Citation
Assessing whitebark pine vigor and facilitation roles in the alpine tree ecotone

Material Information

Title:
Assessing whitebark pine vigor and facilitation roles in the alpine tree ecotone
Creator:
Blakeslee, Sarah C
Place of Publication:
Denver, Colo.
Publisher:
University of Colorado Denver
Publication Date:
Language:
English
Physical Description:
1 electronic file. : ;

Thesis/Dissertation Information

Degree:
Master's ( Master of Science)
Degree Grantor:
University of Colorado Denver
Degree Divisions:
Department of Integrative Biology, CU Denver
Degree Disciplines:
Biology
Committee Chair:
Tomback, Diana F.
Committee Members:
Wunder, Michael B.
Bruederle, Leo P.

Subjects

Subjects / Keywords:
Whitebark pine ( lcsh )
Whitebark pine -- Ecology ( lcsh )
Mountain ecology ( lcsh )
Genre:
non-fiction ( marcgt )

Notes

Review:
Whitebark pine (Pinus albicaulis) is an upper subalpine and treeline conifer of the higher mountains of the western United States and Canada. At treeline on the Eastern Front of the Rocky Mountains, whitebark pine appears to facilitate tree island development. It is currently declining at treeline from infection by white pine blister rust, caused by Cronartium ribicola. We are studying how whitebark pine facilitates tree island formation and how blister rust mortality may affect these processes in two treeline study areas in Montana: Divide Mountain, Glacier National Park and Blackfeet Indian Reservation; and Line Creek Research Natural Area, Custer National Forest. We tested three hypotheses: 1) Whitebark pine is hardier than other treeline conifer species, as demonstrated by more vigorous growth and survival at treeline, 2) whitebark pine provides more favorable leeward microsites for tree island recruitment than other conifers or microsites, and 3) death of windward whitebark pine in established tree islands leads to vigor loss in leeward conifers. We found support for each hypothesis. Whitebark pine was significantly more numerous than both spruce (Picea engelmannii) and fir (Abies lasiocarpa) among solitary trees. Solitary, krummholz whitebark pine trees produced significantly longer annual shoots than both spruce and fir, indicating faster branch growth and canopy area increase under harsh conditions. These results indicate higher vigor and potentially higher survival rate than spruce and fir. Germinated spruce seeds had higher summer survival, and planted fir and spruce seedlings had greater vigor, when leeward of whitebark pine compared to spruce, rock, or exposed microsites, suggesting that whitebark pine microsites provided more protection. In established tree islands, the presence of a windward whitebark pine was associated with greater general vigor, longer shoot lengths, and lower shoot mortality in leeward trees than under experimental conditions where the windward whitebark pine was girdled and defoliated. Because whitebark pine is better able to survive and grow in the alpine treeline ecotone than other conifer species, this may, in part, explain its greater prevalence. Whitebark pine is more likely to facilitate tree island development, and provide a better microsite for seedling establishment.
Thesis:
Thesis (M.S.)--University of Colorado Denver. Biology
Bibliography:
Includes bibliographic references.
General Note:
Department of Integrative Biology
Statement of Responsibility:
by Sarah C. Blakeslee.

Record Information

Source Institution:
|University of Colorado Denver
Holding Location:
|Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
860847889 ( OCLC )
ocn860847889

Downloads

This item has the following downloads:


Full Text
ASSESSING WHITEBARK PINE VIGOR AND
FACILITATION ROLES IN THE ALPINE TREELINE ECOTONE
by
SARAH C. BLAKESLEE
B.S. BiologyUniversity of Colorado Colorado Springs2008
A thesis submitted to the
Faculty of the Graduate School of the
University of Colorado in partial fulfillment
of the requirements for the degree of
Master of Science
Biology
2012


This thesis for the Master of Science degree by
Sarah C. Blakeslee
has been approved for the
Department of Integrative Biology
by
Diana F. TombackChair
Michael B. Wunder
Leo P. Bmederle
16 November2012
li


Blakeslee, Sarah, C. (M.S., Department of Integrative Biology Master of Science)
Assessing Whitebark Pine Vigor and Facilitation Roles in the Alpine Treeline Ecotone
Thesis directed by Professor Diana F. Tomback
ABSTRACT
Whitebark pine (Pinus albicaulis) is an upper suoalpine and treeline conifer of the
higher mountains of the western United States and Canada. At treeline on the Eastern
Front of the Rocky Mountains, whitebark pine appears to facilitate tree island
development. It is currently declining at treeline from infection by white pine blister mst,
caused by Cronartium ribicola. We are studying how whitebark pine facilitates tree
island formation and how blister mst mortality may affect these processes in two treeline
study areas in Montana: Divide Mountain, Glacier National Park and Blackfeet Indian
Reservation; and Line Creek Research Natural Area, Custer National Forest. We tested
three hypotheses:1)Whitebark pine is hardier than other treeline conifer species, as
demonstrated by more vigorous growth and survival at treeline, 2) whitebark pine
provides more favorable leeward microsites for tree island recruitment than other conifers
or microsites, and 3) death of windward whitebark pine in established tree islands leads
to vigor loss in leeward conifers.
We found support for each hypothesis. Whitebark pine was significantly more
numerous than both spmce (Picea engelmannii) and fir (Abies lasiocarpa) among solitary
trees. Solitary, kmmmholz whitebark pine trees produced significantly longer annual
shoots than both spmce and fir, indicating faster branch growth and canopy area increase
under harsh conditions. These results indicate higher vigor and potentially higher
survival rate than spmce and fir. Germinated spmce seeds had higher summer survival,


and planted fir and spmce seedlings had greater vigor, when leeward of whitebark pine
compared to spmce, rock, or exposed microsites, suggesting that whitebark pine
microsites provided more protection. In established tree islands, the presence of a
windward whitebark pine was associated with greater general vigor, longer shoot lengths,
and lower shoot mortality in leeward trees than under experimental conditions where the
windward whitebark pine was girdled and defoliated. Because whitebark pine is better
able to survive and grow in the alpine treeline ecotone than other conifer species, this
may, in part, explain its greater prevalence. Whitebark pine is more likely to facilitate
tree island development, and provide a better microsite for seedling establishment.
The form and content of this abstract are approved. I recommend its publication.
Approved: Diana F. Tomback
IV


DEDICATION
I dedicate this work to Logan, in appreciation of all his love and support.


ACKNOWLEDGMENTS
I would like to thank my advisor Dr. Diana F. Tomback for her willingness to
share her knowledge and experience and for assisting me throughout this entire process.
Also invaluable to this work are Jill C. Pyatt and Libby R. Pansing, who provided
outstanding support both in the field and back in lab. I am also appreciative of the field
assistance provided by Logan Wealing, Soledad Diaz, and Aaron Wagner. I thank my
committee members Dr. Leo P. Bmederle and Dr. Michael Wunder for their advice and
assistance throughout this project. A special thank-you goes to our collaborators in the
Dr. Lynn Resler and Dr. George Malanson labs because without them this project never
would have been possible. I also appreciate the help provided by numerous folks and
various agencies, especially Kent Houston of the Shoshone National Forest, Custer
National Forest, the Blackfeet Tribal Nation, Colorado State Forest Service Nursery,
Glacier National Park, and the ecology and evolutional biology group members at CU-
Denver.
vi


TABLE OF CONTENTS
CHAPTER
I. WHITEBARK PINE BACKGROUND............................................1
Taxonomy and Distribution...........................................1
Whitebark Pine Seeds: Seed Dispersal and Food Source................1
The Alpine-Treeline Ecotone and Whitebark Pines Roles..............3
Threats to Whitebark Pine...........................................6
Global Climate Change and Treeline Impacts..........................9
Figures and Tables.................................................12
II. INTRODUCTION........................................................14
Background.........................................................14
Conceptual Framework...............................................17
Hypotheses for Testing.............................................18
Figures and Tables.................................................21
III. METHODS.............................................................22
Study Areas........................................................22
Relative Vigor Study...............................................23
Field Methods..................................................24
Data Analysis..................................................26
Planting and Sowing Study..........................................28
Field Methods..................................................28
Data Analysis..................................................30
Girdling Study.....................................................31
Field Methods..................................................31
Data Analysis..................................................32
vii


Figures and Tables.....................................................34
IV. RESULTS..................................................................40
Relative Vigor Study...................................................40
Transects...........................................................40
Shoot Lengths and Shoot Growth Rate Comparisons.....................41
Other Small Tree Measurements.......................................43
Results Summary.....................................................44
Planting and Sowing Study..............................................44
Seedling Survival...................................................44
Seed Germination and Summer Survival................................46
Results Summary.....................................................47
Girdling Study.........................................................48
Leeward Conifer Vigor...............................................48
Shoot Lengths.......................................................48
Shoot Mortality.....................................................49
Results Summary.....................................................50
Figures and Tables.....................................................51
V. SYNTHESIS AND DISCUSSION.................................................61
Study Conclusions......................................................61
Potential Implications for Whitebark Pine Decline at Treeline..........65
Figures and Tables.....................................................69
viii


REFERENCES.........................................................70
APPENDIX
I. Small Tree Measurements.........................................77
II. Small Tree Analyses............................................80
III. Planting Study Microsite Heights..............................82
ix


LIST OF TABLES
Table
1.1 US and Canadian conifers susceptible to white pine blister mst infection.13
111.1 Qualitative vigor categories........................................35
111.2 Sample sizes of small kmmmholz trees in the relative vigor study....35
IV.1 Species abundances of solitary conifers in transects.................52
IV.2 Kmmmholz shoot lengths...............................................53
IV.3 Kmmmholz tree shoot growth rates........................................54
IV.4 Upright upper subalpine conifer shoot lengths...........................55
IV.5 Upright shoots with minimum needle lengths subtracted................56
IV.6 Small shoot lengths vs. upright shoot lengths: proportions...........57
IV.7 Summer 2012 survival advantage and relative death risk of seed germinants on
Divide Mountain...........................................................59
AI.l Small Tree Measurement Summaries........................................77
AI.2 Divide Mountain Small Tree Measurements by Site.........................78
AII.l Change in kmmmholz tree stem diameters.................................80
AIL 2 Krummholz tree canopy areas............................................81
AIL 3 Kmmmholz tree heights...............................................81
AIII.l Planting and sowing study microsite heights........................82
x


LIST OF FIGURES
Figure
1.1 Distribution of Pinus albicaulis in North America.............................12
1.2 Image of active blister mst stem canker on an infected whitebark pine........13
11.1 Overall conceptual model.....................................................21
111.1 Research study areas........................................................34
111.2 Planted seedlings at the Line Creek RNA.....................................36
111.3 Germinated seeds on Divide Mountain.........................................37
111.4 Example of before and after girdling and defoliation treatment..............38
111.5 Leeward shoot vs. exposed shoot sampling areas..............................39
IV.1 Solitary kmmmholz tree density by species on Divide Mountain and Line Creek
RNA...............................................................................51
IV.2 One year post-planting seedling survival per microsite.......................58
IV.3 2012 Divide Mountain seed germination counts.................................59
IV. 4 Girdling Study leeward conifer shoot length trends over time...............60
V. l Potential consequences of blister mst to alpine treeline dynamics.........69
xi


LIST OF ABBREVIATIONS
1. ATE
2. SF
3 WP
4. ES
5. RNA
Alpine Treeline Ecotone the region between the subalpine forest
and alpine tundra where conifers are kmmmholz or dwarfed
Subalpine Fir {Abies lasiocarpa)
Whitebark Pine (Pinus albicaulis)
Engelmann Spruce (Picea engelmannii)
Research Natural Area; in reference to the Line Creek Natural
Area located on the Beartooth Plateau, Montana
Xll


CHAPTER I.
WHITEBARK PINE BACKGROUND
Taxonomy and Distribution
Whitebark pine (Pinus albicaulis) is one of several stone pines comprising
subgenus Strobus, section Strobus, subsection Cembrae (Price et al,1998). Since
monophyly of Subsection Cembrae is unsubstantiated, it has been proposed that
subsection Cembrae be merged with subsection Strobi into a new subsection Strobus
(Liston et al,1999; Gernandt et al,2005). This classification has yet to be officially
recognized. The stone pines of subsection Cembrae are characterized as having five
needles per fascicle and indehiscent cones with wingless seeds that are dispersed by
nutcrackers {Nucifraga spp.) (McCaughey and Schmidt, 2001).
Whitebark pine is distributed from the southern Sierra Nevada of California north
through the Cascade and coastal ranges into British Columbia; and from the Greater
Yellowstone region of Wyoming north through the Rocky Mountains of British
Columbia and Alberta Canada (Figure 1.1).Whitebark pine is limited to upper subalpine
and treeline forests in high elevation mountains from 37 to 55N (Arno and Hoff, 1990).
It is often a dominant treeline species, except at its most northern limits and in the
snowiest regions of the southern Canadian Rockies and coastal ranges (Arno and
Hammerly, 1984). Whitebark pine assumes a kmmmholz growth form at treeline in the
drier mountain ranges (Arno and Hammerly, 1984).
Whitebark Pine Seeds: Seed Dispersal and Food Source
Because whitebark pine has indehiscent cones with wingless seeds, it relies on a
co-evolved mutualism with Clark5s nutcracker {Nucifraga columbiana) for seed dispersal
1


(Tomback, 1982). It is possible that this method of seed dispersal evolved as a
consequence of both genetic drift in small populations and seed selection choice by
Clarks nutcracker (Tomback and Linhart1990). Every year from late summer to early
fall, these birds gather seeds from cones, carry them within their sublingual pouch, and
cache them throughout the subalpine and treeline terrain. In many regions nutcrackers
typically select for seed caching steep, south facing slopes that accumulate minimal
snowpack (Tomback, 1982). Distances from the cache to the original seed source can
vary from a few meters up to 29 km in distance and 307 meters in elevation (Lorenz and
Sullivan, 2009). The seeds that are not later consumed germinate, thereby regenerating
the species.
Resler (2004) observed whitebark growing at treeline and found that many cache
sites selected by Clark5 s nutcracker were sheltered and important for whitebark pine
seedling survival. These microsites, which are often terraces or boulders, may facilitate
the germination and growth of whitebark pine seedlings. Cache sites selected by Clark5 s
nutcracker and the natural hardiness of whitebark pine account for a large majority of the
spatial distribution and population genetic structure of the species (Tomback2001).
Seeds from whitebark pine cones are also an important food source for other
wildlife. Grizzly bears {Ursus arctos) and pine squirrels {Tamiasciurus spp.) rely on
these seeds (Mattson et al,1992; McKinney and Fiedler, 2010). In the Greater
Yellowstone Area, pine squirrels store cones in middens that are later raided by the bears.
These seeds are a large part of the grizzlys food source in the spring and summer of
good cone crop years (Matson and Reinhart, 1994). Other wildlife species that consume
these seeds include small mammals, such as chipmunks {Tamius spp.), and golden mantle
2


ground squirrels {Spermophilus lateralis), and some small birds, such as woodpeckers,
nuthatches, finches, and Steller5s jays (Cyanocitta stelleri) (Tomback, 1978; Hutchins
and Lanner1982; Tomback and Kendall2001).
The Alpine-Treeline Ecotone and Whitebark Pines Roles
The transition between the subalpine forest and alpine tundra is referred to as the
alpine-treeline ecotone (ATE). This high elevation zone is characterized by krummholz
conifers, dry, windswept slopes, and cold temperatures (Marr, 1977; Arno and
Hammerly, 1984; Finklin, 1986). Grace et al(2002) describe the 'climatological
bottleneck5 that results in kmmmholz growth. Trees in the alpine treeline seldom
produce cones with viable seeds, so this tree community is generated by seeds coming
from the subalpine zone. Therefore, trees in the alpine treeline ecotone must physically
adapt to survive the harsh climate (Malanson et al., 2007). Kmmmholz growth forms
result when wind-blasted snow and ice particles kill upright growth. Consequently, the
only branches able to survive are those that grow low to the ground. Kmmmholz trees
often have foliage surface temperatures 5-10 higher than ambient temperature. Taller
trees have surface temperatures 5"lower than ambient temperatures. Seedlings in the
alpine treeline ecotone may be sheltered and in favorable microclimates, but as they grow
taller, their growth rate is reduced and direction of growth altered by wind and
desiccation; thus, they become dwarfed or kmmmholz (Grace et al, 2002).
Survival in the alpine treeline ecotone is often increased by the formation of tree
islands. Tree islands are kmmmholz mats containing one or more individual trees
growing in close proximity (Marr, 1977). Solitary tree islands are comprised of one tree.
3


Multi-tree islands are comprised of two or more individual trees or many branches that
growing in layered form due to adventitious roots (Benedict, 1984).
Foundation species are highly abundant ecosystem components that exert much
influence on ecosystem function and stability (Ellison et al.2005). Keystone species
promote and support the biodiversity of their ecosystems (Soule et al,2003). Foundation
and keystone species in forest ecosystems have the ability to maintain biotic and abiotic
ecosystem components. If a keystone or foundational species declines, there could be a
resulting trophic cascade, with a loss of biodiversity or ecosystem function (Ellison et al.,
2005). Throughout its range, whitebark pine is both a foundation and keystone species
(Tomback and Achuff, 2010). Whitebark pine acts as a facilitator or 'nurse tree5 by
creating protective microsites for less hardy conifer species (Callaway, 1998). In the
alpine treeline ecotoneone of whitebark pines most important functional roles is in tree
island initiation. Whitebark pines growing in sheltered microsites can facilitate
community development by mitigating the harsh conditions on their leeward side. Resler
and Tomback (2008) found that whitebark pine was the windward tree island initiator for
nearly half the multi-tree tree islands among two study sites east of the Continental
Divide. They also found that whitebark pine was an important component of tree islands
in this region: 255 out of 266 tree islands sampled contained whitebark pine.
Research has demonstrated just how important tree islands can be for survival of
less hardy conifers. Hattenschwiler and Smith (1999) studied distributions of subalpine
fir and Engelmann spmce in the central Rocky Mountains to determine locations with
greatest survival. Although Engelmann spmce appears to germinate quickly and at lower
temperatures than subalpine fir, no seedlings of either species could survive on the
4


windward side of tree islands. In the alpine treeline ecotone, the most frequent location
for seedling establishment was on the leeward side of tree islands where snow
accumulation is maintained at a moderate depth of 0.5 -1.5 m, thus offering protection.
Germino et al(2002) found that the microsites with windward protection were associated
with a 20% higher survival rate of Engelmann spmce seedlings. Additionally, seedling
survival was 70% higher when microsite features, such as branches, were located directly
above the seedlings. They claim that close proximity to tree islands and overhead
structuressuch as branchesmay moderate solar and long wave radiationreduce
daytime temperature extremes, and maintain snowdrift accumulation. These factors
increase seedling survival by making environmental conditions more moderate.
Whitebark pines role in establishing tree islands is an important ecological
function, both for community development at treeline and for the provision of ecosystem
services to people (Resler and Tomback 2008). Tree islands provide important
ecosystem services. Tree islands are involved in watershed hydrology through the
maintenance of snowpackwhich regulates the rate of snowmelt run-off (Holtmeier and
Broil, 1992). Conifer roots also help stabilize soil erosion (Tomback et al.2001).If no
tree islands are present to perform these functions, erosion and summer drought may
result. Farmers and ranchers with land in the valley bottoms and on the plains,
downstream of these mountains, rely on regulated water from snowmelt to fill streams
and creek beds necessary for crops and livestock. Municipal water reservoirs are
sometimes kept at appropriate levels by snowmelt (Smith et al., 2009). Late summer
shortages could lead to rationing or the costly service of transporting water supply to the
region.
5


Threats to Whitebark Pine
Currently, there are several threats to whitebark pine throughout its range. Fire
suppression, mountain pine beetle outbreaks, and white pine blister mst are compounding
factors in whitebark pines decline. Global climate change may increase the magnitude
of some threats, making it challenging to predict how whitebark ecosystems will respond
(Tomback and Achuff2010).
Fire is a natural occurrence in forest ecosystems. A burned area generates
openings in the foresteffectively setting back the successional clockTomback et al.
2001).Clarks nutcracker is known to select post-bum sites for seed caches (Tomback
2001), which means that whitebark pine is an important species for forest regeneration
after a fire. New whitebark growth creates sheltered microsites for subalpine fir and
Engelmann spruce to growincreasing conifer biodiversity along with forest regeneration
(Tomback and Resler2007).
Whitebark pine is less shade tolerant than the species that it shelters (Amo 1986).
Fire exclusion practices in the 20th century have led to successional replacement of
whitebark pine by subalpine fir and Engelmann spruce (Arno 1986). This has changed
the structure of subalpine forests because conifer biodiversity is lost and the landscape
becomes homogenous. Many stands at a landscape level are now solely comprised of
late serai stage subalpine fir and Engelmann spruce (Keane2001).
Mountain pine beetles {Dendroctonusponderosae) are native to western North
America. These beetles episodically attack large, mature pines with thick bark, resulting
in major outbreaks (Cole and Amman, 1969). They naturally occur in lodgepole pine
forests, but during outbreaks the insects spread to whitebark pine communities (Arno,
6


