THE INFLUENCE OF WARMING, SITE CHARACTERESTICS, AND HOST PLANT ON
ROOT-ASSOCIATED FUNGAL COMMUNITIES FROM ALEXANDRA FIORD IN THE
CANADIAN HIGH ARCTIC
by
Kei E. Fujimura
B.S., University o f Wisconsin, 1991
M.S., Oregon State University, 1999

DISSERTATION SUBMITTED IN PARTIAL FULFILLMENT OF
THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
in
NATURAL RESOURCES AND ENVIRONMENTAL STUDIES

THE UNIVERSITY OF NORTHERN BRITISH COLUMBIA
August 2005

© Kei E. Fujimura, 2005

1^1

Library and
Archives Canada

Bibliothèque et
Archives Canada

Published Heritage
Branch

Direction du
Patrimoine de l'édition

395 W ellington Street
Ottawa ON K 1A 0N 4
Canada

395, rue W ellington
Ottawa ON K 1A 0N 4
Canada
Your file Votre référence
ISBN: 978-0-494-28458-2
Our file Notre référence
ISBN: 978-0-494-28458-2

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
Arctic systems are expected to be impacted earlier and more severely by global
warming than temperate ecosystems. However, much o f the research on the impact o f
warming on arctic ecosystems has centered on plant communities. One objective o f this
thesis was to examine how passive warming would impact the root-associated fungal
community at Alexandra Fiord, Nunavut. The root-associated fungal community consists
mostly o f mycorrhizal, dark-septate and hyaline-septate fungi, which are considered
important mutualists in arctic ecosystems. The objective was to compare the fungal
community from plots warmed by open-top chambers to ambient plots, using two
methodologies: 1) fungal DNA extraeted directly from root tips with terminal restriction
fragment length polymorphisms (T-RFLPs) used to estimate variation, and 2) fungal cultures
isolated from root tips to which PCR-RFLP techniques were applied to assess variation.
T-RFLPs were used to examine the root-associated fungal community on Salix
arctica. Differences between the communities were analyzed using canonical
correspondence analysis (CCA). Genotype diversity was tested using a 2-way, 2-stage,
nested ANOVA. Warming did not significantly change genotype cumulative frequency or
diversity o f the root-associated fungal community, but cumulative frequency tended to
increase on the warmed plots. Genotype richness was significantly different according to
site, which was correlated with differences in soil chemistry.
Again site, not warming, was the main factor that distinguished the root-associated
fungal community of Salix arctica, Saxifraga oppositifolia, Cassiope tetragona, and Dryas
integrifolia based on fungal cultures. Warming did not have a detectable impact on
cumulative frequency and diversity, based on CCA and a nested, 3-way ANOVA. Fungal
cultures were identified based on sequence analysis and morphology. Phialocephala fortinii
ii

was the most frequently identified taxon, but almost half o f the fungal isolates remained
unknown.
The root-associated fungal community was examined along a glacier forefront
characterized by a directional, non-replacement primary plant succession pattern. CCA was
used to examine genotype frequency; linear regressions were used to test for changes of
cumulative frequency and diversity as succession advanced. The fungal community on only
one o f the host plants increased in frequency and richness as succession advanced. The darkand hyaline-septate endophyte communities were distinct on different host plants, providing
evidence for host specificity and higher diversity than previously reported.

Ill

Table o f Contents
Abstract ...............................................................................................................................................ii
Table o f Contents......................................................................................................................... iv
List of Tables............................................................................................................................. viii
List of Figures.............................................................................................................................. xi
Acknowledgements................................................................................................................... xiii
1.

Introduction................................................................................................................................. 14
A.

Rationale........................................................................................................................... 14

B.

Literature review .............................................................................................................. 15

1.

Concepts in community ecology....................................................................................15
a)

2.

Definition o f community................................................................................................. 15
Biodiversity...................................................................................................................... 18

a)

Definitions......................................................................................................................... 18

b)

Assessing biodiversity....................................................................................................21

c)

Rank/abundance curves..................................................................................................24

3.

Methods to assess root-fungal com m unities................................................................28

4.

Concepts in mycorrhizal fungal community ecology..................................................32
a)

Mycorrhizal fungal community structure o f arctic and alpine System s.................. 36

b)

Mycorrhizal fungal succession in relation to plant community succession

39

5.

Global climate change: effects on aboveground plant community structure

41

6.

Impact o f global climate change on root fungal communities.................................. 48

C.

Research objectives and hypotheses............................................................................. 49

D.

Thesis organization......................................................................................................... 52
iv

IL Impact o f warming on the frequency and diversity o f the root-associated fungal
community on Salix arctica from the Canadian High A rctic............................................... 53
A.

Introduction.........................................................................................................................53

B.

Materials and m ethods...................................................................................................... 54

1.

Study S ite ............................................................................................................................54

2.

V egetation.......................................................................................................................... 55

3.

Field collection...................................................................................................................56

4.

Sampling from roots.......................................................................................................... 57

5.

DNA extraction and ITS-T-RFLP analysis.................................................................... 58

6.

Matching root tips with sporocarps................................................................................. 59

7.

Data analysis...................................................................................................................... 60

C.

Results................................................................................................................................. 64

1.

Soil properties.................................................................................................................... 64

2.

Genotype frequency.......................................................................................................... 65

3.

D iversity..............................................................................................................................66

4.

Community composition.................................................................................................. 70

D.

Discussion...........................................................................................................................71

III. How

host plant, warming, and site affect the culturable root-associated fungal community

from the Canadian High Arctic..................................................................................................82
A.

Introduction.........................................................................................................................82

B.

Materials and m ethods...................................................................................................... 83

1.

Field collection...................................................................................................................83

2.

Sampling from roots.......................................................................................................... 84

3.

DNA extraction and ITS-RFLP analysis........................................................................ 84

4.

Data analysis...................................................................................................................... 85

5.

DNA sequencing............................................................................................................... 87

C.

Results................................................................................................................................. 91

1.

IT S-R FL Ps......................................................................................................................... 91

2.

DNA sequencing............................................................................................................... 96

3.

Culture m orphology..........................................................................................................97

4.

Identifieations based on moleeular techniques and m orphology.............................. 102

D.

Discussion.........................................................................................................................105

IV. The root-associated fungal community along a directional, non-replacement succession
chronosequence........................................................................................................................ 112
A.

Introduetion.......................................................................................................................112

B.

Materials and m ethods.................................................................................................... 113

1.

Study site ...........................................................................................................................113

2.

Field eollection.................................................................................................................114

3.

Sampling from roots........................................................................................................ 114

4.

Root m icroscopy............................................................................................................. 115

5.

DNA Extraction and ITS-T-RFLP analysis.................................................................115

6.

Matching root tips with sporocarps...............................................................................117

7.

Statistical analysis........................................................................................................... 117

C.

Results............................................................................................................................... 119

D.

Discussion.........................................................................................................................126

V. Synthesis o f results...................................................................................................................133

VI

VI. Conclusion............................................................................................................................... 138
Vil.

Literature cited................................................................................................................ 144

V lll.

Appendix for chapter 2 ......................................................................................................162

VIII.

Appendix for chapter 3 .................................................................................................. 182

IX. Appendix for chapter 4 ............................................................................................................ 199
X. Appendix for chapter 5............................................................................................................. 214

Vll

List o f Tables
Table 1.1 Summary o f experiments o f warming on arctic plants used for present study........ 47
Table 2.1 2-way, 2-stage nested ANOVA for the effects o f site, treatment, treatment
replicate, and plant specimen replicate on genotype frequency...........................................67
Table 2.2 Summary o f Tokeshi’s model used to describe the rank abundance curves at each
plot. 1.............................................................................................................................................70
Table 3.1 Number o f successful samples isolated from root tips based on RFLPs.................. 92
Table 3.2 Species richness o f cultures according to host plant, site, and treatment based on
R FLPs.......................................................................................................................................... 95
Table 3.3 Species richness o f cultures according to host plant, site, and treatment based on
identification made morphologically........................................................................................95
Table 4.1. Number o f successful extractions according to the number o f attempts at
extraction and total number of genotypes for each host plant.............................................121
Table A2.1 2-way, nested ANOVA table for effects o f site, treatment, and treatment replicate
on soil properties...................................................................................................................... 159
Table A2.2 2-way ANOVA table for the effects o f site, treatment, and treatment replicate on
genotype richness.................................................................................................................... 162
Table A2.3 2-way ANOVA table for the effects o f site, treatment, and treatment replicate on
genotype evenness.................................................................................................................... 163
Table A2.4 Relative abundance of genotypes from control to OTC for each primer-enzyme
combination................................................................................................................................164
Table A3.1 Comparison o f identification (based on molecular techniques) with
corresponding morphological identification - Phialocephala.............................................182

V lll

Table A3.2 Comparison o f identification (based on molecular techniques) with
corresponding morphological identification - Helotiales.................................................. 184
Table A3.3 Comparison o f identification (based on molecular techniques) with
corresponding morphological identification - unknown.................................................... 186
Table A3.4 Comparison o f identification (based on molecular techniques) with
corresponding morphological identification - Hymenoscyphus......................................... 193
Table A3.5 Comparison o f identification (based on molecular techniques) with
corresponding morphological identification -

194

Table A3.6 Comparison o f identification (based on molecular techniques) with
corresponding morphological identification - Dothideales................................................ 195
Table A3.7 Comparison o f identification (based on molecular techniques) with
corresponding morphological identification - Phoma..........................................................195
Table A3.8 Comparison o f identification (based on molecular techniques) with
corresponding morphological identification - Cryptosporiopsis........................................ 196
Table A3.9 Comparison o f identification (based on molecular techniques) with
corresponding morphological identification - Cadophora.................................................. 197
Table A3.10 Comparison o f identification (based on molecular techniques) with
corresponding morphological identification - Colispora.....................................................197
Table A 3.11 Comparison o f identification (based on molecular techniques) with
corresponding morphological identification - Agaricales................................................... 197
Table A3.12 Comparison o f identification (based on molecular techniques) with
corresponding morphological identification -Ceratobasidium........................................... 198
Table A4.1 Genotype richness for host plants at chronosequence zone...................................199

IX

Table A4.2 Sorensen’s Quantitative Index between host plants and chronosequenee zones200
Table A5.1 Comparison o f LS means o f soil chemistry between succession study (control
plot only) and warming studies (T-RFLP-OTC and culture studies)................................217

X

List o f Figures
Fig. 1.1 Example o f how T-RFLP is obtained in comparison to RFLP..................................... 33
Fig. 2.1. CCA on warming treatments and site with soil properties as biplots....................... 67
Fig. 3.1 Species cumulative frequency for Salix arctica and Saxifraga oppositifolia on
different sites............................................................................................................................... 92
Fig. 3.2 Species richness for Salix arctica and Saxifraga oppositifolia on different sites. .. 93
Fig. 3.3 NMS for host plant and warming treatments per site based on phi distances............ 94
Fig. 3.4 Neighbor-joining best tree

based on ITS and the Dothideales.................................. 96

Fig. 3.5 Neighbor-joining best tree based on ITS and Helotiales........................................... 98
Fig. 3.6 Neighbor-joining best tree based on LSU and the Agaricales.................................. 99
Fig. 3.7 Neighbor-joining best tree based on LSU and the Ceratobasidiomycetes..............99
Fig. 3.8 Neighbor-joining tree based on LSU and ascomycetes.............................................. 100
Fig 3.9 Abundance o f Phialocephala fortinii based on number of morphologically identified
cultures differentiated by host plant and treatment...............................................................101
Fig 4.1 Canonical Correspondence Analysis for effects o f host plant and chronosequence
zone............................................................................................................................................. 122
Fig 4.2 Cluster analysis using Ward’s method based on Euclidean distances.......................... 123
Fig. A2.1 NMS for site and warming treatment effect..............................................................165
Fig. A2.2 Cluster analysis o f plots, treatments and plant specimen based on frequency of
genotypes....................................................................................................................................169
Fig. A2.3 Least square means for soil properties.........................................................................171
Fig. A2.4 Least square means for genotype frequency for each treatment per plot................173
Fig. A2.5 Rank abundance curves for lowland sites................................................................... 174

XI

Figure A2.6 Rank abundance curve for highland dolomitic site ................

176

Fig. A2.7 Rank abundance curves for highland granitic site..................................................... 178
Fig. A2.8. Least square means o f genotype richness for each treatment per site...................180
Fig. A2.9. Least square means o f genotype evenness measured by Hurlbert’s PIE index.. 181
Fig. A4.1 NMS of host plant and chronosequence zones..........................................................206
Fig A4.2 LS means o f soil property for each chronosequence zone.........................................207
Fig. A4.3 Simple regression for genotype richness for Luzula confusa...................................209
Fig. A4.4 Simple regression o f genotype cumulative frequency of Luzula confusa.............. 210
Fig. A4.5 Rank abundance curve for each chronosequence zone for Luzula confusa

211

Fig. A4.6 Rank abundance curve for each chronosequence zone for Papaver lapponicum. 212
Fig. A4.7 Rank abundance curve for each chronosequence zone for Salix arctica............... 213
Fig. A4.8 Rank abundance curve for each chronosequence zone for Saxifraga oppositifolia
............................................................................................................................................................214
Fig. A4.9 Rank abundance curve for each chronosequence zone for Cassiope tetragona... 215
Fig. A4.10 Rank abundance curve for each chronosequence zone for Dryas integrifolia. ..216

Xll

Acknowledgements
I would like to thank my advisor, Keith Egger, for all his support, guidance, and
patience. 1 would like to thank my supervisory committee members, Randy Currah, Art
Fredeen, Kathy Lewis, and Hugues Massicotte for all their support and help. Greg Henry
was generous with his field sites and field consultation, and Cecilia Alstrom-Rapaport was
invaluable for statistical guidance.
1 am indebted to Randy Currah who graciously accepted my cultures and to Melissa
Day for the laborious work o f identifying all the cultures morphologically.
1 would also like to thank Clive Owens o f the Ministry o f Forests Analytical Research
Laboratory (Victoria, BC) for processing my soil samples, and Lito Arocena and Yuriko
Yano for their insightful soil data interpretation.
1 am grateful to Brent Murray for all his advice in the lab and the following people for
their laboratory assistance (without their help I would still be writing my thesis!):
Tamara Bereck - exhaustive isolations from root tips
Jennifer Catherall - started extractions and RFLPs for cultures
Susan Gibson - started sequencing o f cultures, RFLPs
Sara Jamieson - preparing roots for extractions
Edward Kraay - database creation and technical support
Linda Rehaume - help finishing sequences from cultures and helping in lab
Rachel Botting, Ld Kraay, Susan & Garth Frizzell, Susan Gibson, Nabla Kennedy,
Brian Milakovic, Tim Phaneuf, Jennifer Psyllakis, Linda Rankin, Susan Robertson, Laura
Ryser, Linda Tackaberry, Andrew Walker, and members o f the Lgger/Massicotte/Murray lab
were wonderful for their insights and camaraderie.
I would finally like to thank my parents, Bob and Shigeko Fujimura, for all their
support, encouragement, and guidance.

X lll

I.

Introduction

A. Rationale
The root-associated fungal community includes mycorrhizal fungi, dark and hyaline
septate fungi, and possibly parasitic or pathogenic fungi. O f these, the mycorrhizal and darkand hyaline-septate fungi are the most abundant members o f this community. Mycorrhizal
fungi are known to have mutualistic relationships with vascular plants, and are important
components o f most ecosystems. Dark and hyaline septate fungi have been reported to be
both pathogenic and mutualistic. Their role as mutualists is hypothesized to be greater in
arctic ecosystems with the absence o f arbuscular mycorrhizae (Bledsoe et al. 1990).
Global warming is an important source o f disturbance of arctic ecosystems. The
primary effect o f global warming is the increase in mean temperature. Global warming is
also significant because o f its secondary effects; it has been linked to the increase o f sea
levels, hurricane occurrence, fire and insect outbreak in coniferous forests, and species
extinction. These problems are compounded by the release o f greenhouse gases.
Arctic environments are opportune ecosystems to examine ecological questions about
the root-associated fungal community. These environments have low plant species diversity,
which simplifies the complexity o f examining the fungal community found on roots. In
addition, arctic ecosystems are expected to be impacted more severely by global warming,
and in advance o f other ecosystems. By simulating warming in arctic environments using
open-top chambers, insights can be gained into how warming may impact the root-associated
fungal community in other environments.
Warming indirectly affects arctic ecosystems by causing glaciers to recede. This
provides an opportunity to examine changes in the belowground community in response to a
14

unique form of primary plant succession that occurs in the high arctic, directional nonrcplaccmcnt succession, where host plants arc not replaced as succession proceeds so
diversity increases along the chronosequence.
Alexandra Fiord, Nunavut provides an excellent opportunity to study the effects of
climate change on the root-associated fungal community. The biology and autecology of
plants in this area have been studied extensively, and passive warming experiments are part
o f the International Tundra Experiment (ITEX), which was created to monitor the effects of
warming in arctic regions. Alexandra Fiord also hosts a unique type o f primary plant
succession, which provides a natural experimental design for studying how the rootassociated fungal community on a common suite o f host plants responds to an increase in
plant diversity along a chronosequence.

B. Literature review

1. Concepts in community ecology
a)

Definition of community

There has been considerable debate among ecologists regarding the conceptual
definition o f “community” . Wilson (1991), for example, questions whether plant
communities are really integrated, discrete entities. He argues that plant communities do not
exist unless the definition is delimited by a list o f criteria including assembly rules, niche
limitation, discreteness, discontinuity, and integratedness (Wilson 1991, Palmer and White
1994).
Looijen and van Andel (1999) asserted that the problem with the definition of
community is that the term is too ambiguous: 1) it can be applied to different levels o f taxa.

15

2) no objective boundaries can be made, and 3) communities are heterogeneous with respect
to species composition. They suggest the following definition: “community may be defined
as a set o f individuals o f two or more species that occur in the intersection o f areas occupied
by populations o f these species” (Looijen and van Andel 1999). This limits the definition of
communities to be used only for coexisting species belonging to a single taxonomic group,
such as phyla or class, that has a static boundary (Looijen and van Andel 1999).
Parker (2001) refutes Looijen and van Andel’s definition because o f scale limitations
and unidentified assumptions. He argues that the scale limitation leads to ambiguity or
conflict with respect to what organisms belong to a given community (Parker 2001). He
identifies three assumptions from Looijen and van Andel’s model which are often violated in
community ecological studies: 1) there must be a unique underlying process, 2) there must be
consistency o f processes among replicates and, 3) there must be independence from other
communities. Instead, Parker’s definition combines concepts from Brand and Parker (1995)
and Pickett et al. (1992): ‘communities are continuous in time and space, and processes
underlie composition and dynamics’. His definition includes a conceptual model where the
community focuses on a single individual and its interactions with other members o f the
community (i.e. consumers, symbionts, pathogens, mutualists, and competitors); each
individual o f the community has its own set o f interactions (Parker 2001).
Some argue to forgo the conceptual definition (McCune and Grace 2002, Palmer and
White 1994) and use an operational definition (Palmer and White 1994), such as a
“collection o f organisms found at a specific place and time” (McCune and Grace 2002). This
operational definition is similar to Parker’s definition (2001) but does not include his
conceptual model, which implies that interactions are necessary in a community. The
16

operational model assumes that variation in species composition is random spatially and
temporally (McCune and Grace 2002). With the operational definition, conceptual or
theoretical implications are circumvented (McCune and Grace 2002, Palmer and White
1994).
To avoid the ambiguous use o f the term community, Fauth et al. (1996) proposed a
more restricted set o f definitions. ‘Taxa’ is reserved for phylogenetically-related species
regardless o f where they occur; ‘communities’ include all species co-oecurring in one place
(i.e. corresponds to the operational definition Palmer and White (1994)); and ‘guilds’ are sets
o f species exploiting the same resources (Fauth et al. 1996). Other concepts are used to
describe overlapping combinations o f these three terms. ‘Assemblages’ are
phylogenetically- restricted groups that occur in a community (i.e. overlap between taxa and
communities). When guilds and taxa overlap, then a group o f related species exploit the
same resource. ‘Local guilds’ are formed from the intersection of communities and guilds
and are groups o f species that share a common resource and occur in the same community
(Fauth et al. 1996). When all three overlap, then the term ‘ensemble’ is used; this is a
taxonomically restricted group o f species that exploit the same resources, and are located in
one place (Fauth et al. 1996).
The term that perhaps best applies to this study is the ‘ensemble’ according to Fauth
(1996), because this study incorporates phylogenetic, community and guild perspectives.
However, operationally, I will use community in the sense o f Palmer and White (1994),
which corresponds to Fauth’s definition.

17

2. Biodiversity
a)

Definitions

Biodiversity is a loosely applied term, and has been used to describe diversity from
the genetic level to the biome level (Hooper et al. 2005). Biodiversity can be described as
the number o f different genotypes, species, ecosystem types, etc. and includes the evenness
o f tbeir distribution (Hooper et al. 2005). Species richness is only part o f this definition and
refers to the number o f taxonomic units (usually species or genotypes).
Ecosystem function is the effect o f the activities o f organisms on the physical and
chemical processes o f an environment. A functioning ecosystem is characterized by these
processes (Naeem et al. 1999). According to Naeem et al. (1999), ecosystem functioning can
be measured by quantifying rates o f movement, such as nutrient transportation, or by
measuring growth or production, such as plant stem growth or seed production.
Disturbance used to be limited to events that were massively destructive and rare
(Rykiel 1985). This definition is no longer acceptable because disturbances are not always
catastrophic and can be recurring events in ecosystem. Rykiel (1985) attempted to formulate
a general definition by defining disturbance as a physical force, agent, or process that causes
a perturbation in an ecological component or system. Disturbance can be abiotic or biotic
and: 1) can cause destruction, where biomass is quantitatively reduced; 2) can cause
discomposition, where certain populations are eliminated, reduced, added, or expanded; 3)
can cause interference, where matter, energy and/or processes are hindered; or 4) can be
caused by suppression, where natural disturbances are prevented (Rykiel 1985). The
outcome of a disturbance is perturbation, which is a deviation of values that are used to
describe the properties o f the ecological component or system (Rykiel 1985). Rykiel’s

18

definition assumes that reference conditions must be known in order to understand
disturbance (Pickett et al. 1989).
White and Jentsch (2001) argued that one general definition is unachievable for
disturbance. They argued that disturbance should incorporate four different areas: 1)
variation in disturbance events, which would include the timing and intensity o f the
disturbance; 2) variation in the disturbance effects within an ecosystem, which would cover
spatial and temporal variation; 3) variation in ecosystem response by including differences in
biota and physical environments; this would include the rates of response and species
adaptation would be included; and 4) inferences o f scale o f observations and measurements,
which are affected by observations, sampling, and analysis by the researcher. These four
topics that they argue should be included are covered in the definition proposed by Pickett et
al. (1989).
The definition by Pickett et al (1989) consists o f identifying the object disturbed,
distinguishing what is and is not disturbed, and recognizing the minimal level o f hierarchical
organization. The following levels are their recommended hierarchical organization:
individual, population, community, ecosystem, and landscape. Disturbance, which is defined
as an external force o f a given level, would affect the structure, function, and attributes of
that level. For example, at the community level, structural disturbance would include effects
on vertical and horizontal patterns, species composition, or functional groups; functional
disturbance would include effects on resource levels, competition, or mutualistic interactions;
and attribute disturbance would include effects on coexistence, evenness, or dominance. At
the ecosystem level, structural disturbance would include effects on functional groups,
functional disturbance would include effects on fluxes in the ecosystem, and attribute

19

disturbance would include effects on resistance and resilience of the ecosystem. In this
hierarchical organization, temporal and spatial scales are also incorporated and are defined
within the context o f each level. Temporal and spatial effects that occur on a broader scale
would be called ‘disturbance regime’ such as the fire regime in grasslands, where the
reoecurrenee o f the disturbance is every few years.
Stability refers to how an ecosystem or community responds to disturbance. For
ecosystems, stability would apply to populations or communities and their abiotic
environment, such as analysis o f nutrient dynamics (Barbour et al. 1999). For communities,
stability is measured by determining how the community composition and diversity responds
to disturbance (Barbour et al. 1999). Stability has two components, resilience and resistance,
and the overall response to disturbance is determined by the interaction between the two
(Barbour et al. 1999). On the ecosystem level, resilience is the ‘ability o f an ecosystem to
return to predisturbance conditions’ (Barbour et al. 1999), which may take a long time.
Resistance is the ‘ability o f an ecosystem to resist changes in response to disturbance’
(Barbour et al. 1999). These terms can also have community definitions. Community
resilience is when the community returns to the same species composition after a disturbance,
and community resistance is where species composition does not change due to disturbance
(Tokeshi 1999). However, there are problems with the community resistance/resilience
concepts. One problem is determining how much change in the community composition is
needed before the community is considered ‘disturbed’. Another problem is determining
whether it is necessary for the community to return to its ‘original’ composition for stability
to return (Tokeshi 1999). Also, community composition changes over time without
disturbances (Tokeshi 1999), so these terms may be difficult to apply.

20

Although ecosystems may become unstable, they can still continue to function. How
well an ecosystem responds to disturbance depends on its resilience. Processes may be
retarded, but if the ecosystem is resilient, then these processes can return to pre-disturbance
conditions.

b)

Assessing biodiversity

Assessing biodiversity has become an important issue because o f the increased rate of
loss of diversity due to anthropogenic activities. One argument used to support conservation
is that preservation o f biodiversity will maintain ecosystem functioning. Maintenance o f
biodiversity has become a surrogate for ecosystem function (Naeem 2002). For example,
Naeem et al. (1995) found that higher diversity correlated with an increase in community
respiration, productivity and nutrient retention in a mesocosm study, and therefore alteration
o f biodiversity can affect ecological processes.
Tilman et al. (1997a) examined how plant species diversity, functional diversity, and
functional composition affect plant productivity, %N in plants, total N in plants, soil NH4,
soil NO3, and light penetration. They found that functional diversity, but not plant diversity,
significantly impacts these functional variables by positively affecting plant productivity and
total N and negatively affecting soil NO3, soil NH4, plant %N, and light penetration. They
also found that many species in monocultures have comparatively less biomass than when
they are found in multifunctional group plots, which supports the hypothesis that higher
diversity increases ecosystem productivity. They concluded that: 1) functional composition
and diversity are significant determinants in grassland ecosystem processes and 2) not all
plant species are equal, so the loss o f one species may be more deleterious than another.

21

Higher species richness alone is not sufficient to explain the impact o f biodiversity on
ecosystem functioning. When examining competition in a resource model for plants, the
variance o f the model is explained more by species identification than species richness
(Tilman et al. 1997a). Likewise, Hooper and Vitousek (1997) found that the identification o f
the functional groups explains more variance than species richness, and that species and
combinations o f species, rather than species richness, control yields and nutrients.
Another issue eoneeming how biodiversity affects ecosystem functioning is
complementary effects versus selection effects (Hooper and Vitousek 1997, Cardinale et al.
2002). Complementary effect theory attempts to explain how resource use by organisms
affects ecosystem processes (Cardinale et al. 2002, Loreau and Hector 2001). According to
this theory, species diversity can increase while avoiding competition, and species are able to
co-exist, especially in environments with limiting resources, by: 1) partitioning resources,
where each species can use nutrients, water, or other resources differently instead o f all
species competing for or using the same resources; or 2) niche differentiation, where
different species avoid using the same resources as other organisms in time and/or space
(McKane et al. 2002). Loreau et al. (2001) found that feeding performance by caddisfly
larvae improves in the presence o f other taxa. They concluded that the increase in species
diversity o f other aquatic arthropods leads to interspecific facilitation.
In contrast, selection effect theory is applied when species diversity is correlated with
the probability that a dominant species uses most o f the available resources (Cardinale et al.
2002), so the formation o f the community is heavily dependent on these dominant species.
In a study where a mathematical model was used to test for complementary effects and
selection effects on monocultures and mixed species o f grasses, results explained by the
22

selection effect theory were not as reliable as by the complementary effect theory (Loreau et
al. 2001). In complementary effect theory, the performance o f communities can go beyond
the additive effects o f individual species (Loreau et al. 2001).
Schwartz et al. (2000) criticized studies that link greater biodiversity with an increase
o f ecosystem productivity. In both observational and experimental studies, conflicting
results have been reported, with some studies having negative or no results and others that
are variable through time and space (Schwartz et al. 2000). In addition, other problems with
experimental studies include; 1) hidden manipulations such as weeding, which would change
the diversity and composition o f the experiment; 2) addition o f species to poor environments,
which probably would not occur in nature; and 3) extrapolation o f results to the whole
ecosystem when only one trophic level has been included (Schwartz et al. 2000). With
theoretical models, the role o f rare species may be missed in stabilizing ecosystems. Also,
models may assess stability and function on the wrong scale; models may apply only at a
local rather than an ecosystem scale.
Schwartz et al. (2000) concluded that where the relationship between biodiversity and
ecosystem function are positive, the relationship is not linear as studies suggest, but that the
function saturates after a few species or functional groups, creating more o f an asymptotic
relationship. For the 23 observational and experimental studies they examined where they
could graph biodiversity against ecosystem function, they found that 60% o f these studies
produced the asymptotic graph. Although the relationship between biodiversity and
ecosystem function is an important question, doubts about this relationship are prevalent
because of conflicting results in finding this asymptotic relationship, and also because the
number o f species or functional groups to fulfill the functions of an ecosystem is not known.

23

However, as Loreau et al. (2001) suggested, biodiversity may not be as important in the
maintenance o f an ecosystem as it is in helping to facilitate changes in the environment.
Bengtsson (1998) argued that biodiversity is not mechanically linked to ecosystem
functions. Simple measures o f species richness assume that all species are equal in their
function. This is unlikely to be true; therefore measuring diversity is pointless unless the
function of species is known. He contends that knowing the species and their functions will
explain the processes and stability o f an ecosystem. However knowing the functions o f all
species is currently impossible. Even though all the functions of species are not known,
linking biodiversity to ecosystem functioning remains important because more diverse
ecosystems may include redundant species that could fulfill ecosystem functions when
dominant species are lost. This would increase the probability o f withstanding or rebounding
from disturbances.

c)

Rank/abundance curves

The shape o f rank/abundance plots (a.k.a. dominance/diversity curve or Whittaker
plots) is used to determine which species abundance model best describes the data (Magurran
2004). These rank/abundance plots are helpful in that: 1) different patterns o f species
richness are easily shown, 2) the relative abundance o f species-poor communities is easily
seen, and 3) emphasis is placed on the differences in evenness for contrasting communities
(Magurran 2004).
Species abundance models can be categorized into two main groups - biological and
statistical (Tokeshi 1999). The biological models are also called niche apportionment models
and include the following; geometric series, broken stick, MacArthur fraction, dominance
pre-emption, random fraction, dominance decay, random assortment, composite, and power
24

fraction (Tokeshi 1999). These biological models are based on the assumption that species
divide the niche space among species that live in a community in different ways (Magurran
2004). These models have been criticized for possibly being too simplistic and confusing in
terms o f bow the niches are apportioned, but they can be valuable tools for understanding
niche differentiation (Magurran 2004).
Depending on which model fits the shape o f the rank/abundance plot, interpretations
can be made about bow the niche space is divided in the community. Rank/abundance plots
that fit a geometric series model often describe communities that are species-poor, such as
those found in harsh environments or in early stages o f succession (Magurran 2004). Those
communities fitting the Mac Arthur's broken stick model (or random niche boundary
hypothesis) are interpreted as having their species competing equally for one resource.
However, Mac Arthur’s broken stick model assumes that the niche space is partitioned
simultaneously, which probably does not happen in nature (Magurran 2004).
Tokeshi developed a set o f niche apportionment models that forgo the assumption of
simultaneous niche partitioning o f the broken stick model (Magurran 2004). Two o f his
models examine extreme cases when the least or most abundant species are invaded by new
species (Tokeshi 1999). The dominance pre-emption model is where new species invade
niche spaces occupied by the least abundant species in an existing community (niche
fragmentation), or alternatively where a new species takes approximately half o f a new niche
space (niche filling) (Tokeshi 1999). In these cases, the dominant species remain so. The
dominance decay model is a model o f the other extreme where the largest niche space,
instead of the smallest, is appropriated by new species (Tokeshi 1999).

25

Tokeshi has three models that examine how niche apportionment occurs when new
species invade all potential niches and not just the ones occupied by the least and most
abundant species. The Mac Arthur fraction model is similar to the broken stick model, but it
assumes that the niche spaces are invaded sequentially rather than simultaneously; however,
the same conclusions can be drawn from both models (Tokeshi 1999). In this model, the
probability o f a community being invaded depends on species abundance or niche size, so
niche space o f more abundant species will likely be invaded before less abundant species.
This model implies a uniform distribution and may be applicable only to small communities
with related species (Magurran 2004).
In the random fraction model, all species have the same probability o f being invaded
by a new species, so the abundance o f species does not influence the chances o f being
selected (Tokeshi 1999). This model fits situations where new species compete for niche
spaces randomly over an existing niche that is already occupied by an assemblage o f species
(Magurran 2004). Magurran (2004) finds this model to be innovative with a wide range o f
applications. The power fraction model is similar to the random fraction model, but it is used
for species rich assemblages because most o f the niche apportionment models are applicable
to communities with small species assemblages (Magurran 2004).
The random assortment model assumes that there is no relationship, or a weak one,
between niehe apportionment and species abundance (Magurran 2004), so the abundances of
species are independent o f each other (Tokeshi 1999). This may be used for situations where
communities are in a state a flux from major environmental changes and competition is not
limited by species abundance (Magurran 2004).

26

The composite model is achieved by taking two or more o f these niche apportionment
models into account to describe how a niche is divided. Tokeshi realized that using only one
model may be too simplistic for a community; however, knowing where to set the boundary
between more and less abundant species may be problematic (Magurran 2004).
Statistical models were initially created so researchers could objectively compare
species abundance between communities (Magurran 2004). Even though some o f these
models have been labeled as statistical, ecological implications have been drawn from these
statistical models. For example, the log normal model is a statistical model, but the
ecological implication is that it explains situations where new species come to a niche in a
random order rather than in fixed intervals such as in the geometric series model (Magurran
2004). The log normal model has been found to fit many datasets and is commonly used
(Magurran 1988). However this model has been criticized because it requires a large number
o f species so the log normal distribution may be a mathematical artifact o f a large sample
size and so may have few biological implications (Magurran 1988, Tokeshi 1999).
These models are useful in assessing plant communities based on species abundance
distributions. Arctic ecosystems are considered harsh environments and, as expected, plant
communities in the arctic fit the geometric series model (McKane et al. 2002). These species
abundance models can be used to assess the root-associated fungal community and test if
these communities fit models similar to their plant counterparts. These models will also be
helpful in assessing how the root-associated fungal community responds to the direct, non­
replacement succession (see below) found on Alexandra Fiord.

27

3. Methods to assess root-fungal communities
The techniques used in this study to examine the root-associated fungal communities
included morphotyping, PCR-RPLP, T-RFLP, and DNA sequencing. The term ‘rootassociated fungal communities’ will be used because some o f the techniques do not
differentiate mycorrhizal, endophytic, and pathogenic fungi. Because the techniques used are
commonly applied for examining mycorrhizal fungal communities, much o f the review will
be based on this group o f fungi.
Sporocarp collections and morphotyping have been commonly used to assess
mycorrhizal fungal communities. However, sporocarp production was found to be an
inaccurate estimate o f composition o f ectomycorrhizal fungi found on root tips below ground
(Gardes and Bruns 1996a, Dahlberg et al. 1997, Jonsson et al. 1999), and corresponds to only
approximately 20% (Jonsson et al. 1999) to 30% (Dahlberg et al. 1997) o f the belowground
ectomycorrhizal fungi. Often the most abundant sporocarps do not correspond with the most
abundant mycorrhizae (Gardes and Bruns 1996a). Ectomycorrhizal fungal species may
rarely, or never, fruit, or may produce small or hypogeous fruiting bodies that are missed in
surveys, which could potentially lead to inaccurate estimation of mycorrhizal fungal species
found on root tips.
Ectomycorrhizal communities described by morphotyping is limiting in that at least
half o f the species are completely unknown (Gardes and Bruns 1996a), especially if the
morphotype comes from the field (Karen et al. 1997). Morphotyping also requires a high
level o f skill (Kârén et al. 1997), and so often takes more time to learn than molecular based
techniques (Dahlberg 2001). Better results in identification through morphotyping require
more phenotypic characteristics, but then fewer samples can be examined (Horton and Bruns
2001). The efficiency of distinguishing mycorrhizal taxa improves with RPLP analysis over
28

morphotyping, circumventing phenotypic plasticity, where multiple species may be grouped
as the same morphotype (Horton and Bruns 2001). Jonsson et al. (1999) distinguished 20
morphotypes from 7152 mycorrhizae but found 42 RFLP-types from 212 root tips that
successfully amplified.
PCR-RFLP has been a useful tool in researching mycorrhizal fungal communities.
The internally transcribed spacer (ITS) region o f the nuclear-encoded ribosomal RNA
(nrDNA) gene repeat is often used because; 1) it is readily amplified with fungal-specific
primers, allowing amplification o f fungal DNA from mixed genomes, such as plant and
fungal DNA in ectomycorrhizal root tips (Gardes and Bruns 1996a) and 2) it is divergent
enough for identifying species within a genus (White et al. 1990). The nrDNA is often used
for fungal studies because DNA sequences are highly conserved among organisms,
variability is high between species and minimal within a species (Egger 1995), and the
ribosomal repeat is a multi-copy gene (Gardes and Bruns 1993), making it easier to amplify.
Studies involving PCR-RFLP o f the ITS region have focused on comparing above
and belowground fungal species composition (Gardes and Bruns 1996a, Dahlberg et al. 1997,
Jonsson et al. 1999); describing differences in composition due to treatment or changing
environments (Horton et al. 1999, Erland et al. 1999, Kemaghan 2001); or describing
changes in composition due to succession (Nara et al. 2003). In addition to the confirmed
lack o f correlation between above and belowground fungal composition, these studies find
that a few widespread species generally account for most o f the abundance o f mycorrhizal
fungi (Gardes and Bruns 1996a, Dahlberg et al. 1997, Jonsson et al. 1999, Horton et al. 1999,
Erland et al. 1999, Kemaghan 2001) and that spatial variation is a large determinant for
species composition, at least for Swedish forests after a low intensity fire (Dahlberg 2001).

29

Terminal Restriction Fragment Length Polymorphism (T-RFLP) analysis is a
relatively rapid and accurate, PCR-based tool to identify taxa (Avaniss-Aghajani et al. 1996,
Martfnez-Murcia et al. 1995). It has been used for examining microbial communities found
in sludge (Marsh et al. 1998, Liu et al. 1997), termite guts (Liu et al. 1997), aquifer sand from
groundwater (Liu et al. 1997), and for determining the effects o f temperature on the
microbial community in rice fields (Chin et al. 1999). A few more ectomycorrhizal fungal
community studies have been based on T-RFLP analysis, such as analyzing how the increase
of CO 2 would change the mycorrhizal fungal community (Klamer et al. 2002) and
determining the soil vertical distribution o f ectomycorrhizal fungi (Dickie and Koide 2002).
T-RFLPs are helpful when organisms have indistinct morphologies (Avaniss-Aghajani et al.
1994), and this technique avoids creating time-consuming cloning o f organisms, which may
not work for all species (Bruce 1997).
T-RFLPs are similar to restriction length polymorphisms (Clement et al. 1998) in that
they are both used to characterize the ITS region o f the nrDNA for differentiating taxa in
describing community diversity. However, the protocol for T-RFLPs differs from RFLPs in
that each primer is fluorescently labeled with a different dye, and fragments are separated on
a 6 % polyacrylamide gel rather than an agarose gel (see Fig. 1). The fluorescence allows the
samples to be detected by automated DNA sequencers/fragment analyzers and because o f the
polyacrylamide gel has higher resolution (Avaniss-Aghajani et al. 1996), allowing detection
o f fragments that are only 1 or 2 base pairs different (Totsch et al. 1995). Once the ITS
region is amplified with the fluorescent dye-labeled primers, restriction endonuclease
enzymes are used to digest the region, as with the RFLP methodology. Instead o f visually
detecting multiple bands as with RFLP analysis, only one fragment, the terminal fragment, is

30

detected because the laser only detects the digested product with the fluorescent dye (Fig. 1).
If both primers are labeled, two sets o f results are generated: one from the forward primer
and the other from the reverse primer. Multiple bands in PCR products, which indicate the
presence o f more than one organism, are problematic in RFLP analysis because assigning
fragments to their respective fungus is difficult or impossible. In RFLPs, individual taxa are
indicated by a unique pattern o f multiple bands, but when multiple taxa o f fungi are on the
root, fragment patterns become too complex for analysis. T-RFLPs circumvent this problem
by detecting only terminal fragments, so in theory, each fragment should represent a unique
taxon. Several restriction enzymes may be needed to differentiate taxa that share restriction
sites, but multiple restriction enzymes are used for RFLP analyses as well.
Clement et al. (1998) list potential problems with T-RFLPs: 1) PCR primers may
differentially amplify certain species, therefore, measurements of relative abundance in a
community may not be accurate; 2 ) unequal relative abundance may occur due to different
optimum annealing temperatures for different species; 3) evidence o f fragments may be
limited by electrophoresis technology; and 4) accurate community analysis needs multiple
digestive enzymes. The problems listed can also be applied to RFLP analysis as well, and
cannot be resolved without more advanced technologies that reduce the number o f samples
that are analyzed.
T-RFLPs is a valuable tool in mycorrhizal fungal community analyses. Like bacterial
systems, mycorrhizal fungi are often morphologieally indistinct. In addition, this method
circumvents the phenotypic plasticity o f the mycorrhizal fungi on different hosts. Unlike
RFLP analyses, this tool can detect and distinguish multiple mycorrhizal fungi on the same
root tip.

31

Although RFLP and T-RFLP analyses are powerful tools for assessing communities,
they are limited in identifying taxa. One way to identify taxa using RFLPs and T-RFLPs is
to compare fragments with a database that already exists. However, comparing RFLP
fragments with those in databases created by other researchers is often not feasible because
restriction enzymes and primers are not standardized. DNA sequencing is important in
filling this gap, especially when the goal is to identify unknown taxa. Some reasons why
sequencing is more successful are: 1) there is a central database (GenBank) where scientists
deposit their sequences, and 2 ) sequences are not restricted by choice o f endonucleases or
primers; as long as the unknown sequence has an overlapping segment in GenBank, then
identification to at least order is plausible. The caveat with comparing sequences from
GenBank is that submitted sequences are not checked for accuracy, so the identification of
the sequences may not be reliable (Bridge et al. 2003).

