An Investigation Of Natural Hybridization Between
Jack Pine (Pinus banksiana) And Lodgepole Pine
(Finns contorta var. latifolia)
In Northern British Columhia

Lisa Wood
B.Sc., NRM (Forestry), University O f Northern British Columbia, 2003

Thesis Submitted In Partial Fulfillment Of
The Requirements For The Degree Of
Master O f Science
in
Natural Resources And Environmental Studies
(Forestry)

The University Of Northern British Columbia
April 2006

© Lisa Wood (nee: VanArendonk), 2006

1^1

Library and
Archives Canada

Bibliothèque et
Archives Canada

Published Heritage
Branch

Direction du
Patrimoine de l'édition

395 Wellington Street
Ottawa ON K1A 0N4
Canada

395, rue Wellington
Ottawa ON K1A 0N4
Canada
Your file Votre référence
ISBN: 978-0-494-28357-8
Our file Notre référence
ISBN: 978-0-494-28357-8

NOTICE:
The author has granted a non­
exclusive 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 non­
commercial purposes, in microform,
paper, electronic and/or any other
formats.

AVIS:
L'auteur a accordé une licence non exclusive
permettant à la Bibliothèque et Archives
Canada de reproduire, publier, archiver,
sauvegarder, conserver, transmettre au public
par télécommunication ou par l'Internet, prêter,
distribuer et vendre des thèses partout dans
le monde, à des fins commerciales ou autres,
sur support microforme, papier, électronique
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 propriété du droit d'auteur
et des droits moraux qui protège cette thèse.
Ni la thèse ni des extraits substantiels de
celle-ci ne doivent être imprimés ou autrement
reproduits sans son autorisation.

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

Conformément à la loi canadienne
sur la protection de la vie privée,
quelques formulaires secondaires
ont été enlevés de cette thèse.

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
This study was conducted in order to find evidence of hybridization between jack pine
{Pinus banksiana) and lodgepole pine {Pinus contorta var. latifolia) in Northeast British
Columbia (EC) through genetic identification of paternity and maternity o f each sample,
while relating differences in morphology, wood and fibre traits, and chemical makeup to
genetic identity. Adding to this, an attempt was made to determine if wood and fibre traits
can be used as distinguishing features between jack pine, lodgepole pine, and their hybrids,
and to determine if a “chemical extractive footprint” can be used to differentiate between
species groups.
Thirty samples of pure lodgepole pine, 30 of pure jack pine, and 30 potential hybrid
samples were collected from the Prince George area of EC, the Smoky Lake area of Alberta,
and the Fort Nelson region of EC, respectively. Data was analyzed by comparing
characteristics in order to establish trends and interactions between characteristics, site
conditions, and differences among samples.
Needles from foliar regions and inner cambial layers were harvested for DNA
analysis. Chloroplast DNA and mitochondrial DNA were used to determine hybridization
between lodgepole pine and jack pine by restriction fragment length polymorphism (RFLP).
Gross morphological characteristics of each tree were measured from collected cone and
needle samples, including, cone length and orientation, and needle length and position. Tree
height and diameter at breast height (DEH) were recorded at time of sampling. Two 10 mm
cores (bark to bark) were taken from eaeh tree and analyzed for fibre length and coarseness
through Fibre Quality Analysis, and for microfibril angle, basic density, earlywood :
latewood ratios, modulus of elasticity (MCE), and cell wall configuration using SilviScan
technology. Chemical extractive makeup of each sample was determined using gas
chromatography- mass spectroscopy.
Eased on the genetic evidence, 16 out of 24 samples from the introgression zone
clearly show hybridization, with lodgepole pine maternity and jack pine paternity, occurring
in the Fort Nelson region. Morphological characteristics that have been identified to
distinguish among species groups include: needle V width/length, cone angle of attachment,
cone curvature, cone length, and DEH/age. Cell and fibre traits that best differentiated
between pure jack pine, lodgepole pine, and hybrids were moisture content (MC), MOE, cell
wall thickness, and fibre coarseness, while only MC and MOE displayed hybrid intermediacy
clearly.
Preliminary chemical analysis of the samples indicates that quantitative variability does
exist between jack pine, lodgepole pine, and hybrid sample wood extractives; however, a
clear “chemical extractive footprint” can not be determined. Further manipulation of
chromatographs and more extensive investigation of mass spectra are required.
Revealing how natural hybrids are different from pure species, and what this means to
the scientific and industrial communities as well as forest managers is of primary importance.
Supportive evidence of hybridization and introgression in the Fort Nelson region of British
Columbia will provide information for proper management of forests in this region, with
regard to silvicultural practices and tree breeding, and aid in optimization of processing and
manufacturing to improve wood and product quality.

11

Table of Contents

A bstract................................................................................................................................................. il
List of Tables..................................................................................................................................... vil
List of Figures..................................................................................................................................... ix
Chapter 1: Introduction....................................................................................................................... 2
1.1 Introgression and Hybridization............................................................................................... 2
1.1.1 Characterization of Hybrid Zones................................................................................. 2
1.1.2 Evolutionary Significance.............................................................................................. 6
1.1.3 Hybridization vs. Divergence.......................................................................................8
1.1.4 References to Lodgepole x Jack Pine Hybrids.......................................................... 9
1.1.5 Evidence of Jack Pine and Lodgepole Pine H ybridization.................................... 11
1.1.6 Experimentation with Cross-Pollination between Species...................................... 12
1.1.7 Other Hybrids Occurring in Nature........................................................................... 13
1.2 Thesis Objectives.......................................................................................................................13
1.2.1 Introduction to Species..................................................................................................15
1.3 Thesis M ethodology................................................................................................................. 17
1.4 Site Descriptions....................................................................................................................... 20
1.4.1 Locations and M ap s.....................................................................................................20
1.4.2 EEC Information.......................................................................................................... 22
1.4.3 Description of Soil and Vegetation............................................................................23
1.5 References.................................................................................................................................. 25
Chapter 2: Genetic Analysis of Jack Pine, Lodgepole Pine and their H ybrids....................... 28
2.1 Chapter Objectives.................................................................................................................... 28
2.2 Support from Existing Literature............................................................................................ 28
2.2.1 Description of CpDNA and MtDNA M arkers......................................................... 28
2.2.2 Critical Assessment of Various M arkers................................................................... 29
2.3 Experimental Design and Hypotheses................................................................................... 30

111

2.4 M ethodology............................................................................................................................... 32
2.4.1 Protocols for DNA Extraction......................................................................................32
2.4.2 DNA Quality and Q uantity......................................................................................... 33
2.4.3 PCR Analysis............................................................................................................33
2.4.4 Restriction D igest......................................................................................................... 37
2.4.5 Imaging...........................................................................................................................38
2.5 R esults........................................................................................................................................38
2.5.1 DNA Extraction............................................................................................................ 38
2.5.2 PCR Products................................................................................................................39
2.5.3 Restriction D igests....................................................................................................... 41
2.6 Discussion and Conclusions....................................................................................................49
2.7 References.................................................................................................................................. 52
Chapter 3: Morphological Analysis of Jack Pine and Lodgepole P in e .....................................54
3.1 Chapter Objectives..................................................................................................................... 54
3.2 Species Characteristics According to Existing Literature....................................................55
3.3 Experimental Design and Hypotheses................................................................................... 57
3.4 Methodology..............................................................................................................................58
3.4.1 N eedles...........................................................................................................................58
3.4.2 Cones..............................................................................................................................59
3.4.3 Height, DBH, and A ge................................................................................................ 59
3.5 R esults.........................................................................................................................................60
3.5.1 N eedles...........................................................................................................................60
3.5.2 Cones..............................................................................................................................61
3.5.3 Height, DBH, and A g e................................................................................................ 65
3.5.4 Statistical A nalysis....................................................................................................... 68
3.6 Discussion and Conclusions................................................................................................... 71
3.6.1 Needle Characteristics..................................................................................................71
3.6.2 Cone Characteristics.....................................................................................................72
3.6.3 Site Considerations........................................................................................................ 74
3.6.4 Growth Rate Characteristics - Height/Age and DBH/Age......................................75
3.6.5 Cluster A nalysis........................................................................................................... 77

IV

3.6.6 Summary of Gross Morphological Characteristics................................................ 78
3.7 References.................................................................................................................................. 80
Chapter 4: Fibre and Cell Analysis of Jack Pine, Lodgepole Pine, andtheir H ybrids.............. 81
4.1 Chapter Objectives..................................................................................................................... 81
4.2 Wood and Fibre Characteristics................................................................................................ 82
4.3 Experimental Design and Hypotheses................................................................................... 87
4.4 Methodology..............................................................................................................................87
4.4.1 Fibre Quality A nalysis.................................................................................................87
4.4.2 SilviScan.................................................................................................................... 89
4.4.3 Electron M icroscopy....................................................................................................89
4.5 R esults........................................................................................................................................90
4.5.1 Solid Wood Analysis....................................................................................................90
4.5.2 Fibre A nalysis:............................................................................................................109
4.6 Discussion and Conclusions..................................................................................................124
4.6.1 Discussion of Solid Wood Properties.......................................................................124
4.6.2 Discussion of Fibre Properties..................................................................................133
4.6.3 Conclusions................................................................................................................. 137
4.7 References................................................................................................................................ 138
Chapter 5: Chemical analysis of Jack Pine, Lodgepole Pine and their Hybrids......................140
5.1 Chapter Obj ectives.................................................................................................................. 140
5.2 Wood Chemical Extractives from Literature.......................................................................140
5.3 Experimental Design and Hypotheses................................................................................. 142
5.4 Methodology............................................................................................................................ 143
5.4.1 Chemical Extraction Using GC-MS.........................................................................143
5.5.1 PLS-DA Output...........................................................................................................147
5.5.2 Chromatography..........................................................................................................148
5.5.3 Mass Spectral A nalysis............................................................................................. 152
5.5.4 Statistical Analysis of Chromatographs.................................................................. 158

5.6 Discussion and Conclusions................................................................................................... 161
5.7 References................................................................................................................................. 168
Chapter 6: Discussion of and Applications for Hybridization of Lodgepole and Jack P ine. 169
6.1 Hybridization............................................................................................................................ 169
6.1.1 Predictive Abilities of Characteristics......................................................................171
6.1.2 Maternity and Paternity Issues: Effects on Productivity
...............................173
6.2 Applications.............................................................................................................................. 176
6.2.1 Hybridization Effects on Management for Wood Quality.................................... 176
6.2.2 Hybridization effects on Manufacturing and Prrocessing..................................... 178
6.3 References................................................................................................................................. 180
A ppendices:......................................................................................................................................182
Appendix 1 :......................................................................................................................................182
Site m aps................................................................................................................................ 182
References:............................................................................................................................. 187
Appendix 2: Haplotype Tables...................................................................................................... 188
Appendix 3: Statistics Tables....................................................................................................... 192

VI

List of Tables
Table 1:

Suggested Classification for Hybrid Zones.................................................................4

Table 2:

Sample numbers as collected from each sampling area.......................................... 38

Table 3;

Paternity haplotypes for each species by collection site.........................................48

Table 4:

Maternity haplotypes for each species by collection site........................................48

Table 5:

High and low range needle measurements displaying
the variation within and among species.................................................................... 60

Table 6:

Average cone angle for each species/sample group according
to both methods of categorizing sam p les..................................................................61

Table 7:

Cone length averages and range of variation.......................................................... 63

Table 8:

Average height, DBH, and age values, and range in height,
DBH and a g e ................................................................................................................ 65

Table 9:

Pairwise Comparisons for m orphology..................................................................... 68

Table lOa-c: Cluster Analysis for Morphology........................................................................... 70
Table 11 :

Average densities and moisture contents..................................................................90

Table 12:

Pairwise Comparisons of Calculated density and moisture content..................... 95

Table 13a-c: Cluster Analysis for moisture contentand density................................................ 96
Table 14:

Pairwise Comparisons for wood characteristics.................................................... 107

Table 15a-c: Cluster Analysis for wood characteristics............................................................ 108
Table 16:

Pairwise Comparisons of fibre length and coarseness for wood
age 40-80.................................................................................................................114

Table 17:

Pairwise Comparisons for fibre length and coarseness of wood
age 20-40.................................................................................................................115

Table 18a-c: Cluster Analysis for fibre length and coarseness at 40-60 and
60-80 y r s ................................................................................................................117
Table 19:

Pairwise Comparisons of MFA and coarseness..................................................... 122

Vll

Table 20 a-c: Cluster Analysis for MFA and coarseness.......................................................... 123
Table 21:Average Extractive % per sample for each sample site............................................ 146
Table 22a: Library search results of potential chemical compound matches
for the mass spectra in Figure 87a..................................................................... 152
Table 22b: Library search results of potential chemical compound matches
for the mass spectra in Figure 87b...................................................................... 153
Table 22c: Library search results of potential chemical compound matches
for the mass spectra in Figure 87c...................................................................... 154
Table 22d: Library search results of potential chemical compound matches
for the mass spectra in Figure 87d...................................................................... 155
Table 22e: Library search results of potential chemical compound matches
for the mass spectra in Figure 8 7 e ..................................................................... 155
Table 22f: Library search results of potential chemical compound matches
for the mass spectra in Figure 8 7 f ...................................................................... 156
Table 23:

Pairwise Comparison of Chromatographic peak regions with
most variability between species g ro u p s.......................................................... 158

vm

List of Figures
Figure 1;

Schematic diagrams illustrating allopatric and parapatric hybrid zones,
and three types of sympatric hybridization...................................................... 3

Figure 2:

Ranges of jack pine and lodgepole p in e .......................................................... 16

Figure 3:

Sample area locations; 3 sites are located within each a re a ......................... 22

Figure 4:

Extraction of DNA from sam ples................................................................... 39

Figures 5-6:

CpDNA PCR P roduct.......................................................................................40

Figure 7:

MtDNA N adi PCR product............................................................................ 41

Figure 8:

CpDNA PCR product digested with Snabl enzyme...................................... 41

Figures 9-10:

CpDNA PCR product digested with H hal enzym e...................................... 42

Figure 11:

Nadi PCR product trial digestion using R sal, M bol, and H h a l ................ 43

Figure 12:

N adi PCR product digestion using R s a l ...................................................... 44

Figure 13:

N adi PCR product digestion using M b o l......................................................45

Figures 14-15:

N adi PCR product digestion using H h a l...................................................... 46

Figure 16:

Needle Forms for jack pine and lodgepole pine............................................ 56

Figure 17:

Cone Forms for jack pine and lodgepole pine................................................57

Figure 18:

Needle measurements........................................................................................58

Figure 19:

Cone Measurements.......................................................................................... 59

Figure 20:

Average ratio of needle V width over needle length for each species/
sample group by area......................................................................................... 60

Figure 21a:

Mean cone angle for all species groups according to sampling area.......... 61

Figure 21b:

The range in angles that the cones form with the branch for
each species/sample group according to sample area................................... 62

Figure 22:

Qualitative classification of cones based on observed curvature................ 63

IX

Figure 23:

Cone length as a funetion of site index for sample groups according to
Sampling a re a .................................................................................................... 64

Figure 24:

Cone length as a function of age class............................................................... 64

Figure 25:

Growth rates for all sample groups according to area.....................................66

Figure 26:

Growth rate (DBH/age) as a function of site index......................................... 67

Figure 27a:

Jack pine mature wood, earlywood density profile moving from inner
wood to bark........................................................................................................91

Figure 27b:

Lodgepole pine mature wood, earlywood density profile moving from
inner wood to bark.............................................................................................. 91

Figure 27c:

Hybrid pine mature wood, earlywood density profile moving from inner
wood to bark...................................................................................................... 92

Figure 28a:

Calculated density and DBH growth rate based on ring count age, measured
DBH, measured volume, and measured weight of
cores.....................................................................................................................93

Figure 28b:

Density of samples measured by SilviScan that were over 60 years of
age plotted against DBH growth rate.............................................................. 94

Figure 29:

Ratios of earlywood to latewood in each sample........................................... 97

Figure 30a:

Earlywood - latewood transition as shown in potential hybrid pine
Sample #64..........................................................................................................98

Figure 30b:

Jack pine sample cross section with longitudinal view of trachieds.............98

Figure 31a:

The relationship between earlywood/latewood contributions and cell
wall thickness..................................................................................................... 99

Figure 3 lb:

Average radial and tangential diameter measurements for samples
measured by SilviScan......................................................................................100

Figure 32:

Wall thickness profile for mature wood moving from inner wood to
bark..................................................................................................................... 101

Figure 33:

Cell wall thickness as a function of density................................................... 102

Figure 34:

Wall thickness as a function of modulus of elasticity...................................103

Figure 35:

Modulus of Elasticity (MOE) profile from pith to bark.............................. 104

Figure 36:

Density as a function of M OE........................................................................ 105

Figure 37:

MOE and growth rate (DBH/age).................................................................. 106

Figure 38:

Length weighted fibre length from pith to bark............................................109

Figure 39:

Average number of fibres per fibre length interval for wood rings
aged 40-60........................................................................................................ 110

Figure 40:

Fibre coarseness as a function of fibre length...............................................I l l

Figure 41 :

Coarseness profile from FQA measurements moving from pith to
bark.................................................................................................................... 112

Figure 42:

Coarseness as a function of density................................................................113

Figure 43:

Microfibril angle profile moving from pith to bark..................................... 118

Figure 44:

Average mierofibril angle for select samples representing sampling
areas...................................................................................................................119

Figure 45:

Mierofibril angle as a function of DBH growth rate....................................120

Figure 46 :

Mierofibril angle and M OE.............................................................................121

Figure 47:

PLS-DA model output of the regression coefficient for lodgepole
pine.................................................................................................................... 147

Figure 48:

Aligned Chromatogram for range 825-1380.................................................148

Figure 49a:

Chromatogram of jack pine sample 2 ............................................................ 149

Figure 49b:

Chromatogram of lodgepole pine sample 4 3 ................................................ 149

Figure 49c:

Chromatogram of hybrid pine sample 68...................................................... 150

Figure 49d:

Chromatogram of hybrid pine sample 84......................................................151

Figure 50a:

Mass spectra for sample 08, peak region 11 at a retention time of
18.395 minutes.................................................................................................152

Figure 50b:

Mass spectra for sample 37, peak region 11 at a retention time of
18.275 minutes.................................................................................................153

XI

Figure 50c:

Mass spectra for sample 49, peak region 11 at a retention time of
18.361 minutes..................................................................................................154

Figure 50d:

Mass spectra for sample 56, peak region 9 at a retention time of
17.679 minutes..................................................................................................155

Figure 50e:

Mass spectra for sample 43, peak region 5 at a retention time of
16.102 minutes..................................................................................................155

Figure 50f:

Mass spectra for sample 43, peak region 5 at a retention time of
15.912 minutes..................................................................................................156

Figure 51 :

Sample groups, based on the discriminant variable value and species
group................................................................................................................... 157

Figure 52a:

An example of a compound similar to the stilbene identified as a
candidate match...................................................................................................... 164

Figure 52b:

An example of a compound similar to the phenanthrene identified as
a candidate m atch............................................................................................. 165

Figure 52c:

An example of a compound similar to the naphthalene identified as
a candidate match............................................................................................. 165

Figure 52d:

An example of a compound similar to the phenanthrene identified as
a candidate m atch............................................................................................. 166

X ll

Acknowledgements :
Many people contributed their time and effort to initiating, carrying out, and
concluding this study. Without these individual’s contributions and guidance this project
could not have been completed.
Firstly I would like to thank my supervisor. Dr. Ian D. Hartley who provided
invaluable assistance, and committed resources, time, and energy to my educational and
career goals. He not only provided a direction for my growth but also gave me the freedom
to contribute my ideas and methods to the project.
I would also like to thank my thesis committee. Dr. Paul Watson and Dr. Brent Murray,
who helped me shape my hypotheses into a manageable and realistic project. Their guidance
and resources allowed for the development of applicable project results.
Through these three individuals, I was able to solicit the assistance of others who also
contributed to this study. I am grateful for the help from Dr. Craig Newton, who provided
me with assistance in developing my protocol, and initiating my laboratory analysis for the
genetics component of this study. I am also thankful to Dr. Paul Bicho, and Maxwell
McRae, who assisted with the wood chemical extraction, and GC-MS analysis. The project
could not have been carried out without these individuals.
Thanks to Lars Wallbacks and James Drummon, who also contributed to the project by
assistance with PCA of the GC-MS output, and providing exquisite electron microscope
photographs, respectively.
Finally I would like to thank those people who provided site information and who
trekked through the forest and snow with me in order to colleet samples in temperatures as
low as -30°C. These people include Mark Petrovcic, Adam Lloyd, Bob Krahn, and
especially Tennessee Trent who spent hours coring frozen trees for me.
Thank you to everyone, this projeet has been a great experience for me, and I am
indebted to all of you.

Dedication:
I would like to dedicate this very long piece o f work, and with it more than 2 years o f time,
energy, and life experience, to my husband, Keith, my mom, June, and my dad, Joe.
Keith, thank you fo r encouraging and pushing me to strive fo r my goals and fo r providing me
with an example o f work ethic that is untouched by anyone else I know.
Mom and dad, thank you fo r always supporting me and allowing me to believe that I could
meet any challenge that comes my way.
I love you all very much!

X lll

Chapter 1: Introduction
1.1 Introgression and Hybridization
Introgressive hybridization can be defined as

. .the infiltration of genes of one species

into another through repeated backcrossing of hybrids to one or both parental species.”
(Wheeler and Curies, 1987). Introgression can be seen as a step in evolution; increasing
genetic information and providing new gene interactions (Wheeler and Curies, 1987).
Another definition describes hybridization is the “ .. .interbreeding of two populations, or
groups of populations, which are distinguishable on the basis of one or more
characters...’’(Woodruff, 1973). Woodruff (1973) furthers this definition by saying that
hybridization produces individual hybrids that are recognized by one or more taxonomic
features. Even with these definitions in mind, classifying hybrids using the taxonomy that
exists today can be challenging. Hybrid pines that are under investigation in this study will
fall into the same taxonomic “grey area”. If a lodgepole pine {Pinus contorta var. latifolia) x
jack pine {Pinus banksiana) hybrid is found to exist not only morphologically, but
genetically and based on wood, fibre, and chemistry traits, then a system of classification
must be established in order to properly identify these hybrid individuals.

1.1.1 Characterization of Hybrid Zones
The definition of species, “ . ..as a set of populations delimited by genetic barriers to
gene exchange..

or populations that naturally interbreed and are reproductively isolated

from other populations, is contrasted by the existence of hybridization (Barton and Hewitt,

1985).
In order to interpret hybrid zones in terms of common taxonomy, the zones can be
characterized as areas where genetic information is exchanged between two different species

populations to create a range or gradient of genetic forms between the two species. This
genetic gradient is referred to as a “d in e ” (Barton and Hewitt, 1985) in order to distinguish
gradient genotypes and phenotypes from hybrids, which are often seen as only one
recombined phenotype or genotype of the parental species. Clines are not limited to hybrids
of a particular form or morphology; these ‘gradient hybrid zones’ move with reproduction
and dispersal area of seed (Barton and Hewitt, 1985). Clines can be further described as
either allopatric zones, where only the hybrid species is present in a narrow area between
both parental species that do not meet, or as sympatric zones, where hybrids and parental
species are located in the same area heterogeneously (Short, 1969). Sympatric zones can be
further classified as parapatric, if the d in e is quite narrow; see Figure 1 and Table 1
(Woodruff, 1973).

Allopatric Hybridization

Sympatric Hybridization

1. At range periphery
Parapatric Hybridization

2. Widespread

3. Localized
Figure 1 : Schematic diagrams illustrating allopatric and parapatric hybrid zones, and three types o f sympatric
hybridization. Hybrids are indicated by shading, geographic ranges o f parental species are outlined (from
Woodruff, 1973).

Table 1: Suggested Classification for Hybrid Zones by W oodruff (1973).
Distribution o f
Parental Types

Distribution
o f Hybrids

Suggested
Terminology

allopatric

between ranges o f
parental types

intergradation or
allopatric hybridization

parapatric

adjacent to contact

parapatric hybridization

sympatric

associated with
peripheral
populations

peripheral sympatric
hybridization

sympatric

localized

localized
sympatric
hybridization

sympatric

widespread

widespread
sympatric
hybridization

Clines are established when two distinct populations come into contact. Advantageous
alleles will advance. Differences in fitness causes certain individuals to perpetuate better
than others; two populations head towards an equilibrium, or hybrid population, that may be
better evolved than the two original populations for the surrounding environment (Barton and
Hewitt, 1985). It is likely that a gradient of hybridization, or eline, exists where any
introgression zone is verified. Individual hybrids found within a eline may contain genetic
information from eaeh parent species in any percentage, and is not necessarily a generation 1
hybrid; lodgepole pine from one side, jack pine from the other. Therefore genetically
identifying these hybrids is difficult and may not always be possible using the RFLP
(restriction fragment length polymorphism) method that is used in this study.
New hybrids are formed within a eline from the flow of genetic information from
individual to individual. Many factors contribute to the movement of a eline; gene flow does
not always favor the fittest. The most influential factors are stand density and seed dispersal

rates although dominant alleles will spread more frequently than recessive (Barton and
Hewitt, 1985).
Since jack pine stands are segregated into random patches across the landscape in the
western reach of the jack pine range, hybrids or d ines may be more infrequent. Variation
among jack pine may be limited due to patchiness, and hybrid types may be less diverse than
in lodgepole pine stands in Northern BC. Pine stands in BC are large and vast allowing for
increased hybrid gradients and ease in d in e movement over the landscape.
As genetic information flows between populations creating variety in pine species, the
movement of alleles from generation to generation can meet certain physical or genetic
barriers. Physical barriers include environmental changes that affect seed dispersal or density
of a stand. Movement from an even-aged aged stand to an uneven-aged stand can also be a
physical barrier to d in e movement because not all trees would be at an age where they are
able to reproduce. By the same token, d in e movement from a single species stand to a
multiple species stand could also act as a physical barrier; not all trees would be compatible
for reproduction. Other physical barriers that affect hybrid zone movement or formation
include climatic gradients, differences in vegetation, and differences in soil type. Human
development and disturbance can also be seen as a physical barrier to gene flow, producing
variation in hybrids to best suit the surrounding environment (Barton and Hewitt, 1985).
Barton and Hewitt go on to suggest, however, that environmental gradients will not maintain
a hybrid zone, or keep it in the same position. Genetic barriers include the fitness of
individual trees; reproductive capability can be lowered by pathogen or pest infestation
thereby impeding gene flow.

Since barriers to gene flow are random and gene flow is asymmetric, d in es do not
necessarily move in an even pattern (Barton and Hewitt, 1985). This accounts for the
variability in hybrid locations. Jack pine x lodgepole pine hybrids, although approximately
located in the introgression area where population ranges overlap, may be found outside of
this zone, for example in an area northeast of Prinee George, BC due to d in e movement over
time.
Due to the high amount of genetic recombination that is on going within a hybrid zone,
a high percentage of rare alleles can be found. This characteristic is compounded by high
mutation rates and somewhat relaxed selection in some cases (Barton and Hewitt, 1985).
However, this would only apply to hybrid zones that have not reached equilibrium with
surrounding species. Well-equilibrated populations would be more homogenous or allopatric
in nature.

1.1.2 Evolutionary Significance
Species develop as an adaptive response to their environments as well as in
concordance with the gene flow available to them. These responses allow some species to
hybridize and evolve populations with greater fitness. These populations form hybrid zones
that possess a balance between dispersal and selection, and allow a wider range of
environments to be tolerated (Barton and Hewitt, 1985). Evolution of species involves the
recombination of genetic information through reproduction. Various factors are known to
m odify reproductive processes, individuals within a population, or populations them selves.

Modifiers arise due to mutations within a hybrid zone, and decreases in gene flow from
outside of the zone. These modifications to a population are then able to gain proportional
advantages over the gene pool and perpetuate within the area (Barton and Hewitt, 1985).

From an evolutionary standpoint, if a certain zone is modified and barriers exist so that the
gene input from outside the zone is restricted and gene flow is somewhat limited to within
the hybrid population, then the hybrids in question may potentially be referred to as a new
species. Given that species are reproductively isolated from others and display unique and
individual characteristics. We can say that it may be possible to distinguish this type of
population as a new species or subspecies based on the possibility of hybrid isolation. The
classification of the hybrid zone could also aid in the argument of the development of a new
species (Woodruff, 1973). An allopatric population that is segregated from parental types
and reproductively self-sustaining would be more likely to fit the definition of a separate
species than a sympatric population.
Specifically identifying evolutionary steps within the jack pine and lodgepole pine
populations is somewhat challenging. Conifer populations have a large history and tracking
this information in order to establish exact relationships and interactions between these two
closely related pines requires methods to quantify genetic distances and data from the fossil
record. Nei (1973) established a method to measure genetic differences in any species and
establish the “genetic distance” between two different populations. Genetic distance,
expressed as “D”, can be found by the following equation “where Jxy =

Jx =

X y \ and x, and y; denote the frequency of the ith allele in the population x and y
respectively” (Danick and Yeh, 1983):

D = -l0 g e [Jx y /(Jx Jy )''"]

[1]

Jy =

Danick and Yeh (1983) applied this directly to lodgepole and jack pine and found genetic
distances using the Nei equation for Alberta populations using allozyme markers. They
found that the genetic distance between the two speeies was 20 times greater than and
interspecific distance. Therefore, variation within species is much less genetieally
identifiable than variation between jack pine and lodgepole pine. The faet that the genetie
distanee between jack pine and lodgepole pine is greater than the genetic distance between
individuals within species, these genetic distance values being comparable to those observed
in other organisms, supports that jack pine and lodgepole pine are two very distinct species.
These are species that possess extensive variation in range, and that ean not be misinterpreted
as different variations of the same species from any point in history.

1.1.3 Hybridization vs. Divergence
An alternate hypothesis exists to counter the question of jack pine x lodgepole pine
hybridization, namely divergence. The hypothesis states that the jack pine and lodgepole
pine species are not presently hybridizing, or exchanging genetic information over time, but
rather these species are diverging from one original species due to environmental adaptation
and have been since the last ice age occurred approximately 18,000 years ago (Critchfield,
1985; Barton and Hewitt, 1985). Pollen from the fossil record supports the fact that a
variation in species existed, at the time of the last ice age, which is not distinguishable as
lodgepole pine or jack pine. This pollen can not tell us if this species is the primitive
ancestor of the two species that now exist or an evolved hybrid of lodgepole and jack pine
(Critchfield, 1985). To add to this, little historical evidence exists regarding hybrid zone
formation, stability, or movement, but most have been fairly static for at least the last century
(Barton and Hewitt, 1985). With this in mind, it is difficult to determine is lodgepole pine x

jack pine hybrids exist today due to historical divergence or hybridization, however, plenty of
evidence exists to confirm genetic recombination in natural environments today.

1.1.4 References to Lodgepole x Jack Pine Hybrids
According to Zavarin et.al. (1969), lodgepole pine x jack pine hybrids “ .. .are readily
recognized in the field and have been verified by comparison with artificial [crosses of these
species]...” (Wheeler and Guries, 1987). W heeler and Curies (1987) go on to say that there
is concordance between the morphological characteristics of lodgepole x jack pine hybrids
and electrophoretic evidence collected, leading to the assumption that gene exchange does
take place between these pine species.
Since morphology is variable within the same species, the reliability of purely
characteristic data is subjective. However, jack pine species shows no variation in
morphology west of Lake Nipigon, Ontario, until the potential introgression zone with
lodgepole pine (Critchfield, 1985). Danick and Yeh (1983) observed that the variability
within populations o f jack pine and lodgepole pine in Alberta is approximately normal

compared to populations of other conifers. Contrary to Barton and Hewitt (1985), Dancik
and Yeh (1983) believe that high genetic variability exists in conifer populations due to
“ .. .lack of effective barriers to gene flow ...”. Other contrihutors to genetic variability are
population size, breeding systems, and ecological amplitude. Distribution of jack pine x
lodgepole pine hybrids in the Alberta introgression zone is uneven and allele groups are
patchy. This genetic variability and population lay-out is most likely due to restricted seed
dispersal; seed falling close to the maternal parent produces small patches of trees with allelic
similarity. Therefore cross-pollinations in these types of populations can be said to occur

between

. .trees closely related by decent...” or trees that are evolutionarily similar (Dancik

and Yeh, 1983).
The suggestion of hybridization between these two species is a possibility using
historical evidence from pine pollen dating back to the Pleistocene (-18,000 years ago).
Pollen evidence suggests that lodgepole and jack pine may have been exchanging genes since
late-postglacial times.
Hybrids have been noted, described, and studied in Central and Northern Alberta as
well as in the Mackenzie District of the NWT (Critchfield, 1985). This study’s attempt to
describe and identify hybrids in northeastern BC will add further dimension to this known
introgression zone.
Extractive chemistry has been another area of study to support the existence of
lodgepole x jack pine hybrids. Monoterpene analysis has been suggested as a “ .. .sensitive
method for detecting introgression...” (Pollack and Dancik, 1985); extractive compounds are
studied in Chapter 5 in order to establish a chemical footprint that may identify between
lodgepole pine, jack pine and hybrids.
Genetically, hybrid characteristics can be identified through analysis of isozyme
markers, restriction fragment length polymorphisms (RFLPs), chloroplast DNA (Kormutak
et.al., 1993), mitochondrial DNA, allozyme variation, and random amplified polymorphic
DNA (RAPD) markers. Out of these methods, RFLPs of both mtDNA and cpDNA were
chosen for this study in order to give a well rounded description of the DNA inputs and due
to their ease of interpretation.

10

1.1.5 Evidence of Jack Pine and Lodgepole Pine Hybridization
There is extensive evidence regarding morphological characteristics which suggests
hybridization between lodgepole and jack pine. That is, there are trees that display a
combination of morphological characteristics of both species. In a study conducted by
Wheeler and Guries (1987), 4 putative hybrid populations of the 37 pine populations sampled
were identified as hybrids due to cone and branch characteristics. Locations of these
populations surround Blue River, White Court, Grande Prairie, and Wapiti, Alberta. The
hybrid morphological characteristics interpreted in an index were correlated to gene
frequency data to reveal that introgression in these areas was statistically significant (by
least-squares regression and pairwise comparison; all correlation coefficients were greater
than 0.98).

Jack pine populations of eastern Alberta were shown to have a significant level of
influence from lodgepole pine populations by evidence of genetie variation at enzyme loci
(Critchfield, 1985).
Other examples of pines found with morphological characteristics of lodgepole x jack
pine hybrids were between Yellowstone Lake, Wyoming, and Banff, Alberta; these trees
displayed cone angles significantly smaller and more variable than lodgepole pine, but not
completely curved towards the branch as in jack pine (Critchfield, 1985).
Evidence of introgression taken from chemical constituents of tree extractives is
supported by Pollack and Dancik (1985), who tested lodgepole and jaek pine for levels of Pphellandrene and a-pinene. According to the study, P - phellandrene occurs in much higher
concentrations in lodgepole pine then in jaek pine, and a-pinene is found in much higher
concentrations in jack pine than lodgepole pine. At sample sites Two Creeks, Rocky

11

Mountain House, and Twin Lakes, populations were found with moderate concentrations of
both P-phellandrene and a-pinene. These results are supportive of hybridization between
lodgepole pine and jack pine. It has been stated however that “.. .variations in the
composition of wood resin are inconsistent and difficult to interpret” (Critchfield, 1985).
Putative lodgepole x jack pine hybrids from the introgression zone in central Alberta
have been characterized by chloroplast DNA restriction fragments which differ from the
parental species, and may suggest lodgepole pine paternity (Kormutak, et.al., 1993).

1.1.6 Experimentation with Cross-Pollination between Species
According to Critchfield (1985), lodgepole pine and jack pine can be very easily
crossed. Crosses that are performed with lodgepole pine as the female parent result in
approximately 30% germinal seeds. Performing cross-pollination experiments in a
laboratory setting allows reproductive barriers that exist in nature to be overcome. In a study
in 1983, jack and lodgepole pine were successfully crossed in a nursery setting, to form a
hybrid species (Kormutak, et.al., 1993). Jack pine x lodgepole pine crosses preformed by
Dong et. al. (1992) indicate that crossing can be done successfully in either parental
direction; jack pine as mother or father. Twenty-one matings between jack pine and
lodgepole pine were conducted in the study, parents obtained from Alberta, British
Columbia, and the Yukon Territory. One Alberta population sampled for parental DNA of
jaek pine used in Dong et. al. study (1992), was the Smoky Lake population used for
obtaining jack pine samples in this study.
Natural hybridization is deterred between jack and lodgepole pine due to differences in
flowering time; jack pine flowers 2-3 weeks before lodgepole pine. Ecological preference in

12

germinating site also varies between the two pines, making cross-fertilization difficult.
Putative hybrids in Alberta have been observed to often produce aborted pollen grains or
pollen grains that are smaller than non-hybrids. Forty-two percent of putative hybrids have
aborted pollen as compared to only 1-2% in pure species, making hybrid proliferation
difficult (Critchfield, 1985).
Although it is helpful to look at laboratory crosses to determine some hybrid
characteristics, natural hybrid zones give a much better picture of the possibilities of genetic
recombination. Laboratories are limited to only a few generations, where natural hybrid
areas have been recombining genetic information for thousands of years (Barton and Hewitt,

1985).

1.1.7 Other Hybrids Occurring in Nature
There are many other examples of coniferous trees that hybridize naturally, so the
hypothesis that lodgepole and jack pine may hybridize in the study area surrounding Fort
Nelson, EC, is appropriate. Some good examples of other species that are suggested to
naturally hybridize are Sitka x interior spruce (OECD, 2002) and black x red spruce
(Johnsen, et.al., 1998).
Pinus sylvestris x Pinus conforta crosses occur in natural populations in Sweden, and
are studied due to their commercial importance to Swedish forestry (Szmidt, et. al., n.d) for
possible increases in population yield and tree vigor.

1.2 Thesis Objectives
This study was conducted in order to find evidence of hybridization between jack pine
{Pinus banksiana) and lodgepole pine {Pinus conforta var. latifolia) in northeast British

13

Columbia, specifically the Fort Nelson region. Morphological evidence supporting
hybridization in this area has been documented (Critchfield, 1985; Wheeler and Curies,
1987); however, substantial evidence of this interaction is yet to be shown. Using the
principle morphological characteristics of hybrids according to literature, potential hybrid
samples were collected for further investigation.
While attempting to characterize the data sets collected according to all of the objectives,
it is essential that the characteristcs identified are related to one another in order to illustrate a
picture of any hybrid samples that may exist within the data. Revealing how these trees are
different from pure species, and what this means to the scientific and industrial communities
as well as forest managers is o f primary importance. Supportive evidence o f hybridization

and introgression in the Fort Nelson region of British Columbia will provide information for
proper management of forests in this region, and specifications for processing and
manufacturing. Objectives of this study were chosen to support morphological evidence of
natural hybridization between jack pine and lodgepole pine.
The first objective of this study, and the focus of Chapter 2, was to genetically identify
the paternity and maternity of each sample based on restriction fragment length
polymorphisms (RFLPs). In this study, genetic evidence provides a basis for classifying
species groups for further wood and fibre analysis, and wood extractive analysis.
The purpose of Chapter 3 was to present differences in morphology between species for
proper identification of a species in the field, providing visual evidence of hybridization so
noted in the literature, while categorizing species groups for comparison with genetic results
from Chapter 2. Chapter 3 also quantifies variation in the morphology between species,
reveals the relationship of some characteristics to site conditions, and determines the ability

14

of each characteristic to distinguish between species groups. This quantification provides a
basis for comparison with other characteristics, namely wood and fibre traits (Chapter 4) and
extractive chemistry (Chapter 5).
The objective of Chapter 4 was to compare species based on wood and fibre traits and to
correlate these trait differences to genetic evidence and morphological differences between
species and hybrids. Chapter 4 assists in determining if wood and fibre traits can be used as
distinguishing features between jack pine, lodgepole pine, and their hybrids, while
identifying interactions and relationships between wood traits, fibre traits and site conditions.
The primary goal of Chapter 5 was to determine if a “chemical extractive footprint” can
be used to differentiate between jack pine, lodgepole pine and hybrid species. Characterizing
the species in this way allows for further support of hybrid existence, and support the
morphological evidence of hybridization.

1.2.1 Introduction to Species
Pure jack pine {Pinus banksiana) grows naturally from Great Bear Lake in Northwest
Territories diagonally across Canada to southern Ontario, into southern Quebec, New
Brunswick, and Nova Scotia (Farrar, 1995), while pure lodgepole pine {Finns contorta)
ranges over most of British Columbia and into Alberta. Lodgepole pine is found just over the
Yukon border south through BC and well into the United States, and from the east side of the
coastal mountains to western Alberta (Farrar, 1995), see Figure 2.
Both jack pine and lodgepole pine are hard pines that are self-pruning, meaning all
branches are located at the top of the tree (Forintek Canada Corp., 1994). Pines are a primary
successional species or pioneer species that grow best in deep, moist soils where drainage is
good, but are usually out-competed by more shade tolerant species. Therefore these pine

15

species are often found occupying areas that other species are unable to inhabit such as rocky
or very sandy soils (Isenberg, 1951). Lodgepole and jack pine usually inhabit open areas
receiving plenty of direct sunlight and are susceptible to fire as a stand-replacing event every
interval period depending on the location of the stand. Jack pine and lodgepole pine overlap
in range in the northeast corner of BC, the southern Yukon and Northwest Territories, and in
central Alberta (Forintek Canada Corp., 1994).

i4
Figure 2a: Jack Pine Range (Farrar, 1995).

16

I
/ ’fli /’/ ï r ü tT u ii

r '"1^

Figure 2a: Lodgepole Pine Range (Farrar, 1995).

1.3 Thesis Methodology
Pine samples were collected from three different areas. Jack pine samples were
collected from central Alberta, east of the introgression zone; lodgepole pine samples were
collected from the Prince George area of British Columbia, west of the introgression zone;
and potential hybrid samples were collected from the Fort Nelson region of BC, within the
introgression zone (further location and site descriptions can be found in Section 1.4). In
order to classify each sample as lodgepole pine, jack pine, or hybrid, several characteristics
were selected as potential identifying traits for each species group. These traits included
gross morphological characteristics such as needle length, cone length, tree height, and tree
diameter at breast height (DBH); wood and fibre morphology such as earlywooddatewood
ratios, density, moisture content, cell size, modulus of elasticity, microfibril angle, fibre
length, and fibre coarseness; and wood chemical extractive makeup.
To ensure that classification of species group was unbiased and to enable the evaluation
of predictive abilities of certain species traits, the parental lines of each sample were
genetically identified using restriction fragment length polymorphisms from chloroplast and
17

mitochondrial DNA. This allowed for genetic variation in species to be idenitified as well as
possible hybridization.
Evaluation of each characteristic required various methods ranging from qualitative
assessment to minute quantitative measurements. Height and DBH were measured and in the
field, while all other measurements required sample collection for more detailed analysis,
some using specific equipment. Characteristics requiring special preparation for
measurement included wood and fibre traits, wood extractive chemistry, and genetic identity.
Wood and fibre traits were measured through scanning electron microscope, fibre
quality analysis, and SilviScan analysis. Each of these methods required step by step sample
preparation as outlined in detail in Chapter 4. The electron microscope was used to obtain
images of the wood samples in order to distinguish visual differences between wood
characteristics such as latewood content. Fibre Quality Analysis was preformed to obtain
statistically correct measurements of fibre length and coarseness using wood material broken
down into pulp form and then dispersed to a low concentration in water in order to measure
individual fibres. SilviScan technology was used to obtain measurements for other wood
characteristics that are more difficult to obtain, such as microfibril angle, modulus of
elasticity, and average cell dimensions, using a combination of three different instruments, an
x-ray densitometer, diffractometer, and image analysis equipment.
Wood extractive chemistry was investigated in this study through an in-depth process
as described in Chapter 5. Using acetone, chemical constituents comprising the wood resin
were extracted from sample wood cores and analyzed using gas chromatography - mass
spectrometry. This process enables a spectral output to be obtained from each compound
eluted. Each spectral graph displays the quantities of ions at various masses that were

18

present within each compound. This not only allows for possible identification of each
compound based on its molecular weight, but also allows for quantification of each chemical
within a sample. The chemical extractive make up of a particular sample can potentially be
used to identify the wood species being tested. Some tree species produce unique quantities
of certain compounds, which can then be used for identification purposes.
Chapter 2 outlines the protocol followed in order to obtain the genetic identities of each
study sample. Chloroplast DNA from the matk gene (NCBI, 2004) was used to determine the
paternity of each sample, and mitochondrial DNA from the nadl gene (Jaramillo-Correa et
al., 2003) was used to determine the maternity of each sample. DNA was extracted from
each sample, amplified using polymerase chain reaction, and digested using several
restriction enzymes. Parental lines were determined by matching fragment length patterns of
each sample from within the introgression zone, to those collected from outside the
introgression zone (considered pure species).
By analyzing the variation between each sampling group with respect to each of these
characteristics, the strength of the discriminatory ability of the characteristic was determined.
To add to this, the ability of each characteristic to predict hybrid genotypes is discussed in
further chapters.
The application of this study to forest management, breeding, and manufacturing and
processing is discussed in Chapter 6.

19

1.4 Site Descriptions
1.4.1 Locations and Maps
This study involves three main forest stands and nine sites (three sample sites in each
stand) within the species ranges mentioned above, seen in Figure 2. The first stand was
located east of Smoky Lake, Alberta. This stand was a pure jack pine stand, even-aged at
approximately 55 years old. This stand has been used by other researchers to obtain pure
jack pine parent genotype and phenotype information (Dong, et. al., 1992; Danick and Yeh,
1983). The first of the three sites within the jack pine stand was located on Highway 28, east
of Smoky Lake near the Smoky Lake Tree Nursery. Some dwarf mistletoe (Arceuthobium
americanum) and western gall rust (Endocronartium harknessii) were present on trees in the
stand but only endemically. The second jack pine sampling site was located across Highway
28 adjacent to an old burn area. Sampling site #3 was located on the eastern edge of the jack
pine stand. Trees sampled were exposed on the southern face, while north sides of the trees
faced into a densely arranged jack pine stand. Edge trees were more widely spaced, and
possibly affected by wind and weather. Dwarf mistletoe was more prominent at this site than
sites #1 and #2.
The second stand used for sampling was a lodgepole pine stand west of Prince George,
BC. This stand was an even-aged mature lodgepole pine stand with an average tree age of 83
years. Three sites were chosen within this stand (sites #4, #5 and #6); two sites on the
Bobtail Forest Service Road (FSR) (site #4 and 5), and one site on the Gregg Creek FSR (site
6). Site 4 was located 0.5 km south of Highway 16 on the Bobtail FSR. Site 5 within the
lodgepole pine stand area was located at 11 km on the Bobtail FSR, north of Highway 16.
The third site in the lodgepole pine stand (site 6) was located at 1 km on the Gregg Creek

20

FSR, south of highway 16. This site was adjacent to a previous cut-block with trees ages 810 years old. Otherwise this site was similar to sites 4 and 5. All three sites west of Prince
George were subject to fairly heavy mountain pine beetle (Dendroctonus ponderosae)
infestation. Trees were selected that did not demonstrate symptoms of attack. Site 6 was the
most heavily infested of the 3 sites.
The third study area was located north of Fort Nelson, BC, bordering Alberta and the
Yukon Territory. This area is distinct because it is the only area found in BC where the
species range of jack pine and lodgepole pine overlap. Trees in this area ranged in age
between 50 and 120 years old, with an average age of approximately 68 years. Sample sites
for this study (site #7, 8, and 9) are located within this potential hybrid zone for jack pine and
lodgepole pine. The first sampling site was located -5 0 km north of Fort Nelson on FSR
7866-01 (the Patry Mainline) at 70.5 km mark. This site was primarily pine (-95% ) with
some spruce and a small deciduous component. The second sampling site was located five
kilometers west of site #7 on the same FSR at 75.5 km mark, with the same characteristics.
The third sampling site in the Fort Nelson region (site #9) was located approximately 40 km
north of the Patry Mainline turnoff, on Highway 77. This site was a smaller even-aged pine
stand composed of -55% pine. Site maps are located in Appendix 1.

