer temperature has come to
be known as the 'divergence problem' (Briffa et al, 1998; D'Arrigo et al, 2008), and
currently represents a central focus in northern dendroclimatic studies (Lloyd & Bunn, 2007;
Wilson et al, 2007; Esper & Frank, 2009; Loehle, 2009). Drought stress has been the most
79

frequently suggested cause (Barber et al, 2000; Davi et al, 2003; Wilmking et al, 2004;
Driscoll et al, 2005; Pisaric et al, 2007; D'Arrigo et al, 2009; Beck et al, 2011), though
other confounding factors, such as increased plant competition (Hogg & Hurdle, 1995),
insect herbivory (Fleming & Volney, 1995), or increased UV-B levels (Callaghan et al,
2004a) potentially contribute to these patterns as well. Currently, it is not certain why
particular trees respond positively or negatively to warming temperatures, but it is likely that
a combination of environmental, biotic and abiotic factors contribute to this issue.

3.1.1.2 Precipitation
It is important to note that precipitation data are often less reliable than temperature data, and
in the circumpolar region, coverage is sparse and intermittent, measured time spans are short,
and datasets are often replete with missing values. As a result, the closest long-term climate
data is often collected at stations located a significant distance from most study sites (e.g.
Szeicz & MacDonald, 1995a; Wilmking & Juday, 2005). Alternately, interpolated data can
be derived from climate normals and general circulation models to approximate local
conditions and predict future scenarios (Mbogga et al, 2009). Recognizing that all forms of
available precipitation data have inherent limitations, general observations can still be made
that are relevant to current issues of climatic and environmental change in the north.

As evidenced by the aforementioned studies implicating temperature as the prime control
over tree growth in the Arctic, precipitation appears to have less of a controlling effect on
tree growth than temperature, though it may be increasingly important in future (drier)
situations. Precipitation may affect overall soil moisture deficits that can contribute to
drought limitations, but it is equally, if not more likely that increased evaporative demands
80

may be the leading cause (Beck et al, 2011). Significant correlations between precipitation
and tree-ring growth are not commonly found at high latitudes, as temperature is the
dominant climatic limitation. Where significant positive precipitation effects are found,
summer temperatures are often simultaneously found to exert negative effects. For example,
during the latter half of the 20th century in central Alaska, black spruce was found to be
positively correlated with August precipitation, accompanied by negative relationships with
July and August temperatures (Wilmking & Myers-Smith, 2008). Similarly, white spruce
negative responders (to summer temperatures) in the Mackenzie River delta reacted
positively to April precipitation (Pisaric et al, 2007).

Although there are some general notions of the relationship between tree growth and growing
season temperatures and precipitation, a review of current findings clearly shows a more
complex situation than previously assumed. Tree growth response to climate across the
northern treeline is not static or similar at all sites and for all species, yet large-scale
inferences of past climates and projected changes are regularly produced. The fact that so
much irregularity exists in circumpolar tree growth/climate relationships necessitates more
detailed studies to fully understand the nature and implications of these relationships.

3.2 Methods and materials
3.2.1 Field sites
We selected 33 sites along the northern Dempster Highway, 14 sites in Yukon and 19 sites in
the Northwest Territories (NWT). Specific site selection strategies and criteria are outlined in
Chapter 2, Section 2.2.1.

81

3.2.1.1 Site locations
The Yukon sites spanned the region from the Arctic Circle north to the northern extent of tree
growth, just south of Wright Pass (67° 2'56.81" N, 136°12'24.73" W) in the Richardson
Mountains (Figures 2.1, 2.2), and were distributed along a north-south gradient, following the
Dempster Highway along the western edge of the mountains. Continuing along the Dempster
corridor into the NWT, sites were distributed along an east-west gradient, from the eastern
foothills of the Richardson Mountains and across the Mackenzie delta and Arctic Red River
plain nearly to the town of Inuvik (Figure 2.2). Site elevations ranged from 491 m to 758 m
in Yukon, with altitudinal treeline generally observed around 700 m, comprised entirely of
white spruce. Conversely, site elevations in the NWT ranged from 31 m in the delta to 400 m
in the Richardson foothills, where a composition of tamarack, white and black spruce existed
at the altitudinal treeline near 400 m.

3.2.2 Ecoregional classification & descriptions
Detailed descriptions of the British-Richardson Mountains ecoregion in Yukon, as well as the
Richardson Plateau and Arctic Red River Plain ecoregions in the NWT, are given in Chapter
2, Section 2.2.2.

3.2.3 Sampling strategy
Our sampling strategy is described in detail in Chapter 2, Section 2.2.3.

82

3.2.4 Chronology development
Tree cores were prepared according to standard dendrochronological methods (Stokes &
Smiley, 1968). Specific details of sample preparation, measurement, and detrending are
described in Chapter 2, Section 2.2.4.

3.2.5 Data analysis
A detailed description of the methods employed for the analysis of the chronologies is
located in Chapter 2, Section 2.2.5. This same section also describes the application of
principal components analysis on the set of chronologies.

3.2.5.1 Climate data
Climate stations in the far north of Canada are sparsely located, and often do not have
adequately long and/or complete records. The station with the longest record (approximately
100 years) in the area is in Dawson City, Yukon, which is located over 300 km from the
nearest study site. In addition, the Dawson precipitation record is fraught with missing
values. Other stations, such as the ones near Old Crow, Inuvik, and Fort McPherson, are
closer to the study area, but do not have long enough records to be effectively utilized in this
study.

To address this situation, we used climate data derived from the ClimateWNA v.4.52 model,
which is an expanded version of the ClimateBC model (Wang et al, 2006). ClimateBC has
been used in prior studies conducted in British Columbia, Canada, with satisfactory results
(O'Neill et al, 2008; Stoehr et al, 2009; Griesbauer & Green, 2010b), supporting the use of
this data in applied situations. ClimateWNA extracts and downscales monthly PRISM data
83

(Daly et al, 2002) using years 1961-1990 as the reference period. It then calculates monthly
and annual climate data based on latitude, longitude and elevation. Representative locations
were determined for both Yukon (Quarry site) and NWT (Fort McPherson), based on their
average location and elevation amongst the sites of each region. Using all the output from the
model would be prohibitive for interpretation and analysis, so seasonal variables were used in
the analyses. Each variable was checked for normality using the Shapiro-Wilkes test, and log
transforms were applied to those not passing the initial test. The relationship between
modeled and measured data was tested by correlating the station data with modeled data
using the exact coordinates of the particular climate station. However, the output from
ClimateWNA would be expected to correlate highly with station data of the same location, as
the model was partially built using observed data from the entire regional coverage. It is
beyond the scope of this study to fully assess the precision of the ClimateWNA output, and
thus we assumed that the data is a realistic representation of the conditions in our study area.

3.2.5.2 Climate-growth analysis
Chronological scores from the regional PCs were correlated with corresponding
ClimateWNA output using Pearson's simple correlation coefficient. Climate indices
representing general circulation patterns were also correlated to the PC chronologies,
including the Arctic Oscillation, El Nino-Southern Oscillation, Pacific Decadal Oscillation,
and the Pacific North American Oscillation.

3.3 Results
The results of the chronology analyses and subsequent principal components analysis are
found in Chapter 2, Sections 2.3.1 and 2.3.2.
84

3.3.1 Climate data
Monthly temperature values from ClimateWNA correlated extremely well with the station
data; in Dawson (1901-2006), mean monthly temperature correlation coefficients ranged 0.95
to 0.99, and values from Inuvik (1957-2006) ranged 0.94 to 0.99. The Rock River climate
station (1995-2006), nearest to the Yukon sites, had lower but still strong correlations, likely
a reflection of the very short time period. This is not surprising nor entirely helpful, as
ClimateWNA output is largely based on measured station data. Precipitation records are
prone to more uncertainty, due to changing methods of measurement, frequent missing data,
and variable forms of precipitation. In Dawson, correlation coefficients for precipitation
ranged 0.46 to 0.88, while in Inuvik the coefficients range 0.53 to 0.95. The two lowest
values in Dawson correspond to the months of May and October, when precipitation is most
likely changing from snow to rain, increasing the likelihood of measurement errors as well as
confounding the extrapolation ability of the climate model. Similar observations are true for
Inuvik, except the "shoulder" months occur farther apart (February/March and November).