4. Concepts in mycorrhizal fungal community ecology
Mycorrhizal fungal community ecologists often have the triple task o f describing and
interpreting the fungal community structure as well as extrapolating their results to the plant
community. Some topics that are addressed in mycorrhizal fungal community ecology are
complementary effects (e.g. Koide 2000, Perry et al. 1989), community structure (e.g. Horton
and Bruns 1998, Dahlberg et al. 1997, Gardes and Bruns 1996a), and the role o f mycorrhizae
in plant communities’ resistance to change (e.g. Horton and Bruns 1998). Mycorrhizal
fungal community ecologists have approached the issue o f complementary effects by
examining facilitation and niche differentiation (e.g. Dickie and Koide 2002, Helm et al.
1999, Titus and del Moral 1998). Facilitation is an important concept in mycorrhizal
community ecology because o f the guild concept, where the diversity o f the mycorrhizal

32

Fluorescent-dyelabeled primer

Fluorescent-dyelabeled primer
V
18s small
subunit

1r

ITS
1

5.8s
region

ITS
2

I
Fluorescent
tag Cy 5.0

28s large subunit

PCR Reaction

Fluorescent
tag Cy 5.5

ITSl - 5.8s - 1TS2 regions

▼

1r

ITS region cut with
an endonuclease
enzyme

Terminal end that would be
fluorescently tagged, e.g. with
Cy 5.0, for T-RFLP analysis

Appear as on an RFLP gel

Terminal fragment shown on
automated sequencer

Fig. 1.1 Example o f how T-RFLP is obtained in comparison to RFLP

33

fungal community and the plant community can stabilize their plant-soil ecosystem after
disturbance or stress (Perry et al. 1989).
The benefits o f linkages between plants by mycorrhizal fungi, are generally expressed
by bow plant communities may profit rather than by bow the fungal communities may profit.
Possible benefits for plants include: 1) seedlings may benefit by linking into a larger fungal
network via fungal bypbae (Newman 1988); 2) interplant exchange o f nutrients (Newman
1988); 3) interspecific competition between plants may be altered if nutrients are received
from one central network (Newman 1988); 4) competition between plants may be reduced
(Newman 1988); 5) nutrients from dying plants may pass directly to living plants (Newman
1988); 6 ) stabilization o f succession patterns because some fungi eould associate with both
pre- and post-disturbance plant hosts (Horton and Bruns 1998); and 7) improved plant
survival (Trappe and Luoma 1992).
Fungal linkages between plants have been demonstrated both in the laboratory and in
the field. Phosphorus transfer between Pinus sylvestris and Finns conforta via Suillus
bovinus was found in vitro using ^^P, in which the labeled phosphorus did not move only to
the plants but throughout the whole fungal network (Finlay and Read 1986). Simard et al.
(1997) showed a net transfer o f carbon between Be tula papyrifera and Pseudotsuga menziesii
in the field. By shading P. menziesii seedlings and not B. papyrifera, they found that P.
menziesii seedlings are sinks for carbon, and that carbon could transfer from a sink to a
souree or along a nutrient gradient. They also found that earbon exebange occurs between
the two ectomycorrhizal plants {B. papyrifera and P. menziesii), but not with the arbuscular
mycorrhizal plant Thujaplicata. Horton et al. (1999) suggested that linkages between
ectomycorrhizal and arbutoid plants allow outplanted P. menziesii seedlings to survive in an

34

arbutoid stand o f Arctostaphylos. They conclude that linkages probably do not exist between
eeto- and arbuscular mycorrhizal because outplanted P. menziesii seedlings died in
arbuscular stands of Adenostoma.
Koide (2000) offers two strategies for the role o f complementarity in root
colonization by arbuscular mycorrhizae. One strategy is where the fungi are complementary
to each other. Although this would allow fungi to coexist on the same root, it does not
explain why some antagonism happens between fungi (Koide 2000). The other strategy he
proposes is that the function o f the fungus is complementary to those o f the plant, which
would lead to redundant species on roots and may explain why some root-fungus
relationships have a high level o f speeifieity. The latter is similar to the application o f
coexistence theory to ectomycorrhizae, where plants select more beneficial fungi by
lengthening or shortening time o f root tip mortality (Hoeksema and Kummel 2003). These
concepts may apply to ectomycorrhizal fungi as well, but competition (Wu et al. 1999),
different life strategies, such as colonization rate and life spans (Hoeksema and Kummel
2003), must also be considered.
Another form o f complementarity is niche differentiation. Although niche
differentiation is one o f the oldest explanations for biodiversity, it has not been tested in
many ectomycorrhizal fungal communities (Bruns 1995). Recently, niche differentiation was
used to explain the vertical distribution o f ectomycorrhizal fungi. Dickie and Koide (2002)
used cluster analysis and species diversity measures to differentiate six spatial patterns for
fungi. The outcome o f the distribution suggested that niche differentiation explained the
vertical distribution o f ectomycorrhizal fungi.

35

Only a few papers link mycorrhizal fungal diversity to ecosystem functioning. Van
der Heijden et al. (1998) showed that higher species richness o f arbuscular mycorrhizal fungi
increased plant biodiversity, improved plant productivity, lengthened hyphal growth in the
soil, increased phosphorus absorption o f plants, and decreased the amount o f phosphorus in
the soil. They concluded that there is probably a feedback loop where the plant benefits from
the increased amount of phosphorus, and the fungi prosper due to increased carbon, indicated
by more hyphal growth.
Baxter and Dighton (2001) examined if higher diversity o f ectomycorrhizal fungi
would affect plant growth and nutrient acquisition. They concluded that ectomycorrhizal
species diversity is more influential in plant biomass and nutrient uptake than the species
composition or rate o f colonization. Although a pioneering paper in linking mycorrhizal
diversity to ecosystem function, this study examined a community with low diversity o f
mycorrhizal fungi (Leake 2001). This may limit applications due to its simplicity o f being an
in vitro study. Another confounding factor included using peat and vermiculite for the
growth media, which adds excess nutrients. As a result o f these short-comings, it may be
premature to draw conclusions about the link between ectomycorrhizal fungi diversity and
ecosystem functioning (Leake 2001).

a)

Mycorrhizal fungal community structure o f arctic and alpine systems

Examining mycorrhizal fungal communities in alpine and arctic systems is preferable
because confounding factors due to direct human contact, such as logging or fire suppression,
that hamper studies in temperate forests have minimal impacts upon these arctic and alpine
systems (Trappe 1988). Another advantage is that these systems are thought to be relatively
simple in comparison to mycorrhizal fungal communities o f temperate systems (Read 1993).

36

Alpine and arctic systems share similar environmental stresses such as short growing
seasons (Trappe 1988, Haselwandter 1987), low air and soil temperatures, and large seasonal
and diurnal temperature fluctuations (Haselwandter 1987). Because o f harsher
environmental factors found in these systems, traits such as longevity and mycelial spread o f
individual fungal genets may be important (Gardes and Dahlberg 1996).
Dark-septate endophytes (DSE) are ubiquitous in both alpine and arctic systems
(Bledsoe et al. 1990, Cazares 1992). Although Kohn and Stasovski (1990) reported no DSE
were found on root tips o f plants from Alexandra Fiord, samples o f DSE from this area were
found later (see Chapter 3). Hyaline septate hyphae that are reported and found on several
plants may have the same ambiguous function o f being either pathogenic or mutualistic, as
has been found with DSE (Jumpponen and Trappe 1998).
Arbuscular mycorrhizae (AM) are scarce in the arctic (Bledsoe et al. 1990, Kohn and
Stasovski 1990) and higher elevations in alpine systems (Haselwandter 1987). Although,
Dalpé and Aiken (1998) found approximately 10% o f Festuca species are associated with
arbuscular mycorrhiza in the high Arctic, this is contrary to previous studies where no or very
little AM was found (Bledsoe et al. 1990, Kohn and Stasovski 1990). The discrepancy
between these may be due to the small sample size used by Bledsoe et al. (1990) and Kohn
and Stasovski (1990) (Dalpé and Aiken 1998). Regardless, AM appear to be scarcer in arctic
regions and in higher altitudes o f alpine systems. Both AM and ectomycorrhizae have been
found on Salix spp. in alpine studies (Trowbridge and Jumpponen 2004) while only
ectomycorrhizae have been found in arctic Salix systems (Vare et al. 1992).
Because AM are scarce in the arctic in contrast to alpine systems, and DSE are
common in both arctic and alpine, Bledsoe et al. (1990) suggested that DSE may replace the

37

functional role o f AM in arctic systems. Jumpponen (1999) suggested that the DSE,
particularly Phialocephala fortinii, may allow for transport o f carbohydrates between plants
through fungal linkages, as this function has been found for ectomycorrhizal fungi between
Betula and Pseudotsuga (Simard et al. 1997). This was suggested because the same genet o f
P. fortinii is found on nine different plant species that are classified as ecto-, ericoid, and
non-mycorrhizal (Jumpponen 1999).
Ericoid mycorrhizae are found on dw arf shrubs in the high arctic and nutritionally
stressed alpine plant communities (Haselwandter 1987). Haselwandter (1987) suggested that
ericaceous plants are capable o f using more complex nitrogen sources such as proteins or
amino acids, which is supported by findings that ericaceous plants take up amino acids in
alpine regions (Michelsen et al. 1996). Ericoid mycorrhizal fungi have also been found to
access N and P by producing enzymes that break down structural components o f litter such
as pectin and hemicellulose in plant cell walls, monophenols, tannins, polyphenols, and
lignin (Smith and Read 1997). Although ectomyeorrhizal fungi break down these structures
as well, the production o f these enzymes appears to be less common than by ericoid
mycorrhizal fungi (Smith and Read 1997).
The mycoflora o f alpine and arctic systems are similar in many aspects. Sporocarps
of ectomycorrhizal fungi in both systems are sparse in comparison to lower elevation,
temperate environments (Trappe 1988), which may be due to climatic factors which strongly
influence fruiting (Gardes and Dahlberg 1996). Preliminary data from Alexandra Fiord
indicate that several species, such as Russula sp., Cortinarius spp. and Inocybe sp., and
Cenococcum geophilum are dominant. Cortinarius spp. are dominant species (comprising
20% o f the abundance) on Salix arctica and Dryas integrifolia roots from the Canadian arctic

38

archipelago (Gardes et al. 2000). Contrary to the findings o f Kohn and Stasovski (1990),
Cenococcum geophilum was found in the present study as well as fruiting bodies o f
Lycoperdon and Helvella. Lycoperdon spp. are reputed to be myeorrhizal with Picea abies,
Pinus nigra, Pinus strobus, Pinus sylvestris, Pseudotsuga menziesii. Eucalyptus spp., and
Quercus spp. (Trappe 1962). Helvella aestivalis formed myeorrhizae with Dryas octopetala,
and species o f Helvella formed mycorrhizae with Salix reticulata under axenic eonditions
(Weidemann et al. 1998). Helvella crispa is reported to form myeorrhizae with Fagus
sylvatica and Quercus spp., and H infula with Picea abies (Trappe 1962).

b)

Mycorrhizal fungal succession in relation to plant community succession

Glacier forefronts are commonly used for research on mycorrhizal fungi during
primary succession. Primary succession is when pioneer species colonize virgin surfaees
(Frankland 1998), and seeondary succession is when the soil is nutrient poor after a
disturbance (Smith and Read 1997). Mycorrhizal fungi may improve nutrient-poor
conditions for latter species as detected by increasing diversity after volcanic disturbances
(Titus and del Moral 1998) and glacial tills (Helm et al. 1999). This facilitative nature o f
mycorrhizal fungi is inferred by the successional pattern described for primary succession,
which starts with non-mycorrhizal plants, AM plants, then ECM plants (Read 1993) and/or
ericoid plants (Cazares 1992). Ectomycorrhizae are thought to colonize in older soils
because they have access to nutrients contained in organic residues that are more abundant in
later stages after accumulation o f organic matter (Read 1993), such as the increase o f
nitrogen and organic matter (Jumpponen et al. 1998) along a chronosequence, which is a
sequential change o f related variables in certain properties, from an alpine glacial forefront.

39

Changes in carbohydrates supplied by the host (Dighton and Mason 1985), nitrogen
availability (Baar 1996), and soil conditions (Termoshuizen 1991, Kranabetter and Wylie
1998) are some o f the mechanisms suggested for primary succession to progress. For plant
communities, primary succession can depend on life history traits, such as seed size and
growth rate, maximum height o f the plant, seed rain, and competition. Facilitation and initial
site conditions are important for the rate o f change and for species composition and
productivity (Chapin et al. 1994). As found with plant communities, primary succession for
mycorrhizal fungi probably is not dependent on a single variable. Several researchers found
that changes in one variable are not enough to describe fungal succession (Termoshuizen
1991, Baar 1996, Helm et al. 1999, Kranabetter and Wylie 1998).
Svoboda and Henry (1987) described three types o f succession; 1) directional,
replacement succession with low resistance; 2 ) directional, non-replacement succession in
high resistance environments; and 3) non-directional, non-replacement succession in extreme
environments. In directional replacement succession, succession goes through serai stages
with species replacement until a relatively stable ecosystem is reached. In directional, non­
replacement succession, species are not replaced but live in co-existence with the invading
species, which expand slowly. In these systems, in which polar semi-deserts are an example,
there is enough space for expansion. Non-directional, non-replacement succession is found
in extreme environments, such as polar deserts where few species survive. Several species
may invade repeatedly but fail to establish permanently.
The lowlands o f Ellesmere Island fit the directional, non-replacement succession
description. The mycorrhizal guild system may play an important role in plant competition
(Newman 1988, Horton and Bruns 1998) if co-existence between plants is typical and

40

expansion is slow. Although Kropp and Trappe (1982) suggested that pioneer plants may be
more host-specific, the case on Ellesmere may be different because succession does not
follow the replacement o f plant species but the co-existence o f additional species. Moving
away from the glacial forefront, the plant community starts with Papaver lapponicum and
Luzula confusa. Salix arctica, Saxifraga oppositifolia, Cassiope tetragona, and Dryas
integrifolia eventually appear, and all six species are found not only on the glacier forefront
but the rest o f the lowlands o f Alexandra Fiord. Van der Heijden and Vosatka (1999)
showed that with AM, the increase o f AM fungal composition and number leads to an
increase in plant diversity as well because more variety o f AM fungi allow different plants to
establish themselves. Perhaps the increase o f ectomycorrhizal and ericaceous mycorrhizal
fungi will have a similar capacity o f increasing plant diversity and stability.
Understanding succession on Ellesmere Island will be different from other studies,
including those that are conducted in the Arctic. Previous successional studies in the Arctic
occurred in the low arctic where trees still grow (Helm et al. 1999, Helm et al. 1996,
Brubaker et al. 1995, Chapin et al. 1994) while Ellesmere Island is located in the high arctic
where only low shrubs are found.

5. Global climate change: effects on aboveground plant community structure
Global warming is a complex type o f disturbance because not only can it have direct
effects on an organism or ecosystem, it can also lead to other disturbances. For example,
warming has been linked to increase fire frequency (He et al. 2002) and more intense
hurricanes (Shen et al. 2000). For this study, the effects o f warming will be examined on the
community level even though this disturbance is classified at the ecosystem or landscape

41

level, according to the hierarchical organization of disturbances described by Pickett et al.
(1989).
Global circulation models predict that arctic systems will not only experience
warming before other ecosystems, but also undergo the greatest increase in surface
temperatures due to the doubling o f CO 2 (Shaver et al. 1992, Oecbel et al. 1993, Henry and
Molau 1997). Climate change will have a more dramatic effect on the arctic than other forms
of disturbance, mostly due to its spatial isolation, so findings from the arctic can be used to
predict bow other systems may respond (Shaver et al. 1992). By the year 2100,
approximately 63% o f biodiversity will be altered due to climate change in the arctic,
compared to other human-induced disturbances such as changes in land use (15%),
introduction o f exotic species (4%), and changes in atmospheric CO 2 and/or nitrogen
deposition (18%) (Chapin et al. 2000). Warming is expected to increase more in winter
months (up to 17° C) than during the summer months (~4° C) (Oechel et al. 1993, Edlund
1992), thus lengthening the growing season o f plants (Edlund 1992, Henry and Molau 1997)
and altering plant communities through changes in the distribution o f snow in the winter,
persistence o f snowbeds, and pattern o f snowmelt (Edlund 1992). Global climate change
will likely amplify in arctic regions due to positive feedback loops that include; 1) ice and
snow melt that would decrease surface albedo; 2 ) stabilization of the atmosphere that may
trap temperature anomalies near the ground surface; 3) cloud dynamics that may amplify
change (Overpeck et al. 1997); and 4) the permafrost layer melting sooner (Oechel et al.
1993). Warming in the arctic affects lower latitudes by possibly changing river run-off and
the circulation o f the atmosphere, and increasing atmospheric concentrations o f CO 2 and CH4
(Overpeck et al. 1997, Henry and Molau 1997).

42

By understanding how plants established historically, predictions o f how climate
change will affect plant species evolutionarily and geographically may be more accurate
(Murray 1995). For example, in the early Holocene period, warming increased the number
o f shrubs, which parallels the present spread o f dwarf shrubs {Salix spp., Betula nana, and
Alnus crispa) in Alaska (Sturm et al. 2001). This is indirect evidence that these regions in
Alaska may adapt relatively quickly to climate change (Sturm et al. 2001). Also, fossil
records of some species such as Dryas integrifolia and Saxifraga oppositifolia indicate that
these plants have existed since the Tertiary period (Murray 1995) and, therefore, have
survived temperature fluctuations for at least 1.8 million years.
Most o f the present arctic flora established approximately 6000-3000 b.p. (Brubaker
et al. 1995) and originated from: 1) survivors from Tertiary forests, northern réfugias from
Quaternary glaciation, and Pleistocene migration from Asia; 2) plants that returned during
interglacial and post-glacial time from unglaciated areas; and 3) newly evolved species from
the Pleistocene and Holocene (Murray 1995). According to Late Quaternary pollen records,
species found in the arctic tundra are thought to have expanded southward into much of
Canada (Brubaker et al. 1995).
Arctic systems are carbon sinks. Current carbon sinks are the wet and moist tussock
tundra of arctic systems (Oechel et al. 1993, Shaver et al. 1992). Arctic systems have three
times more soil carbon than alpine systems but only 13% o f the plant species richness, which
indicates active accumulation o f soil organic matter and little disturbance (Chapin and
Komer 1995). Release o f carbon to the atmosphere is predicted to be caused indirectly and
not directly from the increase o f temperature (Oechel et al. 1993). Researchers have
suggested that the cause o f the loss o f carbon from arctic systems to the atmosphere is due to

43

enhanced drainage and soil aeration, deerease in the water table (Oechel et al. 1993, Shaver
et al. 1992, Billings et al. 1983), and increase in respiration, especially from the soil
microbial community (Schimel 1995, Billings et al. 1983) o f whieh myeorrhizal fungal
hyphae are thought to be a large eontributor (Rygiewicz and Andersen 1994). Change in
carbon storage is somewhat constrained by the nitrogen cycle because nitrogen is the primary
limiting faetor in aretie systems (Shaver et al. 1992, McKane et al. 1997).
With enhaneed drainage and soil aeration, decomposition and release o f carbon will
likely occur in systems that have large amounts o f earbon storage such as high latitudinal
bogs, and boreal and arctic systems (Oeehel et al. 1993). Billings et al. (1983) found that a 48 ° C warming decreased the net carbon storage in the wet sedge tundra rather than increased

net primary production, which they attributed to greater inerease in soil respiration.
However, loss from carbon storage may be for the short-term, and eventually increase in
above ground plant biomass may compensate for the earbon loss (Oechel et al. 1993).
Many factors influence the impacts o f global climate change on above ground plant
growth such as water availability, nutrient availability, summer warmth, snowfall, (Edlund
1992, Field et al. 1992), light, and CO 2 levels (Field et al. 1992). For arctie systems,
warming may first impact individual plants, indicated by an increase o f vegetation growth
(Edlund 1992). Response by plant communities would depend on the eombination of
summer warming, snowfall in the winter, possible drought in the summer (Edlund 1992), and
resource availability (Field et al. 1992). Global warming may result in major reorganization
o f plant communities (Brubaker et al. 1995); however these changes for the plant eommunity
may take centuries (Edlund 1992).

44

Several studies have examined the impacts o f global warming on arctic plants in situ
by manipulating temperatures with greenhouses (Hobbie and Chapin 1998, Havstrom et al.
1993) or open-top chambers (OTCs) (Henry and Molau 1997, Stenstrom et al. 1997, Jones et
al. 1997) placed over plots. Open-top chambers have some advantages over closed
greenhouse systems by allowing in more direct solar radiation; lessening the chance of
overheating; allowing herbivores and pollinators to the plants; and avoiding decreased
relative humidity (Marion et al. 1997). Problems o f both systems include increasing
temperature extremes rather than lowering the range o f diurnal temperatures, altering o f wind
patterns around the plant, and disturbing the sites (Marion et al. 1997, Hobbie and Chapin
1998). Problems that are unique to OTCs consist o f snow accumulation, disturbance by
animals (Marion et al. 1997), and only a small area can be uniformly warmed (Shaver et al.
2000 ).
Table 1.1 summarizes experiments o f warming on dominant plants o f arctic systems
that will be used in this present study. Experiments using greenhouses to increase air
temperature find no significance o f warming on Cassiope tetragona (Hobbie and Chapin
1998, Havstrom et al. 1993), which lead researchers to conclude that perhaps nutrients rather
than temperature affect C. tetragona growth (Hobbie and Chapin 1998). Their findings are
contrary to what is found when OTCs are used, where warming did increase different factors
measuring plant growth (Henry and Molau 1997). This discrepancy may be because major
changes in the arctic tundra from warming o f the last glaciation have little similarity in
different circumarctic sites (Brubaker et al. 1995), as the two studies are in Alaska and
eastern Canada. Another explanation may be that the greenhouse experiments do not allow
for enough time for temperature increase to show significant differences as is found with an

45

OTC study on Salix arctica, where Henry and Molau (1997) found no significance after two
years but did after four years.

46

Table 1.1 Summary of experiments o f warming on arctic plants used m present study.
Plant
Reference
Location
Temperature
Plant Growth
Manipulation
Cassiope
Hobbie and
Alaska,
Greenhouse
Biomass decrease.
tetragona
Chapin (1998) USA
Havstrom et al. Spitsbergen, Greenhouse
Significant increase from
(1993)
Norway
temperature for leaf indices
and shoot growth index.
Henry and
Ellesmere
OTC
Significant increase in plant
Molau (1997)
Island,
growth.
Canada
Stenstrom et
Saxifraga
OTC
Ellesmere
Increase in flowering
oppositifolia
al. (1997)
Island,
frequency and reproductive
Canada
success.
Salix arctica
Jones et al.
OTC
Increase o f plant growth after
Ellesmere
(1997)
Island,
3-4 years o f warming.
IncreaseW in seed production
Canada
first years o f warming.
Stronger impact on male
willows than females.
Dryas
Henry and
OTC
Ellesmere
Significant but not strong
integrifolia
Molau (1997)
Island,
increase in vegetative and
Canada
reproductive phenologies,
seed set and weight,
germination.
NA: not available

Predictions of Adaptation
Due to Global Warming
NA
Will probably not migrate
south as air temperature
increases.
NA

Will probably be
outcompeted by graminoids
and forbs.
May adjust easily to global
warming because it adapts
to broad range o f
ecosystems.

NA

6 . Impact o f global climate change on root fungal communities

Much o f the research on global climate change on mycorrhizae has been indirect
focusing on the effects o f elevated CO 2. Results have been conflicting showing decreased
(Fitter et al. 2000), increased, and no difference for ectomycorrhizal and AM colonization
due to elevated CO 2 levels (Fitter et al. 2000, Treseder and Allen 2000). The impact o f
elevated CO 2 on mycorrhizal growth and colonization seems to depend on; 1) mycorrhizal
fungal species because some species are more sensitive than others to elevated carbon (Fitter
et al. 2000); 2) availability o f N, where additional N can negate the effects o f CO 2 on
mycorrhizal biomass for some systems (Treseder and Allen 2000); 3) growth rate o f plants
because larger plants need more roots (Fitter et al. 2000, Treseder and Allen 2000); and 4)
roots lengthening which would increase mycorrhizal colonization (Eissenstat et al. 2000).
In a review by Fitter et al. (2000), only two studies that examine the increase o f
temperature on AM colonization are published and no studies have currently been published
on the effects on ectomycorrhizae. Perry et al. (1990) speculated on the role o f mycorrhizal
fungi in climate change in that plant species would migrate during climate change and that
sharing mycorrhizal fungi would help with the transition.
Although there is a lack o f experiments that examine the impact o f global warming on
mycorrhizae, the rhizosphere will probably be an important factor in how ecosystems adjust.
The major effects o f global warming may be from its impact on soil processes rather than the
increase of biomass o f plants in the tundra (Hobbie and Chapin 1998). Boone et al. (1998)
suggested that the rhizosphere would be more sensitive to warmer temperatures than above­
ground plant parts, and that variation in soil respiration is determined by responses o f root
respiration and heterotrophs to temperature change. Warmer soil temperature may influence

48

root growth, cell elongation, initiation o f new lateral roots, increase in root respiration and
ion uptake, interaction with water and nutrient availability, more N mineralization, less water
availability, and earlier initiation o f root growth in the spring (Pregitzer et al. 2000).
Although global warming is suspected to have significant impacts on the rhizosphere,
much o f implications have been speculative. In addition, information on the impacts on the
mycorrhizal community, in particular ectomycorrhiza, is scarce.

C. Research objectives and hypotheses
Alexandra Fiord provides an opportune site to examine root-associated fungal
communities. The role o f these fungi in plant establishment in primary succession increases
the understanding of how plants and fungi adapt to nutrient-poor conditions. Future
ecological conditions are examined by use o f OTCs to simulate potential global warming
scenarios.
This study will be one o f the first to examine the impact o f global warming on the
root-associated fungal community. Because global warming impacts arctic systems more
intensely than temperate environments, this may give insight to the role o f root-associated
fungi in facilitating changes to the plant community. Although global warming is suspected
to have significant impacts on the rhizosphere, much o f the implications have been
speculative. This study examines how warming may impact the root-associated fungal
community by using both PCR-based techniques and isolating fungi from root tips.
Although there should be some overlap, the PCR-based techniques are more likely to favor
mycorrhizal fungi; whereas, fungal isolations would favor faster-growing root endophytes
that do not fit the morphological definition o f mycorrhizae.

49

By examining the changes o f the root-associated fungal community in a direct, non­
replacement succession, insight will be gained as to how this community would adjust to a
changing plant community. This type o f succession also has the unique characteristic where
the increase in biodiversity o f plants happens in vivo while retaining all the same plants. This
study can determine if the root-associated fungal community follows a similar trend. Also,
this will be the first study to examine the root endophytic community for this type o f
succession.
Given what the literature indicates about fungal community structure, I expect that
the root-associated fungal community, based on direct DNA extraction, will differ according
to site and treatment (passive warming versus ambient). The following null hypotheses will
therefore be tested:
Ho 1.1 The root-associatedfungal community, based on DNA directly extracted from root
tips from Salix arctica, will not differ between warmed plots and ambient plots.
Ho 1.2 The root-associatedfungal community, based on DNA directly extracted from root
tips from Salix arctica, will not differ due to site.

Because culture studies may reveal a different perspective than direct DNA
amplification studies, I expecte that the root-associated fungal community will differ
according to site, treatment (passive warming versus ambient), and host plant. Therefore, the
following null hypotheses will be tested:
Ho 1.3 The root-associated fungal community, based on cultures isolatedfrom root tips, will
not differ between warmed and ambient plots.

50

Ho 1.4 The root-associated fungal community, based on cultures isolated from root tips, will
not differ according to the host plants Cassiope tetragona, Dry as integrifolia, Salix
arctica, and Saxifraga oppositifolia
Ho 1.5 The root-associated fungal community, based on cultures isolatedfrom root tips, will
not differ due to site.

Because eulture studies only assess culturable fungi, 1 expeet that the fungal
communities described by the two methods (direet DNA amplifieation from roots versus
culturing) will be different (despite revealing the same patterns aeeording to site, treatment,
and host plant). Therefore, the following null hypothesis will be tested:
Ho 1.6 The root-associated fungal communities described by the two methods (direct
extraction versus culturing) will not differ.

The unusual directional, non-replacement sueeession pattern found in high arctic
systems permits me to examine how diversity on different host plants varies along a
ehronosequenee, without the confounding factor o f host plant replacement. My objective
was to examine how the root-associated fungal community changes during a directional, non­
replacement primary plant succession, using Cassiope tetragona, Dryas integrifolia, Luzula
confusa, Papaver lapponicum, Salix arctica, and Saxifraga oppositifolia as host plants
Therefore, the following hypotheses for this objeetive will be tested:
Ho 2.1 The root-associated fungal communities will not differ along a ehronosequenee.
Ho 2.2 The root-associated fungal communities will not differ according to host plant.

51

D. Thesis organization
This thesis starts with a literature review to provide background on the concepts,
theories, and techniques associated with this thesis. This provides more comprehensive
information that may not be covered in subsequent chapters.
The next two chapters address the first research objective, using two different
methods that collectively provide a more complete assessment o f the root-associated fungal
community. One method, directly amplifies fungal DNA from root tips, and the second
involves isolations o f fungi from roots. These two methods have described different fungal
communities found on the same plant host in previous studies. Chapter 2 addresses how
warming will impact the root-associated fungal community detected by direct extraction o f
fungal DNA from root tips o f one host species and covers hypotheses 1.1-1.2. Chapter 3
addresses the question based upon fungal cultures from several host species, aseptically
isolated from plant roots, and will cover hypotheses 1.3-1.5.
Chapter 4 covers research objective 2 and chapter 5 is a synthesis o f the research
findings. This synthesis will cover hypothesis 1.6 , and will attempt to tie the three studies
together. The final chapter is a summary o f the thesis.

52

IL Impact of warming on the frequency and diversity of the root-associated fungal
community on Salix arctica from the Canadian High Arctic

A. Introduction
Global circulation models predict that arctic systems experience will experience a
greater effeet o f global warming before other ecosystems due to increased surface
temperature and CO 2 levels (Shaver et al. 1992, Oechel et al. 1993, Henry and Molau 1997),
decreased surface albedo, alterations in cloud dynamics that may amplify change (Overpeek
et al. 1997), and melting o f the permafrost layer (Oechel et al. 1993). Plant communities
may ehange due to lengthening o f the growing season (Edlund 1992, Henry and Molau 1997)
by inereasing air and soil temperature (Oechel et al. 1993, Edlund 1992) and altering water
distribution by changing the dispersal o f snow, increasing the persistence o f snowbeds, and
modifying the pattern o f snowmelt (Edlund 1992).
The objective o f this study was to examine the impacts of experimental warming on
the root-associated fungal community o f Salix arctica in the Canadian high arctic. Although
there have been numerous studies on the impact o f experimental warming on arctic plants
(e.g. Chapin et al. 1995, Henry and Molau 1997, Hobble and Chapin 1998, Jones et al. 1997,
Sturm 2001), no studies have examined how warming may impact the root-associated fungal
community even though the rhizosphere may play an important role in plant response
(Hobble and Chapin 1998, Boone et al. 1998). To date, only one study has examined the
impact of warming on an ectomycorrhizal fungal community from Douglas-fir {Pseudotsuga
menziesii Mirb. Franco) seedlings (Rygiewicz et al. 2000), where they found warming to
increase species richness. Perry et al. (1990) speculated that the mycorrhizal fungal
community may help facilitate migration o f plants. Hobble and Chapin (1998) suggested that
53

the major effects o f global warming will result from the influences o f increased soil
temperatures and the subsequent soil processes. Increased soil temperature may influence
root growth by increasing cell elongation, initiating new lateral roots, increasing root
respiration and ion uptake, interacting with water and nutrient availability, increasing N
mineralization, decreasing water availability, and initiating root growth in the spring earlier
(Pregitzer et al. 2000). Other studies have indirectly examined the impact o f warming on
mycorrhizal fungi by studying how CO 2 fluxes may impact the community (Oechel et al.
1993). Some researchers have suggested that the main cause of the loss o f carbon from arctic
systems to the atmosphere is enhanced drainage and soil aeration, decrease in the water table
(Oechel et al. 1993, Shaver et al. 1992, Billings et al. 1983) and increase in respiration,
especially from the soil microbial community (Schimel 1995, Billings et al. 1983) o f which
mycorrhizal fungal hyphae are thought to be a large contributor (Rygiewicz and Andersen
1994).
We chose three distinct sites, a lowland site, highland granitic site, and highland
dolomitic site, to see how soil type interacts with warming to influence the abundance,
composition, and biodiversity o f these communities. In order to address these questions, we
used terminal restriction fragment length polymorphism (T-RFLP) analysis, a technique
commonly used in prokaryotic systems and but only used recently in mycorrhizal community
analysis (Dickie et al. 2002, Klamer et al. 2002).

B. Materials and methods

1. Study Site
Samples were collected from three sites at Alexandra Fiord on Ellesmere Island,
Nunavut, Canada, 78° 53’N, 75° 55’W. One site was located on the lowland (valley bottom)
54

and two on a mountain plateau. The lowland mesic site was at or near sea level and enelosed
by 450-700 m-high plateaus to the east and west, a glaeial forefront to the south, and the
oeean to the north. The elimate at the lowland site was relatively warm due to frequent
sunny skies, relatively warm air masses from the west and south (Labine 1994), and
reflection o f sunlight from the surrounding cliffs and oeean (Freedman et al. 1994). The soil
was granitic, which is mostly composed o f quartz, feldspar, and mica. The highland sites
were located on top o f the western plateau at 450 m. These sites were xeric with sparse
vegetation and have desert and semi-desert arctic conditions (Batten and Svoboda 1994),
which accounted for the decrease o f diversity o f vascular plants by approximately 40% when
compared to lowland areas (Batten and Svoboda 1994). The two highland sites were
distinguished by soil type, dolomitic, which is distinguished by high amounts o f calcium
magnesium carbonite (CaMg ( € 0 3 )2), and granitic for the other site.

2. Vegetation
Alexandra Fiord has been described as a ‘polar oasis’ because it comprises an 8 -km^
pocket of arctic shrubs, mosses, lichens, and sedges nested within vast ice fields. Dryas
integrifolia Vahl, Cassiope tetragona (L.) D.Don, and Salix arctica Pall, are the most
prominent (Freedman et al. 1994) o f the 92 species o f vascular plants found on the lowlands
(Ball and Hill 1994) as well as the less vegetated and less diverse uplands (Batten and
Svoboda 1994).
The lowland site has been described as a dwarf shrub-cushion plant community (Muc
et al. 1994) and is dominated by Salix arctica and Dryas integrifolia, along with Saxifraga
oppositifolia L., Cassiope tetragona, Papaver, Pedicularis, sedges, and mosses. The granitic
upland site, described as a S. arctica-C. tetragona dominant community (Batten and Svoboda

55

1994) includes S. oppositifolia and D. integrifolia. The upland dolomitic site is a S.
oppositifolia-àommstQdi community (Batten and Svoboda 1994) with no D. integrifolia
present.
To simulate the increased temperatures predicted by global climate models, open-top
chambers (OTCs), which covered 0.8 m^ and were 0.3 m high, were used to increase air
temperatures by 1-4° C, the predicted temperature range o f global maean increases for the
middle of the 21®' century for the Canadian high arctic due to climate change (Henry and
Molau 1997). Three 1-m diameter OTCs were placed on each of the three sites; in 1995
(GHR Henry 2000, pers. com.) for the lowland site, and 1993 for the two highland sites
(Stenstrom et al. 1997). Plants that were found between 0.5 m and 1.5 m from the OTCs
were harvested as controls for a total o f 18 plots.

3. Field collection
Two specimens o f Salix arctica were harvested from each o f the plots in August
2000. Plants and surrounding soil were kept in Ziploc® bags (18 x 20 cm) in a permafrost
cooler while in the field and in a 4° C refrigerator once back at UNBC until processed.
Two 300 gram samples of soil were collected in August 2001 from each plot. Soil
was collected no more than 1 m from the harvested plant. Soils were dried and separated in a
2-mm sieve to remove rocks from the samples. One hundred grams from each o f the two
replicates were mixed and sent to the Ministry o f Forests, Research Branch Laboratory,
Analytical Chemistry Section in Victoria, BC for the following analyses: pH in water, total C
and N using combustion elemental analysis, cation exchange capacity and exchangeable
cations using 0.1 N barium chloride extraction, available NH 4-N and NO 3-N extracted with
2N KCl, and available P using the Mehlich III protocol.

56

Fungal sporocarps from Alexandra Fiord were collected and tentatively identified to
genus; however, no sporocarps were found in any o f the plots per se. A 2- x 2- mm^ piece
was extracted from each sporocarp and stored in 50% EtOH for DNA extraction. The
remaining sporocarp tissue was dried for storage.

4. Sampling from roots
Plant roots were immersed in water for at least 24 hours at 4° C. The root systems were
gently cleaned with water and collected in a 0.5 mm sieve (No. 35 USA standard testing
sieve, W.S. Tyler, Inc.). Root systems were placed on a numbered grid for random selection
o f root tips. Numbers from a random-numbers table were used to select grids for sampling.
The root tip that traversed or was closest to the grid was selected, and a 2-cm root section
was sampled. Fifteen root tips were randomly selected, described morphologically, placed in
a 1.5 ml microcentrifuge tube, and frozen at -40° C until DNA was extracted. Root tips were
assessed morphologically (morphotyping) based upon characteristics such as color, tip
ramification, and presence and absence o f rhizomorphs, loosely following the techniques o f
Agerer (1987-1998) and Goodman et al. (1995). An additional forty root tips were randomly
selected, frozen, lyophilized, and stored at -40° C for further DNA extractions if needed. O f
these additional tips, approximately 15 were later extracted to increase number o f samples.
The remaining root system was frozen and lyophilized. Preparation for DNA extraction was
done first for the highland sites because the plants did not appear as robust as at the lowland
site. Plant tissue from stems or leaves was also sampled to determine if plant DNA would
amplify with the chosen primers.

57

5. DNA Extraction and ITS-T-RELP analysis
DNA o f root tips was extracted using the CTAB protoeol o f Gardes and Bruns
(1996b), which was modified by excluding the freeze-thaw procedure and by including
another purification step o f adding phenolxhloroform-isoamyl aleohol (1:1) (Lee and Taylor
1990). Individual frozen tips were ground in 300 pL o f 2X CTAB buffer (100 mM Tris at
pH 8 , 1.4 M NaCl, 20 mM EDTA, 2% CTAB, and 0.2% p-mercaptoethanol) and then
incubated at 65° C for one hour. Equal volumes o f phenol: chloroform-isoamyl alcohol
(24:1) were added, vortexed, and centrifuged at 13,000 ref for 15 minutes. The supernatant
was then transferred to fresh tubes. A second wash o f chloroform-isoamyl alcohol (24:1)
was added (v/v), vortexed and eentrifuged for 5 minutes. Again the supernatant was
removed and placed in fresh tubes. Nucleic acid was precipitated by adding 500 pL of
isopropanol and ineubated for at least 3 hrs at -20° C. This was centrifuged for 15 min, and
the isopropanol was removed. The remaining pellet was washed with 70% ethanol and then
centrifuged for 10 minutes. The ethanol was removed and allowed to evaporate before the
pellet was resuspended in 50 pL o f TE -8 buffer. Plant tissue and sporoearps were extraeted
with the same protocol except the extra phenol :chloroform-isoamyl alcohol purification step
was not included, and pellets were resuspended in 100 pL o f TE -8 buffer instead o f 50. All
extractions were stored in -20° C until use.
Amplification o f the nrDNA ITS region was done using lOX Buffer (200 mM TrisHCl [pH 8.4], 500 mM KCl) (Invitrogen), 2X dNTPs, 25 mM MgCb, 0.5 pM o f ITS 1
(White et al. 1990) dye labeled with Cy 5.0, 0.5 pM o f ITS 4 (Gardes and Bruns 1993) dye
labeled with Cy 5.5, 5 U o f Platinum Taq DNA Polymerase (Invitrogen), and 0.1 OX DNA
template . The following program was used for PCR on a MJ Research thermocycler PTC100: 94 °C (4 min); 48° C (1 min); 72° C (2min); [94 °C (30 sec); 48° C (30 sec); 72° C (1
58

min 30 sec) x 34 cycles]; 72°C (6 min 30 sec). Samples were run on a 0.7% agarose gel to
confirm amplification. Unsuccessful amplifications were redone with no dilution and 1:10
dilutions o f the DNA template.
Digestions using either Alul or Hin^{ (Invitrogen) and 6 pL o f the PCR produet were
completed in 10 pL reactions following the manufacturer’s recommendations and incubated
at 37° C for at least 3 hours. One microliter o f the digested samples was mixed with 1.65 pL
o f loading dye mixture that contained formamide and two sets o f internal markers at 101 ,
200, and 351 bp; one set labeled with Cy 5.0, the other with Cy 5.5. The samples containing
the loading dye mixture were denatured at 80° C for two minutes and then quenched on ice.
Two microliters were then loaded onto a 6 % polyacrylamide gel with the laser power set at
50%, temperature at 53° C, and current at 1250 V and ran for 60 min. on an OpenGene
System Long Tower Sequencer (Bayer International). Fragments for each primer-enzyme
combination (i.e. ITS \-Alul, ITS \-HinQ., ITS A-Alul, and ITS 4-7/mfl) were determined
using GeneObjects 3.1 fragment analysis software.

6 . Matching root tips with sporocarps

Terminal restriction fragments (T-RFs) from root tips were compared to those from
sporocarps using TRAMP (T-RFLP analysis matching program) (Dickie et al. 2002) to check
for matches. However, because most o f the root tips exhibited multiple fragments,
determining T-RFs for individual genotypes was difficult. Identification o f T-RFs for
individual genotypes was attempted by matching all possible combinations o f at least three
fragments against the known fragments from sporocarps.