1986). This can result in wide scale mortality. In the 20th century, mountain pine beetle
outbreaks killed many mature whitebark pine trees in Idaho and Montana (Bartos and
Gibson1990; Jenkins et al.2001).
Because fire suppression results in higher density and greater age of late
successional forests, this practice may increase the scale and abundance of mountain pine
beetle outbreaks (McGregor and Cole, 1985). Climatic warming facilitates pine beetle
population growth and may reduce whitebark pine defenses (Raffa et al., 2008).
Currently, mountain pine beetles are again in outbreak mode throughout the West, but at
a geographic scale considered unprecedented (Logan et al., 2010). This outbreak is
driven by milder winter temperatures (Logan and Powell 2001;Logan et al.2010).
Whitebark pine stressed by competition from fire suppression may be even more
vulnerable to pine beetle attacks generated by warming trends (Logan et al., 2010; Raffa
et al., 2008).
A third cause of decline in whitebark pine is white pine blister mst, a disease
caused by the exotic pathogen Cronartium ribicola. This fungal pathogen, which infects
five-needle white pines of subgenus Strobus (McDonald and Hoff, 2001), was
inadvertently introduced to western North America in the early 19005s through
importation of infected nursery seedlings from western Europe (Spaulding, 1909, 1911,
1922 as cited in McDonald and Hoff, 2001). Cronartium ribicola has evolved with
Eurasian pine species, which have resistance to this disease. Since its introduction to
North America, C. ribicola has exploited a range of host pine species with low natural
resistance (McDonald and Hoff, 2001) (Table 1.1). Although many North American five-
needle pines are susceptible to infection by C. ribicola, they vary in susceptibility and
7


extent. Whitebark pine populations are currently being infected by C. ribicola nearly
range wide, with high infection levels in some areas; the resulting mortality is impacting
ecosystems throughout whitebark pines range (Tomback and Achuff2010).
Cronartium noicola relies on both five- needled white pines and alternate hosts to
complete its life cycle (McDonald and Hoff, 2001). Ribes spp., the gooseberries and
currants, have long been recognized as alternate hosts, but recent research has discovered
that herbaceous plants in the genera Pedicularis and Castilleja may also act as hosts
(McDonald et al., 2006).
McDonald and Hoff (2001) describe the specific mechanism of white pine blister
mst transmission. Five-needle white pines are infected by C. ribicola when wind-blown
basidiospores from alternate hosts enter the stomata of pine needles. Rust mycelia grow
from the needle into the living wood of the pine tree and eventually produce a fmiting
canker, which leads to swellings on branches or stems of the tree (Figure 11.2). The
canker spomlates, producing sacs of aeciospores, and these sacs ultimately burst through
the surface of seemingly healthy bark to release spores into the environment. Some
spores inevitably reach an alternate host and complete the cycle. Cankers eventually
girdle the branch or stem of the pine, cutting off the supply of water and nutrients. Since
seed cones are produced at branch tips, the accumulating dead branches reduce seed cone
production long before the tree itself dies.
Blister mst mortality is especially detrimental to whitebark pine, which requires
up to 50 years to reach reproductive maturity (McCaughey and Schmidt, 1990). Loss of
mature trees means a loss of cone production that potentially takes decades to replace.
However, seedlings, saplings, and smaller kmmmholz whitebark pine are also affected by
8


C. ribicola, and die more rapidly from infection than their larger counterparts (Tomback
et al.1995). This reduces the number of young trees available to regenerate the species.
Kmmmholz whitebark pines are also affected by white pine blister mst. It was
once thought that this pathogen could not survive the extreme winter temperatures of the
alpine treeline ecotone (Campbell and Antos, 2000), but high numbers of infected
individuals have recently been discovered among kmmmholz whitebark pine (Resler and
Tomback, 2008), suggesting that C. ribicola can reproduce and survive under the most
extreme conditions.
White pine blister mst may affect the keystone and foundational roles that
whitebark pine plays within the alpine treeline ecotone. Resler and Tomback (2008)
discovered that 33.7% of the whitebark pine in their sampled tree islands showed
evidence of white pine blister mst infection. This has serious implications because global
climate change is predicted to alter treeline dynamics (Tomback and Resler, 2007). If
whitebark pines in the alpine treeline ecotone are succumbing to blister mst at significant
levels, it may affect the way treeline is able to respond to climate warming and
potentially rising treeline elevations (Tomback and Resler, 2007; Resler and Tomback,
2008).
Global Climate Change and Treeline Impacts
Treeline forests are indicators of global climate change. These so called
'bellwether5 ecosystems are often the first to show symptoms of stress (Smith et al.,
2009). Treeline is known to be dependent on several factors, including temperature, wind
speeds, nitrogen deposition, and concentration of carbon dioxide (Grace et al, 2002).
Treelines have responded to temperature fluctuations since the last glacial maximum
9


(Lloyd and Graumlich, 1997). During the early Holocene, temperatures were 1.4 C
greater than the present, and treelines were on average 200 m higher than they are today
(Grace et al., 2002). Average temperatures are conservatively predicted to increase by
more than 2.5 C over the next century (Easterling, 2005). Warmer temperatures are
expected to cause an upward shift in treeline (Millar et al., 2004), with an estimated
elevation gain of 140 700 m (Grace et al.2002).
Climate models have been generated to predict how whitebark pine will respond
to warming trends (Hamann and Wang, 2006; McKenney et al., 2007; Warwell et al.,
2007). These models are generally in agreement that whitebark pine will see a shift from
its current range. These models do not account for fine scale habitat features or
ecological processes, such as topography, soil nutrients, seed dispersal and germination,
or disturbance regimes (Loehle, 1996). However, they do provide a coarse estimate of
changes in range and habitat area. Some models show that while whitebark pine will lose
current distribution in the U.S., it will gain new habitat at higher latitudes and elevations.
Warwell et al.(2007) predict a 97% loss of suitable whitebark habitat in the U.S. by
2090. Hamann and Wang (2006) found that 73% of whitebark pines habitat will be lost
by 2085, but it should gain 76% of the original area at northern latitudes. McKenney et
al. (2007) predicted that by the end of the centurywhitebark pines current range will be
reduced by 42%. However, whitebark pine will gain an expected 7.8% new habitat by
moving north approximately 6.4. With global climate change shifting treeline to higher
latitudes and elevationsthere is concern about whitebark pines ability to respond and
maintain its ecological roles due to stresses brought on by blister mst and other threats
10


such as mountain pine beetle. If whitebark pine mortality is widespread, the ability of
treeline as a whole to move upwards could be compromised.
Clark5 s nutcracker plays an important role in upward movement of whitebark pine
by caching seeds in the alpine tundra and treeline ecotone (Tomback, 1998; Tomback,
2001). Frequently, seeds are cached next to microsites such as rocks or ground
topography which potentially act as 'nurse objects5, facilitating germination and
sheltering developing whitebark seedlings (Resler, 2004). Because whitebark pine then
in turn creates favorable microsites for less hardy conifers facilitating tree island
development, warming trends should result in kmmmholz tree islands shifting upwards in
elevation (Resler et al,2005). With blister mst killing whitebark pine at treeline, the
number of favorable whitebark pine microsites available to subalpine fir or Engelmann
spruce is reduced. This may affect the ability of treeline to move upwards in the manner
predicted (Tomback and Resler, 2007; Resler and Tomback, 2008).
The more we can learn about whitebark pines ecological functionsthe better we
can predict how ecosystems will respond to whitebark pine mortality. Presently, very
little is known about the mechanisms behind facilitation roles whitebark pine plays in the
alpine treeline ecotone, or how whitebark pine ecosystems will respond to increased
blister mst mortality and global warming trends. Our research will contribute to a better
understanding the role of whitebark pine in two study areas in Montana, facilitates tree
island initiation and maintenance within the alpine treeline ecotone.
11


Figures and Tables
Figure LI Distribution of Pinus albicaulis in North America
(Tomback and Achuff2010)
12


Figure 1.2 Image of active blister rust stem canker on an infected whitebark pine.
Light colored sections on stems are sacs containing aeciospores (Photo by Sarah
Blakeslee).
Table 1.1 US and Canadian conifers susceptible to white pine blister rust infection
(McDonald and Hoff2001)
North American Tree Hosts
Whitebark Pine (Pinus albicaulis)
Foxtail Pine (Pinus balfouriana)
Rocky Mountain Bristlecone Pine
(Pinus aristata)
Great Basin Bristlecone Pine
Southwestern white pine {Pinus strobiformis)
Limber Pine {Pinus flexilis)
Eastern White Pine (Pinus strobus)
Western White Pine (Pinus monticola)
Sugar Pine (Pinus lambertiana)
(Pinus longaeva)
13


CHAPTER II.
INTRODUCTION
Background
Certain plant species may act as keystone and foundational ecosystem
components by facilitating stability and biodiversity (Ellison et al.2005). Research has
indicated the importance of facilitative plant interactions for survival and regeneration in
stressful environments (Lortie et al2004; Brooker et al2008). This is particularly true
for high elevation sites where abiotic stress is high (Callaway et al.2002). Calloway et
al.(2002) examined 115 plant species in 11 mountain sites across the globe to determine
whether elevation, and thus environmental stress, changed plant community interactions.
They generally observed competitive interactions at lower elevations where
environmental conditions were moderate. At higher elevations, species interactions
predominantly switched to facilitation, whereby one competitor provided shelter for
another. Harsh environmental conditions may be moderated when facilitative species
provide a protective microclimate for germination and establishment, e.g., protection
from solar radiation or shelter from wind (Germino et. al.2002; Baumeister and
Callaway, 2006).
As more examples of plant facilitation in high elevation communities are
discovered, it is increasingly apparent that facilitation is important to community
development in these extreme environments. For example, Cavieres et al.(2005, 2007)
observed plant interactions at the upper limit of vegetation in the Chilean Andes. Their
studies indicate that a cushion plant (Azorella monantha) moderates substrate and air
temperatures and enhances soil moisture and nutrients for both the native Andean
14


cauliflower (Nastanthus agglomerates) and the invasive field chickweed (Cerastium
arvense). Batllori et al.(2009) found that survival of Pinus uncinata seedlings planted in
the alpine treeline ecotone was increased when seedlings were located on the leeward
side of kmmmholz conifers, likely due to retention of sheltering snowpack in this
location during winter months.
The process of seedling establishment is important for long-lived plants such as
conifers. Years with successful seed germination are more frequent than years with both
high seed germination and high seedling survival (Cui and Smith, 1991). Seedling
establishment is particularly difficult in the alpine treeline ecotone because high winds,
variable temperatures, poorly developed soils, and intense solar radiation (Marr, 1977;
Arno and Hammerly, 1984; Finklin, 1986; Maher et al., 2005) make establishment a
challenge.
The likelihood of seedling survival is improved in the alpine treeline ecotone
when harsh climatic conditions are mitigated by rocks, topographic niches, and other
objects acting as protective microsites or nurse objectsproviding windward shelter
(Germino, 2002; Resler, 2004; Batllori et al., 2009). Survival is further facilitated when a
solitary conifer establishes and other conifers grow in its lee, resulting in two or more
conifers growing together in close proximity as a multi-tree tree island. In the alpine
treeline ecotone, tree islands facilitate the survival of conifers species such as Engelmann
spruce (Picea engelmannii) and subalpine fir {Abies lasiocarpa) (Resler and Tomback,
2008). Both species are less likely to be found on the windward rather than leeward side
of tree islands (Hattenschwiler and Smith1999)and Engelmann spruce seedlings have
15


been found to be associated with higher survival rates when windward or overhead
shelter, such as branches, is present (Germino et al., 2002).
In the cool, dry, and windy northerly eastern slope faces of the alpine treeline
ecotone on the eastern Rocky Mountain front, whitebark pine (Pinus albicaulis) is a
dominant ecosystem component (Smith et al.2011;Resler and Tomback2008). This
species is dispersed by Clark5s nutcracker (Nucifraga columbiana) (Tomback, 1982), and
is tolerant of drought and high levels of solar radiation (Amo and Hammerly, 1984;
Maher et al.2005; Tomback et al.2001).Because whitebark pine often grows as a
solitary conifer (Maher et al,2005; Resler and Tomback, 2008), it may act as a 'nurse
tree5 by facilitating the survival of less hardy subalpine fir and Engelmann spmce on
harsh sites in the subalpine zone and in the alpine-treeline ecotone, where it facilitates
development of multi-tree tree islands (Callaway, 1998; Resler and Tomback, 2008).
Resler and Tomback (2008) found that within two study areas east of the Continental
Divide, 95.9% of multi-tree tree islands sampled included whitebark pine. Of these tree
islands48.5% had whitebark pine as the windward initiatorindicating the importance
of whitebark pine in facilitating the establishment of leeward conifers in this region.
Whitebark pine is presently designated a candidate for endangered species listing
by the U.S. Fish and Wildlife Service (USFWS, 2011). Fire suppression leading to
successional replacement by shade tolerant conifers, mountain pine beetle {Dendroctonus
ponderosae) outbreaks, and the disease white pine blister mst, caused by the invasive
fungal pathogen Cronartium ribicola, are the major factors in the decline of whitebark
pine. Within the alpine treeline ecotone, the most immediate threat to whitebark pine
populations is damage and mortality resulting from white pine blister mst. Infected small
16


diameter trees can progress from showing no outward disease symptoms to death within a
few years (Tomback et al.1995). This rapid mortality may reduce the chances for
establishment of new tree islands, and also the health of existing tree islands by exposing
formerly leeward conifers to the wind.
Global climate change may increase the frequency and severity of fire regimes,
accelerate the rate and spread of pine beetle outbreaks, and likely alter the distribution
and infection rates of blister mst, making it challenging to predict how whitebark
ecosystems will respond (Tomback and Achuff, 2010). Warmer temperatures are also
expected to cause an upward shift in treeline (Millar et al., 2004), with an estimated
elevation gain of 140 700 m (Grace et al.2002). As whitebark pine numbers decline
this may impact the frequency of tree island establishment in the upper alpine treeline
ecotone, possibly altering the response of treeline as a whole to warming trends
(Tomback and Resler 2007).
Conceptual Framework
Although research has indicated that whitebark pine is an important component of
tree islands, very little is known about the specific mechanisms of facilitation leading to
tree island formation or about whitebark5s role in established tree islands. The overall
objectives for this study are to determine empirically and experimentally the attributes
and ecological interactions that enable whitebark pine to facilitate tree island
development, and to address how the mortality of whitebark pine from blister mst may
impact these ecosystem functions.
17


Hypotheses for Testing
This is an NSF-supported project with objectives already articulated. For my
master5s project I clarified goals and hypotheses, designed experiments, developed field
protocols, and had oversight responsibility for a series of experiments and empirical
studies. I had the support of Dr. Diana F. Tomback and other students in the lab and
fieldso often refer to the work as we I am testing three separate hypotheses that
together support the overall research objective in a logical sequence (Figure II.1).My
hypotheses are as follows:1)Whitebark pine is hardier than other alpine treeline ecotone
conifer species, as demonstrated by more vigorous growth and higher survival at treeline;
2) whitebark pine provides a more favorable microsite for tree island recruitment than
other common alpine treeline ecotone microsites; and 3) death of windward whitebark
pine in established tree islands leads to loss of vigor in leeward conifers.
The first hypothesis addresses whether there is differential growth vigor, if any, of
whitebark pine in comparison to other treeline conifers. Because tree islands usually
migrate leewardthe most windward tree is often the oldest (Holtmeier and Broil
1992). The dominant presence of whitebark pine in this position suggests that this
species is hardy and serves an important role in recruiting tree islands through subsequent
facilitation by mitigating conditions for a leeward conifer. In order to establish a tree
island, whitebark pine seedlings must first become established in the harsh, and exposed
areas within the alpine treeline ecotone and then survive these conditions. This first
hypothesis is addressed through an empirical studythe Relative Vigor Study that
identifies and compares differences in survival and vigor among small kmmmholz
whitebark pine and the other two dominant treeline conifer species, subalpine fir and
18


Engelmann spruce. In Figure II-1,the importance of the initial establishment and
survival of whitebark pine to subsequent facilitation is illustrated in a visual model.
The second hypothesis states that because whitebark pine often facilitates tree
island development, it may provide more favorable growing conditions for leeward
conifers than other common treeline microsites. An alternative to this hypothesis is that
whitebark pine trees are simply more numerous at treeline, and this means that there are
more opportunities for tree islands to form in whitebark pine microsites. In the Planting
and Sowing Studywe test whether whitebark microsites are associated with higher
conifer germination and/or seedling survival rates than other common treeline microsites
in order to determine whether conditions are in fact more favorable. This would imply
that whitebark pine may offer facilitation or a higher quality of facilitation than other
microsites. The other alpine treeline ecotone microsites investigated include rocks,
another conifer Engelmann spmce for consistency and exposed sites with no apparent
shelter. In this study, we planted conifer seeds and seedlings leeward of four microsite
types and compared seed germination and seedling survival rates. This mechanism for
tree island development is represented in the #2 position of Figure II-1.
The last hypothesis directly tests the facilitation function of whitebark pine in
established tree islands. Since whitebark pine is often the intiating or most windward
conifer within a muilti-tree island, it may provide important leeward shelter to other
conifer species, mitigating the harsh wind and particle-blast of treeline environments
(Habeck, 1969; Resler, 2004). Blister rust is currently infecting and killing many
whitebark pine in some regions. We hypothesize that the loss of these windward
whitebark pines will result in exposure and thus damage to leeward conifers. As
19


indicated in the #3 position of Figure II.1the Girdling Study simulates the effects of
blister mst on windward whitebark pine, and we monitored the growth and vigor of the
non-whitebark conifer immediately leeward.
20


Figures and Tables
Figure 11.1 Overall conceptual model
The role of whitebark pine in tree island formation can be explained by 1,the
establishment of a solitary conifer in an exposed area without shelter from other tree
islands. To accomplish this, the species must be hardy and vigorous to withstand the
harsh treeline climate. We test this hypothesis in the Relative Vigor Study. As the
conifer establishes and grows, it generates a sheltering leeward microsite (indicated by
star), 2, where other conifers can germinate, eventually leading to the formation of a tree
island. We examine this hypothesized process by testing whether whitebark pine
microsites are associated with the highest conifer germination and survival rates in the
Planting and Sowing Study. Blister mst is currently killing whitebark pine at treeline,
potentially exposing leeward conifers to harsh wind and ice particles. The impact of this
windward shelter loss, 3, on established tree islands is unknown. We simulate blister mst
on windward whitebark pine (x indicates blister mst simulation) in the Girdling Study
and monitor impacts on the newly exposed leeward conifer.
21


CHAPTER III.
METHODS
Study Areas
This research was conducted over three field seasons within a bioclimatically
induced kmmmholz treeline at two separate study areas (Figure 111.1).These study areas
were selected because whitebark pine is a major ecosystem component and also because
of the accessibility of treeline. The northern study area includes Divide and Whitecalf
Mountains, Montana. Divide Mountain is located on both Blackfeet Tribal Land, as well
as on the east slope (Rocky Mountain eastern front) of Glacier National Park at
approximately 48 39' 25" N and 113 23' 45" W. Treeline occurs at approximately 2200
m elevation. Whitecalf Mountain is located within the east slope of Glacier National
Park at 48 38' 20" N and 113 24' 08" W. Treeline occurs at approximately 2100 m
elevation. Divide and Whitecalf Mountains are characterized by steep slopes (x = 25.7)
and poorly developed soils with limestone bedrock. Mountain avens {Dryas octopetala)
and bearberry {Arctostaphylos uva-ursi) are the dominant herbaceous understory
vegetation in this study area. Willows (Salix spp.) and junipers (Juniperus spp.) are also
distributed in patches. Subalpine firEngelmann spruceand whitebark pine are the
dominant trees on Divide Mountain, but there is a noticeable absence of Engelmann
spruce on Whitecalf Mountain. Found in small numbers are limber pine (Pinus flexilis),
lodgepole pine (Pinus contorta), and Douglas-fir {Pseudotsuga menziesii).
The southern study area is located 530 kilometers directly southeast of the
northern study area on the Beartooth Plateau5 s Line Creek Research Natural Area (RNA),
22


