Genecology of 20 Paper Birch (Betula papyrifera Marsh.)
Provenances from British Columbia and Northern Idaho

Nicole Balliet
B.S.F., University of British Columbia, 1994

Thesis Submitted in Partial Fulfillment of
The Requirements for the Degree of
Master of Science
in
Natural Resources and Environmental Studies
(Forestry)

The University of Northern British Columbia
April 2009

© Nicole Balliet, 2009

1*1

Library and
Archives Canada

Bibliotheque et
Archives Canada

Published Heritage
Branch

Direction du
Patrimoine de I'edition

395 Wellington Street
Ottawa ON K1A0N4
Canada

395, rue Wellington
Ottawa ON K1A0N4
Canada

Your file Votre reference
ISBN: 978-0-494-48783-9
Our file Notre reference
ISBN: 978-0-494-48783-9

NOTICE:
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.

AVIS:
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 Plntemet, prefer,
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 imprimes 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.

Canada

ABSTRACT

Twenty provenances of paper birch (Betula papyrifera Marsh) were collected from five
regions in BC and Idaho. Seedlings were grown at three nurseries and planted in three
common gardens in BC and Idaho.

Geographic variation in the timing of bud burst in paper birch is under genetic and
environmental control. It follows climatic clines based on latitude, longitude and
elevation. The signal for the onset of spring bud flush is determined by an interaction
between air and soil temperature and photoperiod. Long distance displacement of provenances from their site of origin can be detrimental to phenology, survival and
growth. Differences in stock handling among nurseries and nursery displacement effects
also influenced growth and survival of some provenances.

When developing seed zones and seed transfer guidelines for paper birch in BC, these
factors need to be considered. The impacts of future climate change on birch deployment
must also be considered.

TABLE OF CONTENTS

Abstract

ii

Table of Contents

iii

List of Tables

v

List of Figures

vi

Acknowledgement

viii

Chapter One Introduction and Literature Review
1.1 Introduction
1.2 Purpose of Research
1.3 Development of Genecology
1.4 Geographic Variation in Paper Birch
1.5 Tree Improvement, Seed Zones and Seed Transfer Guidelines
1.6 Nursery Effects
1.7 Phenology
1.7.1 Bud Flush
1.7.1.1 Impacts of Climate Change
1.7.1.2 Birch Dieback
1.8 Height Growth
1.9 Conclusion

1
3
4
6
7
9
10
11
13
14
15
17

Chapter Two Research Design
2.1 Birch Genecology Study - Overall Study (48 Paper Birch Provenances)
2.2 Birch Genecology Study - 20 Paper Birch Provenances
2.3 Tables and Figures

18
19
21

Chapter Three Bud Flush
3.1 Introduction
3.2 Methods
3.2.1 Sample Population
3.2.1.1 Red Rock
3.2.1.2 Skimikin and Idaho
3.2.2 Survey
3.2.3 Climatic Data
3.2.4 Statistical Methods
3.3 Results
3.4 Discussion
3.4.1 Garden Effects
3.4.2 Regional Effects
3.4.3 Nursery Effects

25
27
27
27
27
28
28
29
31
32
33
36
39

111

3.4.4 Garden by Region Interaction
3.4.5 Nursery by Region Interaction
3.4.6 Conclusion and Research Pitfalls
3.5 Tables and Figures
Chapter Four Height Growth
4.1 Introduction
4.2 Methods
4.2.1 Sample Population
4.2.1.1 Red Rock
4.2.1.2 Skimikin
4.2.2 Survey
4.2.3 Statistical Methods
4.3 Results
4.4 Discussion
4.4.1 Garden Effects
4.4.2 Regional Effects
4.4.3 Garden by Nursery Interaction
4.4.4 Garden by Region Interaction
4.4.5 Conclusion

4.5 Tables and Figures

40
40
41
43
49
52
52
52
52
53
53
54
55
55
57
59
60
61
64

Chapter Five Summary Discussion
5.1 Summary of Key Findings
5.2 Seed Zones and Seed Transfer Guidelines
5.3 Potential Impacts of Climate Change

69
71
72

References

76

Appendices
Appendix A
Appendix B

Results from Initial Bud Flush and Height Growth Analysis
Results of 2001 and 2002 Bud Flush Study

IV

83
87

LIST OF TABLES

Table 2. la
Table 2.1b
Table 2.2
Table 3.1
Table 3.2
Table 3.3

Table 4.1
Table 4.2a
Table 4.2b
Table 4.2c
Table 4.3

Southern provenances of paper birch collected for larger genecology
study and germination percentages
Northern provenances of paper birch collected for larger genecology
study and germination percentages
Nursery and common garden locations
Provenances chosen for bud flush survey at Red Rock, Skimikin and
Idaho; and the height growth survey at Red Rock and Skimikin
ANOVA results for Days to 20% and 80% Bud Flush in
2003
Mean Julian Days to 20% and 80% Bud Flush by Garden,
Nursery and Region in 2003
Height Growth Results from Univariate Repeated Measures Analysis
inGLM
Summary of Means (+ SEM) for Total Height (cm) by Year
Summary of Means (+ SEM) for Total Height (cm) by Year and
Nursery at Red Rock and Skimikin
Summary of Means (± SEM) for Total Height (cm) by Year and
Region at Red Rock and Skimikin
Health Survey Results at Red Rock in 2002 (Damaged category
includes trees with minor and or major top and or basal damage
due to abiotic and biotic factors)

21
22
23
43
44
44

64
65
65
66
66

Table A-1
Table A-2

ANOVA results for days to 20% and 80% Bud Flush in 2003
Height Growth Results from Univariate Repeated Measures Analysis
inGLM

83
85

Table B-l
Table B-2
Table B-3

ANOVA results for Days to 20% and 80% Bud Flush in 2001
ANOVA results for Days to 20% and 80% Bud Flush in 2002
Mean Julian Days to 20% and 80% Bud Flush by Garden, Nursery
and Region in 2001
Mean Julian Day to 20% and 80% Bud Flush by Garden, Nursery
and Region in 2002

87
88
88

Table B-4

v

89

LIST OF FIGURES

Figure 2.1
Figure 2.2

Interlocking block design used at each common garden
Collection and Common Garden Locations - 20 Seed Source Trial

23
24

Figure 3.1

Growing Degree Days (GDD 5°C Base) and Julian Days to 20% and
80% Bud Flush ±SEM by Garden Compared (RR=Red Rock; ID=Idaho;
SK=Skimikin)
Mean Julian Days to 20% Bud Flush ±SEM by Region in each
Garden in 2003 with Nursery pooled (FN = Fort Nelson; ID = Idaho;
PG = Prince George; PR = Prince Rupert; SA = Salmon Arm;
ID = Idaho; SK = Skimikin and RR = Red Rock)
Mean Julian Days to 80% Bud Flush +SEM by Region in each
Garden in 2003 with Nursery pooled (FN = Fort Nelson; ID = Idaho;
PG = Prince George; PR = Prince Rupert; SA = Salmon Arm;
ID = Idaho; SK = Skimikin and RR = Red Rock)
Mean Julian Days to 20% Bud Flush +SEM by Region in each
Nursery in 2003 with Garden pooled (FN = Fort Nelson; ID = Idaho;
PG = Prince George; PR = Prince Rupert; SA = Salmon Arm;
LN = Landing Nursery; NW = J.D. Little Forestry Centre;
UI = University of Idaho Nursery)
Mean Julian Days to 80% Bud Flush ±SEM by Region in each
Nursery in 2003 with Garden pooled (FN = Fort Nelson; ID = Idaho;
PG = Prince George; PR = Prince Rupert; SA = Salmon Arm;
LN = Landing Nursery; NW = J.D. Little Forestry Centre;
UI = University of Idaho Nursery)
Effective Photoperiod (Day length (minutes) including civil twilight)
by Julian Day for each Garden and Fort Nelson for comparison
Day length (minutes) vs. Mean Monthly Temperature (°C) by Garden
(RR=Red Rock; SK=Skimikin; ID=Idaho) and the Fort Nelson
Region (FN) for comparison
Period of Bud Flush ±SEM by Garden and Region in 2003 with Nursery
pooled (ID=Idaho; SA=Salmon Arm; PG=Prince George; PR=Prince
Rupert; FN=Fort Nelson)

45

Figure 3.2

Figure 3.3

Figure 3.4

Figure 3.5

Figure 3.6
Figure 3.7

Figure 3.8

Figure 4.1

Mean Total Height (cm) +SEM by Garden and Nursery in 2006
(SA=Salmon Arm; FN=Fort Nelson; PR=Prince Rupert;
PG=Prince George; UI=University of Idaho Nursery;
LN=Landing Nursery; NW=JD Little Forestry Centre)

vi

45

46

46

47

47
48

48

67

Figure 4.2
Figure 4.3:

Figure A-l:
Figure A-2:
Figure B-l

Figure B-2

Mean Total Height (cm) +SEM by Garden and Region in 2006
(SA=Salmon Arm; FN=Fort Nelson; PR=Prince Rupert;
PG=Prince George; RR=Red Rock; SK=Skimikin).
Mean minimum and maximum daily air temperature (°C) for Prince
George (PG) and Salmon Arm (SA) for the period of April 1
(90 Julian Days) to May 30 (150 Julian Days) for the years 2001
to 2003.

68
68

Mean Julian Days +SEM to 20% and 80% Bud Flush by Provenance
in 2003 with Garden and Nursery pooled.
Mean Total Height (cm) ±SEM by Provenance in 2006 with Nursery
and Garden pooled.

84

Period of bud flush (days) +SEM by Garden and Region in 2001 with
Nursery pooled (Region: ID=Idaho; SA=Salmon Arm; PG=Prince
George; PR=Prince Rupert; FN=Fort Nelson; Garden: ID=Sandpoint,
Idaho; SK=Skimikin; RR=Red Rock)
Period of bud flush (days) ±SEM by Garden and Region in 2002 with
Nursery pooled (Region: ID=Idaho; SA=Salmon Arm; PG=Prince
George; PR=Prince Rupert; FN=Fort Nelson; Garden: SK=Skimikin;
RR=Red Rock)

89

vn

86

90

ACKNOWLEDGEMENT

Dedicated to my husband, Gary, for believing in me.

I would like to thank my supervisor, Dr. Chris Hawkins, for his guidance, support and
assistance throughout my years as a graduate student. I would also like to thank my
committee members, Dr. Michael Carlson, Dr. Kathy Lewis and Dr. Kevin Keen for their
time and commitment.
I would also like to extend many thanks to Vicky Berger, Danny Barney and Ron Mahoney
for their time and assistance with this thesis. Additionally, I would like to thank my research
assistants, Jennifer Lange, Tracy Rompre and Cindy Baker-Hawkins, as well as Patience
Rakochy and Kirstin Campbell.
Finally, sincere thanks must go to Gary Balliet for his encouragement, support and
understanding and Kevin Wilder for his time and assistance with this thesis. Without their
support and assistance, this thesis could not have been completed.
Funding for this research was provided by the FRBC - Slocan Chair of Mixedwood Ecology
and Management, the BC Ministry of Forests and Range and the University of Idaho.

viii

CHAPTER ONE INTRODUCTION AND LITERATURE REVIEW

1.1 Introduction
Until recently, paper birch (Betula papyrifera Marsh.) in Western Canada has been looked
upon more as a weed species than as a tree worthy of a breeding program (Eriksson and
Jonsson 1986). Few experiments have been conducted because birch plantation establishment
is rare and consequently, tree improvement has had a low priority (Morgenstern 1996). As a
result, very little is known about its pattern of genetic variation (Carlson et al. 2000a) and
evolutionary strategies (Rehfeldt 1993). Conversely in Finland, silver birch (Betula pendula
Roth) has been recognized for its favourable impacts on soils; the enhanced biodiversity and
forest health benefits it brings to mixed species stands (Jones et al. 1998) as well as its
economic value (Koski and Rousi 2005). As a result, birch in Finland has a long history of
tree-breeding and tree improvement efforts (Koski and Rousi 2005).

Paper birch is a very versatile species. Its wood is becoming increasingly valuable,
particularly in the value added sector. Although its wood is often used for lumber, veneer
and pulpwood (Morgenstern 1996), its strength, high density and excellent machining and
finishing properties, make it highly desirable for furniture, flooring and turning (Nielson
2000) as well as doors, window frames, toothpicks, ice cream sticks, toys, butcher blocks and
musical instruments (Jozsa 2000).

Additionally, "paper birch can be managed to maintain or enhance slope stability" (Campbell
2000 p. 30). It is also commonly used in landscaping and the sap can be used to make syrup,
wine, beer or medicinal tonics (Safford et al. 1990). Paper birch is an important species for

1

wildlife as well. Its seeds, buds and bark are an important food source for birds and small
mammals, whereas deer and moose use it primarily for browse and cover (Safford et al.
1990). Paper birch also plays an important role in nutrient cycling due to its deciduous
nature, rapid rate of litter decomposition and high nutrient concentrations in its foliage
(Simard 1996), leading to improved nitrogen and pH levels in the soil (Jones et al. 1998).
Because paper birch is resistant to root disease (immune to Phellinus ignarius; tolerant of
Armillaria sp.) (Simard 1996), it can reduce the incidence of these pathogens where it is
grown in conjunction with conifers such as Douglas-fir {Pseudotsuga menziesii Franco)
which is susceptible to root disease (Baleshta et al. 2005). Furthermore, the presence of
paper birch in the stand results in an increased diversity of tree species, stand structural
diversity and below ground diversity (Jones et al. 1998).

Although paper birch has many benefits, it suffers from many forest health issues. The
presence of fungi such as Armillaria sp., Fomes, Phellinus ignarius, Piptoporus betulinus
and Inonotus obliquus lead to stain and decay in paper birch resulting in the loss of wood
quality and biomass (Peterson et al. 1997). Canker-causing fungi such as Nectria spp. can
weaken birch, leaving it susceptible to attack by other agents (Peterson et al. 1997).
Although wood quality is generally unaffected, redheart discolours the central column of
living birch trees, making it less desirable for some manufactured products. The birch
skeletonizer (Bucculatrix canadensisella Chambers), the forest tent caterpillar (Malacosoma
disstria Hubner) and the gypsy moth (Lymantria dispar Linnaeus) are common birch
defoliators which cause reduced annual growth and weaken the trees, leaving them
susceptible to secondary attack by bark beetles, ambrosia beetles, and several borer species

2

(Peterson et al. 1997). The bronze birch borer (Agrilus anxius Gory) is the most damaging.
It attacks the crown and advances down the bole; its larvae can girdle the tree and kill it.
Birch leaf miners discolour and defoliate birch leaves (Peterson et al. 1997). However, as the
biodiversity and versatility of paper birch is realized, forest managers and manufacturers
alike are becoming more and more interested in the species and the time has come to
recognize paper birch as an economically and ecologically valuable tree species in British
Columbia.

There are many issues, however, that must be addressed in order to manage paper birch more
effectively as a high yield species, including the genecology of the species, as well as an
assessment of the potential impacts of climate change. Additionally, it is helpful to have a
clear understanding of the patterns of geographic genetic variation of the species before
developing seed zones and seed transfer guidelines. This information will also be of use if
and when a breeding program is established. Research into these areas is currently in its
infancy.

1.2

Purpose of Research

The purpose of this study was to address some of the above stated issues by examining the
genecology of 20 paper birch provenances from British Columbia and Northern Idaho.
Previous research into the geographic genetic variability of 18 paper birch provenances from
across British Columbia suggested some regional and population differentiation for certain
traits (Carlson et al. 2000a). Furthermore, there were some initial differences in stock size
resulting from the different nurseries (M. Carlson personal communication Feb. 16, 2009).

3

Despite promising results from this initial research, the populations in that study were not
collected along elevational, latitudinal or longitudinal transects, which in turn limited the
usefulness of the data in understanding patterns of genetic diversity across the species range
(Carlson et al. 2000b). "The importance of physiographic and climatic variables needs to be
understood before the patterns of geographic genetic variation in paper birch can be
adequately described for the species range" (Carlson et al. 2000b p. 1).

The specific objective of this study was to determine if the phenotypic variability observed in
paper birch for certain traits was due to genetics, the growing environment or an interaction
between the two by examining phenological (bud flush) and morphological (height growth)
traits. These two traits were chosen as test parameters because of their suspected strong
heritabilities (M. Carlson personal communication, May 2001). Whether or not there was a
carryover effect from nurseries was also examined.

1.3

Development of Genecology

The term genecology was originally coined and defined by Turesson as "the study of the
[intraspecific] variation of plants in relation to environment" (Heslop-Harrison 1964 p. 159).
Turessonian genecology has three basic propositions: a) that "wide-ranging plant species
show spatial variation in morphological and physiological characteristics"; b) that "much of
this [intraspecific] variation can be correlated with habitat differences"; and c) that
"ecologically-correlated variation is not simply due to plastic response to environment, it is
attributable to the action of natural selection in moulding locally adapted populations from

4

the pool of [genetic] variation available to the species as a whole" (Heslopp-Harrison 1964 p.
160).

Turesson's experiments were highly significant in that he was the first to bring plants of a
single species naturally growing under different environmental conditions into a common
garden, noting, "that certain morphological types tend to be associated with particular
habitats" (Clausen et al. 1940 p. 403). Kerner, on the contrary, was the first to carry out
scientific transplant experiments in varied environments, noting morphological differences in
the same plant species' population grown at different sites (Clausen et al. 1940).

Other influential researchers, including Langlet (Heslop-Harrison 1964), Bonnier, Clements
and MacDougal (Clausen et al. 1940) also contributed significantly to the study of
genecology. Langlet, for instance, was the first to point out that adaptive variation in species
tended to follow continuous variation in habitat (Heslop-Harrison 1964). More recent
studies, however, "have shown that [this] variation may be either continuous or
discontinuous" depending on the forces acting on the species (Heslop-Harrison 1964 p. 161).
Bonnier was the first to introduce the method of clone transplanting; and in separate
experiments, Bonnier and Clements noted transformations in species brought from low
elevations to high elevations as well as limitations in a plant's ability to adapt beyond the
natural range of the species (Clausen et al. 1940). Finally, MacDougal, in an attempt "to
determine the factors involved in the dissemination, establishment and adaptation of plants to
new environments," grew corresponding sets of different species under different climates
(Clausen et al. 1940 p. 401).