59

7. Data analysis
Frequencies o f genotypes generated from the T-RPLP analysis were tabulated for
each cnzymc-primcr combination. First, genotypes were binned and the average o f the
binned numbers was used to identify that genotype, then the number o f genotypes in each bin
was entered into frequency tables. Each different primer-enzyme combination was treated as
an independent database for determining if site and treatment bad changed the rootassociated fungal communities; therefore all analyses except for the ordination analyses were
done in quadruplicate.
Non-metric multidimensional sealing (NMS) was used to analyze the frequency
tables generated from the four primer-enzyme combinations using PC-ORD, version 4.25
(McCune and Mefford 1999) as an initial exploratory tool to visualize the overall effects of
site and treatment on genotype frequency. The four individual primer-enzyme frequency
tables were merged into a single table for the ordination, and Beal’s transformation was used
to alleviate problems associated with databases containing numerous zeroes (McCune and
Grace 2002). The following parameters for NMS were used: Sorensen was used for
measuring distance; the configuration for the first run was randomly selected; for all
subsequent runs the best configuration from the previous run was used, as recommended by
the program; 40 runs were conducted; dimensionality was assessed by examining the scree
plot; and 50 randomized runs were used in the Monte Carlo test.
Canonical correspondence analysis (CCA) in PC-ORD (McCune and Mefford 1999)
was used to examine the influence o f soil properties (i.e. pH, CEC, available NH4, available
NO 3, available P, and C:N ratio, and exchangeable cations) in explaining the differences in
genotype occurrence among sites. Genotype frequencies were transformed by adding one (to
remove zero frequencies), and soil properties were transformed by log 10 except for pH and
60

C:N ratio. Log transformation was used on soil properties because results from the different
properties greatly varied so the log transformation compressed high values and spread the
low values (McCune and Grace 2002). A Monte Carlo test was used to test for linear
relationships between the genotype abundance and soil properties. This test was used
because a large number of reiterations were needed to gain a more precise p-value (McCune
and Grace 2002). The ordination diagram was based on LC (linear combination) scores,
which are linear combinations o f soil properties (McCune and Grace 2002).
To further investigate the impacts o f site and treatment on genotype frequency, a
nested, two-way analysis o f variance (ANOVA) in Statistica, vers. 6.1 (StatSoft) was used.
Because the number of fungi extracted from root tips varied from each plant specimen,
genotype frequency was standardized by;
total number genotypes per plant specimen^
.
. . .
. This standardization
[1]------------------------------ —— - — ------total number o f root tips used from plant specimen^
r,,

assumes that genotypes were not present in unsuccessful amplifications.

The statistical

model used was:
[2] Yijki = p + Si + Tj + Rk(j) + Pi(jk) + (ST)ij + S(ijki)m, where
yijki = genotype frequency
Si = effect o f site, i = 1,2,3;
Tj = effect o f treatment (OTC or control), j = 1,2;
Rk(j) = replication effect, k = 1,2,3 nested within theyth treatment.
Pi(jk) = plant specimen effect nested within the Ath replicate and jth treatment, 1 = 1,2 .
ANOVA was also used to test if warming affected soil properties, which could
explain some o f the variation due to site or treatment. The model used was:
[3] yijk = p + Si + Tj + Rk(j) + Si*Tj + £(ijk)j., where
61

yyk = the amount (ppm for P and available N ) , concentration (exchangeable cations and
CEC), or ratio (C;N ratio, pH) o f soil properties;
Si = effect o f site, i = 1,2,3;
Tj = effect o f treatment (OTC or control), j = 1,2;

Rk(j) = replication effect, k = 1,2,3 nested within the treatment.
Genotype diversity was investigated by examining genotype abundance curves,
genotype richness, and evenness according to site and treatment. Genotype abundance
curves were made by taking the natural log o f genotype frequency and plotting it against the
arithmetic ranking o f frequency. Once a model was chosen from the genotype abundance
curves, niche apportionment analyses were conducted using PowerNiche, an Excel-based
macro that uses niche division algorithms (Drozd and Novotny 1999), which helps determine
if abundance is associated with random fraction, power fraction, broken stick, or other niche
apportionment models based on the works o f Tokeshi, Sugihara, and Mae Arthur (see
Magurran 2004). The selection and division exponents were varied for each test o f the power
model with 250 replications. Different models were tested by changing the selection (= k)
and division exponent (= m), which is indicative o f which niche apportionment model is
chosen, e.g. to test for the broken stick model, one is chosen for both exponents (i.e., k = 1 , m
= 1). Other than the broken stick model, the random fraction model (k=0, m =l), the power
fraction model (0<k< 1, m =l), and Sugihara’s sequential breakage model (k=0,
m=0:25:0:75) were tested (Drozd and Novotny 1999).
Genotype richness and genotype evenness were tested using EcoSim (Gotelli and
Entsminger 2001). Genotype richness measured the number o f genotypes in a rarefied
sample, i.e. different samples had the same abundance level before the number o f different

62

genotypes was counted (Gotelli and Entsminger 2001). Hurlbert’s probability o f
interspecific encounter (PIE) index was used to measure evenness (Gotelli and Entsminger
2001). ANOVA was performed for both richness and evenness using the model:
[4] yijk = p + Sj + Tj + Rk(j)+ Si*Tj + sykjj), where
yijk = genotype richness (number o f different genotypes) or evenness (probablilty),
Si = effect o f site, i = 1,2,3;
Tj = effect of treatment (OTC or control), j = 1,2;

Rk(j) = replication effect, k = 1,2,3 nested within the treatment.
Similarity in community structure was examined using Sorensen’s quantitative index
and analysis o f molecular variance (AMOVA). Sorensen’s quantitative index was used to
examine the similarity o f community composition between warmed and ambient plots in
each site using Estimates vers. 6 .Obi (Colwell 1997). This index examines the similarity
between communities by comparing genotype frequency distributions, and gives more
weight to frequent genotypes (Magurran 1988). For each plot, the average Sorensen’s
quantitative index for all four primer-enzyme combinations was calculated.
Variations in genotype frequency due to site, treatment, and within plots were tested
with AMO VA using Arlequin, vers. 2.0 (Schneider et al. 2000). These parameters were also
tested without rare genotypes (those found less than 5% o f the sample). Although AMOVA
is normally used for population studies, it was used here to test for variation in genotype
frequency. Genotype frequency variation between plants was analyzed using a Euclidean
distance measure and Ward’s method for linking groups for cluster analysis in PC-ORD
(McCune and Mefford 1999).

63

Changes in genotype composition were examined by tracking the relative frequency
o f dominant genotypes, (i.e. genotypes that occurred on more than five root tips for a plot).
Genotypes that were dominant for one plot were assessed in all plots for each primer-enzyme
combination.

C. Results
Overall 1105 extractions were attempted; on average 30 extractions per plant. For the
four primer-enzyme combinations, 1058 genotypes were found for the ITS \-Hinü, 1047 for
ITS 4-HinÛ, 985 for ITS \-Alu\, and 894 for ITS 4-Alul. None o f the aboveground plant
tissues amplified with the given primers. No sporocarp T-RFs matched the fragments from
root tips.
Site was the main determining factor in differentiating genotype composition and
frequency according to the results from NMS and cluster analysis (see Figs. A2.1 and A2.2).
Whereas genotypes clustered according to site, there were no clear patterns o f differentiation
o f genotypes by temperature (i.e. warmed or ambient). The root-associated fungal
community from the granitic site was more similar to the community found on the lowland
site than on the dolomitic site. The variation was largest for the dolomitic site, as indicated
by the higher scattering o f the plots than the other two sites. The overall variance was
explained by r^ =0.894 with two axes; axis 1 explained 64.9% o f the variance and axis 2
explained 24.5%. Thirty-seven iterations were needed for the final solution. Final stability
o f the model was met with the final stress = 12.1 (p= 0.0196) and final instability = 0.00363.

1. Soil properties
The soil conditions at the different sites were distinguished by CCA. The lowland
site had higher amounts o f Fe, K, and NO 3 than the other two sites (see Fig. 2.1); the
64

dolomitic site had higher C:N ratio, pH, CEC, and Ca; and the granitic site had higher levels
of NH 4. For the final CCA analysis, Al, Ca, and K were excluded beeause o f their high
eorrelation with other variables: Al with NO 3 (r = 0.91), Ca with CEC (r = 0.94), and K with
NO 3 (r = 0.91). The dimensions used for the final analysis were 18 plots and 95 genotypes
for the main matrix and 18 plots and 8 variables for the environmental faetors matrix. Three
axes were interpreted; the first axis explained 15.8% o f the varianee, the seeond 12.7%, and
the third 8.2%. Results from the Monte Carlo test indicate that genotype occurrence and soil
properties were related as indieated by eigenvalues that were higher than expected by chance:
0.071 (p = 0.02) for axis 1 and 0.057 (p = 0.01) for axis 2.
Results from ANOVA on the different soil properties confirm the outcome from the
CCA analysis. Site, not warming, was important in distinguishing the different soil
properties except for Al, Na, and Fe, which were too variable to interpret (see Fig. A2.3
Table A2.1). Soil properties did not change due to warming. The soil properties for the
lowland and granitic sites were more similar to each other than to the dolomitic site. The
dolomitic site was differentiated from the others by higher CEC, Ca, pH, and C:N ratio (Fig.
A2.3 c, d, i, j) and with lower K, Mn, NO3, and P (Fig A2.3 f, h, k, m). The granitic site had
higher NH 4 than the other two sites, but had a lower amount in the warmed plot compared to
the ambient one (Fig. A2.3 1). There was lower Mg on the lowland site than the granitie (Fig.
A2.3 g).

2. Genotype frequeney
The genotype frequeney was significantly lower for the dolomitie site compared to
the other two sites (p = 0.003, p < 0.001, p < 0.001, and p = 0.007 for ITS X-HinU, ITS 4HinÜ, ITS l-Alul, and ITS 4-Alul respeetively) for each o f the primer-enzyme eombinations.

65

Genotype frequency also increased significantly in warmed plots (p = 0.017, p = 0.029, p =
0.003, and p < 0.001 for ITS \-HinÛ, ITS

ITS \-Alu\, and ITS A-AM respectively)

(see Fig. A2.4, Table 2.1). However, after further examination o f the least square means with
their confidence intervals, only the warmed plots o f the granitie site had significantly higher
frequency than the ambient plot (see Fig A2.4). There were no significant interactions
between site and treatment for any o f the primer-enzyme combinations, and there were no
significant differences between the two plant specimen replicates.
The distribution patterns o f genotypes were log normal for most o f the treatments on
each site (see figs. A2.5-A2.7). Further analyses with the power niche model indicated that
these graphs fit the power fraction model for the lowland and granitic sites and broken stick
for dolomitic site for all primer-enzyme combinations except for ITS A-Alul (see Table 2.2).
ITS A-Alul differed from the other datasets in that the broken stick model fit best for the
lowland control and OTC plots and for the granitic control plot.

3. Diversity
Genotype richness (i.e. the number o f unique genotypes) varied according to site (p
0.002 ITSl-/7mfI, p <0.001 ITS A-Hinfl, p = 0.038 IT S I-^ M , and p < 0.001 ITS A-AIul) (see
Fig. A2.8, Table A2.2) but did not change due to warming. Genotype richness was lowest on
the dolomitic site, averaging 10 unique genotypes for the control and 20 genotypes for the
OTC plots. Genotype richness was about the same for the lowland and granitic sites with the
average of 26 unique genotypes for the control and 32 for the OTC. Genotype evenness did
not change due to warming and was significantly different for site according to the ITS I Hinfl analysis (p = 0.012), where the granitic site had lower evenness, and ITS A-Alul (p =
0.026), where the lowland site had higher evenness (see Fig. A2.9, Table A2.3).

66

Table 2.1 2-way, 2-stage nested ANOVA for the effects o f site, treatment, treatment replicate, and plant specimen replicate on
Source of
variation

Effect
(fixed/
random)

Intercept
Site

Fixed
Fixed

Treatment
Site*
Treatment
Replicate
(treatment)

Fixed
Fixed

Plant
specimen
(replicate)
Error

Random

a\

Random

Degrees
of
freedom

1
2
1
2

4

Sum o f squares

Mean square

F

IT SlHinfL
34.077
1.825

ITS4Hinfi
33.614
2.517

IT Sl7/mfI
34.077
0.912

1TS4HinÜ
33.614
1.259

IT S lHinü
766.403
7.742

1TS4HinÜ
670.630
11.509

IT SlHinÜ
< 0.001
0.003

1TS4Hinü
< 0.001
< 0.001

0.696
0.209

0.562
0.007

0.696
0.105

0.562
0.004

15.661
0.886

11.215
0.034

0.017
0.428

0.029
0.966

0.178

0 .2 0 0

0.044

0.050

0.612

1.171

0.670

0.411

0.436

0.257

0.072

0.043

0.616

0.391

0.715

0.876

2.357

2.187

0.118

0.109

6
20

p-value

Table 2.1 2-way, 2-stage nested ANOVA for the effeets o f site, treatment, treatment replicate, and plant specimen replicate on
Source of
variation

Effeet
(fixed/
random)

Intercept
Site

Fixed
Fixed

Treatment
Site *
Treatment
Replicate
(treatment)

Fixed
Fixed

Plant
specimen
(replicate)
Error

Random

00

Random

Degrees
of
freedom

1
2
1
2

4

Sum o f squares

Mean square

F

IT SlAlul
29.507
2.918

ITS4Alul
24.832
1.592

IT SlAlul
29.507
1.459

1TS4Alul
24.832
0.796

IT SlAlul
1122.557
13.268

1TS4- A M

1.03
0.170

0.866
0.017

1.034
0.085

0 .8 6 6

0.105

0.037

0.557

2 .2 0 0

2666.863
6.539

IT SlAM
< 0.001
< 0.001

1TS4AM
< 0.001
0.007

0.009

39.356
0.773

92.986
0.071

0.003
0.475

< 0.001
0.931

0.026

0.009

0.283

0 .1 1 2

0.879

0.974

0.499

0.0928

0.083

0.844

0.683

0.551

0.665

2.435

0 .1 1 0

0 .1 2 2

6
20

p-value

Axis 2
5301

5303

T:N

\ \ ^ ^ l o g H xch I e
\ l ( \ H xcHNa
\

\
^4-N

\N 0 3 )N

51C1

, 5103

viPg A v a il P \
\lo g Exch M n

540

54C2

Fig. 2.1. CCA on warming treatments and site with soil properties as biplots.
Image rotated at 15°. Coding is as follows: first two characters represent site, SI (lowland),
S3 (highland dolomitic), S4 (highland granitic); the last two characters signify the treatment
and the replicate o f that treatment, e.g. C l would be replicate one o f control, 0 3 would be
replicate three o f OTC. Site is clustered together with the lowland and highland granitic sites
clustering closer together than the lowland and highland dolomitic sites. Soil properties are
the biplot vectors radiating from the center. Biplots were used to assess the relationship
between genotype and soil properties.
69

Table 2.2 Summary o f Tokeshi’s model used to describe the rank abundance curves at each
Primer-Enzyme

Site

ITS 1Alul
Lowland control PL
Lowland OTC PF
Dolomitic control BS
Dolomitic OTC BS
Granitic control PF
Granitic OTC PF

ITS 4Alul
BS
BS
BS
PF
PF
PF

ITS 1Hinü
PF
PF
BS
BS
PF
PF

ITS 4HinÛ
PF
PF
BS
BS
PF
PF

However, genotype evenness was not significantly different when analyzed for the other
three primer enzyme combinations.

4. Community composition
Based on the Sorensen’s quantitative index, the similarity o f the genotypes from the
root-associated fungal communities between the control and the OTC for each site ranged
from 52%- 65% for the lowland site, 35%-55% for the dolomitic site, and 42%-66% for the
granitic site. The similarity was generally lower between sites, which ranged from 26%-59%
(see Table 2.3).
Results from AMOVA indicate that most o f the variation in genotype frequency is
explained by within plot variation (variation within plots averaged 0.95 for all four primer
enzymes - 0.85 without rare genotypes).
When comparing the increase in genotype frequency from control to OTC, most of
the additional genotypes came from increases in the frequency of dominant genotypes (the
three genotypes most frequently found in each plot) (see Table A2.4). For the lowland site,
the dominant genotypes remained dominant from the control to the OTC plot, except when
analysis was done with ITS4-v4M. The dominant genotypes remained dominant from the
control to the warmed plots also in the dolomitic site for all primer-enzyme pairs except for

70

IT S l-^ M . For the granitic site, results were less conclusive. When analyzed with IT SlHin^l and ITSl-^/wI, the relative frequency o f genotypes from the control to the OTC
increased from genotypes that had low relative frequencies ( 1-2 %) to mid-ranged frequencies
(8-9%). When the granitie site was analyzed with ITSA-Hinü and YYSA-Alul, the dominant
genotypes remained dominant between the ambient and warmed plots.
Approximately 7-11 genotypes were found on all plots; 6-8 genotypes were found
only on the lowland and granitic site; 0-5 genotypes were found on the two highland sites;
and 0-1 genotype was found only on the dolomitic site. Only 0-2 genotypes increased in
relative abundance from the control to the OTC plot; 0-1 genotypes decreased in all plots; 05 genotypes increased from the control to OTC on the lowland and granitie sites; and 1-3
genotypes decreased from the control to OTC on the lowland and granitic sites.

D. Discussion
Genotype composition, cumulative frequency, and richness differed primarily
according to site and not by warming. The dolomitic site had the lowest cumulative
frequency and richness, while the granitic and lowland sites were higher and more
comparable to each other. Reasons for the low richness and abundance for the dolomitic site
include a high C:N ratio, which is not conducive to the breakdown o f organic matter, and low
levels o f NH 4, NO 3, and P. NH 4 and NO 3 are forms o f nitrogen that are absorbed by
ectomycorrhizal fungi, which are either assimilated into the fungal mycelia or transformed
into glutamine and transported to plants (Martin et al. 2000), making available nitrogen a
possible limiting factor for these dolomitic sites. The low frequency o f genotypes found on
the dolomitic plots was not due to an effect o f storage since these samples were the first to be
prepared for DNA extraction.
71

Table 2.3 Sorensen’s quantitative index for treatments per site differentiated by primerenzyme combinations
For Hinû\
ITS 4
ITS 1
Lowland
Control
Lowland
OTC
Highland
Dolomitic
Control
Highland
Dolomitic
OTC
Highland
Granitic
Control
Highland
Granitic
OTC

0.65

Highland
Dolomitic
Control
0.28

Highland
Dolomitic
OTC
0.39

Highland
Granitic
Control
0.41

Highland
Granitic
OTC
0.5

0.61

0

0.28

0.39

0.56

0.59

0.33

0.3

0

0.51

0.32

0.29

0.32

0.37

0.59

0

0.34

0.28

0.43

0.45

0.33

0.39

0

0.66

0.4

0.44

0.33

0.34

0.42

0

Highland
Dolomitic
OTC
0.3

Highland
Granitic
Control
0.36

Highland
Granitic
OTC
0.35

0.29

0.35

0.44

0.55

0.31

0.33

0

0.26

0.33

0.3

0

0.5

0.3

0.56

0

Lowland
Control

Lowland
OTC

0

Table 2.3 Sorensen’s quantitative index For
Highland
Lowland
ITS 4 Lowland
Dolomitic
OTC
Control
Control
ITS 1
0.52
0
0.29
Lowland
Control
0.62
0
Lowland
0.29
OTC
0.2
0
Highland
0.18
Dolomitic
Control
0.23
Highland
0.23
0.35
Dolomitic
OTC
0.35
0.24
Highland
0.25
Granitic
Control
Highland
0.36
0.25
0.35
Granitic
OTC

72

Dolomitic sites are characterized by having CaMg(C 0 3 ) 2; the high pH o f 8 was due to
excess Ca and Mg. The high concentration of Ca on these sites also explains the low
availability o f P because Ca reacts with P to create calcium phosphate, which is insoluble.
Although not statistically significant, P levels dropped to negligible amounts in the warmed
plots on the dolomitic site, where the warmer temperatures may have accelerated the
chemical reaction to form calcium phosphate. The lack o f available P could explain the
paucity of plants and mycorrhizal fungi present at this site. However, some ectomycorrhizal
fungi can solubilize calcium phosphate (Martin et al. 2000) so that fungi that can fulfill this
role may have a competitive advantage over other fungi that cannot solubilize calcium
phosphate. This could play an important role in plant establishment and maintenance on this
dolomitic site.
Genotype frequency tended to be greater on the warmed plots compared to the
ambient ones but was significantly different only at the granitic site. Genotype diversity did
not increase due to warming. The lack o f significance to the response o f warming could be
compounded by different plant age, plant size, sampling effects, amount o f time plots were
allowed to warm, or the increase in soil temperature was too small to affect the rootassociated fungal community. Chapin et al. (1995) found significant increases in plant
biomass after nine years o f experimental warming in the low arctic. Even though our plots
have been warmed for 5-7 years, this may not have been sufficient to detect significant
changes. This may be particularly true for arctic Salix species because any excess energy
goes to seed production, so it may take a few years for the benefits o f excess energy to be
realized belowground (Jones et al. 1997). Further, Salix may be constrained in its response to
warming due to its tightly controlled meristem activity (Arft et al. 1999). Researchers also

73

found an increase in vegetative growth in the 1^*year o f experimental warming, but the effect
was no longer significant by the 4*’’ year, which may be due to limitation o f resources other
than temperature, soil nutrients, and decrease in litter quality indicated by a high C:N ratio
(Arft et al. 1999). There was no significant difference between warmed and ambient plots
for soil properties in the present study, which may indicate that our plots had acclimated to
temperature change. Some reasons as to why the increase o f genotype frequency from
ambient to warmed conditions was not found on the lowland and dolomitic site include:: 1)
the granitic site had harsher conditions than the lowland site, so the lowland site may be more
resistant to change; and 2 ) the dolomitic site was too harsh for many root-assoeiated fungal
species to survive
In cases where amplification was unsuccessful, the distinction between true ‘zeroes’
and failed amplification of a genotype was impossible to demonstrate conclusively. The
assumption that genotypes were not present in unsuccessful amplification was made because
numerous attempts were made to amplify these samples with and without diluting the DNA
template. Also, the results of the least squared means were more consistent between the four
different primer-enzyme combinations than by assuming that unsuccessful amplifications
were null. The alternative assumption, that the failed amplifieations were nulls, would have
favored plots for which amplification was difficult, i.e. the dolomitic site.
Our finding that there was no ehange in genotype richness due to warming is contrary
to the study by Rygiewicz et al. (2000) where experimental warming was found to increase
the genotype richness on Douglas-fir {Pseudotsuga menziesii) seedlings grown in
environmentally eontrolled chambers. This is probably due to difference o f the host plants
and their ecology. Douglas-fir grows in temperate forests, where the growing season is

74

longer the seedlings had more time to react to passive warming in the same time frame.
Other reasons include that seedlings were used for their study while we sampled mature
shrubs; the age o f the plant may determine what fungi are colonizing the roots (Fleming et al.
1984); and that we used field samples, which would increase the range o f possible fungi that
may colonize the roots. The lack of change in genotype richness may indicate that our sites
were more stable and less susceptible to warming (Rygiewicz et al. 2000, Kârén and Nylund
1997), which is reasonable considering these plant communities have adapted to past
warming events.
Similarity in community structure was greatest within sites as supported by results o f
Sorensen’s quantitative index and the ordination analyses. Although the results from
AMO VA indicate that the genotype composition is the same between sites and treatments, by
examining the ordination graph from NMS (Fig A2. 1), the dolomitic site had high variation,
which could have overshadowed any significant differences that the AMO VA may have
found. The large variation found on the dolomitic sites as compared to the other two is
probably there was less diversity. The same number o f root tips was sampled, but
amplification was not as successful as the other two sites. Perhaps sampling from many
different plants would have captured greater diversity for the given sites (Gehring et al.
1998). However, this was not possible because many o f the OTCs had only a few plants in
them, and these ecosystems have low plant diversity and abundance.
Part of the reason for studying arctic sites was the assumption that these ecosystems
would have a ‘simpler’ mycorrhizal community (Read 1993). This was not the case; these
systems had 25-35 genotypes for the lowland and granitic sites, which is comparable to other
studies using T-RFLP analysis such as Klamer et al. (2002) who found between 19-35

75

genotype in an oak shrub community, and Dickie et al. (2002) who observed 26 genotypes in
a 60 year old pine plantation, as well as studies using RFLP analysis in temperate forests
(Kârén and Nylund 1997, Gardes and Bruns 1996a, Horton and Bruns 1998, Dalhberg et al.
1997). Our findings were less comparable to other arctic and alpine studies. We had
comparable numbers o f genotypes associated with S. arctica with an alpine study, where 34
genotypes were found on four plant species and with an arctic site in Greenland where 137
genotypes were found on 10 plant species; however;our results were less comparable to one
study where 60 ectomycorrhizal fungi were found on S. herbacia (Gardes and Dahlberg
1996).
Arctic plant systems have been reported to conform to a geometric series model
(McKane et al. 2002), which is commonly associated with species-poor, harsh environments,
or species found in early succession (Magurran 2004). Arctic plant communities fit the
geometric series model because o f their paucity compared to temperate ecosystems; only 4%
o f known vascular plants are found in the arctic (Chapin and Komer 1994). In contrast our
mycorrhizal fungal communities, including the dolomitic site, have a log normal distribution,
which is usually associated with more diverse ecosystems such as temperate and tropical
forests (McKane et al. 2002), and found in ectomycorrhizal fungal communities o f plant
communities with more extensive root systems than arctic shrubs such as Douglas-fir, bishop
pine (Pinus muricata), Arctostaphylos spp., ponderosa pine {Pinusponderosa), Norway
spruce (Picea abies), and western hemlock {Tsuga heterophylla) (Horton and Bruns 2001).
The log normal model has been criticized for possibly being a statistical artifact
without any biological meaning (Tokeshi 1999). Further testing based upon Tokeshi’s
analyses indicated that the power fraction model best explained the distribution o f genotypes

76

in the lowland and granitic sites. Although Tokeshi’s models often discuss the mechanisms
by which a niche can be fragmented for different species, these models can also be used to
describe how a niche can be filled (Tokeshi 1999, Magurran 2004). The latter fits better with
mycorrhizal fungi because once a niche is occupied by a mycorrhizal fungus, another fungus
is challenged to outcompete and take the existing mycorrhizal fungus, which would be
difficult due to the intimate contact o f the Hartig net and mantle. This is not necessarily true
for eetendomycorrhizas where ectomycorrhizas can outcompete established
ectendomycorrhizas (Wilcox 1971) or DSE, which have little or no mantle; niche
fragmentation in these systems may be more easily visualized. In the power fraction model,
abundant species are more likely to successfully invade or inhabit niche space (Tokeshi
1999) or in this case colonize a segment o f root. This may help explain the patchiness o f
species found on root tips, where 1-2 dominant types are found in one area but may not be
dominant or present a few centimeters away (Horton and Bruns 2001). These clusters of
fungi may outcompete other fungi and dominate a certain niche. When examining the
increase in cumulative frequency o f genotypes due to warming, the dominant genotypes, for
the most part, were the ones that increased in number, more so than the rare or subdominant
genotypes. However, this is probably not the whole picture as niche allocation for dominant
species may be different than for rare and subdominant species. Tokeshi’s composite model
takes into account more than one niche allocation model because the assemblage o f species
probably requires more than one process (Tokeshi 1999). The random assortment model
may be more appropriate to explain the resource use o f subdominant and rare species. In this
model, invading species exploit resources not used by existing species (Tokeshi 1999).

77

The broken stick model best fits the dolomitic sites. The broken stick model is
usually based on one resource, evenly divided among species (Magurran 1988). This may be
applied to the dolomitic sites because this site is limiting to both the plant and fungal
community as seen by its limited diversity compared to the other sites. Those fungi that
tolerate the limitations of this site may be sharing rather than competing for what little
resources are available.
Applying niche apportionment concepts to mycorrhizal fungal communities is
difficult because the availability o f a niche may be determined by the host plant. Hoeksma
and Kummel (2003) suggested that plants may increase mortality o f root segments that are
colonized by mycorrhizal fungi that are consuming too much carbon from the plant. Another
possibility would be that this niche has yet to be filled; more fungi may colonize these roots.
Perhaps because the conditions are harsher on the dolomitic site, colonization is retarded in
comparison to the other two sites.
Magurran and Henderson (2003) suggest that communities can be divided into two
groups, core genotypes and occasional genotypes. Core genotypes are persistent and
abundant while occasional genotypes are infrequent and low in abundance. Most o f the
dominant genotypes remained so in both ambient and warmed plots and can be classified as
core genotypes. That we were able to distinguish between core and occasional genotypes,
and the persistence o f the dominant genotypes, may indicate that these few dominant
genotypes contribute the most to ecosystem function, a hypothesis suggested by Walker et al.
(1999). These dominant genotypes, whose functions are assumed to be very different from
each other, may maintain ecosystem functioning (Walker et al. 1999), which may explain the
lack o f changes in soil properties between the warmed and ambient conditions. They also

78

suggest that the many oeeasional genotypes may be functionally similar to the few dominant
genotypes, so if the dominant genotype is lost, then hypothetically, the abundance o f an
occasional genotype will increase to fill the void, thus adding to the resilience o f the
ecosystem (Walker et al. 1999). This may be an important factor for these sites. The
lowland and highland granitic sites may prove to be more resistant to warming because o f
their high diversity, while the dolomitic site, with its low genotypic richness, may be less
resistant to change.
T-RFLPs proved to be a useful tool for analyzing root-associated fungal communities
by providing a community fingerprint for the different sites and treatments. It allowed us to
analyze multiple fungi that colonize single root tips, and to distinguish morphologically
indistinct tips. However, this technique was not without problems.
One o f the difficulties was interpretation o f the four datasets that were produced;
results with different enzyme-primer combinations did not always concur. The differences
between the primer-enzyme combinations may be because the restriction site for one o f the
enzymes may be in a more variable region, which may explain why the forward primer
would detect more rare genotypes in the warmed plots than the reverse primer. YISA-Alul
had the fewest genotypes compared to the other three combinations, which suggests that this
combination cuts in a more conserved area so detection o f genotypes is less discriminating.
In addition, the dye used to label the reverse primer, Cy 5.0, had a weaker reading than the
Cy 5.5 dye, which labeled the forward primer. More bands labeled with the Cy 5.0 dye may
have been too weak to be detected. The combination o f the weaker dye and the more
conserved region may explain why the reverse primers, and especially Alul, showed fewer
genotypes than the other three primer-enzyme combinations.

79

Another factor affecting the reliability o f T-RPLP analysis is the generation o f
pseudo-fragments, PCR artifacts where secondary structures are formed from single stranded
DNA. Single stranded DNA may be created by intrastrand annealing (Egbert and Friedrich
2003). These single strands may form palindromic secondary structures so that the
restriction enzyme would still have the required double strand to cut (Egbert and Friedrich
2003). Creation o f these secondary structures is more common for some restriction enzymes
than others (Egbert and Friedrich 2003), which may be the ease with Hinii. More fragments
were generated from the //m fl enzyme than Alul, which could have resulted in more
variation and noise than the Alul digestions. Klamer et al. (2002) also found i/m fl to
produce more fragments than the other endonucleases that he used {Taql and Msel).
Finally, genotype frequency classifications may be misleading for mycorrhizal fungal
communities because mycorrhizal root tips tend to be clustered, so sampling may miss or
overestimate genotypes. For example, there were some dominant fungi that were found only
on 1-2 plants from the same plot (e.g., a genotype found 18 times from one plant), so they
may appear to be dominant genotypes but are not distributed among all o f the plots.
How global warming will affect the root-associated fungal community on Salix
arctica is difficult to predict. These communities are relatively stable as indicated by the
lack o f change in genotype richness, but if the trend towards increasing frequency of
dominant genotypes continues, then changes in the root-associated community may become
more apparent. Because the soil properties did not change much for the granitic and lowland
sites, these communities may remain more stable. The dolomitic site will likely experience a
decrease in plants due to its increase in C:N ratio and possibly reduced available phosphorus.

80

Because of the changes in these conditions, this site may not increase in plant or fungal
richness.

81

III. How host plant, warming, and site affect the eulturahle root-associated fungal
community from the Canadian High Arctic

A. Introduction
Fungi isolated from within root tips represent a different community than one
determined using PCR-based methods. Cultures tend to favor fast-growing saprotrophs or
endophytes over mycorrhizal fungi, and those species isolated from roots were rarely found
in PCR-based (e.g. RFLPs, sequencing, T-RFLPs) studies (Kernaghan et al. 2003). Allen et
al. (2003) also found different communities based on direct PCR and isolating fungi from
cricoid plant roots, noting that Sebacina was the dominant fungus with direct PCR but was
not found in any o f their cultures. Part o f the problem with culturing is that complete
information on the nutritional requirements for ectomycorrhizal fungi is not known, so
isolating ectomycorrhizal fungi is very difficult, and success is low. With surface sterilization
o f the root tips the endophytic fungal community seems to flourish when cultured.
Endophytic hyphae may not be as abundant as ectomycorrhizal fungi, and as a result they are
not easily amplified. Although ectomycorrhizal fungi can be classified as endophytic fungi,
the distinction used here is that ‘endophytic fungi’ do not have a mantle and Flartig net while
‘ectomycorrhizal fungi’ do.
Dark-septate endophytes (DSE) are ubiquitous in both arctic and alpine systems
(Bledsoe et al. 1990 Gardes and Dalberg 1996, Cazares 1992) and are found associated with
ecto-, ericoid, and non-mycorrhizal plants (Jumpponen 1999). Hyaline septate hyphae have
been reported and found on several plants and may have the same ambiguous function as
DSE o f being either pathogenic or mutualistic (Jumpponen and Trappe 1998). Because
arbuscular mycorrhizae (AM) are scarce in the arctic, in contrast to alpine systems,
82

(Haselwandter 1987, Bledsoe et al. 1990), DSE may replace the function o f AM fungi
(Bledsoe et al. 1990). In addition, Phialocephala fortinii Wang & Wilcox, a commonly
found DSE from arctic and alpine environments, is thought to transport carbohydrates
(Jumpponen 1999), a function demonstrated by mycorrhizal fungal linkages (Simard et al.
1997).
The objective o f this study was to test for the impacts o f site, host, and warming on
the root endophytic fungal community using culture-dependent techniques combined with
morphological and molecular identification. Molecular identification included grouping taxa
according to RFLP patterns and DNA sequencing. Dissimilarity matrices, which were used
for clustering the RFLP patterns, were also the basis o f a method to ordinate the data using
the Phi index, which is a diversity index.

B. Materials and methods

1. Field collection
To assess the root endophytic fungal community found on these roots, two specimens
each of Dryas integrifolia Vahl, Salix arctica Pall., Cassiope tetragona (L.) D.Don, and
Saxifraga oppositifolia L. from each open top chamber (OTC) and ambient plot were
destructively harvested in late July, 2000 along with surrounding soil and then placed in
Ziploe ® bags. To increase the sample size for identification by sequencing or culture
morphology, extra specimens o f C. tetragona and D. integrifolia were harvested from the
designated plots; these samples were not used to statistically test for differences according to
treatment or plot. The samples were stored in a permafrost cooler while in the field and
placed in a 4° C walk-in cooler at University o f Northern British Columbia until they were
processed.
83

2. Sampling from roots
Soil was loosened from plant roots by immersing in water for at least 24 hours at 4°
C. Roots were gently cleaned with water and were collected in a 0.5 mm sieve (No. 35 USA
standard testing sieve, W.S. Tyler, Inc.). Segments o f roots were randomly selected by
placing the root system on a numbered grid. Root segments that intersected a randomly
selected numbered grid were sampled. Fungi were isolated from five o f these randomly
selected root tips. From each root section, 3-6 mm was excised, surface-sterilized with 30%
H 2O 2 following Danielson (1984), and placed on modified Melin Norkrans (MMN) agar
(Marx 1969) with streptomycin sulphate and chlorotetracycline added to minimize bacterial
growth. The protocol o f Monréal et al. (1999) was used for root tips that were from the
ericaceous plant Cassiope tetragona. These tips were plated on 1/3 concentration potato
dextrose agar. All plates were incubated at room temperature (20-22° C). Fungi were
purified by re-isolation from each colony onto fresh agar plates. Once these re-isolated
cultures grew, a 0.5 mm plug was taken for DNA extraction and frozen at -20° C. Two more
plugs were placed on separate agar slants for storage at 4° C until being sent to Dr. R. Currah
at the University o f Alberta for identification based on morphology.

3. DNA extraction and ITS-RFLP analysis
DNA from cultures and sporocarps, and plant tissue from stem and leaves, was
extracted using the CTAB miniprep protocol o f Gardes and Bruns (1996b) except samples
were not freeze-thawed prior to crushing. All DNA samples were reconstituted in 100 pL o f
TE-8 buffer.
The ITS region o f the nrDNA was amplified in a 35 pL reaction containing I OX
buffer, 2X dNTPs, 25 mM M gCb, 0.45 pM o f the primer pair ITSI/ITS4 , and 2.5 U of
Platinum Taq (Invitrogen). PCRs were done on MJ Research PTC-100 thermocyelers with
84

the following program: 94 °C (30 sec); [93 °C (35 sec); 50° C (53 sec); 72° C (30 + 5
sec/cycle) x 34 cycles]; 72°C (5 min). PCR amplification was confirmed on a 0.7 % agarose
gel. Unsuccessful amplifications were redone with 1:50 dilutions o f the DNA template.
PCR products were digested with the endonucleases Alul, Hinü (Invitrogen), and
MboW (Fisher), as suggested by manufacturers, using 8 pL o f the PCR produet in a 15 pL
reaction with 10 U o f enzyme/pL for 5 hrs at 37° C. Digested products were separated on a
1% low-melting point agarose/1.5% agarose lOX TBE gel containing ethidium bromide and
photographed using a Biophotonics 2000i imaging system (BioCan Scientific). RFLP
patterns and fragment sizes were determined using Gene Profiler version 4.02 (Scanalytics)
with a 1 KB ladder standard (Invitrogen).

4. Data analysis
Cluster analysis using the “Neighbor” algorithm in PHYLIP (Felsenstein 2004) was
performed using Gene Profiler based upon the total number o f polymorphic bands as the
distance measure. Based on the resulting tree (not shown), samples representing commonly
occurring RFLP groups were selected for sequencing.
In order to ordinate treatments without subjective assignment o f isolates into taxon
groups, we developed an ordination method using a distance measure based upon the Phi
index (Mah et al. 2001). To calculate the Phi index for each plot, a pairwise distance matrix
based on Dice’s index [(2 x the number o f common bands)/ (2 x the number o f common
bands + number o f polymorphic bands)], was generated for eaeh plot database using Gene
Profiler, with a 5% match tolerance to account for inter-gel variation in band sizes. From this
distance matrix, the Phi Index, was ealculated by squaring each distance, summing the
distances for each sample, and dividing the summation by n-1. The totals o f the different

85

distances for each sample were then summed and divided by n (Mah et al. 2001, Khetmalas
et al. 2002):

S
7=1

[1]

o = -

/=]___

n -\
n

A Phi distance was calculated for pairs o f treatments by combining the two databases,
then recalculating the Phi index o f the merged database. If the two merged databases were
identical, then the minimum possible Phi distance value can be estimated as a weighted
average o f the two individual Phi index scores:
[2]

Min„,« =

_
l«l + «2 ) - («1 + «2 )

If the two merged databases shared no RFLP fragments, then the maximum Phi index value
can be estimated as

[3]

.
_ [Phly (», - « ,) ] + [Phlj («2 - «2 )] + 2(«, * «2 )
Max(i,2) -------------------- r------------------------ r----------------- .
l«l + «2 j - («1 + «2 )

The Phi index o f the merged database must fall between the minimum and maximum values
estimated from the Phi index values o f the two unmerged databases, therefore a standardized
Phi distance can then be calculated as:

[4]

Phi dist(i,2)

Phi^ 2 - Mm, 2

MaXj 2 - M w ,

.

The Phi distances were placed in a distance matrix, and a non-metric multidimensional
scaling (NMS) ordination was run with a starting configuration using the standard GuttmanLingoes and 50 iterations in Statistica vers 6.1 (StatSoft).
86

Nested, two-way analysis o f variance (ANOVA) in Statistica was used to test for the
impacts of site and treatment on species richness and cumulative frequency. The following
model was used for host plants that had successful isolation from all plots:
yijk = p + Si + Tj + Rk(j) + Si*Tj + 8ijk(j), where
yijk =species richness or cumulative frequency
Si = effect o f site, i = 1,2,3;
Tj = effect o f treatment (OTC or control), j = 1,2;
Rk = replication effect, k = 1,2,3;
A nested, three-way ANOVA was conducted to test for differences in the cumulative
frequency for P. fortinii that were identified according to morphology. Although
morphological identification was done on all isolates, including replicates made for storage,
for this analysis only one representative o f eaeh sample was included. The model used for
the analysis was:
yijki = p + Pi + Sj + Tk + Ri(k) +?i*Sj*Tk + Sj*Tk + 8ijki(k), where
yijki = cumulative frequency o f P. fortinii.
Pi = effect o f host plant, i = 1, 2, 3, 4, and the other variables are as in the above model.