MT, in Custer National Forest at 45 01'47.45" N and 109 24' 09.22" W. Subalpine fir
is nearly absent from the ecosystem and is only commonly found in the shelter of willow
patches or within dense tree islands. Engelmann spmce and whitebark pine are the two
most common species. Lodgepole and limber pine are found in very small numbers.
Kmmmholz rapidly grades into erect trees on the northeast-facing slope. Like the Divide
Mountain and White Calf study areas, the Line Creek RNA is characterized by many
solitary kmmmholz whitebark pine. The terrain is largely open with gentle slopes (-20).
Sedges (Carex spp.), mountain meadow cinqueroil (Potentilla diversifolia.), American
bistort (Polygonum bistortoides), and silvery lupine {Lupinus argenteus) are common
among the herbaceous groundcover.
Relative Vigor Study
Two studies were conducted to examine relative vigor in different ways. One was
an observational study addressing our first hypothesis by comparing shoot lengths, shoot
growth rates, and vigor of the three most common treeline species in our study areas:
whitebark pine, subalpine fir, and Engelmann spmce. For the first study, we used
measurements representing readily obtainable characteristics of growth and vigor. We
originally intended to use seedlings for each species, but when selecting trees for this
study, we found very few seedlings at either study area. Instead, we selected small
kmmmholz trees that were already established. These trees ranged roughly from 1 to 40
years of age, based on stem constriction counts (Appendix I, Tables 1 and 2). These
small trees were unlikely to die during the short-term three year observation period. Our
measurements act as a surrogate for survival and are also important clues to any species
differences in ability to utilize resources from the environment. Shoot lengths are
23


particularly relevant to vigor and survival. Ability to increase photosynthetic canopy
rapidly is important at treeline, where growing seasons many only last a few months in
the summer. The second study was a random sampling effort to determine the relative
numbers of solitary kmmmholz trees of each species in our study area.
Field Methods. These studies were conducted both on Divide Mountain and at
the Line Creek RNA. In July 2010,100 small solitary krummholz conifers (<30 cm
high) of the three major conifer species were haphazardly selected based on exposed
growing conditionsi.e.unsheltered by a tree island or other large microsites (Table
111.2). We marked conifers with a tagged leeward nailspike and monitored them from
July 2010 to September 2012. Stem diameter at ground level, length from 3-5 shoots
(defined as the total length of the new branch elongation plus extending needles), canopy
area (calculated using longest dimension and the dimension immediately perpendicular to
longest dimension), and woody tissue height were measured annually in July. The only
exceptions were shoot length measurements. In 2011 and 2012 we measured five shoots,
if available, from each tree, first in July near the beginning of the growing season and
then again in September after shoots were fully extended. The difference between the
July and September mean shoot length per tree was used to calculate the shoot growth
rates with the following formula: (September Mean Shoot July Mean Shoot)/No. Days
between Measurements for each conifer (mm/day). Canopy areas were calculated using
the formula for the area of an ellipse x a x b).
All tree geographic locations were marked with a Trimble Geo XT handheld GPS
unit (GeoExplorer 2008 series). Measuring tapes were used to measure height and
canopy area to the nearest half centimeter. Mitutoyo (500-195-20) digital calipers with a
24


precision of a hundredth of a millimeter were used to measure shoot lengths. Care was
taken to ensure consistency of remeasurement from year to year. In 2011we marked
five branches on each conifer with zip ties so that the same terminal branch shoots were
remeasured. Each conifer stem was marked with tree specific paint to ensure consistency
in caliper placement for stem diameter measurements.
The shoot lengths of subalpine forest conifers, which are taller trees with large
diameters and upright growth forms, were measured for comparison with kmmmholz
conifers to see how shoots growing in less harsh conditions might differ. In both study
areas, larger stature trees occurred in sheltered sites and at the lower limit of the alpine
treeline ecotone. In 2011,five shoots each of 10 haphazardly selected conifers of each
species at each study area were measured in September. In 2012, we increased the
sample size to 20 conifers of each species at each study area. The measurement
procedure was identical to that of the small kmmmholz trees, except tree branches were
not marked, so the same trees were not necessarily revisited from year to year.
In order to determine the relative abundance and density of solitary kmmmholz
whitebark pine, subalpine fir, and Engelmann spmce, growing at treeline, transects 50 m
long transects with 10 m wide belts (500 m2) were established using a subset of randomly
generated GIS points within each study areas. We sampled twenty transects at each study
area. On each transectsolitary krummholz conifers growing in largely unsheltered
conditions were measured for species, height, canopy area, microsite, and qualitative
vigor. Qualitative vigor was determined using a four category ranking scale of poor to
excellent based on characteristics of windward needle death, health of new annual shoots,
and needle color (Table 111.1).
25


Data Analysis. All data analyses were completed using R (GUI statistical
software program version 2.11.1). We compared kmmmholz shoot lengths among
species using a Kmskal-Wallis rank sum test (because data were not normally distributed
and sample sizes were unequal). Because of an unequal number of shoots per tree
resulting from small tree size and shoot death on some marked branches, one shoot per
tree was selected for this analysis using randomly generated numbers. Wilcox-signed
rank post-hoc tests with Bonferroni corrections were used for paired comparisons of
shoot lengths between species (a = 0.05).
Our shoot measurements do not separately account for branch extension length
and needle length, but rather the sum of both. We were interested in seeing if species
differences remained when needle length was subtracted from our measured shoot
lengths. This would compensate for any length attributed to the needle extending beyond
the branch extension point. We obtained minimum needle lengths for each of the three
species from Flora of North America (eFloras, 2008) and subtracted them from the mean
shoot lengths of the upright upper subalpine conifersas follows: Engelmann spruce -16
mm, subalpine fir -18 mm, and whitebark pine 30 mm. The reduced shoot lengths
were then compared by species in a series of Wilcox-signed rank tests with Bonferroni
corrections (a = 0.05) by year and study area. Nonparametric statistics were used
because data were not normallydistributed. Kmmmholz conifer shoots were not assessed
in this manner because their needle lengths were generally shorter than the minimum
needle lengths suggested by Flora of North America. This analysis would have resulted
in negative shoot lengths for kmmmholz individuals.
26


Other measurements analyzed with Kruskal-Wallis rank sum tests and Wilcoxon-
signed Rank post hoc tests with Bonferroni corrections (a = 0.05) were kmmmholz shoot
growth rates, kmmmholz canopy area, kmmmholz height, kmmmholz stem diameter, and
qualitative vigor of trees sampled in the transects. Non-parametric statistics were used
because of unequal sample sizes and data that were not normally distributed.
In order to use the mean shoot for each individual upright tree, we verified that
the within tree variation in shoot lengths was smaller than among species variation as
follows: The variation of the five measured upright subalpine conifer shoots was
compared within and among species with a Nested ANOVA. Mean shoot length was
calculated for each tree and compared with a One-way ANOVA and Tukey5s post hoc
test to determine species differences. Parametric statistics were used because data were
normally distributed and sample sizes were equal. Significance level was set at a = 0.05.
We used a Chi-Square test (a = 0.05) to analyze differences in microsites
responsible for establishment for the conifers sampled within the transects. In this
comparison there were two categories for microsite: the proportions of conifers found in
relatively unsheltered, unknown, or small microtopographic ground depression microsites
were compared to the proportions of conifers found established leeward of more
substantial shelter, such as rocks or vegetation. These proportions were also compared
between species at each study area.
We used a Binomial Distribution test (a = 0.05) to determine the proportional
occurrence of whitebark pine in relation to the number of all solitary conifers sampled for
the transects at each study area. This test was performed for each of the 20 transects at
each study area. The theoretical probability was considered to be an equal proportion of
27


each species (0.33). We determined how many of the 20 transects at each study area had
solitary whitebark pine in a significantly greater proportion than expected. The mean
density of each species was also calculated as the number of conifers per square meter for
each transect. These densities were compared visually in a figure.
Planting and Sowing Study
This study addresses hypothesis 2 by comparing the survival rates of planted seedlings
and germination and survival rates of sown seeds leeward of four common treeline
microsites.
Field Methods. Study areas were Divide Mountain and the Line Creek RNA.
For the Divide Mountain seed sowing and seedling planting study, we collected
Engelmann spruce and subalpine fir cones at Divide Mountain in September2010. The
Colorado State Forest Service Nursery found that the Engelmann spmce cones contained
non-viable seeds, so only subalpine fir seeds were available for the study. The quantity
of subalpine fir seeds was adequate only to produce enough seedlings for the seedling
planting component of the study. Engelmann spmce seeds from the same seed transfer
zone (McDonald Pass, Helena National Forest, 6300 m elevation) were provided by the
USDA Forest Service Nursery in Coeur d'Alene, ID.
On the entire Beartooth Plateau there was no cone crop for either spmce or fir in
2010. Engelmann spruce seeds collected by the Dubois Ranger District of the Shoshone
National Forest were provided by the USDA Forest Service Nursery in Bessey, South
Dakota. These seeds were collected at an elevation of 2712 m, and were the highest
elevation Engelmann spmce seeds available in the same seed transfer zone as the
28


Beartooth RNA. These seeds were used for both the direct seed sowing experiment and
the seedling planting experiments.
Seeds for the direct sowing component of the study were chilled for 4 months at
approximately 35 before planting. Seedlings were grown by the Colorado State Forest
Service Nursery in Fort Collins, CO.
Planting and sowing took place in July 2011.In each study area, we located 20
replicates each of the four microsites kmmmholz whitebark pine, kmmmholz spmce,
rock, and exposed site for the seedling planting study, and at an additional 20 replicates
of each microsite for the seed sowing study. Microsite geographic locations were marked
with a Trimble Geo XT handheld GPS unit (GeoExplorer 2008 series). Microsites
ranged in height from 4 66 cm. In general, conifer microsites were taller than rocks.
For each study, spmce and whitebark microsites were selected based on similar heights
(Appendix III). Once suitable microsites were marked with a numbered nailspike, either
5 seeds or 2 seedlings were placed immediately leeward of the microsite object or in the
middle of the exposed microsite. Leeward direction was determined by observing
dominant flagging of kmmmholz conifers. Seedlings were labeled with colored zip ties
placed at the base of the stems for 2012 identification and planted in 25 cm holes dug to
fit the container substrate (Figure 111.2). Seeds were planted 0.5 cm deep. Seedling sites
received 1liter of water at the time of planting, and seed sites received !/2 liter of water at
the time of planting. Germination and survival were assessed in July 2012. At this time,
terminal shoot lengths were measured and qualitative vigor on a poor to excellent four
level categorical scale assessed for surviving seedlings. Germinated seeds were located
29


and counted by microsite type (Figure 111.3). Germination sites were revisited September
2012 to document germinant survival over the summer months.
Data Analysis. The results of this study were analyzed separately by planting
type, microsite type, and study area. Data that were not normally distributed were
analyzed with non-parametric statistics. Significance levels were set at a = 0.05. The
resulting analyses included one-year seedling survival rates and seedling vigor
assessments, and seed germination and new seedling summer survival for both Divide
Mountain and Line Creek RNA.
One year seedling survival was examined with a Pearsons Chi-squared test of
independence to assess whether survival differed among microsite types. Qualitative
vigor was also analyzed with a Pearsons Chi-squared test to determine differences
among microsite types. Terminal shoot lengths were compared among microsite types
with a Kruskal-Wallis rank sum test.
July, 2012, seed germination among microsite types was compared with a
Fishers Exact probability test (a = 0.05). September2012 summer survival numbers of
these initial germinants were also analyzed with a Fishers Exact probability test to
examine survival associated with the different microsite types. If any Fishers exact test
results were statistically significant, we compared the observed survival values to Chi-
square expected survival values to determine which microsite type(s) contributed the
most towards statistical significance. Lastly, the relative risk of survival or death by
microsite type was calculated using an odds ration calculation comparing actual
germinant summer survival numbers to expected values. A ratio near 1.0 was interpreted
as relative risk of survival or death near expected.
30


Girdling Study
This study addresses the third hypothesis by simulating whitebark pine mortality
caused by blister rust infection.
Field Methods. The girdling study was conducted on both Divide and Whitecalf
Mountains. In July 2010, we selected tree islands with whitebark pine as the windward
species, and placed them into either control or experimental groups. In order for a site to
be classified as experimental, the windward whitebark tree had to be infected with blister
rust. This was a condition of the research permits issued by both Glacier National Park
and the Blackfeet Indian Reservation. There were a total of 44 sites in this study, with an
equal number of control and experimental sites. All site locations were marked with a
Trimble Geo XT handheld GPS unit (GeoExplorer 2008 series).
At each site, the conifer species immediately leeward of the whitebark pine was in
most cases subalpine fir (n = 40) with the remaining sites Engelmann spmce. We
collected baseline measurements of height, vigor, shoot lengths, and canopy on both the
windward whitebark pine and the immediately leeward conifer. Heights were measured
to the nearest half centimeter with a metric tape measure. Shoot lengths were measured
using Mitutoyo (500-195-20) digital calipers with a hundredth of a millimeter precision.
After baseline measurements were collected, the whitebark pine at all experimental sites
was defoliated and girdled (Figure 111.4). This was accomplished by manually removing
all foliage from the tree and sawing deep grooves completely around the trunk to ensure
no future growth. This simulated the effects of blister mst infection, which reduces small
trees to branch and stem skeletons in a short time period (Tomback et al., 1995). In sites
that had extensive layering and tree islands larger than the windward whitebark pine, only
31


the discrete section which was sheltered by the whitebark pine was used for
measurements.
In 2011 and 2012, we monitored the effects of girdling on the conifer
immediately leeward of the whitebark pine. Subsequent measurements were taken to
determine whether exposure on the experimental sites impacted the leeward conifer
measurements of length of new shoots, shoot mortality, and qualitative tree vigor (Table
III.1),differently than control sites. In July 2011, zip-ties were placed on the branches
immediately leeward of the whitebark in both the control and experimental sites for
repeated measurements. At this time, we also marked shoots from fully wind exposed
subalpine fir or Engelmann spruce conifers in the same tree island that were not
associated with our experimental or control conifer (Figure 111.5). These trees represent a
natural measure of exposure to windward conditions in the same tree island and are of the
same leeward species we are investigating.
Data Analysis. The difference in 2010 and 2012 vigor of the leeward conifer was
determined and categorically ranked as follows: loss of vigor, no change in vigor, or
increase in vigor. Categories were compared between control and experimental site
leeward conifers with a Fishers Exact Probability Test. If this test was statistically
significant, the categories contributing most to significance were determined using
observed values vs. Chi-square expected values.
The changes in leeward conifer shoot lengths over the 2010 2012 time period
were compared in a Before-After-Control-Impact (BACI) analysis. One measured shoot
length per leeward conifer per year was randomly selected for comparison between
control and experimental site groups.
32


We also compared shoot length and mortality by treatment type at a tree island
level. We paired the natural, wind exposed shoots to the shoots experimentally exposed
by girdling or sheltered by a control whitebark from the same tree island. We determined
differences between these pairings for 2011 and 2012 shoot lengths and also the
proportion of dead shoots over the 2011-2012 timeframe. One shoot was randomly
selected from the five sampled per tree. The shoot length differences were compared in a
BACI analysis, and shoot mortality differences were examined with a Wilcox-Signed
rank test with a Bonferroni correction to determine whether experimental leeward
conifers had shorter shoot lengths and higher shoot mortality than control conifers.
33


Figures and Tables
Figure 111.1 Research study areas
Divide and Whitecalf Mountains, MT, in east Glacier NP and Blackfeet Reservation
(4839 25 N113 23 45 W)and Line Creek Research Natural AreaCuster National
ForestMT (1130 01 47 N1090 24 09 W). Basemap from Montana Government
Natural Resources: http://nris.mt.gov/gis/gisdatalib/mtmaps.aspx.
34


Table 111.1 Qualitative vigor categories
This table demonstrates the characteristics responsible for classifying conifers into a
particular vigor category. A conifer is assigned to a category based on meeting the
majority of specified criteria.
Excellent Good Fair Poor
Windward Appearance Tree may be flagged, but no obvious windward damage Minimal windward damage, but only on a few branches Most windward branches are damaged to some extent Tree is extensively flagged with lots of windward die-off
Needle Health and Color Needles are long & numerous; color is characteristic of a healthy specimen by species (i.e., dark blue green for whitebark) Needles generally healthy, but may have slightly yellowish color due to drought conditions Some needles have been blasted and/or yellow due to drought Red or brown dying and dead needles are numerous
New Shoot Status Numerous new shoots throughout entire tree; shoots are fully developing and healthy Many shoots present, but some may be underdeveloped New shoots developed, but were blasted and/or are underdeveloped New shoots generally absent from branch tips
Table 111.2 Sample sizes of small krummholz trees in the relative vigor study
Whitebark Pine Subalpine ^ir Engelmann Spruce
Divide Mountain 17 15 15
Line Creek RNA 21 12 20
35


Figure 111.2 Planted seedlings at the Line Creek RNA
Nursery grown Engelmann spruce seedlings planted on the leeward side of the four
experimental microsites at the Line Creek RNA. Microsites are as follows: a. whitebark
pine, b. Engelmann spmce, c. rock, and d. open or unprotected. These images were taken
in July 2011 at the time of planting. (Photo credits Sarah Blakeslee)
36


Figure 111.3 Germinated seeds on Divide Mountain
Representative Engelmann spruce seedling cluster on Divide Mountain showing
germination in three out of five sown seeds; this image was taken in July 2012 shortly
after germination and early in the treeline growing season. (Photo credit Sarah Blakeslee)
37


Figure 111.4 Example of before and after girdling and defoliation treatment
Representative girdling treatment site. In image a, the whitebark pine is sheltering
the windward edge of the tree island. Image b shows the subsequent exposure after
sawing through the main stem and defoliating the tree. Leeward conifer measurements
were taken in the areas previously sheltered by the whitebark pine. (Photo Credit: Sarah
Blakeslee)
38


Figure 111.5 Leeward shoot vs. exposed shoot sampling areas
This image shows representative sampling locations for leeward shoot and exposed
shoot measurements in the girdling study. The leeward shoots were measured on the
subalpine fir immediately leeward of the whitebark pine. Exposed shoots were measured
on a conifer of the same species located elsewhere on the same tree island but without
windward whitebark protection. The exposed shoots represent natural exposure and
serve as a baseline comparison to the initially protected leeward shoots. This procedure
was done for control as well as experimental tree islands. (Photo Credit: Sarah Blakeslee)
39


CHAPTER IV.
RESULTS
Relative Vigor Study
Transects. On Divide Mountain, 487 solitary, wind-exposed (unsheltered)
kmmmholz conifers were sampled within 20 transects. Species composition comprised
64% whitebark pine (n = 312)23% subalpine fir (n =111)and 13% Engelmann spruce
(n = 64). At the Line Creek RNA, 209 solitary exposed krummholz conifers were
sampled. We found species composition to be 83% whitebark pine (n =174),15%
Engelmann spruce (n = 32)and 1.4% subalpine fir (n = 3). Binomial tests of individual
transects indicated that sampled whitebark pine was present at statistically significant
higher abundances than expected at both Divide Mountain (79% transects with solitary
trees, n =15/19) and the Line Creek RNA (80% transects with solitary trees, n =12/15)
(Table IV.1).Transects with no solitary trees present were not included in these
analyses.
Based on transect data for solitary trees, tree density per square meter was
calculated by species for each study area (Figure IV.1).Divide Mountain had greater
densities for all three species than Line Creek RNA. At both study areas whitebark
densities were the highest of the three species (Divide = 0.031 0.03 trees/m2; Line
Creek RNA = 0.017 0.02 trees/m2). Whitebark pine densities with respect to other
species were as follows: Divide Mountain 5 times spruce and 3 times fir; Line Creek
RNA 5.5 times spmce and 58 times fir.
40