21

Lodgepole Pine

FavtNeIc

0 1 Jack Pine

|||l\Aenüdf%dHid;&ane
#

Samples Areas

Figure 3; Sample area locations; 3 sites are located within each area
( from Sharma and Potter, 2000)

1.4.2 EEC Information
British Columbia is divided into 13 distinct biogeoelimatic (EEC) zones. Alberta does
not bave the same classification system, so the same descriptive facts do not exist for the
sampling areas identified in AE.
• Smoky Lake sites selected for sampling are classified in the Eellis Lake and Eellis
North Natural Areas which are described further in Section 1.4.3.
• Bobtail and Gregg creek FSR sites (west of Prince George) fall within the Sub-Boreal
Spruce EEC zone, subzone moist-cool with a variant of l(S E Sm kl) (Meidinger and
Pojar, 1991). Site series for the Prince George study sites is 03.
• The Fort N elson sites selected for this study all fall within the B W B Sm w 2

biogeoelimatic zone. Boreal White and Black Spruce zone, moist and warm subzone
with a variant of 2. It was determined using vegetation, soil descriptions, climate

22

information and topography of the area. Site series for the areas sampled is 02,
representing the lodgepole pine forest cover of the area (Delong, et.al., 1990).
Site index was determined for the above sites in order to approximate the productivity
levels of the stands being sampled. The site index value represents the average height of a
dominant or co-dominant tree, in meters, growing on a particular site after 50 years (SteamsSmith, 2001). Site index was determined to be 12.0 for the potential hybrid sites, 15.0 for the
jack pine sites and 16.1 for the lodgepole pine sites, which suggests similar productivity.

1.4.3 Description of Soil and Vegetation
The first jack pine site in Smoky Lake, AB, has a very open understory with minimal
vegetation composed of grey reindeer lichen {Cladina rangiferina), kinnikinnick
{Arctostaphylos uva-ursi), and several types of moss. The second jack pine site had an
understory composed of some deciduous and brush species including balsam poplar {Populus
balsamifera) and alder (Alnus spp.). Vegetation on site #2 was composed almost entirely of
grass species, with a small amount of grey reindeer lichen and moss. This site was slightly
denser in stems per hectare than site #1 in the jack pine stand. The third site was quite
densely populated with jack pine, and therefore little understory was present. Vegetation was
almost entirely grass species.
All three jack pine sites were dry with well-drained, sandy soils. The first and third
sites in the Smoky Lake area fell on the south side of Highway 28 and are therefore classified
in the Beilis Lake Natural Area. The second study site for jack pine fell on the north side of
the highway and was classified under the Beilis North Natural Area. Both natural areas
contain soils of the Nestow soil series described as degraded Dystric Brunisols of loamy sand
texture and aeolian origin, which are rapidly drained and have a fine layer of pine needles

23

overlying a variable Ae horizon. The soil profile for this area consists of loamy sand to sand
with no discernable structure. The B horizon is brown, and lightens as it moves down to the
C horizon. The pH of these soils is slightly acidic (Earnhardt, 2005).
The three sites west of Prince George were characterized as follows: Site 4 had an
understory composed of some alder and trembling aspen {Populus tremuloides). Other brush
species present included high bush cranberry {Viburnum edule), and red-osier dogwood
{Cornus stolonifera). Vegetative species included fireweed {Epilobium angustifolium), and
clasping twisted stalk {Streptopus amplexifolius). This was a mesic site with rolling terrain,
slope varied between 0 and 10%. Soil drainage was moderate and of a finer texture than
found in the Smoky Lake sites. Site #5 was very similar to site 4 only less dense. The
understory vegetation was comprised of moss and grass species as well as bunchberry
{Cornus canadensis) and Labrador tea {Ledum groenlandicum). Slope on this site was 0%
throughout and drainage was moderate to poor. Soil had a greater clay content than site 4.
Site 6, located at Gregg Creek, had a vegetative understory composed of clasping twisted
stalk, grass species, and fireweed. Brush component included highbush cranberry and alder.
Sampling site 6 was a mesic site with 0% slope throughout. Soils were moderately welldrained, and contained some clay content, similar to site 5. Nutrient regimes are classified as
poor to very poor for the SBSmkl site series 03 areas, and moisture regimes are subxeric
(Delong et. al., 1993).
Potential hybrid sites #7, #8 and #9, located outside of Fort Nelson, BC, were
characterized by a site series 02; dominant vegetation cover as primarily lingonberry
{Vaccininum vitis-idaea) and velvet-leaved blueberry {Vaccinium myrtilloides). Tree cover
was on average 40% composed of lodgepole pine, trembling aspen, black spruce {Picea

24

mariana), white spruce {Picea glauca) and jack pine. The shrub layer was representative at
30% cover, herb layer 50% cover, and moss layer 50% cover. Lodgepole pine was common
on wetter sites in this area, in combination with black spruce, or in well-drained higher
elevation sites. Moisture regime was classified as xeric to sub-xeric and nutrient regime is

poor to very poor (Delong, et.al., 1990). Soils at study site 7 and 8 were classified as
Kiwigana soils or eluviated dystric brunisols, associated with calcareous, silty, sandy, or
gravelly fluvial materials. Soil texture was sand overlying very gravelly sand and drainage is
rapid. Soils at study site 9 were classified as Sikanni soils or gleyed gray luvisols, associated
with calcareous moderately fine-to fine-textured morainal materials. Soil texture was
characteristic of silt loam overlaying clay loam. Drainage in this area was imperfect
(Kowall, 1982).

1.5 References
Bamhardt, Leonard. Personal Communication. May 9, 2005. Alberta Tree Improvement
and Seed Centre, Government of Alberta.
Barton, N.H., and G.M. Hewitt, 1985. “Analysis of Hybrid Zones”. Ann. Rev. Ecol. Syst.
16:113-148.
Critchfield, William B., 1985. “The late quaternary history of lodgepole and jack pines”.
Can J o f For Res. 15: 749-772.
Dancik, B.P. and F.C. Yeh, 1983. “Allozyme variability and evolution of lodgepole pine
{Pinus contorta var. latifolia) and jack pine {Pinus banksiana) in Alberta” Can. J.
Genet. Cytol. 25: 57-64.
Delong, C. A. MacKinnon, and L. Jang, 1990. A field guide for identification and
interpretation of eeosystems of the northeast portion of the Prince George forest
region. BC Ministry of Forests, pp.9-11, 21,35.
Delong, C., D. Tanner, and M.J. lull, 1993. A field guide for site identification and
interpretation for the southwest portion of the Prince George forest region. BC
Ministry of Forests, pp. 174-176.

25

Dong, J., D. Wagner, A. Yanchuk, M. Carlson, S. Magnussen, X. Wang, and E. Szmidt,
1992. “Paternal chloroplast DNA inheritance in Pinus contorta and Pinus banksiana:
independence of paternal species or cross direction”. Journal of Heredity 83: 419-422.
Farrar, J.L., 1995. Trees in Canada. Canadian Forest Service Natural Resources Canada,
Canadian Cataloguing in Publication Data, Markham, Ontario, pp.
Forintek Canada Corp., 1994. Properties of Jack Pine. Alberta Department of Economic
Development and Tourism, Edmonton, AB. pp.
Isenberg, I.H., 1951. Pulpwoods of United States and Canada. The Institute of Paper
Chemistry, Appleton, Wisconsin.
Jaramillo-Correa, J.P., J. Bousquet, J. Beaulieu, N. Isabel, M. Perron, M. Bouille. 2003.
“Cross-speiees amplification of mitochondrial DNA sequence-tagged-site markers in
conifers: the nature of polymorphism and variation within and among speices in
Picea” Theor. Appl. Genet. 106: 1353-1367.
Johnsen, K.H., J.F. Major, J. Loo, and D. McPhee, 1998. “Negative heterosis not
apparent in 22-year-old hybrids of Picea mariana and Picea rubens”. Can J of Bot.
76: 434-439.
Kormutak, A., R. Matusova, A. Szmidt, and D. Lindgren., 1993. “Karyological,
anatomical, and restriction fragment length polymorphism characteristics of the
interspecific hybrid Pinus banksiana x Pinus contorta"'. Biologia (Bratislava) 48: 95100.
Kowall, R.C., 1982. “Soils of the Fort Simpson Trail Area”. APD Bulletin 25, report
number 58, British Columbia soil survey, BC Ministry of Environment,
Assessment and Planning Division.
Meidinger, D., and J. Pojar, 1991. Ecosystems of British Columbia. BC Ministry of
Forests. Victoria, BC, Canada.
National Center for Biotechnology Information (NCBI). 2004. “Pinus banksiana chloroplast
matk gene for maturase, complete cds.” Accession no. AB080922.
www.ncbi.nlm.nih.gov/entrez/viewer.fcgi?db=nucleotide&val=22775513, accessed
Oct. 27, 2004.
N ei, M. 1973. “A nalysis o f gene diversity in subdivided populations.” Proc. Nat. Acad. Sci.

70(12-1): 3321-3323.
Organisation for Economic Co-operation and Development (OECD). 2002. “Consensus
document on the biology of Picea Sitchensis (Bong.) Carr. (Sitka spruce)”, series on
harmonization of regulatory oversight in biotechnology, No. 21.

26

Pollack, J.C. and B.P. Dancik, 1985. “Monoterpene and morphological variation and
hybridization of Pinus contorta and P. banksiana in Alberta”. Can J of Bot. 63: 201210 .
Sharma, S. and S. Potter, 2000. “Use of a DNA-Based Hybridization Marker for Hybrid
Pine Discrimination”. Pulp and Paper Research Institute of Canada, unpublished.
Vancouver, BC.
Steams-Smith, S. 2001. Making sense of site index estimates in British Columbia: a quick
look at the big picture. BC Journal of Ecosystems Science and Management 1 (2); 14. http://www.siferp.org/jem/2001/voll/no2/art6.pdf
Szmidt, A., J. Lidholm, and J-E. Hallgren, not dated. “DNA extraction and preliminary
characterization of chloroplast DNA from Pinus sylvestris and Pinus contorta’'.
Dept, of Forest Genetics and Plant Physiology. Swedish University of
Agricultural Sciences. Umea, Sweden.
Wheeler, N.C., and R.P. Curies, 1987. “A quantitative measure of introgression between
lodgepole and jack pines”. Can J of Bot. 65:1876-1885.
Woodruff, D. 1973. “Natural hybridization and hybrid zones” . Syst. Zool. 22: 213-218.

27

Chapter 2: Genetic Analysis of Jack Pine, Lodgepole Pine and their Hybrids
2.1 Chapter Objectives
The objective of this chapter was to genetically identify the paternity and maternity of
each sample based on restriction fragment length polymorphisms (RFLPs). The
identification of paternity and maternity of each sample provided recognition of first
generation jack pine x lodgepole pine hybrids when results produced samples with divergent
ancestry. After lineage identities were assigned to each sample the genetic information
gathered was used to support the evidence of hybridization based on characteristics measured
in the following three chapters.
Distinguishing hybrids from the pure species samples based on genetic analysis verifies
the morphological evidence of hybridization in the Fort Nelson region and supports the
presence of an introgression zone that has been previously documented in literature.

2.2 Support from Existing Literature
2.2.1 Description of CpDNA and MtDNA Markers
The markers used for determining hybridization between lodgepole pine and jack pine
are polymorphisms located in the matk chloroplast gene and in the mitochondrial gene, nadl.
Polymorphisms are identified as a nucleotide difference at the same physical location on a
genome. These differences allow inheritance of specific regions of the chromosome to be
identified and traced as a genetic marker (Miesfeld, 1999). The chloroplast deoxyribonucleic
acid (cpDNA) determines the paternity of the sample; inheritance of cpDNA is
predominantly paternal in both lodgepole pine and jack pine. The mitochondrial
deoxyribonucleic acid (mtDNA) determines the maternity of the sample; inheritance of
mitochondrial genes is mainly maternal in these species (Dong et. ah, 1992).

28

2.2.2 Critical Assessment of Various Markers
Many methods exist to analyze and compare hybridization between species. Random
amplified polymorphic DNA (RAPD) markers, allozyme variability, the Eco2.0 probe, and
chloroplast/mitochondrial DNA markers are all examples of ways to test for hybridization;
these are discussed below.
RAPD markers are very abundant and accessible in most species, but limited in the
amount of information derived from them due to their very short sequence length. These
markers are however applied extensively to create genetic maps in many species (Yin, et.al.,
2001).
Allozyme variation examined through electrophoresis can be used to identify evolution
and variability in tree populations. Haploid tissue within the megagametophyte can be used
for direct analysis and does not rely on controlled crosses (Danick and Yeh, 1983).
Eco2.0 is a probe developed to identify polymorphisms, and has been successfully
applied to Sitka x Interior spruce hybrids (Potter, et.al., 2001). However, this probe has no
assay developed for PCR amplification and therefore requires very high quality DNA from
samples and analysis through Southern blotting, which has largely been replaced by
polymerase chain reaction (PCR), and can be very time consuming.
By analyzing cpDNA and mtDNA through restriction fragment length polymorphisms
(RFLPs) parental lines can be identified. CpDNA identifies father and mtDNA identifies
mother, giving the ability to determine if the species in question is a hybrid or if it is pure.
Because two types of DNA are required for this method, it can be more time consuming than
using one region of DNA that can potentially give the same result, however, results are less
prone to misinterpretation. In a study conducted on Pinus contorta and Pinus sylvestris

29

(Szmidt et. a l, n.d.), it is suggested that little variation exists between individuals of the same
population when looking at restriction patterns of cpDNA. However these areas are different
when comparing between different species. This indicates that cpDNA restriction patterns
are potentially useful for determining relationships in conifer phytogeny. Szmidt also
supports that the chloroplast chromosome is ideal for determining phytogeny in plants
because of its small size (only 150 kilobase pairs in total) and its absence of molecular
heterogeneity. Perhaps the most convincing characteristic that allows the chloroplast genome
to be a desirable area for phylogenetic study is its conservation over evolutionary time
(Szmidt, et. al., n.d.)

2.3 Experimental Design and Hypotheses

Needles from foliar region of pines and inner cambial layers were harvested for DNA
analysis. DNA was extracted from tissues and tested according to protocols listed in Section
2.4.
Thirty samples of pure lodgepole pine, 30 of pure jack pine, and 30 potential hybrid
samples were collected from areas around Prince George, B.C., Smoky Lake, Alberta, and
Fort Nelson region, B.C., respectively, yielding a total of 90 samples tested.
DNA sequences utilized throughout this procedure were obtained through GenBank
Blast searches on the National Center for Biotechnology Information website (Oct.27, 2004),
and from literature with the assistance of Dr. Craig Newton of Vizon SciTec Inc.
It is the belief of many forest scientists, and stated in numerous articles (Ye et. al, 2002;
Kormutak et. al, 1993; Wagner et. al., 1992; Govindaraju et. al., 1988; W heeler and Guries,
1987; Critchfield, 1985; Dancik and Yeh, 1983) that an introgression zone of jack pine and

30

lodgepole pine exists around the Alberta/British Columbia border, and that in this area,
hybrids of these two species exist. Based on these previous assumptions, it is hypothesized
that the parental lines of the samples eolleeted within the introgression zone area will display
that the maternity belongs to one species (either jack pine or lodgepole pine), and the
paternity belongs to the other; thus indicating that the samples in the introgression zone are
most likely FI generation hybrids of the two pines.
More specifically, based on observations made at the time of sample collection in
accordance with the morphology of the samples, it is believed that the samples collected from
sites 7 and 8 in the Fort Nelson sampling area are more like jack pine; where as samples
collected from site 9 in the Fort Nelson sampling area are more like lodgepole pine. It is
questionable if this observation will be shown in genetic evidence or not.
The genetic analysis in this study was preliminary, and therefore, complex evolutionary
details can not be identified from the results. If genetic recombination took place Fn
generations ago, there may no longer be evidence of it at this analysis level. Therefore the
percent make-up or contribution from each species to the potential hybrids will not be
determined, the only question that will be answered is whether or not the parental lines of the
samples are different species (indicating hybridization) or the same species (indicating pure
species). This may or may not correlate to morphological evidence, as some characteristics
will be strongly influenced by the environment, and some traits may not show up in hybrids
after the FI generation, due to back-crossing.

31

2.4 Methodology
2.4.1 Protocols for DNA Extraction
Using the protocol from the Qiagcn DNcasy® Plant Mini Extraction Kit (Qiagcn,
2004), needles and inner bark tissue were ground to a fine powder using liquid nitrogen and a
mortar and pestle. Four hundred microlitres of Buffer A PI were added to 1 ml of ground
tissue, and mixed. The mixture was incubated for 10 minutes at 65°C and inverted 2 or 3
times during incubation. Following incubation, 130 pi of Buffer AP2 was added to the
mixture, and the tubes were placed on ice for 5 min. The mixture was then centrifuged for 5
minutes at 14,000 rpm to separate lysates from precipitates. Lysate was then applied to the
QIAshredder Mini Spin Column and centrifuged for another 2 min. at 14,000 rpm. Flow­
through was transferred to a new tube without disturbing cell-debris pellets that may have
formed during centrifuging. One and a half times volume of Buffer AP3/E was then added to
the solution and mixed by pipetting. A maximum of 650 pi was then added to the DNeasy
Mini Spin Column and was centrifuged at 8,000 rpm for 1 min. Flow through was discarded.
This was then repeated until all of the solution had been run through the DNeasy Mini Spin
Column. The column was then placed in a new 2 ml collection tube and 500 pi Buffer AW
was added. The tubes were spun for 1 min. at 8,000 rpm. Flow-through was discarded and
another 500 pi was added to the DNeasy Mini Spin Column and spun down for 2 min. at
14000 rpm to dry the membrane. The DNeasy Mini Spin Column was then transferred to a
new tube and 100 pi of Buffer AE was pipetted directly onto the membrane. Tubes were
incubated for 5 min. at room temperature and centrifuged for 1 min. at 8,000 rpm to elute.
The elution was repeated to yield a DNA dilution in 200 pi of Buffer AE.

32

Using the phenol - chloroform protocol for DNA extraction according to Newton
(2004), needle and inner bark tissue was ground to a fine powder under liquid nitrogen,
adding polyvinyl-polypyrrolidone as needed. One thousand micro-liters of extraction buffer
were added to the ground material and vortexed to mix. The mixture was incubated at 65 °C
for 15 min. Seven hundred and twenty micro-liters of chloroform isoamyl alcohol were then
added and the mixture was centrifuged at 14,000 rpm for 5 min. Aqueous layers were
carefully removed from the mixture and transferred to a new 1.5 pi tube. Approximately
10% volume of ammonium acetate was then added and the solution was mixed by inversion.
Isopropanol was added to the mixture at 1.1 x volume and the solution was again mixed by
inversion. The tubes were spun down at 7,500 rpm for 5 min. Isopropanol was removed
from the tube being careful not to disturb the pellet formed at the bottom of the tube. The
pellet was washed with 70% EtOH, tubes were spun down, EtOH was removed, and tubes
were left to air dry. Product left in the tube was then dissolved in 100-200 pi TE buffer.

2.4.2

DNA Quality and Quantity
All DNA samples were assessed for the quality of DNA present in the sample and for

the quantity of DNA available within the sample. Quality of DNA was assessed through
electrophoresis, quantity was assessed through testing of various dilutions of DNA in PCR
amplifications.

2.4.3

Polymerase Chain Reaction Analysis
Polymerase chain reaction (PCR) amplification was used to copy select sequences so

larger quantities of DNA could be used for further processes. Lodgepole pine and jack pine
sequences were compared, and primer pairs and restriction enzymes chosen to be species

33

specific markers. In order to identify the portion of DNA needed for amplification, several
primers were used. The cpDNA was processed using two different primer pairs obtained
through GenBank Blast searches; one primer pair to amplify cpDNA Section I, from
nucleotide 1 to 540 on the matk gene (See Appendix 2), and one to amplify cpDNA Section 2
from nucleotide 1081 to 1660 on the matk gene (See Appendix 2). Restriction enzymes cut
jack pine cpDNA while leaving lodgepole pine epDNA in one full sequence. The mtDNA
was processed using three different primer pairs obtained from literature. The first was used
to amplify a portion of the nadl gene, the second was to amplify a section of the nad4 gene,
and the third was to amplify a portion of the nadS gene. An attempt was made to optimize
each PCR in order to gain the best result possible and yield the most DNA for digestion.
Only nadl gene material was successful during amplification and used for obtaining digested
fragment patterns which were species specific.
The protocol established by Newton (2004) for PCR was used for the cpDNA
reactions. Protocols from literature (Jaramillo-Correa et. al., 2003; Kubo et. al., 2000;
Demesure et. al., 1995) were used initially for the n adl, nadA, and nadS regions, with
minimal to no results, therefore requiring optimization. Optimization of the MgCl2
concentration was preformed first. Trials of coneentrations ranging from 1.6 mM to 2.4 mM
of MgCl2 in a 12.5 pi reaction were preformed. Following this, the optimal MgCl2
concentration was utilized while annealing temperatures were varied for optimization. A
PCR trial was set up with samples subjected to annealing temperatures that varied on a
gradient from 42°C to 55°C. The annealing temperature that produced the best product was
utilized for the full reaction of all samples in nadl using the PCR protocol by Newton (2004),

34

however, no results were obtained for the nadA and nadS reactions even after optimization,
so these gene sequences were not used.

CpDNA primers for PCR amplifications are as follows:
•

PbMatSnaBF - 5’ ATGGATGAGTTCCATAGATG 3’

•

PbMatSnabR - 5’ AACAGATCGTAATGGGTGCA 3’

•

PbMatHhaF - 5 ’ ATTCTGTGACATATCAGGGC 3 ’

•

PbMatHhaR - 5’ TTCTCATTGCACACGGCTTT 3’

MtDNA primers for PCR amplifications are as follows:
•

nadSaF - 5 ’ AGTCCAATAGGGACAGCAC 3 ’

•

nadSaR - 5’ ACCCGACGATAACTAGCTTC 3’ (Jaramillo-Correa et. al., 2003)

•

nad4L-orf25F - 5’ TATTACTTTCCGAGTCCGGGG 3’

•

nad4L-orf25R - 5’ TCTTCTTCGAACTTGATGCAC 3’(Kubo, et. al., 2000)

•

nadlexonBF - 5’ GCATTACGATCTGCAGCTCA 3’

•

nadlexonCR - 5’ GGAGCTCGATTAGTTTCTGC 3’(Demesure et. al., 1995)

A. Primer Preparation:
DNA/RNA free H 2 O was added to the primer tubes at ten times the initial primer
volume to resuspend the oligonucleotide. Five pi of 50nM forward primer and 5 pi of 50nM
reverse primer were mixed together and 90 pi of H 2 O was added to further dilute the primer
mixture to 0.5 nM.
B. DNA Preparation:
DNA was further diluted to 1/100 gl in DNA/RNA free H 2 O. This diluted DNA
concentration was used for chloroplast PCR amplifications and the initial nadl PCR
35

amplification. Secondary nadl PCR amplifications were preformed using DNA
eoncentrations in their original TE dilutions, with no further dilution due to some DNA
degradation over the time frame in which the DNA was used.
C. Amplification:
For 12.5 pi PCR reactions, 1.25 pi of lOx buffer was mixed with 0.75 pi 25 mM
MgCla (varying depending on optimized concentration), 0.2 pi dNTPs (20 mM each), 1.0 pi
DNA at varying concentrations depending on optimal, 1.0 pi primer (F/R mix at 0.5nM), 0.3
pi platinum taq polymerase (5 units/pl), and 7.5 pi of H 2 O (or amount to top to 12.5 pi).
Some reactions were made at 50 pi (same proportions as above) in order to create enough
product for digestion by restrietion enzymes. Master mix of buffer, MgCli, dNTPs, taq, and
water was made based on number of samples prepared. DNA and primer were added

separately to each tube. CpDNA primers were used to successfully amplify product using the
following procedure: 95°C for 3 minutes, 95°C for 30 seconds, 55°C for 15 seconds, 72°C
for 1 minute, repeat steps 2-4 for 35 cycles, 72°C for 5 minutes (Newton, 2004). MtDNA
nadl primer product was successfully amplified using the following optimized PCR
procedure: 95°C for 3 minutes, 95°C for 30 seconds, 55°C for 15 seconds, 72°C for 1
minute, repeat steps 2-4 for 35 cycles, 72°C for 5 minutes. MtDNA nadA and nadS primer
products were unsuccessful in amplification after several attempts using procedures from
literature, variations in these procedures regarding annealing temperature and extension time,
and variations in MgCla concentration and DNA concentration. Procedures according to
Newton (2004) were also unsuccessful for nadA and nadS. Negative control samples were
run to ensure mixtures were not contaminated.

36

2.4.4 Restriction Digest
For the cpDNA matk gene, restriction enzymes SnaBl and H hal were used to cut the
DNA at TACGTA and GCGC, respectively (See Appendix 2). Digestion of the mtDNA
nadl fragment was necessary in order to interpret the difference between the species after
PCR amplification of the large fragment. This digestion was preformed using three
restriction enzymes: R sa l, M bol, and H hal.
Jack pine cpDNA displayed 2 DNA fragments at different base pair weights when cut
with SnaBl and H hal, while lodgepole pine cpDNA remained in one fragment. The variation
in number of fragments after digestion identified the paternity of the samples.
The nadl gene of mtDNA was digested using R sal, M b o l, and H hal. Each formed a
unique banding pattern which identified variation between species. The variation in banding
pattern after digestion identified the maternity of the samples. Banding patterns formed from
the three digests were analyzed by comparing bands with those that existed in the original
amplified segment of DNA. Smaller sized bands that were conserved from the original DNA
segment and appear in all the digested material were eliminated from the analysis. These
bands are most likely caused by non-specific primer binding with other sites in genome that
results in amplified by-product, and was not the targeted material.
In order to ensure that digests were complete, all true product band lengths were added
to equal the length of the original PCR product. This length for the cpDNA products was
540 base pairs (bp) for cpDNA Section 1, and 580 bp for cpDNA Section 2; mtDNA product
was approximately 2000 bp.

37

After determining the maternity and paternity of each sample based on the restriction
fragment length polymorphisms it was possible to piece together the parental line of each
sample and identify if any hybrids existed within the sample sets.

2.4.5 Imaging
After running DNA samples on agarose gels stained with EtBr, the electrophoresis
output was determined using an ultra violet image analysis tool. This allowed digital images
of the gels to be viewed and saved electronically.

2.5 Results
The following images display the gel electrophoresis output from DNA extraction,
amplification, and digestion. Illuminated bands indicate a presence of DNA at a certain
weight in base pairs.
Table 2: Sample numbers as collected from each sampling area.

Sample #
1 -30
31 -60
61 - 90

Species Area
Jack Pine
Lodgepole Pine
Potential Hybrid

2.5.1 DNA Extraction

Samples 1 through 23 and 27 were initially extracted using the chloroform/phenol
method outlined in Section 2.4.1. Bands were faint using this extraction procedure indicating
less than optimal extraction amounts, therefore a Qiagen® plant mini DNA extraction kit was
used for the remainder of the samples, and only samples with acceptable output were kept
from the chloroform/phenol method. All samples extracted using Qiagen® plant mini DNA
extraction kit yielded good DNA output. Figure 4 displays DNA obtained from this

38

extraction along with labeling for sample numbers. Figures thereafter are not labeled, but
sample numbers are given in the caption and numbering follows the same method, left to
right and top to bottom.
The only samples without visible product from extraction were 33, 36, 37,42, 57, 58,
and 60. Sample 11 was lost and therefore none of the results include sample 11. All samples
were used for PCR regardless of product visibility in initial extraction output.

Row 1: Sample # 2 6 27 28 29 30 31 32 33 34 35 ladder

37

38 39 4 0 41

42 43 4 4 45

ladlSl

Figure 4: Extraction o f D N A , row 1 left to right; samples 26-35, X Hind III ladder, and row 2 left to right; 36-45,
X Hind III ladder.

2.5.2 PCR Products
DNA extractions were diluted to 1/10^ and 1/100* the concentration and amplified as
outline in Section 2.4.2. Most samples were amplified successfully. Some DNA samples
degraded over time and were therefore not amplified and exeluded from the digestions, these
samples are; 1, 3, 4, 6, 7, 10, 12, 17, 33, 34, 57, 63, 64, 74, 76, 82, and 84. For subsequent
statistical analysis, it was assumed, due to similarity in sampling location and therefore
pollen and seed sources, that these samples follow similar fragment patterns to those digested

39

and displayed in Section 2.5.3. Some amplifications revealed faint, non-specific banding at
lower fragment base pair weights. However this does not interfere with the product
distinction and were taken into account in the analysis of restriction fragment patterns.

Figure 5: C pDNA Section 1 PCR product (refer to Appendix 2). W ells left to right; 100 bp ladder, samples 72-

86 .

Figure 5 displays the PCR product for cpDNA Section 1, which was subsequently
digested with the SnaBX enzyme. In most samples, the amplified fragment of DNA showed
up well on the gel output. The desired fragment was 540 base pairs in size, as expected.

R ow 1:

R ow 2:

Figure 6; C pD N A Section 2 PCR product (refer to Appendix 2). R ow 1, w ells left to right; lOObp ladder,
sample 2-21 (no 11). R ow 2, wells left to right; 100 bp ladder, samples 22-38 (no 31), negative control, 100 bp
ladder.

40

Figure 6 shows an example of the PCR product obtained for cpDNA Section 2, which
was subsequently digested with the H hal restriction enzyme. Most of the samples amplified
well, and produced a clear fragment band 580 base pairs in length.

Figure 7: N a d \ PCR product (optimized annealing temperature o f 55°C). W ells left to right; 1 kb ladder,
sample 13, 14, 43, 44, 73, 74, 15, 16, 45, 46, 75, 76, 17, 18, 47, 48, 77, 78.

Figure 7 shows the nad\ PCR product with the optimized DNA concentration and annealing
temperatures. Most samples produced a clear band around 2000 base pairs in size. The
bands show no distinguishable size difference between species groups. This trial was
preformed at PCR conditions as per Newton (see section 2.4.2), with a 55°C annealing
temperature.

2.5.3 Restriction Digests

Figure 8: C pD N A Section 1 PCR product digested with SnaBl enzyme. W ells left to right; sample 59-77, 1 kb
ladder.

41

Figure 8 shows the digestion of Section 1 of cpDNA matk gene PCR product by the
restriction enzyme SnaBl. Samples 1 through 30, and 61 through 80 were digested into 2
small size fragments the first weighing approximately 320 base pairs, and the second
weighing approximately 220 bp, as shown in Figure 8. Samples 31 through 60, and 81
through 90 were not digested and remain in one product band. The undigested bands are of
the same size as the original PCR product (540 bp). Some bending in the wells occurred
during electrophoresis, which accounts for the uneven placement of the bands. This is a
source of experimental error and was probably due to impure TBE buffer in which the
electrophoresis was preformed.

Figure 9: C pDNA Section 2 PCR product digested with H h a\ enzyme. W ells left to right; sample 21-39, 1 kb
ladder.

Figure 10: C pD N A Section 2 PCR product digested with H h a l enzyme. W ells left to right; 1 kb ladder,
sample 78-90.

42

Digestion of Section 2 of the matk gene with the H hal enzyme supports the Section 1
digestion results, showing the same digestion pattern of samples (Figure 9 and 10). Samples
digested compared to those undigested are clearly seen in these figures. Samples 1 through
30 and 61 through 80 were digested into 2 small size fragments, 300 bp and 280 bp; samples
31 through 60 and 81 through 90 were not digested. The large band remaining in the
undigested samples is the same size as the band in the original PCR product (580 bp). The
results from this digestion coordinate exactly with the results from the SnaBl digestion,
shown in Figure 8.
As expected, the jack pine samples were all digested in both the cpDNA reactions.
None of the lodgepole pine samples were digested in these reactions. This variation in
digestion provides a basis for comparison of putative hybrid samples. The potential hybrid
pine samples showed varying results. Samples 61 through 80, from the first 2 sites in the
Fort Nelson area, show paternity of jack pine (digested bands); samples 81 through 90, from
the third site in the Fort Nelson area, show paternity of lodgepole pine (undigested bands).
Each sample was assigned a haplotype in order to compare banding patterns for proper
identification of paternity as well as to identify any unique samples, refer to Table 3.

Figure \ ï \ N a d \ PCR product trial digestion using R s a l, M b o l, and H h al. W ells left to right; Ikb ladder,
sample 20 H, 46 H, 73 H, 81 H, 20 R, 46 R, 73 R, 81 R, 20 M, 46 M, 73 M, 81 M.
H = H h a l, R = R^al, M = M b o l

43

Figure 11 shows the digestion o f the nad\ gene PCR product by R sa \, M b o l, and H hal

restriction enzymes. Each enzyme cut the PCR product into several fragments, forming
different banding patterns depending on the enzyme. Based on this trial, sample 20 (jack
pine) displays a different banding pattern then samples 46 (lodgepole pine), 73 (potential
hybrid), and 81 (potential hybrid). This indicates that the potential hybrid samples match the
lodgepole pine banding pattern for the mitochondrial gene (nadl).
Some sample digestions have additional bands which can not be explained without
further analysis. For example, sample 59 contains a fragment at approximately 400 bp which
is found in every digest (R sal, M bol, and H hal). This fragment is not found in the original
PCR product, but could be mtDNA material which was not intended to be amplified. It is not
considered to be a digested product fragment, forming an alternate haplotype, because this
band was eonsistantly found throughout each digestion, and the addition of this fragment
would bring the total digested fragment weight to well over that of the original amplified
product (-2000 bp). Other bands were incompletely digested, also evident because of
expected total band weight for each sample. These “suspicious” bands were not considered in
the DNA analysis.

Figure 12: N a d l PCR product digestion using R s a l (left to right): 100 bp ladder, sample 24, 25, 26, 27, 28, 29,

30, 31, 32, 35, 36. 37, 38, 39,40, 41,42,43,44.

44

The digestion using restriction enzyme R sa\ reveals that different haplotypes exists for
each species group; haplotypes are displayed in Table 4. There are distinctive bands present
to distinguish between speeies, and the absence o f a bright band at approximately 360 base

pairs distinguishes most jack pine samples (samples #1-30) from the rest of the samples.
However, the 360 bp band is present in 4 of the jack pine samples (sample 19, 27, 28, and
29). All the potential hybrid samples (61-90) have banding patterns most closely matching
the lodgepole pine samples (31-60).
The band present at -500 bp in almost all samples was disregarded due to its presence
in the original PCR amplified product, indicating that it is not a product of the intended
digest.
The presence of an extra band in the lodgepole pine and hybrid samples (as well as the
four unique jack pine samples) makes the total weight of the digested fragments to -2250
base pairs, which is larger than the total weight of the digested jack pine fragments (-1900
bp). This difference could be accounted for by smaller fragments in the jack pine samples
which are not visible due to the expected fainter staining of smaller sized bands and/or the
migrations of the EtBr stain which runs in opposite direction to the samples, and therefore
does not highlight lower regions of the gel where smaller bands would be located.

Figure 13: N a d l PCR product digestion using M b o l (left to right): 100 bp ladder, sample 24, 25, 26, 27, 28,

29, 30, 31, 32, 35, 36, 37, 38, 39, 40, 41, 42,43,44.

45

Most jack pine samples have a distinctive band at approximately 1200 base pairs,
where as all lodgepole pine samples and the potential hybrid pine samples have a band at
approximately 1100 base pairs. Jack pine samples 19, 27, 28, and 29 have the same band
pattern as the lodgepole pine samples, as previously seen in the R sal digestion (Figure 12).
In this digest the jack pine samples have a larger fragment present in the digest than
lodgepole pine samples (1200 bp and 1100 bp, respectively). This may indicate that the
original PCR product for jack pine was slightly larger than lodgepole pine, however was not
distinguishable based on the size of the fragment and combined electrophoretic output. It is
also possible that smaller bands, whose presence in lodgepole pine samples would make the
total weight of the digested bands equivalent to that of jack pine, are not distinguishable in
the digests. This is also true for the digest using H hal, as seen in Figures 14 and 15.
As stated in the digest for R sal, the band present at -500 bp in almost all samples was
disregarded due to its presence in the original PCR amplified product, indicating that it is not
a product of the intended digest.

Figure 14: N a d l PCR product digestion using H h a l (left to right): sample 24, 25, 26, 27, 28, 29, 30, 31, 32,
35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 100 bp ladder.

46

•

-

#W«,

#

Figure 15: N a d l PCR product digestion using H h a l (left to right): 100 bp ladder, sample 45, 47, 48, 49, 50,
51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 65, 66.

Digestion with the H ha\ restriction enzyme revealed species specific banding patterns.
Jack pine samples have a distinctive band at 1350 base pairs, and lodgepole pine samples
have a band at approximately 1250 base pairs.
Potential hybrid samples have the same band as the lodgepole pine samples, paralleling
the results of the previous two restriction enzymes (Rsal and M bol). The same four jack
pine samples display lodgepole pine banding as seen in R sal and M bol digestions. As stated
for the other digests, the band present at -500 bp in almost all samples was disregarded due
to its presence in the original PCR amplified product, indicating that it is not a product of the
intended digest.
Jack pine banding patterns can be differentiated from lodgepole pine banding patterns
according to the results from the mtDNA nadl gene digestions. All of the potential hybrid
samples most closely match the lodgepole pine bands indicating that the maternity of the
potential hybrids is lodgepole pine. It is also evident that the maternity banding pattern of
jack pine samples 19, 27, 28, and 29 is unique, with bands that are found in both pure
species.
All nadl digestions were classified according to haplotype, shown in Table 4. Each
haplotype identifies a different banding pattern. Based on these groupings, parental identity
and similarities or differences between samples can be easily distinguished.

47

Haplotypes;
Table 3: Paternity haplotypes for each species group broken down by collection site.
Number indicated in the “sites” column represents the number of the corresponding
haplotype that was found within the samples from that site. For example. Hybrid site 8 has 7
samples with the haplotype AA.
Jack pine
sites
Haplotype
Number

cp1
cp2
cp3
cp4

SnaBl
A
B
C
D

Hha^
A
A
A
B

Lodgepole pine
sites

Hybrid pine
sites

1

2

3

7

8

4

4
3

6

8

8

9

4

5

6

8

8

10

10

4

Table 4: Maternity haplotypes for each species group broken down by collection site.
Number indicated in the “sites” column represents the number of the corresponding
haplotype that was found within the samples from that site. For example, Jack pine site 1 has
4 samples with the haplotype AAA.
Jack pine
sites
Haplotype
Number

mtl
m12

Rsa^
A
B

Mbol

A
B

Hha-\
A
B

Hybrid pine
sites

Lodgepoie pine
sites

1

2

3

7

8

9

4

5

6

4

6
1

7
3

8

8

8

8

10

10

Tables 3 and 4 reveal similarities between maternity and paternity within some
samples. There are also a few samples with distinct haplotypes. Hybrid samples from site 9
have both maternity and paternity matching lodgepole pine samples, which has unique
implications. Jack pine samples with an mt2 haplotype are unique; these findings will be
discussed in Section 2.6. In total 16 out of 24 hybrid samples had a cpl paternal haplotype
and an mt2 maternal haplotype, matching jack pine paternity and lodgepole pine maternity.

48

2.6 Discussion and Conclusions
It was found in the cpDNA digestions that samples 1-30 possess a confirmed paternity
of jack pine and samples 31 -60 posses a confirmed paternity of lodgepole pine. These
samples also have a maternity matching their respective species with the exception of sample
19, 27, 28, and 29, which were expected to have a maternity of jack pine but in fact seem to
have banding patterns that indicate a lodgepole pine maternity. These samples (19, 27, 28,
and 29) have genetic patterns of possible hybrids, and for statistical analysis of traits they
will be classified as species group 4. Due to the fact that samples 19, 27, 28, and 29 all
possess a paternity of jack pine and banding patterns with lodgepole pine maternity (mt2
haplotype), it can be assumed that these specific samples may also be natural hybrids,
perhaps from an introgression event that has been developing over centuries. It is possible
that these trees underwent hybridization many generations ago, and since then have been
back-crossed with pure jack pine trees, losing the distinct FI polymorphism that is used to
distinguish hybrids through restriction digestion in this study; although these trees have all
morphological traits resembling only jack pine which distinguishes them from other hybrids
studied. This may also be an indication that hybrids and introgression took place further east
in Alberta than documented.
Samples 61-80 (excluding samples 63, 64, 74, and 76) have paternity matching the jack
pine samples and maternity matching the lodgepole pine samples indicating that these species
are most likely a natural hybridization of the two pine speeies. Excluded samples are
expected to follow the same trend in analysis and were excluded because of degradation of
DNA during laboratory processes. These hybrids can be interpreted as FI generation hybrids
because they show direct lineage to pure species groups. However, these trees would also

49

display a jack pine paternity and lodgepole pine maternity in 50% of the F2 genotypes, and
50% of backcrossed genotypes.
Samples 81-90 (excluding samples 82 and 84) have a paternity matching lodgepole
pine samples and a maternity also matching lodgepole pine, indicating that these samples,
although morphological resembling natural hybrids for some characteristics such as cone
angle and curvature, do not have the genetic banding patterns to support hybridization in the
FI generation. Excluded samples are expected to follow the same trend in analysis and were
excluded because of degradation of DNA during laboratory processes. More genetic analysis
would have to be conducted on samples 81-90 in order to determine if any hybridization
occurred to form the physical characteristics that these samples possess. It is possible that
these trees are hybrids but do not represent the FI generation, and therefore do not possess
the distinet polymorphism that was used for restriction digestion in this study. These samples
may have been produced from hybrids that were back-crossed with lodgepole pine trees in
the surrounding area, and therefore genetic evidence of hybridization is indistinguishable at
this level of analysis. For statistical analysis of traits, samples 81-90 will be classified as
species group 5.
The results for samples 19, 27-29, and 81-90 leave room for interpretation, as the
banding patterns visible for these samples are unique and could represent within species
variation or a possible past introgression event; a more in depth genetic analysis is required in
order to further explain these specific sample cases (Refer to Appendix 2). Evidence that
indicates that these four samples are different from other jack pine samples is clear in Tables
3 and 4. Samples 19, and 27-29 have cpDNA haplotypes matching other jack pine samples
(cpl), but mtDNA haplotypes matching lodgepole pine samples (mt2). Although these

50

samples seem like FI generation hybrids, they are morphologically identical to other pure
jack pine samples, and therefore suggest a past introgression with pure jack pine. Evidence
of introgression in this area indicates that the Alberta introgression zone may extend further
East than previously thought. Samples 81-90 are also good examples of possible past
hyrbidization with back-crossing was lost likely with lodgepole pine based on banding
patterns and trees found in the immediate surrounding area.
Slight within species variation was expected and shown in results as variation in
haplotype; cpl, 2 and 3 show variation in jack pine samples (Table 3). Variation in
haplotype could be due to individual tree differences caused by adaptation to environmental
gradients, or due to possible introgression of genes over generations. Back crossing in
populations of both species could potentially disguise genetic inheritance. Hybrids could be
present that do not show up from digestion because of too much introgression, and
dominance by one species over the other. For these hybrids to be identified, further genetic
investigation is required. It is not surprising that some samples were found with morphologic
characteristics of hybrids but a pure species genotype. Introgression and creation of dines
(as discussed in Section 1.3.1) means that a genetic gradient exists within the hybrid area
(Woodruff, 1973). For this study, the d in e can be most closely characterized as a parapatric
zone because while jack pine populations remain localized from lodgepole pine populations;
hybrid stands form between parental pure species stands, and are somewhat integrated (refer
to Figure 3 and Table 1). According to Barton and Hewitt (1985), a parapatric population
indicates that full equilibrium between pure species populations and hybrid has not been
reached, which is supported by the existence of FI generation hybrids in the Fort Nelson
region, that must have been crossed within the last 150 years.

51

The genetic evidence in this study supports the presence of hybrid jack pine x
lodgepole pines. It can be concluded based on the genetic evidence, that there is

hybridization between lodgepole pine and jack pine occurring in the Fort Nelson region of
British Columbia. Assumptions can be made that species crosses exist where morphological
evidence occurs, and that tree displaying morphology of both species could eventually be
identified as genetic hybrids with the appropriate laboratory analysis.
In following chapters, morphology, wood chemistry, and wood and fibre properties are
discussed. Samples from each sampling area were used for comparison of various tree
characteristics. For this comparison, it is useful to look at the variation among species groups
as distinguished by area: 1) Smoky Lake - jack pine 1-30, 2) Prince George - lodgepole pine
31-60, 3) Fort Nelson - hybrid pine 61-90; and also as distinguished by genetic RFLP group:
1) jack pine 1-18, 20-26, 30, 2) lodgepole pine 31-60, 3) hybrid pine 61-80, 4) unique jack
pine haplotypes 19, 27-29, and 5) possible Fn generation hybrids 81-90.

2.7 References
Barton, N.H., and G.M. Hewitt, 1985. “Analysis of Hybrid Zones”. Ann. Rev. Ecol. Syst.
16:113-148.
Critchfield, 1985. “The late quaternary history of lodgepole and jack pines” Can. J. For. Res.
15: 749-772.
Danick, B.P. and F.C. Yeh, 1983. “Allozyme variability and evolution of lodgepole pine
{Pinus contorta var. latifolia) and jack pine (Pinus banksiana) in Alberta” Can. J.
Genet. Cytol. 25: 57-64.
Demesure, B., N. Sodzi, and R.J. Petit, 1995. “A set of universal primers for amplification
of polymorphic non-coding regions of mitochondrial and chloroplast DNA in plants”.
Mol. Ecol. 4:129-131.
Dong, J., D. Wagner, A. Yanchuk, M. Carlson, S. Magnussen, X. Wang, and E. Szmidt,
1992. “Paternal chloroplast DNA inheritance in Pinus contorta and Pinus banksiana:
independence of paternal species or cross direction”. Journal of Heredity 83: 419-422.

52

Govindaraju, D R, D. Wagner, G. Smith, and B. Dancik. 1988. “Chloroplast DNA variation
within individual trees of a Pinus banksiana - Pinus contorta sympatric region” Can.J
For. Res. 18: 1347-1350.
Jaramillo-Correa, J.P., J. Bousquet, J. Beaulieu, N. Isabel, M. Perron, and M. Bouille,
2003. “Cross-species amplification of mitochondrial DNA sequence-tagged-site
markers in conifers; the nature of polymorphism and variation within and among
species in Piced”. Theor. Appl. Genet. 106:1353-1367.
Kormutak, A. R. Matusova, A. Szmidt, and D. Lindgren. 1993. “Karyological, anatomical,
and restriction fragment length polymorphism characteristics of the interspecific hybrid
Pinus banksiana x Pinus contorta” Biologia, Bratislava 48/1: 95-100.
Kubo, T., M.P. Yamamoto, and T. Mikami, 2000. “The nad4L-orf25 gene cluster is
conserved and expressed in sugar beet mitochondria”. Theor. Appl. Genet. 100:
220.

214-

Miesfeld, R.L., 1999. Applied Molecular Genetics. Wiley-Liss Inc. Toronto, ON.
Newton, C. personal communication. Oct. 5, 2004. Vizon SciTec.
Potter, S., V. Hanemaayer, S. Pitts, and P. Watson, 2001. “Fibre Quality Prediction in
B.C. Spruce with DNA-Based Methods”. Pulp and Paper Research Institute of
Canada, PPR 1528. Pointe-Claire, QC.
Szmidt, A., J. Lidholm, and J-E. Hallgren, not dated. “DNA extraction and preliminary
characterization of chloroplast DNA from Pinus sylvestris and Pinus contorta”. Dept,
of Forest Genetics and Plant Physiology. Swedish University of Agricultural
Sciences. Umea, Sweden.
Wheeler, N., and R. Guries. 1987. “A quantitative measure of introgression between
lodgepole and jack pines” Can. J. Bot. 65: 1876-1885.
Woodruff, D. 1973. “Natural hybridization and hybrid zones”. Syst. Zool. 22: 213-218.
Ye, T. R. Yang, and F. Yeh. 2002. “Population structure of a lodgepole pine {Pinus contorta)
and jack pine {P. banksiana) complex as revealed by random amplified polymorphic
DNA” Genome 45: 530-540.
Yin, T., M. Huang, M. Wang, L.Zhu, Z. Zeng, and R. Wu, 2001. “Preliminary
interspecific genetic maps of the Populus genome constructed from RAPD markers”.
Genome 44: 602-609.

53

Chapter 3: Morphological Analysis of Jack Pine and Lodgepole Pine
3.1 Chapter Obj ectives
The main purpose of documenting differences in morphology between species is for
proper identification of a species in the field and also because of its traditional importance in
biological study. It is important to distinguish between lodgepole pine and jack pine in order
to recognize the presence o f a hybrid lodgepole pine x jack pine tree. Literature available

that described hybrid characteristics in AB and NWT was used in order to identify areas
where hybridization between jack pine and lodgepole pine was plausible in Northern B.C.
Based on the literature, morphological characteristics that would most easily distinguish
possible hybrid individuals were identified, such as crown shape and cone orientation. Since
tree samples in the introgression zone around Fort Nelson were selected for their
morphological traits, resembling possible hybrids, it is necessary to measure how strong the
hybridization actually is by comparing these possible hybrid samples with samples collected
based on “pure species” characteristics. It is possible that some level of intermediacy will be
identified within the hybrid characteristics, indicating the presence of both jack pine and
lodgepole pine influence.
Morphology of pure species and hybrids are investigated to measure ability to
distinguish between pure species and hybrids, and provide a basis for comparison with other
characteristics, namely wood and fibre traits, extractive chemistry, and genetic variability.