5. DNA sequencing
Sequencing was done following the manufacturer’s instructions for the Thermo
Sequenase Cycle Sequencing Core Kit (Amersham Life Science). Samples were first
amplified with primer pair ITSl (White et al. 1990) and NL8 (Egger 1995) with the program:
94° C (35 sec); [93° C (35 sec); 54° C (53 sec); 72° C (2 min 10 sec) x 35 cycles]; 72 °C (3
min), and amplification was confirmed on 0.7% agarose gels. Amplicons were purified with
Qiaquick PCR purification kit (Qiagen) according to manufacturer’s instructions. Cycle

87

sequencing was done by adding PCR product to 3 U o f Thermo Sequenase, 0.34 pmol/pL o f
Cy 5.0 labeled primer (for forward primers) or Cy 5.5 labeled primer (for reverse primers),
lOX sequencing buffer diluted to one-tenth the volume, 15% dimethylsulfoxide, sterile
dH20, and 0.1 mM o f one the four ddNTPs; cycle sequencing was completed for each o f the
four ddNTPs for every sample. Cycle sequencing used the following program: 94° C (2
min); [94° C (20 sec); annealing temp (45 sec); 70° C (varied) x 25 cycles]; [94° C (25 sec);
70° C (2 min) x 15 cycles]; 72° C (6 min). Annealing temperature and the first extension
time varied according to the primer used: ITSl (52-54° C, 75 sec); ITS4 (54° C, 1 min); NL5
(50° C, 70 sec); NL6c (52° C, 75 sec); NL7 (50 24° C, 70 sec); and NL8 (52° C).
Samples were run on an OpenGene System Long Tower Sequencer (Bayer
International). Three microliters o f PCR products were added to 4.5 pL o f formamide
loading dye, denatured between 70 to 80° C for 2 minutes, quenched on ice, and loaded onto
a 6% polyacrylamide gel. Samples ran for 50 min. with the temperature on sequencing
control set at 50° C, gel voltage at 1250V, and laser power for 50%.
Sequences were edited in Sequencher (Gene Codes, corp). These edited sequences
were compared in GenBank (BLAST search <www.ncbi.nlm.nih.gov>) to identify fungi that
were most closely related to the sequence and to ensure that sequences were fungal. Edited
sequences were imported into Mac Vector (Oxford Molecular) and were aligned with the
most closely related sequences downloaded from GenBank. Sequences downloaded included
Epicoccum nigrum AF455455, Colispora elongata AY148102 Leaf litter Ascomycete
AF502763, Epacris microphylla root associate AY268216, Dothideales sp. AY465446,
Fungal endophyte MUT 2723 AF373055, Fungal endophyte MUT 585 AF373051,
Ascomycete sp. AJ279473, Phoma glomerata AY618248 (Fig. 3.4); Zalerion varium

88

A F169303, Glarea lozoyensis A F169304, Hymenoscyphus epipiphyllus AY348581,
Hymenoscyphus monotropae AF 169309, Lachnum bicolor AJ430394, Phialocephala virens
AF486132, Xenochalara juniperi AF 184889, Phialophora sp. AY465463, Cadophora sp.
AY371512, Phialophora finlandia AF486119, axenic ericoid root isolate AJ430119,
ectomycorrhiza cf. Hymenoscyphus ericae (= Rhizoscyphus ericae (Zhang and Zhuang
2004)) AJ430150, Hymenoscyphus sp. AY354244, ectomycorrhizal isolate (Helotiales)
AJ430410, Cadophora luteo-olivacea AY249069, Leptodontidium orchidicola A F214578,
Cadophora gregata AY249071, Cadophora malorum AY249063, Cadophora sp.
AY371506, ectomycorrhiza (cf. Phialocephala fortinii) AJ430214, Phialocephala fortinii
crypt sp. AY347405, Phialocephala fortinii AY078138 (Fig 3.5); Inocybe angustispora
AY380360, Favolaschia cf. sprucei AF261420, Favolaschia cf. calocera AF261419,
Panellus serotinus A F518633, Ripartites metrodii AF042012, Mycena haematopoda
AJ406590, Mycena leaiana AF261411, Omphalina rivulicola U66451, Collybia tuberosa
AY639884, Clitocybe subvelosa AY647208 (Fig. 3.6); Phlebia lindtneri AF141623,
Ceratobasidium goodyerae-repentis AY243523, Uthatobasidium fusisporum A F518664,
Thanatephorus cucumeris AF354062, Ceratobasidium sp. AGO AF354094, Ceratobasidium
sp. AGL AF354093, Inonotus weirii AY059040, Phlebiopsis gigantea AF141634 (Fig 3.7);
Alternaria helianthi A Y \5 4 7 \3 , Ampelomyces sp. AY293794, Ascomycota sp. AJ301960,
Byssoascus striatosporus AB040688, Cadophora luteo-olivacea AY249087, Cadophora
malorum AY249086, Cenococcum geophilum A Y l 12935, Chalara fungorum AF222462,
Chalara kendrickii AF222464, Chalara longispes AF222466, Chalara microchona
AF222468, Chalaraparvispora AF222473, Chalara strobilina AF222477, Cryptosporiopsis
sp. AY442321, Cryptosporiopsis ericacea AY442323, Cudonia lutea AF433139, ericoid

89

mycorrhizal sp PPO-7 AY599245, ericoid mycorrhizal sp PPO-8 AY599246, ericoid
mycorrhizal sp. AY599244, ericoid mycorrhizal sp. PPO-5 AY599243, ericoid mycorrhizal
sp. PPO-2 AY599240, euascomycete RFLP type A A F127116, euascomycete RFLP type B
AF 127117, euaseomycete RFLP type C AF 127118, Fabrella tsugae AF356694, Graphium
rubrum AY266313, Hymenoscyphus ericae (= Rhizoscyphus ericae (Zhang and Zhuang
2004)) AY284122, Hymenoscyphus sp. UBC tra 1436 AY219881, lodosphaeria sp.
AF452045, Lachnum cf. bicolor AY544674, Lachnum virgineum AY544646, Lecythophora
sp AF353607, Leptosphaeria doliolum U43475, mycorrhizal sp. AF081443,
Mycosphaerella mycopappi U43480, Oidiodendron tenuissimum AB040706, Phialocephala
fortinii AF326082, Phialocephala dimorphospora AF326081, Phialophora gregata
AF222502, Phialophora sp. AF 156922, Phoma herbarum AY293788, Phoma sp. 199
AY293785, Pleomassaria siparia AY004341, Pleospora herbarum U43476, Rhytisma
acerinum AF356696, Setomelanomma holmii AF525678, Shiraia bambusicola AB105798,
Trematosphaeria heterospora AY016369 (Fig 3.8).
Neighbor-joining analysis was used for all trees, with 1000 repetitions for bootstrap
values. The Kimura 2-parameter model was used for the distance measure, the
transitiomtransversion ratio was estimated, and gaps were distributed proportionally.
Bootstrap values greater than 50 were reported. Identifications from samples that were
sequenced were applied to those samples with matching RFLP patterns. Those that clustered
according to RFLP data were verified by viewing composite gels o f likely samples in Gene
Profiler database, vers. 4.02 (Scanalytic, Inc.). These samples were then compared to those
identified based on morphological traits.

90

c. Results
1. ITS-RFLPs
Out o f 720 isolated cultures, 432 were amplified successfully and digested with all
three restriction enzymes for RFLP analysis. For statistical analyses, 30 o f the isolates were
deleted for a total o f 402. These isolates were not included for statistical analyses because
they were from plants that were not sampled within the experimental design (i.e. they were
sequenced to increase the chances o f identification by providing more isolates from the
harsher highland sites). The species cumulative frequency for each plot is illustrated in Table
3.1. Isolating fungi from root tips was most successful for Salix arctica and Saxifraga
oppositifolia. The average number o f isolates was highest from S. arctica growing in control
plots at the granitic site (p=0.02) (Fig 3.1). Although not statistically significant, more
isolates from S. oppositifolia were also recovered from the granitic site, averaging 15 isolates
from both plant specimens for the granitic site compared to six for the other two (Fig. 3.2).
Fungal isolation was particularly unsuccessful for Cassiope tetragona from the warmed plots
for all three sites, as well as for the control from the dolomitic site; fungi were only
successfully isolated from one plot on the granitic site. In contrast to Salix arctica and
Saxifraga oppositifolia, no fungi were successfully isolated from roots o f Dryas integrifolia
from the granitic site.
Results from NMS showed that the site was the most influential factor in
differentiating the root endophytic fungal community, more so than warming or host plant
(Fig 3.3). The lowland site had a wider range o f variation when compared to the other two
sites. The dolomitic site also had large variation, but the isolates were more similar at this

91

Table 3.1 Number of successful samp es isolated from root tips based on R "LPs.
OTC
Control
Treatment
Replicate
Replicate
2
Sum
1
Sum
2
3
1
3
(cumulative
(cumulative
Site
frequency)
frequency)
Lowland
2
C. tetragona
0
0
8
8
8
6
16
21
4
6
7
17
10
5
6
D. integrifolia
10
4
23
4
Salix arctica
9
5
6
15
Saxifraga
5
8
3
16
10
1
6
17
oppositifolia
24
22
64
27
24
Sum (species
18
18
69
abundance)
Dolomitic
C. tetragona
0
0
0
0
0
0
0
0
D. integrifolia
5
3
5
13
2
8
5
15
2
4
Salix arctica
7
8
17
4
7
15
1
12
Saxifraga
3
8
4
2
17
23
oppositifolia
21
42
Sum (species
15
6
10
14
29
53
abundance)
Granitic
0
0
C. tetragona
0
0
10
0
0
10
D. integrifolia
0
0
0
0
0
0
0
0
Salix arctica
17
6
5
24
12
28
16
52
11
Saxifraga
16
16
43
8
13
20
40
oppositifolia
Sum (species
17
21
71
42
33
36
25
103
abundance)

(a) Salix arctica

(b) Saxifraga oppositifolia

p = 0 .0 1 7

20.00 -

A 15-00

□ Control

10.00

a O TC

Lowland

DoionntK

Granitic

□ C ontrol

10.00

OTC

Lowland

Dolomitic

Granitic

Site

Fig. 3.1 Species cumulative frequency for Salix arctica and Saxifraga oppositifolia on
different sites. LSM ± SE
92

(a) Salix arctica

p=0.02

12.00

(b) Saxifraga oppositifolia
14.00

10.00

12.00

1.00

10.00

6.00
4.00 2.00

□ C ontrol

8.00 -

□ C ontrol

HOTC

6.00

HOTC

2.00 ’

0.00

0.00
Lowland

D obm itic

Site

Granitic

Lowland

Dolomitic

Granitic

Site

Fig. 3.2 Species richness for Salix arctica and Saxifraga oppositifolia on different sites.
LSM ± SE.

site than in the community from the lowland site. The genotypes within the granitie site
clustered tightly together in the ordination and appeared to have a similar composition.
Species richness ranged from 0-13 for Cassiope tetragona, 0-18 for Dryas integrifolia, 11-27
for Salix arctica, and 10-31 for Saxifraga oppositifolia (Table 3.2) for eaeh o f the two
replicate plants for each treatment. Beeause fungi were not recovered from all o f the sites for
Cassiope tetragona and Dryas integrifolia, ANOVAs were only done for Salix arctica and
Saxifraga oppositifolia. There were significantly more different types o f cultures isolated
from Salix arctica in the granitic site (p=0.02) (Table 3.2) than the other two sites. Although
not statistically significant, more diversity in fungal species was found (Table 3.1, Table 3.3)
in the granitic control than the other plots for Salix arctica. For Saxifraga oppositifolia,
diversity was greatest (p=0.017) in the granitic site, and a trend showing higher cumulative
frequency in the granitic site was found as well. None o f the plant tissues amplified with the
primer pair ITS 1/4.

93

Final Configuration, dim ension 1 v s. dim ension 2
2.0

3cb

0.5

Icb

30 #
4cd

0.0

3cd
iOb

lOb

3cc
Icc

lcd

•0.5

lOd

10c
-

1.0

-1.5
-1.5

30c

-

1.0

-0.5

0.0

0.5

1.0

1.5

2.0

2.5

Dim ension 1

Fig. 3.3 NMS for host plant and warming treatments per site based on phi distances. The
first number represents the site where 1 = lowland, 3 = highland dolomitic, and 4 = highland
granitic; the 2nd letter is e for control or O for OTC, and the last letter represents the plant, a
= Cassiope tetragona, b = Dryas integrifolia, c = Salix arctica, and d = Saxifraga
oppositifolia.

94

Table 3.2 Species richness o f cultures according to host plant, site, and treatment based on
RFLPs
OTC
Control
1
2
Treatment
3
Total for 1
2
3
Total for
species
species
Site
Lowland
C. tetragona
0
0
6
6
1
6
6
13
2
D. integrifolia
5
7
16
8
5
5
18
5
7
3
15
4
4
S. arctica
5
13
3
S. oppositifolia 4
3
10
10
1
5
16
Dolomitic
C. tetragona
0
0
0
0
0
0
0
0
4
3
5
13
2
D. integrifolia
6
5
13
5
2
7
16
4
2
S. arctica
5
11
1
S. oppositifolia 3
6
10
4
1
10
15
Granitie
0
0
0
0
C. tetragona
9
0
0
9
0
0
0
0
D. integrifolia
0
0
0
0
S. arctica
10
6
5
21
11
10
6
27
9
11
31
S. oppositifolia 11
8
7
7
22

Table 3.3 Species richness of cultures according to host plant, site, and treatment based on

Treatment
Site
Lowland
C tetragona
D. integrifolia
S. arctica
S. oppositifolia
Dolomitic
C tetragona
D. integrifolia
S. arctica
S. oppositifolia
Granitic
C. tetragona
D. integrifolia
S. arctica
S. oppositifolia

1

2

OTC
3
Total for
species

0
3
3
8

0
6
4
10

1
1
3
5

1
10
10
23

0
3
5
9

4
4
2
2

3
5
5
10

7
13
12
21

0
4
3
3

0
2
1
3

0
4
3
4

0
10
7
10

0
0
2
5

0
3
4
2

0
4
1
6

0
6
7
12

0
0
8
12

0
0
7
8

0
0
6
8

0
0
21
28

6
0
8
8

0
0
8
14

0
0
7
8

6
0
23
30

1

2

Control
3
Total for
species

95

2. DNA sequencing
O f the 50 samples that were sequenced, 43 produced high quality results for analysis.
Approximately 330 bp were used for phylogenetic analysis o f the ITS region (ITS) and 350
bp o f the nuclear large subunit rRNA gene (LSU). Five trees were formed based on primer
pairs and alignment consensus, which were the ITS and the Dothideales; ITS and the
Helotiales; LSU and the Ceratobasidiomycetes; LSU and the Agaricales; and LSU and
aseomycetes (Figs 3.4-3.8 respectively). Sequencing was able to classify 37.9% o f the
samples to at least Order.
The ITS - Dothideales tree had two unknown samples (see Fig. 3.4). One o f the
unknown samples was affiliated with Colispora elongata although there was <50% bootstrap
support, and the other unknown sample was nested in the Dothideales.

E picoccum nigrum
1-S3-C -D rin 5.2
C olispora elongata
l-S l-C 3 -S a o p 4 .1
81

Epacris m icrophylla root assoc
L e a f litter A sco m y cete

87

D oth id eales sp.
97

99

I Fungal endophyte M U T 2723
Fungal endophyte M U T 585

A sc o m y ce te sp.
Phom a glom erata

l_om

1

Fig. 3.4 Neighbor-joining best tree based on ITS and the Dothideales. Bootstrap values
(1000 replications) >75 are reported.

96

All unknown cultures from the ITS - Helotiales (Fig. 3.5) were found to be affiliated
with Cadophora, Phialocephala fortinii, or Hymenoscyphus. One sample was assoeiated
with species of Cadophora with a bootstrap value o f 100%; three cultures were clustered
with P. fortinii supported by a bootstrap value o f 81%; and two samples were affiliated with
Hymenoscyphus even though the bootstrap value was <50%.
Cultures affiliated with basidiomyeetes were associated with the Agaricales, whieh
had <50% bootstrap support, and with Ceratobasidium (77% bootstrap value) and (Fig. 3.6
and 3.7 respeetively). From the LSU -aseomyeete tree (Fig. 3.8), four eultures were
affiliated
with Cryptosporiopsis (99% bootstrap value); five cultures with Mycosphaerella (85%
bootstrap); three with Phoma (84% bootstrap); two with Cadophora (88% bootstrap); three
with Hymenoscyphus (90% bootstrap); and four with Phialocephala (95% bootstrap).
Identities o f ten eultures were not resolved from this analysis. Not all samples were resolved
to genus and are identified by their family names. One sample could not be resolved even to
order and was left as an unknown.

3. Culture morphology
Culture morphology was based on 1347 eultures (usually two tubes for eaeh isolate).
Forty pereent o f the cultures were identified to genus, whieh included Acremonium,
Cryptosporiopsis, Geomyces, Leptodontidium, Monodietys, Phialocephala fortinii,
Scytalidium, Sebacina, Staphylotrichum, Trichocladium, Trichoderma, Trichosporiella, or
Xenosporium. O f these, Phialocephala fortinii was the most abundant taxon isolated,
accounting for 85.7% o f the identified eultures. O f the unknowns, cultures with hyaline
hyphae were the most abundant (23%) followed by other isolates with dark hyphae other than

97

Xenochalara juniperi
Cadophora malorum
" Cadophora luteo-olivacea

100

Cadophora

1-S4-0TC l-Saop4.5
Cadophora gregata
Cadophora sp.
Leptondontidium orchidicola

_&L

2-S4-Cl-Saar4.2

100

2-S4-C t-Saar 5.9 {P h ia lo c e p h a la fo r tin ii
2-Sl-Cl-Sær 2.3 {Phialocephala fortinii )
ectomycorrhizal root (cf. P. fortinii)

Phialocephala fortinii

r l-Sl-C3Saar 1.3
|| Phialocephala fortinii crypt sp
Phialocephala fortinii
Zalerion varium
Laehnum bicolor
Glarea lozoyensis
Hymenoscyphus epiphyllus
74

H yy menoscyphus monotropae
1-Sl-C3-Saar 1.2

Hymenoscyphus

2-S4-OTCl-Saar 1.9
Phialophora finlandia
100 1 Axenic ericoid root isolate
Ecto cf. Hymenoscyphus ericae (= Rhizoscyphus ericae)
Eeto isolate (Helotiales)
Phialocephala virens
Phialophora sp.
0.02

Fig. 3.5 Neighbor-joining best tree based on ITS and the Helotiales. Bootstrap values (1000
replications) >75 are reported. Names in parenthesis after samples are identifications based
on morphology.

98

In ocyb e angustispora
I F avolasch ia cf. sprucei
F avolasch ia cf. calocera
I

Panellus serotinus
Ripartites m etrodii

M ycen a haem atopoda
M ycen a leaiana
1 -S l-C l-D r in 1.6
Orphalina rivu licola
~

C ollyb ia tuberosa

C litocyb e su b velosa
0.02

d

Fig. 3.6 Neighbor-joining best tree based on LSU and the Agaricales.

Phlebia lindtneri
99

C eratobasidium goodyerae-repentis
U thatobasidium fusisporum

99
8 9 C eratobasidium sp. A G O
S9

94

77

Ceratobasidium sp. A G L
Thanatephorus cucum eris
l-S l-O T C 2 -S a a r 2 .1

Inonotus w eirii
P hleb iop sis gigantea

Fig. 3.7 Neighbor-joining best tree based on LSU and the Ceratobasidiomycetes. Bootstrap
values (1000 replications) >75 are reported.

99

--------------------- 2-S4-C 3-Saar2.1
Chalara parvispora
Cudonialutea
Cryptosporiopsis ericacea
A scom ycota sp.
Cryptosporiopsis sp
Helotiales
^
99 l - S 4 - C 2 - S a o p 3.3
2 -S l-C l-S a o p 4.1
I l-S 4 -C l-S a o p 4.5
l-S4-C 2-Saop 4.5
Ericoid mycorrhizal sp PPO-8
Ericoid mycorrhizal sp PPO-7
Cenococcum geophilum
T rematosphaeria heterospora
90
lodosphaeria sp.

98

_|

Pleomassaria siparia

a s I h ^ y c o sn h a e re lla m v c o n a n n i

Euascomycete RFLP type B
Phialophora sp.
/ 1-S4-C l-C ate 5.8
1-Sl-C 2-D rin 2.2 (P h ialoceph alafortin ii)
96
1| l-S 4-O T C l-S aop 4.1 {P h ialoceph alafortim i)

Phialocephala

U Phialocephala fortinii
1-S I-C l-S aar 4.4 {P h ialoceph alafortim i)
1-S I-C l-C a te 2.3
Lachnum cf. bicolor
Mycorrhizal sp.

aD2i

Fig. 3.8 Neighbor-joining tree based on LSU and ascomycetes. Species in parentheses are based on morphology. Bootstrap values
(1000 replications) >75 are reported.

100

p. fortinii (15%).
Success o f isolation o f Phialocephala fortinii differed according to host plant species
(p < 0.001), site (p < 0.001), and from which site the host plant was taken (p < 0.001). P.
fortinii was found on Salix arctica more than other host plants (60.5%), but the number o f
successful isolations plummeted for S. arctica on the dolomitic site (Fig 3.9). The
cumulative frequency of isolates for the lowland OTC plots was as high as 34, but only 1-2
isolates were found on the dolomitic plots. With Saxifraga oppositifolia as the host, isolation
o f P. fortinii was most successful from the granitic site (~ 6 isolates), whereas the number of

1

20

□ Cate control
M Cate OTC

I

? 15
V
%
« 10
G
3
Z

S Drin Control

-1

1

S D rinO T C

U

■ Saar OTC
B Saop control

m TTinUi kû
lowland

[D Saar control

dolomitic

M

B Saop OTC

granitic

Site

Fig 3.9 Abundance o f Phialocephala fortinii based on number of morphologically identified
cultures differentiated by host plant and treatment. Cate = Cassiope tetragona, Drin = Dryas
integrifolia, Saar = Salix arctica, and Saop = Saxifraga oppositifolia. OTC = open top
chamber, and control refers to ambient conditions.

isolates was low (less than 2) on both the lowland and dolomitic sites. Isolation o f P. fortinii
from Cassiope tetragona was low from all sites. Isolation from Dryas integrifolia was most

101

successful from the lowland site with 14 isolates from the control and seven from the OTC
plots. Isolating P. fortinii from the two highland sites was not very successful, with only two
from the control and one from the OTC from the dolomitic site and no isolates from the
granitic site.

4. Identifications based on molecular techniques and morphology
When combining sequence and morphological data, 193 (44.7%) o f the cultures
remained unknown or inconclusive, 63 (14.6%) were identified as Phialocephala fortinii, 44
(10.2%) as members o f the Helotiales, 29 (6.7%) closest to Mycosphaerella, 23 (5.3%)
closest to members o f Cryptosporiopsis, 21 (4.9%) closest to Hymenoscyphus, 12 (2.8%)
closest to Phoma, 9 (2.1%) closest to Ceratobasidium, 10 (2.3%) closest to Cadophora, 9
(2.1%) as members o f the Dothideales, 7 (1.6%) closest to Geomyces, 4 (0.9%) as members
o f the Agaricales, 2 (0.5%) closest to Colispora, 2 (0.5%) closest to Trichoderma, and 1 eaeh
(0.2%) closest to Monodictys, Pénicillium, Sebacina, or Trichocladium. Samples whose
morphological identifications were incongruent for the two replicates were considered
inconclusive. Samples identified based on molecular data and corresponding morphological
identification are listed in A3.1-A3.12.
Morphological identification as Phialocephala fortinii matched 34 o f 52 samples
(65.4%) that were shown to be affiliated with P. fortinii from sequence/RFLP analyses
(Table A3.1). An additional five isolates that were affiliated with the Helotiales based on
sequence analysis were morphologically identified as P. fortinii (Table A3.2). When isolates
with RFLP patterns that matched these five Helotiales were included in the P. fortinii
complex, an additional 15 samples were included in the P. fortinii complex. Morphological
identification also suggested that 11 samples that could not be identified from

102

sequence/RFLP analyses were P. fortinii (Table A3.3). Each of these 11 samples had a
unique RFLP pattern. There were seven different RFLP types based on identification by
sequence/RFLP and morphological analyses, which would total 18 different RFLP patterns
for all putatively identified P. fortimi. There were some discrepancies between molecular
and morphological identifications. P. fortinii was identified morphologically for samples
that were affiliated with Mycosphaerella and members in the Dothideales according to the
molecular analysis. The RFLP patterns for these samples matched others in the same cluster
but were identified differently morphologically.
There were also discrepancies between the two identification methods for
Hymenoscyphus, Mycosphaerella, and the Dothideales. One sample, identified
morphologically as Leptodontidium, was affiliated with Hymenoscyphus based on sequence
data, and belonged to an RFLP pattern that had a wide range o f morphological descriptions
(Table A3.4). Another discrepancy included a sample that was identified morphologically as
Phialocephala fortinii, whose RFLP pattern matched two samples that were affiliated with
Hymenoscyphus according to sequence data. In total, four RFLP types were found for this
group.
Twenty-nine samples were affiliated with Mycosphaerella, which formed five
different RFLP patterns (Table A3.5). Two discrepancies between morphological and
molecular techniques were found; one sample was identified as Phialocephala fortinii and
the other Staphylotrichum, both o f which have RFLP patterns matching several samples
associated with Mycosphaerella.
Members that were shown to be affiliated with the Dothideales by the molecular
methods had some incongruity with the morphological identification (Table A3.6). Two

103

samples were morphologically identified as Phialocephala fortinii (Helotiales) and one as
Acremonium, a member o f the Sordariomycetes. For all the isolates that grouped in the
Dothideales, there was only one RFLP pattern, and all the samples were isolated only from
Saxifraga oppositifolia.
Ten samples were affiliated with Phoma according to sequence and RFLP data, with
two different RFLP patterns. Morphologically, Monodictys and Trichocladium were
identified for four o f these samples (Table A3.7). Phoma belongs to the mitosporic
Ascomycetes, while both Monodictys and Trichocladium are teleomorphs that are members
o f the Sordariomycetes. Two o f the Monodictys had the same RFLP pattern as
Trichocladium and other members o f the Phoma clade, while the third Monodictys had a
RFLP pattern that matched the second RFLP type for Phoma.
Four samples were associated with Cryptosporiopsis from sequence analysis, but
when samples with matching RFLP patterns were included, the number o f samples increased
to 23 (Table A3.8). Three o f these samples were supported by morphology. There were two
RFLP types for this clade, which had only a difference o f 50 bp for one o f the fragments in
MboW', all other fragments in the three enzymes matched.
Ten samples were affiliated with Cadophora according to the molecular analysis,
with three RFLP types. Three o f these samples were morphologically identified as P. fortinii
(Table A3.9).
There were only two groups o f cultures associated with Basidiomycetes, the
Agaricales and Ceratobasidium (Tables A3.11-A3.12). All the morphological identifications
were labeled as ‘hyaline sterile’, with the exception o f one o f the replicates, which was
inconclusive.

104

Most o f the samples that were not identified by sequencing/RFLP data remained
unknown based on morphology, but a few samples were identified. Seven unknown samples
that were morphologically identified as Geomyces were probably the same species because
they all had the same RFLP pattern. Although three samples were identified as
Trichocladium, each had a distinct RFLP pattern, which may be an artifact o f the small
number o f samples.

D. Discussion
In this study, site was the most influential factor in distinguishing the root endophytic
fungal community composition, more so than warming or host plant, which was evident from
the NMS analysis. Most of the members consisted o f DSE, o f which Phialocephala fortinii
was the most common. This was similar to the findings o f Stoyke et al. (1992) when they
isolated fungi from 10 different subalpine host plants, but contrary to Kohn and Stasovski
(1990) who reported no DSE from the same study area. In concurrence with previous
research (Kohn and Stasovski 1990, Stoyke et al. 1992), hyaline septate fungi were found in
the present study. Thirty-one different species were found on the granitic - OTC plot based
on RFLP analysis and 30 species with morphological identification, which was comparable
to the species richness o f isolated root endophytes found in temperate grasslands
(Wilberforce et al. 2003).
Differences in species cumulative frequency depended on site and host plant but not
due to passive warming. The dolomitic site had significantly fewer isolates than the other
two sites. This may be in part due to the soil conditions o f this site, where CEC, pH, Ca, and
C:N ratio were higher than from the other two sites (see Chapter 2), the pH substantially
higher, pH 8, versus pH 6 for the other two sites. Interestingly the host plants Salix arctica
105

and Saxifraga oppositifolia yielded the most isolates from the granitie site, but at the lowland
site, Dryas integrifolia and Cassiope tetragona produced the most isolates.. Although the
granitic site had more isolates than the other two sites, it was the most homogenous, as
indicated by the tight clustering found in CCA. This may be because isolates from the
granitic site were from S. arctica and S. oppositifolia, with the one exception o f isolates from
C. tetragona that were found on one o f the plots; whereas, the other sites had isolates from
the other host plants. The high variability from the dolomitic site may be due to the low
number o f isolates that were found from this site.
Even though root tips from Cassiope tetragona were isolated using a protocol that
supposedly favors erieaceous plants, the success rate was low when compared to Salix
arctica and Saxifraga oppositifolia. Fungal isolates from Dryas integrifolia, a reputed
ectomyeorrhizal plant, were also low. The incubation temperature may also have been too
high. Although unlikely, the possibility that isolates from D. integrifolia and C. tetragona
may favor colder conditions than those found on the other host plants is possible.
Species richness tended to be higher for most plots when assessed by RFLP analysis
compared to morphological identification, which was in part because cultures
morphologically identified as Phialocephala fortinii had 18 different RFLP types. Many o f
the cultures that were identified as P. fortinii had unique RFLP types; therefore, the ITS
region for these P. fortinii isolates was highly polymorphic, as were strains from Europe and
North America (Harney et al. 1997, Grünig et al. 2002a). Our results were contrary to
studies that found morphological differences but no or little RFLP differences for the rDNA
ITS region (Addy et al. 2000, Jumpponen 1999). Although Addy et al. (2000) covered a
larger geographic range, we had 70 isolates o f P. fortinii compared to their 33 isolates and 34

106

for Jumpponen (1999), which may partially explain the wide range o f RFLP types in our
study. Our strains may have more characters that were polymorphic as Grünig et al. (2002a)
found with their isolates, which had over five times more polymorphic characters than Addy
et al. (2000).
Phialocephala fortinii is part o f a species complex composed o f cryptic species,
which are species that are morphologically similar but have unique genotypes (Grünig et al.
2004, Piercey et al. 2004). Cryptic species are often found in morphologically asexual fungi
that are genetically isolated (Taylor et al. 1999), which would likely be the case for these P.
fortinii from Ellesmere Island. Grünig et al. (2004) even found different cryptic species on
the same root and clusters of cryptic species that were morphologically indistinct but had
high diversity at the population level. In addition, genetic variation o f P. fortinii increases at
higher latitudes (Ahlich and Sieber 1996, Piercey et al. 2004); which is consistent with the
high number o f RFLP types found for P. fortinii at our site. However, some samples may
have been misidentified, which is particularly true for those with RFLP patterns that matched
other species, and samples with unique RFLP patterns that were not supported by molecular
analyses. This inflated the number o f ribotypes and may be an indication o f phenotypic
plasticity for culture identification.
Most o f the Phialocephala fortinii were isolated from Salix arctica. Although P.
fortinii has been reported to be non-host specific, it was isolated more frequently from S.
arctica and Saxifraga oppositifolia than from Cassiope tetragona and Dryas integrifolia.
This difference may be a combination o f two factors: 1) P. fortinii is found commonly on
roots o f ectomyeorrhizal plants (Stoyke et al. 1992, Jumpponen 1999, Addy et al. 2000,
Grünig et al. 2001), which would explain its absence on the ericoid C. tetragona, and 2)

107

although D. integrifolia is an ectomyeorrhizal plant, the root systems from Salix arctica and
Saxifraga oppositifolia were larger. Also, P. fortinii grew more frequently on the granitie
and lowland sites. The number o f isolates decreased significantly from plants grown on the
dolomitic site, perhaps due to the different soil chemistry properties found on this site.
Surprisingly, isolates from Saxifraga oppositifolia did not follow the same trend as those
from Salix arctica. Even though the abundance o f P. fortinii was comparable from the two
hosts from the granitie site, there were significantly more isolates found on Salix arctica on
the lowland site. Therefore, both host plant and site appear to affect the abundance o f P.
fortinii.
Twenty fungal isolates were potentially affiliated with Cryptosporiopsis. Only three
o f these samples were identified morphologically; the other samples were identified from
RELP patterns. These isolates varied in color, suggesting that morphological differences
could be due to different stages o f development. Verkley et al. (2003) found that
Cryptosporiopsis isolates from Erieaceous plants that grow on MEA and oatmeal agar
change colors at different stages o f development; even though our isolates were grown on
MMN, this change could occur on this medium as well. Also identification could have been
hampered because maeroeonidia are rarely produced in culture (Verkley et al. 2003).
Although Verkley et al. (2003) found that isolation o f Cryptosporiopsis is rare after surface
sterilization, we were successful even though we surface sterilized the roots, as did Ahlich
and Sieber (1996), who isolated C. radicola from Fagus sylvatica, Abies alba, Picea abies,
and Pinus sylvestris. Cryptosporiopsis has teleomorphs in Pezicula and Neofabracea.
However, our sequences were based on the primer pair for the large subunit o f the rRNA
gene and sequences for this region were not available in GenBank for Pezicula and

108

Neofabracea. Therefore, we could not resolve the teleomorph with which our samples were
affiliated.
Five isolates were affiliated with Mycosphaerella with an 85% bootstrap value and an
additional 23 samples were included with matching RFLP patterns. Mycosphaerella has not
been reported from roots in Arctic and Antarctic systems, but Cladosporium, which has a
Mycosphaerella teleomorph (Wirsel et al. 2002) and is described as an oligotrophic,
melanized fungus (Wirsel et al. 2002, Gunde-Cimerman et al. 2003), is commonly found in
the Arctic and Antarctica. In particular C. sphaerospermum, C. herbarum (Gunde-Cimerman
et al. 2003, Bergero et al. 1999), and C. cladosporoides (Widden and Parkinson 1979,
Robinson 2001, Tosi et al. 2002) are found. Cladosporium has been reported to have a wide
range o f functions, from saprotrophic, to epiphytic and endophytic in aboveground plant
tissues, and pathogenic and mycoparasitic (Held et al. 2005, Wirsel et al. 2002, David 1997).
Our samples were probably pathogenic or were soil fungi that were not killed by surface
sterilization, because there have been no reported cases o f root endophytic Cladosporium.
However, this needs to be investigated further.
Other common fungi found in the Arctic include Geomyces and Phoma (Bergero et
al. 1999, Robinson 2001, Tosi et al. 2002). We found seven samples that were
morphologically identified as Geomyces with matching RFLP patterns. Ten samples were
affiliated with Phoma after sequencing and RFLP analyses. Three o f these samples were
morphologically identified as Monodictys. This discrepancy may be due to Phoma having
Monodictys - like conidia. When Grondona et al. (1997) described Pyrenochaeta dolchi,
which belongs to Phoma section Paraphoma, they found this species as having Monodictys like conidia on lateral branches and also stated that the conidia were similar to Phoma cava

109

and Phoma tracheiphla. These fungi were found in soil by previous researchers (Bergero et
al. 1999, Robinson 2001) or from moss (Tosi et al. 2002). Because we surface sterilized our
roots, these fungi may be endophytes colonizing the root or saprotrophs not adversely
affected by the hydrogen peroxide used for sterilization.
Our fungal community may be overly representative o f psychrotrophic fungi.
Psychrotrophic and psychrophilic fungi can grow at 0 °C, but psychrotrophic fungi can grow
above 20°C, while psychrophilic fungi have an optimum growth temperature at 15 °C or
below and cannot grow above 20°C (Robinson 2001). We grew our cultures at room
temperature, which may have excluded some o f the psychrophilic fungi. Another factor in
determining what fungi were successfully isolated is pH, because it influences the growth of
cultures (Taber and Taber 1984). Our isolates were grown on MMN, which has a pH o f 5 to
6 . However, we found later that the pH o f our sites was higher at 6 to 8 . The high pH o f 8

was from the dolomitic site, which may help explain the low number o f fungi that
successfully isolated from roots.
Discrepancies between morphological and PCR-based identification can be the result
o f two factors. The morphological identification may be incorrect, or the identity from the
sequence database may be incorrect. Bridge et al. (2003) found that approximately 20% of
sequences available in GenBank are unreliable for the ITS region o f rDNA for Amanita and
Phoma and the rRNA small subunit for members o f the Helotiales. Our findings, however,
were probably accurate because many o f the similar taxa downloaded for analysis had more
than one sequence for a species. Problems can arise if the unknown sequence is closely
matched to an incorrectly identified sequence. Knowledge o f phylogenetic relationships and
taxonomy is important to find these possible misidentified sequences.

110

Overall, warming and host plant did not have a noticeable impact on the composition
o f the root endophytic fungal community, which was dictated primarily by site. These plots
have only been warmed for approximately five years, which may be too short a time to find
differences, especially in the arctic where processes are much slower than those found in
temperate forests. Examining the fungal communities after another 5 to 10 years may show
more effects o f warming on the root endophytic fungal community. Although host
specificity was not found for the overall community, Phialocephala fortinii was
preferentially isolated from Salix arctica and Saxifraga oppositifolia over other host plants.
Our study identified many o f the isolates; however, most o f the isolates remain
unidentified. Even samples that have been sequenced remained unknown, which implies that
there are many fungi that have yet to be identified in these ecosystems. The role o f these
fungi was not determined although many have speculated that root endophytes play an
important role for nutrient uptake by plants in harsh environments. Bioassay studies using
some o f these isolates would help elucidate some o f the roles o f these fungi.

Ill

IV. The root-associated fungal community along a directional, non-replacement
succession chronosequence from the Canadian High Arctic

A. Introduction
Facilitation is a type o f complementary effect, a theory to explain how resource use
by organisms can direct ecosystem processes (Cardinale et al. 2002, Loreau and Hector
2001). In this theory, species diversity can increase while avoiding competition, and species
are able to co-exist especially in environments with limiting resources by: 1) partitioning
resources, where each species can use nutrients, water, or other resources differently instead
o f all species competing for or using the same resources; or 2 ) niche differentiation, where
different species avoid using the same resources as other organisms by means o f time and/or
space (McKane et al. 2002). In contrast, selection effect theory is where species diversity is
correlated with the chance that a dominant species uses most o f the available resources
(Cardinale et al. 2002), so the formation o f the community is heavily dependent on this one
central species.
Facilitation is thought to play an important role for the primary succession of
mycorrhizal fungi in that nutrient poor conditions are improved by pioneer species, as
indicated by increasing diversity o f arbuscular mycorrhizal fungi after volcanic disturbances
(Titus and del Moral 1998), and o f ectomycorrhizae on glacial tills (Helm et al. 1999), and by
the increase o f nitrogen and organic matter along a chronosequence from a sub-alpine glacier
(Jumpponen et al. 1998). This facilitative process is also inferred by the successional pattern
in the mycorrhizal status of plants, which begins with non-mycorrhizal plants, to arbuscular
mycorrhizal, ectomyeorrhizal (Read 1993), and/or ericoid mycorrhizae (Cazares 1992).
Given that complementary effect is the main theory to explain resource use by the root112

associated fungal community, selection effect theory cannot be totally discarded. The
possiblility that one or two dominant species may use most o f the resources and influence
how the community structure will develop subsists.
The objectives o f this study were to examine how the mycorrhizal fungal community
composition, species diversity and abundance change according to host plant and
chronosequence in a directional, non-replacement succession. Directional, non-replacement
succession has been used to describe a type o f plant succession, where species found in the
youngest plots are also found in older plots; they are not replaced by different plant species
(Svoboda and Henry 1987). This type o f succession serves as a natural laboratory where the
increase of biodiversity can be examined. This study is the first to our knowledge that
examines the root-associated fungal community in this type o f succession. Also, previous
research on succession in the arctic occurred in the low arctic where trees are present (Helm
et al. 1999, Helm et al. 1996, Brubaker et al. 1995, Chapin et al. 1994). Ellesmere Island is
located in the high arctic where only low shrubs grow.

B. Materials and methods

1. Study site
Samples were collected from the receding western lobe o f the Twin glacier located at
the south end o f an 8 -km^ lowland high arctic plant oasis on Ellesmere Island, Nunavut,
Canada (78°53’N, 75°55’W) at the end o f July, 2001. The western lobe o f the Twin glacier,
which started to advance since the Little Ice Age and began to recede approximately 1960, is
diminishing approximately 10 m/yr (GHR Henry, pers. comm.). Plots were placed within
zones delineated by the time since exposure: 1990 to present [labeled 1990 plots], 1980-90
[1980 plots], 1970-80 [1970 plots], and 1960-70 [1960 plots]. The area before 1959 was not
113

covered by the western lobe o f the glacier (GHR Henry, pers. eomm.), so plants collected
from this area were used as controls.

2. Field collection
Plants were seleeted that fulfilled the directional, non-replacement succession model
(i.e. once the plant appeared within the chronosequence, it was present for all remaining
chronosequence zones including the eontrol). The plants chosen were Luzula confusa
Lindeb., Papaver lapponicum (Tolm.) Nordh., Salix arctica Pall., Saxifraga oppositifolia L.,
Cassiope tetragona (L.) D.Don, and Dryas integrifolia Vahl. Luzula confusa and P.
lapponicum were first present in the 1990 plots (Y90), and they were the dominant plants for
1990, 1980, and 1970 plots. S. arctica first appeared in the 1980 plots (Y80), became more
abundant and larger in the 1970 plots (Y70), and was dominant in the 1960 plots (Y60). S.
oppositifolia was first noted in the 1970 plots, and C. tetragona and D. integrifolia in the
1960 plots. The control (<Y50) was dominated by C. tetragona and S. arctica. Epilobium
latifolium Pursh was also found, but only in the 1980 plots. Moss was present in all
chronosequenee zones. Three plant speeimens o f each target species were collected, and
when possible, approximately 300 g o f soil were collected adjacent to the plant samples.
Both plant and soil samples were stored in a permafrost cooler while in the field and at 4° C
at UNBC until processed.

3. Sampling from roots
Soil was loosened from plant roots by storing roots in water for at least 24 hours at 4°
C. A 0.5 mm sieve (No. 35 USA standard testing sieve, W.S. Tyler, Inc.) was used to
capture roots that were gently cleaned with water. Fifteen root tips were randomly sampled
by placing root systems on a numbered grid. Grid locations were selected with a random

114

number generator, and 2-cm segments o f root closest to the grid location were selected. Root
segments were described morphologically based on color, tip ramification, and presence and
absence of rhizomorphs (Agerer 1987-1998, Goodman 1995). Tips were then placed in a 1.5
ml Eppendorf tube and frozen in a -40° C freezer until DNA extractions were done.
Remaining root systems were frozen and lyophilized. To test if primers would amplify plant
DNA, plant tissue from stems or leaves was also collected.

4. Root microscopy
Root tips o f Papaver lapponicum and Luzula confusa were cleared and stained
according to Brundrett et al. (1996) to confirm that they were non-mycorrhizal as reported by
Kohn and Stasovski (1990). If roots o f other species appeared to be non-mycorrhizal when
randomly selecting for DNA extraction, a subsample was collected for clearing and staining
as well. Roots were cleared by autoclaving in 10% KOH for 30 min for P. lapponicum, 60
min for L. confusa, and 80 min for Saxifraga oppositifolia and Cassiope tetragona. Cleared
roots were rinsed with water and then stained by autoclaving for 15 min in 0.03% Chlorazol
Black E (CBE) in lactoglycerol. Roots were stored and mounted in lactoglycerol and
observed microscopically.