High correlations between the station and modelled data would be expected, however, as the
modelled data is derived mainly from the station data (Wang et al, 2006). Corrected
temperature data (to account for measurement equipment upgrades and/or station relocation)
from available stations in the study area (Dawson City, Old Crow, Rock River, Inuvik) were
correlated against each other using the longest possible common time periods, to assess the
relative regional cohesiveness of the climate variability (Table 3.1). Strong positive
correlations suggest that climate conditions fluctuate similarly across the study area, though
the absolute values may differ between the stations. This lends credibility towards using the
modelled data, as it likely portrays the common climate variation reflected by the area's
85

climate station records. The modelled data also allow for more accurate absolute temperature
(and to a lesser extent precipitation) values based on the coordinates and elevation that
represent the Yukon and NWT study areas. Temperature appears to be more accurately
predicted by the model, which allowed us to use temperature parameters with a greater
degree of confidence than those based on precipitation. Precipitation values, though more
variable and generally less correlated with station data than temperature, are still useful due
to uncertainties in both station and modelled data. The modeled data from ClimateWNA
provided a localized approximation of the conditions at the sites, and were thus presumed to
represent meaningful values.

3.3.2 Climate-growth relationships
3.3.2.1 Yukon
Spruce growth on low-gradient slopes (represented by YTPC1) correlated negatively (p <
0.05) with the previous season's winter, current summer and annual precipitation. Strong
positive correlations (p < 0.01) were found with previous autumn temperatures and the
associated previous season's frost-free period. Continentality, a measure of the difference
between the warmest and coldest monthly temperatures, also had a positive effect on tree
growth at these sites. A weak but significant positive correlation with the current season's
mean warmest month temperature (MWMT, usually July) was also present (Table 3.2, Figure
3.1). No climate index (e.g., PDO) was significantly correlated with this PC.

Spruce growth on high-gradient slopes, represented by YTPC2, was negatively affected most
strongly by the previous summer's temperatures, with the maximum and mean temperatures
highly significant (p < 0.01). A weaker positive correlation is associated with the current
86

spring precipitation, likely falling as snow, although no individual chronology represented by
this PC showed a significant correlation to this variable (Table 3.2, Figure 3.1). No climate
index was significantly correlated with this PC.

3.3.2.2 NWT
Tree growth within the Mackenzie River delta (NTPC1) was primarily affected by previous
summer temperatures, particularly the MWMT. This pattern was evident in negative
correlations to the previous summer heat-moisture index and continentality (difference
between warmest and coldest monthly mean temperatures, often inferred as a measure of
winter severity), as well as the beginning Julian date of the frost-free period.

Positive

correlations were weakly attributed to previous autumn and winter temperatures. No climate
index was significantly correlated to this PC (Table 3.2, Figure 3.1).

Tree growth in the foothills of the Richardson Mountains (NTPC2) was positively correlated
to current summer temperatures, including the number of frost-free days and continentality;
all of these variables were highly significant (p < 0.01). Previous summer frost-free period
also had a positive effect. Weaker negative correlations were attributed to previous MWMT
and MCMT variables, and previous season continentality. No climate index was significantly
correlated (Table 3.2, Figure 3.1).

Tree growth on the Peel River plateau (NTPC3) had the fewest significant correlations.
Previous summer temperatures were negatively correlated to growth, while previous winter
temperatures had an opposite positive effect. No climate index was significantly correlated
(Table 3.2, Figure 3.1).
87

3.4 Discussion
3.4.1 Climate and tree growth in Yukon
In general, YTPC1 variation (representing low slope gradients) appeared to be temperature
limited, while the high gradient group (YTPC2) appeared to be more limited by moisture,
supporting the finding in Chapter 2 that tree growth patterns do indeed appear to be
differentially affected by slope gradient. The strongest climate-growth associations for
YTPC1 appeared to favour warm temperatures in late summer in the prior season. Warmer
temperatures in the autumn would intrinsically be associated with a longer frost-free period
by delaying the onset of freezing temperatures late in the growing season. In turn, trees can
accumulate additional carbohydrate resources into the fall before shutting down for the
winter season, and utilize these resources the following year for more productive growth
(Kozlowski et al, 1991). This is important on low gradient slopes, where slower runoff rates
and decreased solar insolation levels favour the retention of soil moisture, in which case
potentially heightened photosynthetic activity would allow excess resources to be allocated
partially towards radial growth (Waring & Pitman, 1985; Barnes et al, 1998). Also for this
group (YTPC1), weaker positive correlations were observed for the mean temperature of the
warmest month (in this region, always July) and continentality (a measure of the disparity
between the warmest and coldest monthly mean temperatures). The association with
continentality was likely a reflection of the correlation to warmer summer temperatures;
however, winter temperatures are warming at a greater rate than summer temperatures
throughout Yukon, leading to a decreased continental effect (Figure 3.2). Thus, the positive
correlation to continentality was likely to be increasingly weakening in light of the
decreasing continentality trend.

88

Mean annual and summer precipitation also had a weak but negative effect on tree growth on
low gradients. Significant positive correlations to precipitation are occasionally found at
other circumpolar sites, but negative responses are extremely rare in the literature. In areas of
continuous permafrost, it is possible higher precipitation levels could accumulate in the
shallower active layers and temporarily waterlog sites on low gradients. This would be
particularly detrimental to white spruce, such as those in this region, which are not wellsuited to growing in hydric soil conditions. Low gradient slopes are also subject to lower
intensities of solar radiation due to the extremely obtuse angle of incoming radiation at high
latitudes, particularly on southerly aspects; wetter soils on these slope types are generally
colder during the growing season and require greater heating loads to warm, and will thus
conceivably be less prone to evaporative drying (Lee, 1964). Combined with low runoff
capability, these conditions promote greater moisture retention in the soil.

Forest growth on lower gradient slopes will potentially respond positively to predicted
warming temperatures in the Arctic. Conditions favourable to soil moisture holding capacity
may prevent drought-like conditions to prevail, and thus facilitate increased radial growth in
response to increasing growing season temperatures. However, it is possible that some trees
may not possess the adaptive capacity to positively respond to warming temperatures
(Callaghan et al, 2004b). Heightened temperatures may also impose evaporative demands
that overcome the available soil moisture (Beck et al, 2011), creating drought-like conditions
for individual trees. Though low gradient sites may support positive tree growth, it is likely
that future growth responses will vary according to the capabilities of individual trees to
respond.

89

Fewer significantly correlated climate variables were associated with YTPC2 (high
gradients). The strongest relationship was a negative correlation to summer temperatures in
the previous season. A weaker positive association with current year spring precipitation was
also evident, though this variable did not correlate significantly with any individual
chronology represented by this PC. Negative associations with temperature coupled with
positive precipitation correlations have also been reported in central Alaska and within the
eastern Mackenzie delta, though black spruce was the species under consideration in both
instances (Pisaric et al, 2007; Wilmking & Myers-Smith, 2008).

Steeper slopes are prone to increased rates of drainage due to gravity, and consequently are
likely to be moisture limited during the growing season (Bonan, 2008). They are also subject
to higher levels of solar radiation-induced evaporation, particularly at such high latitudes,
leading to increased moisture deficiencies as well (Barnes et al, 1998). High summer
temperatures are frequently correlated to overall drier conditions, meaning that trees at these
site types will have a reduced capacity to store adequate reserve carbohydrates to be used for
new growth in the following early growing season. Thus, the following season's growth will
be limited from the very beginning. However, greater precipitation in the spring (falling
predominantly as snow in this region) will normally extend the date of snowmelt, thus
effectively shortening the growing season on the early end. In this case, a shortened growing
season may diminish the period of water loss by extending the time of snow cover, which in
turn would negate some of the effects of diminished resource stores caused by prior growing
season temperature extremes. Increased duration and amount of snow cover will also create a
water reservoir for moderating early season soil moisture. However, with increasing
temperatures in the future, forest stands on steeper slopes may experience reduced growth as
90

the growing season becomes longer in response to earlier snow melt. Longer growing
seasons may ultimately exacerbate the already strong negative effect of prior season
temperatures.