54

3.2 Species Characteristics According to Existing Literature
Conifers can be classified by several distinguishing features. Morphological
characteristics that are most commonly used to identify species of conifers, specifically
pines, include; cones, needles, bark, and height and crown shape.
Lodgepole pine is characterized by cones oriented perpendicularly from the stem
(Farrar, 1995). Some cones are serotinous; cone scales are fixed together with resin, so seeds
are sealed until released via a heat interaction of 45-50° C, refer to Figure 17 (Koch, 1996).
The needles of lodgepole pine are found in pairs, 3-7 cm in length, sometimes twisted,
varying in color from light-yellowish green to dark green, with in tact bundle-sheath. Needle
pairs run parallel rather than shaped in a “V” (Farrar, 1995). Koch (1996) found that the
average length of air-dry needles was 5.34 mm which fits into the appropriate size range
stated by Farrar, refer to Figure 16 (1995).
The bark of lodgepole pine is thin and orange to brown and grey in eolor (Farrar,
1995). These trees grow to 30 m in height, with a diameter at breast height of up to 60 cm,
and live to an average age of 200 years. The stem of the tree tapers from base to crown, and
the live crown is conical and located mostly at the apex of the stem (Farrar, 1995).
Jack pine is characterized by very distinct cones, which point away from the stem of
the tree, and curve in towards the branch. These cones range from 3-7 cm long, are found in
clusters of 2 or 3, and are serotinous (Farrar, 1995). Some non-serotinous cones are
produced in the southern extent of the jack pine range (Forintek Canada Corp., 1994). Jack
pine serotinous cones open in heat related events such as fire or direct sunlight, as seen in
lodgepole pine. Cone scales in jack pine are thick and smooth, refer to Figure 17 (Farrar,

1995).

55

Needles of jack pine are paired and usually range from 2 to 4 cm long. Jack pine is the
only short-needled pine in Canada with paired needles. These needles are usually straight but
sometimes twisted, light green in color and spread apart to form a “V”, with bundle sheath in
tact, refer to Figure 16 (Farrar, 1995).
Jack pine bark is thin, reddish brown in color, and has scales that are flaky in a young
tree, and grow into thicker irregular plate-like shapes in mature trees (Farrar, 1995). Jack
pines grow up to 20 m in height, with a diameter at breast height of up to 30cm, and they live
to 150 years of age on average. The stem tapers from base to crown (Farrar, 1995) and live
crown is wide and irregular in shape (Forintek Canada Corp., 1994).
These characteristics are similar to what was found in the samples collected in this

study.

Figure 16: N eedle forms for jack pine (left 3 needle pairs) and lodgepole pine (right 3 needle pairs).

56

Figure 17: Cone forms for jack pine (left) and lodgepole pine (right) (from Farrar, 1995).

3.3 Experimental Design and Hypotheses
Gross morphological characteristics of each tree were measured, including: cone
length and orientation, and needle length and position. Tree height and diameter at breast
height (DBH) were recorded at time of sampling. These phenotypic characteristics were
compared among the samples.
Thirty samples of pure lodgepole pine, 30 of pure jack pine, and 30 potential hybrid
samples were collected from the areas of Prince George, BC, Smoky Lake, AB, and Fort
Nelson BC, respectively. Each site was chosen for its similarity in vegetation and
moisture/nutrient regimes. Sites for lodgepole pine and jack pine were chosen as to be
similar distances from the potential hybrid zone.
The cones, needles, and heights of jack pine and lodgepole pine were different enough
between samples for field identification of each species. The most prominent difference was
the cone orientation. The potential hybrids had a combination of the characteristics of both
species, for example, an average needle length in between that of the short jack pine needles
and long lodgepole pine needles.

57

Some within species variation was expected due to inevitable site variation, but by
using averages of all samples collected, and by comparing species based on site index, an
appropriate representation of each species was gained.

3.4 Methodology
All sample trees were measured for gross morphology. Data was analyzed by
comparing characteristics in order to establish trends and interactions between variables.
General linear multivariate models were used to test the statistical significance of the results
and pairwise comparisons were preformed in order to compare each species group with
respect to each characteristic variable using SPSS version for Windows. Also, cluster
analysis was preformed to establish sample groups that were the most similar to each other
based on the characteristics identified.

3.4.1 Needles
Needle pairs were measured for length (dashed line on Figure 18) and width of “V”
formation (solid line on Figure 18) using electronic calipers. A ratio of needle V width over
needle length was used to compare samples; the width of the V is dependant on the length of
the needles. Five needle pairs per sample were measured to obtain average dimensions; 150
needles per sampling area were measured in total.

Figure 18: N eedle Measurements

58

3.4.2 Cones
Cones were measured for length; curvature of the cones was noted for comparison.
Cone length was measured from the tip of the cone to the base where the cone met the
branch, not including any cone curvature (dashed line on Figure 19). Measurements of cone
height from the branch (solid line on Figure 19) and length along the branch (dotted line on
Figure 19) were also recorded in order to calculate the angle that the cone formed with the
branch (Figure 19). Cone curvature was given a qualitative descriptor; slightly curved,
moderately curved, largely curved, or not curved. An average of two cones per sample were
measured depending on availability at time of sampling. In total, approximately 60 cones per
sampling area were used to calculate dimensions.

Figure 19: Cone Measurements

3.4.3 Height, DBH, and Age
Tree height was measured using a clinometer and measuring tape, and tree diameter at
breast height was measured using a DBH tape. Age was determined within a 5 year error by
counting the growth rings of each sample. Tree height, DBH and age were used to calculate
growth rates of each tree sampled. Using growth rate instead of absolute height and DBH
values allows comparison of various sites by factoring in stand age.

59

3.5 Results
3.5.1 Needles:
Results of mean length, V width, and range of lengths and V widths found are shown in
Table 5. Figure 20 shows the average ratios (V width / needle length) for each sample group
classified by sampling area.
Table 5: High and low range needle measurements displaying the variation within and

Avg. needle length (cm)
Avg. needle V width (cm)
Length min. value (cm)
Length max. value (cm)
V width min. value (cm)
V width max. value (cm)
Number of samples

jack pine
3.015
1.457
1.900
4.700
0.200
4.000
150

lodgepole pine
5.397
0.986
2.900
10.600
0.000
3.900
150

hybrid pine
4.107
1.372
2.700
6.200
0.100
2.900
150

0.6
_

E

I

0.5

0.4

'T 0.3
rc

DC 0.2

0.1

0
jack pine

lodgepole pine

hybrid pine

Figure 20: Average ratio o f needle V width over needle length for each
species/ sample group by area with standard error bars.

Figure 20 shows significant differences among the species based on the ratio of needle
V width / needle length. Jack pine needles were characteristically much shorter and split in a
wider “V” than lodgepole pine needles which are longer and oriented in a much tighter “V”
formation. The hybrid pine needle measurements were found to be in between those of the
other species, moderate in length and “V” width.
60

3.5.2 Cones
Table 6: Average cone angle for each species/sample group. A ngles represent the
number o f degrees that the cone deviated from the branch. Negative values indicate
that the cone curved over the branch to the opposite side from which it began

Jack
Pine
Mean Cone Angle (°) by
sample area
Number of samples

Lodgepole Pine

Hybrid Pine

57.88
52

47.67
66

-4.26
77

60
50

œ
2

Ï

40

g
S

_o) 30 05

<
g

20

1

0

10
0

Î
-10

A Jack Pine

■ Lodgepole Pine

a Hybrid Pine

Figure 21a: Mean cone angle for all species groups according to sampling area with standard error bars.

Table 6 and Figure 21a display mean cone angles. The table gives a comparison
between the cone angle averages calculated. Figure 21a shows that mean cone angles are
lowest in jack pine and highest in lodgepole pine. Hybrid pine samples have intermediate
cone angles. There are significant differences between species groups indicated by the
standard error bars that are present.

61

100

80
E
(D
œ
E

60

I

20

40

o

E

-20

Low R a n g e V alue

High R a n g e V alu e

A

-40
-60 -J
A Jack Pine

Lodgepole Pine

▲Hybrid Pine

Figure 21b: The range in angles that the cones form with the branch for each species/sam ple group
according to sample area. Low range value is the smallest or m ost negative angle observed, and high
range value refers to the largest angle observed in that sample group. Negative values indicate that the
cone curved over the branch to the opposite side from which it began growing.

It was observed that generally the lower or more negative the cone angle value, the
more curvature the cone displayed. As seen in Figure 21h, jack pine cones tend to have
lower angle values, than the other species groups, in both the high-range (highest angles
found) and low-range (lowest angles found) categories. The lower the cone angle value, the
closer the cone tip is to the branch. Clearly there is greater within “species” variability for
the hybrid samples as compared to the jack pine or lodgepole pine samples for cone angle as
a trait.

62

no c u rv e

slight c u rv e

sm all c u rv e

m o d e ra te
cu rv e

la rg e c u rv e

C u rv a tu re C lassification
□ Jack Pine

■ Lodgepole Pine

■ Hybrid Pine

Figure 22: Qualitative classification o f cones based on observed curvature.

Cones were classified based on comparative observation. As seen in Figure 22, less

curvature was observed in lodgepole pine cones than in jack pine cones. Hybrid cones have
high variability in cone curvature, with the highest percentage of cones being “slightly
curved”. As this is a purely qualitative observation it does not have statistical significance,
but when added to the quantitative results, may illustrate a clearer pieture of the differences
between species groups.

Avg. Length (mm)
Shortest cone (mm)
Longest cone (mm)
Number of samples
Cone length SE

Jack Pine
36.55
21
58.5
77
1.260

Lodgepole Pine
36.2
23
44.5
52
1.135

Hybrid Pine
42.87
29
54.5
66
1.029

Average cone lengths and variation in eone lengths are shown in Table 7. The average
lengths were similar among species, hybrids display longer eones on average, possibly due to
factors that will be diseussed in Section 2.5. Jack pine cones had a higher range of cone
length values than lodgepole pine.
63

20
CD

I

15
I— A— I

I

10

0

5

_c

0
34

36

38

40

42

44

Cone Length (mm)
Figure 23: Site index as a function o f cone length for sample groups according to sampling area
A Jack pine ■ Lodgepole pine ▲ Hybrid pine.

Figure 23 shows the effect that site index has on cone length. Lodgepole pine and jack
pine site indices and cone lengths are similar; hybrid average cone length is significantly
greater even though the site index value is the lowest for these samples.

45.00

A

A
E 40.00
E

I
É

A

i

É

03

c 35.00

I
A

0)

Ii

O 30.00

m
25.00

Age C lass
A Ja ck

Pine ■ Lodgepole Pine a Hybrid Pine

Figure 24: Cone length as a function o f A ge Class

Figure 24 shows cone length as a function of age class. Age classes were defined and
given a representative number from 1 - 7 . Age class 1 represents trees aged 20-39, age class

64

2 represents trees aged 40-59, age class 3 = trees aged 60-79, age class 4 = trees aged 80-99,
age class 5 = trees aged 100-119, and age class 6 = trees aged 120-139. Jack pine samples
fall into the lower age classes, while lodgepole pine samples occupy the older age classes.
Hybrid samples range in ages and cone lengths. Even though jack pine samples are younger,
their cones demonstrate a similar average length, and a high-range length value greater than
lodgepole pine. Hybrid cones on average are longer than the other species groups, but
samples are not necessarily from older age classes. Focusing only on age class 3, where all
three species groups have representative samples, and based on Figure 24, lodgepole pine
cones are shorter than the other species groups while hybrid cones are the longest. Cone
length statistics are shown in Table 9.

3.5.3 Height, DBH, and Age
Table 8; Average height, D BH and age values for the given sample groups, and range in

Avg. height (m)
Shortest Tree (m)
Tallest Tree (m)
Avg. DBH (cm)
Smallest DBH (cm)
Largest DBH (cm)
Avg. Age (yrs)
Youngest (yrs)
Oldest (yrs)
Number of samples

Jack Pine
15.34
10.1
23.06
24.94
16.1
36.6
53.31
35
69
30

Lodgepole Pine
22.46
18.58
27.04
22.52
16.3
30.3
83.18
62
140
30

Hybrid Pine
17.86
11.64
22.9
23.92
17
41.3
68.59
44
120
30

Table 8 displays the average height, DBH, and age for each species group according to
sampling area, as well as the range in these values found within each group. As shown, jack
pine has the shortest average height, but also the shortest range in height among all samples.
Lodgepole pine has the tallest average height and the tallest range among samples. This
correlates well to the age information, which shows that the lodgepole pine trees were the

65

oldest of all trees sampled and the jack pine trees were the youngest. Therefore variation in
height can be explained by age differences. In order to compare samples for height
differences without having an age bias, growth rate (height divided by age) was used for
further comparison.
The same issue exists when dealing with DBH, as intuitively, DBH changes with the
age of the tree. Therefore, growth rate (DBH divided by age) is used for comparing DBH
between samples to eliminate any age biases. Observations of DBH suggest that jack pine
has that greatest average diameter, and lodgepole pine has the smallest average diameter.
Hybrid samples show the widest range in DBH.
Figure 25 displays average growth rates for all species groups, and Figure 26 shows
how the differences in growth rates for height and DBH might be explained by site index
differences.

0.6 n
0) 0.5 rS
CC 0.4
x:
1 0.3
O
d) 0.2 >
< 0.1
0
Ja c k Pine

Lodgepole Pine

■ Avg. Growth rate (height/age)

Hybrid Pine

□ Avg. Growth rate (DBH/age)

Figure 25: Growth rates for all sample groups according to area.

In Figure 25 there is no significant difference shown among speeies’ growth rates with
respect to height.

66

Growth rates for DBH show a significant difference between all three species groups.
Jack pine has the fastest growth rate based on DBH, lodgepole pine has the slowest growth
rate, and the hybrid samples display an average growth rate in between those of the pure
species. Further investigation is required to help explain these growth rate differences.
Comparing DBH growth rate to site indices may offer an explanation.

17 n
16 CD
_)
CO
X
CD
T3
C

15

AA A A A A m /A

AV A / V W A A

A

A

A

14
13

B 12 -

Aàk

CO

11 -

in 0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

DBH/age Growth rate (cm/yr)
A jack pine

■ lodgepole pine

a hybrid pine

Figure 26: Site index as a function o f Growth rate (DBH/age)

Figure 26 shows that DBH growth seems relatively unaffected by differences in site
index value for these sample groups. Jack pine displays the highest rate of DBH growth, and
lodgepole pine shows the lowest growth rate. The average growth rate for hybrids is
moderate among these samples, even though they were found on sites with the lowest index
potential.

67

3.5.4 Statistical Analysis
Table 9: Pairwise Comparisons for morphology. Species 1 = jack pine, species 2 = lodgepole pine, species 3 :
hybrids, species 4 = hybrids from jack pine sampling area, species 5 = lodgepole pines from hybrid sampling
area

Dependent
Variable

(1) species

(J) species

Mean Difference
(l-J)

Std. Error

Sig.(a)

95% Confidence Interval
for Difference(a)
Lower
Bound

needle w
1

1

2

.3190

.033

.000

.081 (*)
.006

.033

4

.057

.265(*)

.040

.000

.186

.345

2

5
1

-.3190

.033

.000

-.385

-.253

-.2380
-.3130
-.054

.036
.059
.042

3

3
4
5
1

-.0810

.033

.000
.000
.206
.017

-.309
-.431
-.138
-.146

-.167
-.196
.030
-.015

2

.2380
-.075

.036

.000

.167

.309

-.193

.042

1
2
3

.059
.042
.057
.059
.059

.207

.1850
-.006
.3130
.075

.000
.920
.000
.207

.101
-.120
.196

.268
.109
.431

-.042

.193

5

.2600

.063

.000

.134

.385

1

-.2650
.054
-.1850
-.260(*)

.040

.000

-.345

-.186

.042
.042
.063

.206
.000
.000

-.030
-.268
-.385

.138
-.101
-.134

3.071

1.759

.085

-.438

6.580

3

-3.431

1.759

.055

-6.940

.078

4

8.6800

3.065

5
1
3
4

-2.911
-3.071
-6.502(*)
5.610

2.129
1.759
1.897
3.146

.006
.176
.085
.001
.079

2.568
-7.158
-6.580
-10.286
-.665

14.793
1.336
.438
-2.718
11.884

5

-5.982(*)

2.245

.010

-10.459

-1.505

1

3.431

1.759

.055

-.078

6.940

2
4

6.502(*)
12.112(*)
.520
-8.680(*)

1.897
3.146
2.245
3.065

.001
.000
.818
.006

2.718
5.837
-3.957
-14.793

10.286
18.386
4.997
-2.568

-5.610
-12.112(*)

3.146

3

3.146

.079
.000

-11.884
-18.386

.665
-5.837

5

-11.5920

3.367

.001

-18.307

-4.876

1

2.911

2.129

.176

-1.336

7.158

3

4
5
4

5

2
3
4
cone
length
(mm)

Upper
Bound

1

2

3

4

.385

.017

.015

.146

.920

-.109

.120

2

5
1
2

5

.253

68

Dependent
Variable

(!) species

(J) species

Mean Difference
(l-J)

Std. Error

Sig.(a)

95% Confidence interval
for Difference(a)
Lower
Bound

2

Height/Age

1

5.982(3

2.245

.010

1.505

10.459

3

-.520

2.245

.818

-4.997

3.957

4

11.592(3

3.367

.001

4.876

18.307

2

-.005

.016

.770

-.036

.027

3
4
2

5
1
3
4
5

3

1

4

2
4
5
1

5

1

.028

.425

-.079

.025
-.017
-.022
-.045
.003

.020

.219
.282
.205
.120
.874

-.015
-.049
-.056
-.101
-.037

.065
.014
.012
.012
.044

.325

.082

-.028
-.034

3

.045
.048
-.020
-.025

.028

.120

-.012

.101

.030
.019
.020
.020
.030
.026

-.013
-.059
-.065
-.044

.108
.018
.015

-.003
-.048
.187(3

.120
.291
.219
.874
.120
.000

-.108

.153(3
.039
.070(3

.026
.045
.031

.000
.389
.027

.135
.102
-.050
.008

.013
.238
.204
.128
.132

-.187(3
-.033
-.148(3

.026

.000

-.238

-.135

.028

.235

-.088

.022

.046

.002

-.239

-.057

1
2

-.116(3
-.153(3
.033

.033
.026
.028

-.182
-.204
-.022

4

-.115(3

.046

.001
.000
.235
.014

-.206

-.051
-.102
.088
-.024

5
1
2
3
5
1

-.083(3
-.039
.148(3
.115(3
.031
-.070(3
.116(3

.033
.045
.046
.046
.049
.031
.033

.013
.389
.002
.014
.524
.027

-.149
-.128
.057
.024
-.066
-.132

.001

.051

-.018
.050
.239
.206
.129
-.008
.182

.083(3

.033

.013

.018

.149

-.031
-61.269(3

.049
4.394

.524
.000

-.129
-70.032

.066
-52.506

5
1
2

2

1

2

1

-.023

.028
.059
.036
.056
.034

.425

3
4
cone angle

.049

.028

5

5

-.014
-.082
-.018
-.027
-.012

.023

3
4

4

.282
.325
.291
.770
.205

.027

4
5

3

.016
.028
.019
.016
.017

2

3

2

.017
-.027
.020
.005
.022

.016
.017
.028
.020
.028

3
4
DBH/Age

Upper
Bound

2

.079

.037

69

Dependent
Variable

(I) species

(J) species

Mean Difference
(l-J)

3

2

Sig.(a)

95% Confidence interval
for Difference(a)
Lower
Bound

Upper
Bound

-44.877(*)

4.394

.000

-53.639

-36.114

4

7.630

7.654

.322

-7.635

22.896

5
1

-64.680D
61.269(*)

5.318
4.394

.000

-75.286

-54.073

.000

52.506

70.032

3

16.3920

6.943

25.841

68.899(*)
-3.411
44.8770

4.738
7.857
5.606
4.394

.001

4

.000
.545
.000

53.229
-14.591
36.114

84.569
7.770
53.639

4.738
7.857

.001
.000

-25.841
36.837

-6.943
68.177

5
1

3

Std. Error

2
4

-16.3920

5

-19.8030
-7.630
-68.8990
-52.5070

5.606

.001

-30.983

-8.623

7.654
7.857
7.857

.322
.000
.000

-22.896
-84.569
-68.177

7.635
-53.229
-36.837

5
1

-72.3100

8.409

.000

-89.081

-55.539

64.6800

5.318

.000

54.073

75.286

2

3.411

5.606

.545

-7.770

14.591

3
5.606
.001
8.623
19.8030
4
8.409
.000
55.539
72.3100
Based on estimated marginal means
* The mean difference is significant at the .05 level.
a Adjustment for multiple comparisons; Least Significant Difference (equivalent to no adjustments).

30.983
89.081

4

52.5070

1
2
3

5

Cluster Analysis:
Table 10a: Final Cluster Centers for Morphology

Cluster
1

2

3

Needie_w_l
cone length (mm)

.3923
41.0

.2319
37.7

.4846
37.4

Height/Age

.2650

.2844

.2941

DBH/Age

.3580

.3366

.4842

60.6495

-7.8530

Gone angle

26.0606

Table 10b: Distances between Final Cluster Centers for morphology

Cluster
1

1

2

34.750
34.099

3

2
34.750

3
34.099
68.503

68.503

70

Table 10c: Number o f Cases in each Cluster for morphology

Cluster

Valid
Missing

1

17.000

2

32.000

3

26.000
75.000
16.000

Further statistical data can be found in Appendix 3. The cluster analysis in Table 10a,
b, and c, reveals that most jack pine samples fall into as cluster 3, while lodgepole pine
samples were categorized as cluster 2 and the majority of hybrid samples were categorized as
either cluster l o r 2. A few jack pine samples fell into cluster 1; sample 5, 8, 10, and 25.
Only sample 63, out of the hybrid samples, was identified as part of cluster 3. Samples 61,
62, 64, 65, 67, 68, 69, 71, 75, 76, 80,81, and 90 fell into cluster 1; samples 66, 72, 74, 77, 78,
79, and 82 to 89 fell into cluster 2.

3.6 Discussion and Conclusions
3.6.1 Needle Characteristics
Needle V width and needle length are two characteristics that are identifiably different
among speeies of pines in this study (Table 5), and supports work by Farrar (1995), and Koch
(1996). Wheeler and Curies (1987) used Lubischew’s coefficient of determination in order
to evaluate how effective a particular morphological trait was at distinguishing between
species. The greater the coefficient, K, the more “discriminating power” the trait has; K =
(X a-X b)^2 * (sa*sb) where X is the mean of a certain characteristic in speeies A and B, and s
is the standard deviation of that characteristic in species A and B (Christensen, 2003).
Needle length was used in the study by Wheeler and Curies as one of the traits that had
significant ability to distinguish between species. Other traits that are also used in the study
by W heeler and Curies (1987) include cone curvature, angle of cone attachment, and cone

71

length. Each of these traits allows for supportive species identification, and when observed
in combination allows for accurate assessment of species type. The information provided by
Lubischew’s coeffient of determination for the characteristics analyzed by Wheeler and
Guries indicates that cone angle of attachment and needle length, among others, are very
useful for discriminating between hybrids and pure pine species. This provides support for
the use of these characteristics in this study.
To add to this, needle V width used as a ratio with needle length provides further
support in species identification. As seen in the results in Figure 20, needle V width/length
produced a significant difference among all species groups (jack pine, lodgepole pine, and
hybrids). Figure 20 shows that the ratio for needle V width/length is highest in jack pine,
lowest in lodgepole pine, and intermediate in hybrids, indicating that this trait was modified
in the hybrids due to genetic recombination to produce intermediate needle formations. In
further support that the pure species and hybrids show variation, a pairwise comparison
between species groups revealed that the mean differences between all three main species
groups, when compared to one another, were statistically different for needle width/length
(Table 9). Species groups 4 and 5 showed that jack pine or hybrid samples were not
significantly different from group 4, and lodgepole pine samples were not statistically
different from group 5.

3.6.2 Cone Characteristics
Cone curvature, angle of cone attachment, and cone length are also morphological
characteristics that distinguish between species (Wheeler and Guries, 1987). In this study,
significant differences in cone angles were found among all species groups; variation in

72

angle, i.e. the total range within all samples, was also found to be unique for jack pine and
lodgepole pine. This indicates that cone angle is probably one of the largest differences in
morphology between these two pines. In fact, according to Wheeler and Guries (1987), the
coefficient of determination for angle of cone attachment is one of the most discriminating
factors between jack pine and lodgepole pine. In their study, lodgepole pine cone angles
were an average of 120° +/- 22.6, jack pine cone angles were 16.5° +/- 29.5 and hybrid cone
angles were an average of 52.8° +/- 41.0. The same pattern was found in this study,
lodgepole pine angles of cone attachment were large, jack pine angles are small, hybrid
angles are intermediate and contain the most variability. The pairwise comparison revealed
that cone angles were significantly different among all three main species groups (Table 9).
As seen in the needle V width/ length characteristic, species group 4 is not significantly
different from jack pines, and species group 5 is not significantly different from lodgepole
pines.
The result suggests that hybrid pine cone angles are a characteristic feature that may be
genetically controlled and may have been influenced by both pure species, forming an
intermediate, genetically recombined, characteristic cone angle. This finding parallels needle
V width/length ratio, which also suggested a recombined intermediate characteristic.
Wheeler and Guries (1987) described cone curvature in a quantitative way by
measuring the arc of the curve. In this study, cone curvature was not investigated to its full
extent, but simply given a qualitative value based on comparative observation. As shown in
Figure 22, lodgepole pine cones sampled were much straighter than jack pine cones. Hybrid
cones formed intermediate curves and contained the highest variability, similar to the cone’s
angle of attachment as discussed above, providing supportive evidence of “intermediacy” in

73

the hybrids studied, as compared to the pure species samples of jack pine or lodgepole pine.
Since the cone angle and curvature were found to be so different among these pines, cones
can provide a very important distinguishing feature for field identification of hybridization.
Cone length is another characteristic that has been identified as a possible
distinguishing feature among the pine species. The average cone length was similar for jack
pine and lodgepole pine in this study, but hybrid cones were longer on average (Table 7).
Variability in cone length was fairly consistent throughout the species groups. The statistical
pairwise comparison in Table 9 revealed that the cone length for hybrids was significantly
different from jack pine and lodgepole pine, but the pure species were not significantly
different from each other, indicating that cone length does not act as a characteristic
displaying intermediacy. However, this result does assist in distinguishing hybrids from pure
species within these samples.

3.6.3 Site Considerations
Characteristics mentioned such as needle V width/length, cone angle of attachment,
and cone curvature may be unaffected by site, because there would be little advantage for a
tree to develop these characteristics in one way or another with changes to its environment.
Needle length on its own may be affected by light exposure on some sites, however, when
combined as a ratio needle with V width, the value represents total needle development rather
than just growth. Alternatively, cone length may be affected by site conditions.
For this reason, cone length was investigated as a function of site index in Figure 23,
and as a function of age class in Figure 24. The expected trend, the greater the site index
value the greater the eone length, was not observed in these figures. Hybrids, with the lowest

74

site index value, have the longest cones on average. Therefore differences in cone length
can not he explained by site index variation; the expected outcome, that cones may grow
larger on good sites to increase reproduction, was not observed. Age class did not have a
significant impact on cone length. By selecting only age class 3, where all species groups
were represented, comparing cone length showed that hybrid cones are longer than both jack
pine and lodgepole pine; the same outcome as comparing overall averages. Wheeler and
Guries (1987) identified cone length as a possible distinguishing feature between jack pine,
lodgepole pine, and hybrids, however did not perform full analysis using this trait because it
scored fairly low as a test for Lubischew’s coefficient of determination. For the purpose of
this study, cone length provided distinction between hybrids and lodgepole pine, but not
between jack pine and any main species group or as a representation of intermediacy between
the pure species.

3.6.4 Growth Rate Characteristics - Height/Age and DBH/Age
The final two characteristics that were used to identify morphological traits among
pines in this study were height and DBH used as a growth rate ratio over age. Since age
range varied among the species sampling groups and height and DBH are directly related to
age, it was necessary to compare the groups based on growth rates. Table 8 shows the
measured values of DBH, height, and age before ratio calculations.
Figure 25 shows that there are no significant differences in height growth rate for the
three species groups by a pairwise comparison; Table 9 shows no significant difference
between species groups for the height/age variable. Therefore growth rate represented as

75

height/age is similar between species groups, and is not a valuable tool for species and hybrid
identification.
Contrary to this, the growth rate for DBH did show statistically significant variation
between species, both in Figure 25 and in the pairwise comparison (Table 9). Since growth
rate is heavily dependent on site conditions, it is necessary to discuss DBH/age as a function
of site index value. Figure 26 shows that lodgepole pine has the greatest site index value,
which would normally coordinate with the highest growth rate. However, jack pine
demonstrates the highest growth rate while lodgepole pine has the lowest growth rate.
Hybrids demonstrate the lowest site index value but have an intermediate growth rate.
Therefore, the differences in growth rate can not be attributed to site index, or site potential
in this study. The pairwise comparison for DBH/age (Table 9) shows that there are
significant differences between pure species, and between jack pine and hybrids, but not
between lodgepole pine and hybrids. These statistical results indicate that hybrids may be
more similar to lodgepole pines with respect to DBH growth rate.
Although there is sufficient analysis to support the relevance of growth rate according
to DBH in this study, it would be prudent to investigate it further before identifying it as a
strong distinguishing feature between species because of its “heavy” reliance on site. From
preliminary work on DBH growth rate, it has the potential to act as a distinguishing feature to
identify hybridization between jack pine and lodgepole pine by intermediacy, similar to
needle V width/length, cone angle, and cone curvature.

76

3.6.5 Cluster Analysis
Table 10a, b, and c show the results from the cluster analysis; the values for the cluster
centers, distances between cluster centers, and the number of cases in each cluster. Three
clusters were formed from samples that contained values for all the input variables. Fifteen
samples were missing values and therefore not included in the cluster analysis. Individual
samples included in each cluster can be found in Appendix 3 along with an ANOVA table for
the cluster analysis. Based on the ANOVA output, needle V width/length, DBH/age, and
eone angle were significant variables in distinguishing between clusters. Cone length and
height/age were not significant; this is not surprising considering their ambiguity in
distinguishing between species in previous analysis. The samples were grouped in the cluster
analysis to produce the most variability between clusters. So, based on the output, most jack
pine samples varied significantly from the lodgepole pine and hybrid samples according to
morphological traits that were input (forming cluster 3). Lodgepole pine samples varied
from jack pine samples and were grouped into cluster 2. Hybrid samples however, did not
form a whole distinct cluster. Approximately half the hybrid samples formed their own
cluster (cluster 1), and the other half remained in cluster 2 with the lodgepole pine samples,
indicating that there was more variability within the hybrids than between some of the
hybrids and the lodgepole pine samples. This could be due to the fact that samples 81-90
were found to have unique maternal and paternal haplotypes even though resembling hybrids
for some traits, such as cone angle and curvature. Also it indicates that these samples were
probably not first generation (FI) hybrids, and more likely a back-cross of hybrid and
lodgepole pine, causing most of the morphological jack pine traits and genetic information to
be lost.

77

Overall, the cluster analysis does display a large amount of between species variation
based on the morphological traits identified, therefore allowing for a large amount of
distinction between species groups.

3.6.6 Conclusions
It is important to identify morphological characteristics between species and hybrids
for rapid field identification of these trees. During timber harvesting, forest management, or
research of the areas containing hybrids, it is valuable to be able to quickly and accurately
identify species and hybrids.
Several morphological characteristics provide significant differences between the three
species groups identified in this study; jack pine, lodgepole pine, and hybrids. These
differences provide distinction between the speeies groups and allow for field identification,

as well as support the concept of hybridization. Characteristics that have been identified to
distinguish between speeies groups include: needle V width/length, cone angle of
attachment, cone curvature, cone length, and DBH/age. These characteristics not only
distinguish between species but also show that the hybrids are intermediate to the pure
species, with the exception of cone length. Cone length was also identified as a
distinguishing characteristic, however does not identify intermediacy for hybrids.
Intermediacy is an important factor to consider because it indicates that the hybrids are not
only different from the pure species, but contain a combination of their genetic traits.
Therefore hybrids are in fact a “mixture” of both pure speeies.
Identifying the difference between jack pine, lodgepole pine, and hybrids of these
species in the field is possible based on the characteristics investigated in this study, which

78

addresses the first objective of the morphological traits. Some characteristics that had been
used to identify hybrids in Alberta and N.W.T. (Wheeler and Guries, 1987, and Critchfield,
1985) were used in this study to identify hybrids in British Columbia. This study
successfully identified characteristics among species with strong discriminatory abilities
versus characteristics that were weaker or not useful for distinction. Characteristics that are
most useful for distinction between species are needle V width/length and cone angle, which
not only distinguished between pure species and hybrids, but also displayed intermediacy in
hybrids. Furthermore, growth rate by DBH showed the ability to distinguish between jack
pine and hybrids, and indicated that hybrid sample growth may be more like that of
lodgepole pine than jack pine in this study.
Sources of error exist within the morphological study that may have contributed to
some of the variability in the results. Since the trees sampled were of a mature age,
collecting needles and cones from all over the live crowns was difficult. This means that
needles and cones were primarily collected from the bottom half of the live crown which
could have affected average overall size of the needles or cones. It is also possible that
because trees were collected from different sites they were subject to varying conditions
other than the differences explored in this study.
Many characteristics were used in other studies for identification of hybrids that were
not used in this study. In order to build on the information gathered here, and support the
conclusions drawn, these additional characteristics could be studied in jack pines, lodgepole
pines, and their hybrids for their ability to distinguish species groups. Wheeler and Curies
(1987) used leaf serrations, prickle, and seed coat mottling among others, as distinguishing
traits with successful results; Christensen and Dar (2003) used the number of stomata on the

79

adaxial surface as well as other traits for comparisons. Another comparative study in
Juniperus by Adams (2003), used leaf margins (entire vs. serrate), leaf thickness, and length
of stomatal apparatus to distinguish between various species, all of which could be used for
the pine species in this study.

3.7

References

Adams, R.P., 2003. Reconciling differences among morphological, chemical, and molecular
data in the taxonomy of Juniperus. Plant Biotechnology Center, Baylor University.
Acta Hort 615, International Society for Horticultural Science, aecessed online Nov.
12/05, www.actahort.org.
Christensen, K.I., and G. H. Dar. 2003. A morphometric study of hybridization between
Pinus mugo and P. sylvestris {Pinaceae). Botanic Garden, University of Copenhagen.
Acta Hort 615, International Society for Horticultural Science, accessed online
Nov. 12/05, www.actahort.org.
Farrar, J.L., 1995. Trees in Canada. Canadian Forest Service Natural Resources Canada,
Canadian Cataloguing in Publication Data, Markham, Ontario.
Forintek Canada Corp., 1994. Properties of Jack Pine. Alberta Department of Economic
Development and Tourism, Edmonton, AB.
Koch, Peter., 1996. Lodgepole Pine in North America, v. 1. Forest Products Society,
Madison, Wisconsin.
Wheeler, N.C., and R.P. Guries, 1987. A quantitative measure of introgression between
lodgepole and jack pines. Can. J. Bot., 65: 1876-1885.

80

Chapter 4: Fibre and Cell Analysis of Jack Pine, Lodgepole Pine, and their Hybrids
4.1 Chapter Objectives
Since processing and manufacturing of wood and wood products requires specific
adjustments for variations in wood quality and therefore species, it is important to know
where these variations occur and which areas of British Columbia are providing which types
of wood.
Interactions between various wood traits or fibre traits are explored in this chapter to
determine where possible relationships occur or from where they may originate. This aids in
comparison between species as it was previously determined in Chapter 3 that each sampling
area varied slightly in age, growth rate, and site index.
The objective of this chapter was to compare each species group based on wood and
fibre traits in order to identify if there were any significant differences between lodgepole
pine and jack pine, and their hybrids. Using the obvious morphological differences between
species and hybrids, samples were classified according to sampling area. Variation in wood
and fibre traits was compared in these same classifications to determine if these traits can be
used as distinguishing features between jack pine, lodgepole pine, and their hybrids.
Another objective of this chapter was to identify hybrid characteristics closely related
to jack pine, and those characteristics that are more closely related to lodgepole pine.
Inferences can be made about the inheritance of the traits and tree lineage within the hybrid
sampling area. By identifying the differences between species groups with regard to wood
and fibre traits, and comparing these results to genetic outcomes of Chapter 2, it was hoped
that the wood and/or fibre characteristics may then be used to predict hybridization.

81

4.2 Wood and Fibre Characteristics
As various wood and fibre characteristics are investigated, it is important to keep in
mind that three main factors contribute to variation in these traits, and that a difference in
characteristic between species can be due to one of these factors, or more probably due to all
three of them in combination. These factors are: 1) the changes in wood cells and fibres as
they age; 2) the genetic influence of the parental trees; and, 3) the environmental influences
that affect the tree growth (Panshin and de Zeeuw, 1980).
Lodgepole pine has varying angles of grain spirality causing a cross-grained effect
among the tracheids (Koch, 1996), but it is generally considered a straight-grained wood.
Sap wood in lodgepole pine is white to yellow in color, heartwood is yellow to brown. The
wood is fairly light weight and somewhat weak in bending and when compressing end-toend. Lodgepole pine is susceptible to decay when conditions are favorable (Alden, 1997).
Similar to lodgepole pine in most characteristics, jack pine possesses a few slight differences
in its wood properties and appearance. Jack pine sapwood is almost white in color, while the
heartwood is light brown or orange. The wood is resinous and moderately light weight with
low shrinkage properties for drying, and has limited durability. It has low bending and
compressive strength. Jack pine, like lodgepole pine, is also susceptible to decay when
conditions are favorable (Alden, 1997).
Lodgepole pine and jack pine woods are characterized by growth rings that show an
abrupt transition from early wood (light in color) to latewood (dark in color). Even though
the transition between growth rings is equally abrupt in both species, a differentiation is
possible in the amounts of both earlywood and latewood present between species. The ratio
of earlywood to latewood can identify a wood as high or low quality for processing or

82

manufacturing depending on the characteristics desired in the produet. Earlywood (EW) and

latewood (LW) possess very different eharacteristics in hard pines, which is why the ratio of
EW:LW can provide important information about a species. In earlywood, the longitudinal
tracheids usually have a larger diameter in the radial direction and thinner cell walls.
However, in latewood, tracheids have thick walls and are almost flattened radially. This
change in cell shape is due to the seasonal growth changes; large earlywood tracheids are
formed during active internodal elongation, when the growth hormone, auxin, is being
produced at a maximum. As auxin production declines, elongation slows down and cell
diameters decrease yielding latewood tracheids. The abrupt transition from earlywood to
latewood characteristic in pines forms a defined tangential band of latewood with thick
flattened cells, and suggests that auxin production weans quickly. This latewood is therefore
much denser than the earlywood; profiles of wood from pith to bark not only show increases
in density as the wood matures, but also show immense variation in density within rings
(Panshin and de Zeeuw, 1980).
Juvenile wood in lodgepole pine is located near the pith and is less prominent than in
other conifers in North America with the transition to mature wood gradual in increasing
tracheid length and decreasing in microfibril angle (Koch, 1996). There are significant cell
and fibre level differences between juvenile wood and mature wood in all species. Juvenile
wood is a column of inferior wood that is formed from prolonged influences of the apical
meristems during wood formation. This portion of the tree is considered to be substandard
from a wood quality point of view because it is prone to extensive warping when dried as
lumber due to large longitudinal shrinkage (Panshin and de Zeeuw, 1980). Juvenile wood
contains shorter fibres that have greater microfibril angles, and smaller tangential cell

83

diameters (Barbour, 2004). Juvenile wood is also less dense in most cases due to relatively
fast growth rates, and contains shorter fibres which confine juvenile wood to specific paper
products. In hard pines, it is characteristic in most cases that density will increase from pith
to bark (Panshin and de Zeeuw, 1980).
Average density for lodgepole pine has been previously calculated as 39 Ibs/cu.ft. or
0.625 g/cm^ for green wood, and 27 Ibs/cu.ft. or 0.43 g/cm^ for oven dried wood (Isenberg,
1951). Koch (1996) stated that average specific gravity based on green volume and oven
dried weight was 0.41 glcrci. Recorded density is 50 Ibs./cu.ft or 0.801 g/cm^ for jack pine
green wood and 29 Ibs./cu.ft. or 0.465 g/cm^ for oven dried wood (Isenberg, 1951), which is
higher than that of lodgepole pine. In general terms, specific gravity is “ .. .the amount of cell
wall substance present in w ood...” (Panshin and de Zeeuw, 1980), and has a direct linear
relationship with density, which is the ratio of the weight of wood due to cell wall content, to
the volume. Therefore, specific gravity and density yield the same information about a
particular sample or species.
Density changes when the cell wall thickness changes or when cell dimensions change.
Photosynthetic input controls the amount of cell wall material that is produced in the tree
(Panshin and de Zeeuw, 1980). Cell wall thickness therefore has a linear relationship to
density and has the potential to distinguish between species provided that site conditions and
age of samples remain constant. As stated, density increases from pith to bark. It is therefore
intuitive that cell wall thickness also increases from pith to bark. This is due to the increase
in diameter of cell walls as the tree matures, with little increase in tracheid length (Panshin
and de Zeeuw, 1980).

84

A wood quality trait is dependent on the end use of the product. Generally, density is
considered to be the most important wood characteristic for identifying quality because it
suggests thick walled, strong cells that will hold together under stresses such as bending or
compression, which are common in building. Modulus of elasticity (MOE) quantifies the
flexibility of wood under compression parallel to the grain. This characteristic can be
estimated based on mierofibril angle (MFA) or tracheid lengths, indicating that significant
relationships exist between MOE, fibre length, and MFA (Panshin and de Zeeuw, 1980).
Lodgepole pine has varying fibre dimensions. Sharma and Potter (2000) determined
fibre dimensions of 3.1 mm x 35-45 um and 23.0 mg/100 m for lodgepole pine fibres, where
Koch (1996) calculated an average fibre length of 2.2 mm. This is a big difference in length,
however, these measurements are dependent on the age of the wood measured, and the site
conditions under which the samples were grown such as latitude, elevation, moisture and
nutrient regime, as well as the measurement method. Jack pine fibre dimensions are recorded
as 3.5 mm x 28-40 um and 18.0 mg/100 m by Sharma and Potter (2000). This fibre
information can be used to distinguish desirable processing traits and correlate them based on
species group. Fibre length is an important characteristic to consider because pulp products
have different fibre requirements. Longer fibres produce increased bonding area between
fibres which yields paper that is less likely to tear under stress (Panshin and de Zeeuw, 1980).
The same strength standards would not be required for tissue as for writing paper. Fibre
coarseness is the most important characteristic to consider for paper production.
Fibre coarseness is defined as the weight of a fibre per unit length (Robertson, 1998).
This characteristic, like fibre length, is variable between species, age classes, and site
conditions. Trees grown on poorer sites, or trees younger in age will generally have finer

85

fibres than others. Lodgepole pine is said to have coarser fibres on average than other
species such as Engelmann spruce, white spruce, or subalpine fir (Pitts, 2001). Jack pine
may have similar fibre coarseness to lodgepole pine, but coarseness will not be identical.
Fibre yield is important to consider from a quality and quantity point of view. The
amount of pulp that can be derived from a species is a key part of processing, but it is also
valuable to explore the quality of the fibres, in order to predict what products the pulp can be
used for. Fibre quantity depends on the density of the wood, FW:LW ratios, the age of the
tree, the health of the tree, and some others. Fibre quality depends on cell characteristics and
the type of cells present (Panshin and de Zeeuw, 1980).
Another important characteristic for wood manufacturing and processing is mierofibril
angle (MFA). Microfibrils are long strands of polysaccharides found in woody plant cell
walls. These are physically oriented at various angles depending on where they are found
within the cell wall, and help to identify the strength of the wood. The angle is measured
from the cell axis, and is referred to as mierofibril angle (MFA). As the cell matures, the
orientation of the microfibrils changes; giving rise to different characteristics for juvenile and
mature wood. In general, juvenile wood has higher mierofibril angles than mature wood; the
decrease in angle as the cells mature is related to the growth of the cell (Panshin and de
Zeeuw, 1980). MFA is directly related to fibre strength. Increases in MFA result in
decreases in strength, thereby decreasing the strength of a paper product. Therefore, like
fibre length, different standards are required for mierofibril angle, depending on the end use
of the product.
Fach of the wood and fibre properties mentioned plays a key role in determining wood
quality, and therefore affects the manufacturing and processing procedures and end products.

86

It is important to consider site conditions and age of samples when testing wood for quality
purposes. By keeping these variables as consistent as possible, species specific wood and
fibre characteristics may be identifiable.

4.3 Experimental Design and Hypotheses
Stem increment cores were obtained at breast height from samples at each site for fibre
property analysis. Cores were analyzed for fibre length and coarseness at different age
classes through Fibre Quality Analysis (FQA); mierofibril angle, basic density, and cell wall
configuration using SilviScan technology.

Two 10 mm cores (bark to bark) were taken

from each tree sampled.
Thirty samples of pure lodgepole pine, 30 of pure jack pine, and 30 potential hybrid
samples were collected from the areas of Prince George, EC, Smoky Lake, AB, and Fort
Nelson, BC, respectively. Each site was chosen for its similarity in vegetation and
moisture/nutrient regimes. Sites for lodgepole pine and jack pine were chosen as to be
similar distances from the hybrid zone.

4.4 Methodology
4.4.1

Fibre Quality Analysis
The Fibre Quality Analyzer (FQA) is an instrument that determines length, shape,

average coarseness, and fibre curl and kink of a selected wood sample. It can also determine
fibre morphology by combining these wood characteristics, and determine any areas of fibre
damage. The FQA utilizes hydrodynamic flow to orient fibres for measurement (Robertson,
et.al., 1998).

87

Using one of the increment cores obtained from each sample tree, wood age 20-40,4060, and 60-80 was prepared for analysis.

Sample Preparation and Protocol
Cores were cooked in deionized water at 120°C for 4 hours, drained, and macerated in a
solution of 50% acetic acid and 50% hydrogen peroxide at 70°C for 48 hours. After this
period, they were rinsed, filtered, and washed into pulp. Pulp was left to air dry for at least
24 hours, and then select samples were oven dried at ~95°C for another 24 hour period to
obtain an average moisture content for the pulp samples. Using the average moisture
content, oven dry weights were calculated for all samples.
To prepare the samples for input into the FQA, 30 mg of each sample was weighed out
and placed into small vials with deionized water. Samples were allowed to sit and reabsorb
water until saturated. Each sample was then broken up into single fibres by vigorous mixing
and further dilution into 4000 g of water. Concentration of fibre weight (mg) over water
weight (g) was then calculated. Based on the concentration calculated, a desired sample
weight (water and fibres) was established using a chart provided. The chart displays sample
weights for each possible concentration calculated based on a target fibre frequency of 10-20
fibres/second when measured in the FQA, and a sample mass of 1.6 mg for the type of
softwoods being analyzed.
After this is complete the sample was then loaded into the FQA for measurement.
Output from the FQA analysis included: a fibre distribution table, fibre frequencies, average
length weighted fibre length, and average coarseness (mg/m).

88

4.4.2 SilviScan
SilviScan technology is designed for fast, non-destructive testing of wood.
The instrument uses optical microscopy, x-ray diffractometry, x-ray densitometry, image
analysis, applied mathematics and analysis of large data sets, to output wood and fibre
properties. SilviScan can be used to determine variation in wood chemistry and anatomy for
more effective and efficient processing (Paprican press release, 2004).
Using the remaining whole core sampled from select trees, wood was non-destructively
analyzed by SilviScan. Fifteen samples displaying average wood properties for each site
were selected for this analysis.
Sample Preparation
Cores were selected based on site, in order to get equal representation from all samples
locations, and samples were also selected for average representation of fibre traits or
demonstration of fibre trait range. Five cores of each species (jack pine, lodgepole pine, and
hybrid) were chosen for SilviScan.
Cores were debarked and labeled. Samples were submerged in Ethanol for 96 hours
and then air-dried for 24 hours. Cores were shipped to CSIRO in Australia for further
preparation and analysis. Further preparation involved cutting the cores into 2 mm x 7 mm
strips to form samples that are compatible for use in the x-ray diffractometer, x-ray
densitometer, and image analysis systems.

4.4.3 Electron Microscopy
Scanning electron microscopy (SEM) involves scanning an electron beam across a
specimen to produce a digital image of that specimen, in this case wood samples. Electrons

89

are collected from the beam’s interaction with the sample, X and Y position data is produced,
and contrast and brightness information is used to create the image (Goldstein, 1992).

Sample Preparation
Two samples of each species (jack pine, lodgepole pine and hybrids) were sanded to
smooth surfaces for imaging. A razor blade was used to slice a straight cross section surface
from each core. Samples were soaked in Ethanol to remove water and air dried before
imaging.