5. DNA extraction and ITS-T-RFLP analysis
Extraction o f root endophytes followed the CTAB protocol o f Gardes and Bruns
(1996b), modified by excluding the freeze-thaw process and including an extra purification
step with phenokchloroform-isoamyl alcohol (1:1) (Lee and Taylor 1990). DNA was
resuspended in 50 pL o f TE -8 buffer. Plant tissue was extracted with the same protocol
except the extra phenokchloroform-isoamyl alcohol purification step was not included, and

115

DNA was resuspended in 100 pL o f TE -8 buffer instead o f 50. All extractions were stored in
-20° C until use.
The ITS region o f the rRNA gene was amplified using a PCR cocktail composed o f
lOX Buffer (200 mM Tris-HCl [pH 8.4], 500 mM KCl) (Invitrogen), 2X dNTPs, 25 mM
M gCb, 0.5 pM o f ITS 1 (White et al. 1990) dye-labeled with Cy 5.0, 0.5 pM o f ITS 4
(Gardes and Bruns 1993) dye-labeled with Cy 5.5, 5 U o f Platinum Taq Polymerase
(Invitrogen), and 2.2 pL o f DNA template. PCR amplification was accomplished using the
following program on a MJ Research thermocycler PT-100: 94 °C (4 min); 48° C (1 min);
72° C (2 min); [94 °C (30 sec); 48° C (30 sec); 72° C (1 min 30 sec) x 34 cycles]; 72°C (6
min 30 sec). Successful amplification was verified on a 0.7% agarose gel. For all
unsuccessful amplifications, amplifications were retried with the DNA templates diluted to
1 : 10 .

The endonucleases Alu 1 and Hinf \ (Invitrogen) were mixed with 6 pL o f the PCR
product in a 10 pL volume following the manufacturer’s recommendations and incubated at
37° C for at least 3 hours. Digested product was mixed with formamide loading dye (1:1.65
pL) then denatured at 80° C in a heat block for 2 min and quenched on ice before 2 pL were
loaded onto a 6 % polyacrylamide gel. The formamide loading dye contained two sets of
internal markers at 101, 200, and 351 bp, labeled with Cy 5.0 for one set and Cy 5.5 for the
other. Samples were run for 60 min. on an OpenGene System Long Tower Sequencer (Bayer
International), with the laser power set at 50%, temperature at 53° C, and current at 1250 V.
GeneObjects 3.1 fragment analysis software (Bayer International) was used to determine
fragments for each primer-enzyme combination (i.e. ITS \-Alu I, ITS \-H ird \, ITS A-Alu I,
and ITS A-Hinî\).

116

6 . Matching root tips with sporocarps

Fragments generated from root-associated fungi were compared with those of
sporocarps that were collected from Alexandra Fiord in 2000 using TRAMP (T-RFLP
analysis matching program) (Dickie et al. 2002). Multiple fragments were found on many of
the root tips; therefore, a database using every combination o f fragment sizes from the four
primer-enzyme combinations was made to be tested with fragment sizes from sporocarps.

7. Statistical analysis
Frequency tables o f genotypes generated from the T-RFLP analysis were made for
eaeh enzyme-primer combination. To adjust for inter-gel variation, genotypes were
manually binned. These average binned fragment genotypes were used to construet
frequeney tables, which were based on cumulative frequeney of the genotypes for eaeh
replicate o f treatment per site. The different primer-enzyme combinations were used as
individual datasets for examining if root-associated fungal eommunities were more
influeneed by time o f exposure or by host plant; therefore all analyses exeept for the
ordination analyses were done in quadruplicate (once for eaeh enzyme using forward [Cy5]
and reverse [Cy5.5] fragments). Genotypes that matehed fragments from plant tissue (+/- 3
bp) o f Saxifraga oppositifolia and Papaver lapponica were deleted.
The frequency tables made for the four primer-enzyme combinations were merged
into a single frequeney table to use for non-metric multidimensional scaling (NMS) with PCORD, version 4.25 (McCune and Mefford 1999). NMS was used to examine the effects o f
time of exposure and host plant on species frequency. Beal’s transformation was applied to
the merged data to alleviate problems associated with datasets that contain numerous zeroes
(McCune and Grace 2002). The following parameters for NMS were: Sorensen was used for

117

measuring distance; selection for the configuration o f the first run was random, but scores for
the third configuration were used for the consequent runs as recommended by the program;
15 runs were conducted; dimensionality was assessed by examining the scree plot; and 30
randomized runs were used in the Monte Carlo test.
Canonical correspondence analysis (CCA) was used to test for correlations between
soil properties and differences in fungal communities using PC-ORD (McCune and Mefford
1999). Transformation o f genotypes was done by adding one to each frequency to alleviate
the problems associated with too many zeroes in databases. Soil properties (pH, CEC,
available NH4, available NO3, available P, and C:N ratio, and exchangeable cations) were
transformed with log base 10 for each property except for pH and C:N ratio. Linear
relationships between the genotype abundance and soil properties were tested with the Monte
Carlo test. The ordination diagram was based on LC (linear combination) scores; these
scores are linear combinations o f soil properties (McCune and Grace 2002).
To examine how communities would group according to time exposed for a given
host plant, a hierarchical clustering was done using W ard’s method o f clustering based on a
Euclidean distance matrix. Ward’s method is a space-conserving method and minimizes
chaining, which is the addition o f single items to existing groups so distinct branches are lost
from dendrograms (McCune and Grace 2002).
Effects o f soil properties were further examined by simple regression, where
[ 1] p{soil properties|year o f exposure}=|3o + Piyr o f exp + s.
A regression was done for each o f the 13 soil properties: exchangeable Al, exchangeable Na,
exchangeable Fe, exchangeable K, exchangeable Ca, CEC, exchangeable Mg, exchangeable
Mn, pH, C:N ratio, available P, available NO 3-N, and available NH 4-N. To test for the

118

possible changes in soil properties due to presence o f plants, simple regressions were done,
where
[2] p{soil properties! host plant}=Po + PiY host plant + e.
Host plant was limited to plots in which they were present, e.g. to test for all chronosequence
zones, only Luzula confusa, Papaver lapponicum, and Salix arctica were used; whereas for
the control and Y60 all host plants were used because all were present in these plots.
Genotype diversity was examined using two indices (richness and evenness) and rank
abundance curves. Richness was determined as the number o f different genotypes for a
given plot. Evenness was determined using EcoSim (Gotelli and Entsminger 2001) for each
genotype where Hurlbert’s probability o f interspecific encounter (PIE) index was applied to
measure evenness. Simple regression was done to test for differences o f richness and
abundance (the number o f individuals in a plot) for each host plant at each succession zone.
Niche differentiation was analyzed by examining rank-abundance curves. The natural log o f
abundance o f species was graphed against the arithmetic ranking o f abundance. To test for
similarities in community composition, Sorensen’s quantitative similarity index was used in
Estimates, vers.7 (Colwell 1997). Communities from the same host plant as well as different
host plants were tested for shared species.

C. Results
Overall, 53% o f the 930 attempted extractions were successful for all four primerenzyme combinations. Successful amplification differed according to host plant and primerenzyme. Fungi extracted from roots from Papaver lapponicum and Luzula confusa were
most successful (see Table 4.1) and Saxifraga oppositifolia the least. When examining the
number o f extractions per number o f attempts, the percentage of successful amplification was
119

comparable for all primer-enzyme combinations. When compared to total number o f
genotypes generated, up to 16% more genotypes were obtained with the forward primer than
the reverse. Although there were no genotypes that matched with sporocarps, there were
matehes with sequenced cultures o f Phialocephala fortinii. From microscopic analysis, fungi
colonizing Papaver lapponicum and Luzula confusa were endophytic fungi with runner
hyphae and microsclerotia present, charaeteristies o f dark septate endophytes. Cassiope
tetragona and Saxifraga oppositifolia also had runner hyphae and mieroselerotia present for
those samples examined with elearing and staining.
Fungal communities, determined by NMS and CCA, were distinguished by host plant
and not by how long the soil had been exposed in the chronosequence (see Figs. A4.1 and
4.1). For both analyses, the root-associated fungal community found on Papaver
lapponicum, Luzula confusa, and Cassiope tetragona were distinct from each other. The
eommunities found on Salix arctica, and Dryas integrifolia were more tightly grouped. In
NMS, the fungal community from Saxifraga oppositifolia clustered with L. confusa. The
overall variance in NMS was explained by a cumulative r^ =0.925 with three axes; axis 1
explained 47.5% of the variance, axis 2 explained 15.4% o f the variance, and axis 3
explained 29.6% o f the variance. The final solution required 58 iterations. Final stability of
the model was met with the final stress = 8.01 (p= 0.03) and final instability = 0.00009.
Soil was not a determining factor in distinguishing root-associated fungal
communities that are separated aeeording to host plant. From the CCA, available P was
more strongly associated with the Luzula confusa clade, and nitrate was higher for soil
surrounding Papaver lapponicum and L. confusa. However, results from the Monte Carlo
test indicate that there was no relationship between soil variables and the main matrix

120

Table 4.1. Number of successful extractions according to the number of attempts at extraction and total number o f genotypes for each
host plant.

Luzula
confusa
Papaver
lapponicum
Salix
arctica
Saxifraga
oppositifolia
Cassiope
tetragona
Dryas
integrifolia

K)

ITS \-Hinî\
# of
# of
successful extractions
extraction per total #
per # o f
of
attempts
genotypes

ITS 4-Hinf\
# of
# of
successful extractions
extraction per total #
of
per # o f
genotypes
attempts

ITS \-Alul
# of
successful
extraction
per # of
attempts

61%

82%

63%

76%

79%

92%

78%

48%

67%

18%

# of
extractions
per total #
of
genotypes

ITS 4-Alul
# of
successful
extraction
per # o f
attempts

# of
extractions
per total #
of
genotypes

61%

82%

63%

75%

81%

77%

88%

78%

81%

46%

59%

44%

62%

44%

55%

41%

17%

29%

18%

41%

18%

30%

29%

61%

24%

45%

31%

63%

30%

55%

49%

60%

51%

59%

48%

57%

48%

59%

A x is 3

« 5 9 Saar

Y 6 0 D rin

0 P ala
Y 60
'Y 8 0 P ala
' 5 9 D rin
Y 6 ( f laop
Y 90 Lu(
Y 8 0 L u co

A x is 1

10 P ala
Y 70

Y 60 L uco
< 5 9 L u co ,

<?f) P ala

< 5 9 C a te

>60 C a te

Fig 4.1 Canonical Correspondence Analysis for effects o f host plant and chronosequence
zone. Rotated 345°. Number preceded by ‘Y ’ represent the earliest time o f exposure from
the glacier. <Y59 is the control, or réfugia plot that was not covered by the glacier studied.
Plants are: CdX&=^Cassiope tetragona, Drin = Dryas integrifolia, L\xco=Luzula confusa,
Vala^Papaver lapponicum, Saar=5'«/zx arctica, ^ao^=Saxifraga oppositifolia. Biplots were
used to examine the relationship between soil properties and genotypes.

122

Distance
5.8E -01

1.3E +01

2.6E + 01

3.9E + 01

5.1E + 01

Y90 Luco
Y80 Luco
Y70 Luco
Y60 Luco
<59 Luco
Y70 Saop
Y80 Saar
Y60 Saar
<59 Drin
Y70 Saar
Y60 Drin
Y60 Saop
<59 Saar
Y60 Cate
<59 Cate
Y90 Pala
<59 Pala
Y60 Pala
Y80 Pala
Y70 Pala

Fig 4.2 Cluster analysis using Ward’s method based on Euclidean distances. Key same as in
Fig. 4.1

because the eigenvalue oceurrence could be explained by chance; eigenvalue for axis
1=0.106 (p=0.3); axis 2=0.059 (p=0.57); axis 3=0.047 (p=0.38). Calcium and magnesium
were omitted from the final analysis because o f their high correlation with CEC (Ca: r=0.97,
Mg: r=0.962). The final dimensions used were 20 plots for both the main (plot) and
secondary (soil properties) matrices with 288 species and 11 soil properties. Three axes were
interpreted, with axis 1 explaining 19.7%, axis 2 10.3%, and axis 3 8.7% o f the variance.
One o f the differences between the two ordination analyses is that in CCA, both fungal
communities from S. oppositifolia do not cluster with Luzula confusa as in the NMS analysis;
one community associated with S. oppositifolia grouped more closely with S. arctica and D.
integrifolia.

123

Results from our cluster analysis supported those found from the ordination analyses
and were closer to the results from CCA (see Fig. 4.2). Plots clustered by host plant with the
exception o f Salix arctica and Dryas integrifolia grouping together, Y70 Saxifraga
oppositifolia branching with Luzula confusa, and Y60 S. oppositifolia clustering in the S.
arctica-D. integrifolia cluster.
Some o f the soil properties either increased or decreased as the glacier receded (see
Fig. A4.2). Concentrations o f Al (p=0.0005), K (p=0.013), and Fe (p=0.006) increased as the
glacier receded, and Mg (p=0.03), CEC (p=0.09) and NO3 (p=0.009) decreased. Although
not significant, P followed a trend where amounts increased as the glacier receded, and Na
and Ca tended to decrease. Amounts o f NH4 were lower in the glaciated areas than in the
control. The pH was higher in the glaciated areas with a pH ranging from 4.2-4.5 compared
to a pH in the control o f 5. The C:N ratio increased from Y90 to Y80, decreased from Y80 to
Y70 with the introduction o f Salix arctica, and gradually increased with older plots. There
were no differences in soil properties o f soil samples collected near different host plants.
Data from soil samples that were collected around Dryas integrifolia from the control and
from around Papaver lapponicum in Y90, were deleted because less soil was collected from
these plots, which skewed the results.
Genotype richness increased as the glacier receded for Luzula confusa for the reverse
primer datasets (see Fig. A4.3). The regressions testing species richness against time for both
ITS A-HinÛ and IT S4-^M were significant (p=0.005 and p=0.02 respectively). Although
regressions based on the datasets o f the forward primer were not significant, they had a
similar trend where richness increased with increasing time since exposure. Genotype
cumulative frequency also increased significantly for L. confusa as the glacier receded
124

(p=0.008, p=0.003, p=0.02, p=0.02 for IT Sl-//m fI, ITS4-//mf[, YYS\-Alu\, and IT S4-^M
respectively) (see Fig. A4.4). The richness (Table A4.1), evenness, and cumulative frequency
o f the communities on the other host plants did not change significantly with time. Species
evenness did not change significantly between plots for any community.
Similarity based on Sorensen’s quantitative similarity index was highest for
communities found on the same host plant except for Dryas integrifolia, which had a higher
similarity with Salix arctica at 25% (Table A4.2). The root-associated fungal communities
found on Luzula confusa for all 5 plots were more similar than fungal communities found on
other host plants according to Sorensen’s quantitative index, where 40%-60% o f the
communities were similar, depending on which primer-enzyme was used for analysis. For
Papaver lapponicum 32%-40% o f the fungal communities from five plots were similar; for S.
arctica 41%-48% o f the fungal communities found on four plots were similar; for Saxifraga
oppositifolia 28% o f the fungal communities found on three plots were similar; for Cassiope
tetragona 26% o f the fungal communities found on two plots were similar, and D.
integrifolia 15% o f the fungal communities found on two plots were similar.
The resource allocation for the fungal communities differed according to its host
plant. On Luzula confusa, the community from Y90 fit the geometric series model and those
from the control were described by the log normal model, with the communities in the other
plots falling between the two extremes (Fig A4.5). Information for Papaver lapponicum was
more inconclusive and depended upon the enzyme-primer (Fig A4.6 ); most o f the
communities from the different succession zones fit the log normal model, but some were
described by the geometric series model. Most o f the communities o f Salix arctica fit the log
normal model, but results w ith /fM found that the geometric series model described

125

communities from Y70 plots (Fig A4.7). The communities on Saxifraga oppositifolia fit the
geometric series model except in the Y70 plot, where the log normal model better described
the community (Fig A4. 8 ). The time o f exposure made a difference for resource allocation
for communities found on Cassiope tetragona (Fig A4.9). In the most recently exposed plots,
the community associated with C. tetragona fit the geometric series model. However, in the
older plots, the communities fit the log normal model. For Dryas integrifolia, the log normal
model fit communities found in <Y59 plot (Fig A4.10). For communities in the Y60 plot,
either the log normal or geometric series model described the community depending on the
primer; communities described by the forward primer fit the geometric series model and by
the reverse, the log normal model.

D. Discussion
Root-associated fungal community composition depended more on the mycorrhizal
status o f the host plant rather than how long plant communities have colonized the area since
the glacier retreated. The root-associated fungal communities found on the ectomycorrhizal
plants, Salix arctica and Dryas integrifolia, grouped together in both the ordination analyses
and cluster analysis. The root-associated fungal communities found on the only ericoid
plant, Cassiope tetragona, were separated from the other plants. The root-associated fungal
communities on Saxifraga oppositifolia clustered with either the communities found on
Luzula confusa or with the S. arctica - D. integrifolia clade. The mycorrhizal status o f S.
oppositifolia at Ellesmere Island has been reported to be ectomycorrhizal (Kohn and
Stasovski 1990), but runner hyphae and microsclerotia, characteristic o f septate endophytes,
were observed in its roots in the present study. Therefore, its tendency to group with the
ectomycorrhizal plants and non-mycorrhizal, endophytic plants was not surprising.
126

The roots o f Papaver lapponicum and Luzula confusa had both runner hyphae and
microsclerotia, which indicated that both o f these host plants were colonized by septate
endophytes, which included Phialocephala fortinii. Interestingly, the communities on these
host plants were distinct from each other as indicated by the two ordination analyses, the
cluster analysis and the low similarity index from the Sorensen’s quantitative index. Because
septate endophytes are thought to be ubiquitous and not host-specific (Jumpponen and
Trappe 1998, Wilberforce et al. 2003, Harney et al. 1997), finding similar communities on
both hosts was expected. There are at least three explanations for this dissimilarity between
the two communities: The first explanation is that the root-associated communities may have
contained a higher proportion o f parasitic or pathogenic fungi, which generally exhibit higher
levels o f host specificity than expected. Parasitic fungi were found from soil extractions
taken from a glacial terminus (Jumpponen 2003).. Latent pathogens grow along and over the
root surface, eventually the hyphae enter the stele when the plant decreases its resistance, and
invade senescent cortical cells (Garrett 1981). Another explanations is that the distribution of
host plants may be correlated with an environmental variable, e.g. soil moisture, that we did
not measure, but which affected the composition o f the root-associated community. The
third possibility is that there may be more host-specificity in the root-associated community
than previously recorded.
This possibility that there is more host specificity o f DSE than previously thought,
which would also indicate that this group is very diverse in order to have two communities
distinct from the other. Genotype richness was highest for the septate endophyte
communities found on Luzula confusa and Papaver lapponicum. Number o f potential
genotypes for L. confusa reached as high as 39, and its range was comparable to the number

127

of fungal endophytes from grasslands (Wilberforee et al. 2003). Although many o f the
members were thought to be Phialocephala fortinii from microscopic investigations, the
morphological features o f P. fortinii may not be limited to this species and may have broad
taxonomic affiliations to taxa that have not been identified (Sehadt et al. 2001). The high
diversity o f DSE may also be attributed to: 1) P. fortinii is a multi-speeies complex that
contains cryptic species that have been found to occupy the same root tip (Grünig et al.
2002b); 2) additional DSE were present that fit the DSE description but were not culturable;
or 3) intraspecifie variation, which has been reported for the ITS region for P. fortinii
(Harney et al. 1997). Identifying the taxa involved would help clarify the richness and
distinction between the two communities.
Species richness was comparable to other fungal community studies based on RFLP
and T-RFLP for other host plants. However, similarity between communities o f different
host plants was low. Even for communities with the same host plant but on different
specimens, the similarity index was low, with only a couple o f incidences where the
similarity was close to 60% on L. confusa. The low similarity suggests that spores were
aerially deposited (Jumpponen 2003), with ramets establishing from different sources
(Grünig et al. 2002a). The high similarity o f 60% found on L. confusa may be from chance.
High similarity would suggest that the fungi are vegetative; however, L. confusa have small
root systems so the chances o f a root system from one plant contacting another would be
small.
As the plant community became more diverse along the chronosequenee, soil
properties began to change. No host plant significantly altered soil properties, which
indicates that the plant and its root fungal community as a whole, not individually, drove

128

these changes. For many parts o f Alexandra Fiord and other places in the arctic, vegetation
development is assumed to be the main pedogenie factor for these soils (Muc et al. 1994).
Interestingly, the amount o f NO 3 decreased with increasing time o f exposure; total N and
ammonium did not change with time. The decrease in NO 3 may be due to: 1) its status as the
main nitrogen source for plants at this succession site, but this would be surprising because
only some ectomycorrhizal fungi have the ability to reduce nitrate (Paul and Clark 1996); 2)
its highly solubility so it may have leached from the soil; 3) High Arctic processes whose
rates are slower than those found in alpine systems, so not enough time has transpired to test
if the amount o f N will increase; or 4) increases in denitrifiers that reduce nitrate to
ammonium.
The CEC decreased as the chronosequenee and plant diversity increased with time.
This trend was found in Norway and was attributed to the establishment o f vegetation cover
and the development o f nutrient cycling (Matthews 1992). This was contrary to the expected
rise in CEC due to the increase o f organic matter as found with a chronosequenee at Glacier
Bay, AK (Matthews 1992) and Lyman glacier, WA (Jumpponen et al. 1998). However, our
sites did not have relatively high level o f accumulation o f organic matter.
Models for niche apportionment on Luzula confusa fit as expected. The youngest
plots fit the geometric series model, which often describes communities in harsh
environments (Magurran 1988). The fungal community on L. confusa in the control, where
glaciation had not occurred during the little lee Age, fit the log normal model, which often
describes more diverse plant communities such as temperate or tropical rainforests
(Magurran 1988). Cassiope tetragona and Dryas integrifolia also fit the expected models,
where their communities fit the geometric series model in the youngest plots, and log-normal

129

for older areas. Salix arctica for the most part had a log normal distribution, and Saxifraga
oppositifolia fit the geometrie series model. Perhaps when S. arctica establishes in a new
area, it requires a more established mycorrhizal fungal community to survive, as with other
ectomycorrhizal plants, which are often found in older areas during plant succession (Cazares
1992). Saxifraga oppositifolia has been reported to be both ectomycorrhizal and nonectomyeorrhizal; therefore, this plant may be better at adapting to environments where the
ectomycorrhizal fungal community is not well-established and may depend more on other
fungal root endophytes or no fungal endophytes. This may also indicate that those
communities with a log normal distribution have more facilitation between its members;
whereas, communities that fit the geometric series model have a few dominant species
accounting for most o f its resource consumption, which would fit the selection effect theory.
Observing succession following exposure by a receding alpine glacier, Cazares
(1992) reported that the first plants were non-myeorrhizal, followed by arbuscular
mycorrhizal plants, then ectomycorrhizal plants, and finally ericoid plants, with DSEs found
throughout. Our succession follows the same trend except that we did not find any
arbuscular mycorrhizae. Arbuscular mycorrhizae are thought to be rare in the arctic
(Haselwandter 1987, Bledsoe et al. 1990), although Kohn and Stasovski (1990) found
arbuscular mycorrhizae on a fern, Dryopteris fragrans (L.) Schott, from Ellesmere Island.
Arbuscular mycorrhizae may have been present on Salix arctica as found on alpine Salix spp.
(Cazares 1992, Trowbridge and Jumpponen 2004), but colonization on these alpine willows
constituted less than 1% (Trowbridge and Jumpponen 2004) so they could have easily been
missed in this study. Interestingly, a shift in mycorrhizal status was found in Saxifraga
oppositifolia. The communities that had been exposed for 30 years grouped with Luzula

130

confusa, which is associated with septate endophytes. Communities exposed for 40 years
clustered with Salix arctica and Dryas integrifolia, both ectomycorrhizal plants that grouped
separately from the other host plants.
This site was unique in that the direetional, non-replaeement plant succession allowed
for the root-associated fungal community to be assessed as plant biodiversity increased and
host plants were not replaced. This type o f succession is an example o f faeilitation as
demonstrated by the change o f soil properties as the plant eommunity increases in
biodiversity. More plant diversity brought an increase in root-associated fungal diversity
because each plant had a relatively unique fungal community. For individual host plants, the
fungal richness did not inerease with plant diversity except for Luzula confusa and Dryas
integrifolia, and for all eommunities species evermess remained the same. Zak et al. (2003)
found that increases in plant diversity and thus production changed the microbial community
in a field experiment. They hypothesized that because plant species differ in their
bioehemical composition, they ean eontrol the composition and function o f heterotrophic
microbial communities. In our study, we found that the inerease o f speeies riehness in the
plant community may control the eomposition and function o f the root-associated fungal
community because most o f the fungal communities had low similarity to each other. The
inerease in plant diversity may have caused a shift in the fungal community composition as
Zak et al. (2003) suggested for the microbial community they examined.
Niehe differentiation may help explain the diversity o f the fungal community as well.
Species and species eombination are more important in controlling supplies o f nutrients than
species richness (Hooper and Vitousek 1997, Tilman et al. 1997b). Perhaps, the different
fungal members were exploiting different niches and obtaining nutrients at different depths
131

or areas, which do not interfere with the acquisition by fungal communities found on
neighboring hosts. Biodiversity may not be important for maintenance o f an ecosystem, but
it can allow changes to occur (Loreau et al. 2001).
Directional, non-replacement succession at the edge o f a glacier forefront gave a
unique opportunity to examine how the root-associated fungal community responded to an
increase in plant diversity. The diversity o f the root-associated fungal community was higher
than expected for most host plants and for the most part, was not as species-poor as its
aboveground counterparts. Co-existence was evident above-ground as well and for most of
the belowground fungal communities; aboveground because o f the increase o f plant diversity
along the chronosequenee, and belowground by the high genotype richness found on most
host plants. Perhaps for ectomycorrhizal and ericoid plants, an established fungal community
was needed for plants to succeed. These systems may also fit the selection effect model.
Plants with fungal communities that fit the geometric series model may be more dependent
on fewer fungal partners or have more competitive fungi colonizing their roots. These few
dominant species may be important in shaping the fungal community. The possibility that
the septate endophytic community was more diverse that previously assumed, and their role
in plant establishment may be critical. However, more research in identifying the fungal
players would be needed to interpret the richness established by their community fingerprint.

132

V. Synthesis of results
The effects o f site and temperature on the root-associated fungal community o f Salix
arctica was tested hy two methodologies, one based on PCR techniques and the other hy
isolating fungi from the sterilized root surface. Comparison o f fungal genotypes using the
two methods was done by matching T-RFLP fragments generated from fungal DNA
extracted directly from root tips (T-RFLP - OTC study), to fragments from RFLPs produced
from fungal cultures isolated from root tips (culture study) for each o f the 18 plots. If a TRFLP fragment size was within 10% o f the RFLP band size, they were considered a putative
match between the T-RFLP and RFLP datasets. The 10% threshold was to account for the
different algorithms to determine the fragments sizes o f the two methods. Matches were
counted only if each o f the four primer-enzyme combination (i.e. ITS 1 -A lu l, ITS 4 -A lu l,
ITS 1 - HinÛ, and ITS 4 - HinÛ) matched RFLP fragments for the two restriction enzymes.
O f the possible 150 fungal cultures isolated from Salix arctica from all plots, 65
(43%) matched samples from T-RFLP analysis. For the lowland OTC plots, 85% o f the
fungal cultures were also found in the T-RFLP analysis. Most of these lowland isolates were
morphologically identified as Phialocephala fortinii or as hyaline septate endophytes. The
highest percentage o f matches between RFLP hands from cultures and T-RFLPs fragments
were from the lowland plots. The percent o f cultures with matching T-RFLP fragments for
the three sites ranged from 17 to 50%, in which the 17% were found on the dolomitic OTC
plots. This low percentage may be due to the small number o f isolates found (4) on the
dolomitic site, which was the lowest number o f isolates found for all the plots.
The total number o f genotypes from the T-RFLP - OTC study that had matching
cultures characterized by RFLP patterns was 188 (21%). This is out o f 894 possible
133

genotypes, which was the number o f genotypes found by the primer-enzyme combination
ITS 4 - Alul. This primer-enzyme combination was used to make comparisons between the
T-RFLP - OTC and culture studies instead o f the other three primer-enzyme combinations
because the least number o f genotypes were generated from ITS 4 -Alu\. Seventy-nine
percent o f the community from the OTC -T-RFLP study was probably not dark or hyaline
septate endophytes isolated from root tips and were likely to be mycorrhizal fungi and
pathogens. This difference between culturing and PCR-based techniques has been noted for
ericaceous plants where Sebacina was the most dominant fungus when amplifying from roots
and Capronia-likQ fungi from fungal cultures (Allen et al. 2003).
Although Phialocephala fortinii accounted for 31% (58/188) o f the RFLP samples
that were also present in the T - RFLP community, it only accounted for 6.5% (58/894) o f
the whole eommunity from the T - RFLP-OTC study. This percentage was lower than
expected because P. fortinii has been hypothesized to fill the functional void left by
arbuscular mycorrhiza (Bledsoe et al. 1990). This low percent may be due in part because
fewer root tips were used for isolating fungal cultures in comparison to the number o f root
tips used for DNA extraction; only five roots from each plant were used for fungal isolation
in contrast to approximately 30 root tips per plant used for T-RFLP analysis.
Based on the T-RFLP - OTC study, genotype frequency o f the fungal community on
Salix arctica increased in the warmed plots compared to the ambient plots overall; however,
this was only significant on the granitic site when examined for each site. One explanation is
that there was insufficient time for the warming to affect the root system. Initially, S. arctica
uses any excess resources for seed production (Jones et al. 1997). The increase in genotype
frequency may be the beginning o f further changes. The fungal community from the culture

134

study, which used an additional three plant species, did not change due to the warming
treatment. Again, not enough time may have passed for the warming to impact root growth.
As with S. arctica, Dryas integrifolia and Saxifraga oppositifolia initially direct resources to
for reproduction (Henry and Molau 1997, Stenstrom et al. 1997). Changes may not have
been found for the fungal community associated with Cassiope tetragona because not many
endophytes were isolated from its roots. O f the three sites examined, the impaets o f warming
probably will be observed first on the dolomitic site, where resource limitations are maximal.
Although the sample sizes were different for the two studies, some trends were found.
The genotype richness and frequency were different for T-RFLP - OTC and eulture studies.
For the fungal community in the T-RFLP - OTC study, genotype frequency and richness
were comparable between the lowland and granitic sites. In contrast, for the culture study,
genotype riehness and frequency were eomparable between the lowland and dolomitic sites.
Genotype frequency was higher on the lowland and granitic sites than on the dolomitic site
for the T-RFLP - OTC study, and was highest on the granitie site for the culture study.
Although there was a trend where frequency was higher in the warmed plots o f the lowland
and dolomitic sites, frequency was highest in the eontrol plot of the granitic site.
Although sampling may aeeount for some o f the discrepancies between the T-RFLP OTC and culture studies, other possibilities can contribute to these differences. The rootassoeiated fungal community on Salix arctica may be dominated by eetomyeorrhizal fungi,
which are more diffieult to culture, so richness and frequency may be higher when directly
amplifying DNA versus isolating fungi from root tips. The reason why frequency and
richness were high from the granitic site when isolating fungi from root tips may be because
these sites had more root endophytes. Although the soil chemistry o f the two sites was

135

comparable, the granitic site was more arid and had lower plant diversity than the lowland
site, so S. arctica on the upland site may have been dominated by non-mycorrhizal fungal
endophytes. This is supported by the comparison between PCR-RFLP and T-RPLP
fragments; o f the 65 fungal isolates that had matching T-RFLP fragments, 47.7% were from
the granitie site while only 20% were from the lowland plot, suggesting that a higher
proportion o f the root-associated fungal community on the granitic site were endophytes.
Host plant was a more important determinant o f the fungal community in the
suecession study eompared to the culture study. This can be attributed in part to the different
spatial and temporal scales between the two studies. The succession study eneompased a
more restricted area with soils o f similar composition, compared to the culture study; with
the range o f abiotic and soil factors diminished, the effect o f host became more apparent.
Over the larger environmental range and variation in soil parent materials represented in the
culture study, site was more important in determining community structure. Also, since
isolating from root tips favors fungi that grow quickly and have less host specificity
(Kernaghan et al. 2003), the role o f the host plant may have been more difficult to detect in
the culture study.
Genotype richness, in general, was higher for the fungal community in the T-RFLP OTC study in comparison to the fungal community in the succession study. However, the
average genotype frequency o f the control plot o f the succession study, an area which had not
been covered by the retreating glacier, was comparable to the lowland and granitic site o f the
T-RFLP - OTC study, and the earlier successional plots were comparable to the average
genotype frequency found on the dolomitic site. Some o f the differences in genotype
richness between the T-RFLP - OTC and succession studies were correlated with soil

136

chemistry. Higher amounts o f Al, Na, Fe, P, and NH4 were found in the control plot o f the
succession study compared to the lowland and granitic sites o f the warming study (see Table
A5.1). The control site from the succession study had lower amounts o f Mg and NO3 than
the lowland and granitic sites.
By using two methodologies, direet PCR amplification and culturing, to examine the
root-associated fungal community, a more comprehensive interpretation was possible
because each method favoured different groups o f fungi. The succession study provided an
added dimension, and gave insight into the development o f this arctic oasis.

137

VI,

Conclusion

Hypotheses revisited.
Ho 1.1 The root-associated fungal community, based on DNA directly extracted from root
tips from Salix arctica, will not differ between warmed plots and ambient plots.
The root-assoeiated fungal community based on DNA directly extracted from root
tips did not significantly differ between warmed and ambient plots; therefore, there was not
sufficient evidence to reject the null hypothesis. However, although not statistically
significant, the cumulative frequency o f the genotypes from the root-associated fungal
community on Salix arctica tended to be higher in the warmed plots compared to ambient
plots in the T-RFLP - OTC study. The first five to seven years o f passive warming may
have contributed mainly to reproduction (Jones et al. 1997), so the immediate effects of
warming may not have impacted the root system. However, as warming continues, more
carbon may be allocated to growing roots and initiating new lateral roots, increasing root
colonization by fungi (Pregitzer et al. 2000). It may be too soon to tell if the reponses o f the
root-associated fungal community to warming will play a role in plant migration, although
migration o f southern plants to Alexandria Fiord will likely be difficult as it is surrounded by
ice fields and ocean.

Ho 1.2 The root-associated fungal community, based on DNA directly extracted from root
tips from Salix arctica, will not differ due to site.
The root-associated fungal community, based on DNA directly extracted from root
tips from Salix arctica, varied significantly between sites; therefore, the null hypothesis is

138

rejected. The dolomitic site had significantly lower genotype frequency and richness than the
lowland and granitic sites.

Ho 1.3 The root-associated fungal community, based on cultures isolated from root tips, will
not differ between warmed and ambient plots.
The root-associated fungal community, based on cultures isolated from root tips did
not significantly differ between warmed and ambient plots; therefore the null hypothesis is
not rejected.

Ho 1.4 The root-associated fungal community, based on cultures isolated from root tips, will
not differ according to the host plants Cassiope tetragona, Dryas integrifolia, Salix
arctica, and Saxifraga oppositifolia.
The root-assoeiated fungal eommunity, based on cultures isolated from root tips, did
not differ significantly according to host plants; therefore, the null hypothesis is not rejected.

Ho 1.5 The root-associated fungal community, based on cultures isolated from root tips, will
not differ due to site.
The root-associated fungal community, based on cultures isolated from root tips, did
significantly differ among site; therefore, the null hypothesis is rejected. This is consistent
with the results o f the T-RFLP study.

Ho 1.6 The root-associated fungal communities described by the two methods (direct
extraction versus culturing) will not differ.

139

Although not statistically testable, the root-associated fungal community described by
the two methods did differ according to the method used. Not only did the two methods,
DNA extraction directly from roots and fungal cultures isolated from roots, distinguish two
fungal communities, but these communities appeared to respond to warming differently. The
effects o f warming on the root-associated fungal community based on fungal cultures
isolated from root tips were not as obvious as in the T-RFLP - OTC study. Part o f the reason
may be because many o f the isolates were P. fortinii, which accounted for 31% o f the fungal
cultures but only 6.5% o f the genotypes from the T-RFLP - OTC study. Therefore,
ectomycorrhizal fungi probably made up a much larger proportion o f the fungi in the TRFLP - OTC study. Based on the morphotyping done prior to T-RFLPs, most o f the roots
appeared to have at least a mantle.

Ho 2.1 The root-associated fungal communities will not differ along a chronosequenee.
The root-associated fungal communities did not significantly differ along a
chronosequenee; therefore, the null hypothesis is not rejected.

Ho 2.2 The root-associated fungal communities will not differ according to host plant.
The root-associated fungal communities did not significantly differ according to host
plant, but the composition o f the root-associated fungal community did differ according to
the mycorrhizal status o f the host plant during directional, non-replacement succession. This
is indicative o f facilitation, where the non-mycorrhizal plants and their rhizosphere improve
soil conditions for new plants to colonize the land. Also, niche differentiation was evident
because each o f the plant species in a plot hosted a unique fungal community (although there

140

was overlap between the two ectomycorrhizal plant host communities, as indicated in the
ordinations). Interestingly, host specificity was suggested even with the non-mycorrhizal
host plants, L. confusa and P. lapponicum, as each host plant also had unique communities
based on the ordination analysis. Although non-mycorrhizal, these plants were colonized by
endophytes.

General conclusions:
The root-associated fungal community was not as simple as expected. The rank
abundance curves did not fit the geometric series model which typically characterizes arctic
plant communities. The root-associated fungal communities in the T-RFLP - OTC study fit
the log normal model and are as diverse as root-associated fungal communities found in
temperate forests. Further analysis with Tokeshi’s models elucidated a possible mechanism
due to an increased number o f lateral roots being colonized and how these niches are filled.
This may help explain the patchiness o f fungal species found on root tips, where 1-2
dominant types are generally found in any one area (Horton and Bruns 2001). These patches
o f fungi may outcompete other fungi and dominate a certain niche, which is partly supported
by our finding that some dominant genotypes remained dominant in warmed plots even
though cumulative frequency increased. In this case, observing changes in frequency o f
individual genotypes was important in elucidating how the community reacted to a
disturbance, as suggested by previous researchers (Tilman et al. 1997a, Hooper and Vitousek
1997).
Site was the main variable that differentiated the root-associated fungal community.
This was seen from with both DNA directly extracted from roots and by isolating fungal

141

cultures from roots. The dolomitic site had the lowest genotype richness compared to the
other two sites according to both methods. Soil chemistry was a eontributing faetor. In the
two warming studies, differenees in soil chemistry correlated with lower genotype frequency
and richness on the dolomitic site. The differences in soil chemistry may also help explain
why genotype richness was higher from the T-RFLP - OTC study than the eontrol plot o f the
suecession study. According to the ordination analyses, the community based on fungal
eultures had higher variability, but this was probably because there were fewer samples.
However, when the root-associated fungal community was examined at a smaller scale, such
as in the succession study, then the host plant was a bigger factor in defining these fungal
communities.
It remains an open question whether endophytes in these aretie communities replace
AM fungi, as proposed by Bledsoe et al. (1990). Endophytes were eertainly very common
inhabitants o f plants at these high arctic sites, but it is unknown if they are fulfilling
funetional roles similar to AM fungi in other systems.
As with previous fungal community studies based on RFLP analysis (Gardes and
Bruns 1996a, Dahlberg et al. 1997, Jonsson et al. 1999), sporocarps were not indieative o f
which fungi were eolonizing the roots. None o f the genotypes o f fungi found on roots
matched any o f the sporocarps based on either T-RFLP and RFLP analyses.
T-RFLP was a helpful tool in assessing the root-assoeiated fungal eommunity.
Although there were some difficulties, such as interpreting four datasets derived from a
common rDNA fragment, the possible presence o f pseudo-fragments, and different estimates
o f diversity depending upon which restriction enzymes and primers were used, we were able
to use this technique to assess changes in the genotype eumulative frequency and diversity.

142

Other advantages o f this technique were that it allowed comparisons o f T-RFs with RFLPs,
and it allowed detection o f multiple fungi on single root tips.
This thesis has examined how global climate change may affect the root-associated
fungal community. Further studies are needed. Changes may he only beginning, as
indicated by the still limited change in genotype frequency. Genotype richness may change
as warming progresses, and perhaps more effects o f warming may be observed as soil
properties and moisture change with increased warming. As with most community-level
studies on the root-associated fungal community, the functions o f these fungi are not known.
Gene expression techniques using genes involved in particular mycorrhizal functions would
be helpful, such as genes involved in uptake o f ammonium, nitrate or phosphorus.
Further study on the dark- and hyaline-septate fungi is needed. This area is
understudied, and assumptions about this group o f fungi may he inaccurate. Because they
are ubiquitous, more diverse, and more host-specific than previously thought, their function
in the rhizosphere may be more important than currently recognized.

143

VII.