The potential effects of solar radiation and seasonally continuous sunlight are suggested here,
however little research has gone into quantifying the precise nature of its effect on tree
growth at high-latitude areas. A specific band of solar radiation, UV-B, has been proposed as
a contributing factor to the 'divergence issue,' but only as a function of decreasing ozone
levels (Hansell et al, 1998; Callaghan et al, 2004a). There are many other factors that lead
to variable solar radiation levels, with incidence angle only one contributing factor. The
nature of incoming radiation, direct or diffuse, is largely affected by cloud cover and
prevailing atmospheric conditions; indeed, Young et al (1997) showed in the high arctic that
cloudy conditions diminished the effect of slope angle on overall solar input. However, cloud
cover is difficult to monitor and quantify over large areas in the field, with satellite measures
of outgoing long-wave radiation regarded as the only acceptable proxy measure for
cloudiness (Wheeler & Kiladis, 1999). Cloudiness could also be anecdotally inferred from
temperature trends, as cooler seasons would likely be characterized by increased cloud cover
and precipitation (e.g. Figure 3.4). Pohl et al (2006) indicated that fine-scale spatial
variability was highly important when considering solar radiation effects on the landscape,
supporting the notion that many factors, likely difficult to quantify, contribute to actual
received solar insolation at specific sites. However, the evidence from this study and in the
literature still seems to suggest that incidence angle and solar insolation play a significant
role in the multi-faceted system of tree growth and climate interactions.

91

In reported instances of the 'divergence effect,' slope gradient has not yet appeared as a
potential causal factor, though no definitive influence has been determined. However,
drought stress appears to be the leading hypothesis, and sites on steep gradients would
logically be more susceptible to moisture deficits. Though this proposition requires further
investigation, slope gradient appears to be a valid contributing factor towards the 'divergence
effect' through its potential ability to encourage drought-like conditions (see Results in
Chapter 2).

3.4.2 Climate and tree growth in the NWT
The climate-growth associations in the Northwest Territories were much more varied, as the
PC groups were more broadly defined by ecoregions identified in Chapter 2, rather than
finer-scale attributes such as slope gradient, artifacts of the more limited sample range in the
Yukon study area. Ecoregions encompass many distinct ecological characteristics, including
climate, physiography, soils, etc. (Ecological Classification Group, 2007). It is beyond the
scope of this study to link individual climate parameters to specific environmental
characteristics inherent to each ecoregion. However, it is clear that unique relationships
between climate and tree growth exist within distinct ecoregions.

NTPC1 (delta sites) correlated exclusively to prior season climate variables, a possible
indication that current year tree growth is heavily influenced by the conditioning of growth
traits in the prior growing season. The strongest, most significant of these was a negative
correlation with the mean warmest month temperature (again, July in this region) of the
previous year, coupled with slightly weaker negative correlations with other measures related
to growing season temperature. Ground temperatures in and around the delta are higher than
92

adjacent uplands due to the moderating effect of the abundant water bodies (Burn & Kokelj,
2009), which was likely a primary trait distinguishing NTPC1 from the other groups. Tree
root growth would be vigorous in a warm growing season coupled with the warmer soil
temperatures of this region, which may in turn prohibit the sensitive roots from achieving full
cold-hardiness during abrupt transitions into winter (Raghavendra, 1991). This would
manifest in the following growing season as the trees reallocate resources towards regrowing
or repairing damaged root shoots from the previous autumn, thus limiting overall radial
growth. Warmer summer temperatures also tend to result in lower streamflows in the
following season, which may indicate decreases in precipitation and snowpack reservoirs in
years of low streamflows. This was particularly evident in the NWT, where temperature data
from Inuvik was negatively correlated to the 30-year streamflow record of the adjacent east
channel of the Mackenzie River; similar patterns occur in the between the Fort McPherson
temperature record and streamflow records of the upper Peel River and the Arctic Red River
mouth (Figure 3.3).

Another possible contributing factor towards these significant climate relationships is the
desiccating effect of strong winter winds (Baig & Tranquillini, 1980). Winter winds blow
across the frozen Beaufort Sea ice, creating very cold and dry air conditions that can move
unimpeded onto the flat expanse of the Mackenzie delta region and exert a considerable
effect on tree growth and survival (Ecological Classification Group, 2007). Slow diffusion of
water vapour can occur through the closed stomates and cuticle of the trees' needles, and the
moisture cannot be replenished from the frozen ground until the active layer thaws in the
following growing season (Marchand, 1996). This may enhance the negative effect of
warmer and potentially more drought-prone conditions of previous growing seasons, and
93

contribute to a cumulative, prolonged water deficit that becomes apparent in the following
season's growth. However, winter desiccation damage is dependent on many interrelated
factors (Kozlowski et al, 1991), and likely plays a complementary role to other related
processes. Regardless, it would appear likely that wind plays some role in the climate and
tree growth relationships in this region. This hypothesis was not tested in this study, however
could potentially be a worthwhile avenue of further research on limits to tree growth at high
latitudes.

The climate associations here generally agree with the findings of Pisaric et al. (2007), where
a majority (75%) of their tree samples in the Mackenzie delta did not positively associate
with current summer temperatures. This group of "non responders" appeared to lose their
positive association with summer temperature in the 1930s, the same decade when our
analyses begin. The study area of Pisaric et al. was focused on the eastern side of the
Mackenzie delta, whereas our range encompassed much of the southern delta and Arctic Red
River plain. Though this study did not focus on individual tree responses, our similar results
still implicate the delta as a broad region where tree growth does not often respond positively
to summer temperatures, which may lead to continued growth decreases and increased
mortality in response to further warming in the delta region. The aforementioned possible
mechanisms behind this trend suggest that similar results would be found in other large river
deltas throughout the circumpolar region, although supporting evidence is not currently
available.

Sites in the Richardson foothills (NTPC2) exhibit strong and highly significant positive
correlations with current growing season temperatures, implicating these sites as being cold
94

limited. Sites in this area are much more prone to extreme weather affected by the mountains,
including high winds, colder temperatures, and increased precipitation due to orographic
effects (Figure 3.4). Likewise, ground temperatures are generally lower than the adjacent
lowlands in the delta, and soil active layer depths are likely more shallow (Burn & Kokelj,
2009). To compensate for such unfavourable growing conditions, carbohydrate reserves in
spruce are allocated primarily to belowground root systems, at the expense of radial stem
growth (Bonan, 2008; Crawford, 2008). This allows the trees to tolerate and survive
particularly cold years when normal photosynthesis would be diminished (Philipson, 1988).
Warmer years more conducive to growth would then likely lead to increased resource
allocations to less critical parts of the tree, such as radial growth. The conditions here are
affirmed by a different class II ecoregion delineation than the rest of the NWT sites, based on
significant changes in climate and physiography (Ecological Classification Group, 2007).
The highly significant correlations to summer temperature variables reflect the limited
growing conditions here, and is a classic example of temperature-sensitive sites frequently
cited in dendroclimatic studies.

The response of this group was similar to the results found of black spruce in Alaska and
Quebec (Wang et al, 2002; Lloyd et al, 2005). This suggests that black spruce growing
within the FT ecotone, exclusive of river deltas, may exhibit similar positive responses to
warming temperatures across North America, though this limited sample size precludes any
definitive conclusion. Sites like these may represent areas where expanded growth and
treeline advancement will occur.