4.5 Results
4.5.1 Solid Wood Analysis

Avg. calculated density (g/cm3)
Standard Errors for Density
Avg. moisture content (% of OD)
Standard Errors for MC
Site index
Number of samples

Jack Pine
0.41
0.006897
73.58
2.838508
15
30

Lodgepole Pine
0.43
0.008577577
60.17
2.378054
16.1
30

Hybrid Pine
0.4
0.00447093
55.25
2.571498
12
30

Table 11 shows the average density, average moisture content and site index values for
the three species groups. On average densities are comparable, lodgepole pine seems to be
slightly higher than the other species. Jack pine has the highest average moisture content,
and hybrids have the lowest. Site indices for the sampling areas are highest for lodgepole
pine and lowest for hybrid pine.

90

600
CO
E 500
g 400
300
c/3 200
C
CD
"O 100

0

11 11 1 11 1 111 1 1I MI 11 111 n rn

Th
lO
lO

! M

M

I 1 I M

I II

II

M

II

II

I 1 I I II I !

I I I I

I I II

II

II

T T T t

CD T- N CO
CD O CD CO CD 00 CD CO CD
O) CJ) CO CO CD 00 O O CO ID 00 CD O
N
00
N
N
ID ID ID CD CD CD CD
position from pith to bark
sam ple 5 - a — s a m p le s - û —sam ple 30

Figure 27 a: Jack pine mature wood, early wood density profile moving from inner wood to bark.

CO 600
E 500
400
300
(/} 200
c
CD 100
■a
0
' ^ ' « d - l D ' ^ ' ^ C D O l D r ^ O C M O C M C M C D
l D C O O O O C \ I C D C O C D T - r ^ T - l D C D C D C O

I DCDNMCDCDMMOOOOCD( DCDT- C\ I
position pith to bark
sam ple 33

sam ple 35

sam ple 39

-A— sam p le 4 8 -4*— sam ple 60
Figure 27b: Lodgepole pine mature wood, earlywood density profile moving from inner wood to bark.

91

600
E 500
TO 4 0 0
^ 300

200
-g 100

0

II

II i l l

II II I I I

I I I I I I ............................

1 l l

ii ' i i

I I I

II II

I I I

II

!i

III

II i M i i

II

i M i M i

II I I I

II

M

n n i

i r r i

' ^ C n - ^ ' ^ O C D ' ^ O L O O î l O ' ^ O J - i - C D

■^mt^a>T-c\j'^cûN.cx30c\]cocDco

LOl OLOl OCDCOCOCDCOCOI^N-N-h^Nposition pith to bark
- A — s a m p ie 66

- « —s a m p le 67

Figure 27c: Hybrid pine mature wood, earlywood density profile moving from inner wood to bark.

Figure 27a, b, and c show that lodgepole pine has the highest density of the three
species groups, when considering mature and earlywood distributions. Mature wood and
earlywood were separated from the core for consideration because variation exists within the
wood, and considering the whole core at once can give skewed averages; juvenile wood and
earlywood/latewood ratios have large influences on the overall density of the core. This
creates profiles which can be directly compared to each other, irrespective of
earlywood/latewood ratio contribution or amount of juvenile wood. Mature wood was
considered to be wood over the age of 60 to ensure no juvenile wood component, and
earlywood was considered to be wood with a density less than 400 kg/m^; this density was
chosen as the division between the two wood types based on the wood density curves
obtained from the samples via SilviScan.

92

0 .6 0 0 0

0 .5 5 0 0

0 .5 0 0 0
CO

E

ü

03

0 .4 5 0 0 -

I
0 .4 0 0 0 -

0 .3 5 0 0

0 .3 0 0 0
0.1

0 .2

0 .3

0 .4

0 .5

0 .6

0 .7

0 .8

DBH growth rate (cm/yr)

A Jack pine ■ Lodgepole pine a Hybrid pine
Figure 28a: Calculated density and D BH growth rate based on ring count age, measured DBH, measured
volume, and measured weight o f cores.

93

500

450

400

2
g 350

K

—
CO
LU

CD 300

i
o 250

O
)
c
(D
Q 200

-

150

100
0.15

0.2

0.25

0.3

0.35

0.4

0.45

0.5

0.6

0.55

DBH Growth rate cm/yr
A

Pj

■ Pli

A Px

Figure 28b: D ensity o f samples measured by SilviScan that were over 60 years o f age plotted against DBH
growth rate. Only mature (w ood over 60 years o f age) earlywood measurements are displayed.

Density was calculated in two ways for the samples collected. All cores collected were
measured for volume and weight, which yielded a “calculated” density value. Selected
samples that were sent for measurement via SilviScan also returned with a density value, and
were used as a comparison to the calculated values obtained. Figure 28a shows the
calculated densities, and Figure 28b shows the SilviScan densities of sample groups as they
may be related to growth rate. Lodgepole pine and hybrid pine have similar rates of growth
according to these figures, but the hybrid sample densities are lower. Jack pine samples have
faster growth rates than both lodgepole pine and hybrid samples, but a lower density than
lodgepole pine samples. The density values obtained from each method are similar.

94

Table 12: Pairwise Comparisons o f Calculated density and moisture content in all 5 sample groups. Species 1
= jack pine, species 2 = lodgepole pine, species 3 = hybrids, species 4 = hybrids from jack pine sampling area,
species 5 = lodgepole pines from hybrid sampling area

Dependent
Variable

(1)
Species

(J)

MC%
(OD-GR)

1

2

2

Species

22.5350

.000

14.392

30.677

4

-2.959

4.093
7.347

11.657

8.697

.688
.092

-17.575

5

-1.458

18.852

1

-12.9940
9.5410
-15.9530
-4.297

4
5
1

1

1

.001

-20.462

-5.525

3.994
7.293
5.026

.019
.032
.395

1.595
-30.460
-14.296

17.487
-1.445
5.702

-22.5350

4.093

.000

-30.677

-14.392

-9.5410
-25.4930

3.994

.019

-17.487

-1.595

7.473

.001

-40.359

-10.628

-13.8360
2.959
15.9530
25.4930
11.656

5.284
7.347
7.293

.011
.688
.032

-24.349
-11.657
1.445

-3.326
17.575
30.460

7.473

10.628

40.359

8.072

.001
.153

-4.401
-18.852
-5.702
3.326
-27.713

27.713

.092
.395
.011
.153

1.458
14.296
24.349
4.401

2
3
4

13.8380
-11.656

5.105
5.026
5.284
8.072

-.016

.010

.116

-.035

.004

3

.004

.696
.285

.026

.021

.011
.019

-.017

4
5

.0300
.016
.020
.037

.013
.010
.011
.019

.028
.116
.062
.060

-.018
.003
-.004
-.001
-.002

.0460
-.004

.013

.001

.019

.059
.057
.035
.041
.075
.072

.011

-.026

.017

.011
.020
.014
.019

-.041
-.023
-.002
-.059

.001
.056
.054
.018

2

-.020
.017
.026
-.021
-.037

.696
.062
.401
.066
.285

.019

.060

-.075

.002

3

-.017

.020

.401

-.056

.023

5

.009

.021

.663

-.033

.052

2

1
3
4

1
2
4

4

5.105
3.754

-8.697
4.297

5
3

Upper Bound

3

2

2

Lower Bound

20.462

3
5

1

95% Confidence Interval for
Difference(a)

5.525

5

Density
(g/cmS)

Slp.(a)

.001

4

5

Std.
Error

3.754

2

4

(l-J)

12.9940

3

3

Mean
Difference

5
1

95

Dependent
Variable

(1)
Species

(J)
Species

5

1

Mean
Difference
(l-J)

Std.
Error

.013
.013
.014

-.030(1
-.046(1
-.026

2
3

Sip.(a)

.028
.001
.066

95% Confidence Interval for
Difference(a)
Lower Bound

Upper Bound

-.057
-.072
-.054

-.003
-.019
.002

4
-.009
.021
.663
-.052
.033
Based on estimated marginal means
* Ttie mean difference is significant at tfie .05 level.
a Adjustment for multiple comparisons: Least Significant Difference (equivalent to no adjustments).

Cluster Analysis:
Table 13a: Final Cluster Centers for moisture content and density

Cluster
1
67.57130

2

3

MC % (OD-GR)

48.88333

91.63527

Density (g/cm3)

.41507

.42588

.39124

Table 13b: Distances between Final Cluster Centers

1

Cluster
1
2

18.688
24.064

3

2

3

18.688

24.064
42.752

42.752

Table 13c: Number o f Cases in each Cluster

Cluster

1
2
3

Valid
Missing

38.000
37.000
12.000
87.000
3.000

Jack pine samples primarily fall into cluster 1, with a small portion of samples in
clusters 2 and 3. Approximately half of the lodgepole pine samples fall into cluster 1 as well,
and the other half fall into cluster 2 with the exception of samples 44 and 57 that fall into
cluster 3. The majority of hybrid samples fall into cluster 2, a portion of the samples fall into

96

cluster 1. One hybrid sample, 85, falls into cluster 3. Further statistical output is displayed
in Appendix 3.

1.8 n
1.6
1.4 1.2 §

_J

§

1 -

LU

g

0 .8

cr
0 .6 0.4
0 .2 0 -

n-------- 1-------- 1--------1-------- 1-------- 1-------- 1-------- 1-------- 1--------1-------- 1--------1-------- 1-------- 1----

A Jack pine ■ Lodgepole pine a Hybrid pine
Figure 29: Ratios o f earlywood to latewood in each sample. Both Juvenile wood and mature wood were used
for the ratio.

SilviSean average EW:LW ratio values are displayed in Figure 29. In general,
lodgepole pine and jack pine both have lower EW:LW ratios than hybrid pine, meaning that
latewood and earlywood ratios are closer to 1. There is almost 2 times the amount of
earlywood to latewood in hybrid samples. This is probably due to site variation and elimate.

97

Figure 30a: Earlywood - latewood transition as shown in potential hybrid pine sample #64.

Note the variability in latewood thickness depending on growth influences of site or growing
season in Figure 30a.

Figure 30b: Jack pine sample cross section with longitudinal view o f trachieds.

An example of cell wall thickness and cell dimensions can be easily viewed in Figure 30b.

98

s

2.91
2 .8 '
2 .7 '
2 .6 '
o
'rl 2 .5 '
H 2.4
2 .3 '
2 .2 '
13 2.1 '
U 2.U'^
0
3

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

1.8

EW: LW ratio
Ajack pine

■ lodgepole pine a hybrid pine

Figure 31a: The relationship between earlywood/latewood contributions and cell wall thickness.

Cell wall thickness and earlywood : latewood ratios have a linear relationship. In
Figure 31a, as the percentage of earlywood increases in the sample, the average cell wall
thickness decreases. Lodgepole pine samples have the lowest percentage of earlywood, and
the thickest average overall cell wall size. Jack pine samples have variable ratios and wall
thicknesses, while hybrid samples have samples with thinner walls and more earlywood.
Site conditions, specifically latitude and length of growing season are correlated to growth
rate and therefore cell wall thickness and EW:LW ratios.

99

I

28.0
27.5
27.0

Q
13

26.5

1

26.0

▲

A

25.5
(ü

I

2 5 .0 —
24.0

25.0

26.0

27.0

28.0

29.0

30.0

31.0

32.0

33.0

Average Radial Diameter (um)
^ Jack Pine ■ Lodgepole pine^ Hybrid Pine

Figure 31b: Average radial and tangential diameter measurements for samples measured by SilviScan.

Figure 31b shows the average dimensions of the cell walls for 5 samples of each
species group. Jack pine samples have a larger radial diameter but smaller tangential
diameters. Lodgepole pine shows that samples on average have larger tangential diameters
than other samples, but smaller radial diameters. Hybrid samples show average diameters
scattered in between those of the pure species.

100

1525 1535 1545 1555 1565 1575 1585 1595 1605 1615 1625 1635 1645 1655 1665 1675 1685 1695 1705 1715 1725
position number

-8

o— 13 —3K— 15 —o— 30 —#— 33 —A— 35 ■

-39 - * - 4 8

-60

-6 6

-9 0

67 ^ * — 80 - » — 88

Figure 32: Wall thickness profile for mature wood m oving from inner wood to bark. Samples 5-30 are jack
pine, 33-60 are lodgepole pine, and 66-90 are hybrid pine.

Figure 32 shows a profile of cell wall thickness for each species. Jack pine and
lodgepole pine samples display a greater variation in wall thickness based on this profile, as
compared to hybrid samples. It seems that cell walls of mature wood in hybrids may be more
consistently thickened than those of pure species.

101

900 -

400 -

1.4

1.9

2.4

2.9

3.4

3.9

4.4

4.9

5.4

Cell Wall Thickness (um)
o 5 □ 8 o 13 A 15 X 30 • 33 A 35 ■ 39 —48 ♦ 60 ■ 66 A 67 ♦ 80 X 88 • 90
Figure 33: Density as a function o f cell wall thickness. Samples 5, 8, 13, 15, and 30 = jack pine; 33, 35, 39, 48,
60 = lodgepole pine; 66, 67, 80, 88, 90 = hybrids

Little difference exists between overall wall thickness profiles with respect to
earlywood and latewood densities. Figure 33 shows all cell wall thickness profiles as they are
related to density, for comparison between species. The only differences that seem to exist
between species are the concentration of samples that are less-dense with thinner walls, as
compared to the concentration of samples that are denser with thicker walls. Lodgepole pine
has more cells with denser, thicker walls than the other two species. It should be noted that
there is a very strong linear relationship between density and cell wall thickness. As cell wall
thickness increases, the density of the sample increases.

102

7
6
I

5

% 4
<
CD
o 3
CO 2
1

0
-1

0

10

15

20

25

30

MOE (G Pa)
* sam ple 5 « 8 & 13 * 15 * 30 » 33

a 35

" 39 - 48 ♦ 60 « 66 & 67 x 80 « 88 * 90

Figure 34: Wall thickness as a function o f Modulus o f elasticity. A ll wood (juvenile and mature) included in
profile. Samples 5, 8, 13, 15, 30 = jack pine; samples 33, 35, 39, 48, 60 = lodgepole pine; samples 66, 67, 80,
88, 90 = hybrid pine

Modulus of elasticity is measured in Figure 34 as a function of wall thickness.
Lodgepole pine samples have a greater MOE than jack pine, and hybrid pines have a
moderate MOE, in between that of the pure species. However, there does not appear to be
any variation due to the wall thickness, therefore it can be assumed that there is no prominent
relationship between MOE and cell wall thickness for these samples.

103

30
1c 25
CL

0

^ 20
ü

1 15
LU

0
1

3
-O
O

10

5

0
54

254 454 654 854 1054 1254 1454 1654 1854 2054 2254 2454 2654 2854 3054
Position of measurement (pith to bark)

■33

66

-"-8
-1— 35
—* —07

—“12

-# -1 5

— 39
H ' 80

----- 48

— ^

-*-3 0
-♦ -6 0
—#—90

Figure 35: Modulus o f Elasticity (MOE) profile from pith to bark for all samples. Mature wood component
begins after approximately position 1500.

Figure 35 shows the profile of MOE moving from pith to bark. It is visible that there
are some main differences between each species as MOE changes over the growth of the
trees. Jack pine samples have a slow but general increase in MOE as the measurements
move from pith to bark. Lodgepole pine samples start at a low MOE in the juvenile wood,
then increase to a high MOE just before maturity (60 years or -position 1500), and slowly
decrease or level off in MOE as age increases past maturity. Hybrid pine seems to be
between these two patterns, where some samples have an increase around the same period as
lodgepole pine but some maintain a steady increase like jack pine samples. In both cases the
MOE values for hybrid are intermediate of the pure species values.

104

1200

1000
_

800

I

600

w

400

<D
“

200

0 i
0(
-200 J

10

15

20

25

30

MOE (GPa)
♦ sam ple 5 " 8

& 13 x 15 x 30 « 33 + 3 5

- 39 - 48 ♦BO » 66 ♦ 67 x 80 * 88 • 90

Figure 36: Density as a function o f MOE for all samples. A ll wood present in profile
(mature and juvenile).

Figure 36 shows density as a function of MOE. There is no discernable trend between
MOE and density according to this figure, however, it can be noted, that lodgepole pine
occupies the upper range in MOE value, while jack pine occupies the lower range of MOE
value. Hybrid samples measured MOE values fall in between the pure species
measurements.

105

0.6 1
OCXD CD

0.5 -

A /A

A A A

E
0.4 P
2

<X>

A

0.3
5
P
CD 0.2 T
CO
n
0.1

OCDOO

C

A AAA AAJAA

0 10

20

15

25

MOE (GPa)
o sample 5

□8

X 13

A

15

o30

—33

■ 35

A 39

A 48

A 60

A 66

A 67

x80

# 88

m 90

Figure 37: MOE and growth rate (DBH/age) for mature wood MOE measurements.

Figure 37 shows that there is a possible relationship between MOE and growth rate of
samples. The samples with slower growth rates seem to have higher MOE values. The jack
pine samples have faster rates of growth, and have lower MOE values. Lodgepole pine
samples were grown more slowly, and are shown to have greater MOE values on average
(Figure 37). Flybrid pine samples also have lower rates of growth, and show intermediate
MOE values. This suggests that at the same growth rate, MOE is higher in lodgepole pine
than in hybrids. Only mature wood MOE measurements were used in this figure, so juvenile
wood should not have influenced the results, however, there are differences in site conditions
which may have played a role in the variability of MOE.

106

Table 14: Pairwise Comparisons for wood characteristics. Species 1 = jack pine, species 2 = lodgepole pine,
species 3 = hybrids, species 4 = hybrids from jack pine sampling area, species 5 = lodgepole pines from hybrid
sampling area
Mean
Difference
(l-J)

Std. Error

Sip.(a)

2
3

-8.655(*)
-4.182(3

1.415
1.634

.000
.027

Lower Bound
-11.769
-7.778

5

-4.502(3

1.872

.035

-8.621

-.383

1

8.655(3

1.415

.000

5.541

11.769

3
5
1

4.473(3
4.153(3
4.182(3

1.634
1.872
1.634

.019
.048
.027

.877
.034
.587

8.069
8.273
7.778

2

.019

-8.069

-.877

Dependent
Variable

(1)
S pecies

(J)
Species

MOE(ave)

1

2

3

5

1

2

3

5

Wall(ave)

-4.473(3

1.634

-.320

2.042

.878

-4.814

4.175

1

4 .502(3
-4.153(3
.320

1.872
1.872
2.042

.035
.048
.878

.383
-8.273
-4.175

8.621
-.034
4.814

2

.152

.242

.543

-.380

.684

3

-.456

.279

.130

-1.071

.158

5
1
3
5

-.384
-.152
-.608
-.536

.320
.242
.279
.320

.255
.543
.052
.122

-1.088
-.684
-1.222
-1.239

.320
.380
.006
.168

.279

.130

-.158

1.071

1

.456

2

.608

.279

.052

-.006

1.222

5

.073
.384
.536

.349
.320
.320

.839
.255
.122

-.695
-.320
-.168

.840
1.088
1.239

1
2
3

-.073

.349

.839

-.840

.695

2

-.272(3

.084

.008

-.457

-.088

3
5
1

.107
.081
.272(3

.097
.111
.084

.290
.482
.008

-.105
-.163
.088

.320
.324
.457

3

.380(3

.097

.002

.167

.593

5

.353(3

.111

.009

.109

.597

3

1

5

2
5
1

-.107
-.380(3
-.027
-.081

.097
.097
.121
.111

.290
.002
.829
.482

-.320
-.593
-.293
-.324

.105
-.167
.239
.163

2

-.353(3

.111

.009

-.597

-.109

3

.027

.121

.829

-.239

.293

-98.715
-43.472
-39.870

-11.901
56.772
74.974

11.901

98.715

1

2

mw dens

Upper Bound
-5.541
-.587

5
2
3
ew/lw ratio

95% Confidence Interval for
Difference(a)

1

2

2
3
5

-55.308(3
6.650
17.552

19.722
22.773
26.089

.017
.776
.515

1

55.308(3

19.722

.017

107

Dependent
Variable

(1)
Species

3

5

Mean
Difference

(J)

(l-J)

Species

Std. Error

3

61 . 958(1

22.773

5
1
2

72.860(1
-6.650
-61.958(1

26.089
22.773
22.773

5
1

10.902

28.466

-17.552

26.089

Sig.(a)

95% Confidence Interval for
Difference(a)
Lower Bound

Upper Bound

.020

11.836

112.080

.018
.776
.020
.709

15.438
-56.772
-112.080
-51.751

130.282
43.472
-11.836
73.554

.515

-74.974

39.870

2

-72.860(1
26.089
.018
-130.282
-15.438
-10.902
.709
-73.554
3
28.466
51.751
Based on estimated marginal means
* Ttie mean difference is significant at ttie .05 level.
a Adjustment for multiple comparisons: Least Significant Difference (equivalent to no adjustments).

Cluster Analysis:
Table 15a: Final Cluster Centers

Cluster
1
ew/lw ratio

.81309
503.09882

mw dens
MOE(ave)

11.00200

Wall(ave)

2.44608

2
.41212

3
1.12401

564.84911
16.69203
2.78952

454.27874
11.55344
2.34982

Table 15b: Distances between Final Cluster Centers

Cluster
1

1

2

62.014

3

48.824

2
62.014

3
48.824
110.693

110.693

Table 15c: Number o f samples per cluster

Cluster

Valid
Missing

1

8.000

2

3.000

3

4.000
15.000
75.000

The cluster analysis revealed that 4 out of 5 jack pine samples fell into cluster 1. Three
out of 5 lodgepole pine samples fell into cluster 2 and the other 2 fell into cluster 1. Two out

108

of 3 hybrid samples fell into cluster 3 and the other 1 fell into cluster 1. The two samples
representing species group five, which could be hybrids or lodgepole pine samples, had one
sample in cluster 1 and one in cluster 3.
Further statistical output can be found in Appendix 3.

4.5.2 Fibre Analysis:

2.9

2.7

2.5

c 2.3

LL

0

10

30

20

40

50

60

70

80

90

Wood - Pith to Bark
X

Jack

—

Hybrid

-

Lodgepole

4 Hybrid

-------Lodgepole Pine

- - Jack pine

Figure 38: Lengtti weiglited fibre length from pith to bark in all species/sample groups with observed trend
lines.

Figure 38 shows length weighted fibre length as samples move from 0-20 age class (020 rings from the pith) to 60-80 age class (60-80 rings from the pith). Curves for the three
species groups are similar, increasing in fibre length in juvenile wood, and forming an
asymptote as wood becomes more mature. Throughout the age classes, lodgepole pine has
the longest fibres on average, and hybrid fibres are the shortest.

109

140

120
100
c

3
O
O 80
2
& 60

I

40

20

0
0

1

2

3

4

5

7

6

Fibre Length (mm)
■hybrid

lodgepoie

-jack

Figure 39: Average number o f fibres per fibre length interval for wood rings aged 40-60

Fibre length distributions were calculated from analysis with the FQA. Intervals of
fibre lengths were established, starting with very short fibres, shown on the left of Figure
39, and working up to the longest fibres found in the samples on the right. This figure
displays the number of fibres found within each one of these length intervals. The
longest fibres found were around 6 mm in length. The shortest fibres found were 0.05
mm in length. All three species groups have an increase in number of fibres between 2
and 3 mm in length, after which length the fibres beeome fewer and fewer. Jaek pine
samples show more fibres at shorter lengths than lodgepole pine. Hybrid fibres follow a
similar curve to jack pine fibres.

110

0.3 1
0.28
0.26
♦

0.24

A A

A
A*

A

«

^
* A ^ ^ A AA* ^4^^^
lA^A
A
\
A $A A ^
A
4k
AAA^<
O A

^0.22
05

% 0-2 4
c

05
05

0.18 ce
O
ü
0.16

0.14
0.12

0.1

1.5

1.7

1.9

2.1

2.3

2.5

2.7

2.9

3.1

3.3

Length Weighted Fibre Length (mm)
A Jack Pine (40-60)

o Jack Fine (60-80)

A Lodgepole Fine (60-80) A Hybrid Fine (40-60)

A Lodgepole Fine (40-60)
A Hybrid Fine (60-80)

Figure 40: Fibre coarseness as a function o f fibre length. Approximate groupings can be considered as
outlined by ovals. Measurements by FQA.

Figure 40 shows the relationship between coarseness and length weighted fibre
length for all samples in mature wood groups 40-60 years, and 60-80 years. There is
definite variability within sample groups, however a general trend is evident in the
majority of samples. Jack pine samples are, in general, lower in coarseness and fibre
length values than lodgepole pine samples. Hybrid samples seem to be intermediate in
both coarseness and fibre length according to this distribution.

Ill

290
270
250

O) 2 3 0
=3

œ
o 210
ë
m
(ô 190
o
ü
170
150
130
0

20

40

60

80

Pith to Bark Ring Count
♦
- -

Jack
Jack pine

■
—

Lodgepole
-Hybrid

&

Hybrid

------- Lodgepole Pine

Figure 41: Coarseness profile from FQA measurements m oving from pith to bark with observed trend
lines.

A coarseness profile moving from pith to bark is displayed in Figure 41. Based on
the trend lines established, lodgepole pine has higher coarseness values as the tree
matures than those of jack pine or hybrid pine. The coarseness values for jack pine and
hybrids remain relatively unehanged moving from pith to bark (juvenile to mature wood)
in these samples. By the 60-80 year ring count, the jack pine and hybrid eoarseness
values seem to be much lower on average than the lodgepole pine coarseness values.

112

780

680

. A/56 ^ \
^

E 580

05
3
<n

» ja

1 480
05

2

CO

O 380

Ta>KLo*
g . *« *

280

180
180

380

580

780

980

1180

Density (kg/m3)
o sample 5 □ 8 x 13 ° 15 a 30 » 33 + 3 5

- 39 - 48 * 60 * 66 * 67 > 80 * 88 « 90

Figure 42: Fibre coarseness as a function o f density for mature wood only.

Figure 42 shows that with increasing fibre coarseness, density increases. All
samples have a concentration of values at lower coarseness and density, while
distributions become increasingly variable as coarseness and density get higher. Jack
pine has the most variability within samples, while lodgepole pine has the least
variability. At lower values of coarseness and density, hybrid pine samples follow a
lodgepole pine pattern, and jack pine samples are slightly different. Lodgepole pine
samples have the highest density and coarseness values among the three species groups.

113

Table 16: Pairwise Comparisons o f fibre length and coarseness for wood age 40-80. Species 1 = jack pine,
species 2 = lodgepole pine, species 3 = hybrids, species 4 = hybrids from jack pine sampling area, species 5
= lodgepole pines from hybrid sampling area

Dependent
Variable

(1)
sp ecies

(J)
sp ecies

Mean
Difference
(l-J)

Std.
Error

Sig.(a)

95% Confidence
Interval for
Difference(a)
Lower
Bound

Fibre
length (4060)

1

2

3
4
2

3

4

1

2

3

4

Fibre
coarsenes
s (40-60)

1

.072

.848

.099

.082

.231

-.065

.264

.103

-.072

.750

.339

1

.014

.072

.848

-.132

.160

3
4
1

.113
.353
-.099

.066
.198
.082

.093
.082
.231

-.020
-.046
-.264

.246
.752
.065

2

-.113

.066

.093

-.246

.020

4

.240

.202

.241

-.167

.646

1

-.339
-.353
-.240

.204
.198
.202

.103
.082
.241

-.750
-.752
-.646

.072
.046
.167

.004

.074

.957

-.145

.153

3

.114

.084

.181

-.055

.282

4
1
3
4

.047
-.004
.110
.043

.209
.074
.068
.203

.825
.957
.111
.835

-.373
-.153
-.026
-.365

.466
.145
.246
.450

1

2

-.114

.084

.181

-.282

.055

2

-.110

.068

.111

-.246

.026

4
1
2

-.067
-.047
-.043

.206
.209
.203

.746
.825
.835

-.483
-.466
-.450

.348
.373
.365

3

.067

.206

.746

-.348

.483

-.017(3

.007

.021

-.032

-.003

4
1

.001
.036
.017(3

.008
.021
.007

.913
.087
.021

-.016
-.005
.003

.018
.078
.032

2

3
2

-.160

.132

-.014

.204

2
3
Fibre
length (6080)

Upper
Bound

3

.018(3

.007

.008

.005

.032

4

.054(3

.020

.010

.013

.094

3

1

4

2
4
1

-.001
-.018(3
.035
-.036

.008
.007
.020
.021

.913
.008
.092
.087

-.018
-.032
-.006
-.078

.016
-.005
.076
.005

2

-.054(3

.020

.010

-.094

-.013

3

-.035

.020

.092

-.076

.006

114

Dependent
Variable

(1)
species

(J)
species

Mean
Difference

Std.
Error

(l-J)

Sig.(a)

95% Confidence
Interval for
Difference(a)
Lower
Bound

Fibre
coarsenes
s (60-80)

1

2

2
-.011

.008

.139

-.027

.004

3
4

.013
.017

.009
.021

.131
.415

-.004
-.025

.030
.060

1

.011

.008

.139

-.004

.027

.001

.011

.038

1
2

.024(3
.029
-.013
-.024(3

.007
.021
.009
.007

.169
.131
.001

-.013
-.030
-.038

.070
.004
-.011

4

.004

.021

.836

-.038

.047

1

-.017

.021

.415

2

-.029
-.004

.021
.021

.169
.836

-.060
-.070
-.047

.025
.013
.038

3
4
3

4

Upper
Bound

3
Based on estimated marginal means
* The mean difference is significant at the .05 level.
a Adjustment for multiple comparisons: Least Significant Difference (equivalent to no adjustments).

Table 17: Pairwise Comparisons for fibre length and coarseness o f wood age 20-40. Species 1 = jack pine,
species 2 = lodgepole pine, species 3 = hybrids, species 4 = hybrids from jack pine sampling area, species 5
= lodgepole pines from hybrid sampling area

Dependent
Variable

(1)
species

(J)

species

Mean
Difference (l-J)

Std. Error

Sig.(a)

95% Confidence Interval
for Difference(a)
Lower
Bound

Fibre
length (2040)

coarsenes
s (20-40)

1

Upper
Bound

2
-.062

.110

.585

-.304

.180

3

.036

.127

.782

-.244

.316

5

.161

.146

.293

-.160

.482

2

1

.062

.110

3

3
5
1

.098
.223
-.036

.127
.146
.127

-.180
-.182
-.098
-.316

.304
.378
.544
.244

2

-.098

.127

.585
.457
.154
.782
.457

-.378

.182

5

.125

.159

.448

.475

5

1
2
3

-.161
-.223
-.125

.146
.146
.159

.293
.154
.448

-.225
-.482
-.544
-.475

1

2

-.009

.012

.478

-.036

.018

.003
-.024
.009
.012
-.014

.014

.863

-.029

.034

.016
.012
.014
.016

.178
.478
.433
.397

-.060
-.018
-.020
-.051

.013
.036
.043
.022

3
5
2

1
3
5

.160
.098
.225

115

Dependent
Variable

(1)
species

(J)
species

Mean
Difference (l-J)

Std. Error

Sig.(a)

95% Confidence Interval
for Difference(a)
Lower
Bound

3

5

Upper
Bound

1

-.003

.014

.863

-.034

.029

2

-.012

.014

.433

-.043

.020

5

-.026
.024
.014

.018
.016
.016

.172
.178
.397

-.065
-.013
-.022

.013
.060
.051
.065

1
2

3
.026
.018
.172
-.013
Based on estimated marginal means
a Adjustment for multiple comparisons: Least Significant Difference (equivalent to no adjustment

116

Cluster Analysis:
Table 18a: Final Cluster Centers for fibre length and eoarseness at 40-60 and 60-80 yrs.

Cluster
1
Fibre length (40-60)
Fibre length (60-80)

2

3

2.51

2.36

2.82

2.59

2.32

2.79

Fibre coarseness (40-60)

.198

.189

.214

Fibre coarseness (60-80)

.20

.19

.22

Table 18b: Distances between Final Cluster Centers

Cluster
1

1

2

3
.313

2

.313

3

.359

.359
.652

.652

Table 18c: Number o f Cases in each Cluster

Cluster

1

24.000

2

8.000

3
Valid
Missing

18.000
50.000
43.000

A small majority of the jack pine samples fell into cluster 1, and a small majority of
the lodgepole pine samples fell into cluster 3. A large majority of the hybrid samples fell
into cluster 1. Cluster 2 is composed of a few samples from each species group. Further
statistical data can be found in Appendix 3.

117

60

%
g, 50
CD

S 40

m

c 30

<
Ë 20

1 10
0 ^
■ « ^ • C O C M T - O O O O t ^ C O L n - ^ C O C M ' i - O C D O O h ' C D i n - ^ C O

m O) CO N
T-

1- "4- C X J C V J C O O ' ^ O O C M C D O C O t ^ i - L O O C O t ^
1— 1— T— -1— 1— -1— T— C\] C\J C\J C\j C\J C\J C\J C\J

co TT (O M. o o o - ! - c o ' ^ m h ^ c D O - > - c M ' ^ L n c D o o a >

Pith to Bark Position Measurement
•sample 5

-8

-*-1 3

15

30

■33

■35

■48

■60

66

67

— 39
— 80

88

90

43: Microfibril angle profile m oving from pith to bark.

Figure 43 displays a profile of the microfibril angle moving from pith to bark in
select samples. Jack pine samples (samples 5, 8, 13, 15, 30) show higher microfibril
angles than both lodgepole pine (samples 33, 35, 39,48, 60) and hybrids (samples 66, 67,
80, 88, 90) throughout the entire core of most of the samples. Lodgepole pine has the
lowest overall microfibril angle, while hybrid samples contain more variability but
parallel lodgepole pine samples over jack pine samples.

118

40.0 n
Ï

35.0 -

o) 30.0

0)

S 25.0 ^

A
A

20.0 -

® 15.0

g)
2

10.0 -

<

5.0 -

m
A

*
■

0 . 0

I

5, 33, 66

I

8, 35, 67

I

13, 39, 80

I

15, 48, 88

I

30, 60, 90

Sample Number
A Jack pine

■ Lodgepole pine

▲Hybrid pine

Figure 44: Average microfibril angle for select samples representing sampling areas. Pj - sample 5, 8, 13,
15, 30; Pli - sample 33, 35, 39, 48, 60; Px - sample 66, 67, 80, 88, 90.

Figure 44 shows microfibril angles averaged over the whole core of each sample,
for all three species groups. Both juvenile wood and mature wood are included in the
average, so no differentiation can be made between angles of the different wood types.
However, it can be observed that jack pine has an overall higher microfibril angle than
lodgepole or hybrid pine in all the samples measured. This also coordinates with the
results gathered from Figure 43, which displayed that jack pine had a higher microfibril
angle in mature wood as well as juvenile wood. Some samples may contain more
juvenile wood than others, which would make the average microfibril angle higher for
samples in Figure 44.

119

0.6

□□ n o ^

0.5

OOOCO

0^4D

CO

8) 0.4

<X >

'Æ DODaDO

X

m# ^

g 0.3 $

XX
A

O

A

I

O 0.1

10

20
30
40
Microfibril Angle (degrees)

50

60

X5 o 8 A 13 O 15 □ 30 • 33 - 35 ■ 39 A 48 ♦ 60 ■ 66 A 67 + 80 * 88 • 90
Figure 45: D BH growth rate as a function o f microfibril angle. Samples 5, 8, 13, 15, 30 = jack pine;
samples 33, 35, 39, 48, 60 = lodgepole pine; samples 66, 67, 80, 88, 90 = hybrid pine.

Figure 45 displays DBH growth rate as a function of MFA. Jack pine samples have
the highest growth rate and highest MFA according to this figure, while lodgepole pine
has a lower MFA and growth rate. Hybrid pine has the lowest representation of growth
rate, and microfibril angles comparable to those of lodgepole pine. This figure suggests
that microfibril angle may be influenced by growth rate for these samples.

120

30
^

25

&
0

20

1

UJ 15

-

o
3 10

3
E

5 -

A.

A

0 *
10

20

30

40

60

50

Microfibril Angle (d eg rees)
o sam p le 5 o 8

a 13

x 15 ° 30 • 33 ■ 35

a

39 - 48 ♦ 60 ■ 66

a 67

♦ 80 x 88 • 90

Figure 46: MOE as a function o f microfibril angle.

Figure 46 displays microfibril angle and modulus of elasticity. There is a clear
relationship between these two properties as can he seen in this figure. As MFA
increases, MOE decreases in a log rhythmic shape, forming asymptotes at the x and y
axes. Jack pine samples are found at the lower end of the MOE scale, hut at the highest
MFA values. Lodgepole pine samples have low MFA values (as seen in previous
figures) and relatively high MOE values. Hybrid samples are found intermediate of both
species, with moderate values for both MFA and MOE.

121

Table 19; Pairwise Comparisons o f fibre traits. Species 1 = jack pine, species 2 = lodgepole pine, species
3 = hybrids, species 4 = hybrids from jack pine sampling area, species 5 = lodgepole pines from hybrid
sampling area

Dependent
Variable

Crs(ave)

(1)
sp ecies

1

mw mfa

Mean
Difference
(l-J)

Std. Error

Sig.(a)

95% Confidence interval
for Difference(a)

2
3

-36.712(*)
15.176

12.089
13.959

.011
.300

Lower
Bound
-63.320
-15.548

Upper
Bound
-10.105
45.900

5

-9.443

15.992

.567

-44.641

25.756

2

1

36.712(*)

12.089

.011

10.105

63.320

3
3

5
1

51.888(*)
27.270
-15.176

13.959
15.992
13.959

.003
.116
.300

21.164
-7.929
-45.900

82.612
62.468
15.548

2

-51.888(*)

13.959

.003

-82.612

-21.164

5

-24.618

17.449

.186

-63.023

13.786

1
2
3

9.443
-27.270
24.618

15.992
15.992
17.449

.567
.116
.186

-25.756
-62.468
-13.786

44.641
7.929
63.023

2

13.857(1

2.971

.001

7.318

20.395

3

9.742(1

3.430

.016

2.192

17.292

5

3.930
2.971
3.430
3.930

.021
.001
.255
.421

1.926
-20.395
-11.665
-11.931

19.225
-7.318
3.435
5.368

3.430

.016

-17.292

-2.192

5

MFA(ave)

(J)
sp ecies

1

2

1
3
5

10.575(1
-13.857(1
-4.115
-3.282

3

1

-9.742(1

2

4.115

3.430

.255

-3.435

11.665

5

.833
-10.575(1
3.282

4.288
3.930
3.930

.849
.021
.421

-8.604
-19.225
-5.368

10.271
-1.926
11.931

5

1
2
3

-.833

4.288

.849

-10.271

8.604

1

2

8 .906(1

3.474

.026

1.259

16.553

3
5
1

7.485
8.295
-8.906(1

4.012
4.596
3.474

.089
.099
.026

-1.346
-1.821
-16.553

16.315
18.411
-1.259

3

-1.421

4.012

.730

-10.252

7.409

-10.728

9.505

-16.315
-7.409
-10.228
-18.411

1.346
10.252
11.848
1.821

2

5

-.611

4.596

.897

3

1

5

2
5
1

-7.485
1.421
.810
-8.295

4.012
4.012
5.015
4.596

.089
.730
.875
.099

2

.611

4.596

.897

-9.505

10.728

3

-.810

5.015

.875

-11.848

10.228

Based on estimated marginal m eans
* The mean difference is significant at the .05 level.
a Adjustment for multiple comparisons: Least Significant Difference (equivalent to no adjustments).

122

Cluster Analysis:
Table 20 a: Final Cluster Centers for fibre traits
Cluster
1

2

3

334.422

367.676

408.744

20.7126

21.7395

12.1331

17.8418

11.8262

Crs(ave)
MFA(ave)
mw mfa

16.1714

Table 20b: Distances between Final Cluster Centers
Cluster
1

1

2

33.312

3

74.941

2

3

33.312

74.941
42.603

42.603

Table 20c: Number o f Cases in each Cluster
Cluster

1

3.000

2

9.000

3

3.000

Valid

15.000

Missing

78.000

Further statistical data can be found in Appendix 3.

123

4.6 Discussion and Conclusions

Some wood and fibre properties that are characteristics of jack pine, lodgepole
pine, and hybrids of these two species have been identified in Section 4.5. The objective
of this chapter was to identify differences in wood and fibre traits among the species
groups and provide for future prediction of fibre traits based on knowledge of species, or
vise versa. Species groups were broken into 5 categories, as in Chapter 3, according to
genetic results obtained in Chapter 2. Jack pine, lodgepole pine, and hybrids made up the
first three groups; group 4 was composed of samples collected from the jack pine area
that were genetically unique from other jack pine samples, and species group 5 was
composed of samples colleeted from the hybrid area that showed up as genetic lodgepole
pines. Species groups 4 and 5 were differentiated for the statistical purposes in Tables
12, 14, 16, 17, and 19, but left in regular sampling area groups for most of the
comparison analysis; it was found that separation of these groups for visual comparison

did not have a noticeable difference on conclusions.

4.6.1 Discussion of Solid Wood Properties
Average moisture contents, shown in Table 11, were variable depending on age of
samples, site conditions, and specimen health. As a percent of oven-dried, and
considering the whole core was used in the determination, the average moisture content
(MC) of jack pine was the highest, w hile hybrids had the low est moisture content. The
pairwise comparison of moisture content revealed that significant differences exist
among jack pine, lodgepole pine and hybrid groups. However, because MC is lowest in
the hybrid groups, it is not a characteristic that displays the intermediacy of hybrids

124

between pure species. In order to verify this result, it is necessary to consider the
variables that directly affect MC, to ensure that comparisons between samples are a direct
result of species difference and not age, site conditions or tree health. Age differences
that are expressed in Table 8 show that jack pine stands are much younger than lodgepole
pine stands for this study. Hybrid stands are of intermediate age. A possible explanation
for MC variation is as a tree ages, it gains a larger percentage of heartwood, which often
has a lower moisture content.
Jack pine samples are the youngest, and possess the highest MC. However,
lodgepole pine samples are the oldest and do not have the lowest MC, therefore factors
other than age may be influencing MC. Since no obvious difference exists in the health
of the trees sampled, it is logical that site conditions are the explanation for the variation
in MC. Site indices are different for each site indicating small differences in nutrient and
moisture regimes. The higher the site index value, the more suitable the site, i.e. the
higher the nutrient regime and the more access to water sources the trees have. If this is
the case, then it may be assumed that the jack pine sites were the wettest, because the
jack pine samples had the highest M C’s, while hybrid pine sites were the driest, because
hybrid samples had the lowest MC values. Since jack pine sites were of the sandiest soil
type, they were well drained and sites were probably drier than some with higher clay
content; therefore site index may not have a direct relationship to wood MC, other site
variables may provide more insight.
The time of sampling varied between sampling areas. Jack pine samples were not
frozen when sampled in late September, lodgepole pine samples were partially frozen
when sampled in November, and hybrid samples were completely frozen when sampled

125

in January. It is questionable whether or not this has an effect on the MC, but it is
possible that frozen samples might be drier than samples that were not frozen when
collected, due to the trees dormancy.
Average densities of the overall oven dried cores for each species are also listed in
Table 11 and profiles of mature wood densities for each species are shown in Figures
27a, 27b, and 27c. Table 11 shows that lodgepole pine has the highest average density
for the overall core, while hybrid pine has the lowest average density. The same results
can be gathered from the density profiles of mature, early wood in Figures 27 a, b, and c.
Although the average densities are different, only lodgepole pine is significantly different
from the other species groups; hybrid pine and jack pine are not significantly different
from each other with respect to overall average density according to the standard error
comparisons in Table 11. The pairwise comparison in Table 12 indicates that no
significant differences exist between jack pine, lodgepole pine or hybrid groups.
Interestingly, only significant differences were found between species group 5 and the
two pure species, which supports the conclusion in this study, that overall density is
affected more by environmental influences than genetic influences, and cannot be a
reliable characteristic for supporting genetic differences between species groups.
Moreover, overall core density cannot be used to demonstrate intermediacy in hybrid
samples.
A closer investigation of density with regard to mature wood exclusively, in Table
12, reveals that jack pine and lodgepole pine are significantly different and lodgepole
pine and hybrids are significantly different. Jack pine and hybrids are not significantly
different however, which suggests that hybrids are more like jack pine than lodgepole

126

pine when considering mature wood density. Like moisture content, mature wood
density is a characteristic that is heavily dependent on outside factors. So, in order to
determine if the variation in mature wood density is due to species type, and not variables
such as site conditions, comparisons must be made between densities as a function of
growth rate; a variable directly related to site conditions.
Individual core densities were investigated for their relationship to DBH growth
rates in Figure 28a and b. Figure 28a shows the overall density of the whole core of all
samples in relation to DBH growth rate. Figure 28b in contrast shows the densities of
mature wood and earlywood only in select samples. It is expected that samples with a
higher rate of growth would have a lower density, and samples with a lower rate of
growth would have higher density. This is primarily due to a higher percentage of
latewood in samples that grow slowly, and since latewood is denser than earlywood
(Panshin and de Zeeuw, 1980), the sample is denser. Figure 28a and b show that jack
pine samples and lodgepole pine samples were similar in their relationships between
density and DBH growth; lodgepole pine had a lower growth rate and higher density,
where as jack pine had a higher growth rate and lower density on average. Hybrid
samples were slightly different in that they had a lower growth rate than jack pine
samples, but a similar average density. Therefore even though hybrids were growing
slower, perhaps due to site conditions, this does not increase their densities (or latewood
percentage) as high as it would in lodgepole or jack pine samples. This suggests that
latewood in hybrids may be less dense than in pure species, or that hybrids may form less
latewood, in relation to earlywood, than pure species.

127

Table 13a, b, and c, show the results of a eluster analysis for all samples using the
variables moisture content and density. These clusters were formed to maximize the
variability between clusters. According to the majority of samples clustered, hybrid
samples and jack pine samples are significantly different; half the lodgepole pine samples
are similar to hybrids, while the other half of the lodgepole pine samples are more like
jack pine samples. Very few of the samples were categorized into the third cluster,
indicating that really only two groups can be identified as significantly different
according to moisture content and density (see Appendix 3).
Earlywood : latewood ratios were calculated in this study in order to acknowledge
the differences in the wood produced early in the growing season, with wood produced
later in the growing season. The ratio itself provides distinction between different kinds
of wood, and using it in a relationship to other variables such as density or cell wall
thickness can help to explain variability in results. Based on the results, there is a large
distinction between pure species and hybrids when observing average EW:LW ratios
(Figure 29). The ratios for hybrid pine are much higher than both lodgepole pine and
jack pine, suggesting that there is a higher percentage of earlywood present in the
hybrids. Earlywood and latewood percentages are indicative of other characteristics
because of their vast differences. Latewood contains smaller tracheids, with thicker cell
walls and is higher in density than earlywood (Panshin and de Zeeuw, 1980). The
relationship to cell wall thickness is demonstrated in Figure 31a, where an increase in the
percentage of earlywood in a sample leads to a decrease in cell wall thickness,
corresponding to evidence that earlywood cell walls are thinner than latewood cell walls.