Differences in qualitative vigor trends between species were found on Divide
Mountain (Kruskal-Wallis rank sum x2 =18.9, df= 2,=7.8e-5). Whitebark pine had
higher vigor than both fir (W =19557, P = 0.037) and spmce (W=13026.5, P = 2.08e-
5). Fir vigor was higher than spmce (W = 2981.5, P = 0.047). Statistical differences in
species vigor were not observed at the Line Creek RNA (Kmskal-Wallis rank sum % =
0.82, df= 2,= 0.67).
Trends in microsites associated with initial tree establishment were characterized
by species. On Divide Mountain we found a statistical difference among proportions of
species found in unknown or minimally protecting microsites (i.e.small ground terraces)
compared to those leeward of more sheltering rocks or vegetation (x2 = 9.769, df= 2,=
0.008). This difference was due to a larger than expected number of whitebark pine that
established with no clear protective microsite or minimal protection and a greater than
expected number of subalpine fir associated with more sheltering microsites. At the Line
Creek RNA, similar statistical differences were also found (% =11.3217, df = 2,P =
0.003). Statistical significance largely derived trom a proportionally greater than
expected number of spmce and fir in more sheltering microsites.
Shoot Lengths and Shoot Growth Rate Comparisons. Kmmmholz shoot
lengths were compared by year between the species. Trends in length were similar
regardless of year and study area. Kmmmholz whitebark pine shoots were roughly 2 to 3
times longer than both spmce and fir shoots; and spmce and fir shoot lengths were not
statistically different from each other (Taole IV.2).
In nearly all comparisons across study areaswhitebark pines shoot growth rates
from July to September were the highest of all comparisons, representing growth rates on
41


average 2 to 10 times faster than fir and spruce. Fir and spmce growth rates were not
statistically different (Table IV.3). The one exception to this trend was similar growth
rates for subalpine fir and whitebark pine on the Line Creek RNA in 2011.
For upright, subalpine conifers, in all comparisons of shoot lengths by year and
study area, nested ANOVA results demonstrated that within tree variance contributed
only minimally to the overall variance (0.25% +/- 0.19). The largest portion of variance
was described by differences between means of different species (97.6% +/-1.98%).
Because within tree variance was low, we used the mean of the five measured shoots per
individual tree for all upright shoot comparisons for all subsequent analyses. Similar
trends were observed at both study areas (Table IV.4). On Divide Mountain, the same
shoot length trends occurred in both years: whitebark shoots were 1.5 to 3 times longer
than both spruce and fir shoots. In 2011spruce and fir shoots were not statistically
different, but in 2012 spmce shoots were longer than fir shoots. At the Line Creek RNA,
the same trend was observed in both years: whitebark pine shoots were 2 to 3 times
longer than both fir and spmce shoots, and the latter two species were not statistically
different.
Means of upright conifer shoot lengths after the minimum needle length was
subtracted were also compared in a One-Way ANOVA with a Tukeys post hoc by year
and study area to determine if species differences still existed. One-Way ANOVA
analyses were significant (Divide, 2011:F= 7.2835, df = 2,P = 0.003, 2012: F= 54.3, df
=2,= 6.297e-14; Line Creek2011:10.861df= 2,=0.0003, 2012: 197.23
df= 2, < 2.2e-16). Tukeys Post Hoc tests revealed that species mean shoot length
trends to remain mostly unchanged in most comparisons (Table IV.5). Whitebark pine
42


shoots were longer than subalpine fir shoots on Divide and Line Creek RNA. Whitebark
pine shoots were not longer than Engelmann spmce shoots on Divide in 2011,but they
were longer in 2012. At the Line Creek RNA, whitebark pine shoots were longer than
Engelmann spmce shoots in both years. Subalpine fir shoots were found to be equal in
length to Engelmann spmce in all comparisons.
With only one exception, upright shoots were longer than kmmmholz shoots of
the same species at each study area. The one exception was subalpine fir shoots on the
Line Creek RNA in 2012, where kmmmholz shoots did not differ from the upright shoot
counterparts. Proportions of upright to kmmmholz shoots were generally similar
between species for both years at both study areas (Table IV.6).
Other Small Tree Measurements. 2010-2012 increases in small tree stem
diameter, canopy area, and height measurements did not produce consistent statistical
differences in terms of species trends. Kmskal-Wallis rank sum test for differences in
median stem diameter resulted in no statistical differences among species on Divide
Mountain (x2 = 0.83, df= 2,=0.66). At the Line Creek RNAthis test was significant
(X2 =11.9, df= 2,=0.003). Median whitebark pine stem diameter increases were
smaller than both spmce (W= 294.5, P = 0.03) and fir (W= 34, P = 6.1e-4). Spmce and
fir median stem diameter increases were not significantly different from each other (W =
92.5, P = 0.29) (Appendix II, Table 1).
Kmskal-Wallis rank sum analysis of median canopy area increase indicated no
statistical differences among species at the Line Creek RNA =12.029, df = 2, .P =
0.21). On Divide Mountain, whitebark had a greater median increase in canopy area than
Engelmann spmce (W=140, P = 0.02), but not subalpine fir (W=103, P = 0.95).
43


Subalpine fir and Engelmann spmce were not statistically different (W= 20, P = 3.6e-4)
(Appendix II, Table 2).
Analyses of median height increases resulted in no significant differences at either
study area (Appendix II, Table 3).
Results Summary. We observed higher proportions of whitebark pine growing
in a solitary exposed state as compared to subalpine fir and Engelmann spmce.
Whitebark pine was also more common in minimally sheltering microsites than other
species. These results indicate that whitebark pine may be able to survive in harsh
treeline conditions better than spmce and fir.
Proportions of kmmmholz shoot lengths to upright tree shoot lengths were similar
between species, with kmmmholz shoot lengths being generally shorter than upright tree
shoot lengths. This indicates that there is a reduced ability to grow at treeline for all three
species. In terms of growth, whitebark pine produced the longest shoot lengths, both as a
kmmmholz and upright tree. Whitebark shoot growth rates were also generally faster
than spmce and fir. These results indicate the whitebark pine is capable of vigorous
growth during short growing seasons.
Planting and Sowing Study
Seedling Survival. One year after planting on Divide Mountain, the nursery
grown subalpine fir seedlings experienced very high overall mortality (90%). Of the 40
total seedlings planted per microsite type, survival was as follows: whitebark microsites -
12.5% (n = 5), spmce microsites 7.5% (n = 3), rock microsites 5% (n=2), and open
microsites -15% (n=6). Chi-square goodness of fit analysis showed no significant
differences in survival among microsite types (x2 = 2.5, df= 3,=0.47).
44


Survival of the planted Engelmann spmce seedlings was generally higher at the
Line Creek RNA, but overall mortality was high at 63.1%. Of the 40 seedlings planted
per microsite type, survival was as follows: whitebark microsites 32.5% (n=13), spmce
microsites 35% (n=14), rock microsites 42.5% (n=18), and open microsites 37.5%
(n=14). There were no significant differences in survival among microsite types (% =
0.59, df = 3,P = 0.9) (Figure IV.2). A One-Way ANOVA test for differences in mean
apical terminal shoot lengths of surviving seedlings did not show any statistical
differences among microsite type at either study area (Line Creek: F = 0.26, df = 3, .P =
0.85; Divide: F = 2.3, df = 3,P = 0.81).
We found qualitative measurement of seedling vigor to differ statistically among
microsite types in Fishers Exact Tests (Divide =7.3e-4; Line Creek/* = 9.6e-3). On
Divide, whitebark microsites had a greater than expected number of seedlings classified
as excellent vigor. Spmce and rock microsite vigor trends were distributed across
vigor classes as expected. Open microsites had a greater than expected number of poor
and rair vigor seedlings and no seedlings in the good and excellent vigor classes.
At the Line Creek RNA, whitebark pine microsites had a greater than expected
number of excellent vigor seedlings. Spmce microsites had fewer than expected good
vigor class seedlings and a greater than expected number of excellent vigor seedlings.
Vigor of rock microsite seedlings was distributed among classes as expected. Open
microsites had a greater than expected number of good vigor seedlings and a fewer than
expected number of excellent vigor seedlings.
45


Seed Germination and Summer Survival. Sown Engelmann spruce seed
germination totals were counted in July 2012, and revisited in September 2012 to assess
proportion of summer survival by microsite type.
Seed germination numbers at the Line Creek RNA were small. Of the 400 seeds
planted, only 7 (1.8%) germinated. The germinant distribution was as follows:
Whitebark microsites 2, spmce microsites 3, rock microsites none, and open
microsites 2. There were no significant differences in germination among the four
microsite types (Fisher5 s Exact Test, P = 0.44).
On Divide Mountain, 80 out of 400 (20%) seeds germinated. In July, whitebark
microsites had 12 germinants, spmce microsites had 17 germinants, rock microsites had
32 germinants, and open microsites had 19 germinants. Differences in germination
among the different microsite types was statistically significant (Fisher5 s Exact Test, P =
0.01). This is largely due to more germinations than expected in rock microsites and
fewer than expected in whitebark microsites.
When sites were revisited in September 2012, we observed that both study areas
had experienced substantial germinant mortality over the summer months. At the Line
Creek RNA, only 3 out of 7 germinants survived: 2 at whitebark microsites and 1 at a
spmce microsite. Rock and open microsites had no living germinants. Mortality was not
significantly different among microsite types (Fisher5s Exact Test, P = 0.31).
On Divide Mountain, 42 out of 80 germinants survived. The numbers of
surviving germinants are as follows: whitebark microsites -11 seedlings, spmce
microsites 8 seedlings, rock microsites -18 seedlings, and open microsites 5
seedlings (Figure IV.3). Difference in microsite type survival was statistically significant
46


(Fisher5s Exact Test, P = 0.004). A comparison of observed survival numbers vs. Chi-
square expected survival numbers revealed that whitebark microsites were associated
with a 5.7 times greater than expected survival advantage and a very low comparative
risk of death (0.18 times expected). Expected values were at or near 1.0. Open
microsites had the lowest chance of survival after germination at 0.64 times greater than
expected and were associated with a 1.56 times greater than expected relative expected
risk of death. Spruce and rock microsites had relative survival advantages very close to
expected (0.89 and 1.08 times expected, respectively). Similarly, these microsites had
relative death risks close to expected (1.12 and 0.93, respectively) (Table IV.7).
Results Summary. The seedling planting experiment did not have significant
results in terms of survival or terminal shoot lengths by microsite type. However, conifer
microsites were generally associated with higher overall health and vigor of the leeward
seedlings.
For the seed sowing experiment, only the results at the Divide Mountain study
area were instructive. Very few seeds germinated at the Line Creek RNAand thus
differences among microsite types were not observed. On Divide Mountain, rock
microsites initially favored germination. However, whitebark pine microsites had the
highest seedling summer month survival, indicating better shelter and perhaps more
favorable growing conditions than the other microsites examined. Whitebark microsites
also had the highest summer survival at the Line Creek RNA, but small sample sizes did
not yield statistical significance.
47


Girdling Study
Three ungirdled control whitebark trees (1 on Divide Mountain and 2 on White
Calf) were infected by blister mst and died over the course of the study. Their death
resulted in canopy defoliation. As a result, they no longer provided windward shelter to
the leeward conifer. These three sites were removed from analyses.
Leeward Conifer Vigor. There was a statistically significant difference between
control and experimental groups in the change in leeward conifer vigor over the 2010 to
2012 time period (n = 44, Fisher5s Exact Test, P = 0.002). This difference was attributed
to the number of experimental sites that lost vigor (77%; n =17 of 22), and the number of
control sites that remained the same or increased vigor over the course of the study (79%;
n =15 of 19).
Shoot Lengths. A BACI analysis comparing leeward conifer shoot lengths for
experimental or control treatment type indicated statistical differences over the course of
the study for treatment in terms of year (F = 8.17, df=2,P = 0.005). Initially in 2010,
there were no differences in shoot length by treatment site type (Wilcox Signed Rank, W
=236, P = 0.89). Overall, shoot lengths generally decreased over time, with
experimental sites experiencing greater decline in length than control sites (Figure IV.4).
The greatest difference in shoot lengths occurred over the 2011 to 2012 time period,
where shoot mortality was also highest (Wilcox Signed Rank, JV= 74.5, P = 0.0003). A
partial explanation for this decline in overall sample mean shoot lengths is the mortality
of some marked shoots over the course of the study.
We compared naturally exposed and leeward conifer shoot lengths for 2011 vs.
2012. We analyzed these shoots lengths by computing the difference between leeward
48


and exposed shoots for each tree island in the study. The differences were then compared
in a BACI analysis as groups of tree islands with a either control or girdled whitebark
pine for 2011 and 2012 (Figure 111.4). Results did not show different trends of mean
shoot length changes in terms of treatment type from year to year {F =1.4, df = 2, .P =
0.24). However, there was a difference in shoot lengths based solely on treatment type.
Control and experimental tree island shoot length differences were significant (F= 26.2,
df = l,P = 2.3e-6). There was a significantly greater difference in exposed vs. leeward
shoot lengths for tree islands with a control whitebark than a girdled whitebark: leeward
conifers in control tree islands had on average of 10 cm longer shoot length compared to
exposed shoots. In experimental tree islands, this difference was roughly 1 cm with
exposed shoots slightly longer than leeward shoots. This indicated that tree islands with
a dead windward whitebark pine will have reduced shoot length growth similar to areas
of the tree island with no windward protection. Results also indicated that the presence
of a windward whitebark pine is associated with longer annual shoot growth.
Shoot Mortality. The proportional differences in mortality from 2011 to 2012
between naturally exposed shoots and experimental or control leeward shoots from the
same tree island were statistically significant (W =129.5, P = 0.05). There were similar
differences in mortality between naturally exposed shoots with no windward protection
and the shoots leeward of a girdled whitebark pine. Leeward conifers in control sites had
an overall lower proportion of mortality than experimental leeward conifers. Mean shoot
mortality was 18.9% (SE = 5%) for control leeward conifers and 59% (SE = 7.7%) for
experimental leeward conifers, indicating that presence of windward whitebark pine
reduced shoot mortality in leeward conifers.
49


Results Summary. After losing windward whitebark pine shelter, leeward
conifers in experimentally girdled sites lost health and vigor over the course of the study.
These conifers also experienced shorter shoot lengths and higher shoot mortality than
leeward conifers in control sites with shelter from a windward whitebark pine, which
supports the hypothesis that loss of a windward whitebark from blister mst will be lead to
decreased health of the immediately leeward conifer(s).
50


Figures and Tables
0
01
d
WB ES SF
Figure IV.1 Solitary krummholz tree density by species on Divide Mountain and
Line Creek RNA
The number of conifers per square meter was calculated using the mean from 20
transects at each study area. We found that Divide Mountain had the highest solitary
kmmmholz conifer densities. Whitebark pine had the highest density at both study areas.
S Divide
Line Creek
so
so 300 5.0
(CNIE / S99J - #) Al!su(DQJ(Di!uoo
51


Table IV.l Species abundances of solitary conifers in transects
The number of solitary kmmmholz conifers is shown by transect per study area.
P- values are the result of binomial distribution tests comparing whitebark pine to an
expected equal distribution of 33%. The total number of transects that had a higher than
expected number of solitary whitebark pine is shown at the bottom of the P value
column. Bolded P values are transects that had a higher abundance of solitary
whitebark pine than expected based on an equal distribution of the three species.
a.
b.
Divide Mountain
Transect ID #WB #SF #ES # Total P- Value
1 18 6 8 32 0.0037
2 26 6 6 38 6.71e-10
3 4 0 0 4 0.012
4 25 3 6 34 1.31e-6
5 39 10 19 68 2.073e-5
6 14 1 6 21 0.001
7 10 0 0 10 1.53e-5
8 15 0 0 15 5.99e-8
9 19 15 1 35 0.0048
10 3 16 4 23 0.02
11 3 0 0 3 0.036
12 0 0 0 0 n/a
13 2 8 1 11 0.16
14 17 9 0 26 0.006
15 22 5 1 28 8.71e-7
16 1 0 0 1 0.33
17 5 8 1 14 0.21
18 10 1 2 13 0.0013
19 75 23 9 107 3.95e-15
20 4 0 0 4 0.012
TOTAL: 312 111 64 487 15/19
Line Creek RNA
Transect ID #WB #SF #ES # Total P- Value
1 26 0 0 26 3.03e-13
2 3 0 0 3 0.036
3 1 0 0 1 0.33
4 11 0 0 11 5.05e-6
5 1 0 0 1 0.33
6 4 0 0 4 0.012
7 25 0 0 25 9.18e-13
8 6 1 1 8 0.02
9 49 8 0 57 1.71e-16
10 0 0 0 0 n/a
11 12 12 2 26 0.059
12 4 1 0 5 0.04
13 12 9 0 21 0.013
14 0 0 0 0 n/a
15 0 0 0 0 n/a
16 0 0 0 0 n/a
17 7 0 0 7 0.00043
18 8 1 0 9 0.00085
19 0 0 0 0 n/a
20 5 0 0 5 0.0039
TOTAL: 174 32 3 209 12/15
52


WB>ES 205 5.3e-4
WB>SF 190 0.006
SF = ES 149 0.14
WB >ES 349 6.1e-6
WB >SF 192 8.4e-5
ES: =SF 82 0.36
2011 WP 22.0 (2.93)
ES 8.8 (0.85)
SF 11.3(1.56)
2012 WP 28.7 (3.35)
ES 9.3 (0.92)
SF 11.0(1.15)
a. __________________________
Line Creek RNA
Year Length Spp., Mean (SE) n Species Comparisons W- Statistic P- Value
2011 WP 48.1 (4.25) 21 WB>ES 36 2.8e-5
ES 22.9(2.45) 12 WB>SF 36 4.0e-4
SF 27.6 (3.34) 20 SF = ES 149 0.36
2012 WP 70.16 (3.67) 21 WB>ES 16 1.0e-8
ES 26.24 (3.74) 12 WB > SF 220 0.0004
SF 23.99 (4.3) 20 ES = SF 90 0.35
Table IV.2 Krummholz shoot lengths
2011 and 2012 shoot lengths (mm) were compared using Kmskal-Wallis Analysis
Rank Sum Test and Wilcox Post Hoc with a Bonferroni correction at a, Divide Mountain
and b, Line Creek RNA. All P values and test statistics shown are from the Wilcox
Post Hoc.
__________________________Divide Mountain___________________________
Year Length n Species W- P-
Spp Mean (SE) Comparisons Statistic Value
7 5 5
111111
7 5 5
111111
53


WB>ES 5.50 24 2.8e-5
WB>SF 4.04 24 3.8e-4
SF = ES -1.20 20 0.69
WB>ES 4.59 30 1.5e-5
WB>SF 3.88 30 2.1e-4
ES = SF -2.01 28 0.73
Line Creek RNA
2011 WP 0.21 (0.02)
ES 0.069 (0.01)
SF 0.093 (0.06)
2012 WP 0.15(0.03)
ES 0.03 (0.007)
SF 0.05 (0.007)
a. __________________________
Year Growth Rate Species t- P-
S.Mean (SE) n Comparisons Statistic df Value
2011 WP 0.52 (0.06) 20 WB>ES 9.22 38 0.0021
ES 0.25 (0.41) 20 WB = SF 1.57 28 0.80
SF 0.42 (0.03) 10 SF>ES -13.16 28 0.069
2012 WP 0.21 (0.03) 21 WB>ES 4.93 39 6.1e-6
ES 0.033 (0.01) 20 WB>SF 5.37 30 4.3e-5
SF 0.022 (0.007) 11 ES = SF 0.87 29 0.96
Table IV.3 Krummholz tree shoot growth rates
20112012 shoot growth rates units are mm/day. Tukeys post hoc test p-values
are shown for a, Divide Mountain and b, Line Creek RNA. All P values and test
statistics shown are from the Tukey5s Post Hoc. Sample sizes are shown for trees that did
not display a negative growth rate (i.e., loss of shoot growth from wind blasts or
experimental error in caliper placement)
________________________________Divide Mountain________________________________
Year Growth Rate Species t- P-
Spp., Mean (SE) n Comparisons Statistic df Value
5 15
111111
7 5 5
111111
54