5

These early studies lead to two general conclusions: 1) "that plants can modify their form to a
certain extent when exposed to different environments," and 2) "that the range of this
modification is governed by the hereditary constitution of the plant" (Clausen et al. 1940 p.
407).

1.4

Geographic Variation in Paper Birch

Although it is found mainly in Canada, paper birch is the most widely distributed birch
species in North America. The northern limit of its natural range "closely follows the
northern limit of tree growth from Newfoundland and Labrador west across the continent
into Alaska" (Safford et al. 1990 p. 158). Its range extends south into Washington, "the
mountains of northeast Oregon, northern Idaho, and western Montana" (Safford et al. 1990 p.
158) as well as into "North Dakota, the Black Hills of South Dakota, Wyoming,
Nebraska.. .the Front Range of Colorado.. .Minnesota and Iowa, through the Great Lakes
region into New England... [and] down the Appalachian Mountains from central New York
to western North Carolina" (Safford et al. 1990 p. 158).

Paper birch can be found throughout British Columbia, except on the Queen Charlotte
Islands, and only sporadically on south eastern Vancouver Island (Peterson et al. 1997). Its
broad geographic distribution is due to "its ability to be a prominent species on sites of poor
quality...its ability to regenerate on sites after fire and harvesting disturbances; its high
resistance to growing season frost and its ability to begin early season growth while [air]
temperatures are still below freezing" (Peterson et al. 1997 p. 13). Paper birch also has an

6

"ability to tolerate solidly frozen ground in the dormant season" (Peterson et al. 1997 p. 13)
and it is flood and drought tolerant (Peterson et al. 1997).

Paper birch also shows a great deal of variation in its commercial attributes, including:
growth and yield, stem form and quality as well as in its physical attributes such as wood
density (Jozsa, 2000). The geographic variation in paper birch's physical traits emphasizes
the need for research into the origin (genetic or environmental) of this variation prior to
developing seed zones and seed transfer guidelines. It also suggests there is potential for tree
improvement through a selective breeding program. The first step, however, is provenance
testing. In Finland, the success of their tree-breeding and tree improvement program has
come through comprehensive scientific knowledge (Koski and Rousi 2005).

1.5

Tree Improvement, Seed Zones and Seed Transfer Guidelines

"A provenance test is [a study] in which seeds are collected from a number of widely
scattered stands (usually natural), and the seedlings are grown under similar conditions"
(Wright 1976 p. 253). Provenance testing is an important first step in tree improvement
activities as it is widely used to "screen the naturally available genetic variation [within a
species] and to choose the best available types for reforestation or further breeding work"
(Wright 1976 p. 254).

There are three main factors known to influence the amount of phenotypic and genotypic
variability observed in a species: a) the size of the species' range; species with very large
natural ranges contain much more genetic diversity than do species with very limited ranges

7

(Wright 1976), b) the amount of environmental diversity within the species' natural range
and c) the extent of range discontinuities (Wright 1976).

Provenance testing of paper birch in British Columbia is currently in its early stages. In
1994, seed was collected from 18 stands of paper birch across five forest regions. This seed
was germinated and grown at two different nurseries in 1995 and planted at six different field
sites in 1996. Early results of this study revealed that height growth in paper birch did not
differ significantly among provenances; differences between individual stands within regions
appeared to be greater than the differences among regions, suggesting that paper birch has a
generalist evolutionary strategy towards height growth (Carlson et al. 2000a). Generalist
species do not exhibit strong genetic variation in growth or other adaptive traits along
environmental gradients, whereas specialists do (Aitken 2004). Frost tolerance however,
appeared to be well differentiated at the regional level (Carlson et al. 2000a).

Seed zones and seed transfer guidelines generally stem from information obtained from
provenance experiments and have been developed for the majority of conifer tree species in
British Columbia. Seed zones and seed transfer guidelines control the movement of seed to
ensure "good growth and hardiness of planted forests" (Morgenstern 1996 p. 137). "Seed
zones are geographic subdivisions of the range of a species based on ecological and genetic
criteria" (Morgenstern 1996 p. 137). Seed transfer guidelines determine "the distance of
movement of seed and planting stock from the place of origin to the plantation area"
(Morgenstern 1996 p. 137).

8

Although provenance testing will provide forest managers and researchers alike with key
information on the geographic variation of paper birch, the importance of initiating a
selective breeding program and developing seed zones and seed transfer guidelines for paper
birch in British Columbia will be determined by how quickly the species is recognized as a
commercial species in the province with harvesting and artificial reforestation (population
deployment) activities.

1.6

Nursery Effects

Nursery practices including seed storage, stratification, sowing, fertilization, watering,
weeding, lifting, culling and packing may lead to the survival of some seedlings over others,
altering a population's genetic structure which may ultimately influence the genetic
adaptation of forest trees to plantation sites (Campbell and Sorensen 1984). Other indirect
effects such as changes in growth-rhythm rates (i.e. bud flush and height growth) may not
appear until several years after planting (Campbell and Sorensen 1984). Consequently, any
nursery with an environment greatly different from that of the seedlot origin may alter
seedling germination behaviour in the nursery as well as seedling growth and development
after planting (Campbell and Sorensen 1984; Hawkins 1998). Furthermore, the timing of
preconditioned developmental stages in the home environment may become out of phase in
the new environment (Rowe 1964). "Sensitive periods for preconditioning seem to be at the
time of initiation and formation of buds and seeds, and also at the time when growth
commences following a dormant or resting stage" (Rowe 1964 p. 402). For example, when
northern provenances grown at southern nurseries are planted in the north, their flushing
phenology may be out of phase with that of the local populations (Hawkins 1998) because

9

bud flush phenology depends not only on the present (growing) environment, but also on the
environment in which the buds were formed (Rowe, 1964). However, whether these effects
are temporary or persistent is yet to be determined (Hawkins 1998). Nursery effects are still
observed in hybrid spruce (Picea glauca x Picea englemanii) 10 years after planting
(Hawkins 2005).

1.7

Phenology

Phenology is defined as the study of the timing of recurrent biological events over the course
of a year, particularly in relation to climate and other environmental factors (Martin and Hine
2000).

Intraspecific variations in phenological characteristics, such as bud flush and leaf drop, are
known to be under genetic control (Lechowicz 1984) and are highly adaptive (Vaartaja 1959;
Lechowicz 1984; Kramer 1995; Baliuckas et al. 1999). Such variations have also been
observed between co-occurring species growing on the same site (Lechowicz 1984). The
simplest explanation for the observed variation in these characteristics is that it evolved as a
measure to avoid unfavourable conditions (Kramer 1995) and involves a trade-off between
maximizing growing season length and protection against the potential damaging effects of
early and late season frosts (Vaartaja 1959; Nienstaedt 1974; Lechowicz 1984; Cannell and
Smith 1986; Hanninen 1991; Kramer 1995; Baliuckas et al. 1999).

Lechowicz (1984) suggests three other broad possible explanations for the observed variation
in leaf phenology between species: phylogenetic, historical and adaptive. "Phylogenetic

10

explanations predict that closely related species will be similar in leaf phenology"
(Lechowicz 1984 p. 823), whereas "historical explanations predict that contemporary
phenology will reflect adaptation to the paleoenvironments in which the taxon or its
immediate ancestors evolved" (Lechowicz 1984 p. 824) and that insufficient time has passed
for them to have evolved to present conditions (Lechowicz 1984). Finally, "adaptive
explanations predict correlations of leaf phenology with other traits such that the coordinated
suite of traits together improves survival and reproduction over other possible combinations
of traits" (Lechowicz 1984 p. 824).

1.7.1

Bud Flush

The geographic variation in bud flush in many tree species is thought to be under genetic
control (Nienstaedt 1974; Lechowicz 1984; Sulkinoja and Valanne 1987; Chmura and
Rozkowski 2002). In provenance experiments with beech (Fagus sylvatica L.), Chmura and
Rozkowski (2002) determined that the origin of beech populations influenced bud flush and
noted that due to the stability of bud flush across different environments, bud flush was
indeed under genetic control. Additionally, in a review of many early studies of North
American and Lappish birch species, Sulkinoja and Valanne (1987) concluded that bud flush
in these species is under genetic control "and show cline-type variation along latitudinal and
[elevational] gradients" (p. 32).

Although bud flush is largely under genetic control, there are several environmental factors
including air temperature, soil temperature and photoperiod, which direct or signal the onset
of bud flush (Rohrig 1991). Of these environmental factors, air temperature appears to be the

11

most significant (Hanninen 1991; Rohrig 1991; Kramer 1995) as annual "variation in the
timing of tree phenological stages is low when related to thermal sums rather than calendar
days" (Lechowicz 1984 p. 821). Furthermore, bud flush in most temperate deciduous trees is
controlled by the cumulative heat sum (degree-hours, degree-days) to which buds are
exposed after a prerequisite chilling period (Lechowicz 1984; Myking and Heide 1995).
Accordingly, heat sums are often used to predict the onset of phenological events such as bud
flush (Rohrig 1991).

Because the timing of bud flush is under strong genetic control and is highly adaptive, there
is considerable variation between species and among populations within a species,
particularly in those with broad geographic ranges (Vaartaja 1959; Nienstaedt 1974;
Lechowicz 1984; Sulkinoja and Valanne 1987; Chmura and Rozkowski 2002). In Europe
and North America, bud burst generally progresses from south to north and from coastal to
continental climates for trees grown in their natural environments (Morgenstera 1976;
Wright 1976). In provenance studies with Lappish and North American birches (Sulkinoja
and Valanne 1987), Scandinavian birches (Myking and Heide 1995) and sugar maple (Acer
saccharum Marsh.) (Nienstaedt 1974), northern provenances flushed earlier than southern
ones; with southern provenances requiring a greater accumulation of heat than northern ones
(Baliuckas et al. 1999). Further well known trends include high elevation and interior
provenances flushing before low elevation and coastal provenances respectively when grown
at a continental site (Morgenstern 1976; Chmura and Rozkowski 2002). This variation in
bud flush has strong implications for survival. For example, trees that flush too early in the
spring are often susceptible to late spring frosts, which lead to reductions in growth and poor

12

stem form (Chmura and Rozkowski 2002). Consequently, seed zones and seed transfer
guidelines have been developed to prevent such losses from poorly adapted seed sources
(Nienstaedt 1974; Morgenstern 1996).

1.7.1.1 Impacts of Climate Change
Despite natural adaptation and efforts to minimize maladaptation through human
intervention, climate change may significantly influence the timing of bud flush. For
example, with warmer winter temperatures, chilling requirements might not be met, delaying
bud flush or alternatively, where chilling requirements are currently exceeded, bud flush may
occur earlier, leaving trees susceptible to spring frosts (Cannell and Smith 1986; Hanninen
1991; Heide 1993; Kramer 1995; Myking and Heide 1995). This may lead to changes in the
competitive balance between plant species and individuals within species as those species
and individuals that can exploit the milder climate will have greater fitness over those who
cannot (Billington and Pelham 1991; Kramer 1995). Although Kramer (1995) suggests that
trees have a certain amount of plasticity to accommodate such a change in temperature,
Billington and Pelham (1991) suggest that many woody species including silver birch (Betula
pendula Roth) will not be able to exploit milder climates due to physiological control
mechanisms. This is because chilling in winter and heat accumulation in spring influence
bud flush, which implies that "an increase in mean annual temperature reduces the number of
chill days which increases the thermal time required to flush, so that the overall timing of
budburst remains approximately constant" (Billington and Pelham 1991 p. 403). However,
in a field experiment with silver birch and downy birch (Betula pubescens Ehrh) populations,
high autumn temperatures increased bud dormancy and delayed bud burst the following

13

spring (Heide 2003). This suggests that high autumn temperatures may delay the timing of
spring bud burst, offsetting the adverse effects of rising winter temperatures (Heide 2003).
However, observed responses are likely species and population dependent (Hawkins personal
communication, February 20, 2009).

1.7.1.2 Birch Dieback
Forest or birch dieback is a phenomenon which has been observed in stands of birch and
other hardwood species in eastern Canada and the north eastern United States (Auclair et al.
1997), regardless of age, vigour or stand condition (Braathe 1995). Forest dieback is defined
as "the development of symptoms associated with the unnatural mortality of leaves, buds,
twigs, and branches" (Auclair et al. 1997 p. 176), leading to stem dieback, early leaf
colouration and leaf fall, mortality of fine roots and reduced radial growth (Auclair et al.
1997).

The causes of birch dieback are unknown. Although birch dieback commonly occurs in
association with one or more insect infestations or incidences of disease, it can occur
independently of these factors (Auclair et al. 1997). Recent research has shown that
prolonged winter thaw (reducing cold hardiness) followed by freezing temperatures are key
processes leading to dieback in hardwood species (Braathe 1995; Auclair et al. 1996; Zhu
2001; Zhu 2002). Frost is likely the cause of the initial tissue damage (Zhu 2002), leaving
the trees more susceptible to drought, insect infestations and attack by pathogens (Auclair et
al. 1996). Additionally, changing global temperatures may prevent plants from maintaining
maximum cold hardiness in winter (Ogren 2001). Moreover, increased extremes in weather

14

such as warmer spring temperatures may lead to early bud burst, increasing the risk of frost
damage (Ogren 2001). These conditions in combination with drought stress could lead to the
complete dieback of some tree species (Auclair et al. 1996).

Recovery from dieback can occur in the absence of frost and drought stress (Auclair et al.
1996). The extent of recovery, however, will depend on the severity and duration of the
damage to the tissues (Auclair et al. 1997) as well as the impact of secondary damaging
agents such as insect infestation or incidences of disease (Jones et al. 1993).

1.8

Height Growth

Woody plant species exhibit free (indeterminate) growth or fixed (determinate) growth
patterns. Paper birch exhibits free growth. In free growth, "primordia initiation and
internode elongation occur simultaneously during shoot elongation" (Junttila and Nilsen
1993 p. 44) whereas in fixed growth, growth stops once a genetically predetermined structure
has been formed (Tirri et al. 1998). Although species exhibiting free growth are strongly
influenced by many environmental factors; temperature and light (day length) appear to be
the most important (Callaham 1962; Junttila and Nilsen 1993); with shoot elongation
continuing in free growing species as long as both of these factors remain favourable (Junttila
and Nilsen 1993).

Because temperature and light play such a critical role in shoot growth of free growing
woody plants, there is a great deal of variation in the initiation, rate and duration of growth,
particularly among and within species with broad geographic ranges (Kozlowski 1971).

15

Shoot growth has been shown to vary seasonally, diurnally and with elevation and latitude
(Kozlowski 1971). For example, in northern provenances of birch, populations from lower
elevations showed superior growth rates over those from higher elevations (Junttila and
Nilsen 1993). Also, decreased growth rates have been observed at the northern limit of a
species' range (Loehle 1998). Generally shoot elongation is driven by photoperiod, with
growth continuing as long as photoperiod is longer than critical day length (Junttila and
Nilsen 1993); the longest day under which stem elongation ceases (Vaartaja 1959).

Cessation of shoot elongation is particularly important for plant species from northern
climates, as the cessation of growth has been linked to the development of cold hardiness and
dormancy, and is key to their survival (Junttila and Nilsen 1993; Li et al. 2003). Yet,
cessation of growth varies widely with the latitude of origin of the seed source (Junttila and
Nilsen 1993). For example, in a recent study by Li et al. (2003 p. 129), "three ecotypes of
silver birch showed close adaptations to the length of the frost-free growing season
characteristic of their local environment," by exhibiting large ecotypic differences in height
growth. The range of growth was 445.9cm ± 9.3cm, 369.5cm ± 8.8cm and 244.1cm ± 5.9cm
for the populations from southern, central and northern ecotypes, respectively. In a separate
study by Heide (2003), silver birch and downy birch species revealed similar trends in
growth cessation with high latitude populations ceasing growth and shedding their leaves
earlier than those from mid and low latitudes.

Although temperature and photoperiod are crucial environmental factors in the regulation of
growth and development in trees, the mechanisms behind this regulation are still poorly

16

understood. It is believed that plant hormones or plant growth regulators (PGR's) are largely
responsible for controlling shoot elongation and dormancy (Junttila and Nilsen 1993).
Abscisic acid (ABA) for example has been linked to growth cessation, the development of
freezing tolerance and bud dormancy, "however, little is known about seasonal changes in
ABA levels and how they are related to height growth, dormancy and freezing tolerance in
trees" (Li et al. 2003 p. 128). Li et al. (2003) began to explain differences in dormancy
development and release among northern and southern ecotypes of trees by correlating these
differences with variations in ABA levels. Northern ecotypes, for example, developed
dormancy earlier in the fall and had higher levels of ABA than southern ecotypes, however,
by spring, northern ecotypes had lower levels of ABA than southern ecotypes permitting
earlier dormancy release.

1.9

Conclusion

Paper birch is an ecologically valuable species with a broad geographic distribution and
increasingly valuable commercial attributes. Its diversity, versatility and enduring nature
make it an ideal candidate for a selective breeding program. However, an understanding of
the genecology of the species and an understanding of the potential impacts of climate
change will be fundamental to its success.

This thesis began with a general introduction (Chapter 1). An overview of the research
design (Chapter 2) follows. Methods, results and discussion sections are subsequently
presented for bud flush (Chapter 3) and height growth (Chapter 4). This thesis finishes with
a summary discussion (Chapter 5).

17

CHAPTER TWO RESEARCH DESIGN

2.1

Birch Genecology Study - Overall Study (48 Paper Birch Provenances)

The 20 provenances of paper birch examined in this study are a subset of a larger trial
designed by Mike Carlson and Vicky Berger of the BC Ministry of Forests and Range, Chris
Hawkins of the University of Northern British Columbia and Ron Mahoney of the University
of Idaho. The objectives of the overall study were to determine the geographic genetic
variability within and among paper birch populations throughout the range of the species
(along elevational, latitudinal and longitudinal transects) by examining phenological and
morphological traits as well as to determine the importance of nursery environment and
practices on the growth and development of seedlings after planting (Carlson et al. 2000b).
A description of the larger trial follows.