128

A similar relationship may be assumed to exist between EW:LW ratios and density.
Density increases with increases in cell wall thickness, therefore an increase in the
percentage of earlywood in a sample would mean a decrease in the density of that
sample. The percentage of EW to LW in a sample may be partially due to genetic
inheritance within species. If factors such as age of samples and site conditions were
controlled, it may be possible to support between species variation based on EW/LW %.
In a study by Ivkovich et al. (2002), LW % in spruce was observed to have high
heritability in both East Kootenay and Prince George study sites. This observation
suggests that EW;LW ratios could be used as a species or population identifier.
However, the pairwise comparison conducted in this study, shown in Table 14, revealed
that no significant differences exist between the two pure species, or between jack pine,
lodgepole pine, and the hybrids. This characteristic may be more valuable for
distinguishing spruce species as indicated by Ivkovich’s study (2002) due to the slower
transition between earlywood and latewood in spruce as opposed to the abrupt transition
in pines (Panshin and de Zeeuw, 1980). From the discussion it can be concluded that for
this characteristic, hybrids did not demonstrate intermediacy between pure species.
Figures 30a and b are electron microscope pictures that display a) the ring profile
of a hybrid sample, and b) the cross section of a jack pine sample. The ring profile gives
an example of the transition from earlywood to latewood in pines, and also illustrates the
percentage of earlywood to latewood that can be found in an average hybrid sample. The
latewood reveals thicker cell walls than the earlywood; cell lumens are large and visible
in the earlywood, but not in the latewood. The cross section of magnified cells in Figure
30b illustrates cell wall thickness, and cell dimensions. A wood ray is visible in the

129

center of the electron micrograph, and pits can be seen connecting tracheids, in the top,
left comer of the picture.
Cell size is illustrated in Figure 31b. Lodgepole pine and jack pine cells have
similar sizes with differences being that lodgepole pine cell diameter is larger in the
tangential direction and jack pine cell diameter is larger in the radial direction. Hybrid
samples seem to be of intermediate size in both directions; however differences in cell
size are not significant enough to be considered a method of distinguishing between
species groups.
Cell wall thickness was discussed previously in its relationship to EW; LW content.
This can be further illustrated by Figure 32, which shows a profile of mature wood wall
thickness moving towards the bark of the tree. Wall thickness values that are higher
represent the latewood component of the growth ring, and lower wall thickness values
represent the earlywood component. On average, lodgepole pine cell wall thickness is
approximated at 3.02pm (Forintek, 1994), which is a representative value and because it
is an average it does not incorporate the variation that is seen in Figure 32. The SilviScan
values obtained from this study indicate that average cell wall thickness for jack pine is
2.43 pm, for lodgepole pine is 2.7 pm, and for hybrids is 2.4 pm. The lodgepole pine
samples have higher cell wall thickness values than the others, and the hybrid samples
have the lowest cell wall thickness value. The pairwise comparisons of each species
group (Table 14) showed that the average cell wall thickness is significantly different
between jack pine and lodgepole pine, and between lodgepole pine and hybrid pine, but
not between jack pine and hybrids, suggesting that hybrid average cell wall thickness is
more like that of jack pine rather than lodgepole pine. Cell wall thickness can therefore

130

be used to distinguish between some samples, but not others, and can not be used as an
indicator of intermediacy in hybrids.
The relationship between cell wall thickness and density is expressed in Figure 33.
As cell wall thickness increases, density increases. This is intuitive because the more cell
wall material that is present in a measured area of wood, the denser that area would be,
forming a direct linear relationship (Panshin and de Zeeuw, 1980). This relationship is
evident in every one of the samples measured, and little difference exists between species
groups. Figure 33.
Figure 34 shows the relationship between cell wall thickness and modulus of
elasticity (MOE). No strong linear relationship is present between these two variables
due to the large cell wall thickness variations from earlywood to latewood, but based on
this figure and Figure 35, differences can be distinguished between species groups with
regard to MOE. Lodgepole pine has the highest MOE, jack pine has the lowest, and
hybrids show intermediate MOE. According to Table 14 differences in MOE are
significant between all species groups except group 3 (hybrids) and species group 5
(samples from hybrid sampling area with lodgepole pine genetic output). Therefore,
modulus of elasticity is a characteristic that can distinguish between species groups in
this study, and can be used to indicate intermediacy of hybrids. In order to extend this
into a general statement regarding distinction between lodgepole pine, jack pine, and
hybrids outside of this study, factors including site conditions and sample age must be
considered. The results of MOE also support the theory that species group 5 are in fact
hybrids but may have lost jack pine genetic material over generations of back-crossing,
these groups do not display significant variation in MOE.

131

Figure 37 shows the relationship between MOE for mature wood only, which
addresses the issue of age difference, and growth rate according to DBH, which is a

direct reflection of site conditions. Figure 37 shows that there is a distinction between
species groups. Jack pine has a lower MOE on average, which is associated with a
higher growth rate. This is an expected trend, since as trees grow faster, they produce
wider rings that are less dense and the stems contain more juvenile wood (Panshin and de
Zeeuw, 1980). Therefore, wood has lower MOE. Lodgepole pine has a lower growth
rate and higher MOE. Hybrids are intermediate with similar growth rates to lodgepole
pine but lower MOE. Since this figure reflects the same results as those found in Figure
35, a difference in MOE between species groups is due to genetic variation and therefore
can be extended to the pine populations in these areas.
Tables 15a, b, and c reflect the cluster analysis preformed in order to group samples
together based on mw density, EW:LW ratio, MOE, and cell wall thickness. Since only
five samples from each of the pure species groups, 3 samples from the hybrid group, and
2 samples from species group 5 were analyzed for these characteristics, the cluster
analysis is limited in its statistic representation. However, cluster formations do indicate
that some samples are more alike than others, with respect to these characteristics. The
samples forming cluster 1 consist of 50% jack pine samples. Cluster 2 is made up of
100% lodgepole pine samples, and cluster 3 contains 50% hybrid samples. Even though
there is not an overwhelming trend separating the three main species groups from one
another, only one species groups occupies the majority of each cluster, indicating that
there is a possibility of species distinction based on these characteristics if more samples
were tested.

132

In order to better support the conclusions drawn from the wood analysis portion of
this study, alternate eharaeteristies could be looked at to achieve greater confidence in the
variability between species. It was suggested in a study by Christensen and Dar (2003)
that number of resin ducts can be distinctive of species. This could be further
investigated and related to percentage of EW and LW in this study, in order to draw more
conclusive evidence of differentiation between species groups and intermediacy of
hybrids.

4.6.2

Discussion of Fibre Properties
Fibre properties are directly correlated to several wood characteristics that have

already been identified. Fibre length and microfibril angle are both a reflection wood
maturity (Barbour, 2004), and fibre coarseness is directly correlated to cell wall
thickness, and therefore FW:LW ratio (Potter et al., 2001). Although relationships exist
between these characteristics, differences can be noted in their significance for
differentiation between species groups investigated in this study.
Figure 38 shows the length weighted fibre length from pith to bark in all samples.
The relationship between fibre length and wood as it matures is represented well. It
shows that as wood matures from pith to bark, fibre length increases until a pure mature
wood stage is reached at which point the increase levels off and fibre length becomes
relatively constant. Lodgepole pine has the greatest fibre length at all stages of wood
development (Figure 38). Jack pine fibre length is shown to be intermediate, and hybrid
fibres are shown to be the shortest of the three species groups. Species groups 4 and 5
were not separated out for this analysis. As explained earlier, little variation occurred in

133

conclusions drawn from sample separation. According to Figure 38, it can be concluded
that there is a difference in fibre length among species groups, but hybrids do not display
intermediacy between the two pure species. Even though Figure 38 presents evidence of
species differentiation, it is important to consider influence of site conditions and
statistical reliability.
The number of fibres sampled per fibre size interval is reflected in Figure 39.
There is a large sample size represented here, which enhances the reliability of the data,
based on whieh a pattern of fibre lengths can be differentiated. There is a larger amount
of short fibres in jack pine, and a larger number of longer fibres in lodgepole pine. The
largest number of fibres in hybrids is intermediate of the two pure species. Even though
the sample size considered here is large enough for statistical reliability, the pairwise
comparisons conducted in Table 16 and 17 show that there are no significant differences
between any species group with respect to fibre length at any stage of wood maturity.

Therefore fibre length can not be used as a discriminating characteristic between species
groups or support intermediacy in hybrids; however, the fibre length distribution curve
(Figure 35) is useful for displaying intermediacy of hybrid fibre lengths.
Another characteristic investigated in this study is fibre coarseness, and is largely
dependent on fibre length (Figure 40). Lodgepole pine samples have the largest fibres,
both in length and coarseness. Jack pine and hybrid samples have shorter fibres
compared to lodgepole pine, and are made distinctive of each other based on coarseness.
Flybrid samples are slightly coarser than jack pine samples. Lodgepole pine samples are
coarser than hybrid and jack pine samples from 40-80 years from the pith (Figure 41).
Table 16 shows that the difference in coarseness between jack pine and lodgepole pine is

134

significant, as well as the difference between lodgepole pine and hybrids at 40-60 years
from the pith. At 60-80 years from the pith there is a significant difference between
lodgepole pine and hybrids, but no difference between jack pine and the hybrids or
lodgepole pine samples.

There is no significant difference between the fibre coarseness

of samples at 20-40 years from the pith (Table 17). With that, it is possible to use
coarseness to distinguish between lodgepole pine and jack pine, and lodgepole and
hybrids, and therefore hybrid coarseness values may be more closely related to jack pine
samples than lodgepole pine samples. On the other hand, average overall coarseness in
Table 19 displays that there are signifieant differences between jack pine and lodgepole
pine, and jack pine and hybrids, indicating that hybrids are more closely related to
lodgepole pine. Due to this discrepancy, a greater sample size should be used in the
overall coarseness analysis (Table 19), because only 15 separate samples were measured
at multiple positions for the overall measurement by SilviScan. Coarseness profiles in
Table 16 are from FQA analysis of over 80 samples and therefore reflect a more accurate
representation. Coarseness does not indicate intermediacy in hybrids by either
calculation. Figure 42 investigates the relationship between coarseness and density. This
figure shows a direct correlation, where an increase in coarseness leads to an increase in
density. Lodgepole pine samples are have coarser fibres, and are denser; hybrid fibres
are the finest and samples are not as dense. Coarser fibres therefore lead to denser wood,
which is usually wood that has grown more slowly, and developed a higher percentage of
latewood (Panshin and de Zeeuw, 1980).
Table 18a, b, and c show the cluster analysis results for fibre length and coarseness.
No conclusive results are reflected from this analysis with respect to species groupings.

135

Clusters 1, 2, and 3 all contain samples from all three species groups. Based on these
results, some samples can be seen as more related to hybrids than others, but it is not
dependent on species group.
Microfibril angle was investigated in order to determine its ability to distinguish
between species groups. Figure 43 shows the profiles of each sample from pith to bark
with respect to MFA, and average MFAs over the whole core is displayed in Figure 44.
Jack pine samples show a profile with greater MFAs and lodgepole pine samples show
profiles with lower MFAs. Hybrid samples show MFA profiles with intermediate values.
These relationships are reflected almost continuously over the whole profile, indicating
that even mature wood displays differences in MFA between species groups; differences
are not only due to juvenile wood percentage or the variation in age of sample. The same
trend is observed in both Figure 43 and 44. Therefore, it can be assumed that jack pine
has higher MFAs in general than lodgepole pine, and hybrids display intermediacy for
this trait. In order to verify this trend, and ensure that variation in MFA is not due to site
conditions, MFA versus growth rate is shown in Figure 45. This relationship reveals that
some jack pine samples have faster growth rates corresponding to higher microfibril
angles. However, other jack pine samples reveal the same MFAs as lodgepole pine or
hybrid samples for varying growth rates, indicating that growth rate does not have an
affect on the MFA of every sample in this study. Table 19 displays statistical
significance in a pairwise comparison for both mature wood MFA and overall MFA. For
overall MFA, jack pine is significantly different from all samples, all other samples are
not significantly different from each other. Mature wood MFA, which is a more accurate
way to compare species groups as it is not dependent on age, reveals that jack pine and

136

lodgepole pine are significantly different from each other. However, hybrids are not
significantly different from either species group.
Considering all relationships, MFA can be used as a characteristic to distinguish
between jack pine and lodgepole pine pure species, but not hybrids. However, MFA can
be used to support the theory of intermediacy in hybrids. To add to this, the relationship
between MFA and MOE is shown in Figure 46, which can give insight into using MOE
to predict MFA, or vise versa.
Table 20a, b, c show the results of a cluster analysis preformed to determine how
samples can be grouped when considering influences of overall coarseness, overall MFA,
and mature wood MFA. Four out of five jack pine samples fell into cluster 2, three out of
five lodgepole pine samples fell into cluster 3, and three out of five hybrid samples fell
into cluster 2. This indicates only that hybrid samples may be more related to jack pine
samples than lodgepole pine samples. However, considering that only five samples were
available for clustering per species group, it would be more statistically reliable to study
a larger sample size in the future.

4.6.3 Conclusions
In summary, several wood and fibre quality traits are useful in combination to help
identify species groups and support hybridization. Moisture content and MOE were the
most valuable characteristics for differentiating between pure jack pine, lodgepole pine,
and hybrids of the two species in this study, as well as acting as traits to support the
theory of intermediacy in hybrids; hybrids show intermediate traits of the pure species
they are derived from. In addition to these characteristics, cell wall thickness and fibre
coarseness can be used to differentiate between lodgepole pine and jack pine, and

137

lodgepole pine and hybrids. Because these traits indicate no significant difference
between jack pine and hybrids, the hybrids have characteristics like jack pines. Finally,
fibre length distribution and MPA, although not statistically different for hybrid
differentiation from pure species, do show intermediacy of hybrid characteristics, further
supporting this theory.
Possible sources of error that may have skewed results include site conditions
unaccounted for such as stand density and position, wood types such as reaction wood
that may have be present in the stems, and any undiagnosed disease or pest attack. These
variables were controlled as well as possible with factors such as growth rate and site
index taken into consideration. The objective of this Chapter, to compare species based
on wood and fibre traits in order to identify if any significant differences, was met;
results indicate that some traits are more distinctive than others for differentiation of
hybrids in the Fort Nelson region of British Columbia.

4.7 References
Alden, H.A., 1997. Softwoods of North America. Gen. Tech. Rep. FPL-GTR-102. US
Department of Agriculture, Forest Service, Forest Products Laboratory. Madison,
WI.
Barbour, J. 2004. Wood Quality. Encyclopedia of Forest Science pp. 1840-1847,
Elservier Major Reference Works, Oxford.
Christensen, K. I. and G.H. Dar. 2003. A morphometric study of hybridization between
Pinus mugo and P. sy Ives tris (Pinaceae). Acta Hort 615, International Society of
Horticultural Science.
Goldstein, Joseph I., 1992. Scanning electron microscopy and x-ray microanalysis : a text
for biologists, materials scientists, and geologists. 2nd ed.. Plenum Publishing
Corp.
Isenberg, I.H., 1951. Pulpwoods of United States and Canada. The Institute of Paper
Chemistry, Appleton, Wisconsin.

138

Ivkovich, M., G. Namkoong, and M. Ko shy. 2002. Genetic variation in wood properties
of interior spruce 2. Tracheid characteristics. Can. J. For. Res. 32: 2128-2139.
Koch, Peter., 1996. Lodgepole Pine in North America, v. 1. Forest Products Society,
Madison, Wisconsin.
Panshin, A.J., and C. de Zeeuw. 1980. Textbook of Wood Technology, 4* Edition.
McGraw-Hill Inc. Toronto, CA.
Pitts, S., P. Bicho, and P. Watson. 2001. Rapid Prediction of Fibre Properties:
Opportunities Using Forest Site Classification Methods. Pulp and Paper Research
Institute of Canada, PPR 1532. Vancouver, BC.
Potter, S., V. Hanemaayer, S. Pitts, and P. Watson. 2001. Fibre Quality Prediction in
B.C. Spruce with DNA Based Methods. Pulp and Paper Research Institute of
Canada, PPR 1528. Vancouver, BC.
Pulp and Paper Research Institute of Canada (Paprican). 2004. “Paprican to Market
CSIRO’s SilviScan Services in North America”. Press Release. Montreal, QC.
Robertson, G., J. Olson, P. Allen, B. Chan, and R. Seth, 1998. Measurement of Fibre
Length, Coarseness, and Shape with the Fibre Quality Analyzer. Pulp and Paper
Research Institute of Canada, PPR 1361. Pointe-Claire, QC.
Sharma, S. and S. Potter, 2000. “Use of a DNA-Based Hybridization Marker for Hybrid
Pine Discrimination”. Pulp and Paper Research Institute of Canada, unpublished.
Vancouver, BC.

139

Chapter 5: Chemical Analysis of Jack Pine, Lodgepole Pine, and their Hybrids
5.1 Chapter Objectives
The objective of this chapter is to determine if a “chemical extractive footprint” can
be identified for each of the three species groups: jack pine, lodgepole pine, and hybrids,
and to determine if this footprint can be used to differentiate between species. The
creation of an extractive footprint by gas chromatography - mass spectrometry enables
the identification of some of the extractives within wood that are characteristic to one or
more of the species groups in question, or allow for characteristic quantities of the same
extractives to be identified within a species group. Characterizing the species in this way
allows for further support of hybrid existence, and support the morphological evidence of
hybridization that is prominent in scientific literature and found in Chapter 3.

5.2 Wood Chemical Extractives from Literature
Wood extractives usually account for less than 5% of wood volume and are not
considered a structural wood component. Wood extractives can be categorized as
triglycerides, resin acids, fatty acids, terpenoids, waxes, phenolic acids, or others,
depending on the literature. These compounds are found throughout the wood at
different locations depending upon the chemical (Sjostrom, 1993). Large amounts of
chemical extractives are found in the cell lumens, but extractives can also be found in the
cell walls (Higuchi, 1985). Resin acids are found within the resin ducts and fatty acids
are found in the ray parenchyma. Most species have a distinct pattern of chemical
extractive types, amounts, and locations, although similar within families (Sjostrom,

1993).

140

Functions of a cell within a specific species are detectable based on the chemical
extractive makeup found within the wood. Fats provide energy for the cells, and
terpenoids, resin acids, and phenolic acids aid in defense mechanisms to protect the tree
from pests and pathogens (Sjostrom, 1993). Therefore, if a tree has particularly high
levels of resin acids, we may be able to assume that the speeies has a need to defend
itself, and therefore has a larger percentage of pests and pathogens that can harm it
compared to another species with lower resin acid levels. Links ean then be made from
this idea to environmental and biological reasons that the tree is more susceptible to
attack, such as thinner bark, or favored climate or elevation range.
Chemical extractive footprint can be used to identify similarity between species. In
this case, determining the differences, if any, between lodgepole pine and jack pine will
aid in the identification of a possible hybrid between these two species.
The ratio of extraetive and tissue amounts varies between species and families and
is therefore the basis of chemotaxonomy that is, to identify plant material based on
chromatographic output. Polyphenol eompounds are most commonly used for
chemotaxonomy due to their distinetive nature, and have been used to group speeies in
Pinus, Prunus, Acacia, and Eucalyptus genera (Higuchi, 1985).
The amount of extractive varies widely between speeies, and have extensive effects
on wood quality. For example, extractives can be a hindrance in pulping processes, and
can affect wood color in furniture or instrument manufacturing (Higuchi, 1985).
Terpenoids are composed of hydrocarbons and their derivatives with over 7500
different structures divided into subgroups according to isoprene unit; hemi, mono.

141

sesqui, di, sester, tri, tetra, and poly prefixes indicate the number of carbon and isoprene
units found in the terpene compound (Sjostrom, 1993).
Softwoods house fats and waxes in ray parenchyma while hardwoods contain fat
and wax abundantly throughout all parenchyma. More than 30 fatty acids have been
identified in softwoods and hardwoods; the most common unsaturated fatty acids are
oleic and linolenic acid and pinolenic acid is a major fat in pines and spruces. Waxes can
be categorized as free fatty alcohols, esters of higher fatty acids, or terpene alcohols
(Sjostrom, 1993).
Phenolic compounds are a large category of extractive chemicals also known as
complex aromatic extractives. Compounds in the phenol family ean be further broken
down into subgroups; stilbenes, lignans, hyrolyzable tannins, flavanoids, and condensed
tannins. Typical phenolic compounds found in pines are the flavanoid known as chrysin,
the stilbene known as pinosylvin, and the lignan, pinoresinol. Hyrdolyzable tannins are
uncommon in wood and condensed tannins occur mostly in hardwood barks (Sjostrom,

1993).
A small percentage of the extraetive component of wood is composed of inorganic
compounds (<1%). Such inorganics include: silicates, oxalates, phosphates, calcium,
potassium, magnesium, iron, cobalt, and manganese (Sjostrom, 1993).

5.3 Experimental Design and Hypotheses
A core sample (pith to bark) from each tree was used for the acetone extraction,
and chemical extractive makeup of each sample was determined using gas
chromatography- mass spectroscopy (GC-MS).

142

Thirty samples of pure lodgepole pine, 30 of pure jack pine, and 30 potential
hybrid samples were used for this analysis; from the Prince George area of BC, the
Smoky Lake area of Alberta, and the Fort Nelson region of BC, respectively. Each site
was chosen for its similarity in vegetation and moisture/nutrient regimes. Sites for
lodgepole pine and jack pine were chosen as to be similar distances from the potential
hybrid zone.

5.4 Methodology
5.4.1 Chemical Analysis Using GC-MS
Gas chromatography-mass spectroscopy (GC-MS) was used to analyze the
chemical extractive components of the wood cores. The GC uses helium gas moving
through a capillary column 10 meters long and only 0.25 mm in diameter to separate each
chemical component individually for analysis. As each chemical moves though the
column, it either has an affinity for the column or an affinity for the helium. Depending
on its level of affinity for the stationary phase (the column) or the mobile phase (the
helium), which is generally individually distinctive, the chemicals gradually separate.
This separation means the chemicals move through the column at different speeds and
therefore a retention time, how long the chemical spent in the column, can be measured.
Chromatograms are created based on these retention times, each chemical substance or
component represented by a peak at the specific retention time measured. The peak size
or area represents the amount of chemical present, or the strength of the measurement.
Once the substances are separated through the column, and retention times are
recorded, the substances move into the mass spectrometer (MS), where they are
analyzed. The MS consists of an electron beam which comes in contact with each

143

chemical and forces ions to break off, which are counted in an ion trap. This allows the
chemical mass to be identified based on how many ions of the substance are present in
relation to each other, and gives a spectral analysis output which contains the chemical
mass, the “parent ion” whieh is the chemical or chemical component most represented in
the mass spectra, and other ions, that may be other chemicals present in the substance or
degraded chemical constituents.

Chemical Extraction Protocol
Cores were oven-dried and weighed. Two-hundred milliliters of acetone was
measured into round bottom flasks. Each sample was individually extracted with the
acetone using a heating and cooling apparatus to cycle the acetone and draw chemicals
from the wood over a five hour period.
The 200 ml of acetone plus extractives was then condensed down to 5 ml of liquid
using a Rotavapor® heating, cooling, and vacuum system, or condensed down to 5 ml of
liquid using a TurboVap® with heat and nitrogen gas.
After the acetone solution was brought to a volume of ~5 ml, the liquid was poured
into small pre-weighed vials. The vials were then placed under a low heat and nitrogen
gas to blow down the remaining liquid to approximately 1ml, composed of chemical
extractives from the wood samples while minimizing loss of volatile extractives. Vials
were then sealed with Parafilm®, and placed in a -86°C freezer. Once samples were
frozen they were freeze dried.

144

Dried samples were weighed to determine the percentage of the original core that
was extracted. The dried samples were then re-suspended at 5.000 mg/ml, in an internal
standard solution of Ethyl acetate containing 0.250 mg/ml heptadecanoic acid.
One microliter of each sample was then programmed into the GC using a run-time
of 42 minutes.
Chromatography- Identification and Quantification of Compounds
Extractive profiles (mass spectral output) were established after MS was
completed. Each chemical profile was composed of a series of peaks, representing how
the ions were broken during the MS.
Chromatographs were aligned subsequent to initial output, based on the internal
standard used, in order to properly compare the peaks between samples. After alignment
the chromatographs were still subject to slight differences due to stretch. This variability
is due to the GC analysis; it is very difficult to maintain exactly the same conditions
within the column being utilized. Because of this “stretch” variability, the
chromatographs were used to form 20 peak regions for analysis, instead of identifying
each individual peak. Therefore it was possible to compare regions between samples,
without having to be concerned that any between sample variability was caused by
stretch in the chromatogram. These peak regions were analyzed using partially least
squares discriminant analysis (PLS-DA) in order to identify regions that were the most
variable between species. The peak regions that caused the most variability between
species groups were then identified by mass spectra and compared to a mass spectral
library where extractive profiles are matched to the closest possible found in the library
or authentic known standards. If compounds did not return an appropriate match, the

145

chemical substance was classified based on broad spectral features. The library returns
chemical matches based on the purity of the samples, the fit, and the retro-fit. These
identifiers have a maximum value of 1000, which would indicate a perfect match. Purity
indicates the number of extra peaks that occur in the sample as compared to the candidate
substance from the library. The fit indicates how closely the sample mass spectra fits
with the mass spectra in the library, and the retro-fit indieates how closely the library
mass spectra for the candidate substance fits with the sample mass spectra.
The 20 peak regions analyzed by PLS-DA were also used to create a discriminant
variable with which to compare the samples. The discriminant variable was used to
identify how alike the samples were to lodgepole pine.

5.5

Results
Table 21:

Jack pine site 1
Jack pine site 2
Jack pine site 3
Lodgepole pine site 4
Lodgepole pine site 5
Lodgepole pine site 6
Hybrid pine site 7
Hybrid pine site 8
Hybrid pine site 9

% Extractives
4.466
4.826
2.456
1.691
2.580
4.125
4.152
4.815
1.884

Table 21 show the average amount of extractives that were found in each sample
according to collection site (%).

146

5.5.1 PLS-DA Output

0 .4

- ■

0.2 - ■

5 0.0

-0 .2 H

■0 . 4

Regions o f scanning range
Figure 47: PLS-DA model output o f the regression coefficient for lodgepole pine. The greater the
coefficient value, the greater the contribution o f that peak region to the variability between the sample
groups. A negative coefficient value indicates a negative correlation to lodgepole pine.

Chromatogram regions are indicated on the x-axis of Figure 47. These regions
were used in order to eliminate variability due to stretching in the chromatographs.
There are several peak regions on the chromatograms that contribute to the variability
between samples based on chemical analysis (Figure 47). Out of the 20 regions
indicated, 3 are responsible for a large amount of variability, represented by regression
coefficients greater than +!- 0.28. These 3 regions showing the most variability are
region 5 (representing scan range from 975-1005), region 9 (representing scan range
from 1090- 1110), and region 11 (representing scan range from 1130-1150) and are
shown on the chromatogram shown in Figure 48. Region 9, at 1131, indicates the most
variability between sample groups. Region 5, at 976, is responsible for the second
highest amount of variability between groups, and region 11 is the coefficient showing
the third highest variability.

147

5.5.2 Chromatography
8000000
7000000

CO

6000000

c
2 5000000
3
o
c
0 4000000 -

1
2

3000000 -

"c

§ 2000000
cc

Region 1
-

#

•

Region 2 Region 3
♦------------»

4------------ »

1000000

C \ l T t C O O O O C \ l - ^ C D O O O C g - ^ C O O O O C J - ^ C D O O O C \ l - ^ < £ ) O O O C J ' ^ < 0

o o o o o o o o œ a î a ) 0 ) 0 ) 0 0 o o o i - T - - i - T - T - c \ i c \ i c M c \ i c \ i c o c o n c o

Ordinal Scan Range
Figure 48: Aligned Chromatogram for range 825-1380, where most variability in peaks was determined to
be. Regions o f variability according to the PLS-D A output are indicated.

The chromatogram shown in Figure 48 shows the regions that are responsible for
the most variability between species groups according to the PLS-DA conducted and
displayed in Figure 47. Within these regions, main peaks were identified in order to
analyze the potential extraetive chemicals by mass spectrometry. More than one main
peak is identifiable in regions 5 and 11, and region 9 has one main peak for analysis.
These peaks were all analyzed by mass speetra in Figures 50a-f.

148

600000
« 500000
3 400000
-

300000

200000
100000

15.72

16.03

16.34

16.66

16.96

17.28

17.58

17.89

18.20

Retention Time (mins)
Figure 49a: Chromatogram o f jack pine sample 2. Peak regions used for PLS-D A are highlighted by
circles.

600000
500000
400000
~ 300000
200000
cc 100000

15.68 15.99

16.31

16.62

16.93

17.24 17.55 17.87 18.19

Retention Time (mins)
Figure 49b: Chromatogram o f lodgepole pine sample 43. Peak regions used for PLS-DA are highlighted
by circles.

149

600000
500000
400000
~ 300000

200000
100000

15.71

16.03

16.35

16.66

16.97

17.28

17.59

17.90

18.21

Retention Time (mins)
Figure 49c: Chromatogram o f hybrid pine sample 68. Peak regions used for PLS-DA are highlighted by
circles.

150

600000

1

2
500000
CO
c

^ 400000
ü
c
c 300000
o
ü
200000
c
o

I
ü

CD

CC

100000

15.67

15.99

16.30

16.61

16.93

17.25

17.56

17.88

18.19

Retention Time (mins)
Figure 49d: Chromatogram o f hybrid pine sample 84. Peak regions used for PLS-DA are highlighted by
circles.

Figures 49a-d show the chromatograms for single samples from each species group
according to retention time. Figure 49a represents the typical output for jack pine
samples (samples 1-30), Figure 49b represents the output for lodgepole pine samples (3160), and Figure 49c is representative of hybrid samples (samples 61-80). Figure 86d is
included also because it represents the samples that were collected from the hybrid
sampling area but were unique possible Fn generation hybrids back-crossed with
lodgepole pine, according to the genetic RFLP output (samples 81-90). The same
comparisons can be made between Figures 49a and b as can be seen in the regression
coefficient output in Figure 47. It is noticeable that peak region 5 (circled on the far left
of the chromatograms) is much larger in Figure 49a (jack pine) than the others; region 9
(circled in the centre of the chromatograms) is larger in the lodgepole pine profile (Figure

151

86b); and region 11 is the largest in Figure 49c (hybrid samples), but larger in lodgepole
pine than jack pine. The size of the peak indicates quantity of chemical present.

5.5.3 Mass Spectral Analysis

T a r g e t S p e c tru m .................................................................... ..... ................................................................................................................................
Target
BP 240 (8833=100%) I08.sms

18.395 min. Scan: 1150 Chan: 1 Ion: 2306 us RIG: 79867 BC
240

100 % -

75% -

50% 209
25% -

165
91
400

0%

ïo^

260

360

460

560

560

Figure 50a: Sample 08, peak region 11 at a retention time o f 18.395 minutes. The parent ion is shown at a
mass o f 240 m/z.

Table 22a: Library search results o f potential chemical compound matches for the mass spectra in Figure
50a.

1.
2.
3.
4.
5.
6.
7.

Purity Fit
470
729
450
715
444
692
439 700
430
693
419
665
721
409

RFit

528
504
476
501
514
545
443

Entry #
91361 A
91374 A
19505 B
91313 A
91368 A
91278 A
91289 A

240
240
240
240
240
240

MW. Formula
Name
C16H 20N 2 2 ,6 ,2 ’,6 ’— Tetramethyl
C16H 20N 2 4,6,2',6'— Tetramethyl
[1,1’— Biphenyl]— 4,4
C 16H20N2
[1,1’— Biphenyl]— 4 ,4’
C16H 20N 2
[1,1 ’— B i p h e n y l] ^ ,4 ’— d
016H 20N 2
355
C 16H 29N 04S12
Dimethoxymai
C 16H 1602, N one
Dim ethoxy— stilbene

152

m/2

Target Spectrum
Target
BP 187 (17744=100%) 137.sms

18.275 min. Scan: 1145 Chan: 1 Ion: 888 us RIG: 357931 BC
187

100 % 131

75%^

91

50%^

257

25% 341

ib i

0%T

260

3io

460

560

m /2

Figure 50b: Sample 37, peak region 11 at a retention time o f 18.275 minutes. The parent ion is shown at a
mass o f 187m/z.

Table 22b: Library search results o f potential chemical compound matches for the mass spectra in Figure
50b.

3.
4.
5.
6.
7.
11.
12.

Purity
805
486
474
755
464
841
440
773
726
428
667
420
419
730

Fit
527
539
501
501
501
485
523

RFit Entry #
65379 A
376
304
19779 B
12825 B
286
37145 A
296
94121 A
94122 A
404
30793 A

MW. Formula
Name
C28H 40
Cyclopentane-3 ’— spiropentacy
C 20H 3202 Methandriol
C20H 300
Retinol
C22H32
Cyclodecacyclotetradec
272
C20H32
Phenanthrene, 7— etheny
272
C20H32
Naphthalene, decahydro— 1
C30H44
Dibenzo[a,H]cyclotetra

153

Target Spectrum
T a rg et

18.361 min. Scan: 1149 Chan: 1 Ion: 3468 us RIO: 77023 BC

BP 237 (6639=100%) I49.sms
257

100 % -

75%

50%91
147

25%^

201

y
i6o

0%-

jJUiIuU l i ,..a

341
360

401

sio

m/z

Figure 850c: Sample 49, peak region 11 at a retention time o f 18.361 minutes. The parent ion is shown at
a mass o f 257 m/z.
Table 22c: Library search results o f potential chemical compound matches for the mass spectra in Figure

50c.

1.
2.
4.
5.
6.
7.

Purity
821
661
645
813

638
625
615
577

827
795
817

762

Fit

749
757
729
703
690
664

RFit Entry #
4152 A
94122 A
272
94148 A
384
406
37678 A
94144 A
370
406
39285 A

MW. Formula
Name
290
C20H 340
Verticiol
C20H32
Naphthalene, decahydro— 1
C28H48
Ergost— 14— ene, (5.alph
C27H3403
19-Nor-4— androsten— 3— one
C27H46
Cholest— 14— ene, (5.alp
Nandrolone Phenpropiona
C27H3403

154

Target Spectrum
Target
BP 257 (11490=100%) I56.sms

17.679 min. Scan: 1108 Chan: 1 Ion: 1186 us RIO: 264359 BC
257

100 % -

91

75%

123

50%187
25% -

U
if f

0%

i,Li,,ill

355

3io

200

401

429

4(!lo'

m/z

Figure 50d: Sample 56, peak region 9 at a retention time o f 17.679 minutes. The parent ion is shown at a
mass o f 257 m/z.
Table 22d: Library search results o f potential chem ical compound matches for the mass spectra in Figure
50d.
MW. Formula
Name
Purity Fit
RFit Entry #
1— Phenanthrenecarboxald”
286 C20H300
I.
653 847 695 94154 A
290 C20H 340
Verticiol
2.
598 867 653 94152 A
Naphthalene, decahydro— 1
272 C20H32
3.
574 817
623 94122 A
Phenanthrene, 7— etheny
272 C20H32
94121 A
4.
560
827 631
19— Nor— 4— androsten-3
406 C27H3403
785
5.
537
598 37678 A

T a r g e t S p e c tr u m
Target
BP 239 (4816=100%) I43.sms

16.102 min. Scan: 1006 Chan: 1 Ion: 6045 us RIC: 23618 BC
239

100 % -

T5%-

50%^
25%^

159
91

199

117

281

If ,11, .lii. ilil.ii

^ ^ ff

260

355

327

415

476

460

66^

Figure 50e: Sample 43, peak region 5 at a retention time o f 16.102 minutes. The parent ion is shown at a
mass o f 239 m/z.

Table 22e: Library search results o f potential chem ical compound matches for the m ass spectra in Figure
50e.

1.
2.
3.
5.
6.
16.
23.

Purity
523 718
519
732
743
515
412
630
409
737
374
832

343

639

Fit

685
650
651

488
493
379
376

RFit Entry #
91185 A
314

19478B

314

91068 A
82546 A
91200 A
91204 A

314

82013 A

240
254
254
254

MW. Formula
Name
1— Phenanthrenecarboxy
C21H 3002
1— Phenanthrenecarboxy
C21H 3002
C21H 3002
Phenanthrene— 1— carboxylic e
C16H20N2 9— M ethylene-l-phenyl— 3,6— d
Naphthalene, tris(l— m e’
C19H26
C16H1403
5— Hydroxy— 6— methoxy—
C 12H 1804S1 Benzoic acid, 3,4— d

155

m/z

Target Spectrum
iT a n g e t
BP 197 (11196=100%) I43.sms

15.912 min. Scan: 994 Chan: 1 Ion; 3836 us RIC: 55924 BC
197

100 % -

75% 239

50%^
25%

169
91

0%-

1A 1 1
0

i, H
I
,1. ,ll i l l ,ILi. a ,A .......k i l l .1
1 1 I •• 1 1 1 1 1 1 1 ' 1 1 1 1 1
lèo
200

...

,U

‘••’ U

'o io ..................

400

sio

’oAo’

Figure 50f: Sample 43, peak region 5 at a retention time o f 15.912 minutes. The parent ion is shown at a
mass o f 197 m/z.
Table 22f: Library search results o f potential chem ical compound matches for the mass spectra in Figure
50f.

1.
2.
3.
4.
5.
6.
7.

Purity
559
662
532 700
594
499
494
630
488 807
472
627
467
580

Fit
677

RFit Entry #
314
19478 B

652

82546 A

641
614

91185 A
91068 A
82415 A
82514 A
91350 A

532
565
527

314
314
254
254
240

MW. Formula
Name
1— Phenanthrenecarboxy
C 21H 3002
240
C 16H 20N 2 9— Methylene-1— phenyl1— Phenanthrenecarboxy
C 21H 3002
C 21H 3002
Phenanthrene— 1-carboxylic * * *
C17H22Si
1— T ert— butyl— 1— (naph------C18H 220
Spiro[cyclobutane— 1,1
Chrysene, 1,2,3,4,4a,4
C18H24

Figures 50a-f show the mass spectra output for the peaks identified within the
regions showing the highest variability between species groups. The largest ion percent
in the mass spectral output is referred to as the parent ion. This varies between spectra,
indicating differences between peak chemical compositions. The ions found with the
higher molecular weights can provide indications of the molecular weight of the
compound, as long as the peak is consistent throughout the samples.
The library output for each spectra indicates the name of the chemical that is a
potential match, the purity of that chemical, the fit, the retrofit, the chemical formula and
molecular weight. These are all indicators for chemical matches within the lists.

156

m/z

1.5

“t j r

—

L31
.L46_

1.0 -

’Ï* L 5 6 A

L56

I

.5 -

# * CM
ijnLi.3

L84

■■

--

*;'l53A
l.2:Ë
L27
-k3ê^

# Wÿ
LOI % 7
-JLÛ
0

1.78

2']

t.l7

L65A

I

L77

■ H .,.
___________ 1,68

,

r

-

:_____________L80A

1.12 *

1

1.16

L 68A

D.5L76

- 1 .0 -

[

I

I

I

I

I

I

I

I

I
20

I

I

I

I

I

I

I

I

40

I

I

I

I

I

I

I

I

Sample Number

I

I

I

I

I

I

TT-

I

00

100

80

Figure 51: Sample groups based on the discriminant variable value calculated, and coordinating species
group. The discriminant variable indicates how similar the samples are to lodgepole pine. Samples LOI-30
= jack pine = sampling area, samples L31- 60 = lodgepole pine sampling area, samples L61-90 = hybrid
sampling area.

Figure 51 shows variability between groups based on a prediction coefficient
established from the PLS-DA. The prediction coefficient is used to determine how
closely related each sample point is to lodgepole pine chemical characteristics. This
figure demonstrates that there is a difference between lodgepole pine samples and the
other species groups; groups show distance in coefficient value from lodgepole pine
samples. Lodgepole and jack pine samples are represented by circles, while hybrid
samples are represented by squares. Jack pine and hybrid samples fall mostly on the
lower half of the graph with coefficients of 0.5 or less, while lodgepole pine samples fall
mostly on the upper half of the graph with coefficient values greater than 0.5.

157

5.5.4 Statistical Analysis of Chromatographs
Table 23: Pairwise Comparison o f Chromatographic peak regions with most variability between species
groups. Species 1 = jack pine, species 2 = lodgepole pine, species 3 = hybrids, species 4 = hybrids from
Dependent
Variable

Regions

(1)

(J)

sp ecies

sp ecies

1

2

3

4

5

2

4

95% Confidence Interval for
Difference(a)
Lower Bound

Upper Bound

789220.719(1
760219.272(1
934600.712(1

194973.868
200812.026
399335.578

.000
.000
.021

402251.562
361662.988
142030.132

1176189.876
1158775.556
1727171.291

5

1011188.762(1

276667.804

.000

462079.756

1560297.767

-789220.719(1

194973.868

.000

3

-29001.447

189249.143

.879

1176189.876
-404608.610

4

145379.992
221968.042

393647.903
268393.074

.713
.410

-635902.129
-310717.914

-760219.272(1

200812.026

.000

1

5
1

-402251.562
346605.715
926662.114
754653.999
-361662.988

2

29001.447

189249.143

.879

1158775.556
-346605.715

4

174381.440

396571.976

.661

-612704.157

961467.036

5

250969.490

272663.717

.360

-290192.512

792131.491

-934600.712(1

399335.578

.021

2
3

-145379.992
-174381.440

393647.903
396571.976

.713
.661

1727171.291
-926662.114
-961467.036

5

76588.050

439873.773

.862

-796439.626

276667.804

.000

1

1

4
2
3
4
5

1
3
4
5

3

Sig.
(a)

2

3
1

Std. Error

3
4

2

Region9

Mean
Difference (l-J)

1
2
4
5
1
2
3
5

1011188.762(1
-221968.042

268393.074

.410

1560297.767
-754653.999

-250969.490
-76588.050
272772.100
796426.133(1
904046.365(1
846905.915(1
-272772.100
523654.032(1
631274.265
574133.815(1

272663.717
439873.773
206769.350
212960.704
423494.485
293405.586
206769.350
200698.292
417462.720
284630.253

.360
.862
.190
.000
.035
.005
.190
.011
.134
.046

-792131.491
-949615.726
-137607.824
373758.086
63527.041
264577.046
-683152.024
125323.478
-197273.674
9221.552

-796426.133(1

212960.704

.000

-523654.032(1
107620.233
50479.783

200698.292
420563.692
289159.261

.011
.799
.862

-904046.365(1

423494.485

.035

-631274.265

417462.720

.134

-107620.233
-57140.450

420563.692
466485.151

.799
.903

1219094.179
-921984.587
-727082.279
-523421.307
1744565.689
1459822.204
-942322.744
-982984.345

404608.610

-142030.132
635902.129
612704.157
949615.726
-462079.756
310717.914
290192.512
796439.626
683152.024
1219094.179
1744565.689
1429234.785
137607.824
921984.587
1459822.204
1139046.079
-373758.086
-125323.478
942322.744
624380.873
-63527.041
197273.674
727082.279
868703.445

158

Dependen
t variable

(1)

(J)

sp e c ie s

sp e c ie s

Mean
Difference (i-

Std. Error

Sig.

(a)

J)

5

R eqionll

1

2

3

1
2
3
4
4
2
3
4
5
1
3
4
5
1
2
4

4

5
1
2
3
5

5

-8 4 6 9 0 5 .9 1 5 0
-5 7 4 1 3 3 .8 1 5 0
-50479.783
57140.450
-53575.350
496887.597
1 1 5 1 2 3 4 .2 8 1 0
1298240.635
1 4 2 6 3 8 5 .3 8 5 0

293405.586
284630.253
289159.261
466485.151
282297.045
323529.351
333216.882
662636.390
459087.958

.005
.046
.862
.903
.850
.128
.001
.053
.002

-496887.597

323529.351

.128

6 5 4 3 4 6 .6 8 4 0
801353.038
9 2 9 4 9 7 .7 8 8 0
1151234.281 (*)

314030.044
653198.564
445357.307

.040
.223
.040

333216.882

.001

-6 5 4 3 4 6 .6 8 4 0

314030.044

.040

147006.353

658050.616

.824

275151.103

452443.788

.545

-1298240.635

662636.390

.053

-801353.038

653198.564

.223

-147006.353

658050.616

.824

128144.750

729903.333

.861

459087.958

.002

-929497.788(‘)

445357.307

.040

-275151.103

452443.788

.545

-128144.750

729903.333

.861

1
1 4 2 6 3 8 5 .3 8 5 0
2
3
4

Lower
Bound Ci
95% for
difference
(a)

Upper
Bound Ci
95% for
difference

-1429234.78
-1139046.07
-624380.873
-868703.445
-613856.841
-145228.608
489891.009
-16909.173
515222.870

-264577.046
-9221.552
523421.307
982984.345
506706.141
1139003.801
1812577.553
2613390.442
2337547.899

1139003.801
31083.974
-495065.297
45586.814
1812577.553
1277609.395
1159041.962
-622824.572
2613390.442
2097771.373
1453054.669
1320511.319
2337547.899
1813408.762
1173126.779
1576800.819

(a)

145228.608
1277609.395
2097771.373
1813408.762
-489891.009
-31083.974
1453054.669
1173126.779
16909.173
495065.297
1159041.962
1576800.819
-515222.870
-45586.814
622824.572
1320511.319

Region 5 indicates that jack pine is significantly different from lodgepole pine and
hybrid samples. However, lodgepole pine is not significantly different from hybrid
samples, perhaps indicating that the hybrids are more like lodgepole pine than jack pine
in region 5. Species groups 4 and 5, which were unclear in species classification, show
that they are only significantly different from jack pine.

159

Unlike region 5, region 9 displays no statistical difference between jack pine and
lodgepole pine, but shows that there are significant differences between hybrids and pure
species. Species group 4, made up of samples from the jack pine sampling region that
show hybrid genetic output, show a significant difference from jack pine, but not other
species groups. Species group 5, made up of samples from the hybrid area that show
lodgepole pine genetic output, show a significant difference from both pure species but
not from hybrids or species group 4.
Finally, region 11 shows that there was no significant difference between lodgepole
pine and jack pine, but hybrids show significant difference from both pure species.
Species group 4 show no significant difference from any group, and species group 5
show significant differences from both pure species, much like the hybrids.

160

5.6 Discussion and Conclusions

The primary objective of this chapter was to identify a “chemical extractive
footprint” for the species groups and to determine whether or not this footprint could be

used for differentiation between species. Variability between species ideally would take
place in the chemical components of each species, however due to the similarities
between these species groups variability was only established between quantities of the
chemical compounds found.
Figures 47 through 49 encompass the analysis of variability between species groups
according to chromatographic peak regions. The analysis reveals that not only are there
some regions within the chromatograph (representative of certain chemical groups) that
are found in unique quantities between the species groups, but the variation is sometimes
statistically significant at a = 0.05 according to the pairwise comparison conducted
(Table 20).
Considering the results from all three regions together, hybrids can be considered
significantly different from both jack pine and lodgepole pine according to the chemieal
variability of these regions. However, lodgepole pine and jack pine themselves can not
be distinguished with an a = 0.05 level of confidence. Species group 4 and 5 seem to
have chemical characteristics more like hybrids than the pure species according to these
chemical regions. Therefore, a chemical extractive footprint of distinguishing chemical
quantities may be deduced for hybrid samples, but not for jack pine or lodgepole pine
samples, according to the pairwise comparison.
To add to pairwise analysis, PLS-DA was preformed and Figure 47 was compiled
in order to show how well each sample predicts a lodgepole pine chemical make up. As

161

shown, lodgepole pine samples can be separated from the others based on this predictive
coefficient, indicating that hybrid samples may be more like jack pine samples than
lodgepole pine samples according to chemical make up.
Showing variability between chemical make up among these species groups has
been accomplished, however, several explanations exist to explain this variability. These
differences are not necessarily due to a purely genetic component. Like other
characteristics, age of sample, site conditions and growth rate, and genetics must all be
considered. A significant increase in the amount of extractives present in sapwood is
possible as the sapwood increases in age. There are also lower amounts of extractives in
heartwood of young trees than older trees. In a study conducted on Pinus echinata, it
was concluded that the age of the tree had more influence on extractive quantity than any
other environmental variable (Higuchi, 1985). This is a major factor to consider in this
study because the jack pine trees were younger than the lodgepole pine trees sampled,
and the hybrid samples had more variation in age. With this in mind, and according to
Higuchi (1985), it would be expected that lodgepole pine samples in this study would
contain more extractives than jack pine samples. Fortunately age is not the only
influence on extractive content, and variations on age influence were identified in peak
region 5 and many others not investigated in detail, where jack pine was shown to have
more extractives present than lodgepole pine.
Site conditions have been shown to affect the quantity of extractives present in
some species, but no conclusive correlations have been made to show a direct
relationship between growth rate and extractive content. The quantity of heartwood
extractives is said to be negatively correlated with the rate of diameter growth (Higuchi,

162

1985). If this were an accurate trend, it would indicate that the jack pine samples in this
study would have lower extractive contents than the lodgepole pine samples, due to a
faster DBH growth rate. As concluded in examination of age difference, peak region 5
and many peak regions not investigated in detail show that jack pine has a greater
quantity of extractives than lodgepole pine.
Figures 50a-f and the accompanying library search outputs, indicate the possible
identification of the extractive compounds studied, hi order to interpret the mass spectra
and search output, it is important to factor in the molecular weight of the compounds, the
retention time for the chemical from GC, the parent ion, the overall ion pattern, and the
purity, and fit values given by the search.
All samples had the same mass spectral outputs for each region, indicating that the
chemicals found in the samples were the same. Because the chromatographic peaks were
different, the quantitative analysis is still valid even though the species groups share the
same extractives in their makeup.
According to Fernandez et al. (2000), the chemical classes with GC retention times
approximate to those found in this study, are fatty acids and monoglycerides. The
retention time considered is from -15.0 minutes to -18.5 minutes (Figure 85). Based on
relative retention times with the internal standard, heptadecanoic acid, retention times
were matched to those found by Fernandez et al. (2000), in order to deduce possible
chemical identities of the peaks analyzed by mass spectra in this study. The retention
time of heptadecanoic acid in this study was 16.52 minutes. Peak region 5 contained two
main peak areas, one at a retention time of 15.9 and one at 16.1 minutes; peak region 9
contained one main peak at a retention time of 17.7 minutes; and peak region 11

163

contained three peaks at retention times 18.3, 18.36, and 18.4 minutes. Relative retention
times indicate that candidate chemicals for peaks in region 5 are o-coumaric acid, 4hydroxy-2-methylacetopheone, and 9-oxononanoic acid. The peak in region 9 could be
identified as eicosanoic acid according to Fernandez et al. (2000), and chemical
candidates for peaks in region 11 include only possible broad chemical classes such as;
sterol/triterpenes, diglycerides, or waxes.
Based on the output in Figures 50a-f and the library searches accompanying these
figures, potential chemical identities or chemical classes can be suggested. Figure 50a,
the first of three peaks in region 11, contains a potential match through library searches,
this chemical was identified as a stilbene. This chemical compound has a molecular
weight of 240 g/mol, which is the parent ion weight in the mass spectra. Stilbenes are
naturally oecurring phenolic constituents that originate from cinnamic acids and have
specific chemical types that are commonly found in genus Pinus (Higuchi, 1985;
Sjostrom, 1993).

o

Ph

(Landolt Bornstein, 2004).
Figure 52a: An example o f a compound similar to the stilbene identified as a candidate match.