Table IV.4 Upright upper subalpine conifer shoot lengths
Data are based on 2011 and 2012 measurements of five shoot lengths (mm) from
each of 10 trees per species in 2011 and 20 trees per species in 2012. Results are shown
for aDivide Mountain and bLine Creek RNA. -values and test statistics shown are the
result of a Tukeys post hoc analysis following significance from a One-Way ANOVA.
Divide Mountain
Year Length Spp., Mean (SE) n Species Comparisons t- statistic df P- Value
2011 WP 54.64 (1.6) 20 WB>ES 5.44 38 5.8e-5
ES 33.55 (3.5) 20 WB>SF 7.73 38 9.0e-7
SF 27.06(3.2) 20 SF = ES 1.37 38 0.27
2012 WP 67.12 (3.0) 20 WB>ES 9.76 38 <0.0001
ES 32.66 (1.9) 20 WB>SF 13.42 38 <0.0001
SF 23.21(1.4) 20 ES = SF 3.96 38 0.01
Line Creek RNA
Year Length Spp., Mean (SE) n Species Comparisons t- statistic df P- Value
2011 WP 69.92 (4.1) 20 WB>ES 6.72 38 5.0e-7
ES 34.64 (3.2) 20 WB = SF 5.95 38 0.74
SF 38.4(3.3) 20 SF>ES -0.81 38 3.5e-6
2012 WP 70.16 (1.4) 20 WB>ES 23.55 38 <0.0001
ES 26.24 (1.3) 20 WB>SF 24.10 38 <0.0001
b. SF 23.99 (1.3) 20 ES = SF 1.22 38 0.46
55


Table IV.5 Upright shoots with minimum needle lengths subtracted
Data are based on 2011 and 2012 measurements upright tree shoot lengths with
the minimum needle lengths (as described in Flora of North America) subtracted. All
values shown are Tukeys post hoc results following significance of a one-way ANOVA
at a. Divide Mountain and b. Line Creek RNA.
Divide Mountain
Year Length S.Mean (SE) n Species Comparisons t- statistic df P- Value
2011 WP 54.64 (1.6) 20 WB>ES 1.83 38 0.211
ES 33.55 (3.5) 20 WB>SF 4.37 38 0.002
SF 27.06 (3.2) 20 SF = ES 1.80 38 0.113
2012 WP 67.12 (3.0) 20 WB>ES 5.79 38 <0.001
ES 32.66 (1.9) 20 WB>SF 9.76 38 <0.001
SF 23.21(1.4) 20 ES = SF 4.80 38 0.001
Line Creek RNA
Year Length Spp., Mean (SE) n Species Comparisons t- statistic df P- Value
2011 WP 54.64 (1.6) 20 WB>ES 4.05 38 <0.001
ES 33.55 (3.5) 20 WB>SF 3.68 38 0.001
SF 27.06(3.2) 20 SF = ES -0.38 38 0.935
2012 WP 67.12 (3.0) 20 WB>ES 16.04 38 <0.001
ES 32.66 (1.9) 20 WB>SF 17.84 38 <0.001
SF 23.21(1.4) 20 ES = SF 2.30 38 0.069
56


Table IV.6 Small shoot lengths vs. upright shoot lengths: proportions
Data are based on 2011 and 2012 measurements of small trees and upright trees.
Mean (SE) is shown for each sample population. The proportion of small shoots to
upright shoots is shown for a. Divide Mountain and b. Line Creek RNA.
2011
Small Tree Upright Tree Proportion
Mean (SE) Mean (SE) Small to Up
WP 22.02 (2.9) 54.64(1.6) 0.403
ES 8.8 (0.85) 33.55 (3.5) 0.262
SF 11.3 (1.6) 27.06 (3.2) 0.419
2012
WP 28.66 (3.4) 67.12(3.0) 0.427
ES 9.29 (0.9) 32.66(1.9) 0.284
SF 10.97 (1.1) 23.21(1.4) 0.473
2011
Small Tree Upright Tree Proportion
Mean (SE) Mean (SE) Small to Up
WP 48.08 (4.2) 69.92 (4.1) 0.688
ES 22.87 (2.4) 34.63 (3.2) 0.660
SF 27.56 (3.8) 38.40 (3.3) 0.718
2012
WP 48.18(3.7) 70.16 (1.4) 0.687
ES 21.65 (3.7) 26.24(1.3) 0.825
SF 25.56 (4.3) 23.99(1.3) 1.065
57


Whitebark
Spruce
Rock
Open
20 1S 16 H 12 10 S 6 4 2 C 2 4 6 3 10 12 14 16 1S 20
Line Creek RNA Divide Mountain
Figure IV.2 One year post planting seedling survival per microsite
The total number of seedlings that survived for each microsite type is shown for
the Line Creek RNA (dark grey), and Divide Mountain (light grey).
58


White bark
Spruce
Rock
Open
35 30 55 20 15 10 5 5 tt 15 2D 25 30 35
Jul-12 Sept-12
Figure IV.3 2012 Divide Mountain seed germination counts
Seedling microsites were visited in July to observe initial germination and again in
September 2012 to observe summer drought mortality. While all microsite types
experienced some mortality, conifer germinants leeward of whitebark microsites tended
to have a lower chance of mortality while those in open exposed conditions experienced
the highest mortality. This indicates whitebark microsites may provide more favorable
conditions for seedling establishment.
Table IV.7 Summer 2012 survival advantage and relative death risk of seed
germinants on Divide Mountain
Using Chi-square expected microsite survival totals compared to actual survival
totals per microsite we calculated the relative survival advantage and death risk for each
microsite compared to each other. Of the four types, whitebark microsites are associated
with the highest survival advantage and lowest risk of death. Open microsites have the
lowest survival advantage and highest risk of death.
Microsite Relative Survival Advantage Relative Risk of Death
Whitebark 5.70 0.18
Spruce 0.89 1.12
Rock 1.08 0.93
Open 0.64 1.56
59


0
2010 2011 2012
Year
Figure IV.4 Girdling Study leeward conifer shoot length trends over time
Three year leeward conifer shoot length means for both treatment (girdled) and
control (non-girdled) sites at Whitecalf Mountain and Divide Mountain. Trends generally
reflect a decrease in mean shoot lengths over time, with treatment sites experiencing
shorter shoot lengths than control sites. Note: Control sites where the windward
whitebark died during the study were removed from this comparison.

/
5
2
o
2
5
o
5
(ELLOLI-uedo011 u0591A|
60


CHAPTER V.
SYNTHESIS AND DISCUSSION
Study Conclusions. The overall objectives of this research were to determine
experimentally and empirically the attributes and ecological interactions that enable
whitebark pine to facilitate tree island development, and to address how the mortality of
whitebark pine from blister mst may impact these ecosystem functions. We tested three
hypotheses focused on learning about whitebark pines facilitative functions:1)
Whitebark pine is hardier than other alpine treeline ecotone conifer species, as
demonstrated by more rapid shoot growth and higher survival at treeline; 2) whitebark
pine provides a more favorable microsite for tree island recruitment than other common
alpine treeline ecotone microsites; and 3) blister mst mortality of whitebark pine in
established tree islands will lead to loss of vigor of leeward conifers. Each of these
hypotheses relates to different aspects of whitebark pines role in facilitating formation of
tree islands and the maintenance of established tree islands (Figure 11.1).The results
from these studies provide new insight into whitebark pines role as a keystone and
foundation species at treeline.
First, our results clarify the issues of hardiness concerning the prevalence, and
shoot growth rates of whitebark pine in our two study areas, as illustrated by position #1
in the conceptual model (Figure II.1).This first finding is extremely important, because
whitebark pine in our study area was previously found to be the most common species
initiating multi-tree tree islands (Resler and Tomback2008). Our alpine treeline ecotone
study areas are characterized by harsh climatic conditions consisting or high winds, cold
61


temperatures, and direct solar radiation (Marr, 1977; Amo and Hammerly, 1984; Finklin,
1986; Maher et al.2005). With respect to the first hypothesis, under harsh, treeline
conditions, whitebark pine is the most prevalent conifer growing in exposed sites as a
solitary tree. While seed caching behaviors of Clark5 s nutcracker may lead to greater
seed distribution in open areas at treeline, whitebark pine appears better able to germinate
and establish under challenging conditions than both subalpine fir and Englemann spruce.
This is also demonstrated by the proportionally greater number of solitary whitebark
pines found in minimally sheltering niches or non-sheltering microsites compared to
Engelmann spruce or subalpine fir.
Also indicative of survival and vigor is the ability of whitebark pine to grow
longer shoot lengths, thus potentially expanding canopy biomass more rapidly, than the
other common treeline conifers. Conifer shoots are responsive to a variety of
environmental factors, including length of growing season, soil texture, moisture, and
nutrient levels, temperature, photoperiod, tree vigor, and tree species (Kozlowski, 1964).
Identical trends were found in analyses of shoot lengths both for kmmmholz conifers and
upper subalpine whitebark, spmce, and fir, with whitebark pine shoots longer than spmce
and fir shoots in both growth forms. This suggests that whitebark may have a species
growth advantage in general in the upper subalpine but this is also the case at treeline,
although all implications are not completely clear. The higher growth rates of
kmmmholz whitebark shoots during summer months in comparison with spmce and fir
also supports this finding. In order to produce longer shoots, whitebark pine must be able
to capitalize on the scarce environmental resources found at treeline and allocate them
into annual growth. There could well be trade-offs in growth that we are not aware of,
62


such as differential shoot to root ratios among the conifer species (Tilman 1988). Our
measurements of increases in stem diameter, canopy area, and height did not show
species differences. This is likely due to the short duration of the study. There may be
differences among species growth strategies (i.e.height vs. canopy volume). Longer
shoot lengths of whitebark pine suggest a species strategy for increasing canopy volume.
This trend might be advantageous in harsh treeline environments where upright growth is
often lost by wind and snow blasts, and ground-level canopy growth is favored (Arno and
Hammerly, 1984).
Given the short growing season at treeline, the ability to increase the volume of
photosynthetic biomass appears to support the premise that whitebark pine is hardier.
Whether this can happen may depend on exposure, flagging, water and nutrient
availabilityand annual snowpack depthwhich provides protection.
Because whitebark pine appears better able to survive and grow in the alpine
treeline ecotone than other conifer species, and is thus more prevalent, it is more likely to
initiate tree islands by acting as a nurse object for less hardy species, such as Engelmann
spruce as stated in hypothesis #2. Spmce seedling survival is facilitated with the
presence of overhead branches and windward protection (Hattenschwiler and Smith,
1999; Germino et al., 2002). Hypothesis #2 predicts that whitebark pine is a better
facilitator or nurse object than Engelmann spmce, rocks, or no object. If it is, than
whitebark pine will be more likely to be a tree islana initiator, as depicted in position 2
and the leeward red star in the conceptual model.
Our studies provide some evidence that whitebark microsites facilitate survival of
other conifers better than the other microsites examined. On Divide Mountain, rock
63


microsites better facilitated the initial germination of sown seeds, most likely due to
greater radiant heat. However, seeds germinating in whitebark microsites were
associated with greater survival during the summer months than all other tested
microsites, including Engelmann spmce. While not significant due to small sample sizes,
the same trend was observed on the Line Creek RNA. The summer months represent the
most critical stage for a newly germinated seedling, which must endure periodic drought
and UV radiation exposure. Studies have found minimal seedling mortality over winter
months (Day, 1964; Cui and Smith, 1991), likely because seedlings are covered by
snowpack at that time and have reduced exposure to harsh solar radiation, temperature
extremes, or chilling winds.
Differences in survival for the nursery grown seedlings among the various
microsite types did not show statistical differences, but seedling health and vigor were
greatest when associated with conifer microsites. Of the conifer microsites, whitebark
was the most likely to be associated with excellent seedling vigor, demonstrating that
growing conditions may be more moderate leeward of whitebark as compared to rocks,
Engelmann spmce, and open microsites. Hence, if a seed germinates leeward of a
whitebark pine at treelinethe seedlings chances of summer mortality are lower and it
will likely have greater growth vigor should it become established.
Every established solitary tree potentially could facilitate the establishment of a
leeward conifer, thus starting a tree island. Whitebark pine in our study areas is the most
common tree island initiator, and may create a more favorable microsite for leeward tree
survival. Once a whitebark pine facilitates the establishment of a leeward conifer, other
conifers may continuously establish on the leeward side of the developing tree island.
64


This cycle of establishment often continues until a large multi-tree island has formed.
Even after a tree island becomes established, the most windward conifer still provides a
sheltering role to the leeward individual(s). However, the importance of the windward
whitebark pine in offering protection to established conifers needs to be demonstrated in
order to fully understand the potential effects of mortality from blister mst, as stated in
hypothesis #3 and represented in the conceptual model in position #3. With blister mst
rapidly killing whitebark pine in the alpine treeline ecotone (3 of 22 control whitebark
died from blister mst over the course of our study), these previously sheltered subalpine
fir and Engelmann spmce individuals experience new exposure to wind, snow, and ice
blasts. By simulating blister mst through girdling and defoliating the windward
whitebark pine, we found that exposed leeward conifers were more likely to experience
decreased qualitative vigor, shorter shoot lengths, and have higher terminal shoot
mortality than control sites with a healthy windward whitebark pine. This windward
shelter may be most important in years with harsh winter conditions, low snowpack, and
cold, high winds.
In summary, we have found evidence for the importance of whitebark pine at
every investigated stage of the tree island development: the hardiness of whitebark pine
as an exposed, solitary tree and its role as a facilitator in providing a sheltering microsite,
the protection of an establishing seedling, and the windward role in established tree
islands. With whitebark pine declining at treeline from canopy damage and mortality
caused by blister mst infection, the ecosystem functions that we have studied here may
well be diminished.
Potential Implications for Whitebark Pine Decline at Treeline
65


Several interactive factors currently threaten the regeneration of whitebark pine
and the health of established whitebark pine within the alpine treeline ecotone. At the
same time these kmmmholz conifers are primarily experiencing blister mst mortality,
upper subalpine whitebark pine are declining due to blister mst infection, mountain pine
beetle outbreaks, and fire suppression, leading to the successional advancement of
predominantly subalpine fir and Engelmann spruce stands (Tomback and Achuff2010;
Tomback et al.2011).The loss of these reproductively viable upper subalpine
individuals has a direct impact on the alpine treeline ecotone. Since treeline whitebark
pine rarely produce cones with viable seeds, kmmmholz whitebark pine are regenerated
by upper subalpine elevation whitebark pine seed sources (Malanson et al,. 2007;
Tomback and Resler, 2007). Upper subalpine whitebark pine mortality leads to lower
seed production, and consequently, fewer seeds available for dispersal at treeline by
Clarks nutcracker (Tomback and Resler2007).
Alpine treeline ecotone vegetation dynamics may potentially be impacted by
mortality of treeline whitebark pine and the reduced regeneration due to loss of cone-
bearing upper subalpine individuals (Tomback and Resler, 2007; Resler and Tomback
2008). Since whitebark pine is a dominant alpine treeline ecotone tree island initiator,
fewer tree islands will be facilitated by whitebark (Figure V.l). Potential impacts of
fewer tree islands include increased soil erosion and snowpack melt-off (Smith et al.,
2009; Tomback et al.2011)possibly leading to changes in downstream hydrology.
Another issue likely to further impact alpine treeline ecotone dynamics is global
climate change. Warmer temperatures are expected to cause an upward shift in treeline
(Millar et al, 2004; Schrag et al., 2008), with an estimated elevation gain of 140 to 700 m
66


(Grace et al,2002). As treeline moves upward, new tree islands will form when solitary
conifers are able to establish in areas that were previously solely alpine tundra. These
conifers will act as nurse objects by providing windward shelter for leeward conifers,
ultimately leading to the formation of new tree islands at higher elevations. With
declining numbers of whitebark pine from blister mst at treeline and loss of seed
production, there will be less effective facilitation at the uppermost elevations of suitable
conditions to initiate tree islands. This is because whitebark pine is a dominant tree
island initiator through parts of its range. Thus, the decline of whitebark pine may lead to
a reduced ability of treeline as a whole to respond, possibly leading to the perception that
treeline is not moving up or moving more slowly than suitable temperature zones
(Tomback and Resler 2007). Potential implications for this issue are a loss of treeline
biodiversity as species compositions change (i.e., greater proportions of fir and spmce),
and a reduced range of the alpine treeline ecotone community if tree islands do not form
at higher elevations and the subalpine forest moves upwards.
There is a growing recognition of the importance of plant plant facilitative
interactions in stressful environmentssuch as treeline ecosystems (Calloway et al.2002;
Lortie et al.2004; Brooker et al.2008). Plant facilitators often germinate in harsh
environments with minimal protection. Once established, they provide favorable
growing conditions such as windward shelter, shade, and moisture retention (i.e.,
snowpack) for other species (Calloway, 1998; Lortie et al., 2004; Baumeister and
Calloway2006; Brooker et al.2008; Batllori et al.2009). The facilitation offered by
one species often has the potential to benefit more than one other species (Lortie et al.,
2004; Baumeister and Calloway2006). Here, we have identified the importance of
67


whitebark pine for the establishment of spmce and fir leading to development of tree
islands, and have thus demonstrated the cmcial facilitation role this species plays at
treeline in parts of its range. With whitebark pine currently declining, and climate change
potentially moving treelines upward, these community changes are already taking place
in treeline ecosystems in the Rocky Mountains.
68


Figures and Tables
Figure V.l Potential consequences of blister rust to alpine treeline dynamics
(modified from Tomback and Resler 2007).
69


REFERENCES
Arno, S. F. and R. P. Hammerly1984. Forest
Frontiers. Seattle, WA: The Mountaineers
Arno, S. F.1986. Whitebark pine cone crops-a diminishing source of wildlife food?
Western Journal of Applied Forestry. 1:92-94.
Arno, S. F., and R. J. Hoff. 1990. Pinus albicaulis Egelm. Whitebark pine. Pages 268-279
in R.M. Burns and B. H. Honkala, technical coordinators. Silvics of North America,
Vol.1,Conifers. USDA Forest Service, Agriculture Handbook 654, Washington D.C.
Bartos, D. L., and K. E. Gibson. 1990. Insects of whitebark pine with emphasis on
mountain pine beetle. Pages 171-178 in W. C. Schmidt and K. J. McDonald,
compilers. Proceedings-Symposium on whitebark pine ecosystems: Ecology and
Management of a high-mountain resource. USDA Forest Service Intermountain
Research Station, General Technical Report INT-270, Ogden, Utah.
Batllori, E., J. J. Camarero, J. M. Ninot, and E. Gutierrez. 2009. Seedling recmitment,
survival and facilitation in alpine Pinus uncinata tree line ecotones. Implications and
potential responses to climate warming. Global Ecology and Biogeography. 18:460-
472.
BaumeisterD.and R. M. Callaway. 2006. Facilitation by Pinus flexilis during
succession: a hierarchy of mechanisms benefits other plant species. Ecology.
87(7):1860-1830.
Benedict, J.B. 1984. Rates of Tree Island Migration, Colorado Rocky Mountains, USA.
Ecology, 3:820-823.
BrasierC. M. and K. W. Buck. 2001.Rapid evolutionary changes in a globally invading
fungal pathogen (Dutch elm disease). Biological Invasions 3:223-233.
BrookerR.W.F. T. MaestreR. M. Callawayand 21 others. (2008) Facilitation in plant
communities: the past, the present and the future. Journal of Ecology 96:18-34.
Callaway, R. M.1998. Competition and facilitation on elevation gradients in subalpine
forests of the northern Rocky Mountains, USA. Oikos, 82:561-573.
CallawayR.M.R.W. BrookerP. CholerZ. KikvidzeC.J. LortieR. MichaletL.
PaoliniF.I. PugnaireB. NewinghamE.T. AschehougC. ArmasD. Kikodzeand
B.J. Cook. 2002. Positive interactions among alpine plants increase with stress.
Nature 417: 844-848.
70