95

The third PC group (NTPC3) represents the transition zone between the delta and adjacent
uplands. This PC did not significantly correlate with many climate parameters, creating a
somewhat ambiguous description of the conditions inherent to transitory areas such as this.
However, this PC group shows a somewhat similar pattern to the delta sites, primarily with
negative associations with previous summer temperatures. This suggests that droughtinducing higher summer temperatures stress trees in this area to a point where they cannot
store enough reserves to facilitate early, vigorous growth in the following summer season.
Higher summer and annual mean temperatures in the Fort McPherson area also correspond to
lower June streamflows in the Peel River during the following season, which could feasibly
create moisture limited growing conditions during the subsequent early growing season. The
sites represented by this PC, though distinct, resemble those of the delta (NTPC1) more so
than the upland sites (NTPC2), being farther removed from the effects of the mountains, and
closer to the large water network (e.g. Peel River) of the Mackenzie delta. Though direct
measures were not taken, it would be likely that soil temperatures are warmer here than in the
uplands, implicating similar mechanisms of climate/tree growth interactions inherent to
NTPC1. The relative lack of significant climate correlations highlights the transitory nature
of this PC, where two distinct environments converge and create a small but complex forestclimate system.

3.4.3 Conclusion
The results of this study show how climate can affect tree growth in different ways when
considered in varying contexts. Particular seasonal or monthly climate variables impart
effects on tree growth that may only be observed at specific scales. As the scale of
observation increases, some important climate variables evident at smaller scales may
96

become obscured by the increasing complexity of the climate and environmental system.
This has serious implications when attempting to reconstruct historical (and project future)
climate records using tree ring records, as a single climatic variable must be attributed as the
main source of ring width variation. The correct variable must be chosen based upon the
scale of the sample region, taking into consideration the level of environmental variation
within and between distinct ecoregions.

In Yukon, a single ecoregion (Taiga cordillera) was studied. Slope gradient appeared to be
the primary environmental trait affecting the response of tree growth to climate within this
specific area. Tree growth on steeper slopes appears to be more affected by moisture, as steep
gradients do not favour moisture retention through various aforementioned mechanisms.
Conversely, low gradient slopes were more prone to temperature sensitivity. Therefore,
future studies should identify sample sites with slope gradient in mind, and determine if
varying climate sensitivities are important to the goals of the research. If temperature
reconstruction is the primary goal, then it may be prudent to choose sites from lower gradient
slopes within a specific ecoregion. Slope gradient may or may not be the most important
landscape attribute contributing to tree growth variation, and thus a comprehensive
assessment of the entire landscape would be a reasonable follow up from this baseline study.
However, similar areas to the Taiga cordillera of northern Yukon across the circumpolar
treeline could similarly be affected by slope gradient and its effect upon the extreme solar
patterns inherent to high latitude regions.

Slope gradient may also prove to be a contributing factor in the 'divergence effect' evident
throughout the FT ecotone. Though the effects of gradient were not explicitly investigated in
97

this study, the findings suggest a possible mechanism that may partially explain disparate
growth responses to climate. Based on our conclusions, low-gradient slopes are likely to be
less prone to issues of divergence, and would thus be better suited sites for tree-ring based
historical temperature reconstructions. The findings of this study show a high degree of
variability of responses across the landscape, which would suggest a similar, wide range of
variability of responses to future climate changes.

Once the scale of observation includes multiple ecoregions with greater degrees of
environmental and climatic variation, such as in the NWT component of this study, smallscale landscape traits such as slope gradient become less important considerations. Although
this is likely due to more limited sampling power with each region (as compared to the
Yukon study area), we did find variable growth responses regardless of the spatial scale,
which underscores the highly complex nature of Arctic landscapes the associated growth
responses. The response of tree growth to climate within a broader context is the result of a
complex system of multiple environmental and climatic traits acting upon tree growth. At
this scale in the NWT, functional ecoregions become a primary factor determining distinct
tree growth patterns. Specifically, the distinct PCA grouping of all sites in the Mackenzie
delta suggest that river deltas throughout the circumpolar region are likely to harbour
disparate growing conditions apart from all other sub-arctic landscapes. Sites here are
strongly affected by warmer soil temperatures, deeper active layers, and greater moisture
availability due to the ubiquitous presence of water. It is not likely that these conditions
would be found anywhere outside of a river delta. Therefore, it may not be practical to
equally compare climatically influenced tree growth against that found in other non-delta
environments. Furthermore, river deltas are not practical locations for obtaining reliable
98

climate reconstructions, as tree growth largely appears to not respond positively to current
summer temperatures.

Tree growth responses seem to become more ambiguous and difficult to define where
ecoregions transition from one major type to the next. These transition zones may represent
areas of abrupt or gradual shifts between adjacent regions. This depends largely on the nature
of the criteria used to define ecoregions, such as topography, parent soil types, and
hydrological characteristics. Unless these zones are well documented and delineated, it is
best to avoid sampling trees near these zones, and focus on sites that are categorically
representative of the specific ecoregional traits.

Identifying the specific environment or landscape feature that strengthens the signal of a
specific, desired climate parameter is a necessary step for future dendroclimatic studies
conducted along the sub-arctic treeline. Taking these precautions to ensure a highly robust
and reliable climate record from tree rings will enhance our perception of current and future
changes within the FT ecotone and throughout the circumpolar region. Specific examples
have been shown for Yukon and the Northwest Territories, Canada, but it remains unclear if
the patterns observed here can be attributed to similar high-latitude areas. However, the
evidence suggests that addressing scale-dependent environmental variation can and should be
readily addressed in future studies of tree growth and climate change.

99

3.5 Literature Cited

Baig, M.N., & Tranquillini, W. (1980). The effects of wind and temperature on cuticular
transpiration of Picea abies and Pinus cembra and their significance in dessication
damage at the alpine treeline. Oecologia, 47(2), 252-256.
Barber, V.A., Juday, G.P., & Finney, B.P. (2000). Reduced growth of Alaskan white spruce
in the twentieth century from temperature-induced drought stress. Nature, 405, 668673.
Barnes, B.V., Zak, D.R., Denton, S.R., & Spurr, S.H. (1998). Forest Ecology (4th ed.). New
York: John Wiley & Sons, Inc.
Beck, P.S.A., Juday, G.P., Alix, C , Barber, V.A., Winslow, S.E., Sousa, E.E., et al. (2011).
Changes in forest productivity across Alaska consistent with biome shift. Ecology
Letters, doi: 10.111 l/j.1461-0248.2011.01598.x.
Bonan, G.B., & Sirois, L. (1992). Air temperature, tree growth, and the northern and southern
range limits to Picea mariana. Journal of Vegetation Science, 3, 495-506.
Bonan, G.B. (2008). Ecological Climatology: Concepts and Applications (2nd ed.).
Cambridge: Cambridge University Press.
Briffa, K.R., Schweingruber, F.H., Jones, P.D., Osborn, T.J., Shiyatov, S.G., & Vaganov,
E.A. (1998). Reduced sensitivity of recent tree-growth to temperature at high northern
latitudes. Nature, 391, 678-682.
Brooks, J.R., Flanagan, L.B., & Ehleringer, J.R. (1998). Responses of boreal conifers to
climate fluctuations: indications from tree-ring widths and carbon isotope analyses.
Canadian Journal of Forest Research, 25(4), 524-533.
Bunn, A.G. (2008). A dendrochronology program library in R (dplR). Dendrochronologia,
26(2), 115-124.
Burn, C.R., & Kokelj, S.V. (2009). The environment and permafrost of the Mackenzie Delta
area. Permafrost andPeriglacial Processes, 20(2), 83-105.