The second and third peaks in region 11 (Figures 50b and c) have similar lists of
potential chemicals according to their mass spectral output. Both are suggestive of a type
of phenanthrene, which is related to bibenzyls and the stilbene family of compounds
(Higuchi, 1985). This phenanthrene has a molecular weight of 272 g/mol which is

164

greater than the moleeular weight of the parent ion and reflects a potential match to the
library search.

(Landolt Bornstein, 2004).
Figure 52b: An example o f a compound similar to the phenanthrene identified as a candidate match.

Peak region 9 had only one main peak in which to classify. In the library search,
naphthalene was recognized as a potential chemical candidate. According to Higuchi
(1985), naphthalenes are categorized as a type of quinone. Derivatives of this compound
were found in several gymnosperms, so it is plausible that this compound is found in
Pinus.

X

X

j

' (Landolt Bornstein, 2004).

Figure 52c: An example o f a compound similar to the naphthalene identified as a candidate match.

Lastly, peak region 5, containing two peaks of significance, was identified to
provide variability between species groups in this study. Benzoic acid was identified as a
potential candidate match to Figure 50e, and another phenanthrene is a potential match to
Figure 50f. While the library match to benzoic acid was not the closest match, and does
not show ideal numbers for purity or fit, it is a compound that was found to be present in
the study by Fernandez et al. (2000), and is found at a relatively close retention time to

165

the chemical peak analyzed in this study. The molecular weight of benzoic acid is 254
g/mol which is acceptable for a potential match.
The phenanthrene candidate to Figure 50f is another stilbene (phenolic compound),
but a different phenanthrene than the one identified in region 11. This phenanthrene has
a moleeular weight of 314 g/mol, which is also an acceptable match.

0

O''

(Landolt Bornstein, 2004).

Figure 52d: An example o f a compound similar to the phenanthrene identified as a candidate match.

Analysis on the mass spectrometry output for this study is only preliminary.
Therefore conclusive identities of the compounds can not be deduced. A more detailed
analysis of the mass spectra would allow for identification of species based on a chemical
extractive footprint.
Another possible way to positive identify the compounds in question is to run pure
samples of the candidate chemicals appearing in the library searches through the GC-MS,
and then compare their outputs to the unknowns. A matching retention time would
indicate a chemical match.
Improving the chromatographic warping due to stretch would also improve the
accuracy of this study, and enable more information to be gathered from the
chromatograms. According to Bylund et al. (2002), peak position and peak width often
show variation, other than that caused by sample difference, due to output of a two

166

dimensional figure from a trilinear model. Alignment variation can be compensated for
with slight peak adjustments as were made in this study. However, warping and stretch
are more difficult to balance. Warping is caused by the age of the column used and flowrate variation, therefore affecting the retention time of a chemical compound. Varied
retention time for the same compound over several samples causes peaks that should be
the same to appear skewed or stretched. Correlation optimized warping (COW) of
chromatographic figures was designed to compress or expand peak profiles by linear
interpolation to reduce variation in retention times (Byund et al., 2002). By applying
COW to this study, more accurate representations of chemical profiles may be
determined.
A preliminary assessment of chemical makeup of jack pine, lodgepole pine, and
their hybrids has been provided in this chapter to act as further evidence of differentiation
between species groups and compliment the findings in chapters 2, 3, and 4.
Quantitative variability does exist between jack pine, lodgepole pine, and hybrid
samples. Preliminary analysis indicates that wood extractives may place the samples in
this study more like jack pine than lodgepole pine. However, chemical quantities thus far
do not demonstrate the intermediacy theory in hybrids to an extent that would be
statistically significant. It is possible that site conditions, age of samples, and genetic
heritability all played a role in the established extractive content of the samples
examined; due to which a clear “chemical extractive footprint” can not be determined.
Further manipulation of chromatographs and more extensive investigation of mass
spectra are required.

167

5.7

References

Bylund, D., R. Danielsson, G. Malmquist, and K.E. Markides. 2002. Chromatographic
alignment by warping and dynamic programming as a pre-processing tool for
PARAFAC modeling of liquid chromatography-mass spectrometry data. Journal
of Chromatography A, 961: 237-244.
Fernandez, M., C. Breuil, and P. Watson., 2000. “Determining Wood Extractive
Compounds Simultaneously Using GC-MS” Pulp and Paper Research Institute of
Canada PPR 1487. Pointe-Claire, Quebec.
Higuchi, T. 1985. Biosynthesis and Biodégradation of Wood Components. Academic
Press, Inc. Toronto, CA.
Lewis, I.R., and H.G.M. Edwards, 2001. Handbook of Raman Spectroscopy: from the
research laboratory to the process line. Marcel Dekker, Inc: New York, NY.
Landolt Bornstein, 2004. Landolt Bornstein Organic Online Index. Springer Berlin
Heidelberg. Accessed online at http://lb.chemie.uni-hamburg.de/static/index.html,
on Nov. 28, 2005.
Saariaho, A., and J. Vyorykka, 2002. Raman and IR Methods in Paper Industry.
Spectroscopy seminar, Helsinki University of Technology, Laboratory of Forest
Products Chemistry. Finland.
Sjostrom, E., 1993. Wood Chemistry: fundamentals and applications. 2nd ed.
Academic Press, Inc. Toronto, CA.

168

Chapter 6 : Discussion of and Applications for Hybridization of Lodgepole and Jack Pine

6.1 Hybridization
6.1.1 Predictive Abilities of Characteristics
Chloroplast DNA from the matk gene and mitochondrial DNA from the nad\ gene
both successfully identified parental lineage of the lodgepole pine, jack pine, and hybrid
samples in this study.
Using knowledge of the predominant species for each sampling area as well as the
genetic information gathered using cpDNA and mtDNA, a comparison of species groups
based on other morphological, cell and fibre, and chemical characteristics was attempted.
This comparison was preformed in Chapters 3, 4, and 5 in order to determine if any of
these characteristics are capable of distinguishing between species, or between hybrids
and species. It was also interesting to determine if any of the characteristics placed
hybrid samples as intermediate of the two pure species, thus illustrating what is
happening genetically, the combination of genetic material 50% from jack pine, and 50%
from lodgepole pine. Morphologically, needle pair size represented as a ratio of needle V
width/needle length, and cone angle of attachment were the most useful characteristics
for differentiating between all species groups (including the distinction of hybrids) while
also demonstrating intermediacy in hybrids. Cone length had the ability to distinguish
hybrids from pure species, but no discriminatory abilities were noted for differentiation
between jack pine and lodgepole pine. Cone length did not demonstrate intermediacy.
Wood and fibre traits revealed that modulus of elasticity was the most useful
characteristic for discriminating between species groups, and also showed hybrid samples
to be intermediate. Cell wall thickness and fibre coarseness distinguish pure species from
169

each other, and distinguishes between hybrids and lodgepole pine, but not between
hybrids and jack pine, suggesting that these two latter species may be more similar based
on these two characteristics. Cell wall thickness and fibre coarseness do not demonstrate
intermediacy. Microfibril angle was useful in this study for distinguishing between pure
species but not separating out hybrid samples. Lastly, moisture content, DBH growth
rate, and earlywood/latewood contents were seen to distinguish between species groups
but are subjective in their interpretation due to their strong relationship with site
conditions.
Chemical analysis via GC-MS revealed that peak regions 5, 9, and 11 have
characteristic quantities of chemicals for pure species and hybrids, therefore
distinguishing hybrids from pure species groups. PLS-DA suggested that chemical
makeup of hybrids may he more like that of jack pine than lodgepole pine. No
intermediacy of hybrids was shown through chemical analysis.
Since the characteristics mentioned above possess the ability to distinguish between
species groups, it is possible to use them as predictive tools for hybridization. If a certain
characteristic is measured, and falls within the measurement range that is characteristic
for a hybrid as opposed to a pure jack pine or lodgepole pine, then assumptions can be
derived that the species is a hybrid without having to conduct any genetic analysis. This
information can then be used by managers or manufacturers to alter their practices to best
suite the material they are working with. Managers may adjust growing conditions to
account for trees with faster or slower growth rates, and manufacturers may alter
production to account for varying MOE or latewood content in raw wood. Wood that is

170

being used for pulp products may be sorted according to variation in fibre coarseness or

quantity of chemical extractives in order to be processed more accurately.
Hybridization is not only important to distinguish from a production point of view,
but also from a scientific point of view. Hybrids occur naturally between many different
organisms, and cause controversy over the definition of a species. A species is defined as
a group of organisms that interbreed but are reproductively isolated from other
populations (Barton and Hewitt, 1985). Since hybrids are the product of two supposedly
reproductively isolated populations, it is not certain whether hybrids should be classified
as a new species, or if original taxonomic classification should be considered inaccurate.
Taxonomists and other researchers can use the information in this study to assign
characteristic limitations to new species and subspecies groups being developed.

6.1.2 Hypotheses and Zoning
Two hypotheses exist regarding the presence of a hybrid introgression zone
between jack pine and lodgepole pine in Alberta and Northern EC. Hybridization; the
crossing of genetic information between species, occurring where the two pure species
ranges overlap, is the first hypothesis. The alternate suggests that jack pine and
lodgepole pine diverged from one species into the two species that exist today.
Understanding hybrid zones and species movement helps to better define which
hypothesis may be more accurate.
W oodruff (1973) classified hybrid zones according to the pattern of species
distribution (Figure 1). Allopatric zones feature natural hybrids that occupy a narrow
area between pure parental species; sympatric zones feature hybrids and parental species
in the same areas. Sympatric zones can be broken down further into parapatric zones;

171

parental species remain localized (segregated) while hybrids integrate with both parental
species at their borders. Parapatric hybridization is the closest model to describe the
pattern of species distribution among jack pine, lodgepole pine, and their hybrids. It has
been suggested by Barton and Hewitt (1985) that allopatric zones indicate wellequilibrated populations, while sympatric populations indicate populations that are still
exchanging large amounts of genetic information. This is directly related to the length
time these populations have been hybridizing; stability in populations is said to be
achieved over time. Applying this to parapatric populations in this study suggests that
genetic information is still being exchanged between study groups and equilibrium has
not been reached. This may also mean that these populations have only been hybridizing
for a relatively short period of evolutionary time.
Opposing the hypothesis of hybridization between jack pine and lodgepole pine
exists the hypothesis of species divergence. Divergence implies that over evolutionary
time, jack pine and lodgepole pine were produced from one species, and have been
continually evolving into two species due to environmental influences. This explains the
presence of some trees that exhibit characteristics and genetic information from both jack
pine and lodgepole pine, however, hybridization may better explain this finding.
Based on the evolutionary time required to diverge a species into two separate
species, with only a d in e of hybrids left as a record of the original, it is improbable that
this divergence took place. Barton and Hewitt (1985) suggests that divergence of two
populations could not take place without a “ .. .virtually complete barrier..

the Rocky

Mountains could be seen as such a barrier, however, pines do cover this expanse, and a
larger physical barrier does not exist. Hybridization or divergence could only have taken

172

place since the last ice age. In order to have almost completely diverged, with no solid
genotype or phenotype of a predecessor species left, these pines must have been evolving
very quickly. Hybridization however, could have begun after the last ice age, as jack
pine and lodgepole pine ranges grew closer together, and therefore created the d in e of
phenotypes that are in existence today. Lastly, Danick and Yeh (1983) calculated the
genetic distance of jack pine and lodgepole pine to be much greater than genetic
distances within species, supporting the differentiation between these two pines. If
divergence has occurred over time, it might be expected that the genetic distance between
species would be more similar to within species variation, or smaller than other between
species distances.
These hypotheses cannot be verified without a close inspection of the fossil record,
and in depth genetic analysis of multiple hybrid populations to determine genotype
relationships. Even then it may be impossible to say if one hypothesis is more valid than
another. Based on geographic range alone, it is shown that jack pine and lodgepole pine
having been migrating towards one another, now virtually undivided rather than
migrating away from one another into their own environments over time.

6.1.3 Maternity and Paternity Issues: Effects on Productivity
Maternal and paternal genotypes contribute directly to characteristics such as
germination percentage, rate of growth, and fitness of the tree even though environmental
conditions also play a role. Productivity of a mature tree can be linked to the genetic
information contained within the endosperm, or nutritive tissue, of the seed and within
the embryo. The endosperm is formed from only maternal genes, and regulates the initial
growth rate of the germinant. The embryo is composed of both maternal and paternal

173

genetic information. These genotypes determine growth rates whieh affect the
germinant’s ability to compete for nutrients and light. The fittest trees will out compete
other seedlings and become stronger in later stages, produeing trees with greater volume
at maturity up to a certain age (Fries et. al., n.d.). This supports the idea that individual
trees originating from a hearty, more evolve maternal line have an increased chance of
producing higher volumes of quality wood because of an initial growth advantage.
Therefore, strong, hearty trees that are involved in hybridization will undoubtedly
produce seedlings with these genotypes that can better compete for survival.
Hybrids are produced by evolutionary processes, and recombination of genetic
information allows for improved survival for select populations. These populations go
on to produce Fn generations of individuals with improved tolerance of various
environments as well as a variety of other adaptations to combat pests, pathogens, or
other factors that may threaten survival of the species. In a natural environment, this
process takes thousands if not millions of years to significantly change a population.
However, today, lab controlled crosses can mimic or interpret the advancement of a
species in order to produce populations of hybrids that possess an array of characteristics
for various purposes. Hybrids can provide improved growth rate, improved frost
tolerance for eold climate plantations or general northern environments, improved wood
quality, improved crown shape and size, or improved pest or pathogen resistance
(Savolainen and Kàrkkâinen, 2004). Hybrid vigor is not displayed by intermediate triats,
as were seen in some of the hybrid eharaeteristies identified in Chapters 3 and 4.
Therefore, although intermediacy in eharaeteristies is more effective for hybrid
identification, it is not always desirable for traits that may provide increases in yield.

174

quality, or tolerance when managing for anthropological use. For example, intermediate
MOE is not as desirable as hybrids that demonstrate higher MOE for dimensional lumber
manufacturing.
Natural hybrids can provide seed sources with already evolved genetie material.
The lodgepole pine x jaek pine hybrids that exist in the Northeast corner of British
Columbia, exhibit characteristics varying from the parental species, and may therefore
offer increased value to some products, depending on what traits are being sought. These
trees may have a greater potential for frost resistance than other pines produced due to
the latitude at whieh they were developed. Frost resistance is an important trait to
consider when developing a plantation. Trees that are damaged by frost multiple times
become shrub-like with no vertical growth and very little secondary growth. These
populations are of low quality and therefore low value due to excessive stem
deformation, branchiness, and laek of wood volume. Results from a study conducted in
Sweden revealed that traveling 1° in latitude northward (in the northern hemisphere)
caused a 10 % decrease in survival rate for Scots pine (Pinus sylvestris). Similarly,
increases of 100m in elevation lead to a 3% reduction in survival (Savolainen and
Karkkainen, 2004). For these reasons, it is pertinent for forest managers to select the best
seed possible for reforesting more northern regions of the world, such as the areas
sampled for this study.
One down fall to production of hybrid crosses, both naturally and artificially, is
their reproduetive capabilities. Aeeording to Critehfield (1985), lodgepole pine and jaek
pine can be very easily erossed in a laboratory setting. Crosses that are performed with
lodgepole pine as the female parent result in approximately 30% germinal seeds, which is

175

very low. Also, putative hybrids in Alberta have been observed to produce aborted
pollen grains or pollen grains that are smaller than non-hybrids. Forty-two percent of
putative hybrids have aborted pollen as compared to only 1 - 2 % in pure species, making
hybrid proliferation difficult (Critehfield, 1985).

6.2 Applications
6.2.1 Hybridization Effects on Management for Wood Quality
Information regarding the specific chemistry, wood fibre properties, and growth of
hybrids can be applied to management of cold climate plantations to ensure optimum
yield. Hybrids have the potential to grow faster than predecessor species as they are an
evolved combination of the two parents. Management practices can be altered to
accommodate more specific needs of each species cross rather than assuming lodgepole
pine or jack pine criteria. For instance matching an area with the most adapted variation
of pine hybrid when reforesting, to ensure maximum growth and eventually yield.
Hybrids will be heartier, faster growing, better at adapting to environmental
conditions, and possibly even more resistant to pathogens and pests than pure species are.
Evolving hybrids of two pre-existing species will allow the tree to take on the most
advantageous characteristics of the predecessors.
Wood that is more uniform is generally considered better quality. Uniformity,
meaning the absence of knotting, scarring, or reaction wood, and a consistent pattern of
earlywood versus latewood rings, can be improved with management techniques.
Silvicultural treatments such as site preparation and fertilization allow seedlings and
immature trees to allocate energy towards stem growth rather than defense mechanisms
or root expansion. Fertilizers have generally been used to increase growth rates, and site

176

preparation is generally used to eliminate competitive species and improve soil drainage.
In pines such as jack and lodgepole pine, wood produced from nitrogen fertilization
results in short cells with thin walls and therefore lower density (Zobel, 2004).
Other silvicultural techniques that can be applied to change wood quality for
specific products include spacing and pruning. Varying stand density by
spacing/thinning changes the ring widths (Zobel, 2004) and EW:LW ratios, and therefore
affects manufacturing and processing because ring width is directly associated with wood
density and wood uniformity. Pruning stands eliminates lower branches, forcing the tree
to allocate energy towards stem growth rather than branch growth, resulting in higher
quality, larger stems for solid wood production (Zobel, 2004). Pruning is not as common
in pine stands, because generally pines will self-prune after a certain age.
Wood quality is said to be strongly inherited and therefore using genetically
improved tree stock through methods such as hybridization and selective breeding, in
combination with silvicultural management techniques, has potential to greatly improve
wood uniformity as well as other wood eharaeteristies (van Buijtenen, 2004).
Operationally, several methods can be used to genetically manipulate seed stock for
improved wood quality traits. Using clonal forestry or seed orchards are the most
common approaches. Clonal forestry uses vegetative propagation to reforest areas using
root cuttings or other tissue cultures from one or two parents. This method can be
expensive, so often other methods are sought (van Buijtenen, 2004) however clonal
populations have been used very successfully for low costs with Eucalyptus hybrids
(Potts, 2004). One method that exists for clonal regeneration of improved seedlings is
somatic embryogenesis. This method involves utilization of asexual clone embryos from

177

a single seed source. Embryos are grown on culture plates and then transferred into small
seedling containers. These seedlings are genetically identical and have been used for
successful spruce plantations, and more recently investigated pine plantations
(Klimaszewska, 2005). This method is successful in plantation based areas but is not
implemented in natural reforested areas such as forests in British Columbia for key
reasons. These include the limitations that clonal forestry poses on biodiversity, and
risks associated with reforesting large areas of land with only one or two genotypes, such
as pest or pathogen infestation, or poor resistance to elements such as drought or frost.
Seed orchards also utilize one or two parents with the most desired qualities, but
graft these individuals in an orchard. Orchards are then inspected, and trees displaying
undesirable traits are discarded. Seed produced by the remaining orchard trees is used
for operational needs (van Buijtenen, 2004).
Learning from naturally hybridized forest stands such as the jack pine x lodgepole
pine stands investigated in this study will enable the scientific community to better
understand what these populations have to offer the forest industry, and how they should
be managed for various purposes. Since these natural populations have been growing
and evolving over a long period of time, they lend insight into affects that hybridization
may have on environments and wood quality that are not obvious from newly formed
plantation based progeny.

6.2.2 Hybridization effects on Manufacturing and Processing
Wood fibre properties looked at determined if hybrids were of better quality for
manufacture and processing of wood or pulp than lodgepole or jack pine for certain
products. Knowledge of fibre properties of the hybrids is useful for processing.

178

Manufacturing and processing could be optimized for specific hybrid characteristics of
the wood, rather than using generic settings and assumptions of pure species to
manipulate the boards or fibres.
Equipment and methods for processing and manufacturing wood products is often
altered in order to more efficiently utilize various types of wood. With this increased
efficiency however, costs can increase or the quality of the final products can be
compromised. Production becomes stable and effective when raw materials are relatively
uniform (Zobel, 2004). It may be possible to find and manipulate populations or closely
related species by mimicking natural hybridization as seen in jack pine and lodgepole
pine, in order to produce a hybrid that is genetically improved compared to its
predecessors. If this is possible, this manipulation could lead to more uniform wood for
processing and manufacturing of wood products.
Wood quality needs vary depending on what is being produced. For example, a
low density stem may be considered less valuable than a higher density stem, however, if
the desired product is writing paper or tissue, than low density is actually preferred.
Higher density wood is better for fibre boards or increased pulp yields (Zobel, 2004). All
of the other wood and fibre characteristics investigated in this study have various impacts
on the quality of the end product. Low microfibril angles are important for product
stability in solid wood products and quality of paper products, and thin-walled cells as
well as large lumens are important for production of tissue paper (Zobel, 2004).
Hybridization of species and populations offers evolved variations in wood, allowing for
differences in wood characteristics, which may improve quality for specific products.

179

Wood characteristics under genetic control, as confirmed to date, include latewood
percentage, cell dimensions, chemical properties, and microfibril angle. Other traits are
being investigated for their genetic applicability. Latewood percentage ranges in
heritability depending on the population in question, however little more is known about
EW:LW genetic relationships. Tracheid length, diameter, and cell wall thickness are all
strongly inherited; variation in populations has been noted in lodgepole pine. Chemical
properties have been studied in relation to genetic inheritance since the 1970’s, and
genetic improvements in these areas have been accomplished. Lignin content has been
studied most extensively because of its applications in the paper making process, while
less is known about genetic improvements to cellulose eontent. Lastly, very little is
known about microfibril angle heritability, this trait is still under investigation (van
Buijtenen, 2004).
Knowledge of natural hybridization could potentially lead to management for select
traits in order to obtain the best wood quality output for certain products and for optimal
processing and manufacturing. This could potentially reduce waste wood, increase yield,
and allow for more cost effective production.

6.3 References
Barton, N.H., and G.M. Hewitt, 1985. “Analysis of Hybrid Zones”. Ann. Rev. Ecol.
Syst. 16:113-148.
Critehfield, William B., 1985. “The late quaternary history of lodgepole and jack pines” .
Canadian Journal of Forest Resources. 15: 749-772.
Danick, B.P. and E C. Yeh, 1983. “Allozyme variability and evolution of lodgepole pine
{Pinus conforta var. latifolia) and jack pine {Pinus banksiana) in Alberta” Can. J.
Cenet. Cytol. 25: 57-64.

180

Fries, A., C. Raymond, and Q. Wang, not dated. “Nursery results of controlled crosses
between lodgepole pine plus trees”. Dept, of Forest Genetics and Plant
Physiology. Swedish University of Agricultural Sciences. Umea, Sweden.
Klimaszewska, K. 2005. “Somatic embryogenesis; a conifer reproduction method”.
Canadian Forest Service, Laurentian Forestry Centre. Accessed Dec. 19, 2005 at
http://www.cfl.scf.mcan.gc.ca/CFLLFC/puhlications/eclaircie/emhryogenese_e.html
Potts, B. M. 2004. Genetic Improvement of Eucalyptus. Encyclopedia of Forest Science
pp. 1480-1490, Elservier Major Reference Works, Oxford.
Savolainen, O. and K. Karkkainen, 2004. Breeding and Genetic Resources of Scots Pine.
Encyclopedia of Forest Science pp. 1517-1520, Elservier Major Reference Works,
Oxford.
van Buijtenen, J.P. 2004. Genetics and Improvement of Wood Properties. Encyclopedia
of Forest Science pp. 1466-1472, Elservier Major Reference Works, Oxford.
Woodruff, D. 1973. “Natural hybridization and hybrid zones”. Syst. Zool. 22: 213-218.
Zobel, B. 2004. Biological Improvement of Wood Properties. Encyclopedia of Eorest
Science pp. 1458-1466, Elservier Major Reference Works, Oxford.

181

Appendices:
Appendix
Site maps
AOkm

63

{■'

i Islam ) L a k e

4

I'lainqndOBV
------------

IJslaiK l L a k e S o u th

L a c U B ie h e

■f;

^ C a s la n

E f is c o tt

Lake

x g ) —.

®

°

Ooodtidge
McRae

iNewWook
/
I

o
N##twv

^

Cove

Good Wi Lake^

S t , L in a

I

M allaig

VBfclP6nflL«iw
Wars^

EyMiwnt i. o"^
f.--" Radwy
e d w a te r
, R--------

1

,

-, ^«99l

It. Paul
Andrew

r

i

@

c

. ' »

-

r < % M ic h a e l ^

W lllingdon

Brosseau

H airy Hill

Fort S à s
Mymam

A R iort,

Mundem

Beauwlon
V e g m v llle

Devon

Beaumont

I
^
N e w S a re m ta
i
V tw .
I
I
'
Figure i: Map showing the location o f Smoky Lake (star) with respect to Edmonton, Alberta. Sampling
sites fall east o f Smoky Lake, towards B eilis (MapQuest, 2005).

182

Mkm

h>B eilis

^)S ite # 2
{28 —• - ."
^

Figure ii: An enlarged view o f the jack pine sampling sites, circled. Smoky Lake is west on Highway 28
(MapQuest, 2005).

183

Vandahoof. * Piince
Foœsi*,* George

\bnderhoof

District

Qucuk

Nomwn

15

20

Figure iii: Map o f Bobtail FSR and Gregg Creek FSR west o f Prince George. Sites 4, 5, and 6 are circled
(Sanborn et. al., 2001).

184

Si..*,

JvAtr

1

%KM5
WWWLWk'

A'
!(;fM#o»i wm

&5ZA%*

Figure iv; Site map for Fort N elson area sampling sites. Sites are circled.

185

POINT OF CCmmgNCE
1.30 km i

FJÎR786M1
180X)04)0
!wo m

S ile # 7

JACKPINE REMNANT PROTECTED AREA
WMN##

^

Figure v: An enlarged and more detailed view o f the site 7 location. Forest Service Road (FSR) 7866-01 is
the Patry Mainline.

f
/ 8M74)0t
zaWmt
Site

iMmyHAUNWWE:
f

JACKPINE REMNANT PROTEC

W

I

0

\
\

' ^

------------------ w

.

JACKPINE REMNANT PROTECTED AREA
M#*

,

Figure vi: An enlarged and more detailed view o f site 8.

186

References;
MapQuest® Inc. 2005. www.mapquest.com Accessed November 6 , 2005.
Sanborn, Paul, Rob Brockley, and Caroline Preston. 2001. Effects of Sitka Alder
Retention and Removal on the Growth of Young Lodgepole Pine in the Central
Interior of British Columbia. Establishment Report, EP 1185. Ministry of Forests
Research Program, British Columbia.

187

Appendix 2:
Genetic Sequences for CpDNA matk gene markers in Pinus banksiana (restriction sites
are underlined, primers are italicized):
i.

CpDNA Section 1

1
atggatgagt
61 ccactctttt
121 agttcctcca
181 aaacgtttga
241 gatccaaatc
301 cttacattgg
361 aatgaatcga
421 ccgcattcaa
481 gttcgaacct

ii.

t c c a t a g a t g cggaaaggaa gatagctttt ggcaacaatg ctttttatat
ttaaggaaga tctttacgca atttctcatg
gaccgatgga acatttaagt tccaatgatc
ttggtcaaatacgtaaacaa aatcattcaa
cattagctga tcgcaagaag agtttctatt
tcctggaagt tccgttctct atatggtcaa
agagtttccg gtcgatccat tcaatatttc
attctatatt agatgcacga ataccctatt
ttcgtcgctg gatccgagat gctccctcct

atcattattt ggatgtatca
aattaagttt cctaactgta
ttgttttatt cgygaattgc
ctgaatcggt actagaagca
aatattctgt ggaagggatg
ccttcttaga ggataaattc
ctattcatcc ggaaattttg

tgcacccatt acgatctgtt

CpDNA Section 2

10 81
1141 a t a t c a g g g c
1201 gatcgattcg
1261 gatggtttat
1321 aaacataaaa
1381 tcgttttcaa
1441 tcgcagagag
1501 tggcaaaaga
1561 gtaattgcct
1621 aatagataca

attctgtgac
ggccaattag taaattgtct tggaccagtc
atcaaatttg gagaaatctt tttcattact
atcgtataaa gtatatactt tcattatcat
gtacgatacg tgtagttcgg aaggaattag
aagaacgaga atttgattct ctgcgctttt
aacgaatttg gcattcagat attccccaga
tacaggatct taaaatagaa aacttatttg
cgattcagaa tcatttttat ttttctatcc
ttacatgggg a a a g c c g t g t g c a a t g a g a a

taacagatga tgatatcctc
acagtggatc ctttgatcga
gtgctaaaac tttagcctgt
gtccagaact ctttaaaaaa
catcaaaagc ggcggcccgt
taaatcccct agctaattcc
accaatgaaa tgctctttga
gagaactaaa atgattagga

Table i: Haplotype chart for determining the maternity of samples based on RFLP
categorization from digestions with R sa \, M bo\, and H ha\ using the mtDNA nadl gene.
le

#
2
5
8
9
13
14
15
16
18
19
20
21
22
23
24

s p e c ie s
area
Pj
Pj
Pj
Pj
Pj
Pj
Pj
Pj
Pj
Pj
Pj
Pj
Pj
Pj
Pj

Rsa\

Mbo^

Hha^

A
A
A
A
A
A
A
A
A
B
A
A
A
A
A

A
A
A
A
A
A
A
A
A
B
A
A
A
A
A

A
A
A
A
A
A
A
A
A
B
A
A
A
A
A

Maternity
by
h aplotype
Pj
Pj
Pj
Pj
Pj
Pj
Pj
Pj
Pj
Pli
Pj
Pj
Pj
Pj
Pj

188

le
25
26
27
28
29
30
31
32
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
65
66
67
68
69
70
71
72
73
75
77
78

species
area

Rsa^

Mbo^

Hhs\

Pj
Pj
Pj
Pj
Pj
Pj
PI
PI
PI
PI
PI
PI
PI
Pi
PI
PI
PI
PI
PI
PI
PI
PI
PI
PI
PI
PI
PI
PI
PI
PI
PI
PI
PI
PI
Px
Px
Px
Px
Px
Px
Px
Px
Px
Px
Px
Px
Px
Px

A
A
B
B
B
A
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B

A
A
B
B
B
A
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B

A
A
B
B
B
A
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B

Maternity
by
haplotype
Pj
Pj

Pli
PH
Pli
Pj

Pli
PI
PI
PI
PI
PI
PI
PI
PI
PI
PI
PI
PI
PI
PI
PI
PI
PI
PI
PI
PI
PI
PI
PI
PI
PI
PI
PI
PI
PI
PI
PI
PI
PI
PI
PI
PI
PI
PI
PI
PI
PI

189

Maternity
e
79
80
81
83
85
86
87
88
89
90

species
area
Px
Px
Px
Px
Px
Px
Px
Px
Px
Px

Rsa^

Mbo^

Hha\

B
B
B
B
B
B
B
B
B
B

B
B
B
B
B
B
B
B
B
B

B
B
B
B
B
B
B
B
B
B

by
ha
PI
PI
PI
PI
PI
PI
PI
PI
PI
PI

Table ii; Haplotype chart for determining the paternity of samples based on RFLP
categorization from digestions with Snabl and H ha\ using the cpDNA matk gene.

sample
#
2
5
8
9
13
14
15
16
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
35
36
37
38
39

species
area
Pj
Pj
Pj
Pj
Pj
Pj
Pj
Pj
Pj
Pj
Pj
Pj
Pj
Pj
Pj
Pj
Pj
Pj
Pj
Pj
Pj
Pli
Pli
Pli
Pli
Pli
Pli
Pli

Snabi
A
A
A
A
B
B
A
A
A
B
A
C
C
C
A
A
C
A
A
A
A
D
D
D
D
D
D
D

Hbal
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
B
B
B
B
B
B
B

paternity
by
haplotype
Pj
Pj
Pj
Pj
Pj
Pj
Pj
Pj
Pj
Pj
Pj
Pj
Pj
Pj
Pj
Pj
Pj
Pj
Pj
Pj
Pj
Pli
Pli
Pli
Pli
Pli
Pli
Pli

190

le
#
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
65
66
67
68
69
70
71
72
73
75
77
78
79
80
81
83
85
86
87
88
89
90

species
area
PI
PI
PI
PI
PI
PI
PI
PI
PI
PI
PI
PI
PI
PI
PI
PI
PI
PI
PI
PI
PI
Px
Px
Px
Px
Px
Px
Px
Px
Px
Px
Px
Px
Px
Px
Px
Px
Px
Px
Px
Px
Px
Px
Px
Px

Snabi
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
D
D
D
D
D
D
D
D

Hhal
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
B
B
B
B
B
B
B
B

paternity
by
haplotype
PI
PI
PI
PI
PI
PI
PI
PI
PI
PI
PI
PI
PI
PI
PI
PI
PI
PI
PI
PI
PI
Pj
Pj
Pj
Pj
Pj
Pj
Pj
Pj
Pj
Pj
Pj
Pj
Pj
Pj
Pj
Pj
Pli
Pli
Pli
Pli
Pli
Pli
Pli
Pli

191

Appendix 3: Statistical Output

Morphology:

General Linear Model
Between-Subjects Factors

Descriptive Statistics

n eed le_ w j

cone length
(mm)

Height/Age

DBH/Age

cone angle

Std. Deviation

N

1 sp ecies
1

Mean
.487527

.1258983

25

2

.168409

.0895083

18

3

.406732

.1083569

18

4

.481770

.0850018

4

5

.222111

.0794389

10

Total

.355852

.1704308

75

1

38.326

6.5674

25

2

35.256

5.0294

18

3

41.757

6.1080

18

4

29.646

.9869

4

5

41.238

4.2827

10

Total

38.338

6.4043

75

1

.287649

.0422417

25

2

.292295

.0465218

18

3

.270460

.0724188

18

4

.315020

.0487452

4

5

.267249

.0296795

10

Total

.283378

.0513766

75

1

.485633

.1020937

25

2

.299126

.0645270

18

3

.332194

.0787029

18

4

.446961

.0713601

4

5

.415617

.0651152

10

Total

.392648

.1124738

75

1

-4.630075

9.8775189

25

2

56.638702

6.5662406

18

3

40.246528

20.1071602

18

15.5378457

4

4
5

12.260290
60.049534

Total

29.061819

19.6343089

10

31.8129985

75

192

T e s t s o f B e tw e e n -S u b je c ts E ffe cts

Source
Corrected Model

Intercept

Dependent Variable
needle_w_l
cone length (mm)

Type III Sum
of Squares
1.355(a)

F
29.834

SIg.
.000

4

191.946

5.926

.000

4

.003

1.095

.366

DBH/Age
cone angle

■456(d)
60751.523(e)

4

.114

16.653

.000

4

15187.881

75.180

.000

needie_w_i

6.228

1

6.228

548.559

.000

69203.916

69203.916

2136.517

.000

Height/Age

4.096

1
1

4.096

1559.704

.000

DBH/Age

7.820

1

7.820

1141.145

.000

39137.894

1

39137.894

193.732

.000

needle_w_l

1.355

4

.339

29.834

.000

767.783
.011

4

191.946

5.926

.000

4

.366

4

.003
.114

1.095

.456

16.653

.000

cone angle

60751.523

4

15187.881

75.180

.000

needle_w_l

.795
2267.370
.184

70
70

.011

70

cone angle
needle_w_l

.480
14141.426
11.647

70
70
75

cone length (mm)

Height/Age
DBH/Age

cone length (mm)
Height/Age
DBH/Age

Corrected Total

.339

.011(c)

cone length (mm)

Total

Mean Square

767.783(b)

cone angle

Error

4

Height/Age

cone length (mm)

species

df

113270.260

75

Height/Age

6.218

75

DBH/Age

12.499

75

cone angle

138237.150

needle_w_l

2.149

75
74

cone length (mm)

3035.153

74

Height/Age
DBH/Age

.195

74

.936

74

cone angle

74892.949

74

32.391
.003
.007
202.020

R Squared = .630 (Adjusted R Squared : 609)
R Squared = .253 (Adjusted R Squared : 2 1 0 )
R Squared = .059 (Adjusted R Squared = 005)
R Squared = .488 (Adjusted R Squared 458)
e R Squared = .811 (Adjusted R Squared = .800)

193

Estimated Marginal Means
1. Grand M ean

95% Confidence Interval
Dependent Variable
needle_wj

Mean
.353

.015

.323

.383

cone length (mm)

37.245
.287

.806
.007

35.637

38.852

.272

.301

Height/Age

Std. Error

Lower Bound

Upper Bound

DBH/Age

.396

.012

.373

.419

cone angle

28.009

2.012

23.995

32.022

2. species
Estimates
Dependent
Variable
needle_w_l

cone length
(mm)

Height/Age

DBH/Age

cone angle

95% Confidence Interval
species
1

Mean

Std. Error

Lower Bound

Upper Bound

.488

.021

.445

.530

2

.168

.025

3

.407

.025

.118
.357

.218
.457

4

.482

.053

.376

.588

5

.222

.034

.155

1

38.326

2

35.256

1.138
1.341

.289
40.597

3
4
5

41.757
29.646
41.238

1.341
2.846
1.800

1

.288
.292

.010
.012

.267

.308

2

.268

.316

3

.270

.012

.246

.295

4

.315

.026

.264

.366

5
1
2

.267

.016

.235

.300

.486

.017
.020

.453
.260

.519
.338

3

.332
.447

.020
.041

.293
.364

.371

4

.299

36.056
32.580
39.082
23.970
37.648

37.931
44.433
35.321
44.827

.530

5

.416

.026

1
2

-4.630
56.639
40.247

.363
-10.300
49.957

.468
1.039
63.320

33.565
-26.434

46.928
1.914

51.085

69.014

3
4

-12.260

2.843
3.350
3.350
7.107

5

60.050

4.495

Univariate T ests

194

Sum of
Squares

Dependent
Variable
needle_w_l

df

Mean Square

Contrast

1.355

4

.339

Error

.795

70

.011

con e length
(mm)

Contrast

767.783

4

191.946

Error

2267.370

70

32.391

Height/Age

Contrast

.011

4

.003

Error

.184

70

.003

DBH/Age

Contrast

.456

4

.114

Error
cone angle

.480

70

.007

Contrast

60751.523

4

15187.881

Error

14141.426

70

202.020

F

Siq.

29.834

.000

5.926

.000

1.095

.366

16.653

.000

75.180

.000

The F tests the effect of sp ecies. This test is based on the linearly independent pairwise comparisons
among the estimated marginal means.

Post Hoc Tests
species
Multiple Comparisons

Dependent
Variable
needle_w_l

(1) sp ecies
1

(J) sp ecies
2
3
4

2

4

5

1

Sig.
.000
.114
1.000

95% Confidence Interval
Lower and upper
.226892
.411345
.173023
-.011431
-.154910

.166425

.005758

.0573781

.265417(3

.0398668

.000

.153783

.377050

1

-.319118(3

.0329363

.000

-.411345

-.226892

3

-.238322(3
-.313360(3
-.053702

.0355161
.0588969
.0420233

.000
.000
.705

-.337773
-.478281
-.171373

-.138872
-.148440
.063970

1

-.080796

.0329363

.114

-.173023

.011431

2

.238322(3

.0355161

.000

.138872

.337773

4

-.075038

.0588969

.708

-.239958

.089882

5

.184621(3
-.005758
.313360(3

.0420233
.0573781
.0588969

.000
1.000
.000

.066949
-.166425
.148440

.302292
.154910
.478281

3

.075038

.0588969

.708

-.089882

.239958

5

.259659(3

.0630349

.001

.083151

.436166

1

-.265417(3

.0398668

.000

-.377050

-.153783

2

.053702
-.184621(3
-.259659(3

.0420233
.0420233
.0630349

.705
.000
.001

-.063970
-.302292
-.436166

.171373
-.066949
-.083151

3.071

1.7593

.413

-1.855

7.997

3

-3.431

1.7593

.301

-8.357

1.495

4

8 .680(3

3.0649

.046

.098

17.263

1
2

3
4
cone length
fmml

Std. Error
.0329363
.0329363

5

4
5
3

Mean
Difference (1J)
.319118(3
.080796

2

195

Dependent
Variable

(1) sp ecies

(J) sp ecies
5

2

1
3
4

Mean
Difference (1J)
-2.911
-3.071
- 6 .5 0 2 0
5.610

5

-5.982

1

3.431

2

3

4
4

5
1
2

5

2

3

4

5

DBH/Age

1

3

3.052
1.855
-1.190
14.419

.650
.413
.009
.391

-8.874
-7.997
-11.814
-3.200

2.2447

.070

-12.267

.303

1.7593

.301

-1.495

8.357

6 .5 0 2 0

1.8971

.009

1.190

11.814

1 2 .1 1 2 0
.520

3.1460
2.2447
3.0649

.002
.999
.046

3.302
-5.766
-17.263

20.921
6.805
-.098

3.1460

.391

-14.419

3.200

.002

-20.921

-3.302

- 8 .6 8 0 0
-5.610

2.1295
1.7593
1.8971
3.1460

- 1 2 .1 1 2 0

5

-1 1 .5 9 2 0
2.911
5.982
-.520

3.3670

.006

-21.020

-2.163

2.1295
2.2447
2.2447

.650
.070
.999

-3.052
-.303
-6.805

8.874
12.267
5.766

3.3670

.008

2.163

21.020

1

2

1 1 .5 9 2 0
-.004646

.0158412

.998

-.049003

.039712

3

.017189

.0158412

.814

-.027169

.061547

4
5
1

-.027371
.020400
.004646

.0275967
.0191744
.0158412

.858
.824
.998

-.104646
-.033291
-.039712

.049904
.074092
.049003

3

.021835

.0170820

.705

-.025998

.069667

4

-.022725

.0283272

.929

-.102046

.056595

5

.025046

.0202116

.729

-.031550

.081642

1
2
4
5

-.017189
-.021835
-.044560
.003211

.0158412
.0170820
.0283272
.0202116

.814
.705
.519
1.000

-.061547
-.069667
-.123881
-.053384

.027169
.025998
.034761
.059807

1

.027371

.0275967

.858

-.049904

.104646

2

.022725

.0283272

.929

-.056595

.102046

3

.044560

.0283272

.519

-.034761

.123881

5
1
2

.047771
-.020400
-.025046

.0303175
.0191744
.0202116

.518
.824
.729

-.037122
-.074092
-.081642

.132665
.033291
.031550

3

-.003211

.0202116

1.000

-.059807

.053384

-.132665

.037122

4

-.047771

.0303175

.518

2

.1 8 6 5 0 7 0

.0255890

.000

.114854

.258160

3

.1 5 3 4 3 9 0
.038672
.070016

.0255890
.0445783
.0309734

.000
.908
.170

.081786
-.086154
-.016714

.225092
.163499
.156746

- .1 8 6 5 0 7 0
-.033068

.0255890

.000

-.258160

-.114854

3

.0275933

.752

-.110333

.044198

4

- .1 4 7 8 3 5 0

.0457583

.016

-.275965

-.019704

5

- .1 1 6 4 9 1 0
- .1 5 3 4 3 9 0
.033068

.0326488
.0255890
.0275933

.006
.000
.752

-.207913
-.225092
-.044198

-.025069
-.081786
.110333

4
5
2

95% Confidence Interval
Lower and upper

3.1460

4
1

Sig.

3

2
3
Height/Age

Std. Error

1

1
2

196

Dependent
Variable

(1) sp ecies

Mean
Difference (1J)
-.114767

Std. Error
.0457583

.100

-.242897

.013363

-.083423

.0326488

.090

-.174845

.007998

1

-.038672

.0445783

.908

-.163499

.086154

2
3
5
1

.147835(*)
.114767
.031344
-.070016

.0457583
.0457583
.0489732
.0309734

.016
.100
.968
.170

.019704
-.013363
-.105789
-.156746

.275965
.242897
.168476
.016714

2

.116491 n

.0326488

.006

.025069

.207913

(J) sp ecies
4
5

4

5

cone angle

1

.083423

.0326488

.090

-.007998

.174845

4

-.031344

.0489732
4.393650
9
4.393650
9
7.654143
1
5.318162
4
4.393650
9
4.737795
6
7.856745
2
5.605835
4
4.393650
9
4.737795
6
7.856745
2
5.605835
4
7.654143
1
7.856745
2
7.856745
2
8.408753
1
5.318162
4
5.605835
4
5.605835
4
8.408753
1

.968

-.168476

.105789

.000

-73.571671

-48.965882

.000

-57.179497

-32.573707

.856

-13.802556

29.062987

.000

-79.571276

-49.787940

.000

48.965882

73.571671

.008

3.125622

29.658727

.000

46.898904

90.899080

.973

-19.108028

12.286365

.000

32.573707

57.179497

.008

-29.658727

-3.125622

.000

30.506729

74.506906

.006

-35.500203

-4.105809

.856

-29.062987

13.802556

.000

-90.899080

-46.898904

.000

-74.506906

-30.506729

.000

-95.855618

-48.764028

.000

49.787940

79.571276

.973

-12.286365

19.108028

.006

4.105809

35.500203

.000

48.764028

95.855618

2
3
44.876602(*)
4

7.630215

5
6 4 .6 7 9 6 0 8 0
1
3
4

6 1 .2 6 8 7 7 7 0
1 6 .3 9 2 1 7 5 0
6 8 .8 9 8 9 9 2 0

5
3

1

-3.410831
4 4 .8 7 6 6 0 2 0

2
1 6 .3 9 2 1 7 5 0
4

5 2 .5 0 6 8 1 7 0

5
1 9 .8 0 3 0 0 6 0
4

1

-7.630215

2
6 8 .8 9 8 9 9 2 0
3
5 2 .5 0 6 8 1 7 0
5
7 2 .3 0 9 8 2 3 0
5

95% Confidence Interval
Lower and upper

3

6 1 .2 6 8 7 7 7 0

2

Sig.

1

6 4 .6 7 9 6 0 8 0

2
3
4

3.410831
1 9 .8 0 3 0 0 6 0
72.309823(3

Based on observed m eans.
* The mean difference is significant at the .05 level.

197

Cluster Analysis:

Quick Cluster
Cluster Membership
Case Number
1
2

Cluster

species

Distance
3.902

1
1

3
3

3

1

3

9.978

4

1

3

9.194

5
6

1
1

1

16.475

7

1

3

17.712

8
9

1
1

14.207
5.654
15.589
11.060

9.460

10

1

1
3
1

11

1

3

12

1

3

2.709
23.166

13
14

1
1

3
3

15
16

1

3

1

3

.588

17

1
1

3

4.715

3

1.475

19
20

1

3

1

3

14.610
13.152

21

1

3

5.066

22

1
1

3

1.074
1.682

25
26

1
1
1

3
1
3

27

1

3

28
29
30

1

3

8.263

1
1

3
3

19.719
14.675

31

2

32

2

33
34
35

2
2
2

2
2

4.075
7.408
16.442

36
37

2

38

2

2

5.367

39

2
2

2

7.799

2

10.869

18

23
24

40

3

2

8.733
8.287

6.561
14.514
12.456
19.432

2

198

Case Number
41

2

2

Distance
6,741

42

2

2

8,664

43
44

2

2

3,188

45

2

46
47

2
2

2

10,222

48

2

2

5,512

49

2

2

50

2

2

11,806
9,171

51

2

52

2

53
54

2
2

2
2

7.385
9,012

55

2

56
57
58

2
2
2

2

6,974

59

2

2

12,177

60

2

2

13,612

61
62

3
3

1
1

11,946
6,781

63

3

3

15,977

64

3

1

19,513

65

3

1

17.032

66

3

2

22,107

67

3
3

1
1

3
3

1

9,796
7,105
1,195

3
3

1

14,150

2

2,531
10,435

68
69
70
71
72

species

Cluster

2

73
74

3
3

2

75

3

1

4,820

76
77

3

1

18,258

3

78

3

2
2

4,799
1,737

79
80

3
3
3

2
1
1

14,350
1,339
7,551

3

81
82
83

3

2
2

18,561

84

3

2

11,089

85

3
3

2

18,403
1,448

86

2

4,987

199

Case Number
87

Cluster

species

Distance

3

2

23.331

88
89

3

2

10.673

3

2

14.583

90

3

1

15.021

91
ANOVA

Cluster
needle_w_l
cone length (mm)
Height/Age
DBH/Age

Mean Square
.473

Error
df
2

Mean Square
.017

77.637
.004

2
2

39.998
.003

.169

2

33756.255

2

F

df

Sig.