Literature cited

Addy, H. D., S. Hambleton, and R. S. Currah. 2000. Distribution and molecular
characterization o f the root endophyte Phialocephala fortinii along an experimental
gradient in the boreal forest o f Alberta. Mycological Research 104(10): 1213-21.
Agerer, R. 1987-1998. Colour atlas o f eetomycorrhizae. Munich: Einhorn-Verlag,
Schwabisch Gmiind.
Ahlich, K., and T. N. Sieber. 1996. The profusion o f dark septate endophytic fungi in non
ectomycorrhizal fine roots o f forest trees and shrubs. New Phytologist 132(2): 259-70.
Allen, T. R., T. Millar, S. M. Berch, and M. L. Berbee. 2003. Culturing and direct DNA
extraction find different fungi from the same ericoid mycorrhizal roots. New
Phytologist 160: 255-72.
Arft, A. M., M. D. Walker, J. Gurevitch, J. M. Alatalo, M. S. Bret-Harte, M. Dale, M.
Diemer, F. Gugerli, G. H. R. Henry, M. H. Jones, R. D. Hollister, I. S. Jonsdottir, K.
Laine, E. Levesque, G. M. Marion, U. Molau, P. Molgaard, U. Nordenhall, V.
Raszhivin, C. H. Robinson, G. Starr, A. Stenstrom, M. Stenstrom, O. Totland, P. L.
Turner, L. J. Walker, P. J. Webber, J. M. Welker, and P. A. Wookey. 1999.
Responses o f tundra plants to experimental warming: meta-analysis o f the
International Tundra Experiment. Ecological Monographs 69(4): 491-511.
Avaniss-Aghajani, E., K. Jones, D. Chapman, and C. P. Brunk. 1994. A molecular technique
for identification o f bacteria using small subunit ribosomal RNA sequences.
Biotechniques 17(1): 144-49.
Avaniss-Aghajani, E., K. Jones, A. Holtzman, T. Aronson, N. Boian, M. Glover, S. Froman,
and C. F. Brunk. 1996. Moleeular technique for rapid identifieation o f mycobaeteria.
Journal o f Clinical Microbiology 34(1): 98-102.
Baar, J. 1996. The ectomycorrhizal flora o f primary and secondary stands o f Pinus sylvestris
in relation to soil conditions and ectomycorrhizal succession. Journal o f Vegetation
Science 7: 497-504.
Ball, P., and N. Hill. 1994. Vascular plants at Alexandra Fiord. Ecology o f a Polar Oasis.
(eds) J. Svoboda, and B. Freedman. Toronto: Captus University Publications.
Barbour, M. G., J. H. Burk, W. D. Pitts, F. S. Gilliam, and M. W. Schwartz. 1999. Terrestrial
Plant Ecology, 3rd ed. Menlo Park, CA: Benjamin/Cummings.
Batten, D. S., and J. Svoboda. 1994. Plant communities on the uplands in the vicinity o f the
Alexandra Fiord lowland. Ecology o f a Polar Oasis, (eds) J. Svoboda, and B.
Freedman. Toronto: Captus University Publieations.

144

Baxter, J. W., and Dighton J. 2001. Ectomycorrhizal diversity alters growth and nutrient
acquisition o f grey birch (Betula populifolia) seedlings in host-symbiont culture
conditions. New Phytologist 152: 139-49.
Bengtsson, J. 1998. Which species: What kind o f diversity? Which ecosystem function?
Some problems in studies o f relations between biodiversity and ecosystem function.
Applied Soil Ecology 10: 191-99.
Bergero, R., M. Girlanda, G. C. Varese, D. Intili, and A. M. Luppi. 1999.
Psychrooligotrophic fungi from Arctic soils o f Franz Joseph Land. Polar Biology 21 :
361-68.
Billings, W. D., J. O. Luken D. A. Mortensen, and K.M. Peterson KM. 1983. Increasing
atmospheric carbon dioxide: possible effects on arctic tundra. Oecologia 58: 286-89.
Bledsoe, C., P. Klein, and L. C. Bliss. 1990. A survey o f mycorrhizal plants on Truelove
Lowland, Devon Island, NWT, Canada. Canadian Journal o f Botany 68: 1848-56.
Boone, R. D., K. J. Nadelhoffer, J. D. Canary, and J. P. Kaye. 1998. Roots exert a strong
influence on temperature sensitivity o f soil respiration. Nature 396: 570-572.
Brand, T., and V. T. Parker. 1995. Seale and general laws o f vegetation dynamics. Oikos 73:
375-380.
Bridge, P. D., P. J. Roberts, B. M. Spooner, and G. Panchal. 2003. On the unreliability o f
published DNA sequences. New Phytologist 160: 43-48.
Brubaker, L. B., P. M. Anderson, and F. S. Hu. 1995. Aretie tundra biodiversity: a temporal
perspective from Late Quaternary pollen records. Arctic and Alpine Biodiversity:
Patterns, Causes and Ecosystem Consequences, (eds) F.S. Chapin, and C. Komer.
New York: Springer-Verlag.
Bruce, K. D. 1997. Analysis o f mer gene subclasses within bacterial communities in soils and
sediments resolved by fluorescent - PCR - restriction fragment length polymorphism
profiling. Applied and Environmental Microbiology 63(12): 4914-19.
Brundrett, M., N. Bougher, B. Dell, T. Grove, and N. Malajczuk. 1996. Working with
Mycorrhizas in Forestry and Agriculture. ACIAR Monograph 32. Canberra:
Australian Centre for International Agricultural Research.
Bruns, T. D. 1995. Thoughts on the process that maintain local species diversity o f
ectomycorrhizal fungi. Plant and Soil 170: 63-73.
Cardinale, B. J., M. A. Palmer, and S. L Collins. 2002. Species diversity enhances ecosystem
functioning through interspecific facilitation. Nature 415: 426-29.
Cazares, L. 1992. "Mycorrhizal fungi and their relationship to plant succession in subalpine
habitats." Ph.D. Thesis., Oregon State University.
145

Chapin, F. S., and C. Komer. 1994. Arctic and alpine biodiversity - patterns, causes and
ecosystem consequences. Trend Ecology and Evolution 9: 45-47.
. 1995. Patterns, causes, changes, and consequences o f biodiversity in arctic and
alpine ecosystems. Arctic and Alpine Biodiversity: Patterns, Causes and Ecosystem
Consequences, (eds) F.S. Chapin, and C. Komer. New York: Springer-Verlag.
Chapin, F. S., G. R. Shaver, A. E. Giblin, K. J. Nadelhoffer, and J. A. Laundre. 1995.
Responses of arctic tundra to experimental and observed changes in climate. Ecology
76(3): 694-711.
Chapin, F. S., L. R. Walker, C. L. Fastie, and L. C. Sharman. 1994. Mechanism o f primary
succession following deglaciation at Glacier Bay, Alaska. Ecological Monographs
64(2): 149-75.
Chapin, F. S., E. S. Zavaleta, V. T. Eviner, R. L. Naylor, P. M. Vitousek, H. L. Reynolds, D.
U. Hooper, S. Lavorel, O. E. Sala, S. E. Hobble, M. C. Mack, and S. Diaz. 2000.
Consequences o f changing biodiversity. Nature 405(6783): 234-42.
Chin, K. J., T. Lukow, and R. Conrad. 1999. Effect o f temperature on stmcture and function
of the methanogenic archaeal community in an anoxic rice field soil. Applied and
Environmental Microbiology 65(6): 2341-49.
Clement, B. G., L. E. Kehl, K. L. Debord, and C. L. Kitts. 1998. Terminal restriction
fragment pattems (T-RFLPs), a rapid, PCR-based method for the comparison o f
complex bacterial communities. Journal o f Microbiological Methods 31: 135-42.
Colwell, R. K. 1997. "Estimates : Statistical estimation o f species richness and shared species
from samples. Version 6.Obi." Web page. Available at User's Guide and application
published at: http://viceroy eeb.uconn.edu/estimates.
Dahlberg, A. 2001. Community ecology o f ectomycorrhizal fungi: an advancing
interdisciplinary field. New Phytologist 150(3): 555-62.
Dahlberg, A., L. Jonsson, and J.-E. Nylund. 1997. Species diversity and distribution of
biomass above and below ground among ectomycorrhizal fungi in an old-growth
Norway Spruce forest in south Sweden. Canadian Journal o f Botany 75: 1323-35.
Dalpé, Y., and S. G. Aiken. 1998. Arbuscular mycorrhizal fungi associated with Festuca
species in the Canadian high arctic. Canadian Journal o f Botany 76(11): 1930-1938.
Danielson, R. M. 1984. Ectomycorrhizal associations in jack pine stands in northeastern
Alberta. Canadian Journal o f Botany 62: 932-39.
David, J. C. 1997. Contribution to the Systematics o f Cladosporium. A Revision o f the
Fungi Previously Referred to Heterosporium. New York: CAB International.

146

Dickie, I. A., B. Xu, and R. T. Koide. 2002. Vertical niche differentiation o f eetomyeorrhizal
hyphae in soil as shown by T-RFLP analysis. New Phytologist 156(3): 527-35.
Dighton, J., and P. A. Mason. 1985. Mycorrhizal dynamics during forest tree development.
Developmental Biology o f Higher Fungi, (eds) D. Moore, L. A. Casselton, D A .
Wood, and J.C. Frankland. Cambridge: Cambridge University Press.
Drozd, P., and V. Novotny. 1999. “PowerNiche: Niche division models for community
analysis.” Webpage. Available at: www.entu.eas.cz/png/PowerNiehe.
Edlund, S. A. 1992. Climate change and its effects on Canadian arctic plant communities.
Arctic Environment: Past, Present, and Future, (eds) M.-K. Woo, and D. J. Gregor.
Hamilton, Ontario: McMaster University.
Egbert, M., and M. W. Friedrich. 2003. Formation o f pseudo-terminal restriction fragments, a
PCR-related bias affecting terminal restriction fragment length polymorphism
analysis o f microbial community structure. Applied and Environmental Microbiology
69(5): 2555-63.
Egger, K. N. 1995. Molecular analysis o f ectomycorrhizal fungal communities. Canadian
Journal o f Botany 13{sapf[. 1): S1415-S1422.
Eissenstat, D. M., C. E. Wells, R. D. Yanai, and J. L. Whitbeck. 2000. Building roots in a
changing environment: implications for root longevity. New Phytologist 147: 33-42.
Erland, S., T. Jonsson, S. Mahmood, and R. D. Finlay. 1999. Below-Ground ectomycorrhizal
community structure in two Picea abies forests in southern Sweden. Scandinavian
Journal o f Forest Research 14(3): 209-17.
Fauth, J. E., J. Bernardo, M. Camara, W. J. Resetarits, J. Yanbuskirk, and S. A. Mccollum.
1996. Simplifying the jargon o f community ecology: a conceptual approaeh.
American Naturalist 147(2): 282-86.
Felsenstein, J. 2004. PHYLIP (Phytogeny inference package) version 3.6. Distributed by the
author. Department o f Genome Sciences, University o f Washington, Seattle.
Field, C. B., F. S. Chapin, P. A. Matson, and H. A. Mooney. 1992. Response o f terrestrial
ecosystems to the changing atmosphere: a resouree-based approach. Annual Review
o f Ecological Systems 23: 201-35.
Finlay, R. D., and D. J. Read. 1986. The structure and function of the vegetative myeelium of
ectomycorrhizal plants. New Phytologist 103: 157-65.
Fitter, A. H., A. Heinemeyer, and P. L Staddon. 2000. The impaet o f elevated CO 2 and global
climate change on arbuscular mycorrhizas: a mycocentric approach. New Phytologist
147: 179-87.

147

Fleming, L.V., J. W. Deacon, F. T. Last, and S. J. Donaldson. 1984. Influence o f propagating
soil on the mycorrhizal succession o f birch seedlings transplanted to a field site.
Transactions o f the British Mycological Society 82: 707-711.
Frankland, J. C. 1998. Fungal succession - unravelling the unpredictable. Mycological
Research 102(1): 1-15.
Freedman, B., J. Svoboda, and G. FI. R. Henry. 1994. Alexandra Fiord - an ecological oasis
in the polar desert. Ecology o f a Polar Oasis, (eds) J. Svoboda, and B. Freedman.
Toronto: Captus University Publications.
Gardes, M., and T. D. Bruns. 1993. ITS primers with enhanced specificity for
basidiomycetes-application to the identification o f mycorrhizae and rusts. Molecular
Ecology 2\ 113-18.
----------. 1996a. Community structure o f ectomycorrhizal fungi in Pinus muricata forest:
above- and below-ground views. Canadian Journal o f Botany 74: 1572-83.
. 1996b. ITS-RFLP matching for identification o f fungi. Methods in Molecular
Biology. J.P. Clapp (ed). Species Diagnostic Protocols: PCR and Other Nucleic Acid
Methods, 50. Totowa, NJ: Humana Press, Inc.
Gardes, M., J.-Y. Charcosset, and A. Dahlberg. 2000. High species riehness and dominance
of Cortinarius taxa in belowground communities o f ectomycorrhizal fungi from the
Canadian Arctic. 2000 Mycological Society o f America Annual Meeting. Burlington,
VT: University o f VT.
Gardes, M., and A. Dahlberg. 1996. Mycorrhizal diversity in arctic and alpine tundra: an
open question. New Phytologist 133(1): 147-57.
Garrett, S. D. 1981. Soil Fungi and Soil Fertility, 2nd edition. Toronto: Pergamon Press.
Gehring, C. A., T. C. Theimer, T. G. Whitham, and P. Keim. 1998. Ectomycorrhizal fungal
community structure o f pinyon pines growing in two environmental extremes.
Ecology 19{5)-. 1562-72.
GenBank. www.ncbi.nlm.nih.gov.
Goodman, D. M., D. M. Durall, J. A. Trofymow, and S. M. Berch. 1996. A Manual o f
Concise Descriptions o f North American Eetomycorrhizae. Sidney, BC: Mycologue
Publications.
Gotelli, N. J. and Entsminger, G. L. 2001. "EcoSim: Null models software for ecology.
Version 7.0." Web page. Available at
http://homepages.together.net/~gentsmin/ecosim.htm.
Grondona, L, E. Monte, 1. Garcia-Acha, and B. C. Sutton. 1997. Pyrenochaeta dolichi: an
example of a confusing species. Mycological Research 101(11): 1405-8.
148

Grünig, C. R., B. A. McDonald, T. N. Sieber, S. O. Rogers, and O. Holdenrieder. 2004.
Evidence for subdivision o f the root-endophyte Phialocephala fortinii into eryptic
species and recombination within species. Fungal Genetics and Biology 41: 676-87.
Grünig, C. R., T. N. Sieber, and O. Holdenrieder. 2001. Characterization o f dark-septate
endophytic fungi (DSE) using inter-simple-sequence-repeat-anchor and polymerase
chain reaction (ISSR-PCR) amplification. Mycological Research 105(1): 24-32.
----------. 2002a. Genetic variability among strains o f Phialocephala fortinii and phylogenetic
analysis o f the genus Phialocephala based on rDNA ITS sequence eomparisons.
Canadian Journal o f Botany 80: 1239-49.
. 2002b. Spatial distribution o f dark septate endophytes in a confined forest plot.
Mycological Research 106(7): 832-40.
Gunde-Cimeran, N., S. Sonjak, P. Zalar, J. C. Frisvad, B. Diderichsen, and A. Plemenitas.
2003. Extremophilic fungi in arctic ice: a relationship between adaptation to low
temperature and water activity. Physics and Chemistry o f the Earth 28: 1273-78.
Haselwandter, K. 1987. Mycorrhizal infection and its possible ecological significance in
climatieally and nutritionally stressed alpine plant communities. Angewandte Botanik
61: 107-14.
Harney, S., S. O. Rogers, and C. J. K. Wang. 1997. Molecular characterization o f
dematiaceous root endophytes. Mycological Research 101(11): 1397-404.
Havstrom M., T.V. Callaghan, and S. Jonasson. 1993. Differential growth responses o f
Cassiope tetragona, an arctic dwarf shrub, to environmental perturbations among
three contrasting high- and subarctic sites. Oikos 66: 389-402.
He, H. S., D. J. Mladenoff, and E. J. Gustafson. 2002. Study o f landscape change under
forest harvesting and climate warming-induced fire disturbance. Forest Ecology and
Management 155: 257-270.
Held, B. W., J. A. Jurgens, B. E. Arenz, S. M. Duncan, R. L. Farrell, and R. A. Blanchette.
2004. Environmental faetors influencing microbial growth inside the historic
expedition huts o f Ross Island, Antaretiea. International Biodeterioration and
Biodégradation 55: 45-53.
Helm, D. J., E. B. Allen, and J. M. Trappe. 1996. Mycorrhizal chronosequenee near Exit
Glaeier, Alaska. Canadian Journal o f Botany 74: 1496-506.
----------. 1999. Plant growth and ectomycorrhiza formation by transplants on deglaciated land
near Exit Glacier, Alaska. Mycorrhiza 8: 297-304.
Henry, G. H. R., and U. Molau. 1997. Tundra plants and climate ehange: the International
Tundra Experiment (ITEX). Global Change Biology 3(Suppl. 1): 1-9.

149

Hobbie, S. E., and F. S. Chapin. 1998. The response o f tundra plant biomass, aboveground
production, nitrogen, and CO 2 flux to experimental warming. Ecology 79(5): 152644.
Hoeksema, J. D., and M. Kummel. 2003. Ecological persistence o f the plant-mycorrhizal
mutualism: a hypothesis from species coexistence theory. American Naturalist 162
(supplement): S40-S40.
Hooper, D. U., F. S. Chapin, J. J. Ewel, A. Hector, P. Inchausti, S. Lavorel, J. H. Lawton, D.
M. Lodge, M. Loreau, S. Naeem, B. Schmid, H. Setala, A. J. Symstad, J.
Vandermeer, and D. A. Wardle. 2005. Effects o f biodiversity on ecosystem
functioning: a consensus o f current knowledge. Ecological Monographs 75(1): 3-35.
Hooper, D. U., and P. M. Vitousek. 1997. The effects o f plant composition and diversity on
ecosystem processes. Science 277: 1302-5.
Horton, T. R., and T. D. Bruns. 1998. Multiple-host fungi are the most frequent and abundant
ectomycorrhizal types in a mixed stand o f Douglas-fir and bishop pine. New
Phytologist 139:331-39.
. 2001. The molecular revolution in ectomycorrhizal ecology: peeking into the black
box. Molecular Ecology 10: 1855-71.
Horton, T. R., T. D. Bruns, and V. T. Parker. 1999. Ectomycorrhizal fungi associated with
Arctostaphylos contribute to Pseudotsuga menziesii establishment. Canadian Journal
o f Botany 77: 93-102.
Jones, M. H., C. Bay, and U. Nordenhall. 1997. Effects o f experimental warming on arctic
willows (Salix spp.): a comparison o f responses from the Canadian High Arctic,
Alaskan Arctic, and Swedish Subarctic. Global Change Biology 3(suppl. 1): 55-60.
Jonsson, L., A. Dahlberg, M. C. Nilsson, O. Kârén, and O. Zackrisson. 1999. Continuity o f
ectomycorrhizal fungi in self-regenerating boreal Pinus sylvestris forests studied by
comparing mycobiont diversity on seedlings and mature trees. New Phytologist 142:
151-62.
Jumpponen, A. 1999. Spatial distribution o f discrete RAPD phenotypes o f a root endophytic
fungus, Phialocephala fortinii, at a primary successional site on a glacier forefront.
New Phytologist 141: 333-44.
Jumpponen, A. 2003. Soil fungal community assembly in a primary suceessional glacier
forefront ecosystem as inferred from rDNA sequence analyses. New Phytologist 158:
569-78.
Jumpponen, A., K. Mattson, J. M. Trappe, and R. Ohtonen. 1998. Effects o f established
willows on primary succession on Lyman glacier forefront. North Cascade range,
Washington, USA: Evidence for simultaneous canopy inhibition and soil facilitation.
Arctic and Alpine Research 30(1): 31-39.
150

Jumpponen, A., and J. M. Trappe. 1998. Performance o f Pinus conforta inoculated with two
strains o f root endophytic fungus, Phialocephala fortinii: effects o f synthesis system
and glucose concentration. Canadian Journal o f Botany 76(7): 1205-13.
Kârén, 0 ., N. Hogberg, A. Dahlberg, L. Jonsson, and J. E. Nylund. 1997. Inter- and
intraspecific variation in the ITS region o f rDNA o f ectomycorrhizal fungi in
Fennoscandia as detected by endonuclease analysis. New Phytologist 136: 313-25.
Kârén, O., and J. E. Nylund. 1997. Effects o f ammonium sulphate on the community
structure and biomass o f ectomycorrhizal fungi in a Norway spruce stand in
sovLihwQStcm'&wQàQn. Canadian Journal o f Botany 75(10): 1628-42.
Kernaghan, G. 2001. Ectomycorrhizal fungi at tree line in the Canadian Rockies II.
Identification o f ectomycorrhizae by anatomy and PCR. Mycorrhiza 10(5): 217-29.
Kernaghan, G., L. Sigler, and D. Khasa. 2003. Mycorrhizal and root endophytic fungi o f
containerized Picea glauca seedlings assessed by rDNA sequence analysis. Microbial
Ecology A5{2): 128-36.
Khetmalas, M. B., K. N. Egger, H. B. Massicotte, L. E. Tackaberry, and M. J. Clapperton.
2002. Bacterial diversity associated with subalpine fir {Abies lasiocarpa)
ectomycorrhizae following wildfire and salvage-logging in central British Columbia.
Canadian Journal o f Microbiology 48(7): 611-25.
Klamer, M., M. S. Roberts, L. H. Levine, B. G. Drake, and J. L. Garland. 2002. Influence o f
elevated CO 2 on the fungal community in a coastal scrub oak forest soil investigated
with terminal-restriction fragment length polymorphism analysis. Applied and
Environmental Microbiology 68(9): 4370-4376.
Kohn, L. M., and E. Stasovski. 1990. The mycorrhizal status o f plants at Alexandra Fiord,
Ellesmere Island, Canada, a high arctic site. Mycologia 82(1): 23-35.
Koide, R. T. 2000. Functional complementarity in arbuseular mycorrhizal symbiosis. New
Phytologist 147: 233-35.
Kranabetter, J. M., and T. Wylie. 1998. Ectomycorrhizal community structure across forest
openings on naturally regenerated western hemlock seedlings. Canadian Journal o f
Botany 16(f): 189-96.
Kropp, B. R., and J. M. Trappe. 1982. Ectomycorrhizal fungi o f Tsuga heterophylla.
Mycologia 74: 479-88.
Labine, C. 1994. Meteorology and climatology o f the Alexandra Fiord Lowland. Ecology o f
a Polar Oasis, (eds) J. Svoboda, and B. Freedman. Toronto: Captus University
Publications.
Leake, J. R. 2001. Is diversity o f ectomycorrhizal fungi important for ecosystem function?
New Phytologist 152(1): 1-3.
151

Lee, S. B., and J. W. Taylor. 1990. Isolation o f DNA from fungal mycelia and single spores.
PCR Protocols: A Guide to Methods and Applications, (eds) M.A. Innis, D.H.
Gelfand, J.J. Sninsky, and T.J. White. Berkeley: Aeademie Press, Ine.
Liu, W.-T., T. L. Marsh, H. Cheng, and L. J. Forney. 1997. Characterization o f microbial
diversity by determining terminal restriction fragment length polymorphisms o f genes
encoding 16s rRNA. Applied and Environmental Microbiology 63(11): 4516-22.
Looijen R.C., and J. van Andel. 1999. Ecological communities: conceptual problems and
definitions. Perspectives in Plant Ecology, Evolution, and Systematics 2(2): 210-222.
Loreau, M., and A. Hector. 2001. Partitioning selection and complementarity in biodiversity
experiments. Nature 412: 72-76.
Loreau, M., S. Naeem, P. Inchausti, J. Bengtsson, J. P. Grime, A. Hector, D. U. Hooper, M.
A. Huston, D. Raffaelli, B. Schmid, D. Tilman, and D. A. Wardle. 2001. Ecology biodiversity and ecosystem functioning: current knowledge and future challenges.
294(5543): 804-8.
Magurran, A. E. 1988. Ecological Diversity and Its Measurements. Princeton, NJ: Princeton
University Press.
. 2004. Measuring Biological Diversity. Oxford, UK: Blackwell Science Ltd.
Magurran, A. E., and P. A. Henderson. 2003. Explaining the excess o f rare species in natural
species abundance distribution. Nature 422: 714-16.
Mah, K., E. E. Tackaberry, K. N. Egger, and H. B. Massicotte. 2001. The impacts of
broadcast burning after clear-cutting on the diversity o f ectomycorrhizal fungi
associated with hybrid spruce seedlings in central British Columbia. Canadian
Journal o f Forest Research 31(2): 224-235.
Marion, G. M., G. H. R. Henry, D. W. Freekman, J. Johnstone, G. Jones, M. H. Jones, E.
Levesque, U. Molau, P. Molgaard, A. N. Parsons, J. Svoboda, and R. A. Virginia.
1997. Open-top designs for manipulating field temperature in high-latitude
ecosystems. Global Change Biology 3: 20-32.
Marsh, T. L., W.-T. Liu, L. J. Forney, and H. Cheng. 1998. Beginning a molecular analysis
of the eukaryal community in activated sludge. Water Science Technology 37(4-5):
455-60.
Martin, F., S. Perotto, and P. Bonfante. 2000. Mycorrhizal fungi: a fungal community at the
interface between soil and roots. The Rhizosphere. Biochemistry and Organic
Substances at the Soil-Plant Interface, (eds) R. Pinton, Z. Varanini, and P.
Nannipieri. New York: Marcel Dekker, Inc.

152

Martmez-Murcia, A. J., S. G. Acinas, and F. Rodriguez-Valera. 1995. Evaluation o f
prokaryotic diversity by restriction digestion o f 16s rDNA directly amplified from
hypersaline environments. FEMS Microbiology Ecology 17: 247-56.
Marx, D.H. 1969. The influence o f ectotrophic mycorrhizal fungi on the resistance o f pine
roots to pathogenic infections. I. Antagonism o f mycorrhizal fungi to root pathogenic
fungi and soil bacteria. Phytopathology 59: 153-63.
Matthews, J. A. 1992. The ecology o f recently deglaciated terrain: a geoecological approach
to glacier forelands and primary succession. Cambridge: Cambridge University Press.
McCune, B., and J. B. Grace. 2002. Analysis o f Ecological Communities. Gleneden Beach,
OR: MjM Software Design.
McCune, B., and M. J. Mefford. 1999. PC-ORD: Multivariate Analysis o f Ecological Data,
version 4.25. Gleneden Beach, OR: MjM Software.
McKane, R. B., L. C. Johnson, G. R. Shaver, K. J. Nadelhoffer, E. B. Rastetter, B. Fry, A. E.
Giblin, K. Kielland, B. L. Kwiatkowski, J. A. Laundre, and G. Murray. 2002.
Resource-based niches provide a basis for plant species diversity and dominance in
Arctic tundra. Nature 415(6867): 68-71.
McKane, R. B., E. B. Rastetter, G. R. Shaver, K. J. Nadelhoffer, A. E. Giblin, J. A. Laundre,
and F. S. Chapin. 1997. Climatic effects on tundra carbon storage inferred from
experimental data and model. Ecology 78(4): 1170-1187.
Michelsen A, I. K. Schmidt, S. Jonasson, C. Quarmby, and D. Sleep. 1996. Leaf *^N
abundance o f subarctic plants provide field evidence that ericoid, ectomycorrhizal,
and non- and arbuseular mycorrhizal species access different sources o f nitrogen.
Oecologia 105: 53-63.
Monreal, M., S. A. Berch, and M. Berbee. 1999. Molecular diversity o f ericoid mycorrhizal
fungi. Canadian Journal o f Botany 11 \ 1580-1594.
Muc, M., J. Svoboda, and B. Freedman. 1994. Soils o f an extensively vegetated polar desert
oasis, Alexandra Fiord, Ellesmere Island. Ecology o f a Polar Oasis, (eds) J.
Svoboda, and B. Freedman. Toronto: Captus University Publications.
Murray, D. F. 1995. Causes o f arctic plant diversity: origin and evolution. Arctic and Alpine
Biodiversity: Patterns, Causes and Ecosystem Consequences, (eds) F. S. Chapin, and
C. Komer. New York: Springer-Verlag.
Naeem, S. 2002. Ecosystem consequences o f biodiversity loss: the evolution o f a paradigm.
Ecology
1537-52.
Naeem, S., F. S. Chapin, R. Constanza, P. R. Ehrlich, F. B. Golley, D. U. Hooper, J. H.
Lawton, R. V. O'Neill, H. A. Mooney, O. E. Sala, A. J. Symstad, and D. Tilman.

153

1999. Biodiversity and ecosystem functioning: maintaining natural life support
processes. Issues in Ecology 4: 1-11.
Naeem, S., L. J. Thompson, S. Lawler, J. H. Lawton, and R. M. Woodfin. 1995. Empirical
evidence that declining species diversity may alter the performance o f terrestrial
ecosystems. Philosophical Transactions o f the Royal Society o f London Series B,
Biological Sciences 347(1321): 249-62.
Nara, K., H. Nakaya, B. Y. Wu, Z. H. Zhou, and T. Hogetsu. 2003. Underground primary
succession o f ectomycorrhizal fungi in a volcanic desert on Mount Fuji. New
Phytologist 159(3): 743-56.
Newman, E. I. 1988. Mycorrhizal links between plants: their functioning and ecological
sigmfxcancQ. Advances in Ecological Research 18: 243-70.
Oechel, W. C., S. J. Hastings, G. Yourlitis, M. Jenkins, G. Riechers, andN . Grulke. 1993.
Recent change of Arctic tundra ecosystems from a net carbon dioxide sink to a
source. Nature 361: 520-523.
Overpeck, J., K. Hughen, D. Hardy, R. Bradley, R. Case, M. Douglas, B. Firmey, K.
Gajewski, G. Jacoby, A. Jennings, S. Lamoureux, A. Easca, G. MacDonald, J. Moore,
M. Retelle, S. Smith, A. Wolfe, and G. Zielinski. 1997. Arctic environmental changes
o f the last four centuries. Science 278: 1251-56.
Palmer, M. W., and P. S. White. 1994. On the existence o f ecological communities. Journal
o f Vegetation Science 5: 279-82.
Parker, V. T. 2001. Conceptual problems and scale limitations o f defining ecological
communities: a critique o f the Cl concept (Community o f individuals). Perspectives
in Plant Ecology, Evolution, and Systematics A{2)\ 80-96.
Paul, E. A., and F. E. Clark. 1996. Soil Microbiology and Biochemistry, 2nd edition. San
Diego, CA: Academic Press.
Perry, D. A., M. P. Amaranthus, J. G. Borchers, S. L. Borchers, and R. E. Brainerd. 1989.
Bootstrapping in ecosystems. Bioscience 39(4): 230-237.
Perry, D. A., J. G. Borchers, S. E. Borchers, and M. P. Amaranthus. 1990. Species migrations
and ecosystem stability during climate change: the belowground connection.
Conservation Biology 4(3): 266-74.
Pickett, S. T. A., J. Kolasa, J. Armesto, and S. L. Collins. 1989. The ecological concept o f
disturbance and its expression at various hierarchical levels. Oikos 54: 129-136.
Pickett, S. T. A., V. T. Parker, and P. L. Fiedler. 1992. The new paradigm in ecology:
implications for conservation biology above the species level. Conservation Biology:
The Theory and Practice o f Nature Conservation, Preservation, and Management.
(eds) P. E. Fiedler, and S. Jain. New York: Chapman and Hall.
154

Piercey, M. M., S. W. Graham, and R. S. Currah. 2004. Patterns o f genetic variation in
Phialocephala fortinii across a broad latitudinal transect in Canada. Mycological
Research 108(8): 955-64.
Pregitzer, K. S., J. S. King, A. J. Burton, and S. E. Brown. 2000. Responses o f tree fine roots
to temperature. New Phytologist 147: 105-15.
Read, D. J. 1993. Mycorrhiza in plant communities. Advances in Plant Pathology, (eds) D.
S. Ingram and P. H. Williams. Vol. 9. New York: Academic Press.
Robinson, C. H. 2001. Cold adaptation in Arctic and Antarctic fungi. New Phytologist 151:
341-53.
Rygiewicz, P. T., and C. P. Andersen. 1994. Mycorrhizae alter quality and quantity o f carbon
allocated below ground. Nature 369: 58-60.
Rygiewicz, P. T., K. J. Martin, and A. R. Tuininga. 2000. Morphotype community structure
of ectomycorrhizas on Douglas fir {Pseudotsuga menziesii Mirb. Franco) seedlings
grown under elevated atmospheric CO 2 and temperature. Oecologia 124: 299-308.
Rykiel, E. J. 1985 Towards a definition o f ecological disturbance. Australian Journal o f
Ecology 10: 361-365.
Schadt, C. W., R. B. Mullen, and K. Schmidt. 2001. Isolation and phylogenetic identification
of a dark-septate fungus associated with the alpine plant Ranunculus adoneus. New
Phytologist 150: 747-55.
Schimel, D. S. 1995. Terrestrial ecosystems and the carbon cycle. Global Change Biology 1:
77-91.
Schneider S., D. Roessli, and L. Excoffier. 2000. Arlequin ver. 2.000: A software for
population genetics data analysis. University o f Geneva, Switzerland: Genetics and
Biometry Laboratory.
Schwartz, M. W., C. A. Brigham, J. D. Hoeksema, K. G. Lyons, M. H. Mills, and P. J. van
Mantgem. 2000. Linking biodiversity to ecosystem function: implications for
conservation biology. Oecologia 122: 297-305.
Shaver, G. R., W. D. Billings, F. S. Chapin, A. L. Giblin, K. J. Nadelhoffer, W. C. Oechel,
and L. B. Rastetter. 1992. Global change and the carbon balance o f arctic ecosystems.
BioScience. 42(6): 433-41.
Shaver, G. R., J. Canadell, F. S. Chapin, J. Gurevitch, J. Harte, G. Henry, P. Ineson, S.
Jonasson, J. Melillo, L. Pitelka, and L. Rustad. 2000. Global warming and terrestrial
ecosystems: a conceptual framework for analysis. Bioscience 50(10): 871-82.

155

Shen, W., R. E. Tuleya, and I. Ginis. 2000. A sensitivity study of the thermodynamic
environment on GFDL model hurricane intensity; implications for global warming.
Journal o f Climate 13(1): 109-121.
Simard, S. W., D. A. Perry, M. D. Jones, D. D. Myrold, D. M. Durall, and R. Molina. 1997.
Net transfer o f carbon between ectomycorrhizal tree species in the field. Nature 388:
579-582.
Smith, S. E., and D. J. Read. 1997. Mycorrhizal Symbiosis. 2nd ed. San Diego, CA:
Academic Press.
Stenstrom, M., F. Gugerli, and G. H. R. Henry 1997. Response o f Saxifraga oppositifolia L.
to simulated climate change at three contrasting latitudes. Global Change Biology
3(suppl. 1): 44-54.
Stoyke, G., K. N. Egger, and R. S. Currah. 1992. Characterization o f sterile endophytic fungi
from the mycorrhizae o f subalpine plants. Canadian Journal o f Botany 70: 2009-16.
Sturm, M., C. Racine, and K. Tape. 2001. Increasing shrub abundance in the Arctic. Nature
411: 546-47.
Svoboda, J., and G. H. R. Henry. 1987. Succession in marginal arctic environments. Arctic
and Alpine Research 19(4): 373-84.
Taber, R. A., and W. A. Taber. 1984. Evidence for ectomycorrhizal fungus-mediated nutrient
transfer between Pinus and Tradescantia. Forest Science 30(4): 892-96.
Taylor, J. W., D. J. Jacobson, and M. C. Fisher. 1999. The evolution o f asexual fungi:
reproduction, spéciation, and classification. Annual Review o f Phytopathology 37:
197-246.
Termorshuizen, A. J. 1991. Succession o f mycorrhizal fungi in stands o f Pinus sylvestris in
the Netherlands. Journal o f Vegetation Science 2: 555-64.
Tilman, D., J. Knops, D. Wedlin, P. Reich, M. Ritchie, and E. Siemann. 1997a. The influence
o f functional diversity and composition on ecosystem processes. Science 277: 13001302.
Tilman, D., C. L. Lehman, and K. T. Thomson. 1997b. Plant diversity and ecosystem
productivity: theoretical considerations. Proceedings o f the National Academy o f
Science USA 94: 1857-61.
Titus, J. H., and R. del Moral. 1998. The role o f mycorrhizal fungi and microsites in primary
succession on Mount St. Helens. American Journal o f Botany 85(3): 370-375.
Tokeshi, M. 1999. Species Coexistence: Ecological and Evolutionary Perspectives. Oxford,
UK: Blackwell Science Ltd.

156

Tosi, S., B. Casado, R. Gerdol, and G. Caretta. 2002. Fungi isolated from Antarctic mosses.
Polar Biology 25: 262-68.
Trappe, J. M. 1962. Fungus associates o f ectotrophic mycorrhizae. Botanical Review
October-December: 538-605.
. 1988. Lessons from alpine fungi. Mycologia 80(1): 1-10.
Trappe, J. M., and D. L Luoma. 1992. The ties that bind: fungi in ecosystems. The Fungal
Community: Its Organization and Role in the Ecosystem. 2nd. ed., (eds) G. C. Carroll,
and D. T. Wicklow. New York: Marcel Dekker, Inc.
Treseder, K. K., and M. F. Allen. 2000. Mycorrhizal fungi have a potential role in carbon
storage under elevated CO 2 and nitrogen deposition. New Phytologist 147: 189-200.
Trowbridge, J., and A. Jumpponen. 2004. Fungal colonization o f shrub willow roots at the
forefront o f a receding glacier. Mycorrhiza 14(5): 283-93.
Totsch, M., E. Brommelkamp, A. Stiieker, M. Fille, R. Gross, P. Wiesner, K. W. Schmid,
W. Bocker, and B. Dockhom-Dwomiczak. 1995. Identification o f mycobacteria to
the species level by automated restriction enzyme fragment length polymorphism
analysis. Virchows Archive A21 \ 85-89.
van der Heijden, M. G. A., J. N. Klironomos, M. Ursic, P. Moutoglis, R. Streitwolf-Engel, T.
Boiler, A. Wiemken, and 1. R. Sanders. 1998. Mycorrhizal fungal diversity
determines plant biodiversity, ecosystem variability and productivity. Nature
396(6706): 69-72.
van der Heijden, E. W., and M. Vosatka. 1999. Mycorrhizal associations o f Salix repens L.
communities in succession o f dune ecosystems. II. Mycorrhizal dynamics and
interactions o f ectomycorrhizal and arbuseular mycorrhizal fungi. Canadian Journal
o f Botany 11 {\2)-. 1833-41.
Vare, H., M. Vestberg, and S. Enrôla. 1992. Mycorrhiza and root-associated fungi in
Spitsbergen. Mycorrhiza 1: 93-104.
Verkley, G. J. M., J. D. Zijlstra, R. C. Summerbell, and F. Berendse. 2003. Phytogeny and
taxonomy o f root-inhibiting Cryptosporiopsis species, and C. rhizophila sp. nov., a
fungus inhabiting roots o f several Ericaceae. Mycological Research 107(6): 689-98.
Walker, B., A. Kinzig, and J. Langridge. 1999. Plant attributes, diversity, resilience, and
ecosystem function: the nature and significance o f dominant and minor species.
Ecosystems 2(2): 95-113.
Weidemann, H. M., A. Holst-Jensen, and T. Schumacher. 1998. Demonstration o f Helvella
ectomycorrhizae on Dryas and Salix hosts by means o f taxon-selective Helvella based nrDNA - primers. 2nd International Conference on Mycorrhiza Uppsala,
Sweden: University o f Uppsala.
157

White, P. S. and A. Jentsch. 2001. The search for generality in studies o f disturbance and
ecosystem dynamics. Progress in Botany 62: 399-450.
White, T. J., T. D. Bruns, S. Lee, and J. Taylor. 1990. Amplification and direct sequencing of
fungal ribosomal RNA genes for phylogenetics. PCR protocols: a guide to methods
and applications, (eds) M. A. Innis, D. H. Gelfand, J. J. Sninsky, and T. J. White.
New York, NY: Academic Press.
Widden, P., and D. Parkinson. 1979. Populations o f fungi in a high arctic ecosystem.
Canadian Journal o f Botany 57: 2408-17.
Wilberforce, E. M., L. Boddy, R. Griffiths, and G. W. Griffith. 2003. Agricultural
management affects communities o f culturable root-endophytic fungi in temperate
grasslands. Soil Biology and Biochemistry 35: 1143-54.
Wilcox, H. E. 1971. Morphology o f ectendomycorrhizae in Pinus resinosa. Mycorrhizae.
(ed) E. Hacskaylo.USDA Misc. Publ. 1189.
Wilson, B. 1991. Does vegetation science exist? Journal o f Vegetation Science 2: 289-90.
Wirsel, S. G. R., C. Runge-Frobose, D. G. Ahrén, E. Remen, R. P. Oliver, and K. W.
Mendgen. 2002. Four or more species o f Cladosporium sympatrically colonize
Phragmites australis. Fungal Genetics and Biology 35: 99-113.
Wu, B., K. Nara, and T. Hogetsu. 1999. Competition between ectomycorrhizal fungi
colonizing Pinus densiflora. Mycorrhiza 9(3): 151-59.
Zak, D. R., W. E. Holmes, D. C. White, A. D. Peacock, and D. Tilman. 2003. Plant diversity,
soil microbial communities, and ecosystem function: are there any links? Ecology
84(8): 2042-50.
Zhang, Y-H., and Zhuang, W-Y. 2004. Phylogenetic relationships o f some members in the
genus Hymenoscyphus (Ascomycetes, Helotiales). Nova Hedwigia 78: 475-484.