In 1998, six regional collections of paper birch (Prince Rupert, Prince George, Cariboo,
Kamloops, Nelson and Vancouver) were made in British Columbia along elevational,
latitudinal and longitudinal transects; in total, 44 populations were sampled in British
Columbia, two of which were later dropped from the study due to poor germination. Eight
additional populations were sampled in northern Idaho, three of which were later dropped
from the study due to poor germination (Tables 2.1a and b) (Carlson et al. 2000b). Stands
within each region were chosen within 10-20 m of a pre-determined elevation, beginning at
the bottom of an elevational gradient and continuing every 100 m thereafter. Three to five
non-clonal trees within each stand were chosen for seed sampling. These trees were of good
health, form and had produced seed. Seed was collected and bulked to represent a population

18

(Carlson et al. 2000b). Additionally, seed from a previous collection of a far northern
population (Fort Nelson) was added to the study.

All 48 provenances were grown at three different nurseries: North wood Reforestation Centre
(Prince George) now called the Canfor J. D. Little Forestry Centre, Landing Nursery
(Vernon) and the University of Idaho (Moscow, ID) (Table 2.2). Seed was grown in three
different nurseries to determine the importance of nursery practices on the growth and
adaptedness of the seedlings after planting (Carlson et al. 2000b). Seedlings were hand sown
at each nursery in early May 1999 in PSB 515A styroblocks (Beaver Plastics, Edmonton,
AB), lifted in November 1999 and placed in cold storage until the spring of 2000 when they
were planted. Seedlings were spring planted in 3 common gardens: Late April - Sandpoint,
Idaho, early May - Skimikin (Salmon Arm) and mid May - Red Rock (Prince George) (Table
2.2). Seedlings were planted at 2 m x 2 m spacing in a randomized single tree interlocking
block design (16 trees per provenance and nursery, 4 single-tree plots per provenance, 4
replicates) (Figure 2.1). A demonstration area was also established with two five-tree rows
of each provenance grown at the local nursery (i.e. demonstration areas were established at
Red Rock, Skimikin and Sandpoint, ID with seedlings grown at Northwood, Landing and the
University of Idaho, respectively) (Carlson et al. 2000b).

2.2

Birch Genecology Study - 20 Paper Birch Provenances

A subsample of 20 provenances of paper birch was chosen for this study (Figure 2.2). This
subsample was chosen with an attempt to sample the range of the larger project by choosing
provenances based on their distribution over latitudinal, longitudinal and elevational

19

transects. The Fort Nelson, Prince George, Salmon Arm and Idaho provenances were chosen
for their distribution along a north/south transect to determine the influence of latitude on bud
flush and height growth. The Prince Rupert and Prince George provenances were chosen for
their distribution along an east/west transect to determine the influence of maritime vs.
continental climates on these same two traits. For each of the Prince Rupert, Prince George
and Salmon Arm regions, six population samples were taken along elevational transects to
determine the influence of elevation on the above two traits. The original study did not
include population samples along an elevational transect for the Fort Nelson and Idaho
regions therefore they were not included in this study.

This subsample will not only permit an examination of the genetic differences among
populations for the chosen traits, but will also permit an examination of the environmental
effects on these populations since all provenances were grown under similar conditions in
three different common gardens. It will also permit an assessment of the 'carryover' effect of
nurseries where the seedlings were grown.

20

2.3

Tables and Figures

Table 2.1a: Southern provenances of paper birch col lected for larger g enecology study
and germination percentages (from Carlson et al., 2( 00b)
ID*1
Transect
District
Site name
Elev.
Long.
Germ.
Lat.
(m)
%
11
640
Idaho
Sec. 28
Sandpoint
48°16' 116°33'
6
12
Idaho
Jewel Lake
790
116°43'
6
Sec. 33
48°08'
Idaho
Sec. 16
Wrenco
870
116°42
13
48°13'
26
14
Butler Creek
885
Idaho
Sec. 16
48°06' 116°36'
6
Butler Creek
15
Idaho
885
116°36'
13
Sec. 16
48°06'
21
Nelson
New Denver
560
20
Arrow Lake
50°00' 117°17'
22
Nelson
Arrow Lake
New Denver
740
52
50°00' 117°17'
Nelson
New Denver
23
Arrow Lake
840
10
50°00' 117°17'
Nelson
24
Arrow Lake
New Denver
935
8
50°00' 117°17'
Nelson
25
Arrow Lake
New Denver
1100
6
50°00' 117°17'
28
Vancouver
Chilliwack
Richmond2
3
4
49° 10' 123°07'
Salmon Arm
31
Kamloops
Fly Hills
460
50°42'
119°25'
71
Salmon Arm
32
Kamloops
Fly Hills
640
50°42* 119°25'
72
Salmon Arm
33
Kamloops
Fly Hills
760
50°42'
41
119°25'
Salmon Arm
34
Kamloops
Fly Hills
880
50°42'
119°25'
10
Salmon Arm
35
Kamloops
Fly Hills
985
50°42' 119°25'
46
Kamloops
Salmon Arm
Fly Hills
1100
36
50°42'
119°25'
19
Salmon Arm
21
37
Kamloops
Fly Hills
1200
50°42' 119°25*
41
Horsefly
Raft Creek"
Cariboo
775
52°30' 121°30'
Horsefly
Club Creek
900
42
Cariboo
52°19' 121°01'
78
43
Cariboo
Horsefly
Club Creek
1000
50
52°19' i2i°or
44
Cariboo
Horsefly
Club Creek
1100
41
52°19'
12F01'
Cariboo
Horsefly
Club Creek
1200
46
45
52° 19' i2i°or
46
Cariboo
Horsefly
Club Creek
1250
12F01'
70
52°19'
48
Idaho
Wrenco4
915
116°42
3
Sec. 16
48°13'
4
49
Idaho
Dufort
655
2
Sec. 26
48°09' 116°35'
Idaho
Sec.
Round Lake4
730
0
50
48°09' 116°38'
71
Vancouver
Squamish
Cheakamus
8
River5
72
Vancouver
Chilliwack
265
8
Maple Ridge
49°16'
122°34
Vancouver
Chilliwack
73
Ruby Creek
45
49°21'
121°36'
38
74
Penticton
Kamloops
Trepanier
650
25
49°51'
119°49'
Creek

ID number used in study
Not used in 48 seed source trial
' From 1994 seed collection
1
Dropped from study due to poor germination
1
Added to original table
:

21

Table 2.1b: Northern provenances of paper birch col lected for larger genecology study
and germination percentages (from Carlson et al., 2( 00b)
ID# 6
District
Site name
Lat.
Long.
Germ.
Transect
Elev.
(m)
%
54°17'
4
01
Prince Rupert
Kalum
Exchamsiks
45
129° 19'
7
River
51
Prince
Tabor Lake
700
53°55' 122°28'
54
Prince
George
George
52
Tabor Lake
53°55'
122°28'
Prince
Prince
800
41
George
George
Prince
Tabor Lake
53°55' 122°28'
53
Prince
900
71
George
George
54
Prince
Tabor Lake
53°55" 122°28'
Prince
1000
82
George
George
Tabor Lake
55
Prince
Prince
53°55' 122°28'
1100
51
George
George
56
Prince
Prince
Tabor Lake
53°55'
1200
122°28'
66
George
George
Brown Bear
61
Prince Rupert
Kalum
210
55°47' 128°45*
42
FSR8
62
Brown Bear
55°47'
Prince Rupert
Kalum
310
128°45'
43
FSR
63
Prince Rupert
Kalum
Brown Bear
55°47'
400
128°45'
29
FSR
Brown Bear
64
Prince Rupert
Kalum
500
55°47'
128°45'
15
FSR
Brown Bear
55°47'
65
Prince Rupert
Kalum
600
128°45'
15
FSR
66
Brown Bear
55°47'
Prince Rupert
Kalum
750
128°45'
3
FSR
81
Prince Rupert
Kalum
New Aiyansh
55°16'
129°02'
80
71
82
Dragon Lake
55°22'
Prince Rupert
Kalum
175
128°55'
46
Maroon Ck
54°47
83
Prince Rupert
Kalum
150
128°45'
13
Kispiox
84
Cranberry
55°35'
128°23'
42
Prince Rupert
330
Junction
Moricetown
55°02'
85
Prince Rupert
Bulkley
450
127°18
74
86
Telkwa
54°38' 127°07'
Prince Rupert
Bulkley
530
79
Vanderhoof
Fraser Lake
54°04' 124°35'
87
Prince
700
75
George
Fort Nelson
Beaver Lake 9
91
Fort Nelson
500
59°01'
N/A
i23°ir

ID number used in study
Not used in 48 seed source trial
8
FSR = Forest Service Road
Added to original table

7

22

Table 2.2: Nursery and common garden locations (from Carlson et al., 2000b).
Common
Nursery
Lat.
Long.
Elev.
Lat.
Long.
Elev.
Garden
Northwood
650
54°00'
122°28'
Red Rock
725
53°45' 122°41'
Landing
50° 17'
119°16'
Skimikin
550
50°45' 119°22'
400
46°44'
Sandpoint
U. of Idaho
735
116°58'
640
48°13' 116°40'

->

1 (position 1)

2
3
4
1
2
3
4
1
2
3
4
etc.

I

t

I

etc
4
3
2
1

1
2
3
4
1

etc (position 2,256)
4
3
2
1

4 (position 564) 2

4 (position 1,692)

3
2
1
4
3
2
1

3
4
1
2
3

3
2
1
4
3

etc.

1

t

I

t

4 (position 1 ,128)2

Position 1: start rep. 1
Position 564: end rep. 1
Position 1128: end rep. 2
Position 1692: end rep. 3
Position 2256: end rep. 4
48 sources x 4 trees per
source x 3 nurseries = 564
trees per replication.
Positions for trees within
each replication randomly
selected at time of planting.

Figure 2.1: Interlocking block design used at each common garden (from Carlson et al.
1999).

23

Fort Nelson

®

British Columbia
Legend

*Ǥ)Prince Rupert

<§) Collection Locations
© Common Garden Locations

Prince George j
(5)
Red Rock

©Skimikin
©Salmon Armk

Washington

/(Ifwrenco

Idaho

Figure 2.2: Collection and Common Garden Locations - 20 Seed Source Trial

24

CHAPTER THREE BUD FLUSH
3.1 Introduction
The timing of bud burst depends largely on the amount of chilling received in winter and the
subsequent accumulation of thermal heat sums above a threshold level in the spring
(Lechowicz 1984; Billington and Pelham 1991). Once chilling requirements have been met,
northern (Baliuckas et al. 1999) and high elevation provenances (Morgenstern 1996)
typically require less heat to flush than southern (Baliuckas et al. 1999) and low elevation
provenances (Morgenstern 1996). The timing of bud flush is critical because it determines
the beginning of the growing season and the probability of damage due to late spring frosts
(Cannell and Smith 1986). Accordingly, the timing of bud burst is under strong genetic
control (Nienstaedt 1974; Chmura and Rozkowski 2002) and is typically synchronized with
air temperature at the site of origin (Morgenstern 1976; Hanninen 1991).

Although air temperature appears to be the most significant environmental factor influencing
the timing of bud burst (Hanninen 1991; Rohrig 1991; Kramer 1995), soil temperature and
photoperiod also play an important role. Several studies have shown that low soil
temperatures in spring can contribute to a delay in the initiation of growth in paper birch
(Hawkins manuscript in preparation), yellow birch (Betula lutea Michx.) (Fraser 1956), and
white birch (Betula verrucosa platyphylla) (Chipoulet 1981). In some species, the
temperature response is modified by photoperiod (Wareing 1953), which is defined as "the
response to changes in [day length] that enables plants to adapt to seasonal changes in their
environment" (Thomas 1998 p. 151). Because temperature can be unpredictable, some
researchers have suggested that a light signal may be needed to prevent premature breaking

25

of winter dormancy (Rousi and Pusenius 2005); the proposed date for this critical
photoperiod is the Vernal Equinox (March 21) (Hakkinen 1999). For bud burst and
subsequent shoot elongation to occur, "photoperiod must be above the critical limit for
cessation of growth" (Junttila and Nilsen 1993 p. 45); accordingly "adaptation to photoperiod
is probably the most prominent feature in latitudinal ecotypes of temperate tree species"
(Junttila and Nilsen 1993 p. 49).

Consequently, it is important to remember that in the northern hemisphere, the effective
photoperiod (defined as day length plus civil twilight) is 12 hours on or about March 1 and
October 10. For the period of March 1 to October 10, effective photoperiod is longer for all
points north of a given location, whereas it is longer for all points south of a given location
for the period of October 10 to March 1.

The objective of this study was to determine whether the date to 20% and 80% bud flush
differed significantly among provenances of paper birch when grown in a common garden
and whether or not there was a carryover effect from nurseries where the seedlings were
grown. It was hypothesized that garden and provenance, but not nursery, would be factors in
the date of bud flush for the 20 provenances of paper birch.

26

3.2 Methods
3.2.1 Sample Population
3.2.1.1 Red Rock
Three regional latitudinal/longitudinal transects (Prince Rupert, Prince George and Salmon
Arm) were chosen with six provenance samples along elevational transects. Two additional
regions (Fort Nelson and Idaho) were sampled with one provenance each (Table 3.1).
Sixteen trees per provenance (4 trees per repetition, 4 repetitions), grown at all three
nurseries were surveyed, for a total of 960 trees (4x4x3x20).

3.2.1.2 Skimikin and Idaho
Three regional latitudinal/longitudinal transects (Prince Rupert, Prince George and Salmon
Arm) were chosen with three provenance samples along elevational transects (lowest, middle
and highest elevation from those sampled at Red Rock). Two additional regions (Fort
Nelson and Idaho) were sampled with one provenance each (Table 3.1). Eight trees per
provenance (4 trees per repetition, 2 repetitions), grown at all three nurseries were surveyed,
for a total of 264 trees (4x2x3x11).

This selection of regions and populations within regions (at Red Rock, Skimikin and Idaho)
permitted an assessment of the variation in bud flush across latitudinal, longitudinal and
elevational transects as well as an evaluation of the carryover effect of nurseries where the
seedlings were grown.

27

3.2.2

Survey

The bud flush survey was carried out at Red Rock and Skimikin in 2001, 2002 and 2003.
The bud flush survey was only carried out at Sandpoint, Idaho in 2001 and 2003 due to heavy
moose damage (browse) sustained by these trees in the winter of 2001/2002.

The bud flush survey involved recording bud burst for 10 buds on a single branch, preferably
the terminal. If the terminal did not have 10 buds, then buds on the next highest branch were
examined and so on until a total of 10 buds were surveyed. Surveys were conducted as
required, depending on weather - warmer weather required more frequent surveys (every 2 3 days). For each survey, the total number of buds burst (for each tree) was recorded until all
10 buds burst. "Bud burst was [defined] as when the first green ragged edges visually appear
between the bud scales, almost like the opening of a 'clam shell'" (Berger 2001).

Health and vigour was also assessed at the time of the bud burst survey in 2002 at Red Rock.
Although the health and vigour assessment was subjective, the information may help explain
the results (i.e. heavy deer browsing and vole damage at Red Rock may influence the results
obtained in the study).

3.2.3

Climatic Data

Climate data were gathered from Environment Canada (Salmon Arm 2 for Skimikin and
Prince George Airport for Red Rock) and the University of Idaho research station.

28

3.2.4

Statistical Methods

Because bud flush generally follows a sigmoid curve; beginning slowly, then proceeding
more rapidly before slowing again at the end of the flushing period, the number of days
required to reach 20% (i.e. 2 of 10 buds flushed) and 80% (i.e. 8 of 10 buds flushed) bud
flush were chosen as the points of analysis. 20% bud flush was chosen as the lower threshold
to avoid including those buds that all burst immediately, and 80% bud flush was chosen as
the upper threshold to avoid including those buds that flush very slowly or never flush. The
number of days required to reach 20% and 80% bud flush were calculated for each tree
(when 20% and 80% bud flush did not occur on an observation day) with the following
formulas:
(1)

Total buds flushed (per tree) = # buds flushed/ 10 (total # of buds observed)

(2)

For 80% flush (for example), where 80% did not occur on an observation day:

(3)

0.8-0.7
0.9-0.7

=0.1=0.5x2=1.0
0.2

Where:

- 0.8 is 80% bud flush (would be 0.2 for 20% bud flush)
- 0.7 in numerator and denominator is the number of buds flushed on
the observation day prior to 80%
- 0.9 in denominator is the number of buds flushed on the observation
day post 80%
- where 2 is the difference in the number of Julian Days pre- and post80% bud flush

126 + 1.0 = 127 days to 80% bud flush
Where:

126 is the Julian Day on the observation day prior to 80% bud flush.

The data were analyzed using a general linear model (GLM) with an analysis of variance
(ANOVA). Parametric statistics were used because the populations' residuals are normally
distributed. Data were analyzed using SYSTAT (version 11, SYSTAT Software Inc., 2005).

29

Only data (11 provenances) common to all gardens was used in the analysis. In the early
stages of data analysis, it was determined that although there were significant differences
(p<0.001, final F test) in the timing of bud burst among provenances along an elevational
transect for each of the Prince George, Prince Rupert and Salmon Arm regions, there were no
distinct elevational trends; i.e., bud break did not proceed from lowest elevation to highest
elevation as previously suggested (Sulkinoja and Valanne 1987, Morgenstern 1996, Chmura
and Rozkowski 2002). This was an unexpected result. However, there were distinct regional
groupings (Appendix A: Table A-l and Figure A-l). These provenances were therefore
pooled to form a region and the final analysis was run with 5 regions (Fort Nelson, Prince
George, Prince Rupert, Salmon Arm and Idaho).