72

28.258

.000

72
72

1.941

.151
.190

1.699

.008
72
20.410
.000
102.506
72
329.310
.000
The F tests should be used only for descriptive purposes because the clusters have been chosen to
maximize the differences among cases in different ciusters. The observed significance levels are not
corrected for this and thus cannot be interpreted as tests of the hypothesis that the cluster means are equal.
cone angle

200

Wood Traits:

General Linear Model
Between-Subjects Factors

Descriptive Statistics

MC % (ODGR)

Mean
73.168549
1
60.174814
6
50.634012
0
76.127427
6
64.471713
9
62.942708
3
.4167148
.4324167

Std. Deviation

3
4
5

.3866175
.4163814

1 Species
1
2
3
4
5
Total

Density
(g/cmS)

1
2

Total

N

15.07510868

25

12.58347879

28

12.46159002

20

18.77928001

4

13.02086210

10

15.93643589

87

.03763275
.04538827

25

.4124949

.01938334

20

.3958919

.03233939
.02535922

4
10

.03792872

87

28

T ests of Between-Subjects Effects

Source
Corrected Model

Dependent Variable
MC % (OD-GR)

Type III Sum
of Squares

df

Mean Square

6577.516(a)

4

Density (g/cm3)

.018(b)

4

.005

MC % (OD-GR)
Density (g/cm3)
MC % (OD-GR)

221456.279
8.784
6577.516

1
1

221456.279
8.784
1644.379

Error

Density (g/cm3)
MC % (OD-GR)
Density (g/cm3)

.018
15263.903

Total

MC % (OD-GR)

.106
366516.673

Corrected Total

Density (g/cm3)
MC % (OD-GR)

15.207
21841.419

87
86

Density (g/cm3)

.124

86

Intercept
Species

4
4
82
82

1644.379

.005
186.145

F
8.834

Siq.
.000

3.500
1189.697

.011

6815.716
8.834

.000
.000
.000

3.500

.011

.001

87

a R Squared = .301 (Adjusted R Squared = .267)
b R Squared = .146 (Adjusted R Squared = .104)

201

Estimated Marginal Means
1. Grand M ean

95% Confidence Interval
Dependent Variable
MC % (OD-GR)
Density (g/cm3)

Mean

Std. Error

64.915

1.882

Lower Bound
61.171

Upper Bound
68.659

.409

.005

.399

.419

2. Species
Estimates
Dependent
Variable
MC % (OD-GR)

95% Confidence Interval
Species
1

Mean

2

Std. Error

Lower Bound

73.169

2.729

67.740

Upper Bound
78.597

60.175
50.634

2.578

55.046

65.304

76.127

3.051
6.822

44.565
62.557

56.703
89.698

5
1

64.472
.417

4.314
.007

55.889
.402

73.055
.431

2

.432

.007

.412

.008

.419
.397

.446

3
4

.396

.018

.360

.428
.432

5

.387

.011

.364

.409

3
4
Density (g/cm3)

Univariate T ests
Dependent
Variable
MC % (OD-GR)
Density (g/cm3)

Contrast
Error

Sum of
Squares
6577.516
15263.903

Contrast
Error

.018
.106

df
4
82
4
82

Mean Square
1644.379
186.145

F
8.834

Siq.
.000

.005

3.500

.011

.001

The F tests the effect of Species. This test is based on the linearly independent pairwise comparisons
among the estimated marginal means.

202

Post Hoc Tests
Species
Multiple Comparisons

Dependent
Variable
MC % (ODGR)

Mean
Difference (1(!) Species
1

(J) Species
2
3
4
5

2

J)
12.9937345(1
22.5345371 (*)
-2.9588786
8.6968352

1
12.9937345(1
3
4
5

3

9.5408026
-15.9526130
-4.2968993

1
22.5345371(1
2

-9.5408026

4
25.4934156(1
5
4

1
2
3
5

5

1
2
3

4
Density
(g/cm3)

1

2
3
4
5

2

1
3
4

-13.8377019
2.9588786
15.9526130
25.4934156(1
11.6557138
-8.6968352
4.2968993
13.8377019
-11.6557138

Std. Error
3.7541773
7
4.0930507
3
7.3472509
1
5.1049311
6
3.7541773
7
3.9944085
0
7.2927588
0
5.0261871
4
4.0930507
3
3.9944085
0
7.4728540
4
5.2841057
7
7.3472509
1
7.2927588
0
7.4728540
4
8.0716048
9
5.1049311
6
5.0261871
4
5.2841057
7
8.0716048
9

Sig.

95% Confidence Interval
Lower and Upper

.007

2.5218125

23.4656565

.000

11.1173600

33.9517142

.994

-23.4533387

17.5355816

.437

-5.5428865

22.9365569

.007

-23.4656565

-2.5218125

.129

-1.6012214

20.6828266

.195

-36.2950726

4.3898465

.912

-18.3169720

9.7231734

.000

-33.9517142

-11.1173600

.129

-20.6828266

1.6012214

.009

-46.3382338

-4.6485975

.076

-28.5772142

.9018104

.994

-17.5355816

23.4533387

.195

-4.3898465

36.2950726

.009

4.6485975

46.3382338

.601

-10.8592631

34.1706907

.437

-22.9365569

5.5428865

.912

-9.7231734

18.3169720

.076

-.9018104

28.5772142

.601

-34.1706907

10.8592631

-.0157019

.00987802

.508

-.0432557

.0118519

.0042199

.01076966
.01933214

.995

-.0258210

.0342609

.818

.01343213
.00987802
.01051012
.01918876

.175
.508
.328
.324

-.0331023
-.0073704
-.0118519
-.0093951
-.0170004

.0675649
.0432557
.0492388
.0900500

.0208229
.0300973
.0157019
.0199219
.0365248

.0747481

203

Dependent
Variable

(J) Species
5
1

(1) Species
3

Mean
Difference (1J)
.0457992(3

Std. Error
.01322494

Sig.
.007

-.0042199

.01076966

.995

-.0342609

.0258210

-.0199219

.01051012

.328

-.0492388

.0093951
.0714500
.0646600
.0331023

2
4
4

5

95% Confidence Intervai
Lower and Upper
.0089095
.0826889

5
1

.0166030
.0258773
-.0208229

.01966263
.01390358
.01933214

.916
.346
.818

-.0382441
-.0129054
-.0747481

2

-.0365248

-.0900500

.0170004

-.0166030

.01918876
.01966263

.324

3

.916

-.0714500

.0382441

5

.0092744

.02123806

.992

-.0499672

1
2
3

-.0300973
-.0457992(3
-.0258773

.01343213
.01322494
.01390358

.175
.007
.346

-.0675649
-.0826889
-.0646600

.0685159
.0073704
-.0089095
.0129054

4

-.0092744

.02123806

.992

-.0685159

.0499672

Based on observed means.
* The mean difference is significant at the .05 level.

Cluster Analysis:

Quick Cluster
Cluster Membership
Case Number
1

Cluster

Species
1

Distance
2.372

2

1

3
2

3

1

2

8.018

4

1

1

3.691

5
6

1

2

6.011

7

1
1
1

1

6.102

8
9

1
1

7.719
6.307

10

1

1

2.093

11
12

1

1

3.906

1

.812

7.045

1

1

.600

13

1

14

1

1

3.293

15
16
17

1
1

1
1

3.281
5.834

1

1

1.887

18

1

19
20

1
1

1

6.115
7.367

1

.859

21

1

3

22

1

1

8.423
10.164

204

23

1

1

6.463

24

1

3

3.668

25

1

3

9.595

26

1

3

27.052

27

1

2

1.230

28

1
1

3
3

4.119

29

.307

30

1

1

6.086

31

2

2

6.090

32

2
2

2

33

1

4.437
4.103

34

2

2

15.581

35

2

1

6.633

36
37

2

4.563

2

2
2

38

2

2

5.801

3.135

39

2

1

5.740

40

2

1

41

2

1

5.126
2.062

42

2

2

4.801

43
44

2
2

1

4.203
9.114

45
46
47

2
2
2

1
2

48
49

2
2

1
1

50

2

2

2.039
2.184

51
52

2

53

2

2

2.368

54

2

2

.855

55

2
2
2

2

56
57

2
3

5.679
4.829

58

2

59

2

60
61

2

1

3

2

.779

62
63

3
3

1
1

4.209
7.269

64
65

3
3

2
2

3.059
13.958

66
67

3
3

2

.824
14.111

68

3

2

6.728

69

3

2

16.915

3
2

4.056
7.072
1.690
1.126

2

1
1

2

7.800
.071
4.592
1.619

205

70

3

2

3.513

71

3

2

4.307

72

3

2

5.196

73

3

1

8.274

74

3

2

4.391

75

3

2

5.644

76
77

3

1

1.600

3

1

1.103

78
79

3

2

7.487

3

2

.617

80

3

2

2.829

81
82

3

1

1.833

3

2

83
84

3
3

2
1

2.543
1.451

85

3

86

3

3
2

8.650

87

1

8.424
8.055
4.933
2.651

88

3
3

89

3

1
2

90

3

1

9.588
3.384

ANOVA

Cluster
MC % (OD-GR)

Mean Square
9003.456

Error
df

Mean Square
45.649

df

F

Sig.

84
2
197.233
.000
2
84
.005
.001
4.096
.020
The F tests should be used only for descriptive purposes because the clusters have been chosen to
maximize the differences among cases in different clusters. The observed significance ieveis are not
corrected for this and thus cannot be interpreted as tests of the hypothesis that the cluster means are equal.
Density (g/cm3)

206

Descriptive Statistics
1

MOE(ave)

Mean
7.9652000
16.620460
0
12.147680
0
12.467250
0
12.287056
0
.7239786

Std. Deviation
1.50370234

2

.5721113

3
5

Species
1
2

5

3.25204041

3

.50777338

2

4.16151138

15

.33833379

5

.23184168

5

1.1803463
1.1078037

.33015285

3
2

Total

.8158064

1
2

2.4304920
2.7029860

.42380675
.13864407

3

2.3231033

.13377258
.06165160

5
Total

2.3498650

.19176029

3
2

2.4890953
487.66428
21
542.97266
72
481.01434
49
470.11251
29
502.43018
71

.20003411

15

21.66650158

5

31.51885557

5

23.75394185

3

60.95795948

2

40.94358205

15

5
Total

Wall(ave)

mw dens

5

2.47957977

3

ew/lw ratio

N

1

1
2
3
5

Total

.84581096

15
5
5

T ests of Between-Subjects Effects

Source
Corrected Model

Intercept

Species

Error

Type III Sum
of Squares
187.407(a)

F

3

Mean Square
62.469

.908(b)

3

.303

2.073

Siq.
.001
.162

Wall(ave)

.367(0)

3

.122

6.985

.007

mw dens

12773.402(d)

3
1

4257.801

4.379

.029

1962.728

1962.728

392.210

.000

ew/lw ratio

10.416

1

10.416

71.332

.000

Wall(ave)

77.973

77.973

mw d e n s

3 1 8 4 3 6 8 .4 7 6

.000
.000
.001
.162
.007

Dependent Variable
MOE(ave)
ew/lw ratio

MOE(ave)

df

12.483

MOE(ave)

187.407

1
1
3

62.469

4447.661
3274.913
12.483

ew/lw ratio
Wall(ave)

.908
.367

3
3

.303

2.073

.122

mw dens
MOE(ave)

12773.402

4257.801
5.004

6.985
4.379

55.047

3
11

ew/lw ratio

1.606

11

.146

3 1 8 4 3 6 8 .4 7 6

207

.029

Total

Corrected Total

a
b
c
d

Wall(ave)

.193

11

.018

mw dens

10695.875

11

972.352

MOE(ave)

2507.031

15

ew/lw ratio

15

Wall(ave)

12.498
93.494

mw dens
MOE(ave)

3810010.670
242.454

ew/lw ratio

2.515

15
14
14

Wall(ave)

.560

14

mw dens

23469.277

14

15

R Squared =.773 (Adjusted R Squared = .711)
R Squared =.361 (Adjusted R Squared = .187)
R Squared =.656 (Adjusted R Squared = .562)
R Squared =.544 (Adjusted R Squared = .420)

Estimated Marginal Means
1. Grand Mean
95% Confidence Interval
Dependent Variable
MOE(ave)
ew/lw ratio
Wall(ave)
mw dens

Mean

Std. Error

Lower Bound

12.300

.621

10.933

Upper Bound
13.667

.896
2.452
495.441

.106
.037

.663
2.371

1.130
2.533

8.657

476.386

514.496

2. Species
Estimates
Dependent
Variable

95% Confidence Interval
Species
1
2

Mean

3
5
1

12.148
12.467
.724

2

.572

3
Wall(ave)

5
1

mw dens

2
3
5
1
2
3

MOE(ave)

ew/lw ratio

5

Std. Error

Lower Bound

1.000

5.763

Upper Bound
10.167

1.000
1.292

14.419

18.822

9.305

14.990

1.582
.171

15.949
1.100

.171

8.986
.348
.196

1.180

.221

.695

1.666

1.108

.270

.513

1.703

7.965
16.620

.948

2.430

.059

2.300

2.561

2.703
2.323
2.350

.059
.076
.094

2.573
2.155
2.144

2.833
2.491
2.556

487.664

13.945

456.971

518.358

542.973

13.945

512.279

573.666

481.014

18.003

441.389
421.582

518.643

470.113

22.049

520.639

208

Univariate Tests
Dependent
Variable
MOE(ave)

Sum of
Squares
187.407

Contrast
Error

ew/lw ratio

55.047

Contrast
Error

Waii(ave)

Contrast
Error

mw dens

Contrast
Error

Mean Square

df
3

62.469

11

5.004

.908

3

.303

1.606
.367

11

.146

3

.122

.193
12773.402

11

.018

3

4257.801

10695.875

11

972.352

F
12.483

Sig.
.001

2.073

.162

6.985

.007

4.379

.029

The F tests the effect of Species. This test is based on the linearly independent pairwise comparisons
among the estimated marginal means.

Post Hoc Tests
Species
Multiple Comparisons
Tu key HSD
Dependent
Variable
MOE(ave)

(J)
(1) Species
1

2

Species
2
3

5

-8.6552600(*)
-4.1824800

Sig.
.000
.104

-4.5020500

1.87162981

.133

1.41481915

5
1

8.6552600D
4.4727800
4.1532100
4.1824800

1.63369243
1.87162981
1.63369243

.000
.078
.178
.104

-.4438955
-1.4795494
-.7341955

2

-4.4727800

1.63369243

.078

-9.3894555

5
1

-.3195700

2.04211554

.999

-6.4654144

4.5020500
-4.1532100
.3195700
.1518672

1.87162981
1.87162981
2.04211554

.133
.178
.999
.921

-1.1307094
-9.7859694
-5.8262744

.27907026
.31971514
.24168193
.27907026
.31971514

.400
.639
.921
.189
.380
.400

-1.2962430
-1.3460231
-.8792205
-1.4481102
-1.4978904

5
1

2
3
ew/lw ratio

2

-.4563677

2

3
5
1
3
5
1
2

.6082349

.27907026
.27907026

5
1
2

.0725426
.3838251
.5356924

.34883782
.31971514
.31971514

3
2

-.0725426
-.2724940(1

5

Wall(ave)

(l-J)

1

3

1

95% Confidence Interval
Lower and Upper
-12.9132259
-4.3972941
-9.0991555
.7341955
-10.1348094
1.1307094

Std. Error
1.41481915
1.63369243

3
3

Mean Difference

-.3838251
-.1518672
-.6082349
-.5356924
.4563677

.24168193

4.3972941

12.9132259
9.3894555
9.7859694
9.0991555
.4438955
5.8262744
10.1348094
1.4795494
6.4654144

-.5754861

-.3835076
-.2316404

.8792205
.3835076
.5783729
.5754861
.2316404
.4266057
1.2962430
1.4481102

-.9773015
-.5783729
-.4265057
-1.1223867

1.1223867
1.3460231
1.4978904

.34883782

.189
.997
.639
.380
.997

.08374050

.033

-.5245150

-.0204730

.9773015

209

3
5
1

2

3
5
3

1

5

2
5
1

.3983975
.4140195
.5245150

.3798827(0

.09669520
.11077827

.011
.037

.0888738

.6708915

.09669520
.09669520
.12086900
.11077827

.691
.011
.996
.884

.0197285

.6865135
.1836202
-.0888738
.3369994
.2527655
-.0197285
.3905227

.11077827

.037

.12086900

.996

-.3369994

-55.3083851

19.72158441

.070

6.6499371
17.5517692

22.77252414
26.08920390

.991
.905

55.3083851
61.9583222

19.72158441
22.77252414

.070
.080

72.8601543
-6.6499371

26.08920390
22.77252414

.072
.991

-61.9583222

22.77252414

.080

5

10.9018320

28.46565517

1

-17.5517692

3
5
1
3
5
1
2

3

-.1836202
-.2527655
.0204730

-.3531210(0
.0267617

2

2

.691
.884
.033

-.3983975
-.6708915
-.3905227
-.4140195
-.6865135

3
1

.09669520
.11077827
.08374050

.3531210(0
-.1073887
-.3798827(0
-.0267617
-.0806270

2
mw dens

.1073887
.0806270
.2724940(0

114.6614366
-61.8850634
-60.9649378
-4.0446664

4.0446664
75.1849377
96.0684761

-6.5766783

114.6614366
130.4933228

-5.6565527
-75.1849377

151.3768612
61.8850634
6.5766783

.980

130.4933228
-74.7669187

96.5705827

26.08920390

.905

-96.0684761

60.9649378

-72.8601543

26.08920390

.072

3
-10.9018320
Based on observed means.
* The mean difference is significant at the .05 level.

28.46565517

.980

5

2

151.3768612
-96.5705827

5.6565527
74.7669187

Quick Cluster
Cluster Membership
Case Number
1
2

Cluster

Species

Distance

1

3

1
1

4

1

5

1

6

1

7

1

8

1

9
10
11
12

1

1

13.893

1

19.445

1
1

13

1
1

3

5.575

14
15

1
1

1

12.632

16

1

17

1

210

18

1

19
20

1

21

1

22
23

1
1

24

1

25

1

26

1

27
28

1
1

29

1

1

30

1

31

2

32

2

33
34

2

35
36
37

2
2

38
39

2

40

2

41

2

42

2

43
44

2

45

2
2

46
47
48
49

9.776

2

7.809

2

16.396

1

8.972

2

8.875

1

8.965

2

2
2

2

2
2

50

2
2

51
52

2
2

53
54

2

55
56

2

2

57
58

2
2
2

59
60

2
2

61
62

3

63
64

1

3
3
3

211

11.216
67

15.386

68

73
74
76
77
78
79
5.864

85

86
87

88

27.282

89
10.285
ANOVA

Cluster
Mean Square
ew/lw ratio
mw dens

Error
df

Mean Square

Sig.

df

.434

.137

12

3.168

.079

10483.096

208.590

12

50.257

.000

MOE(ave)

36.788

14.073

12

2.614

.114

Wall(ave)

.182

.002
Ttie F tests stiould be used only for descriptive purposes because the clusters have been chosen to
maximize the differences among cases in different clusters. The observed significance levels are not
corrected for this and thus cannot be interpreted as tests of the hypothesis that the cluster means are equal.
.016

12

11.059

212

Fibre Traits:

General Linear Model
Between-Subjects Factors

Descriptive Statistics

Fibre length
(20-40)

coarseness
(20-40)

Std. Deviation

N

sp ecies
1

Mean
2.4710

.21279

2

2.5330

.16312

5

3

2.4350

.02500

3

5

5

2.3100

.21213

2

Total

2.4630

.17067

15

1

.20020

.027813

5

2

.20930

.015446

5

3

.19767

.008780

3

5

.22375

.003889

2

Total

.20587

.019418

15

T ests of Between-Subjects Effects
Type III Sum

df

Dependent Variable

Corrected Model

Fibre length (20-40)

.074(a)

3

coarsen ess (20-40)

.001(b)

3

Fibre length (20-40)

77.062

1

Intercept
sp ecies
Error
Total
Corrected Total

of Squares

Mean Square

Source

SIq.

F
.813

.513

.000

.922

.462

77.062

2539.487

.000

.025

coarseness (20-40)

.560

1

.560

1459.929

.000

Fibre length (20-40)

.074

3

.025

.813

.513

coarsen ess (20-40)

.001

3

.000

.922

.462

Fibre length (20-40)

.334

11

.030

coarseness (20-40)

.004

11

.000

Fibre length (20-40)

91.403

15

coarsen ess (20-40)

.641

15

Fibre length (20-40)

.408

14

coarseness (20-40)

.005

14

a R Squared = .181 (Adjusted R Squared = -.042)
b R Squared = .201 (Adjusted R Squared = -.017)

Estimated Marginal Means
1. Grand Mean
95% Confidence Interval
Mean

Std. Error

Lower Bound

Upper Bound

Fibre length (20-40)

2.437

.048

2.331

2.544

coarseness (20-40)

.208

.005

.196

.220

Dependent Variable

213

2. species
Estimates
95% Confidence Interval

Dependent
Variable

Mean

Std. Error

Lower Bound

Upper Bound

1

2.471

.078

2.300

2.642

2

2.533

.078

2.362

2.704

3

2.435

.101

2.214

2.656

5

2.310

.123

2.039

2.581

1

.200

.009

.181

.219

2

.209

.009

.190

.229

3

.198

.011

.173

.223

.224

.014

.193

.254

sp ecies

Fibre length
(20-40)

coarseness
(20-40)

5

Univariate Tests
Sum of
Squares

Dependent
Variable
Fibre length
(20-40)
coarseness
(20-40)

Mean Square

df

Contrast

.074

3

.025

Error

.334

11

.030

Contrast

.001

3

.000

Error

.004

11

.000

F

Sig.
.813

.513

.922

.462

The F tests the effect of sp ecies. This test is based on the iinearly independent pairwise comparisons
among the estimated marginal means.

Post Hoc Tests
species
Multiple Comparisons

Dependent
Variable
Fibre length
(20-40)

(1) sp ecies

(J) sp ecies

1

2
3

2

3

95% Confidence interval

Mean
Difference
(l-J)

Std. Error

-.0620
.0360

.11017
.12722

Sig.
.941
.992

Lower Bound
-.3936
-.3469

Upper Bound
.2696
.4189

5

.1610

.14575

.694

-.2776

.5996

1

.0620

.11017

.941

-.2696

.3936

3
5
1

.0980
.2230
-.0360

.12722
.14575
.12722

.866
.454
.992

-.2849
-.2156
-.4189

.4809
.6616
.3469

2

-.0980

.12722

.866

-.4809

.2849

214

5

5

.1250

.15902

.859

-.3536

.6036

1

-.1610
-.2230
-.1250

.14575
.14575
.15902

.694
.454
.859

-.5996
-.6616
-.6036

.2776
.2156
.3536

-.00910

.012385

.881

-.04637

.02817

2
3
coarsen ess
(20-40)

2

1

2

3

5

3

.00253

.014300

.998

-.04050

.04557

5
1
3
5

-.02355
.00910
.01163
-.01445

.016383
.012385
.014300
.016383

.504
.881
.847
.814

-.07286
-.02817
-.03140
-.06376

.02576
.04637
.05467
.03486

1

-.00253

.014300

.998

-.04557

.04050

2

-.01163

.014300

.847

-.05467

.03140

5
1
2

-.02608
.02355
.01445

.017876
.016383
.016383

.492
.504
.814

-.07988
-.02576
-.03486

.02771
.07286
.06376

3

.02608

.017876

.492

-.02771

.07988

Based on observed means.

General Linear Model
Between-Subjects Factors
Descriptive Statistics

Fibre length (4060)

Fibre length (6080)

Fibre coarseness
(40-60)

Fibre coarseness
(60-80)

Std. Deviation

N

1 sp ecies
1

Mean
2.6240

.25967

10

2

2.6380

.17968

26

.16489

13

.19982

50

3

2.5246

4

2.2850

Total

2.5986

1

2.6515

.27564

10

2

2.6475

.17825

26

3

2.5378

.16869

13

4

2.6050

Total

2.6189

.19884

50

1

.19405

.017940

10

2

.21152

.021008

26

3

.19314

.017968

13

4

.15800

Total

.20217

.022014

1

.2025

.02486

10

2

.01923

26

.01854

13

.02240

50

3

.2139
.1894

4

.1850

Total

.2047

1

1

1
50

1

Tests of Between-Subjects Effects

215

Source
Corrected Model

Type III Sum
of Squares

Dependent Variable
Fibre length (40-60)
Fibre length (60-80)

Intercept

sp ecies

Error

Total

Corrected Total

df

F

Mean Square

Sig.

.216(a)

3

.072

1.905

.142

.118(b)

3

.039

.991

.405

Fibre coarseness (40-60)

.006(c)

3

.002

5.117

.004

Fibre coarseness (60-80)

.006(d)

3

.002

4.590

.007

Fibre length (40-60)

83.461

1

83.461

2206.177

.000

Fibre length (60-80)

89.709

1

89.709

2267.886

.000

Fibre coarseness (40-60)

.471

1

.471

1217.269

.000

Fibre coarseness (60-80)

.514

1

.514

1250.289

.000

Fibre length (40-60)

.216

3

.072

1.905

.142

Fibre length (60-80)

.118

3

.039

.991

.405

Fibre coarseness (40-60)

.006

3

.002

5.117

.004

Fibre coarseness (60-80)

.006

3

.002

4.590

.007

46

.038

Fibre length (40-60)

1.740

Fibre length (60-80)

1.820

46

.040

Fibre coarseness (40-60)

.018

46

.000

Fibre coarseness (60-80)

.019

46

.000

Fibre length (40-60)

339.603

50

Fibre length (60-80)

344.874

50

Fibre coarseness (40-60)

2.067

50

Fibre coarseness (60-80)

2.119

50

Fibre length (40-60)

1.956

49

Fibre length (60-80)

1.937

49

Fibre coarseness (40-60)

.024

49

Fibre coarseness (60-80)

.025

49

a R Squared = .111 (Adjusted R Squared = .053)
b R Squared = .061 (Adjusted R Squared = -.001)
c R Squared = .250 (Adjusted R Squared = .201)
d R Squared - .230 (Adjusted R Squared = .180)

Estimated Marginal Means
1. Grand Mean
95% Confidence Interval
Mean

Dependent Variable

Std. Error

Lower Bound

Upper Bound

.054

2.410

2.626

Fibre length (40-60)

2.518

Fibre length (60-80)

2.610

.055

2.500

2.721

Fibre coarseness (40-60)

.189

.005

.178

.200

Fibre coarseness (60-80)

.198

.006

.186

.209

2. species
Estimates
Dependent
Variable

95% Confidence Interval
sp ecies

Mean

Std. Error

Lower Bound

Upper Bound

216

Fibre length (4060)

Fibre length (6080)

Fibre coarseness
(40-60)

Fibre coarseness
(60-80)

1

2.624

.062

2

2.638

.038

2.561

2.715

3

2.525

.054

2.416

2.633

4

2.285

.195

1.893

2.677

1

2.748

2.500

2.652

.063

2.525

2.778

2

2.648

.039

2.569

2.726

3

2.538

.055

2.427

2.649

4

2.605

.199

2.205

3.005

1

.194

.006

.182

.207

2

.212

.004

.204

.219

3

.193

.005

.182

.204

4

.158

.020

.118

.198

1

.203

.006

.190

.215

2

.214

.004

.206

.222

3

.189

.006

.178

.201

4

.185

.020

.144

.226

Univariate Tests
Sum of
Squares

Dependent
Variable
Fibre length (4060)

df

Mean Square

Contrast

.216

3

.072

Error

1.740

46

.038

Fibre length (6080)

Contrast

.118

3

.039

Error

1.820

46

.040

Fibre coarseness
(40-60)

Contrast

.006

3

.002

Error

.018

46

.000

Fibre coarseness
(60-80)

Contrast

.006

3

.002

Error

.019

46

.000

F

Sig.

1.905

.142

.991

.405

5.117

.004

4.590

.007

The F tests the effect of sp ecies. This test is based on the linearly independent pairwise comparisons
among the estimated marginal m eans.

Cluster Analysis:

Quick Cluster
Cluster Membership
C ase Number
1

sp ecies

Cluster

Distance

1

2

1

3

.153

3

1

2

.241

4

1

5

1

6

1

7

1

3

.302

217

C ase Number
8

Cluster

sp ecies

Distance

1

2

.158

9

1

1

.149

10

1

3

.073

11

1

2

.254

12

1

13

1

14

1

15

1

16

1

1

.131

17

1

3

.246

18

1
1

.234

19

1

20

1

21

1

22

1

23

1

24

1

25

1

26

1

27

1

28

1

29

1

30

1

1

.106

31

2

3

.250

32

2

3

.094

33

2

1

.093

34

2

2

.272

35

2

36

2

1

.149

37

2

3

.132

38

2

3

.080

39

2

3

.130

40

2

3

.290

41

2

1

.117

42

2

1

.114

43

2

3

.152

44

2

3

.175

45

2

1

.184

46
47

2
2

3
1

.082

48

2

1

.280

49

2

1

.091

50

2

2

.136

51

2

52

2

53

2

3

.032

.133

218

C ase Number
54

Distance

Cluster

sp ecies
2

1

.078

55

2

2

.069

56

2

3

.155

57

2

58

2

1

.068

59

2

1

.140

60

2

1

.106

61

3

1

.049

62

3

63

3

1

.068

64

3

1

.120

65

3

3

.043

66

3

3

.161

67

3

1

.050

68

3

3

.114

69

3

2

.063

70

3

71

3

1

.154

72

3

73

3

74

3

75

3

1

.131

76

3

77

3

78

3

1

.134

79

3

1

.149

80

3

2

.178

81

3

82

3

83

3

84

3

85

3

86

3

87

3

88

3

89

3

90

3

91
92
93

219

ANOVA
Ciuster
Mean Square

Error
df

Mean Square

F

df

Sig.

Fibre length (40-60)

.733

2

.010

47

70.339

.000

Fibre length (60-80)

.617

2

.015

47

41.328

.000

Fibre coarseness (40-60)

.002

2

.000

47

5.619

.006

Fibre coarseness (60-80)

.003

2

.000

47

7.349

.002

The F tests should be used only for descriptive purposes becau se the clusters have been chosen to
maximize the differences among c a s e s in different dusters. The observed significance ievels are not
corrected for this and thus cannot be interpreted a s tests of the hypothesis that the cluster m eans are equal.

General Linear Model
Between-Subjects Factors

Descriptive Statistics

Crs(ave)

MFA(ave)

mw mfa

1 S pecies
1

Mean

Std. Deviation

N

358.77780

21.400743

5

2

395.49000

20.465508

5

3

343.60200

7.950180

3

5

368.22050

19.627163

2

Total

369.23907

26.706810

15

1

27.590220

4.7595925

5

13.733358

3.7984525

5

3

17.848300

6.8601941

3

5

17.015000

.4792770

2

Total

19.612853

7.3511658

15

1

21.876273

5.7484767

5

2

12.970088

4.5052271

5

3

14.391544

6.8823237

3

5

13.581224

4.8868964

2

Total

16.304592

6.3729798

15

2

220

T ests of Between-Subjects Effects

Source
Corrected Model

Dependent Variable
Crs(ave)

Type III Sum
of Squares

sp ecies

Error

Total

Corrected Total

F

SIq.

5966.601 (a)

3

1988.867

5.444

.015

513.873(b)

3

171.291

7.764

.005

mw mfa

236.625(c)

3

78.875

2.613

.104

Crs(ave)

1742773.595

1

1742773.595

4770.028

.000

MFA(ave)

4706.303

1

4706.303

213.322

.000

mw mfa

3199.656

1

3199.656

106.018

.000

Crs(ave)

5966.601

3

1988.867

5.444

.015

MFA(ave)

513.873

3

171.291

7.764

.005

3

78.875

2.613

.104

MFA(ave)
Intercept

Mean Square

df

mw mfa

236.625

Crs(ave)

4018.952

11

365.359

MFA(ave)

242.682

11

22.062

mw mfa

331.983

11

30.180

Crs(ave)

2055047.877

15

MFA(ave)

6526.515

15

mw mfa

4556.204

15

Crs(ave)

9985.552

14

MFA(ave)

756.555

14

mw mfa

568.608

14

a R Squared = .598 (Adjusted R Squared = .488)
b R Squared = .679 (Adjusted R Squared = .592)
0 R Squared = .416 (Adjusted R Squared = .257)

Estimated Marginal Means
1. Grand Mean
95% Confidence Interval
Dependent Variable

Mean

Std. Error

Crs(ave)

Lower Bound

Upper Bound

366.523

5.307

354.842

378.203

MFA(ave)

19.047

1.304

16.176

21.917

mw mfa

15.705

1.525

12.348

19.062

221

2. species
Estimates
95% Confidence interval

Dependent
Variable

sp ecies

Mean

Crs(ave)

1

358.778

2

395.490

8.548

376.676

414.304

3

343.602

11.036

319.313

367.891

5

368.220

13.516

338.472

397.969

1

27.590

2.101

22.967

32.214

2

13.733

2.101

9.110

18.357

MFA(ave)

mw mfa

Std. Error

Lower Bound

Upper Bound

8.548

339.963

377.592

3

17.848

2.712

11.880

23.817

5

17.015

3.321

9.705

24.325

1

21.876

2.457

16.469

27.284

2

12.970

2.457

7.563

18.378

3

14.392

3.172

7.411

21.373

5

13.581

3.885

5.031

22.131

Multivariate T ests

Univariate T ests
Dependent
Variable
Grs(ave)
MFA(ave)
mw mfa

Sum of
Squares

df

Mean Square

Contrast

5966.601

3

1988.867

Error

4018.952

11

365.359

Contrast

513.873

3

171.291

Error

242.682

11

22.062

Contrast

236.625

3

78.875

Error

331.983

11

30.180

F

Sig.

5.444

.015

7.764

.005

2.613

.104

The F tests the effect of sp ecies. This test is based on the linearly independent pairwise comparisons
among the estimated marginal means.

222

Post Hoc Tests
species
Multiple Comparisons
Tukey HSD
Dependent
Variable
Crs(ave)

(!) sp ecies
1

(J) sp ecies
2
3
5

2

1
3
5

3

1
2
5

5

1
2
3

MFA(ave)

1

2
3
5

2

Mean
Difference (1J)
-36.71220(*)
15.17580
-9.44270
36.712 2 0 n
5 1 .8 8 8 0 0 0
27.26950
-15.17580
-5 1 .8 8 8 0 0 0
-24.61850
9.44270
-27.26950
24.61850
1 3 .8 5 6 8 6 2 0
9.741920
10.575220

1
1 3 .8 5 6 8 6 2 0
3
5

3

1
2
5

5

1
2
3

mw mfa

1

2

-4.114942
-3.281642
-9.741920
4.114942
.833300
-10.575220
3.281642
-.833300
8.906185

Std. Error
12.08899
1
13.95916
4
15.99223
1
12.08899
1
13.95916
4
15.99223
1
13.95916
4
13.95916
4
17.44895
5
15.99223
1
15.99223
1
17.44895
5
2.970657
0
3.430219
2
3.929809
8
2.970657
0
3.430219
2
3.929809
8
3.430219
2
3.430219
2
4.287774
0
3.929809
8
3.929809
8
4.287774
0
3.474492
7

Sig.

95% Confidence Interval
Lower and upper

.048

-73.09460

-.32980

.704

-26.83497

57.18657

.933

-57.57208

38.68668

.048

.32980

73.09460

.015

9.87723

93.89877

.366

-20.85988

75.39888

.704

-57.18657

26.83497

.015

-93.89877

-9.87723

.518

-77.13196

27.89496

.933

-38.68668

57.57208

.366

-75.39888

20.85988

.518

-27.89496

77.13196

.003

4.916528

22.797196

.066

-.581489

20.065329

.084

-1.251731

22.402171

.003

-22.797196

-4.916528

.640

-14.438351

6.208467

.837

-15.108593

8.545309

.066

-20.065329

.581489

.640

-6.208467

14.438351

.997

-12.070961

13.737561

.084

-22.402171

1.251731

.837

-8.545309

15.108593

.997

-13.737561

12.070961

.104

-1.550467

19.362837

223

Dependent
Variable

(J) sp ecies
3

(1) sp ecies

Mean
Difference (1J)
7.484729

5

8.295049

1

2

-8.906185

3

-1.421456

5

-.611136

1

3

-7.484729

2

1.421456

5

.810320

1

5

-8.295049

2

.611136

3

-.810320

Std. Error
4.011998
6
4.596321
8
3.474492
7
4.011998
6
4.596321
8
4.011998
6
4.011998
6
5.014998
3
4.596321
8
4.596321
8
5.014998
3

Sig.

95% Confidence Interval
Lower and upper

.296

-4.589572

19.559031

.322

-5.537802

22.127899

.104

-19.362837

1.550467

.984

-13.495758

10.652846

.999

-14.443987

13.221714

.296

-19.559031

4.589572

.984

-10.652846

13.495758

.998

-14.282558

15.903197

.322

-22.127899

5.537802

.999

-13.221714

14.443987

.998

-15.903197

14.282558

Based on observed means.
* The mean difference is significant at the ,05 level.

Quick Cluster
Cluster Membership
C ase Number
1

sp ecies

Cluster

Distance

1

2

1

3

1

4

1

5

1

1

9.409

2

11.064

6

1

7

1

8

1

9

1

10

1

11

1

12

1

13

1
1

2

24.106

14
15

1

2

8.771

16

1

17

1

18

1

19

1

20

1

224

C ase Number
21
22

.

sp ecies

Cluster

Distance

1
1

23

1

24

1

25

1

26

1

27

1

28

1

29

1

30

1

31

2

32

2

33

2

34

2

35

2

36

2

37

2

38

2

39

2

40

2

41

2

42

2

43

2

44

2

45

2

46

2

47

2

48

2

49

2

50

2

51

2

52

2

53

2

54

2

55

2

56

2

57

2

58

2

59

2

60

2

61

3

62

3

63

3

64

3

65

3

66

3

2

10.252

3

5.517

3

11.155

3

10.440

2

16.589

2

11.717

1

10.099

225

C ase Number
67

sp ecies

Cluster
3

68

3

69

3

70

3

71

3

72

3

73

3

74

3

75

3

76

3

77

3

76

3

79

3

80

3

81

3

82

3

83

3

84

3

85

3

86

3

87

3

88

3

89

3

90

3

Distance
2

18.951

1

8.609

2

16.217

2

15.096

91
92
93

ANOVA
Error

Cluster
Mean Square

Mean Square

df

F

df

Sig.

Crs(ave)

4170.257

2

137.087

12

30.421

.000

MFA(ave)

106.087

2

45.365

12

2.339

.139

mw mfa

40.743

2

40.593

12

1.004

.395

The F tests should be used only for descriptive purposes b ecau se the clusters have been chosen to
maximize the differences among c a s e s in different clusters. The observed significance levels are not
corrected for this and thus cannot be interpreted as tests of the hypothesis that the cluster m eans are equal.

226

Chemical Extractives Analysis:

General Linear Model
Between-Subjects Factors

Descriptive Statistics

Region 1

1 sp ecies
1
2
3
4

Regions

428355.902

26

447762.663

33

301837.362

29
4

506739.541

10

Total

430670.259

102

1

948036.31

349465.907

26

2

881341.76

340557.426

33

3

725037.72

224907.477

29

4

735184.00

380531.451

4

5

853390.90

263895.120

10

Total

315096.810

102

640714.262

26

518643.429

33

375878.762

29

427732.132

4

467214.055

10

533051.376

102

742149.088

26

543850.910

33

442592.632

29

403435.330

4

402503.849

10

599001.854

102

1280142.437

26

386361.838

33

484987.925

29

4

845430.99
1612528.4
2
1448967.2
4
1165961.5
2
1051526.2
5
1320014.8
0
1381968.5
8
1722941.6
9
1485725.7
6
1168943.1
0
1017864.7
5
1177970.7
0
1407607.3
5
1920144.9
6
1130924.2
4
1159925.6
9
985544.25

351493.939

4

5

908956.20

319987.687

10

Total

1312880.9

814736.008

102

1

3
4
5
Total
1
2
3
4
5
Total
Regions

N

761049.831

2

Region4

997978.38
1053199,2
7
860841.72

Std. Deviation

985693.75
1217974.5
0
997940.63

5

Region2

Mean

1
2
3

227

1 species

Std. Deviation

Mean

N

1
Regions

1
2
3
4

Region?

33

271473.828

4

29

195104.922

10

900683.14
9937021.0
4
6061167.1
5
6541355.4
8
6246992.7
5
4615793.7
0
7051238.0
4
3623652.9
2
2018649.1
8
1939260.1
0
1743232.7
5
1449192.6
0
2338566.6
6
1714198.6
2
1441426.5
2
917772.48
810152.25

1242825.080

102

7952795.360

26

3498435.670

33

4182090.094

29

3921525.033

4

2465676.856

10

5341506.677

102

3044049.284

26

1008220.541

33

1175138.581

29

645157.098

4

449563.067

10

1904099.259

102

1099564.966

26

829546.352

33

454975.201
329005.744

29
4

464402.055

10

1

5
Total
1
2
3
4
5
Total
1
2
3
4
5

847018.825

102

825216.445

26

287924.796

3

867292.70
1281031.1
5
1110107.0
0
687199.76
610353.41

33
29

4

559390.25

194684.175

4

5
Total
1

505814.90
750356.21
2284147.3
8
1787259.7
9
1132913.1
0
985906.75

158610.674
516028.510

10
102

1752018.934

26

1236690.029

33

817914.143

29

521702.389

4

Total
1
2

Regioni 1

283325.246
473800.532

550269.90

4

RegionlO

26

Total

3

Regions

620370.25

2309683.350

5

2

Regions

1509703.7
7
708134.91
733267.24

2
3
4

271797.278

228

species
5

Regioni 3

10

1313745.529

102

1294889.292

26

324133.042

33

394551.126

29

4

784444.50

340890.314

4

5
Total

535200.90
931627.82
9658655.2
3
3344723.7
6
4373912.0
0
4622287.2
5
2266540.7
0
5191165.9
2
14100323.
65
3456719.1
8
4501569.1
4
3907738.7
5
1884505.2
0
6330408.4
3
22351318.
65
7512855.9
7
13591462.
34
10963865.
75
4568190.1
0
12870081.
41
13284278.
50
2501626.9
7
3911247.8
6
3848104.2
5
1399430.8
0
5595665.1
4

201585.519
772628.644

10
102

8787816.679

26

1901392.627

33

3645166.747

29

1924071.209

4

1512186.775

10

5617941.768

102

18213747.873

26

2644550.713

33

4784899.196

29

2595726.510

4

1613201.508

10

10606520.232

102

21493927.287

26

5001175.906

33

11702861.422

29

6168584.729

4

3774673.585

10

14222579.213

102

16795088.689

26

1770229.440

33

4111163.343

29

2921946.240

4

1158952.273

10

9843617.906

102

1

4
5
Total
1
2
3
4
5
Total
1
2
3
4
5
Total
Regioni 6

N

801053.34

3

Regioni 5

518772.872

2

1

2

Regioni 4

Std. Deviation

3

Total
Regioni 2

Mean
857762.00
1605324.8
2
1449994.8
5
775934.48

1
2
3
4
5
Total

229

Regioni 7

Std. Deviation

2

Mean
9663135.0
8
3516890.3
9
4708377.5
2
5310747.7
5
2707901.3
0
5413371.0
5
5072989.2
7
1886393.0
3
2743960.6
2
2759073.7
5
1337661.7
0
2922906.4
8
1639267.6
9
679461.61

3
4

890697.38
872146.50

5

542201.40

Total
1

978274.68
2561523.2
3
1141063.3
9
1775778.9
3
1448148.0
0
998828.00
1681697.6
2

1 species
1
2
3
4
5
Total

Regioni 8

1
2
3
4
5
Total

Regioni 9

Region20

1

2
3
4
5
Total

N

7944763.196

26

2121205.540

33

4308624.161

29

2255043.379

4

1579604.357

10

5409370.261

102

4674742.309

26

1057982.563

33

2482266.431

29

1158995.351

4

720726.288

10

3059142.982

102

1641703.648

26

305089.755

33

659672.842
318124.448

29
4

234513.486
994126.018

102

2207244.729

26

613938.531

33

1703171.701

29

493069.229

4

548517.525

10

1584044.604

102

10

230

T ests of Between-Subjects Effects

Source
Corrected Model

Dependent Variable
Regioni
Region2
Regions
Region4
Regions
Regions
Region?
Regions
Regions
RegionlO
Regioni 1
Regioni 2
Regioni 3
Regioni 4
Regioni 5
Regioni 6
Regioni?
Regioni 8
Regioni 9
Region20

Intercept

Regioni
Region2
Regions
Region4
Regions
Regions
Region?
Regions
Regions

Type III Sum
of Squares
11S0S022S11
SO.OSO(a)
?S5S?S1S05?
S.SS5(b)
SS5S49SSSSS
02.S0S(c)
55?S4SSS004
08.180(d)
1S419194441
S44.S50(e)
1S2220SS4S1
101.S50(f)
S1SS09S2S11
0400.500(g)
S02SS549?5S
S42.100(h)
1215291?S1S
508.290(1)
4S0SSSS9S0S
85.9200)
2SSS8341S52
08?.S20(k)
99S888S2S22
24.?90(fi)
?S?S2S518?0
28?0.000(l)
21S0S109241
1?5S2.000(m)
40031?582S2
S9166.000(n)
21234?58310
04049.000(o)
S?591318S80
2?39.000(p)
18181414212
5?10.100(q)
1S4?532S?1S
440.S?0(r)
3490981S209
001.890(3)
5??39468854
3S0.900
3?8S9?S9583
S93.?90
9S0??29S091
201.400
9533468?593
180.900
82244441695
85S.100
3?482384642
182.600
24S1S044932
126??.000
25S104?8185
?150.000
?29S?2039?3

df
4
4
4
4
4
4
A

4

4
4
4
4
4
4
4
4
4

4
4
4
1
1
1
1
1

1

1

1

Mean Square
282S505?2?8
?.522
19S4S829?S4
4.S?4
839S24S?090
0 .7 2 6

13933?41501
02.045
3354?98S103
36.164
33055096152
?5.414
?95??455?7?
600.100
15066637439
210.540
30382293296
2?.0?3
12022159900
96.482
66670854130
21.900
24847208155
56.199
1844066296?
5717.500
54007773102
9383.000
10007939565
67291.000
53086895775
1012.000
16897829670
0684.700
45453535531
427.500
41188316791
10.168
87274540522
50.420
57739468854
360.900
37869769583
693.790
96077296091
201.400
95334687593
180.900
62244441695
856.100
37482384642
182.600
24616044932
12677.000
25610478185
7150.000
72967203973

F

Sig.

1.558

,192

2.062

.092

3.214

.016

4.407

.003

6.068

.000

2.246

.070

3.011

.022

4.777

.001

4.887

.001

5.280

.001

4.380

.003

4.787

.001

7.301

.000

5.693

.000

5.909

.000

6.720

.000

7.191

.000

5.776

.000

4.794

.001

3.874

.006

318.177

.000

397.464

.000

367.778

.000

301.558

.000

148.771

.000

25.464

.000

93.148

.000

81.205

.000

117.360

.000

231

Source

Dependent Variable
RegionlO
Regioni 1
Regioni 2
Regioni 3
Regioni 4
Regioni 5
Regioni 6
Regioni?
Regioni 8
RegionlO
Region20

Species

Regioni
Region2
Regions
Region4
Regions
Regions
Region?
Regions
Regions
RegionlO
Regioni 1
Regioni 2
Regioni 3
Regioni 4
Regioni 5
Regioni 6
Regioni?
Regioni 8

Type III Sum
of Squares
585.200
26609?40006
508.210
10959612182
6841.600
41684030965
13?.360
12991682305
?04?8.000
1?113618300
88?41.000
?6?6929???1
8?310.000
13?284312?5
10619.000
1480814?886
43031.000
4201?2?2142
1110.400
4?169150380
924.300
1385800532?
3649.800
11306022911
50.100
?858?31905?
8.698
33584986836
02.8?4
55?34966004
08.150
13419194441
344.620
13222038461
101.??0
31830982311
0402.600
60266549?56
842.400
1215291?318
508.300
48088639603
85.920
26668341652
08?.?10
99388832622
24.?40
?3?626518?0
28?1.000
21603109241
1?536.000
40031?58262
691?2.000
21234?58310
04040.000
6?591318680
2?38.000
18181414212
5?09.?00

df

Mean Square
585.200
1
1
1
■|
-j
■i

1
■|
1
1
1
1
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4

26609?40006
508.210
10959612182
6841.600
41684030965
13?.360
12991682305
?04?8.000
1?113618300
88?41.000
?6?6929???1
8?310.000
13?284312?5
10619.000
1480814?886
43031.000
4201?2?2142
1110.400
4?169150380
924.300
1385800532?
3649.800
2826505?2?8
7 .5 2 5

19646829?64
4.6?5
8396246?090
0.?19
13933?41501
02.039
3354?986103
36.15?
33055096152
?5.444
?95??455???
600.600
1506663?439
210.600
30382293296
2?.0?6
12022159900
96.482
666?0854130
21.920
2484?208155
56.186
1844066296?
5?1?.900
5400???3102
9384.000
1000?939565
6?293.000
53086895??5
1010.000
1689?8296?0
0684.?00
45453535531
42?.400

F

Siq.