158

Table A2.1 2-way, nested ANOVA
Source of
Effect
Degrees
variation
(fixed/
of
random)
freedom

able for effects o f site, treatment, and treatment replieate on soil properties
Sum o f squares
F
Mean square
p-value

T3

Intercept
Site

Fixed
Fixed

Treatment
Site*
Treatment
Replicate
(treatment)

Fixed
Fixed
Random

Error

Exch Mn

1

Exch
Mg
161.502

0.004

Exch
Mg
161.502

Exch Mn

Exch Mg

Exch Mn

2
1

7.654
0.052

0.002
<0.001

3.827
0.052

2

0.701

<0.001

4

0.590

8

3.921

113.311

Exch
Mg
<0.001

Exch
Mn
<0.001

0.004
<0.001

1094.729

<0.001

7.808
0.351

7.257

0.013

0.016

0.218

0.586

0.665

0.350

<0.001

0.715

0.018

0.518

0.982

<0.001

0.148

<0.001

0.301

0.272

0.869

0.888

<0.001

0.490

<0.001

sa

R"

S'

S'

S?
w

Table A2.1 2-way, nested ANOVA Table for effects o f site, treatment, and treatment replicate on soil properties (cont’d)
Source of
Effect
Degrees Sum o f squares
F
Mean square
p-value
variation
(fixed/
of
random)
freedom

Intercept
Site

Fixed
Fixed

Treatment
Site*Treatm
ent
Replieate
(treatment)

Fixed
Fixed

Error

Random

pH

C:N ratio

pH

C:N ratio

pH

1
2
1

770.936
22.653
0.062

39454.08
40275.61

39454.08
20137.80

9578.161
81.752

479.70

770.936
11.327
0.062

479.70

2

0.100

858.60

0.050

4

0.322

2638.09

8

1.108

5588.41

C:N
ratio
59.822

0.776

28.828
0.727

0.428

C:N
ratio
0.002
<0.001
0.442

429.30

0.360

0.614

0.708

0.565

0.081

659.52

0.581

0.944

0.685

0.486

0.139

698.55

pH
<0.001
<0.001

159

Table 2.1 2-way, nested ANOVA able for effects o f site, treatment, and treatment replicate on soil properties (cont’d)
Source of
Effect
Degree Sum o f squares
F
Mean square
(fixed/
variation
o f Free­
random) dom
Exch Exch
CFC
Exch Exch
CFC
Exch
Exch
CFC
A1
Na
A1
Na
A1
Na
Intercept
Fixed
1
0.013 0.055
1173.54 0.013 0.055
1173.54 9.861
120.11 714.93
Site
Fixed
2
0.007 0.003
44.052
0.003 0.001
1.765 2.445
7.869
22.026
Treatment
Fixed
1
0.003 <0.001 3.846
0.003 <0.001 3.846
1.938 1.086
2.343
Site*
Fixed
2
0.002 <0.001 5.401
0.001 <0.001 2.701
0.571 0.426
0.965
Treatment
Replicate
Random
4
0.005 0.002
6.566
0.001 <0.001 1.641
0.697 0.824
0.586
(treatment)
Error
8
0.015 0.004
22.394
0.002 <0.001 2.799

p-value

Exch
A1
0.034
0.231
0.236

Exch
Na
<0.001
0.148
0.356

<0.001
0.013
0.201

0.587

0.667

0.421

0.615

0.545

0.681

Exch
Fe
0.082
0.295
0.466

Exch
K
<0.001
<0.001
0.611

0.391

0.081

0.187

0.088

Table A2.1-2-way, nested ANOVA Table for effects o f site, treatment, and treatment replicate on soil properties (cont’d
Source of
Effect
Degrees Sum o f squares
Mean square
F
p-value
variation
(fixed/
of
random) Freedom
Exch
Exch
Exch
Exch
Exch
Exch
Exch
Exch Exch K Exch
K
Ca
Fe
Ca
Fe
K
Ca
Fe
Ca
Intercept
Fixed
1
425.853 <0.001 0.252
523.172 5.329 411.127 <0.001
425.853 <0.001 0.252
Site
2
Fixed
31.1634 <0.001 0.039
15.582
<0.001 0.020
14.208
1.428 95.784
0.002
Treatment Fixed
1
2.701
<0.001 <0.001 2.701
<0.001 <0.001 3.318
0.143
0.648 0.304
Site*
Fixed
2
2.621
<0.001 0.001
1.310
<0.001 <0.001 1.1950
1.058 3.508
0.352
Treatment
Replicate
Random
4
3.256
<0.001 0.002
0.814
<0.001 <0.001 0.742
2.006 2.992
0.589
(treatment)
Error
8
8.773
<0.001 0.002
1.097
<0.001 <0.001

160

CFC

Table A2.1 2-way, nested ANOVA Table for effects of site, treatment, and treatment replicate on soil properties (cont’d)
Mean square
F
p-value
Source of
Degrees Sum o f squares
Effect
variation
of
(fixed/
random) freedom
Intercept
Site

Fixed
Fixed

Treatment
Site*
Treatment
Replicate
(treatment)

Fixed
Fixed
Random

Error

1
2
1

NO 3
15444.62

NH 4

NO 3
15444.62

NO 3
91.257
9.482
0.819

NH 4
37.932
7.601
2.812

NO 3
<0.001

1981.76
138.67

NH 4
87.252
26.977
6.468

NH 4
<0.001
0.014

3963.52
138.67

87.252
53.954
6.468

2

109.86

9.138

54.93

4.569

0.263

1.287

0.775

0.328

4

676.97

9.201

169.24

2.300

0.810

0.648

0.552

0.6439

8

1672.02

28.394

209.00

3.549

0.008
0.417

0.169

Table A2.1 2-way, nested ANOVA Table for effects of site, treatment., and treatment rep icate on soil properties (cont’d)
Source o f
Effect
Degrees Sum o f
Mean
F
p-value
variation
(fixed/
of
squares
square
random) freedom
Avail P
68.562

Avail P
68.562
18.615
0.276

Avail P
130.269
61.496
0.524

Avail P
<0.001
<0.001
0.509

Intercept
Site

Fixed
Fixed

1
2

Treatment
Site *
Treatment
Replicate
(treatment)

Fixed
Fixed

1

37.229
0.276

2

0.166

0.083

0.274

0.767

4

2.105

0.526

1.739

0.234

8

2.42156

0.30269

Error

Random

161

Table A2.2 2-way ANOVA table for the effects o f site, treatment, and treatment replicate on genotype richness.
Source of
Effect
Degrees Sum o f squares
Mean square
F
p-value
variation
(fixed/
of
random) freedom

Intercept
Site

Fixed
Fixed

Treatment
Site *
Treatment
Replicate
(treatment)

Fixed
Fixed
Random

Error

ITS4Hin^i
6086.722
320.111
1.389

IT S lHinfl
8756.056

2
1

IT SlHinü
8756.056
561.444
16.056

2

24.111

4
8

1

280.722
16.056

ITS4Hinfl
6086.722
160.056
1.389

IT SlHinÛ
2717.397
15.991

ITS4HinÛ
319.420
25.052

IT S lHinfl
<0.001
0.002

4.983

0.073

0.089

<0.001
0.801

3.444

12.056

1.722

0.687

0.270

0.531

0.770

12.889

76.222

3.222

19.056

0.184

2.983

0.940

0.0881

140.444

51.111

17.556

6.389

ITS4Hina
<0.001

Table A2.2 2-way ANOVA table for the effects o f site, treatment, and treatment replicate on genotype richness (eont’d)
Source of
Effect
Degrees Sum o f squares
Mean square
F
p-value
variation
(fixed/
of
random) freedom

Intercept
Site

Fixed
Fixed

1
2

Treatment
Site *
Treatment
Replieate
(treatment)

Fixed
Fixed

1

IT SlAlul
10368.00
730.33
242.00

2

16.33

4
8

Error

Random

ITS4Alul
7160.056

IT SlAlul
10368.00
365.17

ITS4Alul
7160.056

ITS4- A M

242.00

301.389
40.500

IT SlAlul
1829.647
5.084
42.706

37.000

8.17

18.500

0.114

1.231

0.893

0.342

22.67

28.444

5.67

7.111

0.079

0.473

0.987

0.755

574.67

120.222

71.83

15.028

602.778
40.500

IT S lAM
<0.001

ITS4AM
<0.001

5.695

0.038
0.003

<0.001
0.075

1006.883
20.055

162

Table A2.3 2-way ANOVA table for the effects o f site, treatment, and treatment replicate on genotype evenness
p-value
Source o f
Effect
Degrees Sum o f squares
Mean square
F
variation
(fixed/
of
random) freedom

Intercept
Site

Fixed
Fixed

Treatment
Site *
Treatment
Replicate
(treatment)

Fixed
Fixed
Random

Error

IT SlHinil
15.198
0.002
<0.001

ITS4Hinfl
14.240
0.009
0.002

IT S lHinQ
18484.57
3.01
0.11

ITS4Hinf[
4635.083

<0.001

ITS4HinÜ
14.240
0.018
0.002

2

0.001

<0.001

<0.001

<0.001

4

0.003

0.012

<0.001

0.003

8

0.005

0.009

<0.001

0.001

1
2
1

IT SlHinÜ
15.198
0.004

7.965
0.523

IT SlHinfl
<0.001
0.106
0.759

ITS4HinÛ
<0.001
0.012

1.02

0.243

0.402

0.790

1.22

2.738

0.373

0.105

0.510

Table A2.3 2-way ANOVA table for the effects o f site, treatment, and treatment replicate on genotype evenness (cont’d
Source of
Effect
Degrees Sum o f squares
Mean square
F
p-value
variation
(fixed/
of
random) freedom

Intercept
Site

Fixed
Fixed

Treatment
Site *
Treatment
Replicate
(treatment)

Fixed
Fixed

Error
U)

Random

IT SlAlul
13.090

ITS4Alul
15.327

IT S lAlul
13.090

ITS4Alul
15.327

IT SlAlul
305.448

0.103
0.155

0.010
<0.001

0.051
0.155

0.005
<0.001

2

0.095

<0.001

0.048

4

0.171

0.003

8

0.536

0.007

1
2
1

ITS4- Alul

0.766
3.615

23783.80
5.97
0.08

IT SlAlul
<0.001
0.496
0.130

ITS4Alul
<0.001
0.026
0.794

<0.001

0.711

0.08

0.520

0.925

0.043

<0.001

0.639

0.76

0.649

0.582

0.067

<0.001

Table A 2 A Relative abundance o f genotypes from control to OTC for each primer-enzyme
combination. Numbers in bold indicate a 50% change in relative abundance from the control
to OTC for each site. The three most dominant genotypes in each plot are italicized.
ITS \-H in n
Site
Dolomitic
Granitic
Lowland
Control
OTC
Control
OTC
Control
OTC
Genotype
(%)
(%)
(%)
(%)
(%)
(%)
4.85
0.971
0
2.94
0
82
3.85
110
0.49
0
0
0
0.59
2.74
8.22
139
7.77
2.35
3.88
3.85
3.95
0
0
4.57
150
0
0
0
0
3.15
0
0
166
0
5.83
0
0
210
0
6.39
&5J
0
0
0.49
0
3.88
3.94
238
248
0.97
0
0.59
0.91
2.43
0 .8 6
2.14
0.49
1.94
3.94
2.35
4.57
252
2.91
1.71
4.85
3.94
4.12
0.91
253
2.43
1.71
3.88
7.09
2.35
0
256
0
0.97
2.74
271
0
0
0
274
2.43
20.47
1.18
7.31
2.56
y j .j j
9.4
14.17
10.59
5.94
277
3.88
6.8
280
1.71
1.57
2.91
0
1.18
8.22
4.27
1.94
1.83
283
2.43
0.79
3.53
0.97
0.427
4.72
5.29
286
3.88
0
289
0.49
1.28
0
2.36
10.59
0.46
12.62
0.97
0.79
4.12
292
23.08
1.37
307
4.37
0
0
0
0.59
0.91
2.14
2.94
311
1.94
0
1.83
2.43
314
1.94
2.94
0.97
0
1.83
1.28
317
0
0
16.51
3.85
2.35
2.28
321
0.48
2.99
0
1.58
1.37
2.35
331
2.42
1.28
3.88
6.3
0
0
344
1.94
0.85
0
0
4.71
0.46
347
0.97
0.86
0
0
0
0.46
0
1.71
0.97
11.18
356
0.79
7.37
358
0
0.43
5.83
2.36
0.59
0.46
0.49
0
2.91
1.57
2.94
363
0.46
401
0.49
0.43
0
0
5.83
0

164

Table A2.4 Relative abundance o f genotypes. Numbers in bold indicate a 50% change in
relative abundance from the control to OTC for each site (cont’d). The three most dominant
nS 4-H in a
Genotype
82
91
97
112
178
182
198
215
218
240
262
265
268
271
276
282
305
308
315
318
325
329
335
338
341
344
356
382

Lowland
Control
OTC
(%)
(%)
2 .8 8
0.99
0
0
0.99
0.82
5.76
0.50
1.65
2.97
2.06
4.46
2.06
0
2.06
0
0.41
7.92
70.70
1.48
1.23
7&P3
74. &7
2.47
0
0.99
0.41
0.41
0.50
1.98
2.06
2.47
0
0.99
3.29
1.65
2.97
2.47
2.47
0.41
0
0
0
1.49
2 .8 8
4.95
6.17
2.97
0.82
0.41
3.96
2.97
0.82

Site
Dolomitic
Control
OTC
(%)
(%)
0
0
0
0
6.84
2.30
0
0
2.56
0
8.05
1.15
1.71
0
5.13
5.75
4.27
2.30
0
0
0
6.84
1.15
9.20
20.51
0
1.71
3.42
6.90
0
0
0
0
0
0
0
0
0
0
0
0
0.85
0
9.20
2.56
3.42
3.45
10.34
4.27
3.41
3.45
1.15
0
0
0

Granitic
Control
OTC
(%)
(%)
5.30
0
1.44
3.79
0
0
3.03
0.72
1.44
4.55
3.03
2 .8 8
0
0
1.44
12.88
0
0.72
0
10.07
72.23
P.&j
0
0
5.76
7.Jg
1.52
2 .8 8
0.72
0
0

2 .8 8

0

3.60
5.04
0
0.72
5.76

0.76
0
0

3.06
6.06
3.03
5.30
7.58
0
0.76
0

0
2 .8 8

12.23
12.23
0
0

0

165

Table A2.4 Relative abundance of genotypes. Numbers in bold indicate a 50% change in
relative abundance from the control to OTC for each site (cont’d). The three most dominant
ITS 1-.4M
Genotype
70
77
109
115
127
135
148
169
171
355
358
361
373
386
391
402
416
426
437
441
446
449
462
510
530
533
536
539
559

Lowland
Control
OTC
(%)
(%)
74.09
3.24
3.64
1.62
2.27
0.54
0.91
2.70
3.18
4.32
2.27
&77
5.46
1.08
1.82
0.54
1.82
0
1.36
0.54
0.46
2.70
1.36
0
0
3.24
2.27
0
0
0
0
74.09
1.08
2.73
0
0
0
0
3.24
2.27
3.78
2.27
2.27
1.08
0.54
0
3.24
4.09
0.54
0.91
0.54
1.82
0
0
0.54
0

Site
Dolomitic
Control
OTC
(%)
(%)
0
0
1.74
1.49
0
0
0.87
0
0
0
0
0
0
0
13.43
4.35
0
0
2.99
6.96
7.82
0
2.99
8.70
1.49
2.61
0
0
0.87
0
5.22
4.4&
0
0
1.49
0
0
0.87
0
0.87
5.22
14.93
1.49
0
0
0
0
0.87
0
0
0.87
0
0
0.87
0
5.22
0
0

Granitic
Control
OTC
(%)
(%)
1.75
0.45
0.58
3.13
1.34
1.17
7.02
6.25
0
0
1.17
0
1.75
0
&77
9.j&
8.19
0.45
3.51
1.33
1.75
0.45
1.17
0.45
2.34
2 .6 8
0
1.34
0

2 .6 8

0
0.58

0

0
3.13
0.45
0.89
0.89
5.80
4.01
0.45

0

2 .6 8

&77
&79
4.09
0.59
4.09

&04
9.j,ÿ
5.36
0.45

0

4.09
3.51
1.17
1.17

0

166

Table A2.4 Relative abundanee o f genotypes. Numbers in bold indicate a 50% change in
relative abundance from the control to OTC for each site (cont’d). The three most dominant
ITS 4-Alul
Genotype
70
73
88

94
97
115
130
147
155
170
173
176
184
187
190
193
230
239
246
380
403
442
466
483
486
530
533
536
559
640

Site
Lowland
Control
OTC (%)
(%)
0.60
4.07
10.84
5.43
3.01
2.26
4.22
4.52
1 .2 0
2.71
0
0
7.24
1.81
0.60
1.36
0.60
0.45
2.26
0.60
0.60
5.43
0
0
4.22
9.05
1 .2 0

0

6.63
3.61

1.81
5.43
0.45
0.45
5.43
4.52
2.72
2.71
3.62
3.62
4.07
3.17
2.26
0.91
0
0

6 .0 2

3.01
0

6.63
5.42
0
0

1.81
5.42
3.61
1.81
1.81
0
0

Dolomitic
Control
OTC (%)
(%)
0
0
4.23
2 .8 6
0
0.95
0
0
0
0
0
0
0
0
0
0
4.76
0
2.82
4.76
7.04
7.62
2.82
4.76
5.63
6.67
74.
77.74
P. 52
P. & 6
18.31
3.81
0
0
0
0
0
0
1.41
0
1.41
0
2.82
9.52
0
0
0
0
0
0
1.41
0
0
0
0
0.95
0
0
0
0

Granitic
Control
OTC (%)
(%)
0
0
1.96
1.67
0.65
0.56
1.31
3.89
0.65
0.56
3.27
0.56
0.65
3.89
0
5.56
0
0.56
1.31
0
0
6.67
0
0
0

2 .2 2

0

2 .2 2

5.23
20.92
0
0.65
0
1.31
0
0

12.78
9.44
0

0

2.61
1.31
5.88
77.77
6.54
5.23
0.65

0

0
0

0
0
5.56
0

0.56
8.33
75.33
7.78
0

2.78

167

A xis 2

103
IC I

[3C2

S102

i4C l

S303

A xis 1

S403

S3C1
S401
S4C3

S402

S301

S4C
S302

Fig. A2.1 NMS for site and warming treatment effeet. Rotated at 75°. Coding is as follows:
first two characters represent site, SI (lowland), S3 (highland dolomitic), S4 (highland
granitic); the last two characters signify the treatment and the replicate o f that treatment, e.g.
C l would be replicate one o f control, 0 3 would be replicate three o f OTC. Site is clustered
together with the lowland and highland granitic sites clustering closer together than the
lowland and highland dolomitic sites.

168

(a) ITS \-HinÛ
9.5E-02
S IC IP I
S4C 1P 2
S3C 2P 2
S30T C 2P1
S4C 3P 2
S4C 2P 2
S4C 2P1
S10T C 2P1
S30T C 3P1
S30T C 1P2 b S30T C 3P2
S3C 2P1
S3C 3P1
S30T C 2P2
S1C 1P 2
S3C 1P 2
S30T C 1P1
S3C 1P1
S40T C 1P2
S3C 3P 2
S4C 1P1
S40T C 1P1
S40T C 2P2
S4C 3P1
S10T C 1P2
S40T C 2P1
S1C 2P1
S40T C 3P1
S40T C 3P2
S IO T C IP I
S1C 2P 2
S10T C 2P2
S1C 3P1
S1C 3P 2
S10T C 3P1
S10T C 3P2

(b) ITS A-Hinü
5E+00

7.3E-02
S IC IP I
S3C 3P1
S30T C 1P1
S10T C 1P2
S10T C 2P2
S4C 1P1
S40T C 3P1
S4C 3P1
S10T C 2P1
S4 C 3 P 2
S40T C 3P2
S10T C 3P1
S1 C 1 P 2
S1C 2P1
S10T C 3P2
S40T C 2P1
S1 C 2 P 2
S1C 3P1
S1 C 3 P 2
S IO T C IP I
S3C 1P1
S30T C 1P2
S30T C 2P1
S3 C 3 P 2
S4 C 1 P 2
S40T C 2P2
S3 C 1 P 2
S30T C 3P1
S30T C 3P2
S3C 2P1
S3 C 2 P 2
S30T C 2P2
S4C 2P1
S40T C 1P2
S4 C 2 P 2
S40T C 1P1

3.9E

b-

Fig. A2.2 Cluster analysis o f plots, treatments and plant specimen based on frequency o f genotypes. SI = lowland, S3= dolomitic, and
S4 = granitic. C and OTC represent control or warmed plots, and P I, P2 distinguishes plant specimen replication.

o\
VO

(c) ITS \-Alu\
(d) IT S4-^M
1.5E-01
S lC lP l
S102P1
S401P 1
S 103 P 2
S 102 P 2
S3C3P2
S1C1P2
S1C2P1
S1C2P2
S403P 1
S1C3P2
S1C3P1
S 101 P 2
S103P 1
SIO IP I
S302P 1
S4C2P2
S 403 P 2
S4C1P1
S4C3P2
S4C3P1 Z z h
S 401 P 2
S4C1P2
S4C2P1
S402P 1
S3C1P1
S3C2P2
S 301 P 2
S303P 1
S 303 P 2
S3C1P2
S301P 1
S 402 P 2
S3C2P1
S 302 P 2
S3C3P1

7.8E+C

1.3E-01
S lC lP l
S1C3P1
S1C3P2
S lO T C lP l
S 10T C 1P 2
S 40T C 3P 2
S 10T C 2P 2
S4C1P2
S1C2P2
S 10T C 3P 2
S10T C 2P1
S 30T C 3P 2
S10T C 3P1
S1C1P2
S3C2P1
S3C1P1
S 30T C 1P 2
S3C2P2
S30T C 2P1
S3C3P2
S1C2P1
S30T C 3P1
S3C1P2
S 40T C 1P 2
S30T C 1P1
S 40T C 2P 2
S4C1P1
S40T C 3P1
S40T C 1P1
S40T C 2P1
S4C2P1
S4C2P2
S4C3P1
S4C3P2
S3C3P1
S 30T C 2P 2

Fig. A2.2 (cont’d) Cluster analysis o f plots, treatments and plant specimen based on frequency o f genotypes.

'J
o

6.2E-+

(a) Exchangeable A1

p = 0.23

^ 0.10

0 .1 0

0.08

A

0.08

0.06 -

§. 0.04

p = 0.15

(b) Exchangeable Na

□ Control

0.06

HOTC

0.04 -

□ Control:
H OTC

!

0.02 -

M 0.02

0.00

■g 0.00
Lowland

Dolomitic
h^hland

Lowland

Granitic
highland

Granitic

h^hland

Site

* p = 0.01

(c) CEC
14.00
12.00

p = 0.002

(d) Exchangeable Ca*
10.00

:.oo
□ Control

6.00 -

6.00

□ Control

4.00

HOTC

2.00
0.00
Lowland

Dolomidc

Granitic
highland

Granitic
hi^tland

Lowland

Site

(e) Exchangeable Fe
s

p = 0.33

(f) Exchangeable K

0.010

0.25

0.008

0 .2 0 -

% 0.006

□ Control

Ji 0.004

HOTC

0.15

□ C o ntrol
b

0.10

:

0.00

Lowland

Site

(g) Exchangeable Mg

D olo m itic

G ra n itic

h ig h la n d

h ig h la n d

Site

* p = 0.01

g 5.00
(J

OTC

0.05

1 0.000

s

p = <0.001

(h) Exchangeable Mn

* p = 0.02

0.035
0.030
0.025

4.00

-

3.00

□ Control

0.020

□ Control:

I

2.00

HOTC

0.015 -

HOTC

I 1.00

'

0 .0 1 0 -

0.005

a

« 0.00

0.000
Lowland

Dolomitic
hi^iland

Site

Granitic

higWaitd

Lowland

Dolomitic
hi^iland

Granitic
hi^iland

Site

Fig. A2.3 Least square means for soil properties. Error bar is one standard error. * Significant
for site effect, p-value for site given along with soil property name.

171

p <0.001

(i)pH*
10.00 y

(j) C:N ratio"

p <0.001

60.0 0
140.00
120.00 -

8.00 -

I 6.00 -

□ C ontrol

a

HOTC

4 .0 0 -

100.00

□ C ontrol

2.00 -

0.00
Lowland

D obm itic

Granitic

hçhland

hi^iland

Lowland

Site

p - 0.008

7.00 -

a 40.00

3 0 .0 0 -

0

2 0 .0 0 -

p = 0.01

6.00

5 0 .00 -

if

(1) NH 4"
8.00

70.00 T
6 0 .0 0
h

Granitic

hi^iland

Site

(k) NO]"

D o b rtttic

□ C ontrol
HOTC

5.00
4.00
3.00 -I
2.00

10.00 -

LOO

0.00

0.00
Lowland

Dolomitic

Granitic

highland

highland

HOTC

Lowland

She

(m) Available P*

□ C ontrol

D obm itic

Granitic

h i^ ia n d

h i^ ik n d

Site

p <0.001

5.00

'Ë
'
Q. 4 .0 0 -I
s
a . 3.00
a

□ C ontrol
HOTC

■g 2.00

I

jiL

0.00
Lowland

D obm itic

Granitic

hi^iland

highland

Site

Fig. A2.3 (cont’d). Least square means for soil properties. Error bar is one standard error.
*Significant site effect, p-value for site given along with soil property name.

172

(a) ITS \-H inü

1 .2 0 0

-

LOGO 0.800 0.600 -

□ Control
HOTC

0.400
0.200
0.000
Lowland

Dolomitic

Granitic

Site

(b)ITS A-HinÜ
^

1.600

“

1.200

6

I□ Control

^ \

0.600

r
s

0.400
0.200

^

0.000
Lowland

Dolomitic

Granitic

(c)ITS 1-^ / mI
C 2.000
I

ÿ 1.500
□ Control

t I 1000 -

HOTC

I §■ 0.500

I

Æ

Î - 0.000

Lowland

Dolomitic

Granitic

Site

(d)ITS 4-Alul
1.400
*

.200

r 1000
o 0.800 a

3 0.600

S

0.400

□ Control

Ü A 0.200
0.000
Lowland

Dobmitic

Granitic

Site

Fig. A2.4 Least square means for genotype frequency for each treatment per plot. Error bar
represents one standard error.

173

(a) ITS \-H inû

4 .5

OTC
Control

0 .5

1

4

7

10 13

16 19 2 2 2 5 2 8 31

3 4 3 7 4 0 4 3 4 6 4 9 5 2 55

Rank genotype abundance

(b)ITS
4.5
s
cs
R
-o
S3

OTC

S

S611

Control

O

0.5
1

4

7

10 13 16 19 22 25 28 31 34 37 40 43 46
Rank genotype abundance

Fig. A2.5 Rank abundance curves for lowland sites.
174

(c)ITS \-Alul

3.5
:
■a 2.5
n

— OTC

!

■■ Control

OS

o

0.5
1 4 7 10 13 16 19 22 25 28 31 34 37 40 43 46 49 52 55 58
Rank genotype abundance

(d)ITS A-AM
3.5

OTC
Control

0.5
1 4

7 10 13 16 19 22 25 28 31 34 37 40 43 46 49 52
Rank genotype abundance

Fig. A2.5 (cont’d) Rank abundance curves. Relative abundance is log scaled.
175

(a) ITS l-HinÜ
3.5

I

2 .5

I ^
«

OTC
■■■■Control

1.5

Î >
0 .5

1

3

5

7

9

11

13

15

17

19 21

23

25

27

29

31

33

Rank genotype abundance

(b)ITS A-Hinü
3 .5
3

§
'O
s

t

2 .5

2

OTC

1.5

— Control

CD

o
1-1

0 .5

0
1

3

5

7

9

11

13

15

17

19

21

23

25

27

29

Rank genotype abundance

Figure A2.6 Rank abundance curve for highland dolomitic site. Relative abundance is log
scaled.

176

(c)ITS \-Alu\

OTC
Control

0

t - t -

-

r

:

I

[

I

T

T

1

'

1

'

1

1

'

1

1

1

1

'

r i

'

f

■

I

'

1

1

1

■

1

i

1

i

h

\

Rank genotype abundance

(d)ITS 4-AM
3 .5 --------

—

OTC
C ontrol

0 .5

1

3

5

7

9

11

13

15

17

19

21

23

25

27

29

31

R a n k g en o ty p e ab u ndance

Fig. A2.6 (cont’d) Rank abundance curve for highland dolomitic site. Relative abundance is
log scaled.

177

(a) ITS \-HinÛ
3.5

2.5
OTC
Control

0.5
1 4 7 10 13 16 19 22 25 28 31 34 37 40 43 46 49 52 55
Rank genotype abundance

(b)ITS 4-//m fl

2 .5

§

13

OTC

§

Control

0.5

1

3

5

7

9

11

13

15

17

19 21

23

25 2 7

2 9 31

33

35

Rank genotype abundance

Fig. A2.7 Rank abundance curves for highland granitic site. Relative abundance is log
scaled.

178

(c)ITS \-Alu\
3.5
aw 3
w 2.5
"9
2

I
s

OTC
■Control

1.5

0.5
0

1 4 7 10 13 16 19 22 25 28 31 34 37 40 43 46 49 52 55 58 61
Rank genotype abundance

(d)ITS A-AM
4
3.5
U 3
fl
ft
T3 2.5

e

i

OTC

2

■control

w 1.5
o

1

0.5
0
1 3 5 7 9 11 13 15 17 1921 2 3 2 5 2 7 2931 33 35 3 73 94 1
Rank genotype abundance

Fig. A2.7 (cont’d) Rank abundance curves for highland granitic site. Relative abundance is
log scaled.
179

(a) ITS \-H inû
35.00
« 30.00

I

25.00

•C 20.00

□ Control

*‘ 15.00
10.00
J 5.00 4

BOTC

0.00
Lowland

Dolomitic highland

Granitic highland

Site

(b)ITS A-HinÛ
30.00 « 25.00 e
X 20.00 h
« 15.00
È"
o 10.00

□ Control
BOTC

5.00
0 00 Lowland

Dolomitic h^hland

Granitic hi^iland

Site

40.00
35.00
I

30.00 4

I

25.00
□ Control !

g, 20.00 -

B OTC

I " 15.00
J

10.00
5.00
0.00
Lowland

Dolomitic h i ^ n d

Granitic hi^iland

Site

(d)ITS 4 -^ M
35.00
« 30.00
I

25.00

•C 2 0 .0 0

I □ C o n tro l'

I

BOTC

15.00

I 10.00
^

5.00
0.00
Lowland
Site

Fig. A2.8. Least square means o f genotype richness for each treatment per site. Error bar
represents one standard error.
180

(a) ITS

0.94 S

0.92 -

□ Control
HOTC

Lowland

Dolomitic hi^iland

Granitic hi^iland

Site

(b)ITS A-HinÛ
I.GO
0.95
0.90

□ Control
B O TC

0.85
0.80
0.75
Lowland

Dolomitic hi^iland

Granitic hi^iland

Site

(c)ITS \-Alu\
1.20

1.00

g

0.80
; □ Control

0.60 -

0.20
0.00
Lowland

Dolomitic h^Jiland

Granitic hÿ^tland

Site

(d)ITS A-AM

g

1.00
0.98 '
0.96
0.94
; □ Control
0.90 0.86 0.84
0.82
Lowland

Dolomitic highland
Site

Fig. A2.9. Least square means o f genotype evenness measured by Hurlbert’s PIE index.
Error bar represents one standard error.
181

VIII. Appendix for chapter 3
Table A3.1 Comparison o f identification (based on molecular techniques) with corresponding
morphological identification - Phialocephala. Identifications by sequencing and morphology
ID based on R FL P and sequence
analyses: Phialocephala fortinil
RFLP type Ipf
2-S l-C l-D rin3.2
2-C l-S l-S aar2.3
2-S1-OTC 1-Saar 2.1
2-S1-OTC 1-Saar 2.2
RFLP type 2pf

ID based on m orphology

dark sterile
Phialocephala fortinii
dark sterile
NA

1-Sl-O TCl-Drin 1.1

Phialocephala fortinii

l-Sl-O TC2-Saop 5.5
1-Sl-0TC3-Drin 1.3

Phialocephala fortinii
NA

l-S4-O TC l-Saop4.1
2-Sl-C2-Drin 1.1

Phialocephala fortinii
Phialocephala fortinii

2-Sl-C2-Saar 1.1
2-S 1-OTC 1-Saar 1.1
2-S 1-OTC 1-Saar 1.2
2-S 1-OTC 1-Saar 3.1
2-S 1-OTC 1-Saar 4.1
2-Sl-OTC2-Drin2.1

Phialocephala fortinii
Phialocephala fortinii
Phialocephala fortinii
Phialocephala fortinii
Phialocephala fortinii

2-S l-0T C 2-S aar2.2
2-S4-C3-Saar 4.6
RFLP type 3pf
1-Sl-C l-D rin2.1
1-SI-C l-Saar 4.4
l-Sl-O TC 3-C atc4.1A
1-Sl-0TC 3-C atc4.3
1-S4-Cl-Catc3.8
1-S4-C1-Cate 5.7
1-S4-Cl-Catc5.8
l-S4-C2-Saop 4.2
l-S4-O TC l-Saop4.2
2-S4-C1-Saar 2.5

pigmented warty/ dark sterile
Phialocephala fortinii
Phialocephala fortinii
Phialocephala fortinii
Phialocephala fortinii
Phialocephala fortinii
NA
Phialocephala fortinii
Phialocephala fortinii
NA
Phialocephala fortinii
Phialocephala fortinii
Phialocephala fortinii

182

Table A3.1 Comparison o f identification (based on molecular techniques) with corresponding
morphological identification - Phialocephala. Identifications by sequencing and morphology
ID based on R FL P and sequence
analyses : Phialocephala fortinii
2-S4-Saar 5.18
2-S4-C1-Saar 5.4
2-S4-C1-Saar 5.9
2-S4-C2-Saop 4.1a
1-S4-C3-Saar3.4
2-S4-CTL-Cate 2.2
2-S4-OTC3-Saop 1.3
2-S4-OTC3-Saop 5.3
RFLP type 4pf
1-Sl-C2-D rin2.2
1-Sl-C3-Cate 5.3
1-Sl-0TC 2-Saar3.2
1-Sl-0TC2-Saar4.11
1-Sl-0TC2-Saar4.6
l-Sl-O TC2-Saop 5.5a
1-Sl-0TC 3-C ate5.1
RFLP type 5pf
1-Sl-C3-Saar 1.3
1-Sl-C3-Saar 4.2
1-Sl-O TCl-Saar 2.1
1-Sl-0TC 3-C ate3.1
2-Sl-OTC3-Saar5.1
3-Sl-C3-Saar 4.2
RFLP type 6pf
1-Sl-O TC l-Saar 5.1
l-S4-Cl-Saop2.3
2-S4-Cl-Saar 1.2
2-S4-Cl-Saar4.2

ID based on m orphology
Phialocephala fortinii
Phialocephala fortinii
Phialocephala fortinii
NA
Phialocephala fortinii
NA
dark sterile
Phialocephala fortinii
Phialocephala fortinii
NA
Phialocephala fortinii
Phialocephala fortinii
Phialocephala fortinii
NA
Phialocephala fortinii
NA
Phialocephala fortinii
NA
Phialocephala fortinii
Phialocephala fortinii
NA
dark septate sterile
NA
dark sterile
Phialocephala fortinii

183

Table A3.2 Comparison o f identification (based on molecular techniques) with
corresponding morphological identification - Helotiales. Sequence analysis indicated that
ID based on morphology

ID based on RFLP and sequence
analyses: Helotiales
RFLP Type 1h
1-Sl-C2-Cate 1.4
1-S4-Ctl-Cate 4.2
1-S4-Ctl-Cate 4.5

NA
dark monolioid
NA

2-S4-Cl-Saar 1.1a

NA

2-S4-C3-Saop 3.2

NA

RFLP Type 2 h
l-S4-O TCl-Saop 5.9
1-S4-C2-Saar2.7
RFLP Type 3h
1-Sl-C2-Drin5.2
l-S4-OTC2-Saop2.1
2-Sl-OTC2-Saar2.1
2-S4-C3-Saop3.5
2-S4-OTC2-Saar2
2-S4-OTC3-Saar5.2
3-S l-C 3-Saarl.l
RFLP Type 4 h
l-S4-OTC2-Saop4.2
2-S3-C2-Saop5.1
2-S3-OTCl-Drin 1.2
2-S3-OTCl-Drin 1.4
RFLP Type 5h
1-S4-C2-Saar2.1
1-S4-C2-Saar2.2
RFLP Type 6h
1-S4-Cl-Cate 3.7
l-S4-C l-Saop 1.11

inconclusive
inconclusive
dark monolioid/ mixed
NA
NA
hyaline sterile
NA
mixed/ hyaline sterile
mixed
inconclusive
NA
dark sterile
NA
hyaline conidia/mixed
NA
NA
hyaline sterile

184

Table A3.2 Comparison of identification (based on molecular techniques) with
corresponding morphological identification - Helotiales. Sequence analysis indicated that
ID based on RFLP and sequence
analyses: Helotiales
RFLP Type ? h
1-Sl-C3-D rin2.4
l-Sl-OTC3-Cate-4.2
l-S4-C l-Saop 5.4
1-S4-C2-Saar5.1
l-S4-C3-Saop5.1
1-S4-0TC1-Saar 4.5
l-S4-O TC l-Saop3.7
l-S4-O TCl-Saop 4.4
1-S4-0TC2-Saar2.1
l-S4-OTC2-Saop4.1
l-S4-OTC2-Saop4.3
2-Sl-OTC2-Drin 1.1
2-S4-C1-Saar 5.20
2-S4-C2-Saar 4.9
2-S4-C2-Saop 3.2
2-S4-C2-Saop 3.3
2-S4-C3-Saop 1.1
2-S4-C3-Saop 3.1
2-S4-OTCl-Saar 1.10
2-S4-OTC1-Saar 2.1

ID based on morphology

dark septate monolioid
NA
NA
NA
dark septate monolioid
dark sterile/ mixed
NA
NA
Phialocephala fortinii
dark monolioid
Phialocephala fortinii
Phialocephala fortinii
dark septate monolioid
Phialocephala fortinii
hyaline sterile
inconclusive
dark septate monolioid
Phialocephala fortinii
dark sterile
dark septate monolioid

185

Table A3.3 Comparison of identification (based on molecular techniques) with
corresponding morphological identification - unknown. Identifications by sequencing were
inconclusive.
ID based on RFLP and sequence ID based on morphology
analyses: Unknowns
l-Sl-C 3-Saop 5.4
l-S3-C3-Saop3.3
2-S3-C3-Saar4.1

Ascomycete
brown
brown

2-S4-C3-Saopl.5

brown

1-S4-0TC2-Saar4.1

Cadophora
dark monolioid
dark monolioid
dark monolioid

1-Sl-C3-Drin 1.3
1-S3-C-Drin 1.1
2-Sl-C2-Saar3.1
2-S3-C2-Saop 5.7
2S4-Cl-Saar 1.16
2-S4-OTC1-Saar 4.4
2-Sl-OTC2-Saop 2.1
l-S4-O TC l-Saop3.3
1-Sl-C3-Drin 4.1
1-Sl-0TC1-Saar5.1
l-Sl-O TC3-Saop 2.2
l-S3-C3-Saop5.2
l-S4-C l-Saop2.3
1-S4-Ctl-Cate 2.7
2-Sl-OTC2-Saar 1.1
2-S3-C3-Drin 4.2
2-S3-C3-Saop4.2
2-S3-OTCl-Saar 1.8
2-S3-OTC3-Drin4.1
2-S4-C2-Saarl.2
1-S4-C2-Saar3.2
1-S4-CTL Cate 2.1

dark monolioid
dark monolioid
dark monolioid
dark septate, few monolioids
dark some monolioid
dark sterile
dark sterile
dark sterile
dark sterile
dark sterile
dark sterile
dark sterile
dark sterile
dark sterile
dark sterile
dark sterile
dark sterile
Geomyces
Geomyces

186

Table A3.3 Comparison o f identification (based on molecular techniques) with
corresponding morphological identification - unknown. Identifications by sequencing were
ID based on RFLP and sequence
analyses: Unknowns
1-S4-0TC1-Saar 4.11
1-S4-0TC1-Saar 4.4
l-S4-OTC2-Saop3.1
2-S4-C2-Saop 4.4B
2-S4-C2-Saop 4.8
1-S3-0TC3-Drin 3.3
2-S4-C3-Saar2.1
l-S4-OTC3-Saop 4.3
2-S4-C3-Saar 1.5
3-Sl-C2-Cate-4.1
1-Sl-C3-Cate3.1
1-Sl-C l-D rin2.3
1-S l-C l-D rin2.5
1-Sl-0TC 3-D rin -3.2
1-Sl-0TC 3-Saar3.1
1-Sl-0TC 3-Saar3.2
l-Sl-O TC 3-Saop4.1
1-S3-C2-Drin2.1
1-S3-C-Drin2.1
l-S3-OTC2-Drin2.2B
l-S3-OTC3-Saar 1.2
1-S4-C1-Cate 4.1
1-S4-C1-Saar 3.2
l-S4-C l-Saop 1.11
1-S4-C2-Saar4.2
1-S4-C2-Saar4.3
1-S4-Ctl-Cate 4.7
1-S4-0TC1-Saar 3.10
l-S4-OTC2-Saop 4.4
l-S4-OTC3-Saop4.4

ID based on morphology
Geomyces
Geomyces
Geomyces
Geomyces
Geomyces
Hyaline arthroconidia
hyaline brown; swollen cells
hyaline conidia, monolioid
hyaline hyphae, dark spored
hyaline monolioid
hyaline spherical, ovoid conidia
hyaline sterile
hyaline sterile
hyaline sterile
hyaline sterile
hyaline sterile
hyaline sterile
hyaline sterile
hyaline sterile
hyaline sterile
hyaline sterile
hyaline sterile
hyaline sterile
hyaline sterile
hyaline sterile
hyaline sterile
hyaline sterile
hyaline sterile
hyaline sterile
hyaline sterile

187

Table A3.3 Comparison o f identification (based on molecular techniques) with
corresponding morphological identification. Identifications by sequencing were inconclusive
ID based on RFLP and sequence
analyses: Unknowns
l-S4-OTC3-Saop4.6
l-S4-OTC3-Saop4.9
2-S3-C3-Drin3.3
2-S3-C3-Saop 2.1
2-S3-OTC3-Saar 1.3
2-S4-C1-Saar 3.20
2-S4-C3-Saar 1.3
2-S4-C3-Saop 1.7
2-S4-C3-Saopl.2
2-S4-C3-Saop3.5
2-S4-OTC2-Saop 1.3
2-S4-OTC2-Saop 3.2
2-S4-OTC3-Saar2.1
3-Sl-C2-Cate2.1
3-Sl-OTC3-Drin 1.2
2-S3-OTC3-Saop 4.2
2-S4-OIC3-Saar 2.4
1-Sl-C2-Cate4.1
1-S1-C3-Cate 3.2
1-Sl-C3-D rin2.2
1-Sl-C3-D rin4.2
l-Sl-C 3-Saop2.1
1-S3-Cl-Saar3.4
l-S3-C l-Saop2.2
1-S3-C2-Drin5.5
1-S3-C3-Drin3.1
l-S3-C3-Saop 1.8
l-S3-C3-Saop 4.4
1-S3-0TC1-Saar3.8
1-S4-C1-Cate 4.3
l-S4-C l-Saop 1.7
l-S 4-C l-Saop5.6
1-S4-C2-Saar 1.1
1-S4-0TC1-Saar 4.7