100

Callaghan, T.V., Bjorn, L.O., Chernov, Y., Chapin, T., Christensen, T.R., Huntley, B., et al.
(2004a). Responses to projected changes in climate and UV-B at the species level.
Ambio, 33(1), 418-435.
Callaghan, T.V., Bjorn, L.O., Chernov, Y., Chapin, T., Christensen, T.R., Huntley, B., et al.
(2004b). Biodiversity, distributions and adaptations of arctic species in the context of
environmental change. Ambio, 33(1), 404-417.
Cook, E.R., & Peters, K. (1981). The smoothing spline: a new approach to standardizing
forest interior tree-ring width series for dendroclimatic studies. Tree Ring Bulletin,
41, 45-54.
Cook, E.R., Briffa, K.R., Shiyatov, S., Mazepa, A., & Jones, P.D. (1990). Data Analysis. In
E. R. Cook & L. A. Kairiukstis (Eds.), Methods of Dendrochronology: Applications
in the Environmental Sciences (pp. 97-162). Dordrecht, the Netherlands: Kluwer
Academic Publishers.
Crawford, R.M.M. (2008). Plants at the Margin: Ecological Limits and Climate Change.
Cambridge: Cambridge University Press.
Daly, C , Gibson, W.P., Taylor, G.H., Johnson, G.L., & Pasteris, P. (2002). A knowledgebased approach to the statistical mapping of climate. Climate Research, 22(2), 99113.
D'Arrigo, R.D., Kaufman, R.K., Davi, N., Jacoby, G.C, Laskowski, C , Myneni, R.B., et al.
(2004). Thresholds for warming-induced growth decline at elevational tree line in the
Yukon Territory, Canada. Global Biogeochemical Cycles, 18(3), GB3021,
doi:3010.1029/2004GB002249.
D'Arrigo, R.D., Wilson, R., Liepert, B., & Cherubini, P. (2008). On the 'Divergence Problem'
in Northern Forests: A review of the tree-ring evidence and possible causes. Global
and Planetary Change, 60(3-4), 289-305.
D'Arrigo, R.D., Jacoby, G.C, Buckley, B., Sakulich, J., Frank, D., Wilson, R., et al. (2009).
Tree growth and inferred temperature variability at the North American Arctic
treeline. Global and Planetary Change, 65(1-2), 71-82.
Davi, N.K., Jacoby, G.C, & Wiles, G.C. (2003). Boreal temperature variability inferred from
maximum latewood density and tree-ring width data, Wrangell Mountain region,
Alaska. Quaternary Research, 60(3), 252-262.
101

Driscoll, W.W., Wiles, G.C, D'Arrigo, R.D., & Wilmking, M. (2005). Divergent tree growth
response to recent climatic warming, Lake Clark National Park and Preserve, Alaska.
Geophysical Research Letters, 32(20), L20703, doi:20710.21029/22005GL024258.
Duk-Rodkin, A., Barendregt, R.W., Froese, D.G., Weber, F., Enkin, R., Rod Smith, I., et al.
(2004). Timing and extent of Plio-Pleistocene glaciations in north-western Canada
and east-central Alaska. In J. E. a. P. L. Gibbard (Ed.), Developments in Quaternary
Science: Quaternary Glaciations-Extent and Chronology-Part II: North America (1st
ed., Vol. 2, pp. 313-345). Amsterdam: Elsevier.
Ecosystem Classification Group. (2007). Ecological Regions of the Northwest Territories Taiga Plains. Yellowknife, NT: Department of Environment and Natural Resources,
Government of the Northwest Territories.
Esper, J., & Frank, D. (2009). Divergence pitfalls in tree-ring research. Climatic Change,
94(3-4), 261-266.
Fleming, R.A., & Volney, W.J.A. (1995). Effects of climate-change on insect defoliator
population processes in Canada boreal forest - some plausible scenarios. Water Air
and Soil Pollution, 52(1-2), 445-454.
Fritts, H.C. (1976). Tree Rings and Climate. Caldwell: The Blackburn Press.
Griesbauer, H.P., & Green, D.S. (2010a). Regional and ecological patterns in interior
Douglas-fir climate-growth relationships in British Columbia, Canada. Canadian
Journal of Forest Research, 40(2), 308-321.
Griesbauer, H.P., & Green, D.S. (2010b). Assessing the climatic sensitivity of Douglas-fir at
its northern range margins in British Columbia, Canada. Trees-Structure and
Function, 24(2), 375-389.
Hansell, R.I.C, Malcolm, J.R., Welch, H., Jefferies, R.L., & Scott, P.A. (1998). Atmospheric
change and biodiversity in the Arctic. Environmental Monitoring and Assessment,
49(2-3), 303-325.
Hogg, E.H., & Hurdle, P.A. (1995). The aspen parkland in western Canada - a dry-climate
analog for the future boreal forest. Water Air and Soil Pollution, 52(1-2), 391-400.

102

Holmes, R.L. (1983). Computer-assisted quality control in tree-ring dating and
measurement. Tucson: University of Arizona.
Holtmeier, F.K., & Broil, G. (2005). Sensitivity and response of northern hemisphere
altitudinal and polar treelines to environmental change at landscape and local scales.
Global Ecology and Biogeography, 14, 395-410.
Jacoby, G.C, Lovelius, N.V., Shumilov, O.I., Raspopov, O.M., Karbainov, J.M., & Frank,
D.C (2000). Long-term temperature trends and tree growth in the Taymir region of
northern Siberia. Quaternary Research, 53(3), 312-318.
Juday, G.P., Barber, V.A., Duffy, P., Linderholm, H., Rupp, T.S., Sparrow, S., et al. (2005).
Forests, land management, and agriculture. In Arctic Climate Impact Assessment (pp.
781-862). Cambridge: Cambridge University Press.
Kirdyanov, A., Hughes, M., Vaganov, E., Schweingruber, F., & Silkin, P. (2003). The
importance of early summer temperature and date of snow melt for tree growth in the
Siberian Subarctic. Trees, 17, 61-69.
Kozlowski, T.T., Kramer, P. J., & Pallardy, S.G. (1991). The Physiological Ecology of Woody
Plants. San Diego: Academic Press, Inc.
Lee, R. (1964). Potential insolation as a topoclimatic characteristic of drainage basins.
International Association of Scientific Hydrology. Bulletin, 9(1), 27 - 41.
Lloyd, A.H., Wilson, A.E., Fastie, C.L., & Landis, R.M. (2005). Population dynamics of
black spruce and white spruce near the arctic tree line in the southern Brooks Range,
Alaska. Canadian Journal of Forest Research, 35(9), 2073-2081.
Lloyd, A.H., & Bunn, A.G. (2007). Responses of the circumpolar boreal forest to 20th
century climate variability. Environmental Research Letters, 2(4), 1-13.
Loehle, C. (2009). A mathematical analysis of the divergence problem in dendroclimatology.
Climatic Change, 94(3-4), 233-245.
Luckman, B.H., & Wilson, R.J.S. (2005). Summer temperatures in the Canadian Rockies
during the last millennium: a revised record. Climate Dynamics, 24(2-3), 131-144.

103

MacDonald, G.M., Szeicz, J.M., Claricoates, J., & Dale, K.A. (1998). Response of the central
Canadian treeline to recent climatic changes. Annals of the Association of American
Geographers, 88(2), 183-208.
MacDonald, G.M., Kremenetski, K.V., & Beilman, D.W. (2008). Climate change and the
northern Russian treeline zone. Philosophical Transactions of the Royal Society BBiological Sciences, 353(1501), 2285-2299.
Marchand, P.J. (1996). Life in the Cold: An Introduction to Winter Ecology (3rd ed.).
Hanover: University Press of New England.
Mbogga, M.S., Hamann, A., & Wang, T.L. (2009). Historical and projected climate data for
natural resource management in western Canada. Agricultural and Forest
Meteorology, 149(5), 881-890.
Miyamoto, Y., Griesbauer, H.P., & Green, D.S. (2010). Growth responses of three coexisting
conifer species to climate across wide geographic and climate ranges in Yukon and
British Columbia. Forest Ecology and Management, 259(3), 514-523.
Ogden, A. (2008). Yukon Forestry Monitoring Program: Field Manual & Monitoring
Protocol (1 ed.). Whitehorse: Government of Yukon.
O'Neill, G.A., Hamann, A., & Wang, T. (2008). Accounting for population variation
improves estimates of the impact of climate change on species' growth and
distribution. Journal of Applied Ecology, 45(4), 1040-1049.
Peters, K., Jacoby, G.C, & Cook, E.R. (1981). Principal Components Analysis of Tree-Ring
Sites. Tree-Ring Bulletin, 41, 1-19.
Philipson, J.J. (1988). Root-growth in sitka spruce and Douglas-fir transplants - dependence
on the shoot and stored carbohydrates. Tree Physiology, 4(2), 101-108.
Pisaric, M.F.J., Carey, S.K., Kokelj, S.V., & Youngblut, D. (2007). Anomalous 20th century
tree growth, Mackenzie delta, Northwest Territories, Canada. Geophysical Research
Letters, 34(5), L05714, doi:05710.01029/02006GL029139.
Pohl, S., Marsh, P., & Pletroniro, A. (2006). Spatial-temporal variability in solar radiation
during spring snowmelt. Nordic Hydrology, 37(1), 1-19.