116.868

.000

?2.000

.000

80.299

.000

51.435

.000

18.040

.000

45.331

.000

1?.3?8

.000

63.014

.000

53.390

.000

54.899

.000

61.515

.000

1.558

.192

2.062

.092

3.214

.016

4.40?

.003

6.068

.000

2.246

.0?0

3.011

.022

4.???

.001

4.88?

.001

5.280

,001

4.380

.003

4.?8?

.001

?.301

.000

5.693

.000

5.909

.000

6.?20

.000

?.191

.000

5.??6

.000

232

Source

Dependent Variable
RegionlO
Region20

Error

Regioni
Region2
Regions
Region4
Regions
Regions
Region7
Regions
RegionO
RegionlO
Regioni 1
Regioni 2
Regioni 3
Regioni 4
Regioni 5
Regioni 6
Regioni 7
Regioni 8
RegionlO
Region20

Total

Regioni
Regions
Regions
Region4
Regions
Regions
Region7
Regions

Type III Sum
of Squares
1S47S32S71S
440.S70
3400081S209
001.740
17S02SS17S0
7S1.740
02420127S04
26.200
2S3400220S1
S1S.080
30SSSS28723
103.140
S3S2407SS24
S80.S00
142783003S4
0632.300
2S633012280
0S011.000
30S91844204
6264.700
60308612S74
024.400
2208S063728
118.730
1476S031703
2455.100
S03S3S730S1
046.000
24S00617217
72000.000
02020144000
46860.000
16427281880
682210.000
76631023301
76800.000
22704767616
53380.000
76337970230
1340.000
833416138S4
400.600
218S1011172
1106.300
12031348462
S472.000
82032748077
764.900
223S0101022
S82S.000
23833768824
9026.000
2428S621240
OSOO.OOO
2387S1S0361
1738.000
70531367SSS
45900.000
02401218157

df
4
4
07
07
07
97
07
07
07
07
07
07
07
07
07
07
07
07
07
07
07
97

Mean Square
41188316701
10.169
87274540522
50.430
18146070877
0.740
05278481063.
158
26123734084
0.371
31614050230
0.032
55282553221
3.204
14710000335
01.365
26426713601
710.430
31537083808
80.327
62173827306
8.204
22760034771
2.564
15221682168
29.434
51010000080
4.508
25258368265
601.650
04866128763
369.700
16035342145
0332.100
70001054050
276.300
23400760420
416.380
78608047660
21.030
85010180540
7.213
22527743476
40.272

Sig.
4.794

.001

3.874

.006

102
102
102
102
102
102
102
102

233

Source

Dependent Variable
Region9
RegionlO
Regioni 1
Regioni 2
RegionlO
Regioni 4
Regioni 5
Regioni 6
Regioni?
Regioni 8
Regioni 9
Region20

Corrected Total

Regioni
Region2
Regions
Region4
Regions
Regions
Region?
Regions
Region9
RegionlO
Regioni 1
Regioni 2
Regioni 3
Regioni 4
Regioni 5
Regioni 6
Regioni?

Type III Sum
of Squares
930?.000
23984?69146
638?.000
84324340130
532.900
43?1?95?316
6595.900
1488213581?
3833.900
59364050102
51020.000
15449880646
??9600.000
3?325635252
58?350.000
1298034?930
666380.000
594445??321
21220.000
18166189284
15133.000
19?43311?51
8551.000
54189582940
448?.000
18?33164041
911.830
1002?885941
004.990
28698520?45
118.890
36239125323
511.320
6?0432?1066
034.200
15600603201
0?34.000
2881?010512
06312.000
366184992?0
3106.900
?2461529893
432700
2689482?688
504.660
1?431865868
4542.?00
6029245?213
2?0.800
318?6882404
?4960.000
11362325414
164400.000
2043045??06
951380.000
9?865?81611
80850.000
29553899484
56128.000

df

Mean Square

F

Siq.

102
102
102
102
102
102
102
102
102
102
102
102
101
101
101
101
101
101
101
101
101
101
101
101
101
101
101
101
101

234

Type III Sum
of Squares
94519393451
7050.000
Regioni 9
99816940570
940.300
Region20
25342892793
0108.000
a R Squared = .060 (Adjusted R Squared = .022)
b R Squared = .078 (Adjusted R Squared = .040)
c R Squared = .117 (Adjusted R Squared = .081)
d R Squared = .154 (Adjusted R Squared = .119)
e R Squared = .200 (Adjusted R Squared = .167)
f R Squared = .085 (Adjusted R Squared = .047)
g R Squared = .110 (Adjusted R Squared = .074)
h R Squared = .165 (Adjusted R Squared = .130)
i R Squared = .168 (Adjusted R Squared = .133)
j R Squared = .179 (Adjusted R Squared = .145)
k R Squared = .153 (Adjusted R Squared = .118)
I R Squared = .231 (Adjusted R Squared = .200)
m R Squared = .190 (Adjusted R Squared = .157)
n R Squared = .196 (Adjusted R Squared = .163)
0 R Squared = .217 (Adjusted R Squared = .185)
p R Squared = .229 (Adjusted R Squared = .197)
q R Squared = .192 (Adjusted R Squared = .159)
r R Squared = .165 (Adjusted R Squared = .131)
s R Squared = .138 (Adjusted R Squared = .102)
Source

Dependent Variable
Regioni 8

df

Mean Square

F

Siq.

101
101
101

Estimated Marginal Means
1. Grand Mean
Dependent
Variable

Mean

Std. Error

95% Confidence Interval
Lower Bound

Upper Bound
1136978.771

Regioni

1023137.526

57358.752

909296.281

Region2

828598.138
1319799.647

41561.866

746109.339

911086.937

68820.092

1183210.814

1456388.479

1314689.201
1221099.069
824349.214

1164431.190
1022401.931

1464947.211
1419796.207

6680466.025
2154797.512

75707.288
100113.275
163362.179
692180.247

500120.510
5306679.840

8054252.209

239119.275

1680211.943

2629383.080

1150168.513
694573.064

106169.904
64249.527

1360886.382

1409597.805
869325.615
4853223.788

166122.687
97012.323
676706.374
1311454.44
9
1752244.11
1
1196779.29
5

939450.643
567055.539
1079890.257
676783.008
3510148.957

6196298.618

2967297.132

8173045.238

8319819.048

15275258.079

2613662.061

7364213.291

Region3
Region4
Region5
RegionO
Region7
RegionO
RegionO
RegionlO
Regioni 1
Regioni 2
RegionlO
Regioni 4
Region 15
Regioni 6

5570171.185
11797538.56
4
4988937.676

1148577.918

822090.589
1739305.353
1061868.222

235

Dependent
Variable
Region17

Mean
5181410.408

Std. Error
652723.636

95% Confidence Interval
6476886.141
3885934.674

Region18

2760015.674

377730.527

3509706.208
1986171.774

Region19

924754.916

124808.082

2010325.140
677045.422

Region20

1585068.311

202095.418

1183964.849

1172464.409

2. species
Estimates
Dependent
Variable
species

Mean

Std. Error

95% Confidence Interval
Lower Bound

Regioni

Region2

Region3

Region4

997978.385

83544.025

832166.620

1163790.149

2

1053199.273

74155.796

906020.545

1200378.000

3
4

860841.724

79104.843
212996.308

703840.508
562955.039

1017842.940
1485337.936

RegionO

1408432.461

1217974.500

134710.693

950611.064

948036.308

60535.585

827889.929

1068182.686

2

881341.758

53732.920

987986.730

3

725037.724

57318.975

774696.785
611275.425

4

735184.000

428869.636

1041498.364

5
1
2

154336.063
97610.697

659660.686

1047121.114

1612528.423
1448967.242

100237.668

1413584.400

88973.496

1272379.481

1811472.446
1625555.004

3

1165961.517

94911.454

977588.553

1354334.481

4

1051526.250

255556.912

544316.522

1558735.978

5

1320014.800
1722941.692

161628.383
110268.990

999227.201

1640802.399

1504088.272

1941795.113

1485725.758

97877.552

1291465.910

1679985.605

1168943.103

104409.754

1376167.562

1017864.750

281131.865

961718.645
459895.819

1
2

853390.900

838800.023

1575833.681

5

1177970.700

177803.403

825080.163

1530861.237

1

1920144.962

145816.736

1630739.105

2209550.818

2

1130924.242

129430.633

874040.263

1387808.221

3
4

1159925.690
985544.250
908956.200
1509703.769

138068.640
371761.191
235122.422
237939.671
211201.464

885897.658
247701.195
442303.278
1037458.991
288958.466

1433953.721

225296.734

286115.589

606630.022

1180418.893
1824362.920

5
1
2
3
4

Region7

985693.750

5
1

3
4
Regions

Upper Bound

1

708134.909
733267.241

5

620370.250
550269.900

383666.513

-583622.420
-211201.924

1
2

9937021.038

1008172.637
894879.604

7936077.431
4285078.853

6061167.152

1723387.305
1375609.122
1981948.547
1127311.353

1311741.724
11937964.646
7837255.450

236

Dependent
Variable
species
3
4
5
Regions

Regions

Regioni 2

Regioni 4

Regioni 5

1625629.530

1389369.080

7842218.320

3623652.923

348281.406

2932410.741

4314895.105

2018649.182

3

1939260.103

309143.410
329775.178

1405085.088
1284747.661

2632213.275
2593772.546

4

1743232.750

887946.843

-19095.936

3505561.436

5

1449192.600

561586.893

334598.073

2563787.127

1

1714198.615

154638.322

1407284.353

2021112.878

2

1441426.515

137260.897

1169001.631

1713851.399

3

917772.483

146421.483

627166.370

1208378.596

4

810152.250
867292.700
1110107.000

394251.910
249346.801

27671.343
372408.322

1592633.157

687199.758
610353.414

93580.559
83064.478
88608.077

924375.494
522339.762

559390.250

238584.549

1
2

434490.909
85865.964

1362177.078
1295838.506
852059.753
786215.919
1032914.536

5

505814.900

150894.118

206331.846

805297.954

1

2284147.385

241960.599

1803922.574

2764372.195

2

1787259.788

214770.365

3
4

1132913.103
985906.750

229103.818
616880.908

1361000.065
678205.445

2213519.511
1587620.762

-238431.090

2210244.590

5

857762.000

390149.743

83422.760

1632101.240

1
2
3

1449994.846

141300.146

1169553.171

1730436.521

775934.485
801053.345
784444.500

125421.593
133792.043

527007.341
535513.176

1024861.629
1066593.514

360246.100

69455.713

227839.639
985634.670
874874.333

83002.286

1499433.287
987399.514

7702443.248
1608340.384

11614867.213
5081107.132

4
Regioni 3

2570345.973

95% Confidence Interval
4646733.709
8435977.257
1145567.501 11348417.999

2

3
4
Region11

6246992.750
4615793.700

Std. Error
954602.530

1

5
RegionlO

Mean
6541355.483

5
1
2

9658655.231
3344723.758

535200.900

3

4373912.000

933262.137

2521644.998

6226179.002

4

4622287.250

2512885.208

-365094.285

5
1

1589288.151

-887756.342

9609668.785
5420837.742

1910156.344

10309192.068

17891455.240

2

2266540.700
14100323.65
4
3456719.182

1695503.222

91614.626

3

4501569.138

1808658.569

911882.540

6821823.738
8091255.736

4
5
1

3907738.750
1884505.200
22351318.65
4
7512855.970
13591462.34
5
10963865.75
0

4869962.237
3080034.558

-5757788.217
-4228510.800

13573265.717
7997521.200

2552174.197

17285959.350

27416677.958

2265374.552

3016714.571

12008997.369

2416562.259

8795255.198

18387669.491

6506793.017

-1950317.217

23878048.717

2
3
4

237

Dependent
Variable
species
5
Region 16

1

Mean
4568190.100
13284278.50
0

Region17

RegionIS

Reglon19

Reglon20

Std. Error
4115257.239

95% Confidence Interval
-3599456.359 12735836.559

1743129.976

9824648.020

16743908.980
5572482.736

2

2501626.970

1547246.381

-569228.796

3

3911247.862

7187047.946

3848104.250

1650507.288
4444126.881

635447.778

4

-4972257.413

12668465.913

5

1399430.800
9663135.077

2810712.631

-4179055.728
7776252.012

6977917.328

1
2

3516890.394

950703.392

11550018.142

3

4708377.517

843868.445
900186.962

4

5310747.750

2423827.574

5

2707901.300

1532963.158

1

5072989.269

550171.119

2

3

1886393.030
2743960.621

4

2759073.750

5
1

1337661.700
1639267.692

2

3
4
5

872146.500
542201.400

1

2561523.231

293119.753
294355.512

2

1141063.394

261277.418

622500.044

1659626.744

3

1775778.931

278714.682

1222607.439

2328950.423

4

1448148.000

5

998828.000

750462.249
474634.001

-41311.825
56810.894

2937607.825
1940845.106

1842045.125
2921755.582
500120.965
-334606.222

5191735.663
6494999.452
10121374.535
5750408.822

488345.840

3981051.893
917161.719

2855624.342

520937.311

1710044.262

3777876.980

1402666.636

-24831.245

5542978.745

887124.274

-423034.415

181785.154

1278474.483

3098357.815
2000060.902

679461.606

161357.113

359212.402

999710.810

890697.379

172125.846
463463.023

549075.223

1232319.536
1791992.308

-47699.308
-39560.170
1977309.022

6164926.646

1123962.970
3145737.440

Univariate Tests
Dependent
Variable
Regioni

Contrast
Error

Reglon2

Contrast
Error

Region3

Contrast
Error

Reglon4

Contrast
Error

Region5

Contrast

Sum of
Squares
1130602291
150.100
1760256175
0761.740
7858731905
78.698
9242012750
426.290
3358498683
602.875
2534002206
1515.980
5573496600
408.150
3066562872
3103.140
1341919444

df
4
97
4
97
4
97
4
97
4

Mean Square
28265057278
7.525
18146970877
0.740
19646829764
4.675
95278481963.
158
83962467090
0.719
26123734084
0.371
13933741501
02.039
31614050230
0.032
33547986103

F

SIg.

1.558

.192

2.062

.092

3.214

.016

4.407

.003

6.068

.000

238

Dependent
Variable

Sum of
Squares
1344.630
Error

Regione

Contrast
Error

Region?

Contrast
Error

Regions

Contrast
Error

Regions

Contrast
Error

Region 10

Contrast
Error

Region 11

Contrast
Error

Regioni 2

Contrast
Error

RegionIS

Contrast
Error

Regioni 4

Contrast
Error

RegionIS

Contrast
Error

RegionIS

Contrast
Error

Regioni?

Contrast
Error

RegionIS

Contrast
Error

RegionIS

Contrast
Error

536240?662
4688.500
1322203S46
1101.?S0
142?S3SS35
4S632.300
31S30SS231
10402.?00
25633S122S
0S5S11.000
6026654S?5
6842.400
305S1S442S
46264.?00
12152S1?31
8508.300
603086125?
4324.400
4808863S60
385.320
22085S63?2
8118.?30
2666834165
208?.?10
14?65031?0
32455.100
9938883262
224.?40
503535?3S5
1046.000
?3?626518?
028?2.000
2450061?21
??2090.000
2160310324
11?53?.000
9202014430
046860.000
40031?5826
2631?3.000
1642?28188
0682210.00
0
21234?5831
004042.000
?663102330
1?6800.000
6?53131868
02?39.000
22?34?6?61
653383.000
1818141421
25?09.800
?633?3?323
31340.000
164?5326?1
6440.680
8334161385

df

Mean Square
36.153
3?
4
9?
4
9?
4
3?
4
3?
4
3?
4
3?
4
9?
4
3?

3?
4

A

3?
4
3?
4
A

3?
4
9?
4
3?

55282553221
3.234
33055036152
?5.445
14?13333335
01.365
?35??455???
600.600
26426?13631
?10.430
1506663?433
210.610
3153?383808
89.32?
30382233236
2?.0?6
621?382?396
8.294
12022153300
36.482
22?63034??1
2.564
666?0854130
21.320
15221682168
23.434
2484?208155
56.18?
51310900980
4.538
1844066236?
5?18.000
25258368265
631.650
5400???3102
3384.000
34866128?63
363.?00
1000?333565
6?233.000

F

Siq.

2.246

.070

3.011

.022

4.???

.001

4.88?

.001

5.280

.001

4.380

.003

4.?8?

.001

?.301

.000

5.693

.000

5.303

.000

6.720

.000

7.131

.000

5.776

.000

4.734

.001

16935342145
0332.100
53086835??5
1010.000
?3001054350
2?6.300
1683?8236?0
0684.800
23433?60429
416.380
45453535531
42?.400
?869894?663
21.030
41188316?31
10.1?1
85313183540

239

Dependent
Variable
Reglon20

Sum of
Squares
4499.600

df

Mean Square
7.213

F

SIg.

Contrast

3490981620
87274540522
4
3.874
.006
9001.750
50.430
Error
2185191117
22527743476
97
21106.300
40.272
The F tests the effect of species. This test is based on the linearly Independent pairwise comparisons
among the estimated marginal means.

240

Post Hoc Tests
species
Multiple Comparisons
Tukey HSD
(J)

Dependent
Variable
Regioni

(1) species
1

specie
s
2
3

3

4

-219996.12

158513.643
111708.040

.988

-255300.24

220633.19
365742.02

3
4
5

55220.89
192357.55
67505.52
-164775.23

108428.125
225536.049
153772.732

.395
.998
.821

-109046.22
-559429.74
-592225.95

493761.31
694440.78
262675.50

115052.946

.756

-456955.81

182682.49

2

3

4

-623708.98

648278.25

-660625.42

1

-137136.66
-192357.55

108428.125

.395

-493761.31

109046.22

-124852.03

227211.363

.982

-756444.25

506740.20

5
1

-357132.78

156219.548

-791385.05

77119.50

-12284.63

228794.736

.158
1.00
0
.998
.982

-648278.25

623708.98

1
2

-67505.52

225536.049

124852.03
-232280.75

227211.363
252020.630

-694440.78

559429.74

-506740.20

756444.25

.888

219996.12

158513.643

.637

-932836.69
-220633.19

468275.19
660625.42

164775.23
357132.78
232280.75

153772.732
156219.548
252020.630

.821
.158
.888

-262675.50
-77119.50
-468275.19

592225.95
791385.05
932836.69

2

66694.55

80943.089

.923

-158307.51

291696.62

3

222998.58

83366.791

.065

-8740.78

454737.95

4

212852.31
94645.41
-66694.55
156304.03
146157.76

165783.526

.702

114858.196
80943.089
78566.479
163422.297

.923
.923
.279
.898

-247985.50
-224632.39
-291696.62
-62091.63
-308116.41

673690.11

5
1
3
4
5
1

27950.86
-222998.58

111422.955

.999

-281777.81

83366.791

2
4

-156304.03

78566.479

-10146.28

164636.221

5
1

-128353.18
-212852.31

113195.906
165783.526

.065
.279
1.00
0
.788
.702

2

-146157.76

163422.297

10146.28

164636.221

5

-118206.90

1

-94645.41

3

5

J)

2
4

3
4
1

228794.736

1

5

Region2

12284.63

95% Confidence Interval
Lower and Upper
-365742.02
255300.24
-182682.49
456955.81

5

2
3
5

-55220.89
137136,66

Std. Error
111708.040
115052.946

Big.
.988
.756
1.00
0
.637

4

2

Mean Difference (1-

413923.20
158307.51
374699.70
600431.92
337679.52

-454737.95
-374699.70

8740.78
62091.63

-467794.85

447502.30

-443010.21
-673690.11

186303.86
247985.50

-600431.92

308116.41

-447502.30

467794.85

182612.893

.898
1.00
0
.967

-625826.24

389412.44

114858.196

.923

-413923.20

224632.39

241

(J)

Dependent
Variable

(1) species

specie
s
2
3
4

Regions

1

2

3

4

2

-209007.82

536130.18

.014

62841.97

830291.85

4

561002.17
292513.62
-163561.18

.253
.541
.740

-202074.74
-236161.57
-536130.18

1324079.08
821188.81
209007.82

.198

-78624.09

644635.54

3
4

283005.73

274512.159
190187.602
134029.374
130094.071

397440.99

270602.325

.585

-354767.53

1149649.52

5

128952.44

184499.369

.956

-383910.85

641815.73

2
4
5

-446566.91 n
-283005.73
114435.27
-154053.28

138042.654
130094.071
272612.398
187435.104

.014
.198
.993
.923

-830291.85
-644635.54
-643360.77
-675077.20

1

-561002.17

274512.159

.253

-1324079.08

-62841.97
78624.09
872231.30
366970.63
202074.74

2

-397440.99
-114435.27

270602.325
272612.398

.585

-1149649.52

.993

-872231.30

354767.53
643360.77

-268488.55
-292513.62
-128952.44

.901
.541
.956

-1109028.53
-821188.81
-641815.73

572051.43
236161.57
383910.85

-366970.63
-572051.43

675077.20
1109028.53

5
1

1

3
4

154053.28

302379.016
190187.602
184499.369
187435.104

268488.55

302379.016

.923
.901

2

237215.93

147442.414

.495

-172638.05

647069.92

3
4
5

553998.59(1
705076.94
544970.99

151857.324
301984.065
209220.698

.004
.143
.077

131872.23
-134365.17
-36611.57

976124.95
1544519.06
1126553.55

1

-237215.93
316782.65
467861.01

147442.414

-647069.92

1

.495
.183
.519

-81037.40
-359625.07

172638.05
714602.71
1295347.08

2

307755.06
-553998.59(1
-316782.65

143113.283
297682.953
202963.212
151857.324
143113.283

.555
.004
.183

-256433.22
-976124.95
-714602.71

871943.33
-131872.23
81037.40

4

151078.35

299894.185

.987

-682554.41

984711.11

-9027.60

206192.742

1.00
0

-582193.18

564137.98

-705076.94

301984.065

.143

134365.17

-467861.01
-151078.35
-160105.95
-544970.99
-307755.06

297682.953
299894.185
332639.709
209220.698
202963.212

.519
.987
.989
.077

-1544519.06
-1295347.08
-984711.11
-1084763.29
-1126553.55
-871943.33

256433.22

-564137.98

582193.18

-764551.39
247240.87
202010.77
-175454.89

1331200.56
1318427.77
2044656.31

1

1
2
3
5
1
3

1

118206.90

625826.24

.740

2

Regions

-389412.44

138042.654

5

5

.967

163561.18

5

4

-27950.86
128353.18

446566.91 n

3
4
3

95% Confidence Interval
Lower and Upper
-337679.52
281777.81
443010.21
-186303.86

3

2

1

Sig.
.999
.788

182612.893
134029.374

5

Region4

J)

2

3
5

Std. Error
111422.955
113195.906

Mean Difference (1-

.555

9027.60

206192.742

1.00
0

4

160105.95

332639.709

.989

2

789220.72(1
760219.27(1
934600.71

194973.868
200812.026
399335.578

.001
.002

3
4

.141

359625.07
682554.41
764551.39
36611.57

1084763.29

242

(J)

Dependent
Variable

(!) species

specie
s
5

2

1

1011188.760

Std. Error
276667.804

Sig.
.004

-789220.72(*)

194973.868

-29001.45

189249.143

4
5

145379.99
221968.04

1

-760219.270

393647.903
268393.074
200812.026

29001.45

189249.143

174381.44

396571.976

.001
1.00
0
.996
.922
.002
1.00
0
.992

3

3

2
4
4

2

1239625.24
968035.39
-202010.77

-497065.06

555067.95

-927992.02

1276754.90

-506969.20

1008908.18

-2044656.31
-1239625.24
-1276754.90

175454.89
948865.26
927992.02

76588.05

439873.773

-1146153.85

1299329.95

-1011188.760
-221968.04

276667.804

.141
.996
.992
1.00
0
.004

-1780257.84

-242119.68

268393.074

.922

-968035.39

524099.30

3
4

-250969.49

272663.717

-1008908.18

506969.20

-76588.05

439873.773

-1299329.95

1146153.85

2
3
4

801568.86
776436.53
889333.52

318153.172
327679.722
651625.172

.888
1.00
0
.095
.133
.651

-82819.43
-134433.25
-922025.68

1685957.15
1687306.31
2700692.72

5

959433.87

1

-801568.86

451459.163
318153.172

-25132.33

308811.718

87764.66

642344.177

157865.01
-776436.53

437956.678
327679.722

25132.33

308811.718

112896.99

647115.602

182997.34
-889333.52

444925.400
651625.172

-87764.66

642344.177

-112896.99

647115.602

70100.35

717774.322

1

-959433.87

2

-157865.01
-182997.34

451459.163
437956.678
444925.400

-70100.35

717774.322

1

5
1

5
1
2
3
5

3
4
1

-948865.26
-524099.30
-1318427.77

.888

4

Region?

497065.06

272663.717

2

5

-555067.95

399335.578
393647.903
396571.976

4

4

-247240.87

250969.49

3

3

-1331200.56

-934600.71
-145379.99
-174381.44

2

1

95% Confidence Interval
Lower and Upper
1780257.84
242119.68

1

5

Regione

J)

5
2
3

5

Mean Difference (1-

2
3

3875853.890
3395665.56

1348043.60
9
1388408.46

.218
.095
1.00
0
1.00
0
.996

-295512.59

2214380.33

-1685957.15

82819.43

-883553.66

833288.99

-1697795.63

1873324.95

-1059547.84

1375277.86

-1687306.31

134433.25

-833288.99

883553.66

-1685926.70

1911720.68

-1053786.86
-2700692.72

1419781.54
922025.68

-1873324.95

1697795.63

-1911720.68

1685926.70

-1925137.36

2065338.06

-2214380.33
-1375277.86
-1419781.54

295512.59
1059547.84
1053786.86

-2065338.06

1925137.36

.039

128621.14

7623086.64

.112

-463771.65

7255102.76

.133
1.00
0
1.00
0
.994
.651
1.00
0
1.00
0
1.00
0
.218
.996
.994
1.00
0

243

(J)

Dependent
Variable

(1) species

specie
s

10638548.40

.039

-7623086.64

-128621.14

.996

-4117396.71

3157020.05

1.00
0

-7751405.18

7379753.98

.936

-3712914.30

6603661.20

.112

-7255102.76

463771.65

.996

-3157020.05

4117396.71

1.00
0

-7327415.05

7916140.52

.845

-3314804.12

7165927.68

79389.08

452019.155

1
2
3
5
1
2
3
4

4

3690028.29
5321227.340
-3875853.890
-480188.33
-185825.60
1445373.45
-3395665.56
480188.33
294362.73
1925561.78
-3690028.29
185825.60
-294362.73
1631199.05
-5321227.340
-1445373.45
-1925561.78
-1631199.05

.669
1.00
0
1.00
0
.983
.050

11364920.15

3984863.57

-7379753.98

7751405.18

-7916140.52

7327415.05

-6822802.63

10085200.73
-3906.28

10638548.40

.936

-6603661.20

3712914.30

.845

-7165927.68

3314804.12

.983
.007

6822802.63

10085200.73
310491.83
351118.97

2899515.66
3017666.67

-770933.41
337548.68
-2899515.66

4531773.75
4011371.97
-310491.83

-1177114.03

1335892.18

4

275416.43

940222.975

.288
.012
.007
1.00
0
.998

-2338174.33

2889007.20

5

569456.58

641053.419

.901

-1212515.72

2351428.88

1

-1684392.820

479636.952

-3017666.67

-351118.97

-79389.08

452019.155

-1335892.18

1177114.03

196027.35

947207.086

-2436977.54

2829032.24

5
1

490067.50
-1880420.17

651253.796

.006
1.00
0
1.00
0
.943

953807.912

.288

-1320259.36
-4531773.75

2300394.36
770933.41

2

-275416.43

940222.975

.998

-2889007.20

3

-196027.35

947207.086

1.00

-2829032.24

2338174.33
2436977.54

5
1

3

2
4

4

3906.28

.006

5

3

.050

479636.952

4

2

11364920.15

953807.912
660817.506
465692.587

1
2

1

-3984863.57

1684392.820
1880420.17
2174460.320
-1605003.740

5

Regions

.669

1605003.740

4

5

95% Confidence Interval
Lower and Upper

3

1
3

4

Std. Error

2

5

3

J)

1
2760994.47
4
1912873.08
4
1348043.60
9
1308463.02
8
2721670.06
2
1855661.84
3
1388408.46
1
1308463.02
8
2741887.01
7
1885188.94
5
2760994.47
4
2721670.06
2
2741887.01
7
3041274.37
0
1912873.08
4
1855661.84
3
1885188.94
5
3041274.37
0
465692.587

4

2

Sig.

Mean Difference (1-

244

Dependent
Variable

(!) species

(J)
specie
s
5

5

Regions

1

-2174460.320
-569456.58

3
4

-490067.50

5

3

4

-2626463.23

3214543.53

.012
.901

-4011371.97
-2351428.88

-337548.68
1212515.72

.943

-2300394.36

1320259.36

.999

-3214543.53

2626463.23

.680

206769.350
200698.292

.680
.077

-847540.51
-34238.32

1081546.38

417462.720

.557

-529170.37

1791718.90

574133.82
-796426.130
-523654.03

284630.253
212960.704
200698.292

.266
.003
.077

-217068.93
-1388405.00
-1081546.38

1365336.56
-204447.27
34238.32

4
5

107620.23

420563.692

-1061444.35

1276684.82

50479.78

289159.261

-753312.50

854272.07

1

-904046.37

-2081257.84

2
3
5

-631274.27
-107620.23

423494.485
417462.720
420563.692

.999
1.00
0
.214

-57140.45

466485.151

1

-846905.920

293405.586

2

-574133.82

284630.253

-50479.78

289159.261

57140.45

466485.151

5
1

422907.240
499753.590
550716.75
604292.100
-422907.240

125128.049
128874.793
256280.916
177556.627

3
4

846905.92(3

1

-272772.10

3

523654.03
631274.27

1
2

847540.51

301996.31

.557
.999
1.00
0
.038

-1791718.90
-1276684.82

273165.11
529170.37
1061444.35

-1353855.50

1239574.60

-1662501.95

-31309.88

-1365336.56

217068.93

-854272.07

753312.50

-1239574.60

1353855.50

75081.76
141513.06
-161681.75
110727.94

770732.73
857994.11
1263115.25
1097856.26

125128.049

.266
1.00
0
1.00
0
.009
.002
.208
.008
.009

-770732.73

-75081.76

76846.34

121454.102

.969

414459.15

127809.51
181384.86
-499753.590
-76846.34

252630.747
172246.168
128874.793
121454.102

.987
.830
.002

-260766.46
-574442.41
-297417.52
-857994.11
-414459.15

830061.43
660187.23
-141513.06
260766.46

50963.16

254507.325

-656505.18

758431.51

5

104538.51

174986.931

-381882.52

590959.54

1

-550716.75
-127809.51

256280.916
252630.747

-1263115.25
-830061.43

161681.75
574442.41

-50963.16

254507.325

-758431.51

656505.18

53575.35

282297.045

-731141.65

838292.35

2
3
4

2

.999

1388405.00
2081257.84
1662501.95

796426.130
904046.37

4
1

95% Confidence Interval
Lower and Upper

-301996.31
204447.27
-273165.11
31309.88

272772.10

3

3

Regioni 0

1050632.87
3
660817.506
641053.419

Sig.
0

.003
.214
.038

-294040.15

2

5

4

Std. Error

651253.796
1050632.87
3
206769.350
212960.704
423494.485
293405.586

4
3

294040.15

1
2

4
5
2

Mean Difference (1J)

5
1
2
4

2
3
5

.969
1.00
0
.975
.208
.987
1.00
0
1.00
0

245

(J)

Dependent
Variable

Regioni 1

(!) species
5

1

specie
s
1

3

-604292.100

.008

-1097856.26

-110727.94

-181384.86

172246.168

.830

-660187.23

297417.52

3
4

-104538.51

174986.931

-590959.54

381882.52

-53575.35

282297.045

-838292,35

731141.65

496887.60

323529.351
333216.882

.975
1.00
0
.542
.007
.294

-402445.16
224972.54

1396220,36
2077496.02

-543727.06

3140208.33

2
3
4

662636.390

1

1426385.380
-496887.60

323529.351

3
4
5

654346.68
801353.04
929497.79

1

-1151234.280
-654346.68

1

2

2702538.04
402445.16

314030.044
653198.564
445357.307
333216.882
314030.044

.236
.736
,234
.007

-218580.32
-1014379.80
-308487.01
-2077496.02

1527273.69
2617085.88
2167482.58
-224972.54

.236

-1527273.69

218580.32

147006.35

658050.616

.999

-1682214.01

1976226.72

452443.788
662636.390
653198.564

.973
.294
.736

-982532.38
-3140208.33
-2617085.88

3

-147006.35

658050.616

-1976226.72

128144.75

729903.333

-1900808.66

2157098.16

-1426385.380
-929497.79
-275151.10

459087.958
445357.307
452443.788

-2702538.04
-2167482.58
-1532834.59

-150232.73
308487.01
982532.38

-128144.75

729903.333

.999
1.00
0
.020
.234
.973
1.00
0
.005
.010
.427

1532834.59
543727.06
1014379.80
1682214.01

-2157098.16

1900808.66

148868.03

1199252.69
1189859.81

1
2

2

674060.360

188934.664

3
4

648941.500
665550.35

194591.988
386966.386

5

914793.950
-674060.360

268098.176
188934.664

-25118.86

183387.259

-8510.02

381454.884

5

240733.58

260079.751

1
2

-648941.500

194591.988

25118.86

183387.259

16608.84

384288.385

265852.44
-665550.35

264218.114
386966.386

8510.02

381454.884

-16608.84

384288.385

1

4
5
1
2
3
5
5

150232.73
-1396220.36

275151.10
-1298240.63
-801353.04

4

4

.020
.542

5
1
2

3

3

459087.958

4

3
4
Regioni 2

J)

1151234.280
1298240.63

5
5

95% Confidence Interval
Lower and Upper

2

2

4

Siq.

Std. Error
177556.627

5
2

Mean Difference (1-

1

2

.008
.005
1.00
0
1.00
0
.886

249243.60

426248.934

.010
1.00
0
1.00
0
.852
.427
1.00
0
1.00
0
.977

-914793.950
-240733.58

268098.176
260079.751

.008
.886

108023.19
-410121.91

1741222.61

169546.34
-1199252.69

1660041.55
-148868.03

-534890.75

484653.03

-1068861.64

1051841.61

-482224.75

963691.92

-1189859.81

-108023.19

-484653.03

534890.75

-1051619.23

1084836.92

-468609.53
-1741222.61

1000314.42
410121.91

-1051841.61

1068861.64

-1084836.92

1051619.23

-935624.57

1434111.77
-169546.34
482224.75

-1660041.55
-963691.92

246

(J)

Dependent
Variable

(!) species

specie
s
3
4

Regioni 3

1

2
3
4
5

2

1
3
4
5

3

1
2
4
5

4

1
2
3
5

5

1
2
3
4

Regioni 4

1

2
3
4
5

2

1

3
4
5

Std. Error
264218.114

Sig.
.852

95% Confidence Interval
Lower and Upper
-1000314.42
468609.53

426248.934
1317907.73
6
1357370.22
2
2699271.71
1
1870110.29
9
1317907.73
6
1279211.99
0
2660826.40
6
1814178.03
0
1357370.22
2
1279211.99
0
2680591.40
5
1843045.04
6
2699271.71
1
2660826.40
6
2680591.40
5
2973285.87
5
1870110.29
9
1814178.03
0
1843045.04
6
2973285.87
5
2554100.31
8
2630578.46
9
5231178.59
1
3624266.84

.977

-1434111.77

935624.57

.000

2650469.10

9977393.84

.002

1511584.77

9057901.69

.343

-2466949.64

12539685.60

.001

2193663.59

12590565.47

.000

-9977393.84

-2650469.10

.929

-4585085.87

2526709.38

.989

-8674012.53

6118885.55

.976

-3964789.80

6121155.92

.002

-9057901.69

-1511584.77

.929

-2526709.38

4585085.87

1.00
0

-7699766.17

7203015.67

.783

-3015844.83

7230587.43

Mean Difference (1J)

-265852.44
-249243.60
6313931.47(3
5284743.23(3
5036367.98
7392114.53(3
-6313931.47(3
-1029188.24
-1277563.49
1078183.06
-5284743.23(3
1029188.24
-248375.25
2107371.30
-5036367.98
1277563.49
248375.25
2355746.55
-7392114.53(3
-1078183.06
-2107371.30
-2355746.55
10643604.47(3
9598754.52(3
10192584.90
12215818.45(3
-10643604.47(3
-1044849.96
-451019.57
1572213.98

1

2554100.31
8
2479108.10
5
5156671.73
3
3515870.31
2

.343

12539685.60

2466949.64

.989

-6118885.55

8674012.53

1.00
0

-7203015.67

7699766.17

.932

-5909263.68

10620756.78

.001

12590565.47

-2193663.59

.976

-6121155.92

3964789.80

.783

-7230587.43

3015844.83

.932

10620756.78

5909263.68

.001

3543827.95

17743380.99

.004

2286387.37

16911121.66

.299

-4348816.89

24733986.70

.009

2141239.77

22290397.14

.001

17743380.99

.993

-7936166.40

1.00
0

14785310.45

.992

-8201048.77

-3543827.95
5846466.49
13883271.31
11345476.73

247

(J)
Dependent
Variable

(!) species
3

specie
s
1
2
4
5

4

1
2
3
5

5

1
2
3
4

RegionIS

1

2
3
4
5

2

1
3
4
5

3

1
2
4
5

4

1
2
3
5

5

1
2

Mean Difference (i-

J)
-9598754.52(3
1044849.96
593830.39
2617063.94
-10192584.90
451019.57
-593830.39
2023233.55
-12215818.45(3
-1572213.98
-2617063.94
-2023233.55
14838462.68(3
8759856.31
11387452.90
17783128.55(3
-14838462.68(3
-6078606.38
-3451009.78
2944665.87
-8759856.31
6078606.38
2627596.59
9023272.24
-11387452.90
3451009.78
-2627596.59
6395675.65
-17783128.55(3
-2944665.87

Std. Error
2630578.46
9
2479108.10
5
5194976.22
8
3571814.48
2
5231178.59
1
5156671.73
3
5194976.22
8
5762217.02
7
3624266.84
1
3515870.31
2
3571814.48
2
5762217.02
7
3412552.56
3
3514735.59
3
6989416.89

2
4842410.06
9
3412552.56
3
3312354.90
5
6889867.72
2
4697580.65
5
3514735.59
3
3312354.90
5
6941046.64
4
4772328.08
0
6989416.89
2
6889867.72
2
6941046.64
4
7698941.32
4
4842410.06
9
4697580.65

Sig.
.004

95% Confidence interval
Lower and Upper
16911121.66

.993

-5846466.49

1.00
0

13846937.66

.948

-7311709.97

.299
1.00
0
1.00
0
.997
.009
.992
.948
.997

24733986.70
13883271.31
15034598.43
13994325.70
22290397.14
11345476.73
12545837.85
18040792.80

-2286387.37
7936166.40
15034598.43
12545837.85
4348816.89
14785310.45
13846937.66
18040792.80
-2141239.77
8201048.77
7311709.97
13994325.70

.000

5352398.10

24324527.27

.101

-1010252.20

18529964.81

.483

-8041422.93

30816328.74

.004

4322408.52

31243848.59

.000
.360
.987
.970
.101
.360
.996
.329
.483
.987
.996
.920
.004
.970

24324527.27
15286145.88
22603163.18
10113463.68
18529964.81
-3128933.13
16666821.74
-4242636.93
30816328.74
15701143.62
21922014.93
15005505.22
31243848.59
-

-5352398.10
3128933.13
15701143.62
16002795.42
1010252.20
15286145.88
21922014.93
22289181.42
8041422.93
22603163.18
16666821.74
27796856.52
-4322408.52
10113463.68

248

(J)

Dependent
Variable

(!) species

specie
s

Mean Difference (1J)

Std. Error

Sig.
5

3
4
Region16

1

2
3
4
5

2

1
3
4
5

3

1
2
4
5

4

1
2
3
5

5

1
2
3
4

Region17

1

2
3
4
5

2

1
3
4

-9023272.24
-6395675.65

10782651.53(3
9373030.64(3
9436174.25
11884847.70(3
-10782651.53(3
-1409620.89
-1346477.28
1102196.17
-9373030.64(3
1409620.89
63143.61
2511817.06
-9436174.25
1346477.28
-63143.61
2448673.45
-11884847.70(3
-1102196.17
-2511817.06
-2448673.45
6146244.68(3
4954757.56(3
4352387.33
6955233.78(3
-6146244.68(3
-1191487.12
-1793857.36

4772328.08
0
7698941.32
4
2330766.71
4
2400557.52
3
4773758.04
3
3307356.58
9
2330766.71
4
2262331.91
0
4705766.15
4
3208438.38
3
2400557.52
3
2262331.91
0
4740721.25
8
3259490.72
8
4773758.04
3
4705766.15
4
4740721.25
8
5258361.83
9
3307356.58
9
3208438.38
3
3259490.72
8
5258361.83
9
1271200.57
1
1309264.49
1
2603608.46
7
1803832.85
9
1271200.57
1
1233876.21
7
2566525.67
9

.329
.920

95% Confidence Interval
Lower and Upper
16002795.42
22289181.42
27796856.52

4242636.93
15005505.22

.000

4303688.01

17261615.05

.002

2700065.67

16045995.60

.285

-3833709.88

22706058.38

.005

2691202.29

21078493.11

.000
.971
.999
.997
.002

17261615.05
-7698352.34
14427360.53
-7816480.73
16045995.60

.971

-4879110.55

1.00
0

13114906.31

.938

-6548772.91

.285
.999
1.00
0
.990
.005
.997
.938
.990

22706058.38
11734405.97
13241193.53
12168291.15
21078493.11
10020873.07
11572407.04
17065638.05

-4303688.01
4879110.55
11734405.97
10020873.07
-2700065.67
7698352.34
13241193.53
11572407.04
3833709.88
14427360.53
13114906.31
17065638.05
-2691202.29
7816480.73
6548772.91
12168291.15

.000

2612616.85

9679872.51

.002

1315321.31

8594193.81

.456

-2885009.78

11589784.44

.002

1941017.97

11969449.58

.000

-9679872.51

-2612616.85

.870

-4621362.35

2238388.10

.956

-8928173.35

5340458.64

249

(J)

Dependent
Variable

(1) species

3

specie
s
5
1
2
4
5

4

1
2
3
5

5

1
2
3
4

RegionIS

1

1

5
1

2
4
5
1
2
3
5
5

1191487.12
-602370.23
2000476.22
-4352387.33
1793857.36
602370.23
2602846.45
-6955233.780
-808989.09
-2000476.22
-2602846.45
2329028.650

4

4

-4954757.56(*)

3186596.240

3

3

808989.09

3

5
2

J)

2
4

1
2
3

Sig.

95% Confidence Interval
Lower and Upper

.990

-4055258.84

5673237.03

.002

-8594193.81

-1315321.31

.870

-2238388.10

4621362.35

.999

-7789680.95

6584940.48

.793

-2941171.13

6942123.57

Mean Difference (1-

2313915.52
3735327.570
-3186596.240
-857567.59
-872680.72
548731.33
-2329028.650
857567.59
-15113.13
1406298.92
-2313915.52
872680.72
15113.13
1421412.05
-3735327.570
-548731.33
-1406298.92

Std. Error
1749882.85
2
1309264.49
1
1233876.21
7
2585590.19
8
1777726.81
0
2603608.46
7
2566525.67
9
2585590.19
8
2867911.46
1
1803832.85
9
1749882.85
2
1777726.81
0
2867911.46
1
735642.522
757670.075
1506705.66
2
1043876.30
3
735642.522
714042.955
1485245.88
9
1012655.48
7
757670.075
714042.955
1496278.50
8
1028768.75
8
1506705.66
2
1485245.88
9
1496278.50
8
1659657.54
6
1043876.30
3
1012655.48
7
1028768.75

.456

11589784.44

2885009.78

.956

-5340458.64

8928173.35

.999

-6584940.48

7789680.95

.893

-5369248.58

10574941.48

.002

11969449.58

-1941017.97

.990

-5673237.03

4055258.84

.793

-6942123.57

2941171.13

.893
.000
.022

10574941.48
1141689.28

5369248.58
5231503.19

222890.46

4435166.83

.542

-1874359.08

6502190.12

.005

833605.81

6637049.33

.000
.751

-5231503.19
-2842433.01

-1141689.28
1127297.83

.977

-5001302.38

3255940.94

.983

-2266204.17

3363666.83

.022
.751
1.00
0

-4435166.83
-1127297.83

-222890.46
2842433.01

-4174402.78

4144176.52

.650

-1453427.55

4266025.39

.542

-6502190.12

1874359.08

.977

-3255940.94

5001302.38

1.00
0

-4144176.52

4174402.78

.912

-3192031.51

6034855.61

.005

-6637049.33

-833605.81

.983

-3363666.83

2266204.17

.650

-4266025.39

1453427.55

250

(J)

Dependent
Variable

(!) species

specie
s

Sig.

95% Confidence Interval
Lower and Upper

.912

-6034855.61

3192031.51

.001

284136.80

1635475.37

Mean Difference (1Std. Error

J)

8

2

959806.09(3

1659657.54
6
243067.810

3

748570.31 (*)
767121.19
1097066.29(3

250346.058
497839.147
344913.078

.028
.539
.017

52669.27
-616750.33
138291.98

1444471.35
2150992.71
2055840.60

-959806.09(3
-211235.77

243067.810

.001

-1635475.37

-284136.80

4
Regioni 9

1

4
5
2

1
3
4

1

1293394.39

497839.147

.539

-2150992.71

616750.33

490748.502

.995
1.00
0
.975
.017
.994

-1171476.36

1556846.15

-1392845.32

1355743.56

-1194408.90
-2055840.60

1854299.10
-138291.98

-1067358.98
-1293394.39
-1854299.10

792838.57
596402.43
1194408.90

326382.72
-341093.08
-1127457.97

2514536.96
1912581.68
3354208.44

10200.45

3115190.02

-2514536.96
-1696668.91
-2516001.97
-1363826.54
-1912581.68
-427237.84

-326382.72
427237.84

339921.309

-767121.19

2

192684.89

5
1

2

1
3
4

-18550.88

494393.852

329945.10
-1097066.29(3
-137260.21
-348495.98

548376.844
344913.078

-329945.10
1420459.84(3
785744.30
1113375.23
1562695.23(3
-1420459.84(3
-634715.54
-307084.61
142235.39
-785744.30
634715.54

5
1

334597.232
339921.309
548376.844
393587.419
405372.719
806125.768
558500.315
393587.419
382031.103

.975
.004
.304
.641
.048
.004
.463
.995
.999
.304

382031.103

.463
.994

327630.93

800546.976

5

776950.93

550417.395

1

-1113375.23
307084.61
-327630.93
449320.00

1

.843

794644.245
541796.386
405372.719

4

2
3
5
5

-596402.43

348495.98

2

4

1392845.32

1

5

3

-1355743.56

5

3
4
2

-444594.85

494393.852

2
3
4
Region20

444594.85
1171476.36
1067358.98
-52669.27
867066.39

18550.88

1

3
5

-867066.39
-1556846.15
-792838.57
-1444471.35

235930.975
490748.502
334597.232
250346.058
235930.975

2
4

4

.898
.995
.994
.028
.898
1.00
0
.843

-192684.89
137260.21
-748570.31(3
211235.77

5
3

-1421412.05

-1897694.59

806125.768
794644.245
800546.976
887958.908
558500.315

.622
.641
.995
.994
.987
.048

-753075.31
-3354208.44
-1901832.76
-2552956.45
-2018989.39
-3115190.02

2

-1562695.23(3
-142235.39

541796.386

.999

-1648297.33

3
4

-776950.93
-449320.00

550417.395
887958.908

.622
.987

-2306977.17
-2917629.39

1901832.76
1648297.33
341093.08
1696668.91
2552956.45
2306977.17
1127457.97
2516001.97
1897694.59
2917629.39
-10200.45
1363826.54
753075.31
2018989.39

Based on observed means.
* The mean difference is significant at the .05 level.

251