ID based on morphology
hyaline sterile
hyaline sterile
hyaline sterile
hyaline sterile
hyaline sterile
hyaline sterile
hyaline sterile
hyaline sterile
hyaline sterile
hyaline sterile
hyaline sterile
hyaline sterile
hyaline sterile
hyaline sterile
hyaline sterile
hyaline sterile/ hyaline monolioid
hyaline sterile/ hyaline monolioid
inconclusive
inconclusive
inconclusive
inconclusive
inconclusive
inconclusive
inconclusive
inconclusive
inconclusive
inconclusive
inconclusive
inconclusive
inconclusive
inconclusive
inconclusive
inconclusive
inconclusive

188

Table A3.3 Comparison o f identification (based on molecular techniques) with
corresponding morphological identification - unknowns. Identifications by sequencing were
ID based on RFLP and sequence
analyses: Unknowns

ID based on morphology

l-S4-O TCl-Saop 1.1
l-S4-O TCl-Saop 1.7
l-S4-O TCl-Saop 3.4
l-S4-O TC l-Saop5.5
l-S4-OTC3-Saop4.5
2-S l-C l-D rin4.5

inconclusive
inconclusive
inconclusive
inconclusive
inconclusive
inconclusive

2-Sl-C l-Saop 3.3
2-S l-C l-S aop4.4
2-S4-Cl-Saar 1.12
2-S4-Cl-Saar 1.5
2-S4-Cl-Saar 1.6
2-Sl-O T C l-Saop2.4

inconclusive
inconclusive

l-S4-C2-Saop2.3
l-Sl-C 3-Saop 1.2
1-Sl-Cl-Cate2.1
1-Sl-Cl-Catc2.3
1-Sl-C l-Saop 5.1
1-Sl-C2-Cate 4.2
1-Sl-C3-Catc4.1
1-Sl-C3-Catc5.1
1-SI-OTC 1-Saar 1.3
1-Sl-OTCl-Saop 5.2
1-Sl-0TC2-D rin 1.1
1-Sl-0TC 2-D rin4.1
l-Sl-O TC2-Saop 5.4
1-Sl-0TC3-Cate 5.2
1-Sl-0TC3-Catc4.2
1-Sl-0TC3-Cate 5.2
1-Sl-0TC 3-D rin2.1
l-Sl-O TC3-Saar 1.2
1-S3-Cl-D rin4.2
1-S3-C3-Drin 1.3

inconclusive
inconclusive
inconclusive
light brown sterile
lightly pigmented sterile
Monodictys
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA

189

Table A3.3 Comparison of identification (based on molecular techniques) with
corresponding morphological identification. Identifications by sequencing were inconclusive
ID based on RFLP and sequence
analyses: Unknowns
l-S3-C3-Saop2.1
l-S3-C3-Saop2.3
l-S3-C3-Saop4.1
l-S3-C3-Saop4.2
l-S3-C3-Saop4.3
l-S3-C3-Saop 5.3
1-S3-C-Drinl.2
1-S3-0TC1-Drin2.1
l-S3-OTC3-Drin2.2
l-S3-OTC3-Saop2
l-S3-OTC3-Saop2.1
l-S3-OTC3-Saop4.5
1-S4-Cl-Cate 3.7
1-S4-Cl-Cate 5.2
1-S4-C2-Saar2.2
1-S4-C2-Saar2.8
l-S4-C2-Saop5.1
1-S4-CTL-Cate 1.6
1-S4-CTL-Cate 2.5
1-S4-CTL-Cate 2.6
1-S4-Ctl-Cate 3.3
1-S4-CTL-Cate 3.5
1-S4-CTL-Cate5.1
1-S4-0TC1-Saar 4.15
l-S4-O TCl-Saop 1.5
l-S4-O TC l-Saop5.2
l-S4-OTC3-Saop 5.2
2-S l-C l-D rin4.1
2-S l-C l-D rin4.2
2-S 1-C l-Saar 2.1
2-S 1-Cl-Saar-3.1
2-Sl-C l-Saop 5.2

ID based on morphology
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA

190

Table A3.3 Comparison of identification (based on molecular techniques) with
eorresponding morphological identification - unknowns. Identifications by sequencing were
ID based on RFLP and sequence
analyses: Unknowns
2-S l-C l-Saop5.3

ID based on morphology

2-S3-C3-Saar 3.5
2-S3-C3-Saop 1.1
2-S3-C3-Saop 3.3
2-S3-C3-Saop 5.2
2-S3-OTCl-Saop3.5
2-S3-OTC2-Saar 3.5
2-S3-OTC2-Saar 3.7
2-S3-OTC3-Saar 3.2

NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA

2-S3-OTC3-Saar4.1
2-S3-OTC3-Saar 4.3
2-S3-OTC3-Saop4.1
2-S3-OTC3-Saop 4.4
2-S3-OTC3-Saop 4.5
2-S4-C1-Saar 1.11
2-S4-Cl-Saar 1.14
2-S4-C1-Saar 1.3

NA
NA
NA
NA
NA
NA
NA
NA

2-Sl-C l-Saop 5.5
2-Sl-C 2-D rin2.2
2-Sl-C 2-D rin3.2
2-Sl-C2-Saar2.1
2-S3-Cl-D rin3.1
2-S3-C1-Saar 2.1
2-S3-C1-Saar 3.1
2-S3-Cl-Saar 5.1
2-S3-C2-Drin 5.2
2-S3-C2-Drin 1.1
2-S3-C2-Drin 1.2
2-S3-C2-Drin 1.4
2-S3-C2-Drin 5.2
2-S3-C2-Saar 5.2
2-S3-C3-Drin3.1
2-S3-C3-Saar 1
2-S3-C3-Saar 2.5

191

Table A3.3 Comparison o f identification (based on molecular techniques) with
corresponding morphological identification. Identifications by sequencing were inconclusive
ID based on RFLP and sequence
analyses: Unknowns

ID based on morphology

2-S4-C1-Saar 2.4

NA
NA
NA
NA
NA
NA
NA

2-S4-C2-Saar 2.4
2-S4-C2-Saop 4 .IB
2-S4-C3-Saar 4.3
2-S4-C3-Saar 4.4
2-S4-C3-Saop 2.3
2-S4-C3-Saop 2.4
2-S4-OTC2-Saar 1.2
2-S4-OTC3-Saar 1.2
2-S4-OTC3-Saop 5.6
3-Sl-C2-Catel.3
3-Sl-OTC3-Drin 1.1
3-Sl-O TC3-Drin5.2
1-S4-OTC1-Saar 4.12
1-Sl-C l-D rin4.1
1-Sl-0TC 2-Saar5.5
1-Sl-0TC 3-D rin5.3
l-S3-OTC3-Saop 4.3
l-S4-O TCl-Saop 4.3
l-S4-OTC3-Saop 4.1
2-Sl-O T C l-Saop2.3
2-Sl-O TC2-Saar3.3
2-S3-OTC3-Saar 4.2
2-S4-C1-Saar 5.1
2-S4-OTC3-Saar 3.4
2-S3-O TCl-D rin4.5
2-Sl-C l-Saop 5.6
2-S4-OTCl-Saar 1.5
2-S4-OTC1-Saar 2.5
2-S3-OTC3-Drin3.1
2-S4-C3-Saar 2.3
2-S4-OTC2-Saar 4.3

NA
NA
NA
NA
NA
NA
Pénicillium
Phialocephala fortinii
Phialocephala fortinii
Phialocephala fortinii
Phialocephala fortinii
Phialocephala fortinii
Phialocephala fortinii
Phialocephala fortinii
Phialocephala fortinii
Phialocephala fortinii
Phialocephala fortinii
Phialocephala fortinii
Sebacina
Trichocladium
Trichoderma sporulosum
Trichoderma sporulosum
white
white
white

192

Table A3.4 Comparison o f identification (based on molecular techniques) with
corresponding morphological identification - Hymenoscyphus. Samples showed affiliation
ID based on RFLP and sequence
analyses: Hymenoscyphus

ID based on morphology

RFLP type Iny
l-S4-C l-Saop2.11
1-S4-0TC1-Saar3.5

dark spherical conidia
dark sterile

RFLP type 2 hy
1-Sl-C3-Saar 1.2
1-Sl-0TC 1-D rin2.3

intercalary chlamydospore fried egg
black type

1-Sl-0TC 1-D rin4.1
1-Sl-0TC 2-D rin3.1
1-S4-CTL-Cate 2.4

inconclusive
Leptodontidium

1-S4-0TC2-Saar4.6
2-S l-C l-S aop3.2
2-Sl-C 3-D rin3.2
2-S4-C2-Saar4.10
3-Sl-C3-Saar2.1
RFLP type 3 hy
l-S4-O TC l-Saop5.7
2-S4-OTCl-Saar 1.9
RFLP type 4 hy
1-S l-C l-D rin2.2
1-Sl-0TC3-Cate 4.2a
1-S4-Cl-Saar 4.6
2-S4-Cl-Saar 1.16
2-Sl-C 3-D rin3.2

inconclusive
inconclusive
NA
inconclusive
NA

black type

inconclusive
inconclusive
Phialocephala fortinii
NA
dark sterile
dark monolioid
NA

193

Table A3.5 Comparison o f identification (based on molecular techniques) with
corresponding morphological identification - Mycosphaerella. Samples showed affiliation
ID based on RFLP and sequence
analyses: Mycosphaerella
RFLP type 1m
1-Sl-C2-Cate 5.3
1-S3-0TC1-Drin4.4

ID based on morphology

l-S3-O TCl-Saop 3.4

NA
inconclusive
white

l-S3-O TC l-Saop3.6

inconclusive

2-S3-C3-Saop 1.2

white

2-S4-C2-Saar 5.7
2-S4-C2-Saop 2
2-S4-C2-Saop 2.2
2-S4-C2-Saop 2.4
RFLP type 2 m
l-S3-C l-Saop4.19
l-S4-OTC3-Saop2.1
RFLP type 3 m
l-Sl-C 3-Saop 3.3
1-S3-C Drin 4.9
1-S4-C1-Saar 4.3
1-S4-C2-Saar3.3
1-S4-0TC1-Saar 4.6

inconclusive
NA
Staphylotrichum
inconclusive

2-S4-C2-Saop 44A
2-S4-C2-Saop 4.9
2-S4-OTC1-Saar 4.1
RFLP type 4 m
2-S3-OTCl-Saar 1.9
1 D/G-Saar 3.3
RFLP type 5m
2-S3-C3-Saar 1.7
2-S3-C3-Saar 1.8
2-S3-C3-Saar3.1
2-S3-OTC1-Saar 4.8
2-S3-OTC1-Saar 5.1
2-S3-0TCl-Saar 5.5
1-D/G-Saar 3.4

dark sterile
NA
inconclusive
inconclusive
inconclusive
Phialocephala fortinii
inconclusive
dark septate
inconclusive
NA
NA
dark sterile
NA
NA
NA
NA
NA
NA
dark sterile

194

Table A3 . 6 Comparison of identification (based on molecular techniques) with
corresponding morphological identification - Dothideales. Samples showed affiliation to the
ID based on morphology

ID based on RFLP and sequence
analyses: Dothideales
l-Sl-C 3-Saop 4.1
1-Sl-O TCl-Saop 1.2

NA

1-Sl-O TCl-Saop 2.1

inconclusive

l-Sl-O TC2-Saop 2.2

Phialocephala fortinii

l-Sl-O TC2-Saop 5.3

Phialocephala fortinii

l-S4-O TC l-Saop3.9
2-S l-C l-Saop 1.6
2-Sl-OTC2-Saop 3.3

hyaline Phialocephala fortinii -like
dark sterile
Acremonium

2-Sl-OTC2-Saop5.1

black type

NA

Table A3.7 Comparison o f identification (based on molecular techniques) with
corresponding morphological identification - Phoma. Samples showed affiliation to Phoma
ID based on RFLP and sequence
analyses: Phoma
RFLP type 1?
l-S3-OTC2-Drin2.2A
l-S 4-C l-Saop4.7
l-S4-C2-Saop 2.2
l-S4-OTC2-Saop L IB
l-S4-OTC2-Saop 2.2
l-S4-OTC3-Saop 3.9
l-S4-OTC3-Saop 4.2
2-S3-OTC3-Saar 2.3
2-S3-OTC3-Saar 3.3
RFLP type 2?
l-Sl-O TC 2-Saop3.2
l-Sl-O TC 3-Saop3.2
l-S 3-C l-Saop4.2

ID based on morphology

NA
NA
inconclusive
Monodictys
Monodictys
NA
Trichocladium
hyaline sterile
dark sterile
inconclusive
Monodictys
inconclusive

195

Table A3 . 8 Comparison o f identification (based on molecular techniques) with
corresponding morphological identification - Cryptosporiopsis. Samples showed affiliation
to Cryptosporiopsis according to seq uence and RFLP analyses
ID based on RFLP and sequence ID based on morphology
analyses: Cryptosporiopsis
RFLP type 1l
l-Sl-C 3-Saop 3.2
dark sterile
NA
1-S4-Cl-Cate 2.2
NA
1-S4-Cl-Cate 2.4
1-S4-C2-Saar3.4
Cryptosporiopsis radicola
l-S4-C2-Saop 3.2
brown
inconclusive
l-S4-C2-Saop3.3
inconclusive
l-S4-C2-Saop 3.4
inconclusive
l-S4-C2-Saop4.5
inconclusive
1-S4-C3-Saar3.1
1-S4-C3-Saar5.1
l-S4-C3-Saop3.6
1-S4-CTL-Cate 1.2

dark sterile monolioid
hyaline sterile
hyaline sterile

l-S4-OTC2-Saop 3.2
l-S4-OTC3-Saop3.1
l-S4-OTC3-Saop4.11
1-S4-CTL Cate 3.9
2-Sl-C l-Saop 4.1
2-S3-OTC3-Drin 2.2
2-S4-C3-Saar 3.3
2-S4-CTL-Cate 3.6
RFLP type 2 l
1-S3-C2-Drin2.2
1-S3-C2-Drin5.6

inconclusive
hyaline sterile
inconclusive
Cryptosporiopsis radicola
inconclusive
hyaline sterile
white
white

Cryptosporiopsis radicola
NA

196

Table A3.9 Comparison o f identification (based on molecular techniques) with
corresponding morphological identification - Cadophora. Samples showed affiliation to
ID based on morphology

ID based on RFLP and sequence
analyses; Cadophora
RFLP Type Ica
2-S3-C3-Saop 2.2
l-S4-O TC l-Saop4.5

dark sterile
NA

RFLP Type 2ca
1-SI-C l-Saar 3.6
1-SI-OTC 1-Drin 4.2
1-Sl-0TC 3-C ate4.1

Phialocephala fortinii
Phialocephala fortinii
inconclusive

RFLP Type 3ca
1-S3-0TC2-Drin2.4
1-S4-C2-Saar2.10
1-S4-Ctl Cate 2.1 A
2-S4-CÜ Cate 3.1
2-S4-CÜ Cate 3.3

inconclusive
hyaline sterile
Phialocephala fortinii
inconclusive
NA

Table A3.10 Comparison o f identification (based on molecular techniques) with
corresponding morphological identification - Colispora. Samples showed affiliation to
Colispora according to sequence anc RFLP analyses
ID based on RFLP and sequence ID based on morphology
analyses: Colispora
1-S3-C-Drin5.2
l-S4-OTC3-Saop 5.4

NA
NA

Table A3.11 Comparison o f identification (based on molecular techniques) with
corresponding morphological identification - Agarieales. Samples showed affiliation to the
ID based on RFLP and sequence
analyses: unknown Agarieales
1-SI-C l-D rin 1.6
1-S4-C1-Saar 4.9
2-S4-C1-Saar 3.11
2-S4-C2-Saop 4.3

ID based on morphology
hyaline sterile
hyaline sterile
hyaline sterile
hyaline sterile

197

Table A3.12 Comparison o f identification (based on molecular techniques) with
corresponding morphological identification -Ceratobasidium. Samples showed affiliation to
ID based on RFLP and sequence
analyses: Ceratobasidium
1-Sl-0TC 2-Saar2.1
1-Sl-OTC2-Saar2.10
1-S3-C2-Saar 1.3
1-S3-C2-Saar2.3
1-S3-C2-Saar5.3
l-S3-OTC2-Saop2.1
1-S4-Cl-Cate 5.4
2-Sl-C 2-Saop2.2
2-Sl-OTC2-Drin3.1

ID based on morphology
hyaline sterile
hyaline sterile
hyaline sterile
hyaline sterile
hyaline sterile
hyaline sterile
hyaline sterile
inconclusive
hyaline sterile

198

IX. Appendix for chapter 4
Table A4.1 Genotype richness for host plants at chronosequence zone. Number preceded by
‘Y ’ represent the earliest time o f exposure from the glacier. <Y59 is the control, or réfugia
Luzula confusa
Plot
Y90
Y80
Y70
Y60
control
Papaver
lapponicum
Plot
Y90
Y80
Y70
Y60
control
Salix arctica
Plot
Y80
Y70
Y60
control
Scaifraga
oppositifolia
Plot
Y70
Y60
Cassiope
tetragona
Plot
Y60
control
Dryas
integrifolia
Plot
Y60
control

species richness
IT Sl-//m fI
16
18
14
19
37
species richness

ITS4-//mfI
7

IT S l-^ M

1TS4-^M

22

11

6

22

10

9

8

14

16
30
39

18
24

ITSl-Hinü

ITS4-/7mfI

IT S l-^ M

IT S4-^M

11

11

NA
9
9
NA
5

species richness
IT S l-//w fI

17
22

13

12

12

10

4

27
28

11

13

21

11

14

5

ITS4-//mfl

ITSl-^/wI

12
6

20

IT S4-^M
16
6

14
16

14
18
24

ITSl-7/mfI

ITS4-//mfI

ITSl-yfM

ITS4-v4M

14

6

9

11

8

20

9
14

IT Sl-//m fI

nS4-H infl

IT S l-^ M

IT S4-^M

6

3
14

14
33

8

IT S l-//m fl

nS4-H m fl

ITS 1-^4M

IT S4-^M

13
15

5

12

12

11

14

14

13
22

23
species richness

15
14

species richness

18
species richness

18

199

Table A4.2 Sorensen’s Quantitative Index between host plants and chronosequence zones. I-h = ITS 1-HinfI; 4-h=ITS 4-HinfI; i-a -IT S 1-AluI; 4-a-IT S 4-AluI.
Plot <Y59 is the control. N um ber preceded by ‘Y ’ represent the earliest time o f exposure from the front o f the receding glacier. <Y59 is the control, or réfugia plot that

Y90Pala Y80
Y

90
L
u
c
0
Y

80
L
u
c
0
Y
70
L
u
c
0
Y
60
L
u
c
0
Y
59
L
u
c
0

Y80
Pala

Y80
Saar

Y70
Luco

036

034

Y70

Y60

Y60

Y60

Y60

Cate
0.12

Drin
0.05

Luco

Pala

0.06

Saop
0.11

037

Y70
Saar

038

Y70
Pala
0.18

Y60
Saop
0.14

<Y59

<Y59

<Y59

<Y59

<Y59

Cate
0.18

Drin
0.11

Luco
0.2

Pala
0.19

Saar

038

Y60
Saar
0.2

1-h

0.25

Luco
0.47

4-h

0

0.55

0J3

034

038

0.21

0.27

0.38

0.17

0.23

0.31

0.2

039

034

033

0.08

039

0

0.24

1-a

0.12

0^6

026

032

035

032

0.11

0.13

0.03

0.17

034

0.1

0.2

0.13

0.04

0.14

033

0.04

0.19

4-a

0

039

036

0.18

032

0.07

0

0.24

0.05

0.09

035

0.1

0.07

039

032

0.12

039

0.04

0.07

1-h

0.2

0.14

033

038

0.11

033

0.16

0.17

0.44

0.13

037

0.3

0.21

0.37

045

046

032

4-h

036

0.1

046

0.17

0.19

0.26

0.09

0.16

036

0.1

0.2

0.19

0.17

0.07

0.48

0.05

0.18

1-a

034

0J3

0.4

0.18

0.11

0.06

0.11

0.07

0.43

032

0.1

0.14

035

0.09

0.48

0.19

035

0.14

037

0.04

0.21

0.11

0.02

0.11

038

0.08

0.16

0.4

038

0.17

0.04

0.13

0.08

1-h

0.13

048

0.04

0.09

0.06

042

0.14

0.14

0.1

0.2

0.04

0.43

0.1

0.2

4-h

0.05

0.18

0.16

0.17

0.13

0.61

0.18

0.16

0.14

0.18

0.12

042

0.15

033

1-a

0.15

035

0.06

0.07

0.2

032

0.18

0.18

0.19

0.17

0.07

0.53

0.11

0.24

4-a

0.05

0.17

0

046

0.18

0.61

0.15

0.19

0.14

0.07

0.05

048

0.06

0.2

4-a

0.11

046

1-h

0.11

0.19

0.15

0.21

0.15

0.42

0.1

033

4-h

0.15

033

0.2

0.2

0.1

0.46

0.11

037

1-a

036

0.19

0.19

0.14

0.12

0.47

0.18

0.2

4-a

0.15

035

0.34

0.16

0.08

0.48

0.06

032

1-h

0.06

035

4-h

0.17

0.3

1-a

0.16

0.24

4-a

0.02

0.19

Table A4.2 Sorensen’s Index, l-h = ITS \ - H M ; 4-h=ITS 4-7/wfI; l-a=lT S \-A M \ 4-a=ITS A-AM (cont’d). Num ber preceded by ‘Y’ represent the earliest time o f

1-h

Y80
Pala
0.47

Y80

Luco
0.13

4-h

0

0

Y80
Y
90
P
a
1
a
Y

80
P
a
1
a
Y
70
P
a
1
a
Y
60
P
a
la
<
Y
59
P
a
1
a

Y60

Y60

Y60

Y60

Y60

Cate
0.18

Drin
0.03

Luco
0.08

Pala
0.58

Saar

CU3

Y70
Saop
0.03

0

0

0

0

0.02

0.18

OIK

0

Y70
Saar

0.06

Y70
Pala
0.41

OIK

0.1

Saar

Y70
Luco

OIK
016

032

Y60
Saop
0.04

<Y59

<YM

<Y59

<Y59

<Y59

Cate
0.19

Drin
0

Luco
0.08

Pala
0.27

Saar
0.13

OIK

0

0.03

0.2

0

018

0.11

0.02

0.15

OIK

0.12

OjK

0

0

OIK

0

0.14

0A6

017

0

0.18

0.4

0.18

0.05

0

o jy

016

0

0.07

0

0.02

0.34

0

0

0

0

0.03

0.1

0.05

1-h

0.12

0.14

0.5

CU3

0.15

0.14

0.08

0.11

0.51

0.17

0.13

035

OIW

0.19

0.3

0.2

4-h

OJK

0.3

0.17

CU5

0.24

0.26

0.16

033

0.31

034

033

037

032

0.28

0.3

0.19

1-a

0.13

OJK

0K9

018

0.07

0.1

OIK

032

037

0.04

034

0.18

0.1

0.31

0.3

0.1

4-a

0.11

0J3

0.24

0

0.11

0.15

0

0.28

036

0.04

0.13

0.1

0.13

0.16

0.1

0.08

1-h

0.2

0.17

0.14

033

0.19

0.49

0.25

0.3

0.24

0.2

034

0.3

035

4-h

0.1

035

0.13

0.08

0.04

0.15

0.14

0.19

0.14

0.17

0.1

0.1

0.22

1-a

0.15

0.09

01#

0.15

0.16

037

0.1

039

0.13

0.07

0.21

0.3

032

4-a

0.19

0.18

032

0.03

0.17

035

035

033

0.05

0.08

0.13

0.4

0.3

1-a
4-a

&23
0

1-h

0.15

0.1

032

0.11

0.1

0.3

0.18

4-h

0.2

0.36

038

0.21

0.17

0.2

036

1-a

0.07

0.11

0.16

0.07

0.21

0.6

0.04

4-a

0.08

035

0.13

0.05

0.15

0.3

0.17

1-h

0.18

4-h

0.18

1-a

0.02

4-a

0.08

Table A4.2 Sorensen’s Quantitative Index, 1-h = ITS
4-h=ITS 4-//m fI; 1-a=ITS \-Alul; 4-a=ITS A-Alul (cont’d). Number
preceded by ‘Y ’ represent the earliest time o f exposure from the glacier. <Y59 is the control, or réfugia plot that was not covered by

Y80
Saar

Y70
Saar

Y60
Saar

to

s

1-h

4-h
1 -a
4-a
1 -h

4-h
1 -a
4-a
1 -h
4-h
1 -a
4-a

Y70
Luco
0.14
0.16
0.2
0.17

Y70
Pala
0.17
0.17
0.13
0.13

Y70
Saar
0J8
0.3

Y70
Saop
0.06
0.13

Y60
Cate
0.17
0.09
0.19
0.08

Y60
Drin
0.18
0.16
0.16

0.12

0

0J 3

0.17
0.09
0.46
0.07

0.19

0

025

0.19
0.47
0.25
0.25

0.11
0

0.11

Y60
Luco

Y60
Pala

0.12

0.1

0.19
0.14
0.27

0.26
0.14
0.16

0.05
0.25
0.25

0.14
0.25
0.13

0.2

0.1

Y60
Saar
0.61
0.39
0.19
0.45

Y60
Saop
0.19
0.19
0.23
0.36

<Y59
Cate
0.15
0.25
0.23
0.18

<Y59
Drin
0.09
0.31
0.16

0.35
0.53
0.4
0.48

0.17
0.17
0.19
0.3
028
0.15

0.12

0.1

028
0.07

0.19
0.13
0.09
0.17
0.17
0.25
0.27

0.1
0.21

0.21

0.37
0.07

0.31

0.11

0.11

<Y59
Luco
0.14
0.17
0.18
0.23

<Y59
Pala
0.09
0.16
0.04
0.12

0.16
0.35

0.06
0.06

0.12

0

0.14
0.25

0

0.21

0.23
0.16

0.12

0.08
0.03
0.07

<Y59
Saar
0.3
0.3
0.15
0.29
028
023
028
0.44
025
0.56
0.42
0.48

Table A4.2 Sorensen’s Quantitative Index, 1-h = ITS l-//m fl; 4-h=ITS 4-//mfI; l-a=ITS \-Alul; 4-a=ITS 4-Alu\ (cont’d). Number
preceded by ‘Y ’ represent the earliest time o f exposure from the glacier. <Y59 is the control, or réfugia plot that was not covered by

0

Y60
Drin
0.18
0.42
0.05

0.18

0

Y60
Cate
Y70
Saop

Y60
Saop

1-h
4-h
1 -a
4-a
1 -h
4-h
1-a
4-a

0.1

0.13

Y60
Luco
0.1

0.19
0.04
0.17

Y60
Pala
0.05
0.16
0.02

Y60
Saar
0.14
0.39
0.04

0.06

0.12

Y60
Saop
0.34
0.34
0.09
0.36

<Y59
Pala
0.03
0.08

0

<Y59
Luco
0.25
0.18
0.04

0

0.1

0.24
0.14
0.09
0.13

0.16
0.16
0.25
0.26

0.06
0.08
0.19
0.06
0.05

<Y59
Cate
0.13
0.25

<Y59
Drin
0.33
0.13

0.02

0.08
0.13
0.15
0.17
0.17

0

<Y59
Saar
0.16
0.45
0.03
0.04
0.2

0.32
0.34
0.26

Table A4.2 Sorensen’s Quantitative Index, 1-h = ITS 1-HinfI; 4-h=ITS 4-HinfI; l-a=ITS 1-Alul; 4-a=lTS 4-Alul (cont’d). Number
preceded by ‘Y ’ represent the earliest time o f exposure from the glacier. <Y59 is the control, or réfugia plot that was not covered by
Y60
Drin
Y60
Cate

<Y59
Cate

to

o

1-h

0.1

4-h
1-a
4-a
1 -h
4-h
1 -a
4-a

0.09
0.03
0.03

Y60
Luco
0.08
0.14
0.1

0.19

Y60
Pala
0.17
0.35
0.15
0.18

Y60
Saar
0.21

0.07
0.03
0.19

Y60
Saop
0.19
022
0.18
0.25

<Y59
Cate
0.37
0.18
0.29
0.21

<Y59
Drin
0.11

0.09
0.07
0.05
0.21

0.2
023
0.07

<Y59
Luco
0.17

<Y59
Pala
0.03

0.11

0

0.12

0.08

0.04

0
0.11

0.21
0.2

0.04
0.08
0.03

0.26
0.14

0.09
0.15
0.3
0.17
0.21

028

<Y59
Saar
0.25

0.11

Table A4.2 Sorensen’s Quantitative Index, 1-h = ITS 1-i/mfI; 4-h=ITS 4-//mfI; l-a=ITS l-Alul; 4-a=ITS 4-Alul (cont’d). Number
preceded by ‘Y ’ represent the earliest time o f exposure from the glacier. <Y59 is the control, or réfugia plot that was not covered by

0.12

Y60
Pala
0.13

Y60
Saar
0.27

Y60
Saop
0.39

<Y59
Cate
0.09

<Y59
Drin
0.13

<Y59
Luco
0.17

<Y59
Pala
0.06

<Y59
Saar
0.16

0.24

0.1

0.47

0.13

0.25

0.16

0.17

0

038

0.12

0.07

0J 8

0.15

0.1

0.1

0.24

0

0.25

0.1

0.07

032

0.19

0.06

0.18

0.12

0

032

033

0.03

0.21

0.2

0.16

0.18

0.15

0.02

0.11

0.16

0

0.25

Y60
Luco
Y60
Drin

<Y59
Drin

w
oLA

1h
4h
1a
4a
1h
4h
Ia
4a

A xis 3
Y 90 Pala
Cate

Y 60 Cate
< 59 Pala

Y 60 Pala

Y 80 Pala

<59 Luco
A xis 2

Y70LÏI
Y 60 Drin

'80 Saar^Y 60 Saop

Y 80 Luco

Y 60 »aar
<59 Saar

Fig. A4.1 NMS o f host plant and chronosequence zones. Rotated at 315°. Number preceded
by ‘Y ’ represent the earliest time o f exposure from the glacier. <Y59 is the control, or
réfugia plot that was not covered by the glacier. Plants are: Cate=Cassiope tetragona, Drin
= Dryas integrifolia, Luco=Luzula confusa, Vd\2i=Papaver lapponicum, Saar=Salix arctica,
Saop=5'ox//raga oppositifolia.

206

(b) Exchangeable K

(a) Exchangeable Al

Y70

Xîa95*SE

Y60
Year <fe>posure

Y ear o f exposure

(c)Exchangeable Na

(d) Exchangeable Mg

* 1.6

Y60

<Y59

(f)Exchangeable Mn

(e) Exchangeable Ca

Year o f exposure

(h)pH

(g)Exchangeable Fe

mO.OlO

Y60

<Y59

Y80

Y70

Y60

Fig A4.2 LS means o f soil property for each chronosequence zone. Error bars designate 1
SE. Number preceded by ‘Y’ represent the earliest time o f exposure from the glacier. <Y59
is the control, or réfugia plot that was not covered by the glacier.

207

(i) C:N ratio

G) NO 3 -N
70
60
50

14,6

40

Z13.1

13.2

10

13.0
Y90

Y80

Y70

Y60

<Y59

B Mean

Y90

Year o f exposure

Y80

Y70

Y60

<Y59

B Mean

Year o f exposure

(k) Available P

(1) NH 4-N

10
14

9

12

8
7

10

6

8

>
g5

5

4

4

3

2

2
Y90

Y 80

Y 70

Y 60

< Y 59

■

M ean

0

Y90

Y e a r o f exposure

Y80

Y70

Y60

<Y59

B Mean

Year o f exposure

(m) CEC

(n) Total N

14

0.34

0.30

12

8

0.16
6

0.14
0 , 12

4
Y90

Y80

Y70

Y60

Year cf e.>4) osu e

<Y59

B Mean

0 .1 0

Y90

Y80

Y70

Y60

<Y59

Y earof exposure

Fig A4.2 LS means o f soil property for eaeh ehronosequenee zone (cont’d). Number
preceded by ‘Y ’ represent the earliest time o f exposure from the glacier. <Y59 is the
control, or réfugia plot that was not covered by the glacier.

208

ITS4-///wfi: y= 1.7667+1.0333* x, 0.95 Fred.Int.

y= 5.5556+1.0056*x; 0.95 Pred.Int.

16

10

14

9

8

12

7

10

6
8
5
6

I 4

S

3.
S-

3.

52

I

0
Y 90

Y80

Y70

Y 60

3

2

<Y59

Y90

Y80

Yearofe?q50Sure

Y 70

Y60

<Y59

Y earo fex p o su re

ITSl-^lM l: y = 5.2 6 6 7 + 2 .1333*x; 0.95 Pred.lnt.

AIu 55 = 2 .3 + I.2 3 3 3 * x ; 0.95 Pred.lnt.
14

26
24
12

22
20

10

8
I?• 4

I

2
Y 90

Y80

Y70
Year o f exposure

Y60

< Y 59

Y90

Y80

Y70

Y 60

<Y59

Y ear ofe?q)osure

Fig. A4.3 Simple regression for genotype richness for Luzula confusa. Number preceded by ‘Y ’ represent the earliest time o f
exposure from the glacier. <Y59 is the control, or réfugia plot that was not covered by the glacier.

y = 6.0333+ 6.6333*x; 0.95 Pred.lnt.
ITS4-//jwfl = 3.1667+4.6333*x; 0 95 Pred.lnt.

70
45
60

40
35

50

30
40

25

30

20
15

o 20
10

r

^ 0

Y 90

Y 80

Y6Q

<Y59

I

Y 90

Y 80

Y 70

Y 60

<Y59

year o f exposure

1TS4-/J/W1: y =4.1+4.1*x; 0.95 Pred.lnt.

lTSl-y4/wl: y = 6 .56 6 7 + 6 .5667*x; 0.95 Pred.lnt.
40

70

35

60

30
50
25
40
20
30
15
o 20
I 10

I 5
s
Y90

Y80

Y 70

Y60

<Y59

I 0

Y 90

Y80

Y 70

Y 60

< Y 59

yea ro fe x p o su re

Fig. A4.4 Simple regression o f genotype cumulative frequency o f Luzula confusa. Number preceded by ‘Y ’ represent the earliest time
o f exposure from the glacier. <Y59 is the control, or réfugia plot that was not covered by the glacier.
K)
O

(a) ITS \-H inü

(b) ITS \-A lul

3.5
3

In Y 90

-■ In Y90

2.5

hY 70

hYSO
hY70

- - I n Y60

- - « - - I n Y 60

--In Y S O

2
1.5

1

h<Y 59

— h< Y 59|

0.5

0
I

4

7

10

13

16

19 22 25

28

31

34 37

7

40

10

13

16

19

22

25

28

31

34

37

Rank genotype abundance

Rank genotype abundance

(c) ITS 4-HinÜ

(d) ITS A-AM

3.5

In Y90
-- + --ln Y80
In Y70
—
lnY60
—
hi <Y59

- tc-- In Y 90

— ‘ ù e -‘

I
M 1
0.5

-+ --ln Y 8 0
In Y 70
— • — In Y 60

\

—
* A' ■A-A

1

2

3

4

5

6

7

8

9

10

11

12 13

14

Rank genotype abundance

Fig. A4.5 Rank abundance curve for each chronosequence zone for Luzula confusa.

7

9

11

13

15

17

19

Rank genotype abundance

21

23

25

ln < Y 5 9

(a) ITS \-HinQ.

(b) ITS \-A M
3 .5
-

-■ ln Y 9 0
-In Y S O

C
5

A-

-- ln Y 6 0

In Y 8 0 j

Q

h Y 70

j

— ♦— ln Y 6 0

I

—

— ln < Y 5 9
* -*
7

9

11

13

15

hY 90 I

- '4 —

■■ h Y 7 0

-> 0.5

A-

2 .5

ln < Y 5 9

-a - o -Q- a -a- -o
17

19

21

1

23

3

5

Rank genotype abundance

7

9

11

13

15

17

19

21

23

25

27

R a n k g e n o ty p e a b u n d a n c e

(c) ITS 4-//m fI

(d) ITS A-Alu\
2 .5

2
6-

1.5

In Y 90

ÿ

- • Û - - In Y 9 0
-■ + - - l n Y 8 0

+ --ln Y 8 0

0.5

■a-- - In Y 70

a

- * - - l n Y 60

— InY ôO

ln < Y 5 9

2

0 .5

—

0
3

5

7

9

Rank genotype abundance

3

5

Fig. A4.6 Rank abundance curve for each chronosequence zone for Papaver lapponicum..

w
to

7

9

Rank genotype abundance

13

In Y 7 0

ln < Y 5 9

(a) ITS \-H in ü

(b) ITS \-A lu\
3

2.5

2.5
■-lnY80

2

■a

B
3

• • -H- - • In Y 7 0

■■■lnY70

1.5

--ln Y 6 0

- In You

ec

— h<Y59

h<Y59
-I 0.5
0
1

3

5

7

9

1

11 13 15 17 19 21 23 25 27

3

5

7

9

11

13

15

17

19 21

23

R a n k g e n o ty p e a b u n d a n c e

R a n k g e n o ty p e a b u n d a n c e

(c) ITS 4-7/mfI

(d) ITS A-AM

2 .5

-■+--lnY80
a

In Y70

- - * - - ln Y60
—
- I 0.5

ln<Y59

?

— H - In Y 8 0

1

. ..Q .. . In Y 7 0
— In Y 6 0
— «— In < Y 5 9
0.5

12 13 14 15 16 17
1

2

3

R a n k g e n o ty p e a b u n d a n c e

Fig. A4.7 Rank abundance curve for each chronosequence zone for Salix arctica..
(a) ITS \-HinH
(b) ITS \-Alu\
w
U)

4

5

6

7

8

9

10 11 12 13 14 15 16

Rank genotype abundance

2 .5

s' 2
I ..s

I
J

c
Q

In Y 7 0

— • — In Y 6 0

I

\

0 .5 - I

\

\

k— # —
1

1.5 -

\

■o

\

2

3

4

5

6

^ —#

*

# —#- - - Q- - - □

7

9

10 11 12 13 14 15 16

8

1

\

1

s
5

In Y 7 0

V

--« --ln Y 6 0

«

-1
■Q-•Q-■o --a3

5

7

9

11

13

15

17

19

R a n k g e n o ty p e a b u n d a n c e

R a n k g e n o ty p e a b u n d a n c e

(c) ITS

(d) ITS A-AM

2.5

2

s' 2

1.5
-a-- - In Y 70

i 1
g
J

. \
n,

a
I

- * - - l n Y 60

\
\

5

0.5

0-5 H

0
2

3

4

5

6

R a n k g e n o ty p e a b o n d a n c e

7

in Y70

--•--InYôO

1

2

3

4

5

6

7

8

9

10

R a n k g e n o ty p e a b u n d a n c e

Fig. A4.8 Rank abundance curve for each chronosequence zone for Saxifraga oppositifolia

11

12

13

14

(b) ITS \-A lul

(a) ITS 1-HinfI

■CD

t
Î

3 .5

3

3

2 .5

2 .5

2

2

-0 -- In Y 6 0

1.5

- » — ln < Y 5 9

■ ln Y 6 0
1.5
-ln < Y 5 9

1

1

0 .5

0 .5

0

0
1

3

5

7

9

11

13

15

17

19

21

23

25

1

27

3

5

7

9

11 13 15 17 1 9 2 1 2 3 2 5 2 7 2 9 31 3 3
R a n k g e n o ty p e a b u n d a n c e

R a n k g e n o ty p e a b u n d a n c e

(d) ITS A-AM

(c) ITS A-Hinü

2 .5
2 .5

2
1.5

ee

-e- - • In Y 6 0

hY 60

1!

ln < Y 5 9

- ln < Y 5 9

0 .5

0 .5

0
1

2

3

4

5

6

7

8

9

10 11

R a n k g e n o ty p e a b u n d a n c e

12 13

14

15

1

3

5

7

LA

11

13

15

17

19

R a n k g e n o t y p e a b i u id a n c e

Fig. A4.9 Rank abundance curve for each chronosequence zone for Cassiope tetragona

to

9

21

23

25

(a) ITS \-HinÛ

(b) ITS \-A M

2 .5 3

2

2 .5

1.5

In Y 6 0

2

•— h< Y 59|

1.5

e

1

■ In Y 6 0
- ln < Y 5 9

1H

0 .5
0.5

0

0
1

2

3

4

5

6

7

8

9

10 11

12

13

14

15

3

4

R a n k g e n o ty p e a b u n d a n c e

5

6

7

8

9

10

11

R a n k g e n o ty p e a b u n d a n c e

(c) ITS 4-///«fI

(d) ITS 4-Alul

2. 5

2

2. 5

1. 5

In Y 6 0

1

-h < Y 5 9

• ln Y 6 0
- ln < Y 5 9

0. 5

0
1

2

3

4

5

6

7

8

9

10

11

12

13

14

1

2

R a n k g e n o ty p e a b u n d a n c e

Fig. A4.10 Rank abundance curve for each chronosequence zone for D ryas integrifolia.

K)

3

4

5

6

7

8

9

10

R a n k g e n o ty p e a b u n d a n c e

11

12

13

14

X. Appendix for chapter 5
Table A5.1 Comparison o f LS means o f soil chemistry between succession study (control

Soil
chemistry
A1
K
Na
Mg
Ca
Mn
Fe
pH
C:N ratio
NO 3
P
NH 4
CEC

Succession
site
<Y59

Lowland
Control

OTC

Control

OTC

Control

OTC

0.90
&28
0.76
1.45
150
0.019
0.019
5.6
14.5
18
7.5
9.0
6.3

0.055

< 0.01 0
0.075
0.045
3.00

0.055

0.0 1 0

0.10

0.12

0.070

0.040
0.18
0.045

0.080
3.40
4.00

0.060
4.00
4.20
0.018

Dolomitic

Granitic

220

2.0 0

< 0.0 1 0
0.075
0.040
3.50

4.00
0.025
0.0060

3.80
0.023

8.00

6.00

0 .0 0 1 0

0 .0 0 1 0

0.0 0 2 0

0.0018

0.0018

6.0
20

6.0

8.0

20

50
4.0

40
4.0

8.0
100
10

0.5

1.8

1.0
6.0

1.0
11

0.20

7.0

0.20
0 .0 0 2 2

0 .0 2 0

140

6.0
20

6.0
20

10

40

Negligible
0.90

2.1

6.5

10

8.0

35
1.9
3.5
9.0

217