The model was initially run with growing degree day (GDD) at the source location as a
covariate. GDD source was calculated by Tongli Wang of the UBC Department of Forest
Science, Centre for Forest Gene Conservation, with the software ClimateBC (version 2.0,
2005). GDD source is the total number of growing degree days accumulated at a given
location defined by latitude, longitude and elevation (GDD = sum (mean daily temperature base temperature (5°C)) from an arbitrary start date). GDD source was later dropped from
the model because it was not significant (p=0.242, final F test, for days to 20% bud flush and
p=0.052, final F test, for days to 80% bud flush; where a = 0.05).

The final ANOVA model used in the analysis was: Days to 20% or 80% Bud Flush = G + N
+ R + G*N + G*R + N*R + error, where: G = Garden, N = Nursery and R = Region. A
separate model was run for each year (2001, 2002 and 2003) because there was no data

30

collected in Idaho in 2002 and the results for each year were quite different. Annual
variation in the number of days to bud break appears to be quite variable for these birch
populations (Hawkins manuscript in preparation).

3.3 Results
ANOVA results for 2003 for days to 20% and 80% bud break indicated Garden, Region,
Garden*Region, and Nursery*Region were significant (Table 3.2). Similar results were
obtained in 2001 and 2002; very little variation occurred in the order of bud flush from year
to year. Results for 2001 (Table B-l) and 2002 (Table B-2) can be found in Appendix B.
The mean Julian Days to 20% and 80% bud flush for 2003 by garden increased moving from
south to north, while nursery was not different and region had Fort Nelson, Salmon Arm and
Idaho being first and similar followed by Prince George and then Prince Rupert (Table 3.3).
The 2001 and 2002 results can be found in Appendix B, in Tables B-3 and B-4, respectively.

Significant main effects were found for Garden for both days to 20% and 80% flush in all
years (p < 0.001, final F test) (Figure 3.1). Paper birch planted at Sandpoint, Idaho flushed
before those planted at Skimikin, which flushed before those planted at Red Rock. Although
nursery was significant (a = 0.05) for days to 20% flush (p<0.001, final F test) and days to
80% flush (p=0.024, final F test) in 2001, nursery was not found to be significant (a = 0.05)
for either days to 20% flush (p=0.504, final F test) or days to 80% flush (p=0.110, final F
test) in 2003 or for days to 80% flush (p=0.058, final F test) in 2002. Although days to 20%
bud flush in 2002 appeared to be significant (p<0.032, final F test), pairwise comparisons
were not found to be significant (p=0.068, Scheffe's). Significant main effects were found

31

for Region for both days to 20% flush and days to 80% flush in all years (p < 0.001, final F
test). In 2003, the provenance from the Fort Nelson region was first to flush, followed by
those provenances from the Salmon Arm, Idaho, Prince George and Prince Rupert regions.
There were no significant interactions (a = 0.05) among Garden and Nursery in any year,
however, there were significant interactions (a = 0.05) among Garden and Region for days to
20% flush and days to 80% flush in all years (p < 0.001, final F test) (Figures 3.2 and 3.3).
Provenances from different regions were observed to respond differently (i.e. flushed earlier
or later) at different gardens. There were no significant interactions (a = 0.05) among
Nursery and Region in 2001 or 2002, however, there were significant interactions (a = 0.05)
among Nursery and Region for days to 20% flush (p = 0.023, final F test) and days to 80%
flush (p < 0.001, final F test) in 2003 (Figures 3.4 and 3.5) because bud flush was delayed in
the Fort Nelson provenance when grown at the University of Idaho nursery.

3.4

Discussion

It would have been nice to have been able to consider the influence of total days with
measurable precipitation (>2mm) and total sunshine hours in the model because they are
important factors. However, data was only available at each of the climate recording
stations, one value for each region, and not for the place of origin for each of the
provenances. Consequently, measurable precipitation and total sunshine hours would simply
have been describing region in another way.

32

3.4.1

Garden Effects

Paper birch planted at Sandpoint, Idaho burst bud slightly earlier than those planted at
Skimikin, which in turn burst bud much earlier than those planted at Red Rock; following the
general geographic trend wherein bud flush progresses from south to north (Morgenstern,
1996; Wright, 1976). According to Climate BC
(http://genetics.forestry.ubc.ca/cfgc/ClimateBC/, accessed February 23, 2009), the mean date
of the last spring frost is May 11 at Sandpoint, Idaho, May 7 at Skimikin and May 31 at Red
Rock, suggesting that the order of bud burst in this study may be reflecting the timing of late
spring frost at each of the common garden locations. Ager et al. (1993 p. 1936) advocate that
"an optimal thermal sum would time bud flush to coincide on or about the mean date of the
last damaging spring frost" (even though the last damaging frost may occur earlier than the
last spring frost). Veen (1954) also suggests that the temperature conditions in spring are one
of the main factors influencing phenological events. However, soil temperature and
photoperiod also seem to be playing a role, particularly at the Red Rock site. Given the
ambient air temperatures (Figure 4.3), soils at the Red Rock site appeared to remain very
close to freezing well into the growing season; whereas cold soils were not a concern at the
Skimikin or Idaho sites during bud flush. Results obtained in a laboratory study carried out
by Hawkins (manuscript in preparation) in 2002 to determine the influence of soil
temperature on bud flush and height growth in paper birch support this hypothesis; revealing
that birch with warm roots (mean air temperature 14°C throughout study) flushed earlier and
had better overall height growth than those with cold roots (mean 1°C to early March, soils
then warmed slowly to a mean of 12°C by mid-March). Fraser (1956) also identified low air

33

and soil temperatures in the month of May (in field studies near Chalk River, Ontario) as
possible contributors to a delay in the initiation of growth in paper birch.

Because temperature can vary significantly from year to year, many temperate and boreal
plant species also rely on photoperiod to constrain their development to 'safe periods'. The
significance of this photoperiodic constraint appears to increase with latitude (Morison and
Morecroft 2006). In a recent study involving silver birch, Linkosalo and Lechowicz (2006)
"demonstrated that plants do not necessarily respond to warming as soon as their chilling
requirement is met" (p. 1254). They determined that the initiation of bud development in
silver birch occurs around the spring equinox, months after chilling requirements have been
met. Their research suggests that photoperiod plays a major role in the initiation of bud
development in silver birch in the spring and it may also be important for paper birch.

Although GDD source was not found to be statistically significant, an examination of GDD
garden, revealed important differences. Based on the growing degree day (GDD) data
calculated by Tongli Wang with ClimateBC, the amount of heat (GDD) required for bud
flush to occur at Red Rock, the most northerly site, is 1266 GDD (>5°C); whereas the
amount of heat required at the Skimikin and Idaho sites is 2041 GDD (>5°C) and 1857 GDD
(>5°C), respectively. In the northern hemisphere, northern sites have longer effective
photoperiods throughout the growing season than southern sites (Figure 3.6). Moreover, for
a given temperature, photoperiods at northern sites are much longer than those at southern
sites (Figure 3.7). This may influence the amount of heat (growing degree days) required for
bud flush to occur. Research has shown that plants grown under long days (northern

34

latitudes) typically need less heat to flush than those grown under short days (southern
latitudes) (Morgenstern 1996; Hawkins manuscript in preparation).

Although there is a general latitudinal trend for the amount of growing degree days required
for bud flush to occur, there is also an elevational trend; high elevation sites require less heat
than low elevation sites (Red Rock 725 m; Skimikin 550 m and Idaho 640 m). Studies have
shown that elevation has a similar effect as latitude (Pauley and Perry 1954 cited in Brissette
and Barnes 1984; Morgenstern 1996). Sharik and Barnes (1976 cited in Brissette and Barnes
1984) studied birch in the Appalachian Mountains and determined that a change of 1° latitude
is approximately equivalent to a change of 189 m in elevation.

Period of bud flush (Figure 3.8) was calculated as the time (in days) that it took for trees to
progress from 20% to 80% bud flush. The shortest overall period of flush occurred at the
Idaho garden, followed by the Skimikin garden and lastly the Red Rock garden; trees grown
at the southern gardens (Idaho and Skimikin) beginning bud flush earlier and completing bud
flush more quickly than trees grown at Red Rock. This may be due in part because the
spring climate in the south is warmer than in the north with more heat being available for bud
flush to occur over a shorter period of time. Interestingly, however, the Prince George and
Prince Rupert provenances flushed in a shorter period of time than the Fort Nelson, Idaho
and Salmon Arm provenances at both the Idaho and Skimikin gardens in 2003.

35

3.4.2 Regional Effects
The population from the Fort Nelson region was the first to burst bud, followed by
populations from the Salmon Arm, Idaho, Prince George and Prince Rupert regions. For
continental populations (Fort Nelson, Salmon Arm, Idaho and Prince George), the order of
bud burst appears to correspond closely with elevation at the site of origin: Fort Nelson,
500m; Salmon Arm 810m; Idaho 870m and Prince George 967m. Populations originating
from low elevation sites bursting bud before those from high elevation sites; indicating there
is a genetic component influencing the date of bud flush. Many common garden studies
(Pauley and Perry 1954 cited in Li et al. 2003; Vaartaja 1954; Sulkinoja and Valanne 1987;
Myking and Heide 1995; Chmura and Rozkowski 2002) have shown that spring bud flush is
strongly correlated with latitude, longitude or elevation of origin. In a study by Sulkinoja
and Valanne (1987) involving Finnish and Lappish birch species, low elevation downy birch
provenances burst bud earlier than high elevation provenances. This same trend was
observed in red oak (Quercus rubra L.) by McGee (1969). In a separate study involving
Beech (Fagus sylvatica L.), Chmura and Rozkowski (2002) determined that the elevation of
the provenance influenced the timing of bud flush. Although the relationship was not linear,
lower elevation provenances did flush before higher elevation provenances. Morgenstern
(1996 p. 78) suggests that these "elevational clines are due to adaptation to several
physiological processes involving both photoperiod and temperature". However, because the
natural environment of the provenances is so different (elevation, soil moisture, light, heat...)
from the garden environment in which they were grown, environment is also likely playing a
role.

36

From the perspective of growing degree days (calculated by Tongli Wang with ClimateBC),
northern provenances of paper birch (Fort Nelson 1219 GDD (>5°C); Prince George 947
(±98 SE) GDD (>5°C)) required less heat (GDD) than southern provenances (Salmon Arm
1627 (+204 SE) GDD (>5°C); Idaho 1540 GDD (>5°C)). In the northern hemisphere,
northern sites have longer photoperiods throughout the growing season than southern sites
(Figure 3.6). Moreover, for a given temperature, photoperiods at northern sites are much
longer than those at southern sites (Figure 3.7); which may influence the amount of heat
(GDD) required for bud flush to occur. Research has shown that plants grown under long
days (northern latitudes) typically need less heat (fewer growing degree days) to flush than
those grown under short days (southern latitudes) (Morgenstern 1996; Hawkins manuscript
in preparation). In a reciprocal study carried out by Hawkins (manuscript in preparation),
paper birch provenances chilled at the Red Rock common garden (until 12 hour day length
(including twilight) or effective photoperiod was reached) were subsequently planted at
Skimikin and Red Rock; individuals moved to the Skimikin garden flushed earlier, but
needed more heat than those at the Red Rock garden. Additionally, on a recent spring (early
May 2008) trip to Fort Nelson, we observed that the trees in Fort Nelson were breaking bud
when we arrived and had almost finished by the time we left, however, when we returned to
Prince George, the buds were just beginning to flush (Hawkins personal communication,
February 18, 2009). On May 10, 2008, Fort Nelson had accumulated 184.9 degree days of
heat, whereas Prince George had accumulated 211.0 degree days of heat by May 12, 2008
(source: Canadian Climate Data). This suggests that in addition to temperature, photoperiod
regulates bud flush in response to the acquisition of heat in the spring. This has also been
shown to be true of provenances from high elevation sites versus those from low elevation

37

sites (Morgenstern 1996). This complex interaction between photoperiod and temperature
has likely come about because for a given date, temperature can vary significantly from year
to year whereas photoperiod is consistent; thereby constraining development in temperate
and boreal plant species to 'safe periods' (Morison and Morecroft 2006).

Interestingly, however, the phenological responses (bud flush, leaf drop...) of our northern
provenances (FN) are very different from those of the Finnish northern provenances, which
originate from further north than our provenances. This may be due to the fact that the
climate in northern Finland differs very little from that in southern Finland (except in winter
and in the duration of the growing season, which is shorter in northern Finland) because there
are no high mountains and due to the "modifying influence of the prevailing westerly winds
from the Gulf Stream" (Vaartaja 1954 p. 393). Alternatively, because most of Finland lies
north of 60 degrees latitude, all populations will experience some days with continuous
effective photoperiod. This may result in temperature being a more reliable cue (depending
on the time of year) than photoperiod.

When populations of continental origin (Prince George) were compared with those of
maritime origin (Prince Rupert), the order of bud flush followed a longitudinal or inverse
continentality trend; continental populations bursting bud before maritime populations. This
is because maritime provenances respond more slowly than continental provenances when
brought to a continental site (Morgenstern 1976). These results are supported by Chmura
and Rozkowski (2002) who determined that longitude of origin had the most significant
impact on the timing of bud flush in Beech, with eastern or continental provenances of Beech

38

flushing earlier than western or maritime provenances. Veen (1954) proposed this is because
maritime and lower elevation climates experience gradually increasing temperatures in spring
with long periods of late frosts; whereas continental and higher elevation climates experience
rapidly increasing temperatures with short periods of late frosts. As a result, provenances
from coastal and low elevations need more heat for a given photoperiod and tend to flush
later to avoid the risk of late spring frosts, whereas provenances from interior and high
elevations require a minimum amount of heat once a critical day length has been attained
because there is less chance of late spring frosts.

3.4.3

Nursery Effects

There were significant nursery effects in 2001, but not in 2002 or 2003, implying that nursery
does not play a significant role in the date of bud flush following the initial year or two of
establishment. These results are consistent with those obtained in an 18 seed source trial
with paper birch in British Columbia (Carlson et al. 2000a). In that study, there were no
differences observed between nurseries for tree heights or frost resistance. It was also
determined (in this same study) that nursery did not play a role in the timing of bud flush
following the initial year of establishment (Hawkins personal communication). Interestingly,
however, nursery culture appears to have a much greater influence on conifers. In an
ongoing study involving hybrid spruce (Picea glauca x Picea englemanii), nursery culture
continues to account for significant (P < 0.05) variation in tree survival, height and diameter
growth 10 years after planting (Hawkins 2005). In a separate study involving lodgepole pine
(Pinus contorta Dougl.), early results suggested that nursery culture had a significant (P <
0.01) influence on height and diameter growth two years after planting (Ying et al. 1989).

39

The observed difference may possibly be due to conifers having determinate shoot growth
while birch displays indeterminate shoot growth.

3.4.4

Garden by Region Interaction

This interaction was not deemed to be very significant due to the stability of flushing (order
of the regions is the same) through different environments (at different gardens), suggesting
this feature is under genetic control (Chmura and Rozkowski 2002). The order of flush
follows a latitudinal trend by garden and by region (except for the Fort Nelson provenance)
with southern gardens and provenances flushing before northern ones (Prince Rupert is
actually further north than Prince George and also more maritime). The reason why the Fort
Nelson provenance may not be conforming to the general latitudinal trend is because of its
origin being about 5° north of RR, 9° north of SK and 11° north of ID. Even with the shorter
days, the amount of heat received in the south is far greater for a given photoperiod than it
would receive in Fort Nelson.

3.4.5

Nursery by Region Interaction

This interaction is significant due to the delay in bud flush (to 20% and 80%) in 2003 of the
Fort Nelson provenance grown at the University of Idaho nursery. This trend was not
evident in the Fort Nelson provenance or any other provenance in 2001 or 2002. It is not
known whether this was a one time occurrence or a continuing trend as bud flush surveys
were not carried out after 2003. There is very little difference, however, among nurseries in
the timing of bud flush of the provenances from the other regions; there is greater variability
in the timing of bud flush among the regions (as discussed above). The delayed flushing of

40

the Fort Nelson provenance grown at the University of Idaho nursery may have been
influenced by stock handling; seedlings grown at the University of Idaho were cold stored,
but not frozen. As a result, many of the seedlings were planted mouldy and weak at Red
Rock. More importantly, however, studies have shown that different provenances react
differently to the nursery environment in which they are grown; particularly when
provenances are displaced long distances to completely different temperature and
photoperiodic regimes (Morgenstern 1976): Fort Nelson provenance transferred 11° latitude
south to the University of Idaho nursery. Northern and high elevation provenances grown at
southern and low elevation nurseries have lower heat requirements for bud flush to occur
than local provenances and often flush too early leaving them susceptible to spring frosts
(Morgenstern 1976). Many of the trees from the Fort Nelson provenance (grown at all three
nurseries) planted at the Idaho garden may have died due to spring frost damage.
Furthermore, northern provenances grown in southern nurseries are grown under much
shorter photoperiods than they are adapted to in their natural environments and may be
maladapted to the northern photoperiod when planted in the north (Hawkins 1998).

3.4.6

Conclusion and Research Pitfalls

This study has shown that the geographic variation in the timing of bud flush in paper birch
generally follows latitudinal, longitudinal and elevational clines. Although there is a strong
genetic component (there is little annual variation in the timing of bud flush among regions),
there is an environmental component as well which is demonstrated by the differences in the
number of days and the amount of heat required for bud flush to occur at the different
gardens. Results from this study in conjunction with the literature suggest and that there is

41

no single factor that determines the onset of bud flush in paper birch; rather it is a complex
interaction among population genetics, air and soil temperature and photoperiod. What's
more, there is currently a shortage of information on the modifying effects of photoperiod on
heat sums in paper birch and other tree species. When developing seed zones and seed
transfer guidelines for paper birch, it will be important to consider the influence of these
factors on the timing of bud flush for each provenance. Moving provenances too far from
their place of origin could leave them susceptible to a number of unfavourable biotic
(herbivory, insects and disease) and abiotic (frost) factors which will ultimately lead to poor
growth and possibly even death.