104

Raghavendra, A.S. (Ed.). (1991). Physiology of Trees. New York: John Wiley & Sons, Inc.
Revelle, W. (2011). psych: Procedures for Personality and Psychological Research (Version
1.0-94). Evanston, Illinois: Northwestern University.
Smith, C.A., Meikle, J.C, & Roots, C F . (2004). Ecoregions of the Yukon Territory:
Biophysical properties of Yukon landscapes (PARC Technical Bulletin No. 04-01).
Summerland, British Columbia: Agriculture and Agri-Food Canada.
Stoehr, M.U., Ukrainetz, N.K, Hayton, L.K., & Yanchuk, A.D. (2009). Current and future
trends in juvenile wood density for coastal Douglas-fir. Canadian Journal of Forest
Research, 39(1), 1415-1419.
Stokes, M.A., & Smiley, T.L. (1968). An Introduction to Tree-Ring Dating. Tucson: The
University of Arizona Press.
Szeicz, J.M., & MacDonald, G.M. (1995). Dendroclimatic reconstruction of summer
temperatures in northwestern Canada since A.D. 1638 based on age-dependent
modeling. Quaternary Research, 44, 257-266.
Szeicz, J.M., & MacDonald, G.M. (1996). A 930-year ring-width chronology from moisturesensitive white spruce (Picea glauca Moench) in northwestern Canada. Holocene,
6(3), 345-351.
Tabachnick, B.G., & Fidell, L.S. (1989). Using Multivariate Statistics (2 ed.). New York:
HarperCollinsPublishers, Inc.
Tardif, J., & Bergeron, Y. (2001). Comparative analysis of the climatic response of seven
boreal tree species from northwestern Quebec, Canada. Tree-Ring Research, 57(2),
169-181.
Vaganov, E.A., Hughes, M.K., Kirdyanov, A.V., Schweingruber, F.H., & Silkin, P.P. (1999).
Influence of snowfall and melt timing on tree growth in subarctic Eurasia. Nature,
400(6140), 149-151.
Venables, W.N., & Ripley, B.D. (2002). Modern Applied Statistics with S (4 ed.). New York:
Springer.

105

Wahl, H.E., Fraser, D.B., Harvey, R.C, & Maxwell, J.B. (1987). Climate of Yukon. Ottawa:
Environment Canada.
Wang, L., Payette, S., & Begin, Y. (2002). Relationships between anatomical and
densitometric characteristics of black spruce and summer temperature at tree line in
northern Quebec. Canadian Journal of Forest Research, 32(3), 477-486.
Wang, T., Hamann, A., Spittlehouse, D.L., & Aitken, S.N. (2006). Development of scale-free
climate data for western Canada for use in resource management. International
Journal of Climatology, 26(3), 383-397.
Waring, R.H., & Pitman, G.B. (1985). Modifying lodgepole pine stands to change
susceptibility to mountain pine-beetle attack. Ecology, 66(3), 889-897.
Wheeler, M., & Kiladis, G.N. (1999). Convectively coupled equatorial waves: Analysis of
clouds and temperature in the wavenumber-frequency domain. Journal of the
Atmospheric Sciences, 56(3), 374-399.
Wilmking, M., Juday, G.P., Barber, V.A., & Zald, H.J. (2004). Recent climate warming
forces contrasting growth responses of white spruce at treeline in Alaska through
temperature thresholds. Global Change Biology, 10, 1724-1736.
Wilmking, M., & Juday, G.P. (2005). Longitudinal variation of radial growth at Alaska's
northern treeline - recent changes and possible scenarios for the 21st century. Global
and Planetary Change, 47, 282-300.
Wilmking, M., & Myers-Smith, I. (2008). Changing climate sensitivity of black spruce
(Picea mariana Mill.) in a peatland-forest landscape in Interior Alaska.
Dendrochronologia, 25(3), 167-175.
Wilson, R., D'Arrigo, R.D., Buckley, B., Bantgen, U., Esper, J., Frank, D., et al. (2007). A
matter of divergence: Tracking recent warming at hemispheric scales using tree ring
fate. Journal of Geophysical Research, 112, doi: 10.1029/2006JD008318.
Young, K.L., Woo, M.K., & Edlund, S.A. (1997). Influence of local topography, soils, and
vegetation on microclimate and hydrology at a high Arctic site, Ellesmere Island,
Canada. Arctic and Alpine Research, 29(3), 270-284.

106

Yukon Ecoregions Working Group. (2004). Yukon Taiga Cordillera. In C. A. Smith, J. C.
Meikle & C. F. Roots (Eds.), Ecoregions of the Yukon Territory: Biophysical
properties of Yukon landscapes (pp. 101-106). Summerland, British Columbia:
Agriculture and Agri-Food Canada.

Table 3.1 Climate station monthly mean temperature correlations
Old Crow
Rock River
Inuvik
Jan
Jul
Jan
Jul
Jan
Jul
Dawson City
.633**
.556**
.607*
.945**
.449**
.590**
Old Crow
.599
.887**
.860**
.809**
Rock River
.825**
.934**
** = p<0.01 * = p<0.05

108

Table 3.2 Seasonal climate & PCA correlations
Yukon
YTPC1
Annual
Mean precipitation
ENSOf
Continentality
Previous Annual
Continentality
Spring
Precipitation
Summer
MSP
Precipitation
Max temperature
Mean temperature
Min temperature
MWMT
NFFD
Previous Summer
Max temperature
Mean temperature
Min temperature
MWMT
HM index
FFP
bFFP
eFFP
Previous Autumn
Max temperature
Mean temperature
Min temperature
Winter
MCMT
Previous Winter
Max temperature
Mean temperature
Min temperature

YTPC2

Northwest Territories
NTPC1
NTPC2
NTPC3

-.228*
-.260*
.263*

.360**
-.266*

-.248*

.220*
-.211*
-.244*
.307**
.340**
.341**
.381**
.293**

.231*

-.330**
-.317**
-.249*
-.244*

-.281*
-.267*
-.298**
-.278*

.283**

-.251*
.246*

-.264*
.307**
.259*
.284**
.296**

.233*
-.225*
.236*
.228*
.233*

.245*
.248*

-.227*
Precipitation
** = p < 0.01; * = p < 0.05; MSP = mean summer precipitation (May - Sept); MWMT = mean warmest
monthly temperature; MCMT = mean coldest monthly temperature; HM index = heat-moisture index, a measure
of drought; NFFD = number of frost free days; FFP = frost free period (continuous); bFFP = first Julian day of
frost free period; eFFP = last Julian day of frost free period

109

Figure 3.1 Current and prior year monthly climate correlations
YTPC1 Monthly Mean
Temperature Correlations
04

§

00

02

tI n

02

01

03

Current Year
Prior Year

03

§
o

YTPC2 Monthly Mean
Temperature Correlations

H^L

-0 1

01
0)
o

00

8

01

o

. r ra W¥

-0 3 - ^ H Current Year
I
I Prior Year
-0 4
~i
1
1
r

03
O

<S

Q.

ft

3

u- S < 5

^

=3

<

8
a

CD

CO O

ro (U <9 9- ro E
-> Li- 2 < 2 ->

04

03
02

g
o

-0 1

n n

3
<

l) i? o
CO O z

ctf
Q

^ H Current Year
I
I Prior Year

llf t

02 H

01

Jl

-i

NTPC2 Monthly Mean
Temperature Correlations

NTPC1 Monthly Mean
Temperature Correlations

00

ll. I

-0 2

-0 2

0J
o
it:

I

9>
o
it:
§
o

00

TJ

jJl

-0 2
03

02

• • Current Year
I
I Prior Year

04

5

<

3

tu

<

CO <-> z

j :

>-»

iw

D

a, ro o. oj
ro
-> IL 5 < 5

3
-i

OJ
=3

CL
CD

>
O

O
CD

<

CO <-> Z

O

-ft
>?