CampbellE. M. and J. A. Antos. 2000. Distribution and severity of white pine blister
mst and mountain pine beetle in whitebark pine in British Columbia. Canadian
Journal of Forest Research, 30:1051-1059.
Cavieres, L.A., C. L. Quiroz, M. A. Molina-Montenegro, A. A. Munoz, and A. Pauchard.
2005. Nurse effect of the native cushion plant Azorella monantha on the invasive
non-native Taraxacum officinale in the high-Andes of central Chile. Perspectives in
Plant Ecology, Evolution and Systematics, 7:217-226.
CavieresL.A.E. I. Badano, A. Sierra-Almeidaand M. A. Molina-Montenegro. 2007.
Microclimatic modifications of cushion plants and their consequences for seedlings
survival of native and non-native plants in the high-Andes of central Chile. Arctic,
Antarctic and Alpine Research, 39:229-236.
Chomesky EA, et al.2005. Science priorities for reducing the threat of invasive species
to sustainable forestry. BioScience 55:335-348.
ColeWalter E.and G. D. Amman. 1969. Mountain pine beetle infestations in relation to
lodgepole pine diameters. Res. Note INT-95. Ogden, UT: U.S. Department of
Agriculture, Forest Service, Intermountain Forest and Range Experiment Station, 7.
Cui, M. and W. K. Smith.1991.Photosynthesis, water relations and mortality in Abies
lasiocarpa seedlings during natural establishment. Tree Physiology 8:37-46.
Day, R.J. 1964. The microenvironments occupied by spmce and fir regeneration in the
Rocky Mountains.Can. Dept. Forestry. For. Res. Publ. No. 1037.
Diskin, M., K.C. Steiner and F.V. Hebard. 2006. Recovery of American chestnut
characteristics following hybridization and backcross breeding to restore blight-
ravaged Castanea dentata, For. Ecol. Manage. 223:439-447.
Easterling, W. 2005. Assessing the consequences of climate change for food and forest
resources: a view from the PCC. Climate Change 70:165-189.
eFloras. 2008. Published on the Internet http://www.efloras.org [accessed 7 November
2012]*' Missouri Botanical Garden, St. Louis, MO & Harvard University Herbaria,
Cambridge, MA.
EllisonA. M.M. BankB. ClintonE. ColbumK. ElliottC. FordD. FosterB.
Kloeppel, J. Knoepp, G. Lovett, J. Mohan, D. Orwig, N. Rodenhouse, W. Sobczak,
K. Stinson, J. Stone, C. Swan, V. Thompson, B. Von Holle, and J. Webster. 2005.
Loss of foundation species: Consequences for the structure and dynamics of forested
ecosystems. Frontiers in Ecology and the Environment 3:479486.
71


FinklinA. I.1986. A Climatic Handbook for Glacier National Park with Data for
Waterton Lakes National Park Ogden, UT U.S.D.A. Forest Service, Intermountain
Research Station.
Germino, M. J.W. K. Smithand A. C. Resor. 2002. Conifer seedling distribution and
survival in an alpine-treeline ecotone. Plant Ecology 162:157-168.
GernandtD. S.G. G. LopezO.S. Garciaand A. Liston. 2005. Phylogeny and
classification of Pinus. Taxon 54:29-42.
GraceJ.F. Bemingerand N. Laszlo. 2002. Impacts of Climate Change on the Tree
Line. Annals of Botany 90:537-544.
Habeck, J. R.1969. A gradient analysis if a timberline zone at Logan Pass, Glacier Park,
Montana. Northwest Science. 43(2): 65-73.
Hamann, A., and T. Wang. 2006. Potential effects of climate change on ecosystem and
tree species distribution in British Columbia. Ecology 87:2773-2786.
HattenschwilerS.and W. K. Smith. 1999. Seedling occurrence in alpine treeline
conifers: a case study from the central Rocky Mountains, USA. Acta Oecologica-
Intemational Journal of Ecology 20:219-224.
HoltmeierF.and G. Broil. 1992. The influence of tree islands and microtopography on
pedoecological conditions in the forests alpine tundra ecotone on niwot ridge,
Colorado front range, USA. Arctic and Alpine Research 24:216-228.
HutchinsH. E.and R. M. Lanner. 1982. The central role of Clarks nutcracker in the
dispersal and establishment of whitebark pine. Oecologia 55:192-201.
Jenkins, J., J. Powell, J. Logan, and B. Bentz. 2001.Low seasonal temperatures promote
life cycle synchronization. Bulletin of mathematical biology (0092-8240)63(3):573-
595.
KeaneR. E. 2001.Successional Dynamics: Modeling an Anthropogenic Threat. In D. F.
TombackS. F. Arno, and R. E. Keaneeds.Whitebark Pine Communities: Ecology
and Restoration. Washington, DC: Island Press, 159-192.
KovacsKent F. F.R. G. HaightD. G. McCulloughR. J. MercaderN. W. SiegertA.
M. Liebhold. 2010. Cost of potential emerald ash borer damage in U.S. communities
20092019. Ecological Economics69:569-578.
Kozlowski, T.1964. Shoot growth in woody plants. The Botanical Review 30(30):335-
396.
72


ListonA.W. A. RobinsonD. Pinero, and E. R. Alvarez-Buylla1999. Phylogenetics of
Pinus (Pinaceae) based on nuclear ribosomal DNA internal transcribed spacer region
sequences. Molecular Phylogenetic Evolution 11:95-109.
Logan, J. A., and J. Powell.2001.Ghost Forests, Global Warming and the Mountain Pine
Beetle. American Entomologist, 47(3):160-172.
LoganJ. A.W. W. Macfarlaneand L. Willcox. 2010. Whitebark pine vulnerability to
climate-driven mountain pine beetle disturbance in the Greater Yellowstone
Ecosystem. Ecological Applications 20:895-902.
LortieC J.R.W. BrookerP. CholerZ. KikvidzeR. MichaletF.I. Pugnaireand R.M.
Callaway. 2004. Rethinking plant community theory. Oikos 107:433-438.
Lloyd, A. H. and L. J. Graumlich., 1997. Holocene dynamics of treeline forests in the
Sierra Nevada. Ecology 78:119-210.
Loeh, C.,1996. Forest response to climate change: do simulations predict unrealistic
dieback? Journal of Forestry 94:13-15.
LorenzTeresa J. J.and K. A. Sullivan. 2009. Seasonal Differences in Space Use by
Clark's Nutcrackers in the Cascade Range. The Condor (Los Angeles, Calif.) (0010-
5422)111 (2)p. 326.
MaherE. L.M. J. Germino and N. J. Hasslequist. 2005. Interactive Effects of Tree and
Herb Cover on Survivorship, Physiology and Microclimate of Conifer Seedlings at
the Alpine Tree-line Acadian. Canadian Journal of Forest Research. 35(3):567-574.
MalansonG. P.D. R. ButlerD. B. FagreS. J. WalshD. F.TombackL. D. DanielsL.
M. ReslerW. K. SmithD. J. WeissD. L. PetersonA. G. BunnC. A. HiemstraD.
LiptzinP. S. BourgeronZ. Shenand C. Millar. 2007. Alpine treeline of western
North America: linking organism-to-landscape dynamics. Physical Geography,
28:378-396.
Marr, J. W.1977. The development and movement of tree islands near the upper limit of
tree growth in the southern Rocky Mountains. Ecology 58:1159-1164.
Mattson, D. J., and D. P. Reinhart. 1994. Bear use of whitebark pine seeds in North
America. In W. C. Schmidt and F. K. Holtmeier. Proceedings International
workshop on subalpine stone pines and their environment: The status of our
knowledge. USDA Forest Service, Intermountain Research Station, General
Technical Report INT-GTR-309, OgdenUtah.
McCaughey W. W.and W. C. Schmidt.1990. Autecology of whitebark pine. In:
Schmidt WC, McDonald KJ (eds) Proceedings Symposium on whitebark pine
73


ecosystems: Ecology and management of a high-mountain resource. Gen Tech Rep
INT-270, USDA For Serv Intermountain Res Sta, Ogden, UT, pp 85-96.
McDonaldG.I.and R. J. Hoff. 2001.Blister rust: An introduced plague.. In D. F.
TombackS. F. Arno, and R. E. Keaneeds.Whitebark Pine Communities: Ecology
and Restoration. Washington, DC: Island Press, 193-220.
McDonaldG. I.B. A. RichardsonP. J. Zambino, N. B. Klopfensteinand M. S. Kim.
2006. Pedicularis and Castilleja are natural hosts of Cronartium ribicola in North
America: a first report. Forest Pathology 36:7382.
McGregorM. D.and D. M. Cole. 1985. Integrating management strategies for the
mountain pine beetle with multiple-resource management of lodgepole pine forests.
USDA Forest Service Intermountain Research Station, General Technical Report
INT-174, OgdenUtah.
McKenneyD. W.J. H. Pedlar. K. LawrenceK. Campbelland M. F. Hutchinson. 2007.
Potential impacts of climate change on the distribution of North American trees.
Bioscience 57:939-948.
McKinneyS. T.and C. E. Fiedler. 2010. Tree squirrel habitat selection and predispersal
seed predation in a declining subalpine conifer. Oecologia 162:697-707.
MillarC. 1.R. D. WestfallD. L. DelanyJ. C. Kingand L. J. Graumlich. 2004.
Response of subalpine conifers in the Sierra Nevada, California, U.S.A., to 20th
century warming and decadal climate variability. Arctic, Antarctic, and Alpine
Research 36:181-200.
PriceR. A.A. Listonand S. H. Strauss. 1998. Phylogeny and systematic of Pinus.
Pages 49 68 in D. M. Richardson, editor. Ecology and biogeography of Pinus.
Cambridge University Press, New York.
R Development Core Team 2010. R: A language and environment for statistical
computing. R Foundation for Statistical Computing, Vienna, Austria. ISBN 3-
900051-07-0, URL http://www.R-project.org.
RaffaK. F.B. H. AukemaB. J. BentzA. L. CarrollJ. A. HickeM. G. Turnerand W.
H. Romme. 2008. Cross-scale drivers of natural disturbances prone to anthropogenic
amplification: the dynamics of bark beetle eruptions. Bioscience 58:501-517
ReslerL. M. 2004. Conifer Establishment Sites on a Periglacial LandscapeGlacier
National Park, Montana. Unpublished Ph.D. dissertation, Department of Geography,
Texas State University San Marcos
74


ReslerL. M.D. R. Butlerand G. P. Malanson. 2005. Topographic shelter and conifer
establishment and mortality in an alpine environment, Glacier National Park,
Montana. Physical Geography 26:112-125.
ReslerL. M. and D. F. Tomback. 2008. Blister rust prevalence in krummholz whitebark
pine: Implications for treeline dynamics. Arctic, Antarctic and Alpine Research,
40:161-170.
SchragA. M.A. G. BunnL. J. Graumlich. 2008. Influence of bioclimatic variables on
tree-line conifer distribution in the Greater Yellowstone Ecosystem: implications for
species of conservation concern. Journal of Biogeography 35(4):698-710.
Smith, E.K., L. M. Resler, E. Vance, W. Carstensen, and K Kolivras. 2011.Modeling the
Incidence of White Pine Blister Rust Infection in Whitebark Pine at Alpine Treeline
in the Northern Rocky Mountains using GIS. Arctic, Antarctic and Alpine Research
43(1):107-117.
Smith, William K. K., M. Germino, D. Johnson, and K. Reinhart. 2009. The Altitude of
Alpine Treeline: A Bellwether of Climate Change Effects. The Botanical review 75:
163-190.
SouleM.J. EstesJ. Bergerand C. Martinez del Rio. 2003. Ecological effectiveness:
Conservation goals for interactive species. Conservation Biology 17:1238-1250.
Spaulding, P.1909. European currant mst on the white pine in America. Circ. 36.
Washington, DC: U.S. Department of Agriculture, Bureau of Plant Industry 4 p.
SpauldingP.1911.The blister rust of white pine. Bull.206. WashingtonDC: Bureau of
Plant Industry 88 p.
SpauldingP.1922. Investigations of the white-pine blister rust. Bull. 957. Washington
DC: U.S. Department of Agriculture, Forest Service 100 p.
TilmanD.1988. Plant Strategies and the Dynamics and Structure of Plant Communities.
(Monographs in Population Biology, No. 26). Princeton University Press; Princeton,
New Jersey; First Edition.
TombackD.F. 1978. Foraging strategies of Clarks nutcracker. Living Bird 16:123-161.
TombackD. F.1982. Dispersal of whitebark pine seeds by Clarks nutcracker: a
mutualism hypothesis. J. Anim. Ecol. 51:451-467
TombackD. F.and Y. B. Linhart.1990. The evolution of bird-dispersed pines.
Evolutionary Ecology 4:185-219.
75


Tomback, D. F., J. K. Clary, J. Koehler, R. J. Hoff, and S. F. Amo. 1995. The effects of
blister mst on post-fire regeneration of whitebark pine: The Sundance burn of
northern Idaho (U.S.A.). Conservation Biology 9:654-664.
TombackD. F.1998. Clark's nutcracker (Nucifraga Columbiana). In PooleA. and F.
Gill, editors. The Birds of North America Philadelphia The Birds of North America,
Inc.No. 331.
TombackD. F.A. J. AnderiesK. S. CarseyM. L. Powelland S. Mellman-Brown.
2001.Delayed seed germination in whitebark pine and regeneration patterns
following the Yellowstone fires. Ecology. 82(9):2587-2600.
TombackD. F. 2001.Clarks Nutcracker: Agent of Regeneration. In D. F. TombackS.
F. Amo, and R. E. Keaneeds.Whitebark Pine Communities: Ecology and
Restoration. Washington, DC: Island Press, 88-104.
TombackD. F. and K. C. Kendall.2001.Biodiversity Losses: the downward spiral. In D.
F. TombackS. F. Amo, and R. E. Keaneeds.Whitebark Pine Communities:
Ecology and Restoration. Washington, DC: Island Press, 243-262.
TombackD.F.and L. M. Resler. 2007. Invasive pathogens at alpine treeline:
Consequences for treeline dynamics. Physical Geography. 28:397-418.
TombackD.F; and P. Achuff. 2010. Blister rust and western forest biodiversity:
Ecology, values, and outlook for white pines. Forest Pathology. 40:186-225.
Tomback, D.F., P. Achuff, A. W. Schoettle, J. W. Schwandt, and R. J. Mastrogiuseppe.
2011.The magnificent high-elevation five-needle white pines: ecological roles and
future outlook. In: The future of high-elevation, five-needle white pines in Western
North. USDA Forest Service Proceedings RMRS-P-63. 2011.America: Proceedings
of the High Five Symposium. 28-30 June 2010; Missoula, MT. Proceedings RMRS-
P-63. Fort Collins, CO: U.S. Department of Agriculture, Forest Service, Rocky
Mountain Research Station. 376 p.
U.S. Fish and Wildlife Service. 2011.Endangered and threatened wildlife and plants; 12-
month finding on a petition to list Pinus albicaulis as endangered or threatened with
critical habitat. Federal Register. 76(138):42631-42654.
WarwellM. V.G. E. Rehfeldtand N. L. Crookston. 2007. Modeling contemporary
climate profiles and predicting their response to global warming for whitebark pine
{Pinus albicaulis). In: Whitebark Pine: A Pacific Coast Perspective, Ashland, OR,
2006 August 27-31. Ed. By GoheenE. M.Sniezko, R. A. PortlandOR: US
Department of Agriculture, Forest Service, Pacific Northwest Region, 139-142.
76


APPENDIX
I. Small Tree Measurements
This appendix shows summary data for the Relative Vigor Study (hypothesis 1)small
tree measurements.
Table AI.l Small Tree Measurement Summaries
The 2010 baseline measurements taken for the small trees in the relative vigor
study are shown in the table below. Summaries are presented as means (SE) for each
measurement by species at a. Divide Mountain and b. Line Creek RNA.
Divide Mountain
Species Age (years) Height (cm) Stem Diameter (mm) Canopy Area (cm2)
WP ES SF 14.76 (2.8) 12.73 (1.1) 16.73 (2.5) 6.04 (0.85) 6.67 (0.98) 10.80 (2.34) 5.10(0.92) 6.57(1.46) 4.79 (0.86) 155.5 (58.7) 102.6 (32.7) 209.7 (88.7)
a.
Line Creek RNA
Species Age (years)____Height (cm)__Stem Diameter (mm)____Canopy Area (cm2\
WP 10.55(1.2) 9.31 (1.13) 5.58 (0.73) 157.6 (40.1)
ES 15.80 (1.1) 11.56 (1.43) 7.96 (0.75) 192.3 (34.6)
b. SF 18.4 (0.72) 15.42(1.49) 7.76 (0.84) 234.6 (45.5)
77


Table AI.2 Divide Mountain Small Tree Measurements by Site
The 2010 baseline measurements taken for the small trees in the relative vigor
study are shown for each individual site at a. Divide Mountain and b. Line Creek RNA
(next page).
Site Name Stem Diameter (mm) Height (cm) Canopy Area (cm2) Age (years)
Engelmann Spruce 1 2.70 5.7 42 8
Engelmann Spruce 2 4.33 5.5 172 9
Engelmann Spruce 3 4.81 5.6 36 11
Engelmann Spruce 4 2.06 4.0 26 8
Engelmann Spruce 5 3.40 4.0 26 9
Engelmann Spruce 6 2.03 3.0 8 8
Engelmann Spruce 7 2.33 4.2 10 12
Engelmann Spruce 8 2.81 3.0 43 11
Engelmann Spruce 9 3.36 8.0 105 18
Engelmann Spruce 10 1.86 3.0 31 17
Engelmann Spruce 11 3.29 5.0 22 11
Engelmann Spruce 12 9.78 11 217 15
Engelmann Spruce 13 12.70 11 452 19
Engelmann Spruce 14 8.70 15 280 15
Engelmann Spruce 15 7.76 12 70 20
Subalpme Fir 1 1.68 1.7 3 4
Subalpme Fir 2 1.81 2.4 3 8
Subalpme Fir 3 2.60 4.7 16 8
Subalpme Fir 4 2.18 5.2 19 9
Subalpme Fir 5 4.09 8.5 53 14
Subalpme Fir 6 4.13 9.0 7 13
Subalpme Fir 7 4.56 7.0 54 16
Subalpme Fir 8 5.54 8.0 240 19
Subalpme Fir 9 4.85 10 223 15
Subalpine Fir 10 2.77 5.0 12 17
Subalpine Fir 11 18.36 17.5 410 36
Subalpine Fir 12 13.45 26.0 105 27
Subalpine Fir 13 16.26 29.0 1140 28
Subalpine Fir 14 3.28 3.0 14 6
Subalpine Fir 15 13.06 25.0 848 31
Whitebark Pine 1 4.81 4.0 54 10
Whitebark Pine 2 2.35 6.0 3 8.5
Whitebark Pine 3 1.71 4.0 2 2.5
Whitebark Pine 4 2.07 3.4 5 5
Whitebark Pine 5 3.73 4.7 23 9
Whitebark Pine 6 2.43 4.5 35 8
Whitebark Pine 7 1.76 2.0 8 1
Whitebark Pine 8 4.66 5.5 87 15
Whitebark Pine 9 4.30 7.0 86 13
Whitebark Pine 10 2.65 3.5 14 8
Whitebark Pine 11 4.82 6.5 101 29
Whitebark Pine 12 2.82 3.0 36 7
Whitebark Pine 13 10.63 9.5 707 25
Whitebark Pine 14 12.81 7.0 531 26
Whitebark Pine 15 10.50 13.0 678 33
Whitebark Pine 16 2.82 4.0 35 11
Whitebark Pine 17 11.82 15.0 275 40
78