These findings demonstrate that environmental conditions (i.e., soil moisture, soil
temperature, soil nutrients) are important factors to consider when establishing common
garden trials. The drought at Skimikin and the cold soils and poor soil nutrient conditions at
Red Rock likely influenced the outcome of this study. Additionally, the anticipation and
minimization of herbivory is also crucial to minimize or avoid the potential damage
herbivory can cause. Moose browsing at the Idaho site prevented the collection of bud flush
data in 2002. Deer and vole damage at the Red Rock site also likely influenced the outcome
of this trial. Furthermore, stock handling should be the same at all nurseries to avoid
unknown factors. As well, the potential effects of long distance displacement should be
considered when growing different provenances at centralized nurseries. Alternatively,
because nurseries can't be located to meet the environmental requirements of every seed
source, it may be beneficial to choose a nursery most appropriate for the seed source
(Campbell and Sorensen 1984).

42

3.5

Tables and Figures

Table 3.1: Provenances chosen for bud flush survey at Red Rock, Skimikin and Idaho;
and the height growth survey at Red Rock and Skimikin (Adapted from Carlson et al.,
2000)
Provenance
Region
Site name
Elev. (m)
Lat.
Long.
13*
Idaho
Wrenco
870
48°13'
116°42'
50°42'
31*
Salmon
Fly Hills
460
119°25'
Arm
32
50°42'
Salmon
Fly Hills
640
119°25'
Arm
33*
Salmon
Fly Hills
50°42'
760
119°25'
Arm
35
Salmon
Fly Hills
50°42'
985
119°25'
Arm
50°42'
36
Salmon
Fly Hills
1100
119°25'
Arm
37*
Salmon
Fly Hills
50°42'
1200
119°25'
Arm
Tabor Lk
51*
Prince
700
53°55'
122°28'
George
52
Tabor Lk
53°55'
Prince
800
122°28'
George
Tabor Lk
53
Prince
53°55'
900
122°28'
George
54*
Tabor Lk
Prince
1000
53°55'
122°28'
George
Tabor Lk
55
Prince
53°55'
1100
122°28'
George
56*
Prince
Tabor Lk
53°55'
1200
122°28'
George
61*
Brown Bear
Prince
210
55°47'
128°45'
FSR
Rupert
Brown Bear
62
Prince
55°47'
310
128°45'
FSR
Rupert
63*
Prince
Brown Bear
55°47'
400
128°45'
Rupert
FSR
64
Brown Bear
55°47'
Prince
500
128°45'
FSR
Rupert
Brown Bear
65
Prince
600
55°47'
128°45'
Rupert
FSR
66*
Brown Bear
Prince
750
55°47'
128°45'
Rupert
FSR
Beaver Lake
91*
Fort Nelson
500
123°11'
59°01'
* Provenances surveyed at Skimikin and Idaho. All provenances were surveyed at Red
Rock.
43

Table 3.2: ANOVA results for Days to 20% and 80% Bud Flush in 2003. Significant
(a=0.05) results are in bold and shaded.
F-Ratio
Sum-of-Squares
df
P
z
R : 0.824
Days to 20% Bud Flush
N:784
2
Garden
40028.809
810.284
<0.001
Nursery
2
0.504
33.838
0.685
Region
4
17134.801
173.426
<0.001
Garden * Nursery
51.554
4
0.522
0.720
Garden * Region
1600.819
8
8.101
<0.001
Nursery * Region
448.423
8
2.269
0.021
Error
18648.867
755
Days to 80% Bud Flush
Garden
Nursery
Region
Garden * Nursery
Garden * Region
Nursery * Region
Error

Rz: 0.818
N:802
59088.547
147.116
14894.870
46.814
1828.818
965.429
25657.528

2
2
4
4
8
8
773

890.098
2.216
112.187
0.353
6.887
3.636

<0.001
0.110
<0.001
0.842
<0.001
<0.001

Table 3.3: Mean Julian Days to 20% and 80% Bud Flush by Garden, Nursery and
Region in 2003. Means followed by the same letter within a factor are not significantly
sed on Schef fe's pairwise comparison tests (a=0.05)
Mean JD
Mean JD
Standard
Standard
to 20%
Error of
to 80%
Error of
Bud Flush the Mean Bud Flush the Mean
Garden
102.023 a
106.054 a
0.829
0.682
ID
104.775 b
0.410
108.812 b
0.467
SK
124.271 c
130.967 c
0.358
0.398
RR
Nursery
0.502
109.945 a
114.508 a
0.457
LN
110.475 a
115.703 a
0.478
0.493
NW
110.648 a
115.623 a
0.458
0.510
Ul
Region
104.806 a
1.349
110.223 a
1.163
FN
107.477 a
112.749 a
0.394
0.349
SA
107.963 a
0.612
113.571a
0.690
ID
112.400 b
0.322
116.487 b
0.373
PG
119.135
c
0.332
123.360
c
0.386
PR

44

2500 -T,

140
a-120

2000 f
100
1500
80

it
|

60

£ 1000
3
o

40

5
500

20

RR

ID

SK

Garden
-GDD(>5'C) •

- Days to 20% Bud Flush - A — Days to 80% Bud Flush

Figure 3.1: Growing Degree Days (GDD 5°C base) and Julian Days to 20% and 80%
Bud Flush ± SEM by Garden Compared. Where error bars are not visible, they are
within the plotted symbol. There are no error bars for GDD for any of the gardens
(RR=Red Rock; ID=Idaho; SK=Skimikin).
140

130 -f

120

S.

a
c

a

110
100

90

80
FN

SA

ID

PG

PR

Region
-ID

-SK-A-RRl

Figure 3.2: Mean Julian Days to 20% Bud Flush ± SEM by Region in each Garden in
2003 with Nursery pooled. Where error bars are not visible, they are within the plotted
symbol (FN = Fort Nelson; SA = Salmon Arm; ID = Idaho; PG = Prince George; PR =
Prince Rupert; ID = Idaho; SK = Skimikin and RR = Red Rock).

45

150
140

130 H
in
CO
Q

>

c 120

110
100
90

-SK

-ID-

-RR

Figure 3.3: Mean Julian Days to 80% Bud Flush ± SEM by Region in each Garden in
2003 with Nursery pooled. Where error bars are not visible, they are within the plotted
symbol (FN = Fort Nelson; SA = Salmon Arm; ID = Idaho; PG = Prince George; PR =
Prince Rupert; ID = Idaho; SK = Skimikin and RR = Red Rock).
125

lian Days

190

115 —

-

110

m "'

3 105
100

r^^l>
[
. . _ .

95
FN

SA

ID

F
PG

PR

Region
— • - - LN —•— NW --kr- -Ul

Figure 3.4: Mean Julian Days to 20% Bud Flush ± SEM by Region in each Nursery in
2003 with Garden pooled. Where error bars are not visible, they are within the plotted
symbol (FN = Fort Nelson; SA = Salmon Arm; ID = Idaho; PG = Prince George; PR =
Prince Rupert; LN = Landing Nursery; NW = J.D. Little Forestry Centre; UI =
University of Idaho Nursery).

46

LN

-NW

•III

Figure 3.5: Mean Julian Days to 80% Bud Flush ± SEM by Region in each Nursery in
2003 with Garden pooled. Where error bars are not visible, they are within the plotted
symbol (FN = Fort Nelson; SA = Salmon Arm; ID = Idaho; PG = Prince George; PR =
Prince Rupert; LN = Landing Nursery; NW = J.D. Little Forestry Centre; UI =
University of Idaho Nursery).
1400

200
Julian Day
•Idaho

• Skimikin

• Red Rock•

• Fort Nelson

Figue 3.6: Effective photoperiod (Day length (minutes) including civil twilight) by
Julian Day for each Garden and Fort Nelson for comparison.

47

1300
1250
1200
1150
3 1100
C
1050

|
e

1000 4 —

g

950
900
850
800
10

15

20

Temperature (°C)

-RR

-FN

SK

ID

Figure 3.7: Daylength (minutes) vs. Mean Monthly Temperature (°C) by Garden
(RR=Red Rock; SK=Skimikin; ID=Idaho) and the Fort Nelson Region (FN) for
comparison.
12.0
10.0
•S

8.0

LL

"S
•o
o
»

6.0
4.0

Q.

2.0 4

0.0

ID

SA

PR

PG

FN

Region

ID

•SK

RR

Figure 3.8: Period of Bud Flush (days) ± SEM by Garden and Region in 2003 with
Nursery pooled. Where error bars are not visible, they are within the plotted symbol
(ID=Idaho; SA=Salmon Arm; PG=Prince George; PR=Prince Rupert; FN=Fort
Nelson; Garden: ID=Sandpoint, Idaho; SK=Skimikin; RR=Red Rock).

48

CHAPTER FOUR HEIGHT GROWTH

4.1 Introduction
The geographic variation in height growth of paper birch is influenced by environmental and
genetic factors as well as the interactions between these two factors (Callaham 1962;
Morgenstern 1996). Abiotic factors such as air and soil temperature, light and moisture must
be favourable for growth to occur, however, variation among these factors must fall within
the natural range of tolerance of the genotype or else growth will be unfavourably affected
(Callaham 1962). Biotic factors such as herbivory can also influence height growth.

Temperature influences growth rate and at times growing season length (Callaham 1962).
However, optimal growth, even under favourable climatic conditions, is somewhat limited
because "growth mechanisms of trees have evolved to be in harmony with the environment
in which trees grow" (Callaham 1962 p.319). Therefore a trade off exists between cold
hardiness and optimal growth rate; "adaptations that favour survival under one set of
conditions will generally reduce growth rate or survival elsewhere" (Loehle 1998 p. 737).

There are three major aspects of cold tolerance which can cause reduced growth: i) structural
investments such as adaptations to prevent freezing damage (i.e. increased lignification in
needles); ii) physiological responses such as those involved in cold resistance (i.e.
accumulation of sugars in leaves) and iii) a conservative growth strategy (Loehle 1998).
Although plants growing in less favourable climatic conditions have shorter growing periods,
their reduced growth cannot be solely attributed to a shorter growing season. Studies have

49

shown that plants growing in cold climates have similar photosynthetic capacity as plants
from warmer climates (Korner and Larcher 1988). However, because they have a shorter
growing period, these plants cannot afford to take many risks because re-growth might not be
possible after the damaging event. As a result, these plants tend to invest energy in
components of growth that reduce risk, such as resistance to herbivory and increased frost
tolerance, which further reduces growth rate (Loehle 1998).

Photoperiod; the ratio of light to dark in a 24 hour period, influences the beginning and end
of growth in photoperiodically sensitive plants and in turn determines the duration of the
growing period (Callaham 1962). Studies, among others, have confirmed the existence of
intraspecific photoperiodic ecotypes along latitudinal clines (Vaartaja 1954; Down and
Borthwick 1956; Eriksson and Jonsson 1986 and Li et al. 2003). When trees are exposed to
photoperiods different from those in their natural environment it can lead to early cessation
of growth, abnormal growth or even death (Vaartaja 1954). For instance, northern
provenances transferred south of their natural range (into shorter photoperiods) often survive,
but show abnormal growth and don't take full advantage of the growing season (enter
dormancy too early). Conversely, the longer photoperiods experienced by southern
provenances transferred north of their natural range often prevent proper dormancy and lead
to frost damage or even death (Vaartaja 1954).

Finally, it is thought that "heat stress [or drought] can be a principal limiting factor in the
distribution, adaptability and productivity of wild and cultivated plants" (Ranney and Peet
1994 p. 243). It is believed that moisture stress and not high temperatures often lead to the

50

failure of planting trials where northern provenances are planted at southern sites (Loehle
1998). Furthermore, the lack of moisture either through drought or a lack of availability (i.e.
frozen soils) delays the onset of bud flush and therefore active growth.

Biotic factors such as herbivory and insect infestations can also influence height growth. In
several studies involving silver birch, Rousi et al. (1990), Rousi et al. (1993) and Rousi et al.
(1997) determined that there was substantial variation in resistance among geographic origins
and families of birch due to browsing by herbivores. This genetic variation is thought to be
an evolutionary response to browsing by herbivores; "highly resistant genotypes may
indicate past, more intensive herbivore pressures" (Rousi et al. 1997 pg. 401). These results
are confirmed in a recent study by Bryant et al. (In press) involving Alaska birch (Betula
neoalaskana Sarg.) and paper birch. Their study revealed that Alaska birch is more closely
associated with fire frequency than paper birch (P<0.001). As a result, Alaska birch range
was also more closely associated with higher hare densities than paper birch. Consequently,
Alaska birch had higher gland densities and higher papyriferic acid levels (P<0.001) than
paper birch, resulting in greater antibrowsing defences'.

The objective of this study was to determine whether height growth differed significantly
among provenances of paper birch when grown in a common garden and whether or not
there was a carryover effect from the nurseries where the seedlings were grown. It was
hypothesized that garden and provenance, but not nursery, would be factors in height growth
for the 20 provenances of paper birch.

51

4.2 Methods
4.2.1 Sample Population
4.2.1.1 Red Rock
Three regional latitudinal/longitudinal transects (Prince Rupert, Prince George and Salmon
Arm) were chosen with six provenance samples along elevational transects. Two additional
regions (Fort Nelson and Idaho) were sampled with one provenance each (Table 3.1).
Sixteen trees per provenance (4 trees per repetition, 4 repetitions), grown at all three
nurseries were surveyed, for a total of 960 trees (4x4x3x20).

4.2.1.2 Skimikin
Three regional latitudinal/longitudinal transects (Prince Rupert, Prince George and Salmon
Arm) were chosen with three provenance samples along elevational transects (lowest, middle
and highest elevation from those sampled at Red Rock). Two additional regions (Fort
Nelson and Idaho) were sampled with one provenance each (Table 3.1). Eight trees per
provenance (4 trees per repetition, 2 repetitions), grown at all three nurseries were surveyed,
for a total of 264 trees (4x2x3x11).

This selection of provenances (at Red Rock and Skimikin) permitted an assessment of the
variation in height growth across latitudinal, longitudinal and elevational transects as well as
an evaluation of the carryover effect from nurseries where the seedlings were grown. There
were no height growth measurements carried out at Sandpoint, Idaho for logistical reasons.

52

4.2.2

Survey

Initially, height growth was measured seven times per growing season (early June, early and
mid July, early and mid August, early September and mid October) at Red Rock and
Skimikin in 2001 and 2002. Data from this survey was later dropped due to the heavy and
repeated abiotic (frost) and biotic (herbivory) damage to trees, specifically at the Red Rock
garden. Because annual height growth measurements continued from 2003 through 2006 at
both the Red Rock and Skimikin gardens, the study focused on the cumulative growth of the
birch trees over the entire period from 2000 to 2006.

4.2.3 Statistical Methods
Although there was heavy and repeated damage sustained by many of the birch trees at both
Red Rock and Skimikin over the years, this damage was considered a naturally occurring
event in a common garden trial. Given that the same trees were measured each year, the data
were analyzed using repeated measures analysis of variance with a general linear model.
Parametric statistics were used because the populations' residuals are normally distributed.
Data were analyzed using SYSTAT (version 11, SYSTAT Software Inc., 2005). Total
height growth at the end of each year (2000 to 2006) was used in the analysis. Furthermore,
only data (11 provenances) common to both gardens were used in the analysis. Although
early analysis revealed significant differences (P=0.002, final F test) in height growth among
provenances along an elevational transect for each of the Prince George, Prince Rupert and
Salmon Arm regions, there were no distinct elevational trends beyond 2001; instead there
were distinct regional groupings (Appendix A: Table A-2 and Figure A-2). That is, low
elevation provenances did not show superior growth over high elevation provenances

53

(Morgenstern 1996). This was an unexpected result and the provenances were pooled to
form a region and the final analysis was run with 5 regions (Fort Nelson, Prince George,
Prince Rupert, Salmon Arm and Idaho). Also, because all the trees from the Idaho
provenance grown at one nursery (Northwood) were dead at the Red Rock garden and the
majority of the trees from this same provenance grown at the other two nurseries were also
dead at the Red Rock garden, the Idaho provenance was completely removed from the final
analysis.

The final general linear model used in the repeated measures analysis was:
Height Growth = G + R + N + G*R + G*N + R*N + error
Where: G= Garden, R = Region and N = Nursery.

4.3

Results

Univariate repeated measures analysis results are summarized in Table 4.1. Significant main
effects were found for Garden (P < 0.001, final F test). Paper birch planted at Skimikin had
better overall growth than those planted at Red Rock (Table 4.2). Nursery was not found to
be significant (P = 0.19, final Ftest). Significant main effects were also found for Region (P
< 0.001, final F test). Paper birch populations from Prince George, Prince Rupert and
Salmon Arm exhibited better growth than the population from Fort Nelson (Table 4.2).
There were significant interactions between Garden and Nursery (P = 0.03, final F test)
(Figure 4.1) and Garden and Region (P < 0.001, final F test) (Figure 4.2). There was no
significant interaction between Region and Nursery (P = 0.19, final F test). Within subjects,
Time (P<0.001, final F test) was significant as were the interactions between Time and

54

Garden (P<0.001, final F test), Time and Nursery and Garden (P=0.043, final F test) and
Time and Region and Garden (P<0.001, final F test). There was no significant interaction
between Time and Region (P = 0.199, final F test), suggesting that the genetic influence is
holding over time. There were also no significant interactions between Time and Nursery
(P=0.358, final F test) or Time and Region and Nursery (P=0.836, final F test).

4.4

Discussion

4.4.1 Garden Effects
Paper birch grown at Skimikin exhibited better overall growth than those grown at Red Rock.
This may be due to differences in soils and climate between the two gardens. The soils at
Red Rock are predominantly sand with low organic content; whereas the soils at Skimikin
are sandy loam with high coarse fragment content (Hawkins personal communication Feb.
18, 2009). In general, the climate at Skimikin is more favourable than at Red Rock. The
mean annual temperature at Skimikin is 7.2°C compared to 4.0°C at Red Rock. The mean
annual rainfall at Skimikin is 487mm versus and 419mm at Red Rock. Furthermore, there
are 136 frost free days at Skimkin compared to only 85 at Red Rock (Canadian Climate
Normals: http://climate.weatheroffice.ec.gc.ca/climate normals (August 3, 2005)).