NTPC3 Monthly Mean
Temperature Correlations
03

uT

02
01
CD
o
!t=
§
o

00
01

PT

02
03
-0 4 •

"tr

| ^ H Current Year
I
I Prior Year
§ •§ ro o. 8" §
-J

li_ 2

<

2

">

on Q. t l
3
CD A!

<

CO O

5
O

Z

o
CD

Q

Dashed lines indicate 95% significance level Climate data was derived from ClimateWNA v 4 521

Figure 3.2 Temperature and continentality trends in Yukon study area
YT Mean Summer Temperature
f =-15.004+.013x

2000

1900

YT Mean Winter Temperature
f =-61.8+.021x

o
CD
CD

O)
CD

Q

2000

YT Continentality
f = 56.201 -.0095x
50
48
46
44 42
40
38 36
34
32
30
1900

1920

1940

1960

1980

2000

Year
Both summer and winter mean temperatures are increasing in northern Yukon near our study sites, though
winter temperatures are accelerating more rapidly This results in a decreasing continentality trend, as the
difference between extreme seasonal temperatures grows less (Data obtained from ClimateWNA v 4 52)

111

Figure 3.3 Concurrent river flows and temperature trends in the NWT
Arctic Red River Discharge (mouth)
r = -.430 (p = 0.01)
1200
1000
"g
o
CD
-52
"E

16

River Discharge (August)
Avg Summer Temperature (Ft McPherson)
A

\

800 -

15

'\
I.

14

'

13

600

1

400
200
0

12

Q

V

11

10
9
1980

1970

O
£
a>
§>

2000

1990

2010

Peel River Discharge (above Ft. McPherson)
r = -.472 (p = 0.005)

-C

2400
2200
2000
1800

§

1600

8

1400 -

River Discharge (August)
Avg Sunpmer Temperature (Ft McPherson)

16
t\

- 15

I\
I \

14
13
12

CO

<oE

1200 1000
800
600
400

11
10
9
1980

1970

1990

2000

2010

Mackenzie River Discharge (east channel)
r =-.433 (p = 0.012)
450

16

River Discharge (August)

400
-a
c:
o

350
300
250
200
150
100

1970

1980

1990

2000

2010

River flows frequently negatively reflect the concurrent mean summer temperatures. Warm summers will often
be accompanied by lower moisture levels, indicated by decreased river flows, as a result of decreased
precipitation, earlier and faster snowmelt, and/or increased evaporation. (Hydrologic data obtained from the
Water Survey of Canada (Environment Canada), http://www.wsc.ec.gc.ca/applications/H20/indexeng.cfrn?stype=station)

112

Figure 3.4 Climate trends along the Dempster Highway, NWT
2008 mean summer temperatures
along Dempster Highway, NWT

2008 mean summer PAR along
Dempster Highway, NWT
500

18
• H i June
1
1 July

16

1

14

1 August

400

12
O

co
<D
CP

CD
Q

°

300

1-1

-in
lu

E
-

1-1

r]

200

6
100

4
2
0

36

414

625

Elevation (meters)

36

414

625

Elevation (meters)

Data derived from mobile HOBO climate stations during the summer of 2008, showing decreasing temperatures
and PAR (photosynthetically active radiation) with increasing elevation. The low elevation (36 m) station was
located at the Fort McPherson airport; the mid-elevation station was located near the sites represented by
NTPC2; and the high elevation station is located within the Richardson Mountains near Wright Pass. The
decreasing PAR is likely indicative of increased cloudiness near the mountains due to orographic effects.

113

Chapter 4: Summary
4.1 Introduction
This study aimed to partially fill knowledge gaps concerning radial tree growth and climate
at the latitudinal treeline in northern Yukon and Northwest Territories. Tree ring
chronologies were sampled in two adjacent, environmentally distinct regions within the
forest-tundra ecotone in order to explore the importance environmental variation has on the
climatic growth responses of spruce. Specifically, we sought to a) examine radial tree growth
patterns across an environmental gradient in northern Canada; b) identify key environmental
features that affect how tree growth responds to climate within the forest-tundra ecotone
(Chapter 2); and c) identify important climatic drivers of tree growth as they relate to the
environmental features established in the previous step (Chapter 3). Additionally, we
assessed the feasibility of using modelled climate data in lieu of climate station data, in order
to address issues related to short record time spans, missing values, and long distances
between stations and study sites.

4.2 Key results and implications
Our first research question regarded how climate and tree growth relationships change across
an environmental gradient at the northern treeline in northwestern Canada. We tested the
hypothesis that tree growth would react to climate differently between and within the two
study areas, and found that environmental variation can affect the way trees respond to
climate at the latitudinal treeline. Following this, we asked what particular environmental
characteristics determine the way tree growth responds to climate, hypothesizing that growth
response would be dependent on landscape features and spatial scales unique to each study
area and sampling outcome. We found that growth patterns were indeed affected by the
114

spatial scale at which they were observed, while ecoregional delineations, including
boundary zones, appear to be an important determinant of climate and tree growth
relationships.

In the NWT study area, growth variation was distinctly divided between two adjacent
ecoregions and a third transition area along the boundary between the two regions, which
suggests that the bioclimatic, topographic, and hydrologic conditions unique to any given
northern ecoregion, combine to create distinctive tree-ring records of climate. This is clearly
evident in the study area, as the Mackenzie River delta environment and climate differs
greatly from the surrounding ecoregions.

In the Yukon study area, one ecoregion was sampled much more thoroughly than those in the
NWT study area. Here, the underlying environmental characteristics of this ecoregion were
theoretically similar at all sites, thus enabling the effect of other sub-ecoregional
environmental traits to become apparent in the tree-ring records. Specifically, slope gradient
was found to be a primary determinant of tree growth response to climate, an important effect
that is apparent at a local spatial scale, but is likely obscured - possibly by differences in
sampling intensity in this study - at larger, regional scales, such as the NWT study area.
Steeper slopes likely exacerbate moisture deficiencies, as they receive more direct solar
insolation that in turn warms (and dries) soils. Consequently, trees growing on steeper slopes
in high latitude regions will tend to reflect moisture limitations in their tree-ring patterns.
Conversely, lower gradient slopes tend to have deeper soil active layers, more soil moisture,
and less solar heating, and trees on these sites are more likely to respond to temperature
limitations.
115

Thirdly, we aimed to illuminate the most influential climate components to tree growth
within the context of the environmental variability identified in the previous steps. We
hypothesized that summer temperature is not the only climate parameter to have a significant
effect on growth, as the general paradigm of the climatic control of the northern treeline
suggests. We found that while summer temperature does impart a substantial influence on
tree growth, other variables such as prior season autumn temperatures and frost free period
(i.e. growing season length) often exacted more significant impacts based on certain
environmental settings. Tree growth on high gradient slopes appeared to be more sensitive to
moisture deficits, the steeper slope angles seemed to hold less soil moisture, while
simultaneously being prone to more direct (i.e. increased) solar heating and drying.
Traditional summer temperature limitations were found in the mountain foothills in the NWT
study area, but the effect was lessened within the Mackenzie River delta valley, where
slightly more favourable growing conditions tended to promote negative, lagged relationships
with summer temperatures. This finding underscores the importance of recognizing the
variability of climate and tree growth relationships, and challenges the notion that summer
temperature is the primary control throughout the entire circumpolar region.