Site Name Stem Diameter (mm^
Engelmann Spruce 1 6.27
Engelmann Spruce 2 8.50
Engelmann Spruce 3 7.36
Engelmann Spruce 4 6.31
Engelmann Spruce 5 9.33
Engelmann Spruce 6 5.48
Engelmann Spruce 7 3.35
Engelmann Spruce 8 10.90
Engelmann Spruce 9 12.03
Engelmann Spruce 10 8.05
Engelmann Spruce 11 10.19
Engelmann Spruce 12 3.80
Engelmann Spruce 13 4.79
Engelmann Spruce 14 9.42
Engelmann Spruce 15 9.62
Engelmann Spruce 16 12.19
Engelmann Spruce 17 14.03
Engelmann Spruce 18 2.39
Engelmann Spruce 19 3.64
Engelmann Spruce 20 11.45
Subalpine Fir 1 9.40
Subalpine Fir 2 5.91
Subalpine Fir 3 7.96
Subalpine Fir 4 15.03
Subalpine Fir 5 10.56
Subalpine Fir 6 6.55
Subalpine Fir 7 7.64
Subalpine Fir 8 4.98
Subalpine Fir 9 5.74
Subalpine Fir 10 3.96
Subalpine Fir 11 7.49
Subalpine Fir 12 7.89
Whitebark Pine 1 4.90
Whitebark Pine 2 5.97
Whitebark Pine 3 2.42
Whitebark Pine 4 2.63
Whitebark Pine 5 3.20
Whitebark Pine 6 8.42
Whitebark Pine 7 5.02
Whitebark Pine 8 4.22
Whitebark Pine 9 4.63
Whitebark Pine 10 8.65
Whitebark Pine 11 2.83
Whitebark Pine 12 2.92
Whitebark Pine 13 9.09
Whitebark Pine 14 3.37
Whitebark Pine 15 6.51
Whitebark Pine 16 5.72
Whitebark Pine 17 7.05
Whitebark Pine 18 2.02
Whitebark Pine 19 13.52
Whitebark Pine 20 1.37
b. Whitebark Pine 21 12.65
Height (cm) Canopy Area (cm2) Age (years)
8 99 17
15 258 26
10 94 21
4 56 10
10 165 14
7.5 118 21
4.5 82 10
14 506 19
13 212 13
14 159 16
23 415 18
8 90 9
4 64 15
13 237 23
12 163 18
25.5 560 13
17 189 15
3.5 8 7
5 33 12
20.5 339 19
20 467 17
15 119 20
11.5 82 18
24 490 23
24 440 21
12.5 363 18
15.5 204 20
11.5 121 20
10 121 16
12.5 94 16
9.5 138 14
19 176 18
6 39 9
7.5 113 14
3 21 6
2.5 22 2
8.5 57 12
12 165 12
5 132 6
6 49 5
9 176 9
12 346 15
7 57 7
6 57 8
16 247 14
11 42 9
10 147 18.5
13 247 13.5
13.5 72 11
6.5 61 5
21.5 785 20
1.5 28 2
18 451 23
79


WB=ES 122.5 0.86
WB = SF 149 0.44
SF = ES 95 0.47
WP 0.80 (0.29)
ES 0.98 (0.28)
SF 0.77 (0.37)
II. Small Tree Analyses
This appendix shows results from Relative Vigor Study (hypothesis 1)analyses that did
not show statistical differences among species.
Table AII.l Change in krummholz tree stem diameters
The increase in stem diameters (mm) over the 2010 to 2012 time period is shown
for a, Divide Mountain and b, Line Creek RNA. Analyses were completed with a
Kmskal-Wallis Rank sum test and Wilcox-Signed Rank test with a Bonferroni correction.
All P values and test statistics shown are from the Wilcox Post Hoc.
______________________Divide Mountain _______________________
Increase n Species W- P-
Spp.,Mean (SE) Comparisons Statistic Value
b.
Line Creek RNA
Increase Spp., Mean (SE) n Species Comparisons W- Statistic P- Value
WP 0.99(0.17) 21 WB ES 2.31 (0.46) 20 WB SF 2.58 (0.43) 12 SF = ES 92.5 0.29
7 5 5
111111
80


WB=ES 140 0.78
WB = SF 103 0.91
SF = ES 20 0.56
WP 1.06 (0.74)
ES 0.52 (0.37)
SF 1.40 (0.52)
Line Creek RNA
Area n Species W- P-
Sgp.,Mean (SE) Comparisons Statistic Value
WP 196.6 (29.9) 21 WB = ES 272 0.11
ES 121.3 (18.5) 20 WB = SF 138 0.67
SF 172.6 (38.4) 12 SF = ES 153 0.21
Table AII.3 Krummholz tree heights
Difference between 2012 and 2010 heights (mm) were calculated with a Kmskal-
Wallis Rank sum test and Wilcox-Signed Rank Test with a Bonferroni correction.
Results are shown for a, Divide Mountain and b, Line Creek RNA. All P values and
test statistics shown are from the Wilcox Post Hoc.
________________________Divide Mountain ______________________
Height n Species W- P-
Spp.,Mean (SE) Comparisons Statistic Value
WB>ES 140 0.02
WB = =SF 103 0.95
SF = ES 20 3.6e-4
WP 56.3 (32.8)
ES -1.65 (0.76)
SF -2.3 (63.5)
b.
Line Creek RNA
Height Spp., Mean (SE) n Species Comparisons W- Statistic P- Value
WP 3.66 (0.54) 21 WB=ES 204.5 0.90
ES 4.3 (0.69) 20 WB = SF 168 0.12
SF 1.88 (0.77) 12 SF = ES 163 0.10
Table AII.2 Krummholz tree canopy areas
Canopy area increases (cm2) from 2010 to 2012 were calculated with a Kmskal-
Wallis Rank sum test and Wilcox-Signed Rank Test with a Bonferroni correction.
Results are shown for a, Divide Mountain and b, Line Creek RNA. All P values and
test statistics shown are from the Wilcox Post Hoc.
______________________Divide Mountain _______________________
Area n Species W- P-
Spp., Mean (SE) Comparisons Statistic Value
7 5 5
111111
7 5 5
111111
81


III. Planting Study Microsite Heights
This appendix shows the heights for the seedling planting and seed sowing sites by
microsite type.
Table AIII.l Planting and sowing study microsite heights
The mean and standard deviation (cm) are reported for each microsite type by
seedling and seed sites at a. Divide Mountain and b. Line Creek RNA. All sample sizes
are 20 sites per microsite for both seeds and seedlings at each study area.
a.
b.
Divide Mountain
Site Type Whitebark Spruce Rock Open
Seedlings 19.0 (6.4) 20.9 (6.9) 15.6(5.8) n/a
Seeds 14.1 (5.0) 13.4(4.3) 9.9 (2.7) n/a
Line Creek RNA
Site Type Whitebark Spruce Rock Open
Seedlings 44.3 (10.7) 44.8(12.5) 10.6(3.8) n/a
Seeds 27.2 (10.8) 28.6 (8.2) 7.2 (2.4) n/a
82


Full Text

PAGE 1

ASSESSING WHITEBARK PINE VIGOR AND FACILITATION ROLES IN THE ALPINE TREELINE ECOTONE by SARAH C. BLAKESLEE B.S. Biology, University of Colorado Colorado Springs, 2008 A thesis submitted to the Faculty of the Graduate School of the University of Colorado in partial fulfillment of the requirements for the degree of Master of Science Biology 2012

PAGE 2

ii This thesis for the Master of Science degree by Sarah C. Blakeslee has been approved for the Department of Integrative Biology by Diana F. Tomback, C hair Michael B. Wunder Leo P. Bruederle 1 6 November, 2012

PAGE 3

iii Blakeslee, Sarah, C. (M.S., Department of Integrative Biology Master of Science) Assessing W hite bark Pine Vigor and F acilitation Roles in the Alpine Treeline E cotone Thesis direc ted by Professor Diana F. Tomback ABSTRACT Whitebark pine (Pinus albicaulis) is an upper subalpine and treeline conifer of the higher mountains of the western United States and Canada At treeline on the Eastern Front of the Rocky Mountains, whitebark pine appears to facilitate tree island development. It is currently declining at treeline from infection by white pine blister rust, caused by Cronartium ribicola We are studying how whitebark pine facilitates tree island formation and how blister rust mortali ty may affect these processes in two treeline study areas in Montana: Divide Mountain, Glacier N ational P ark and Blackfeet Indian Reservation ; and Line Creek Research Natural Area, Custer National Forest. We tested three hypotheses: 1) W hitebark pin e is ha rdier than other treeline conifer species as demonstrated by more vigorous growth and survival at treeline 2) whitebark pine provides more favorable leeward microsite s for tree island recruitment than other conifers or microsites, and 3) death of windwar d whitebark pine in establishe d tree islands leads to vigor loss in leeward conifers. We found support for each hypothesis. Whitebark pine was significantly more numerous tha n both spruce (Picea engelmannii) and fir (Abies lasiocarpa) among solitary tree s Solitary, krummholz whitebark pine trees produced significantly longer annual shoots than both spruce and fir, indicating faster branch growth and canopy area increase under harsh co nditions. These results indicate higher vigor and potentially higher s urviva l rate than spruce and fir. Germinated spruce seeds had higher summer survival

PAGE 4

iv and planted fir and spruce seedlings had greater vigor when leeward of whitebark pine compared to spruce, rock, or exposed microsites suggesting that whitebark pine mi crosites provided more protection In established tree islands, the presence of a windward whitebark pine was associated with greater general vigor, longer shoot lengths, and lower shoot mortality in leeward trees than under experimental conditions where the windward whitebark pine was girdled and defoliated Because whitebark pine is better able to survive and grow in the alpine treeline ecotone than other conifer species, this may, in part, explain its greater prevalence. Whitebark pine is more likely t o facilitate tree island development and provide a better microsite for seedling establishment The form and content of this abstract are approved. I recommend its publication. Approved: Diana F. Tomback

PAGE 5

v DEDICATION I dedicate this work to Logan, in appreciation of all his love and support.

PAGE 6

vi ACKNOWLEDGMENTS I would like to thank m y advisor Dr. Diana F. Tomback for her willingness to sha re her knowledge and experience and for assisting me throughout this entire process. Also invaluable t o this work are Jill C. Pyatt and Libby R. Pansing who provided outstanding suppo rt both in the field and back in lab I am also appreciative of the field assistance provided by Logan Wealing, Soledad Diaz, and Aaron Wagner. I thank my committee members Dr. Leo P. Bruederle and Dr. Michael Wunder for their advice and as sistance throughout this project. A special thank you goes to our collaborators in the Dr. Lynn Resl er and Dr. George Malanson labs because without them this project never would have been possible. I also appreciate the help provided by numerous folks and various agencies, especially Kent Houston of the Shoshone National Forest, Custer National Forest, the Blackfeet Tribal Nation, Colorado State Forest Service Nursery, Glacier National Pa rk and the ecology and evolutional biology group members at CU Denver

PAGE 7

vii TABLE OF CONTENTS CHAPTER I. WHITEBARK PINE BACKGROUND ................................ ................................ .......... 1 Tax onomy and Distribution ................................ ................................ .................... 1 Whitebark Pine Seeds: Seed Dispersal and Food Source ................................ ....... 1 The Alpine Roles ................................ .... 3 Threats to Whitebark Pine ................................ ................................ ...................... 6 Global Climate Change and Treeline Impacts ................................ ........................ 9 Figures and Tables ................................ ................................ ................................ 12 II. INTRODUCTION ................................ ................................ ................................ ........ 14 Background ................................ ................................ ................................ ........... 14 Conceptual Framework ................................ ................................ ......................... 17 Hypotheses for Testing ................................ ................................ ......................... 18 Figures and Tables ................................ ................................ ................................ 21 III. METHODS ................................ ................................ ................................ ................. 22 Study Areas ................................ ................................ ................................ ........... 22 Relative Vigor Study ................................ ................................ ............................. 23 Field Methods. ................................ ................................ ................................ 24 Data Analysis. ................................ ................................ ................................ 26 Planting and Sowing Study ................................ ................................ ................... 28 Field Methods. ................................ ................................ ................................ 28 Data Analysis. ................................ ................................ ................................ 30 Girdling Study ................................ ................................ ................................ ....... 31 Field Methods. ................................ ................................ ................................ 31 Data Analysis. ................................ ................................ ................................ 32

PAGE 8

viii Figures and Tables ................................ ................................ ................................ 34 IV. RE SULTS ................................ ................................ ................................ ................... 40 Relative Vigor Study ................................ ................................ ............................. 40 Transects ................................ ................................ ................................ ......... 40 Shoot Lengths and Shoot Growth Rate Comparisons ................................ ..... 41 Other Small Tree Measurements. ................................ ................................ ... 43 Results Summary. ................................ ................................ ........................... 44 Planting and Sowing Study ................................ ................................ ................... 44 Seedling Survival. ................................ ................................ ........................... 44 Seed Germination and Summer Survival. ................................ ....................... 46 Results Summary. ................................ ................................ ........................... 47 Girdling Study ................................ ................................ ................................ ....... 48 Leeward Conifer Vigor ................................ ................................ ................... 48 Shoot Lengths. ................................ ................................ ................................ 48 Shoot Mortality. ................................ ................................ .............................. 49 Results Summary. ................................ ................................ ........................... 50 Figures and Tables ................................ ................................ ................................ 51 V. SYNTHESIS AND DISCUSSION ................................ ................................ .............. 61 Study Conclusions. ................................ ................................ ............................... 61 Potential Implications for Whitebark Pine Decline at Treeline ............................ 65 Figures and Tables ................................ ................................ ................................ 69

PAGE 9

ix REFERENCES ................................ ................................ ................................ ................. 70 APPENDIX I. Small Tree Measurements ................................ ................................ ............................ 77 II. Small Tree Analyses ................................ ................................ ................................ ..... 80 III. Planting Study Microsite Heights ................................ ................................ ............... 82

PAGE 10

x LIST OF TABLES Table I.1 US and Canadian conifers susceptible to white pine blister rust inf ection ................ 13 III.1 Qualitative vigor categories ................................ ................................ ..................... 35 III.2 Sample sizes of small krummholz trees in the relative vigor study ......................... 35 IV.1 Species abundances of solitary conifers in transects ................................ ............... 52 IV.2 Krummholz shoot lengths ................................ ................................ ........................ 53 IV. 3 Krummholz tree shoot growth rates ................................ ................................ ........ 54 IV. 4 Upright upper subalpine conifer shoot lengths ................................ ........................ 55 IV. 5 Upright shoots with minimum needle lengths subtracted ................................ ........ 56 IV. 6 Small shoot lengths vs. upright shoot lengths: proportions ................................ ..... 57 IV. 7 Summer 2012 survival advantage and relative death risk of seed germinants on Divide Mountain ................................ ................................ ................................ ............... 59 AI.1 Small Tree Measurement Summaries ................................ ................................ ....... 77 AI. 2 Divide Mountain Small Tree Measurements by Site ................................ ............... 78 AII.1 Change in krummholz tree stem diameters ................................ ............................. 80 AII.2 Krummholz tree canopy areas ................................ ................................ ................ 81 AII. 3 Krummholz tree heights ................................ ................................ .......................... 81 AIII.1 Planting and sowing study microsite heights ................................ ......................... 82

PAGE 11

xi LIST OF FIGURES Figure I.1 Distribution of Pinus albicaulis in North America ................................ .................... 12 I.2 Image of activ e blister rust stem canker on an infected whitebark pine. .................... 13 II.1 Overall conceptual model ................................ ................................ .......................... 21 III.1 Research study areas ................................ ................................ ................................ 34 III.2 Planted seedlings at the Line Creek RNA ................................ ................................ 36 III.3 Germinated seeds on Divide Mountain ................................ ................................ .... 37 III.4 Example of before and after girdling and defoliation treatment .............................. 38 III.5 Leeward shoot vs. exposed shoot sampling areas ................................ .................... 39 IV.1 Solitary krummholz tree density by species on Divide Mountain and Line Creek RNA ................................ ................................ ................................ ................................ .. 51 IV.2 One year post planting seedling survival per microsite ................................ .......... 58 IV.3 2012 Divide Mountain seed germination counts ................................ ..................... 59 IV.4 Girdling Study leeward conifer shoot length trends over time ................................ 60 V.1 Potential consequences of blister rust to alpine treeline dynamics .......................... 69

PAGE 12

xii LIST OF ABBREVIATIONS 1. ATE Alpine Treeline Ecotone the region between the subalpine forest and alpine tundra where conifers are krummholz or dwarfed 2. SF Subalpine Fir ( Abies lasiocarpa ) 3. W P Whitebark Pine ( Pinus albica u lis ) 4. ES Engelmann Spruce ( Picea engelmannii ) 5. RNA Resear ch Natural Area; in reference to the Line Creek Natural Area located on the Beartooth Plateau, Montana

PAGE 13

1 CHAPTER I. WHITEBARK PINE BACKGROUND Taxonomy and Distribution Whitebark pine ( Pinus albicaulis ) is one of several sto ne pines comprising Pinus subgenus Strobus, section Strobus, subsection Cembrae (Price et al, 1998). Since monophyly of Subsection Cembrae is unsubstantiated it has been proposed that subsection Cembrae be merged with subsection Strobi into a new subsec tion Strobus ( Liston et al, 1999 ; Gernandt et al, 2005). This classification has yet to be officially recognized. The stone pines of subsection Cembrae are characterized as having five needles per fascicle and indehiscent cones with wingl ess seeds that ar e dispersed by nutcrackers ( Nucifraga spp. ) (McCaughey and Schmidt, 2001). Whitebark pine is distributed from the s outhern Sierra Nevada of California north through the Cascade and coastal ranges into British Columbia ; and from the Greater Yellowstone regi on of Wyoming north through the Rocky Mountains of British Columbia and Alberta Canada (Fig ure I. 1). Whitebark pine is limited to upper subalpine and treeline forests in high elevation mountains from 37 to 55 N (Arno and Hoff, 1990). It is often a domin ant treeline species, except at its most northern limits and in the snowiest regions of the southern Canadian Rockies and coastal ranges (Arno and Hammerly, 1984). Whitebark pine assumes a krummholz growth form at treeline in the dri er mountain ranges (Ar no and Hammerly, 1984). Whitebark Pine S eeds : Seed Dispersal and Food Source Because whitebark pine has indehiscent cones with wingless seeds, it relies on a co Nucifraga columbiana ) for seed dispersal

PAGE 14

2 (Tomback, 1982). It is possible that this method of seed dispersal evolved as a consequence of both genetic drift in small populations and seed selection choice by and Linhart, 1990). Every year from late summer to early fall, these bir ds gather seeds from cones, carry them within their sublingual pouch, and cache them throughout the subalpine and treeline terrain. In many regions nutcrackers typically select for seed caching steep, south facing slopes that accumulate minimal snowpack ( Tomback, 1982). Distances from the cache to the original seed source can vary from a few meters up to 29 km in distance and 3 07 meters in elevation (Lorenz and Sullivan, 2009). The seeds that are not later consumed germinate, thereby regenerating the speci es. Resler (2004) observed whitebark growing at treeline and found that many cache seedling survival. These microsites, which are often terraces or boulders, may facil itate the germination and growth of whitebark pine seedling s nutcracker and the natural hardiness of whitebark pine account for a large majority of the spatial distribution and population genetic structure o f the species (T omback, 2001). Seeds from whitebark pine cones are also an important food source for other wildlife. Grizzly bears ( Ursus arctos ) and pine squirrels ( Tamiasciurus spp.) rely on these seeds ( Mattson et al, 1992 ; McKinney and Fiedler, 2010). In the Great er Yellowstone Area, pine squirrels store cones in middens that are later raided by the bears. These seeds are a large part of the good cone crop years (Matson and Reinhart, 1994). Other wildlife species that consume these seeds include small mammals, such as chipmunks ( Tamius spp.), and golden mantle

PAGE 15

3 ground squirrels ( Spermophilus lateralis ); and some small birds, such as woodpeckers, nuthatches, finches, and Cyanocitta stelleri ) (Tomback, 1978; Hutchins and Lanner, 1982; Tomback and Kendall, 2001). The Alpine The transition between the subalpine forest and alpine tundra is referred to as the alpine treeline ecotone (ATE). This high elevation zo ne is characterized by krummholz conifers, dry, windswept slopes, and cold temperatures (Marr, 1977; Arno and e treeline seldom produce cones with viable seeds, so this tree community is generated by seeds coming from the subalpine zone. Therefore, trees in the alpine treeline ecotone must physically adapt to survive the harsh climate (Malanson et al 2007). Kr ummholz growth fo r ms result when wind blasted snow and ice particle s kill upright growth. Consequently, the only branches able to survive are those that grow low to the ground. Krummholz trees often have foliage surface temperatures 5 10 higher than amb ient temperature. Taller trees have surface temperatures 5 lower than ambient temperatures. Seedlings in the alpine treeline ecotone may be sheltered and in favorable microclimates, but as they grow taller, their growth rate is reduced and direction of growth altered by wind and desiccation ; thus they become dwarf ed or krummholz (Grace et al, 2002). Survival in the alpine treeline ecotone is often increased by the formation of tree islands. Tree islands are krummholz mats containing one or more indiv idual trees growing in close proximity (Marr, 1977). Solitary tree islands are comprised of one tree.