Hawkins (Manuscript in preparation) carried out a laboratory study in 2002 with paper birch
to determine the influence of soil temperature on bud flush and subsequently on height
growth. The results of the study revealed that birch with warm roots (mean 14°C throughout
study) flushed earlier and had better overall height growth than those with cold roots (mean
1°C to early March, soils then warmed slowly to a mean of 12°C by mid-March). It was

55

noted that in some years the soils at the Red Rock site remain very close to freezing well into
the growing season. Fraser (1956) also identified low air and soil temperatures in the month
of May (in field studies at Chalk River, Ontario) as possible contributors to a delay in the
initiation of growth in paper birch. Cold soils would not be a concern at Skimikin during the
bud flush and early growth phase. Figure 4.3 demonstrates how the soils at Skimikin have
the potential to warm in early April, whereas the soils at Red Rock potentially warm very
slowly with mean daily maximums < 5°C and minimums below 0°C for much of April.

Even though differences in temperature, precipitation and number of frost free days likely
contributed significantly to the reduced overall height growth at Red Rock, abiotic (frost) and
biotic (herbivory) factors also likely played a significant role. Throughout the duration of the
study, trees grown at Red Rock were repeatedly damaged by frost, deer and voles (Table
4.3). Southern provenances (Salmon Arm and Idaho) received most of the damage, whereas
mid-latitude provenances (Prince George and Prince Rupert) received less damage and the
Fort Nelson provenance received the least damage and had the greatest number of healthy
trees. The rate of mortality for all regions, however, was acceptable at less than 10%.
Although not recorded, damage from abiotic and biotic factors at Skimikin were minimal.

Several studies involving silver birch in Scandinavia (Rousi et al. 1990, Rousi et al. 1993 and
Rousi et al. 1997) revealed that there is substantial variation among geographic origins of
trees in resistance to herbivory and it is thought to be an evolutionary response. The best
predictor of resistance to herbivory (in these studies) was found to be the number of resin
droplets in the bark (Rousi et al. 1990). These results are supported by Bryant et al. (In

56

press) who studied the influence of fire frequency, hare populations and herbivory pressures
in Alaska birch and paper birch. They determined that Alaska birch has more resin glands
and significantly higher concentrations of papyriferic acid (P<0.001) than paper birch and
those higher papyriferic acid concentrations were associated with increasing hare densities
(higher herbivory pressure). They also determined that higher Alaska birch densities are
more closely associated with higher fire frequency than paper birch (P<0.001) suggesting
"the existence of a fire driven birch-hare geographic mosaic of selection" (Bryant et al. In
press p. 16). Based on observations, the Fort Nelson provenance of paper birch (which
overlaps with the natural range of Alaska birch in the Fort Nelson area) has more resin glands
per unit area than the other provenances of paper birch studied and may have higher
concentrations of papyriferic acid than the southern provenances, minimizing the effects of
herbivory. It would be interesting, in a future laboratory study, to determine if the Fort
Nelson provenance of paper birch has higher concentrations of papyriferic acid than more
southerly provenances of paper birch.

4.4.2 Regional Effects
Photoperiod may have played a significant role in the regional differences observed in the
overall height growth of individual populations of paper birch; especially for the population
from Fort Nelson, which exhibited the poorest overall height growth when grown at the
southern gardens. Figure 3.6 illustrates that northern sites have longer photoperiods than
southern sites. When the Fort Nelson provenance was planted in the southern gardens, it was
exposed to shorter day lengths than it was habituated to in its natural environment resulting

57

in reduced growth. The maximum day length at Skimikin was similar to the May day length
in Fort Nelson.

It is well known that short days induce growth cessation, cold acclimation and dormancy
(Vaartaja 1954; Veen 1954; Downs and Borthwick 1956; Li et al. 2003) in photoperiodically
sensitive plants and that "photoperiodic responses often differ between the northern and
southern ecotypes of the same species" (Li et al. 2003 p. 131). Northern ecotypes typically
exhibit earlier growth cessation, cold acclimation and dormancy than their southern
counterparts. However, they also have higher freezing tolerances (Li et al. 2003).
Accordingly, height growth response of birch in the study by Li et al. (2003) was closely
related to the length of the frost-free growing season at their place of origin. The frost free
period at Fort Nelson (Beaver Lake) is 90 days compared to 136 days at Skimikin.
Differences in the length of the growing season were reflected in differences in height growth
between ecotypes; northern provenances being the shortest and southern provenances being
the tallest (Li et al. 2003). Down and Borthwick (1956) studied the effects of photoperiod on
several tree species. Although there were differences in individual species' responses,
growth of the entire tree species tested in the study were influenced by photoperiod; long
days prolonged growth, whereas short days induced dormancy. These results are supported
by Hawkins (manuscript in preparation) who carried out a laboratory study in 2001 with
paper birch to determine the influence of day length on bud flush and subsequently on height
growth. The study revealed that birch grown under long days (16 hours) flushed earlier than
those grown under short days (8 hours); leading to an earlier onset of growth. Air
temperature was the same for both photoperiods.

58

Provenances from the Idaho and Salmon Arm regions were exposed to longer day lengths
than they are habituated to in their natural environments when they were planted at the Red
Rock garden, north of their natural ranges. Due to the longer days, these provenances did
not receive the appropriate day length cues and possibly did not have sufficient cold
hardiness prior to the first fall frosts; as a result many of them were stressed, severely
damaged or killed. In a study with Scandinavian birch species, Vaartaja (1954) observed
similar results; the abnormally long photoperiods experienced by trees transferred north of
their origin did not permit them to develop proper dormancy, leaving them susceptible to fall
frosts. Loehle (1998) noted that the most limiting factor a tree faces when grown north of its
usual range is frost damage. Alternatively, because they were stressed, they may have been
more attractive to pests and disease (Zobel and Talbert 1984); the cause of death in some
individuals may have been either frost or girdling by voles.

4.4.3

Garden by Nursery Interactions

Statistically there was a significant interaction between garden and nursery (P=0.036, final F
test) however there is little difference (within a garden) in growth among the trees grown at
the three nurseries. There is a greater difference in growth between the trees of the two
gardens (as discussed above). One factor that may have had an influence on the outcome of
the study, however, is that the seedlings grown at the University of Idaho were cold stored,
but not frozen. As a result of the longer time in cold (+2 °C) storage, many of the seedlings
were planted mouldy and weak at Red Rock.

59

Additionally, all the trees from the Idaho provenance grown at the Northwood nursery were
dead at the Red Rock garden and the majority of the trees from this same provenance grown
at the other two nurseries were also dead at the Red Rock garden; resulting in the complete
removal of the Idaho provenance from the final analysis, and may be an indication of
displacement and nursery after effects.

4.4.4

Garden by Region Interactions

In general, trees grown at Skimikin exhibited better overall growth (except FN) than those
grown at Red Rock. At Red Rock, provenances from the Prince Rupert, Prince George and
Fort Nelson regions exhibited similar growth; whereas the provenances from the Salmon
Arm region had the poorest growth. In this study it appears that the best growth occurred
with local provenances at local sites, whereas provenances transferred considerable distances
from their places of origin exhibited the poorest growth (Fort Nelson provenance at
Skimikin; Idaho and Salmon Arm provenances at Red Rock). This may be due to light,
temperature and moisture conditions at the common garden sites differing significantly from
those at the site of origin.

Hemery et al. (2005) carried out provenance trials with walnut (Jugulans regia L.) in Great
Britain. Although they observed significant (P < 0.05) linear relationships for height
increment against elevation, latitude and longitude, they noted "reduced tree height
increment with increasing distance of origin from the planting location" (p. 128) and noticed
that provenances from comparable environments to the common garden location performed
best. In a study established in southern Sweden by Eriksson and Jonsson (1986) involving

60

silver and downy birch provenances from a range of latitudes, results showed that there was
considerable growth reduction following a long distance southward transfer. In another
study involving silver birch, results showed that "long distance transfers in both directions
tended to result in growth reductions" (p. 425). However, they noted that short distance
transfers did not influence height growth considerably (Eriksson and Jonsson 1986). Worrell
et al. (2000) observed similar results when they carried out provenance trials with
Scandinavian birch species in Scotland. They observed that Scandinavian provenances had
poorer height growth and survival than the local Scottish controls.

Finally, due to unseasonably dry conditions at Skimikin in 2001 (average of 25% less
precipitation over historical norms for the period of May through September), an irrigation
system was installed to prevent the loss of the entire plantation. This may have had an
influence on the final outcome of the study. Furthermore, due to the poor soil nutrient
conditions at Red Rock, all trees were fertilized in 2000, 2001, 2002 and 2003 with a mixture
of 14-16-10 (N-P-K) and 12% sulphur. The trees each received 25, 30, 60 and 60 grams of
fertilizer in each year, respectively. Table 4.2a shows that growth of the trees at the Red
Rock site started to diverge from those at the Skimikin site in 2004. This may be due to the
discontinuation of the fertilization treatments at Red Rock after 2003 or recovery from the
drought at Skimikin.

4.4.5

Conclusion

This study has shown that the geographic variation in height growth of paper birch does not
definitively follow latitudinal, longitudinal or altitudinal clines. Instead, 'local' provenances

61

showed the best performance. The Prince Rupert and Prince George provenances grew well
outside their natural ranges showing good survival and growth at both the Skimikin and Red
Rock gardens. Salmon Arm and Fort Nelson provenances, however, performed poorly
outside of their natural ranges. What lead to the poor performance of the Fort Nelson
provenance at the Skimikin site and the Idaho and Salmon Arm provenances at the Red Rock
site? Were climatic conditions (temperature, precipitation) too different? Were soil
conditions (nutrients, moisture, freezing) a factor? Were the changes in photoperiod too
great? Was herbivory a factor? One or more of these factors likely played a role in the
performance of the different provenances at the common garden sites.

Although the geographic variation in height growth of paper birch is influenced by many
different environmental factors, the range of responses of a given provenance to these
environmental factors is limited by its genotype. This study involved extreme north south
transfers of provenances; the Idaho provenance was transferred approximately 6 degrees
latitude north to the Red Rock garden and the Fort Nelson provenance was transferred
approximately 10 degrees latitude south to the Skimikin garden. Generally with tree species,
it is easier to move them north than south (O'Neill and Yanchuk 2005). The present seed
transfer guidelines in BC are based on this (Ying and Yanchuk 2006). Therefore it is
interesting to note that the Prince George and Prince Rupert provenances were moved as far
south (to the Skimikin garden) as the Salmon Arm provenance was moved north (to the Red
Rock garden), yet the Prince George and Prince Rupert provenances had fewer differences in
growth performance between the two gardens than the Salmon Arm provenance. This may

62

be illustrating that the response of a given provenance to its environment is limited to a
greater or lesser extent by its genotype.

63

4.5

Tables and Figures

Table 4.1: Height Growth Results from Univariate Repeated Measures Analysis in
GLM, significant (q=0.05) findings are in bold and shaded.
Between Subjects
P
Source
SS
DF
MS
F
Garden
521953.711
1
521953.711
50.284
< 0.001
2
16900.020
Nursery
33800.040
1.628
0.198
Region
161401.228
3
53800.409
5.183
0.002
2
G*N
69448.057
34724.028
3.345
0.036
G*R
468218.910
3
156072.970
15.036
< 0.001
91101.364
R*N
6
15183.561
1.463
0.190
Error
4203981.164
10380.200
405
Within Subjects
DF
F
Source
P
G-G
H-F
SS
MS
9683770.428
1613961.738 1363.572 < 0.001 0.000
Time
6
0.000
292279.049
6
41.156
0.000
T*G
48713.175
< 0.001 0.000
15583.297
12
1.097
0.351
T*N
1298.608
0.358
0.353
T*R
27002.173
18
1500.121
1.267
0.277
0.276
0.199
25559.177
12
0.142
T*N*G
2129.931
1.799
0.139
0.043
T*R*G
165014.337
7.745
0.000
9167.463
18
< 0.001 0.000
T*R*N
32816.783
36
0.770
0.652
911.577
0.836
0.658
Error
2876216.103
2430
1183.628
Grenhouse-Geisser Epsilon:
0.2656
Huynh-Feldt Epsilon:
0.2777

64

Table 4.2a: Summary of Means (±: SEM) for Total Heig it (cm) by"¥ear.
2002
2001
2004
2000
2003
2005
Garden
54.1
88.8
113.4
176.9
RR
155.0
213.7
±2.2
±3.2
±4.4
±5.4
±1.2
±1.9
133.3
218.4
SK
70.9
96.8
167.1
276.0
±2.4
±2.7
±4.0
±5.5
±6.8
±1.5
Nursery
101.7
131.6
167.8
200.8
246.4
LN
73.9
±5.4
±2.4
±2.7
±4.0
±1.5
±6.7
88.7
118.6
157.1
195.5
NW
58.5
243.9
±2.4
±2.7
±5.4
±1.5
±4.0
±6.8
55.2
88.0
119.8
158.3
196.8
244.3
UI
±3.1
±3.5
±5.1
±7.0
±8.7
±1.9
Region
73.3
111.9
171.5
FN
48.8
142.0
212.4
±5.0
±5.7
±8.4
±11.4
±3.1
±14.1
130.2
69.8
101.9
171.6
208.1
PG
258.5
±2.0
±2.3
±3.4
±4.6
±5.7
±1.3
94.2
126.5
65.6
164.6
204.0
253.0
PR
±2.0
±2.3
±3.4
±4.7
±5.8
±1.3
SA
124.8
65.9
101.7
166.0
207.1
255.5
±2.4
±2.8
±1.5
±4.0
±5.5
±6.8

2006
252.8
±6.5
333.3
±8.2
293.1
±8.1
291.3
±8.1
294.6
±10.4
254.6
±17.0
311.1
±6.8
301.7
±7.0
304.7
±8.2

Table 4.2b: Summary of Means (± SEM) for Total Height (cm) by Year and Nursery at
Red Rock and Skimikin.
LN
NW
UI
Year
SK
RR
RR
SK
RR
SK
67.5
80.3
49.7
67.2
45.1
65.2
±2.1
±2.0
±2.0
±2.1
2000
±2.7
±1.9
102.4
83.5
94.0
94.1
101.0
81.8
±3.1
±3.4
±3.3
±3.3
±3.4
±4.3
2001
108.5
128.6
125.2
138.1
106.4
133.3
2002
±3.5
±3.9
±3.7
±3.7
±4.9
±3.9
168.5
152.8
161.4
167.1
145.0
171.6
±5.1
±5.7
±5.5
±5.5
±7.2
2003
±5.7
216.5
177.9
213.0
185.1
167.7
255.8
2004
±7.5
±7.5
±7.0
±7.8
±7.7
±9.8
269.8
216.2
271.5
223.0
202.0
286.7
±9.3
±9.3
±8.7
±9.7
±9.6
± 12.2
2005
324.7
256.1
326.6
348.6
261.5
240.7
±10.4
±11.7
±11.1
2006
±14.6
±11.1
±11.5

65

Table 4.2c: Summary of Means (± SEM) for Total Height (<cm) by Year and Region at
Red Rock and Skimikin.
FN
PR
SA
PG
Year
RR
SK
SK
RR
RR
SK
RR
SK
43.5
54.1
77.7
58.8
61.9
72.3
52.3
79.5
±3.4
±1.7
±2.4
2000
±5.1
±1.6
±1.9
±1.9
±1.8
76.4
70.2
100.1
103.8
90.0
98.5
88.6
114.9
±3.0
±5.4
±8.1
±2.6
±2.8
±3.0
±3.9
±2.9
2001
125.0
98.7
117.9
142.5
115.3
137.8
95.3
154.4
2002
±6.2
±9.3
±3.5
±3.4
±4.4
±3.0
±3.2
±3.3
161.0
122.9
161.4
181.9
156.3
172.9
141.1
190.9
2003
±9.1
±13.6
±4.4
±5.1
±4.7
±5.0
±6.5
±4.8
189.9
153.1
180.2
235.9
183.9
224.2
153.7
260.5
±6.3
2004
±12.5
±18.5
±6.0
±7.0
±8.8
±6.9
±6.6
230.6
194.2
220.1
297.0
221.7
284.3
182.6
328.4
±8.6
2005
±15.4
±7.5
±7.9
±8.5
±10.9
±8.2
±23.0
267.6
241.6
360.2
258.8
344.5
222.7
261.9
386.8
±10.4
±9.4
2006
±18.5
±27.5
±9.0
±10.2
±13.1
±9.8
Table 4.3: Health Survey Results at Red Rock in 2002 (Damaged category includes
trees with minor and or major top and or basal damage due to abiotic and biotic
factors) (SA=Salmon Arm; ID=Idaho; PG=Prince George; PR=Prince Rupert;
FN=Fort Nelson).
Region
Damaged (%)
Dead(%)
Healthy (%)
SA
6.9%
68.1%
25.0%
8.5%
66.0%
ID
25.5%
2.1%
53.8%
44.1%
PG
PR
6.3%
44.7%
49.0%
FN
8.3%
27.1%
64.6%

66

400

350
300

f"

o. 2 5 0
x:
«•»
3
2 200

o
'3 150
x
100

50

LN

NW

Ul

Nursery
• Red Rock •

- Skirrikin

Figure 4.1: Mean Total Height (cm) + SEM by Garden and Nursery in 2006
(LN=Landing Nursery; NW=JD Little Forestry Centre; UI=University of Idaho
Nursery)

67

Figure 4.2: Mean Total Height (cm) ± SEM by Garden and Region in 2006 (SA=Salmon
Arm; PR=Prince Rupert; PG=Prince George; FN=Fort Nelson; RR=Red Rock;
SK=Skimikin).

S

Ol

i

O

Celsius)

30

_/\

15
/ ' " • \ J

3

5

/V

•

,A/\

•
*

5 °
Q-

C".'