Last, we asked how variable landscapes may contribute to cases of disparate tree growth,
where the commonly-held assumption of dominant summer temperature controls was not
evident. We hypothesized that previously ^identified climatic influences, namely drought,
could be attributed to variable landscapes. We found evidence suggesting that environmental
variation may in fact be a contributing cause to previously reported instances of disparate
growth. The results from the Yukon study area suggest that slope gradient is an important
116

consideration when choosing sites for polar-region climate reconstructions. Even within a
single ecoregion, choosing sites from a wide range of slope gradients will lead to a variety of
climate signals amongst the tree-ring records, which may potentially create spurious and
undesirable results, either with weak, washed-out climate signals or altogether false
representations of the actual conditions. This could prove to be an underlying cause of the
reported 'divergence problem' at many northern sites. Many tree-ring studies have reported
disparate growth responses to summer temperatures, but no definitive cause has yet been
found, though it is likely a result of many confounding factors. Thus far, the solution has
been to identify individual trees with positive summer temperature correlations to be used for
climate reconstructions and predictions. We suggest that by restricting the sampling strategy
to low gradient slopes, as well as avoiding climatic "oases" such as large river deltas and
their immediate surrounding landscapes, more positively responding (to summer
temperature) trees will be present. This likely will not fully alleviate the problem of
encountering disparate growth trends; however, it is a feasible, better informed approach that
is less likely to produce inaccurate observations of paleoclimates. Factoring in slope gradient
along with already-established standard site selections criteria, such as maintaining similar
aspects, species, and stem densities, while keeping in mind the suitability and potential
problems of sampling within a given ecoregion, is an essential step for future
dendrochronology studies conducted in high-latitude environments.

In addition to the specific research questions, our findings also indicate a high degree of
spatial complexity within and between high latitude environments. Variable tree growth was
reflected by the degree of biophysical and climatic variability inherent to various spatial
scales. This is a critical notion when considering the response of the northern treeline to
117

climate change, as it is likely to exhibit a range of responses of radial growth, densification,
and recruitment across the entire circumpolar region. Future studies that aim to reconstruct
paleoclimates and make projections about future treeline conditions need to consider to
spatial scale from which they draw their conclusions. If the goal is to make general
assessments and predictions for broad regions, countries or continents, then using the
ecoregion as a unit of observation would be most useful and informative. However, for local
mitigation planning and management purposes, smaller scale monitoring would be more
appropriate. This would require a more thorough understanding of the environment of a
specific ecoregion, and acknowledge that forests will still exhibit variable responses to
climate. For instance, slope gradient was found to be a primary environmental trait affecting
tree growth, although it is unclear if this remains true for other regions in the North.
However, because solar incidence angles are extreme at high latitudes, it is likely that slope
gradient would be an important consideration in most if not all northern sites.

Finally, this study demonstrates the potential utility of modelled climate data in areas without
long-term, representative climate station data. Uncertainties do exist regarding the accuracy
of the modelled data, but it appeared to be an adequate representation of local conditions. For
this study, absolute values were not as important as was capturing trends in the climate data
that could be reflected in the tree-ring chronologies.

4.3 Limitations and future research
This study was conducted in an area that has received relatively little attention from the
scientific community, particularly researchers observing the latitudinal treeline. Much of the
supporting research used in support of our study was conducted in Alaska or eastern Canada.
118

Though similar in regards to general characteristics of sub-Arctic North America, they are
still relatively distinct from our study area. As such, little precedence has been established for
dendroclimatological studies in northern Yukon and NWT. Therefore our investigation was
relatively exploratory in nature, lacking substantial prior knowledge or experience in this
specific region of Canada. As a result, our observations and conclusions are relatively broad,
but will hopefully serve as a viable platform from which future investigations will originate.

Our sampling strategy was also hampered by being mostly confined to the course of the
Dempster Highway. As a result, we were required to sample more intensively within a single
ecoregion in Yukon, allowing us to observe patterns at a smaller, localized scale. In the NWT
however, the highway spanned a greater variety of ecoregions, thus allowing us to capture a
larger range of conditions, at the expense of intensive sampling similar to that in the Yukon
study area. This sampling strategy provided both opportunities and limitations for our
analyses and subsequent interpretations, allowing us to find significant evidence of the role
of environmental variation within different spatial scales, but leaving us unable to compare
the two study areas on an equal basis. Our findings would be greatly substantiated by
applying the sampling strategy from each region to the other; that is, look for and compare
environmental attributes that operate at similar scales across both regions. This would enable
us to better observe spatial transitions of environment and tree growth in this area. However,
doing this would require much more time, effort and funding, as areas beyond the highway
are much more difficult and time consuming to access.

Also, this study was limited by a lack of sufficient, reliable and representative climate data.
The nearest long-term (e.g. longer than 80 years, about the length of our tree ring
119

chronologies) climate station was located over 400 km away in Dawson City, Yukon, which
we determined would not accurately reflect the conditions in the forest-tundra interface in
northern Yukon, let alone across the mountains in the Mackenzie delta region. Other stations
closer by were either relatively short time spans, replete with missing values, located in
environments much different that the study areas (e.g. Old Crow flats), were not accurately
adjusted for changes in measurement devices or techniques, or more often a combination of
most of these factors. Therefore we used modelled climate data from ClimateWNA v.4.52, a
model that was developed in British Columbia and recently updated to include the entire
western North America. It may be likely that the authors of the model did or could not
account for the extreme and variable nature of climate in the Arctic, but we believe that it
could still provide a relatively accurate representation of local conditions. However, as with
any climate study in the far North, there is a large degree of uncertainty regarding the exact
values and precise behaviour of climate, and thus conclusions based on any source of Arctic
climate data should be regarded with a degree of reservation.

This study could be further enhanced by utilizing the absolute ring width measurements and
converting them into basal area measurements, as opposed to converting widths to unit-less
indices as was done in this study. Measurements were standardized in order to better observe
and analyze relative growth trends between sites. Using basal area measurements would
provide a means to assess the actual physical growth gains and declines, which in turn could
facilitate discussions and conclusions regarding the carbon cycling capacity of the boreal
forest. As well, a tangible measure of growth trends would be more useful for forest
managers and users, as it would more accurately represent the physical changes occurring on
the landscape.
120

Much opportunity exists to expand upon this research. Aside from increasing the scope and
sampling power of this initial study, one relevant topic would be to quantify the effect of
various aspects of environmental variation on divergent growth responses. This would entail
categorizing individual trees based on their response to summer temperatures, and then
looking for group differences based on slope gradient or ecoregion. This would lend credible
evidence towards a determination of the driving mechanisms of divergent tree growth.
Another useful future endeavour would be observing the climate and tree growth trends
through time, and assessing the relative stability of these relationships. These temporal trends
could be linked to larger scale climatic patterns, including general circulation patterns such as
the Arctic Oscillation or the El Nino-Southern Oscillation indices. This would enable better
predictive modelling and projections of future conditions by incorporating a greater degree of
variables in the Arctic biosphere.

4.4 Conclusions
This study illustrated the complex nature of climate and tree growth relationships as they
occur amongst varying environments and spatial scales in the Arctic. Our results support an
improved understanding of the northern treeline, and will contribute to the fast expanding
knowledge of latitudinal treeline dynamics. Arctic ecosystems are expected to continue
changing at a rapid pace in response to concurrent climatic change, which will necessitate
prompt, informed, and versatile planning and management efforts to ensure successful
adaptation and mitigation strategies. The Arctic treeline remains a relatively underdeveloped
area of research, yet it will serve as a sentinel of coming environmental change. Thus, it will

121

prove sensible to invest further efforts into understanding the complex ecology of this critical
component of the Arctic biome.

122