1!'l:

•

liu

A

^

/~/~~V'

\AfV

V . ... '..

•

*

iL''i

r.o

140

I:J0

15

-10- •
-15

Julian Days
PG Max. —- P G M i n .

SA Max.

SA Min.

Figure 4.3: Mean minimum and maximum daily air temperature (°C) for Prince
George (PG) and Salmon Arm (SA) for the period of April 1 (90 Julian Days) to May 30
(150 Julian Days) for the years 2001 to 2003 (Source: Canadian Climate Data).

68

CHAPTER FIVE SUMMARY DISCUSSION
5.1 Summary of Key Findings
The overall objective of this study was to determine if the extent of phenotypic variability
observed in paper birch was due to genetics, the local environment, or an interaction between
the two by examining phenological (bud flush) and morphological (height growth) traits. It
is necessary to have a clear understanding of the geographic genetic variation of a species
before developing seed zones and seed transfer guidelines. Furthermore, in order for paper
birch to be managed successfully as a commercial species in British Columbia, we need to: a)
understand the genecology of the species, b) identify potential tree improvement
opportunities, c) develop a selective breeding program and d) forecast the potential impacts
of climate change on its future performance. The results of this study begin to address some
of these issues.

The key findings from this study indicate that although the geographic variation in the timing
of bud burst in paper birch is under genetic and environmental control and generally follows
latitudinal, longitudinal and elevational clines, the signal for the onset of bud flush in the
spring appears to be determined by a complex interaction between air and soil temperature
and photoperiod. Although the amount of chilling a plant receives also influences the timing
of bud flush, it is not a concern in this trial because all three gardens are located on
continental sites which are exposed to 'real' winters; implying that adequate chilling is
received at all three sites. Northern and high elevation provenances require less heat for bud
flush to occur than southern and low elevation provenances. Photoperiod, however, appears
to have a modifying influence on heat requirements with plants growing under long days

69

requiring less heat than those growing under short days. Furthermore, for a given
photoperiod, continental provenances require less heat and flush earlier than coastal or
maritime provenances when grown at a continental site. Additionally, studies, including the
one by Hawkins (manuscript in preparation) have shown that soil temperature can further
influence the timing of bud flush with plants growing in cold soils taking longer to flush than
those growing in warm soils when the above ground environments are similar.

This study determined that long distance displacement of provenances from their places of
origin leads to detrimental abiotic and biotic effects as found in other studies for other
species. Southern provenances grown north of their natural ranges possibly did not receive
the appropriate day length cues and therefore developed insufficient cold hardiness leading to
frost damage. These southern provenances were also more susceptible to herbivory than the
northern provenances. Both factors lead to poor survival and height growth. Northern
provenances grown south of their natural ranges were grown under much shorter
photoperiods compared to their places of origin forcing them into earlier growth cessation,
cold acclimation and dormancy. Although these northern provenances had higher freezing
tolerances and lower susceptibility to herbivory, they also experienced poor survival and
growth. Differences in stock handling among the nurseries and nursery displacement effects,
although not statistically detected, also likely played a role in the poor performance of some
provenances.

Alternatively, differences in incoming radiation may be playing a role. Light is one of many
environmental factors that control plant growth and development. Plants absorb red

70

(wavelength of 650 nm) and far-red (wavelengths of 700-800 nm) light through
photoreceptors called phytochrome. These different wavelengths of light influence plants in
different ways; red light promotes growth whereas far-red light inhibits growth.
Interestingly, in most species, it is the leaves that detect photoperiod; however, in birch as
well as in beech (Fagus sp.), larch (Larix sp.), yellow poplar (Liriodendron tulipifera L.),
sweetgum (Liquidambar sp.) and red oak (Quercus rubra), the buds detect daylength directly
(Salisbury and Ross 1992).

5.2 Seed Zones and Seed Transfer Guidelines
When developing seed zones and seed transfer guidelines for paper birch in British
Columbia, it will be important to take abiotic and biotic factors into consideration.
Furthermore, it will be important to incorporate the elements of climate change. Warming
trends may reduce cold injury for provenances transferred further north, however, it will do
little for Fort Nelson provenances moved south as photoperiod is 'controlling' seasonal
growth. Short distance transfer of provenances to sites with similar climatic and
environmental conditions as their places of origin is possible and may even be beneficial.
However, long distance transfer, particularly northward, of provenances is not recommended
due to the severe impacts on growth and survival. Can sources from southern BC stand
exposures to winter temperature below -30 °C or the probability of frost in any month of the
growing season? In this study, the Prince Rupert and Prince George provenances grew well
outside their natural ranges - showing good survival and growth at both the Skimikin and
Red Rock gardens. Idaho, Salmon Arm and Fort Nelson provenances performed poorly
outside of their natural ranges, yet which factors (temperature, precipitation, nutrients,

71

moisture, frozen soils, photoperiod, herbivory) contributed to the poor survival and growth of
these provenances is unknown. Further research is necessary to pinpoint the exact cause or
causes. This could be determined by repeating this trial and placing greater emphasis on
monitoring these environmental factors. Air temperature and precipitation data could be
collected from nearby Environment Canada weather stations. Soil moisture and temperature
could be monitored on site. Soil samples could be collected for further nutrient analysis and
biotic damage from insects, disease and herbivory could be identified and recorded more
precisely.

5.3 Potential Impacts of Climate Change
Increasing levels of atmospheric CO2, methane and other greenhouse gases are resulting in
changes to our climate which will have far reaching consequences for our forests (Ruel and
Ayres 1996; NRC 2002). These changes may lead to, among others, an increased risk of
forest fires, an increased frequency of extreme weather events, forest dieback, range
shrinkages and an increased risk of insect outbreaks (Loehle 1998; NRC 2002).

The phenological characteristics of a tree species are critical components in its adaptation to
its local climate (Kramer 1995). The timing of phenological events such as bud burst affects
tree productivity and resistance to abiotic and biotic hazards (Rousi and Heinonen 2007).
Recent research has shown that as a consequence of climate warming, phenological events
have advanced, on average, by 2.7 days in the last decade in Canada (Beaubien and Freeland
2000) and 2.1 days (Menzel 2000) to 2.7 days (Chmielewski and Rotzer 2001) per decade in
Europe. Where chilling requirements are presently not met or are barely met, warming

72

temperatures may delay bud flush and the onset of growth. Conversely, where chilling
requirements are currently being met, warming temperatures may lead to an earlier onset of
growth, a longer growing season and increased growth potential. However, early growth
initiation also increases the risk of damage due to late spring frosts (Cannell and Smith 1986;
NRC 2002). The latter is likely the scenario for paper birch, however photoperiod may
further influence the timing of bud flush in northern provenances which may further increase
the risk of late spring frosts.

Changes in temperature may force tree species to migrate slowly northward and upward in
elevation, which may lead to local extinctions (Junttila and Nilsen 1993; Ruel and Ayres
1996; Loehle 1998; NRC 2002). Even though species have migrated in response to climate
change in the past, the predicted speed of change associated with current climate change,
may limit their ability to respond to these changes in the present (Billington and Pelham
1991; Thompson 1998; NRC 2002). Facilitated migration may be required to ensure species'
success (Spittlehouse and Stewart 2004). Fortunately paper birch is a generalist and as a
result should be able to survive in a variety of different climatic conditions and tolerate larger
changes in the environment than specialists. This evolutionary strategy may benefit paper
birch through future climate change. However, if changing temperatures lead to more
thaw/freeze cycles in the winter; resulting in tissue damage, followed by more frequent
drought conditions in the summer, paper birch may be more vulnerable to drought, insects
and disease and subsequently to dieback. Recovery from dieback is possible in the absence
of drought, however, if drought conditions persist; recovery may not be possible which may
lead to the loss of entire populations of paper birch.

73

Hamann and Wang (2006) developed an ecosystem-based climate envelope modeling
approach to assess the impact of climate change on forest ecosystems and tree species in
British Columbia. By the year 2085, significant changes are expected. It is anticipated that
mountainous ecosystems will be particularly vulnerable. Moreover, ecosystems where
interior cedar and hemlock, Douglas-fir and ponderosa pine grow are expected to expand
replacing current sub-boreal and boreal ecosystems. Hamann and Wang (2006) also modeled
changes in the distribution and frequency of British Columbia tree species. Many broadleaf
species, including paper birch, did not appear to be impacted by climate change; many
actually increased their frequency and distribution across the province. Conversely conifer
species are expected to decrease in frequency and lose considerable portions of their habitat.

However, tree species, provenances, families and clones exhibit substantial variation in their
adaptability (Thompson 1998). Species with large geographic ranges (including paper birch)
typically exhibit the highest levels of genetic variation. This variation should allow these
species to adapt to climate change (Thompson 1998; Beaulieu and Rainville 2005).
Additionally, forest managers can assist with adaptation by anticipating these changes and
adjusting their forestry practices accordingly (NRC 2002). Silvicultural practices such as
altering the mixture of species or provenances or planting hybrid species has been suggested.
Also, shorter rotation species may provide greater flexibility in an uncertain climatic future
(Thompson 1998).

74

The results of this study suggest that the observed variation in leaf phenology and height
growth in paper birch may be adaptive in nature. As suggested by Lechowicz (1984), the
adaptive explanation predicts there is a correlation between leaf phenology and other traits
such that the combination of traits improves survival and reproduction over every other
possible combination of traits. The adaptation of some paper birch provenances to herbivory,
frost and drought tolerance leaves them more successful than other less adapted provenances;
however there is a trade-off. Even though the Fort Nelson provenance is very frost tolerant
and highly resistant to herbivory, it may be spending too much energy on these traits to the
detriment of growth, whereas the Salmon Arm and Idaho provenances lack frost tolerance
and resistance to herbivory, which is also detrimental. The Prince George and Prince Rupert
provenances appear to have sufficient frost tolerance and a moderate level of resistance to
herbivory, allowing them to invest more energy into growth and reproduction. Warming
trends may reduce the risk of cold injury for southern (Prince George) and coastal (Prince
Rupert) provenances transferred further north and inland, respectively; however, efforts
should be made to ensure similar climatic and environmental conditions are maintained.

75

REFERENCES
Ager, A.A., P.E. Heilman and R.F. Stettler. 1993. Genetic variation in red alder (Alnus
rubra) in relation to native climate and geography. Canadian Journal of Forest Research.
23(9): 1930-1939.
Aitken, S. 2004. Genecology and adaptation of forest trees. In Encyclopedia of Forest
Sciences. Burley, J., J. Evans and J.A. Youngquist (eds.). Elsevier Academic Press, London,
pp. 197-204.
Auclair, A.D., J.T. Lill and C. Revenga. 1996. The role of climate variability and global
warming in the dieback of northern hardwoods. Water, Air and Soil Pollution. 91:163-186.
Auclair, A.D., P.D. Eglinton and S.L. Minnemeyer. 1997. Principal forest dieback episodes
in northern hardwoods: development of numeric indices of areal extent and severity. Water,
Air and Soil Pollution. 93:175-198.
Baliuckas, V., I. Ekberg, G. Eriksson and L. Norell. 1999. Genetic variation among and
within populations of four Swedish hardwood species assessed in a nursery trial. Silvae
Genetica. 48(1): 17-25.
Beaubien, E.G. and Freeland, H.J. 2000. Spring phenology trends in Alberta, Canada: links
to ocean temperature. International Journal of Biometerology 44:53-59.
Beaulieu, J. and A. Rainville. 2005. Adaptation to climate change: Genetic variation is both
a short- and long-term solution. Forestry Chronicle 81(5): 704-709.
Berger, V.G. 2001. Paper birch genecology trial. Proposed Phenology Survey for Spring
2001. Unpublished.
Billington, H.L. and J. Pelham. 1991. Genetic variation in the date of budburst in Scottish
birch populations: implications for climate change. Functional Ecology. 5:403-409.
Braathe, P. 1995. Birch dieback - caused by prolonged early spring thaws and subsequent
frost. Norwegian Journal of Agricultural Sciences. Supplement No. 20.
Brissette, J.C. and B.V. Barnes. 1984. Comparisons of phenology and growth of Michigan
and western North American sources of Populus tremuloides. Canadian Journal of Forest
Research. 14:789-793.
Bryant, J.P., T.P. Clausen, R.K. Swihart, S.M. Landhausser, C.D.B. Hawkins, M.T. Stevens,
S. Carriere, A.P. Kirilenko, A.M. Veitch, R.A. Popko, D.T. Cleland, J.H. Williams, W.
Jakubas, M.R. Carlson, K.L. Lehmkhul Bodony, M. Cebrian, T.F. Paragi, P. M. Picone, J. E.
Moore, E.C. Packee and T.T. Malone. In press. Fire drives transcontinental variation in tree
birch defense against browsing by snowshoe hares. American Naturalist.

76

Callaham, R.Z. 1962. Geographic variability in growth of forest trees. In Tree Growth.
Kozlowski, T.T. (ed.). Ronald Press Company, New York. pp. 311-325.
Campbell, K.A. 2000. Root strength of paper birch seed sources grown under different
nursery cultures, five years after planting. Proceedings for the 2000 Forest Nursery
Association of BC Meeting, Sept. 18-21, Prince George, BC. pp. 29-37.
Campbell, R.K. and F.C. Sorensen. 1984. Genetic implications of nursery practices. In
Forest Nursery Manual: Production of Bareroot Seedlings. Duryea, M.L. and T.D. Landis
(eds.). Martinus Nijhoff/Dr. W. Junk Publishers, The Hague, pp. 183-191.
Cannell, M.G.R. and R.I. Smith. 1986. Climatic warming, spring budburst and frost damage
on trees. Journal of Applied Ecology. 23:177-191.
Carlson, M.R., C.D.B. Hawkins, V. Berger, R. Mahoney and B.D. Hooge. 1999. Birch
genecology study seedlot numbering scheme. Unpublished.
Carlson, M.R., V.G. Berger and C.D.B. Hawkins. 2000a. Seed source testing of paper birch
(Betula papyrifera) in the interior of British Columbia. Journal of Sustainable Forestry.
10(l/2):25-34.
Carlson, M.R., C.D.B. Hawkins, V. Berger, R. Mahoney and B.D. Hooge. 2000b. A
genecology study of paper birch in British Columbia and Idaho. BC Ministry of Forests and
Range. EP106913. Unpublished.
Chipoulet, J.M. 1981. Observation sur la foliasion et le developpement du feuillage des
bouleaux verruqueux de la butte du Montmorency. Meteorologie. 6:83-86.
Chmielewski, F.M. and T. Rotzer. 2001. Response of tree phenology to climate change
across Europe. Agricultural and Forest Meterology. 108(2):101-112.
Chmura, D.J. and R. Rozkowski. 2002. Variability of beech provenances in spring and
autumn phenology. Silvae Genetica. 51(2/3):123-127.
Clausen, J., Keck, D. and Hiesey, W. 1940. Experimental studies on the nature of species I.
Effect of Varied Environments on Western North American Plants. Carnegie Institution of
Washington Publication No. 520. Washington, D.C.
Down, R.J. and H.A. Borthwick. 1956. Effects of photoperiod on growth of trees. Botanical
Gazette. 117:310-326
Eriksson, G. and A. Jonsson. 1986. A Review of the genetics of Betula. Scandinavian
Journal of Forest Research. 1:421-434.

77

Fraser D.A. 1956. Ecological studies of forest trees at Chalk River, Ontario, Canada. II.
Ecological conditions and radial increment. Ecology. 37. 777-89.
Hakkinen, R. 1999. Analysis of bud-development theories based on long-term phenological
and air temperature time series: application to Betula sp. Leaves. Finnish Forest Research
Institute, Research Paper 754, 81 p.
Hamann, A. and T. Wang. 2006. Potential effects of climate change on ecosystem and tree
species distribution in British Columbia. Ecology. 87(11):2773-2786.
Hanninen, H. 1991. Does climatic warming increase the risk of frost damage in northern
trees? Plant, Cell and Environment. 14:449-454.
Hawkins, C.D.B. Factors influencing bud break in 18 sources of paper birch: air temperature,
photoperiod, and locale. Manuscript in preparation.
Hawkins, C.D.B. 2005. Ten year field performance of six spruce seed orchard and wild
seedlots grown at seven nurseries. Forest Genetics Council Operational Tree Improvement
Program Final Report SPU0403.
Hawkins, C.D.B. 1998. Interactions of forest seedling nurseries and seed orchards. In Tree
Improvement: Applied Research and Technology Transfer. Puri, S. (ed.). Science
Publishers Inc. pp 125-139.
Heide, O.M. 1993. Daylength and thermal time responses of budburst during dormancy
release in some northern deciduous trees. Physiologia Plantarum. 88:531-540.
Heide, O.M. 2003. High autumn temperature delays spring bud burst in boreal trees,
counterbalancing the effect of climatic warming. Tree Physiology. 23:931 -936.
Hemery, G.E., P.S. Savill and A. Thakur. 2005. Height growth and flushing in common
walnut (Juglans regia L.): 5-year results from provenance trials in Great Britain. Forestry.
78(2):121-133.
Heslop-Harrison, J. 1964. Forty years of genecology. In Advances in Ecological Research
Cragg, J.B. (ed.). Academic Press, 159-247.
Jones, E.A., D.D. Reed, G.D. Mroz, H.O. Liechty and P.J. Cattelino. 1993. Climate stress as
a precursor to forest decline: paper birch in northern michigan, 1985-1990. Canadian Journal
of Forest Research. 23:229-233.
Jones, M.D., D.M. Durall, M. Harniman, D. Classen and S.W. Simard. 1998.
Ectomycorrhizal diversity of paper birch and douglasfir seedlings grown in single-species
and mixed plots in the ICH zone of southern British Columbia. BC Ministry of Forests
Research Program, Extension Note 19, February 1998.

78

Jozsa, L. 2000. Separating birch and maple, based on macroscopic features. Crosscut
(FRBC Journal). Kutenai House Press. Fall, 2000.
Junttila, O. and J. Nilsen. 1993. Growth and development of northern forest trees as affected
by temperature and light. In Forest Development in Cold Climates. Alden, J. et al. (eds.).
Plenum Press, New York, pp. 43-57.
Korner, Ch. and W. Larcher. 1988. Plant life in cold climates. Plants and temperature.
Long, S.P. and F.I. Woodward (eds.). Society Experimental Biology, Cambridge, 25-57.
Koski, V. and M. Rousi. 2005. A review of the promises and constraints of breeding silver
birch (Betula pendul Roth) in Finland. Forestry. 78(2): 187-198.
Kozlowski, T.T. 1971. Growth and development of trees. Volume 1: Seed Germination,
Ontogeny, and Shoot Growth. Academic Press Inc., New York.
Kramer, K. 1995. Phenotypic plasticity of the phenology of seven european tree species in
relation to climatic warming. Plant, Cell and Environment. 18:93-104.
Lechowicz, M.J. 1984. Why do temperate deciduous trees leaf out at different times?
Adaptation and Ecology of Forest Communities. The American Naturalist. 124(6):821 -842.
Li, C , A. Vihera-Aarnio, T. Puhakainen, O. Junttila, P. Heino and E. T. Palva. 2003.
Ecotype-dependent control of growth, dormancy and freezing tolerance under seasonal
changes in Betula pendula Roth. Trees. 17:127-132.
Linkosalo, T. and M.J. Lechowicz. 2006. Twilight far-red treatment advances leaf bud burst
of silver birch {Betula pendula). Tree Physiology. 26:1249-1256.
Loehle, C. 1998. Height growth tradeoffs determine northern and southern range limits for
trees. Journal of Biogeography. 25:735-742.
Martin, E and Hine, R.S. (eds.). 2000. A Dictionary of Biology. 4th ed. Oxford University
Press, Oxford.
McGee, C.E. 1969. Bud-break on red oak seedlings varies with elevation of seed source.
Tree Planters Notes. 21(2): 18-19.
Menzel, A. 2000. Trends in phenological phases in Europe between 1951 and 1996.
International Journal of Biometerology. 44:76-81.
Morgenstern, E.K. 1976. The seed source-environment interaction: a factor in nursery
management. The Forestry Chronicle. 52(4): 199-204.
Morgenstern, E.K. 1996. Geographic variation in forest trees genetic basis and application of
knowledge in silviculture. University of British Columbia Press, Vancouver.

79

Morrison, J.I.L. and M.D. Morecroft (eds.). 2006. Plant growth and climate change.
Blackwell Publishing, Oxford.
Myking, T. and O.M. Heide. 1995. Dormancy release and chilling requirement of buds of
latitudinal ecotypes of Betula pendula and B. pubescens. Tree Physiology. 15:697-704.
Nielson, R. 2000. Birch log quality is critical to successful utilization. Crosscut (FRBC
Journal). Kutenai House Press. Fall, 2000.
Nienstaedt, H. 1974. Genetic variation in some phenological characteristics of forest trees.
In Phenology and Seasonality Modeling. Lieth, H. ed. Springer-Verlag, New York. pp.
389-400.
[NRC] Natural Resources Canada, Climate Change Impacts and Adaptation Directorate.
2002. Climate change impacts and adaptation: a Canadian perspective. Forestry. Natural
Resources Canada, Ottawa, Ontario.
Ogren, E. 2001. Effects of climatic warming on cold hardiness of some northern woody
plants assessed from simulation experiments. Physiologia Plantarum. 112: 71-77.
O'Neill, G. and A. Yanchuk. 2005. A primer on seed transfer for compliance and
enforcement in B.C. BC Ministry of Forests and Range, Research Branch. May 2005.
Peterson, E.B., N.M. Peterson, S.W. Simard and J.R. Wang. 1997. Paper birch managers'
handbook for British Columbia. B.C. Ministry of Forests, Research Branch, Victoria, B.C.
Ranney, T.G. and M.M. Peet. 1994. Heat tolerance of five taxa of birch: physiological
responses to supraoptimal leaf temperatures. Journal of the American Society of Horticultural
Science. 119(2):243-248.
Rehfeldt, G.E. 1993. Evolutionary genetics, the biological species, and the ecology of the
interior cedar-hemlock forests. Proceedings of the Interior Cedar-Hemlock-White Pine
Forests: Ecology and Management Conference, March 2-4, 1993, Spokane, Washington,
Washington State University, Pullman, WA. pp. 91-100.
Rohrig, E. 1991. Seasonality. In Ecosystems of the World 7. Temperate Deciduous
Forests. E. Rohrig and B. Ulrich eds. Elsevier Science Publishing Company Inc., New
York. pp. 25-33.
Rousi, M., H. Henttonen and A. Kaikusalo. 1990. Resistance of birch (Betula pendula and B.
platyphylla) seedlots to vole (Microtus agrestis) damage. Scandinavian Journal of Forest
Research. 5:427-436.
Rousi, M., J. Tahvanainen, H. Henttonen and I. Uotila. 1993. Effects of shading and
fertilization on resistance of winter-dormant birch (Betula pendula) to voles and hares.
Ecology. 7(l):30-38.

80

Rousi, M., J. Tahvanainen, H. Henttonen, D.A. Herms and I. Uotila. 1997. Clonal variation
in susceptibility of white birches (Betula spp.) to mammalian and insect herbivores. Forest
Science. 43(3):396-402.
Rousi, M. and J. Pusenius. 2005. Variations in phenology and growth of European white
birch {Betula pendula) clones. Tree Physiology. 25:201-210.
Rousi, M. and J. Heinonen. 2007. Temperature sum accumulation effects on withinpopulation variation and long-term trends in date of bud burst of European white birch
(Betulapendula). Tree physiology. 27:1019-1025.
Rowe, J.S. 1964. Environmental preconditioning, with special reference to forestry.
Ecology. 45(2):399-403.
Ruel, J.R. and M.P. Ayres. 1996. Variation in temperature responses among populations of
Betula papyrifera. Silva Fennica. 30(2-3): 145-158.
Safford, L.O., J.C. Bjorkburn, and J.C. Zasada. 1990. Betula papyrifera Marsh. Paper Birch.
In Silvics of North America. Volume 2. Hardwoods. Burns, R., and B. Honkala, eds.
Agricultural Handbook 654. Forest Service. USDA.
Salisbury, F.B. and C.W. Ross. 1992. Plant Physiology. Wadsworth Publishing Company.
Belmont, California.
Simard, S. 1996. Ecological and Silvicultural Characteristics of Paper Birch in the Southern
Interior of British Columbia. In Ecology and Management of B.C. Hardwoods. FRDA
Report 255, October 1996, pp. 157-167.
Spittlehouse, D.L. and R.B. Stewart. 2004. Adaptation to climate change in forest
management. BC Journal of Ecosystems and Management. 4( 1): 1 -11.
Sulkinoja, M. and T. Valanne. 1987. Leafing and bud size in Betula provenances of
different latitudes and altitudes. Report of the Kevo Subarctic Research Station 20:27-33.
Thomas, B. 1998. Photoperiodism: an overview. In Biological Rhythms and
Photoperiodism in Plants. Lumsden, P J . and A.J. Millar eds. BIOS Scientific Publishers
Ltd., Oxford, pp. 151-165.
Thompson, D.G. 1998. Getting the species and provenance right for climate change. Irish
Forestry. 55(2): 114-121.
Tirri, R., J. Lehtonen, R. Lemmetyinen, S Pihakaski and P. Portin (complied by). 1998.
Elsevier's Dictionary of Biology. Elsevier, New York.

81

Vaartaja, O. 1954. Photoperiodic ecotypes of trees. Canadian Journal of Botany. 32:392399.
Vaartaja, O. 1959. Evidence of photoperiodic ecotypes in trees. Ecological Monographs.
29:91-111.
Veen, B. 1954. General remarks on provenance research in forestry. Euphytica 3:89-96.
Wareing, P.F. 1953. Growth studies in woody species. V. Photoperiodism in dormant buds
of Fagus sylvaticaL. Physiologica Plantarum. 6:692-706.
Worrell, R., E.P. Cundall, D.C. Malcolm and R.A. Ennos. 2000. Variation among seed
sources of silver birch in Scotland. Forestry. 73(5):419-435.
Wright, J.W. 1976. Introduction to forest genetics. Academic Press Inc, New York.
Ying, C.C., C. Thompson and L. Herring. 1989. Geographic variation, nursery effects, and
early selection in lodgepole pine. Canadian Journal of Forest Research 19:832-841.
Ying, C.C. and A. Yanchuk. 2006. The development of British Columbia's tree seed
transfer guidelines: purpose, concept, methodology and implementation. Forest Ecology and
Management. 227(1-2): 1-13.
Zobel, B.J. and J.T. Talbert. 1984. Applied Tree Improvement. John Wiley and Sons, New
York.
Zhu, X.B., R.M. Cox, F.-R. Meng and P.A. Arp. 2001. Responses of xylem cavitation,
freezing injury and shoot dieback to a simulated winter thaw in yellow birch seedlings
growing in different nursery culture regimes. Forest Ecology and Management. 145:243253.
Zhu, X.B., R.M. Cox, C.-P.A. Bourque and P.A. Arp. 2002. Thaw effects on cold-hardiness
parameters in yellow birch. Canadian Journal of Botany. 80:390-398.

82

APPENDIX A: Results from Initial Bud Flush and Height Growth Analysis
Table A-l: ANOVA Results for Days to 20% and 80% Bud Flush in 2003, significant
(q=0.05) results are in bold and shaded.
F-Ratio
Sum-of-Squares
df
P
2
R : 0.0.836
Days to 20% Bud Flush
N:784
Garden
61305.186
2
1278.506
<0.001
Nursery
11.169
2
0.792
5.585
Provenance
17690.107
10
1769.011
<0.001
Garden * Nursery
56.532
4
14.133
0.670
Garden * Provenance
1962.168
20
98.108
<0.001
Nursery * Provenance
651.909
20
32.595
0.135
Error
17382.106
743
23.975
Days to 80% Bud Flush
Garden
Nursery
Provenance
Garden * Nursery
Garden * Provenance
Nursery * Provenance
Error

R2: 0.829
N:802
90815.030
27.467
15545.668
52.439
2274.703
1374.438
24106.352

2
2
10
4
20
20
743

83

1399.539
0.423
47.914
0.404
3.506
2.118

<0.001
0.655
<0.001
0.806
<0.001
0.003

132 j
128 124
CO

120

>O
C

ac 116
.2
3

-s 112

108
104
100
ID

SA
low

SA SA
mid high

PG
low

PG PG
mid high

PR
low

PR PR
mid high

FN

Provenance
• 20% Bud Flush

• 80% Bud Flush

Figure A-l: Mean Julian Days (± SEM) to 20% and 80% Bud Flush by Provenance in
2003 with Garden and Nursery pooled. Where error bars are not visible, they are
within the plotted symbol (ID=Idaho; SA=Salmon Arm; PG=Prince George; PR=Prince
Rupert; FN=Fort Nelson).

84

Table A-2: Height Growth Results from Univariate Repeated Measures Analysis in
GLM, significant (q=0.05) findings are in bold and shaded.
Between Subjects
Source
SS
Garden
1727095.246
Nursery
124373.788
Provenance
277779.598
69071.141
G*N
G*P
629775.338
324328.752
P*N
Error
4014256.001
Within Subjects
SS
DF
Source
1.52746
Time
6
723529.448
T*G
6
T*N
11597.855
12
T*p
99736.544
60
24872.558
12
T*N*G
r ,
J *p*Q
223020.297
60
179287.904
X*P*N
120
Error
2738663.429
2424
Grenhouse-Geisser Epsilon:
Huynh-Feldt Epsilon:

DF
1
2
10
2
10
20
404

MS
1727095.246
62186.894
27777.960
34535.570
62977.534
16216.438
9936.277

MS
2545760.986
120588.241
966.488
1662.276
2072.713
3717.005
1494.066
1129.812
0.2697
0.3008

85

F
2253.261
106.733
0.855
1.471
1.835
3.290
1.322

F
173.817
6.259
2.796
3.476
6.338
1.632

P
<0.001
<0.001
0.593
0.011
0.038
<0.001
0.012

P
<0.001
0.002
0.002
0.032
<0.001
0.042

G-G
0.000
0.000
0.471
0.103
0.135
0.000
0.111

H-F
0.000
0.000
0.481
0.093
0.127
0.000
0.100

400
350

*H^<

<o 300
o
o
CM

£

250

1

'•—

*

•

%

r 200
5>
I 150
2
£ 100
50
0

ID

SA
low

SA
mid

SA
high

PG
low

PG
mid

PG
high

PR
low

PR
mid

PR
high

FN

Provenance

Figure A-2: Mean Total Height (cm) by Provenance in 2006 with Nursery and Garden
pooled (ID=Idaho; SA=Salmon Arm; PG=Prince George; PR=Prince Rupert; FN=Fort
Nelson).

86

APPENDIX B: Results from 2001 and 2002 Bud Flush Study

Table B-l: ANOVA Results for Days to 20% and 80% Bud Flush in 2001, significant
(q=0.05) results are in bold and shaded.
Sum-of-Squares
F-Ratio
df
P
R2: 0.816
Days to 20% Bud Flush
N:930
45318.682
2
1123.812
Garden
< 0.001
Nursery
603.381
2
14.963
< 0.001
10748.722
4
Region
133.273
< 0.001
Garden * Nursery
51.184
4
0.635
0.638
8
30.857
Garden * Region
4977.343
< 0.001
Nursery * Region
170.236
8
1.055
0.392
18166.804
Error
901
Days to 80% Bud Flush
Garden
Nursery
Region
Garden * Nursery
Garden * Region
Nursery * Region
Error

R2: 0.831
N:841
32752.036
162.590
9010.301
48.350
3807.834
243.542
17525.419

2
2
4
4
8
8
812

87

758.745
3.767
104.368
0.560
22.053
1.410

< 0.001
0.024
< 0.001
0.692
< 0.001
0.188

Table B-2: ANOVA Results for Days to 20% and 80% Bud Flush in 2002, significant
(a=0.05) results are in bold and shaded.
F-Ratio
Sum-of-Squares
df
P
Rz: 0.866
Days to 20% Bud Flush
N:681
1
Garden
42578.099
1886.189
< 0.001
Nursery
156.386
2
3.464
0.032
13522.821
4
Region
149.764
< 0.001
Garden * Nursery
35.747
2
0.792
0.453
Garden * Region
4
3054.248
33.825
< 0.001
Nursery * Region
276.335
8
1.530
0.143
Error
14876.007
659
Rz: 0.877
Days to 80% Bud Flush
N:582
1
Garden
41076.766
1799.554
< 0.001
130.654
2
Nursery
2.862
0.058
Region
12488.052
4
136.774
< 0.001
Garden * Nursery
3.361
2
0.074
0.929
4
Garden * Region
3647.539
39.949
< 0.001
Nursery * Region
332.922
8
1.823
0.070
12782.606
560
Error

Table B-3: Mean Julian Days to 20% and 80% Bud Flush by Garden, Nursery and
Region in 2001. Means followed by the same letter within a factor are not significantly
sed on Schefl fe's pairwise comparison tests.
Mean JD
Mean JD
Standard
to 20%
Error of
to 80%
Bud Flush the Mean Bud Flush
Garden
107.432 a
0.322
111.058 a
ID
108.890 b
0.323
113.634 b
SK
125.244 c
0.280
131.557 c
RR
Nursery
112.520 a
0.308
118.021a
Ul
114.326 b
0.321
118.960 ab
NW
114.720 b
0.293
119.269 b
LN
Region
108.239 a
0.501
112.579 a
FN
111.812b
0.303
117.800 b
SA
114.805 c
0.284
119.479 c
PG
114.991c
0.539
119.864 c
ID
119.430 d
0.285
124.027 d
PR

88

Standard
Error of
the Mean
0.336
0.335
0.438
0.344
0.373
0.333
0.523
0.352
0.299
0.759
0.304

Table B-4: Mean Julian Day to 20% and 80% Bud Flush by Garden, Nursery and
Region in 2002. Means followed by the same letter within a factor are not significantly
sed on Schef 'e's pairwise comparison tests.
Mean JD Standard
Mean JD Standard
to 20%
Error of
to 80%
Error of
Bud Flush the Mean Bud Flush the Mean
Garden
ID
SK
RR

-

-

-

-

114.252 a
134.707 b

0.375
0.288

118.456 a
139.926 b

0.377
0.339

124.008 a
124.130 a
125.300 a

0.421
0.408
0.369

128.367 a
129.516 a
129.690 a

0.450
0.431
0.396

115.265 a
123.285 b
123.852 b
127.558 c
132.437 d

0.732
0.690
0.374
0.350
0.350

119.068 a
129.810 be
128.535 b
131.720 c
136.821 d

0.737
0.783
0.427
0.370
0.367

PR

FN

Nursery
Ul
NW
LN
Region
FN
ID
SA
PG
PR

20.0
18.0
16.0
14.0
12.0
10.0
8.0

r

i

6.0
4.0
2.0
0.0 -P™
ID

SA

PG
Region
•ID-

-SK

RR

Figure B-l: Period of Bud Flush (days) + SEM by Garden and Region in 2001 with
Nursery pooled. Where error bars are not visible, they are within the plotted symbol
(Region: ID=Idaho; SA=Salmon Arm; PG=Prince George; PR=Prince Rupert;
FN=Fort Nelson; Garden: ID=Sandpoint, Idaho; SK=Skimikin; RR=Red Rock).

89

14.0

-

-

-

:

lush

12.0
10.0
8.0

O
•a

6.0 <

Q.

4.0

o

it

2.0
nn
ID

SA

PR

PG

FN

Region
-SK

•RR

Figure B-2: Period of Bud Flush (days) ± SEM by Garden and Region in 2002 with
Nursery pooled. Where error bars are not visible, they are within the plotted symbol
(Region: ID=Idaho; SA=Salmon Arm; PG=Prince George; PR=Prince Rupert;
FN=Fort Nelson; Garden: SK=Skimikin; RR=Red Rock).

90