PAGE 16

4 Multi tree islands are comprised of two or more individual trees or ma ny branches that growing in layered form due to adventitious roots (Benedict, 1984 ). Foundation species are highly abundant ecosystem components that exert much influence on ecosystem function and stability (Ellison et al ., 2005). Keystone species promote and support the biodiversity of their ecosystems (Soule et al, 2003). Foundati on and keystone species in forest ecosystems have the ability to maintain biotic and abiotic ecosystem components. If a keystone or foundational species declines, there could be a resulting trophic cascade with a loss of biodiversity or ecosystem functio n (Ellison et al 2005). Throughout its range, whitebark pine is both a foundation and keystone species ( Tomback and Achuff 2010). Whitebark pine acts as a facilitator or creating protective microsites for less hardy conifer species (Ca llaway, 1998). In the island initiation Whitebark pine s growing in sheltered microsites can facilitate community development by mitigating the harsh co nditions o n their leeward side. Resler and Tomback (2008) found that whitebark pine was the windward tree island initiator for nearly half the multi tree tree islands among two study si t es east of the Continental Divide They also found that whitebark pine was an important component of tree islands in this region: 255 out of 266 tree islands sampled contained whitebark pine Research has demonstrated just how important tree islands can be for survival of less hardy conifers. Hattenschwiler and Smith (1999) studie d distributions of subalpine fir and Engelmann spruce in the central Rocky Mountains to determine locations with greatest survival. Although Engelmann spruce appears to germinate quickly and at lower temperatures than subalpine fir, no seedlings of either species could survive on the

PAGE 17

5 windward side of tree islands. In the alpine treeline ecotone, the most frequent location for seedling establishment was on the leeward side of tree islands where snow accumulation is maintained at a moderate depth of 0.5 1 .5 m thus offering protection Germino et al (2002) found that the microsites with windward protection were associated with a 20% higher survival rate of Engelmann spruce seedlings. Additionally, seedling survival was 70% higher when microsite features, such as branches, were located directly above the seedlings. They claim that close proximity to tree islands and overhead structures, such as b ranches, may moderate solar and long wave radiation, reduce daytime temperature extremes, and maintain snowdrif t accumulation. These factors increase seedling survival by making environmental conditions more moderate. ecological function both for community development at treeline and for the prov ision of ecosystem services to people (Resler and Tomback 2008) T ree islands provide important ecosystem services Tree islands are involved in watershed hydrology through the maintenance of snowpack, which regulates the rate of snowmelt run off ( Holtme ier and Broll, 1992 ). Conifer roots also help stabilize soil erosion (Tomback et al. 2001). If no tree islands are present to perform these functions, erosion and summer drought may result. Farmers and ranchers with land in the valley bottoms and on the plains, downstream of these mountains, rely on regulated water from snowmelt to fill streams and creek beds necessary for crops and livestock. Municipal water reservoirs are sometimes kept at appropriate levels by snowmelt (Smith et al 2009). Late sum mer shortages could lead to rationing or the costly service of transporting water supply to the region.

PAGE 18

6 Threats to Whitebark Pine Currently, there are several threats to whitebark pine throughout its range. Fire suppression, mountain pine beetle outbreaks and white pine blister rust are compounding of some threats, making it challenging to predict how whitebark ecosystems will respond (Tomback and Achuff, 2010). Fire is a natural occurrence in forest ecosystems. A burned area generates burn sites for seed caches (Tomback, 2001), w hich means that whitebark pine is an important species for fo rest regeneration after a fire. New whitebark growth creates sheltered microsites for subalpine fir and Engelmann spruce to grow, increasing conifer biodiversity along with forest regeneration ( Tomback and Resler, 200 7) Whitebark pine is less shade tolerant than the species that it shelters (Arno 1986). Fire exclusion practices in the 20 th century have led to successional replacement of whitebark pine by subalpine fir and Engelmann spruce (Ar no 1986). This has changed the structure of subalpine forests because conifer biodiversity is lost and the landscape becomes homogenous. Many stands at a landscape level are now solely comprised of late seral stage subalpine fir and En gelmann spruce (Kea ne, 2001). Mountain pine beetles ( Dendroctonus ponderosae ) are native to western North America. These beetles episodically attack large, mature pines with thick bark resulting in major outbreaks (Cole and Amman, 1969). They naturally occur in lodgepo le pine forests, but during outbreak s the insects spread to whitebark pine communities (Arno,

PAGE 19

7 1986). This can result in wide scale mortality. In the 20 th century, mountain pine beetle outbreak s killed many mature whitebark pine trees in Idaho and Montana (Bartos and Gibson, 1990 ; Jenkins et al 2001). Because fire suppression results in higher density and greater age of late successional forests, this practice may increase the scale and abundance of mountain pine beetle outbreaks (McGregor and Cole, 198 5). Climatic warming facilitates pine beetle population growth and may reduce whitebark pine defenses (Raffa et al 2008). Currently, mountain pine beetles are again in outbreak mode throughout the West but at a geographic scale considered unprecedente d (Logan et al., 2010) This outbreak is driven by milder winter temperatures ( Logan and Powell 2001 ; Logan et al., 2010 ) Whitebark pine stressed by competition from fire suppression may be even more vulnerable to pine beetle attacks generated by warmin g trends ( Logan et al., 2010; Raffa et al 2008). A third cause of decline in whitebark pine is white pine blister rust, a disease caused by the exotic pathogen Cronartium ribicola This fungal pathogen, which infects five needle white pines of subgenu s Strobus (McDonald and Hoff, 2001) was inadvertently introduced to western importation of infected nursery seedlings from western Europe (Spaulding, 1909, 1911, 1922 as cited in McDonald and Hoff, 2001). Cronart ium ribicola has evolved with Eurasian pine species which have resistance to this disease S ince its introduction to North America C. ribicola has exploited a range of host pine spec ies with low natural resistance (McDonald and Hoff, 2001) (Table I.1 ) Although many North American five needle pines are susceptible to infection by C. ribicola they vary in susceptibility and

PAGE 20

8 extent Whitebark pine populations are currently being infected by C. ribicola nearly range wide with high infection levels in so me areas; the resulting mortality is impacting ecosystem s throughout range (Tomback and Achuff, 2010). Cronartium ribicola relies on both five needled white pines and alternate hosts to com plete its life cycle (McDonald and Hoff, 2001). Ribes spp., the gooseberries and curr a nts, have long been recognized as alternate hosts, but recent research has discovered that herbaceous plants in the genera Pedicularis and Castilleja may also act as hosts (McDonald et al 2006). McDonald and Hoff ( 2001) describe the specific mechanism of white pine blister rust transmission. Fiv e needle white pines are infecte d by C. ribicola when wind blown basidiospores from alternate hosts enter the stomata of pine needles. Rust mycelia grow from the needle int o the living wood of the pine tree and eventually produce a fruiting canker which leads to swellings on branches or stems of the tree (Fig ure II. 2). The canker sporulates, producing sacs of aeciospores, and these sacs ultimately burst through the surface of seemingly healthy bark to release spores into the environment. Some spores inevitably reach an alternate host and complete the cycle. Cankers eventually girdle the branch or stem of the pine, cutting off the supply of water and nutrients. Since seed cones are produced at branch tips, the accumulating dead branches reduce seed cone production long before the tree itself dies. Blister rust mortality is especially detrimental to whitebark pin e which requires up to 50 years to reach re productive maturi ty (McCaughey and Schmidt, 1990) L oss of ma ture trees means a loss of cone production that potentially takes decades to replace. However, s eedling s sapling s and smaller krummholz whitebark pine are also affected by

PAGE 21

9 C. ribicola and die more rapidly fr om infection than their larger counterparts (Tomback et al., 1995). This reduces the number of young trees available to regenerate the species. Krummholz whitebark pine s are also affected by white pine blister rust. It was once thought that this pathoge n could not survive the extreme winter temperatures of the alp ine treeline ecotone (Campbell and Antos, 2000), but high numbers of infected individuals have recently been discovered among krummholz whitebark pine (Resler and Tomback, 2008), suggesting that C. ribicola can reproduce and survive under the most extreme conditions White pine blister ru st may affect the keystone and foundational roles that whi tebark pine plays within the alpine treeline ecotone Resler and Tomback (2008) discovered that 33.7 % of the whitebark pine in their sampled tree islands showed evidence of white pine blister rust infection. This has serious implications because global climate change is predicted to alter treeline dynamics (Tomback and Resler, 2007) If whitebark pine s in the alpine treeline ecotone are succumbing to blister rust at significant levels, it may affect the way treeline is able to respond to climate warming and potentially rising treeline elevations (Tomba ck and Resler, 2007; Resler and Tomback, 2008). Glo bal Climate Change and Treeline Impacts Treeline forests are indicators of global climate change. These so called 2009). Treeline is known to be dependent on several fa ctors, including temperature, wind speeds, nitrogen deposition, and concentration of carb on dioxide (Grace et al, 2002). Treelines have responded to temperature fluctuations since the last glacial maximum

PAGE 22

10 (Lloyd and Graumlich, 1997). During the early Hol ocene, temperatures were 1.4 C greater than the present, and treelines were on average 200 m higher than they are today (Grace et al 2002). Average temperatures are conservatively predicted to increase by more than 2.5 C over the next century (Easterl ing, 2005). Warmer temperatures are expected to cause an upward shift in treeline (Millar et al 2004), with an estimated elevation gain of 140 700 m (Grace et al 2002). Climate models have been generated to predict how whitebark pine will respond to warming trends (Hamann and Wang 2006; McKenney et al 2007; Warwell et al 2007). These models are generally in agreement that whitebark pine will see a shift from its current range. These models do not account for fine scale habitat features or ec ological processes such as topography, soil nutrients, seed dispersal and germination, or disturbance regimes (Loehle, 199 6). However, they do provide a coarse estimate of changes in range and habitat area. Some models show that while whitebark pine wil l lose current distribution in the U.S., it will gain new habitat at higher latitudes and elevations. Warwell et al (2007) predict a 97% loss of suitable whitebark habitat in the U.S. by bitat will be lost by 2085, but it should gain 76% of the original area at northern latitudes. McKenney et al ( 2007 ) predicted reduced by 42%. However, whitebark pine will gain an exp ected 7.8% new habitat by moving north approximately 6.4 With global climate change shifting treeline to higher respond and maintain its ecological roles due to stresses brought on by blister rust and other threats

PAGE 23

11 such as mountain pine beetle. If whitebark pine mortality is widespread the ability of treeline as a whole to move upwards could be compromised. ebark pine by caching seeds in the alpine tundra and treeline ecotone (Tomback, 1998 ; Tomback, 2001 ). Frequently, seeds are cached next to microsites such as rocks or ground topography which potentially sheltering developing whitebark seedlings (Resler, 2004). Because whitebark pine then in turn creates favorable microsites for less hardy conifers facilitating tree island development, warming trends should result in krummholz tree islands shifting upwar ds in elevation (Resler et al, 2005). With blister rust killing whitebark pine at treeline, the number of favorable whitebark pine microsites available to subalpine fir or Engelmann spruce is reduced. This may affect the ability of treeline to move upwar ds in the manner predicted (Tomback and Resler, 2007; Resler and Tomback, 2008). ecological functions the better we can predict how ecosystems will respond to whitebark pine mortality. Presently, very little is known about the mechanisms behind facilitation roles whitebark pine plays in the alpine treeline ecotone or how whitebark pine ecosystems will respond to increased blister rust mortality and global warming trends. Our research will contribu te to a bet ter understanding the role of whitebark pine in two study areas in Montana facilitates tree island initiation and maintenance within the alpine treeline ecotone.

PAGE 24

12 Figures and Tables Figure I. 1 Distribution of Pinus albicaulis in North America (Tomback and Achuff, 2010 )

PAGE 25

13 Figure I. 2 Image of active blister rust stem canker on an infected whitebark pine. Light colored sec tions on stems are sacs containing aeciospores (Photo by Sarah B lakeslee). Table I 1 US and Canadian conifers susceptible to white pine blister rust infection (McDonald and Hoff, 2001) North American Tree Hosts Whitebark Pine ( Pinus albicaulis) Foxtail Pine (Pinus b alfouriana) Rocky Mountain Bri stlec one Pine (Pinus aristata) Great Basin Bristlec one Pine (Pinus longaeva) Southwester n white pine ( Pinus strobiformis ) Limber Pine ( Pinus flexilis ) Eastern White Pine (Pinus strobus) Western White Pine (Pinus monticola) S ugar Pine (Pinus lambertiana)

PAGE 26

14 CHAPTER II. INTRODUCTION Background Certain p lant species may act as keystone and foundational ecosystem components by facilitating stability and biodiversity (Ellison et al., 2005). Research has indicated the importance of facilitative plant interactions for survival and regeneration in stressful environments (Lortie et al, 2004; Brooker et al, 2008). This is particularly true for high elevation sites where abiotic stress is high ( Callaway et al 2002). Calloway et al ( 2002) examined 115 plant species in 11 mountain sites across the globe to determine whether elevation, and thus environmental stress, changed plant community interactions. They generally observed competitive interactions at lower elevations where envir onmental conditions were moderate. At higher elevations, species interactions predominantly switched to facilitation whereby one competitor provided shelter for another. H arsh environmental conditions may be moderated when facilitative species pro vide a protective microclimat e for germination and establish ment e.g., protection from solar radiation or shelter from wind ( Germino et. al., 2002; Baumeister and Callaway 2006 ). As more e xamples of plant facilitation in high elevation communiti es are discov ered, it is increasingly apparent that facilitation is important to community development in these extreme environments For example, Cavieres et al. (2005, 2007) observed plant interactions at the upper limit of vegetation in the Chilean Andes. Their st udies indicate that a cushion plant ( Azorella monantha) moderates substrate and air temperatures and enhances soil moisture and nutrient s for both the native Andean

PAGE 27

15 cauliflower (Nastanthus agglomerates) and the invasive field chickweed (Cerastium arvense). Batllori et al. (2009) found that survival of P inus uncinata seedlings planted in the alpine treeline ecotone was increased when seedlings were located on the leeward side of krummholz conifers likely due to retention of sheltering snowpack in this loca tion during winter months. The process of seedling establishment is important for long lived plants such as conifers. Years with successful seed germination are more frequent than years with both high seed germination and high seedling survival (Cui and S mith, 1991). Seedling establishment is particularly difficult in the alpine treeline ecotone because high winds, variable temperatures, poorly developed soils, and intense solar radiation ( Marr, 1977; Arno and Hammerly, 1984; Finklin, 1986 ; Maher et al 2005 ) make establishment a challenge. The likelihood of s eedling survival is improved in the alpine treeline ecotone when h arsh climat ic conditions are mitigated by rocks, topographic niches, and other objects act ing as protective microsit es providing windward shelter (Germino 2002 ; Resler, 2004 ; Batllori et al., 2009 ) Survival is further fac ilitated when a solitary conifer establishes and other conifers grow in its lee, resulting in two or more conifers growing together in close pr oximity as a multi tree tree island. In the alpine treeline ecotone tree islands facilitate the survival of conifers species such as Engelmann spruce ( Picea engelmannii ) and subalpine fir ( Abies lasiocarpa ) (Resler and Tomback, 2008 ). Both species are l ess likely to be found on the windward rather than leeward side of tree islands (Hat tenschwiler and Smith, 1999), and E ngelmann spruce seedlings have

PAGE 28

16 been found to be associated with higher survival rates when windward or overhead shelter such as branches is present ( Germino et al 2002) In the cool, dry, and windy north erly eastern slope faces of the alpine treeline ecotone on the eastern Rocky Mountain front, whitebark pine (Pinus albicaulis) is a dominant ecosystem component (Smith et al 201 1 ; Re sler and Tomback, 2008) This (Nucifraga columbiana) (Tomback, 1982 ) and is tolerant of drought and high levels of solar radiation ( Arno and Hammerly, 1984; Maher et al 2005 ; Tomback et al., 2001 ). Because wh itebark pine often grows as a solitary conifer (Mah er et al, 2005 ; Resler and Tomback, 2008 ), it may facilitating the survival of less hardy subalpine fir and Engelmann spruce on harsh sites in the subalpine zone and in the alpine treeline ecotone, where it facilitates development of multi tree tree isla nd s ( Callaway, 1998 ; Resler and Tomback, 2008 ). Resler and Tomback (2008) found that within two study areas east of the Continental Divide, 95.9% of multi tree tree islands sampled included whitebark pine Of these tree the importance of whitebark pine in facilitating the establishment of leeward conifers in this region. Whitebark pine is presently designated a candidate for endangered species listing by the U S Fish and Wildlife Service (USFWS, 2011 ). Fire suppression leading to successional replacement by shade tolerant conifers mountain pine beetle ( Dendroctonus ponderosae) outbreaks, and the disease whit e pine blister rust caused by the invasive fungal pathogen Cronartium ribicola are the major factors in the decline of whitebark pine Within the alpine treeline ecotone the most immediate threat to whitebark pine populations is damage and mortality re sulting from white pine blister rust Infected small

PAGE 29

17 diameter trees can progress from showing no outward disease symptoms to death within a few years ( Tomback et al., 1995 ) This rapid mortality may reduce the chances for establishment of new tree island s and also the health of existing tree islands by exposing formerly leeward conifers to the wind. G lobal climate change may increase the frequency and severity of fire regimes, accelerate the rate and spread of pine beetle outbreaks, and likely alter t he distribution and infection rates of blister rust making it challenging to predict how whitebark ec osystems will respond (Tomback and Achuff, 2010). Warmer temperatures are also expected to cause an upward shift in treeline (Millar et al 2004), with an estimated elevation gain of 140 700 m (Grace et al 2002). As whitebark pine numbers decline, this may impact the frequency of tree island establishment in the upper alpine treeline ecotone possibly altering the response of treeli ne as a whole to w arming trends (Tomback and Resler 2007) Conceptual Framework Although research has indicated that whitebark pine is an important component of tree island s very little is known about the specific mechanisms of facilitation leading to tree island formatio The overall objective s for this study are to determine empirically and experimentally the attributes and ecological interactions that enable whitebark pine to facilitate tree island development and to address how the mortality of whitebark pine from blister rust may impact these ecosystem functions.

PAGE 30

18 Hypotheses for Testing This is an NSF supported project with objectives already articulated. For my designed experiments, developed field protocols, and had oversight responsibilitiy for a series of experiments and empirical studies. I had the support of Dr. Diana F. Tomback and other students in the lab and I am testing three separate hypotheses that together support the overall research objective in a logical sequence (Fig ure II.1) My hypotheses are as follows: 1 ) W hitebark pin e is hardier than other alpine treeline ecotone conifer species as demonstrate d by more vigorous growth and higher survival at treeline; 2) whitebark pine provides a more favorable microsite for tree island recruitme nt than other common alpine treeline ecotone microsites; and 3) death of windward whitebark pine in established tree i slands leads to loss of vigor in leeward conifers. The first hypothesis addr esses whether there is differential growth vigor if any, of whitebark pine in comparison to other treeline conifers. Because tree islands usually dward tree is often the oldest ( Holtmeier and Broll, 1992 ). The domi nant presence of whitebark pine in this position suggests that this species is hardy and serves an important role in recruiting tree islands through subsequent facilitation by mitigating conditions for a leeward conifer In order to establish a tree island, whitebark pine seed lings must first become established in the harsh and exposed areas within the alpine treeline ecotone and then survive these conditions This first hypothesis is a ddressed through an empirical study that identifies and compare s di fferences in survival and vigor among small krummholz whitebark pine and the other two dominant treeline conifer species, subalpine fir and

PAGE 31

19 Engelmann spruce In Fig ure II 1, the importance of the initial establishment and survival of whitebark pine to subsequent facilitation is illustrated in a visual model. The second hypothesis sta tes that because whitebark pine often facilitate s tree island development, it may provide more favorable growing conditions for leeward conifers than other common treeline microsites. An alternative to this hy pothesis is that whitebark pine trees are simply more numerous at treeline, and this means that there are more opportunities fo r tree islands to form i n whitebark pine microsites. In the whether whitebark microsites are associated with higher conifer germination and /or seedling survival rates than other common treeline microsites in order to d etermine whether conditions are in fact more favorable. This would imply that whitebark pine may offer facilitation or a higher quality of facilitation than other microsites. The o ther alpine treeline ecotone microsites investigated include rocks, anothe r conifer Engelmann spruce for consistency and exposed si tes with no apparent shelter. In this study, w e planted conifer seeds and seedlings leeward of four microsite types and compared seed germination and seedling survival rates Th is mechanism for tree island development is represented in the #2 position of Figure II 1. The last hypothesis directly tests the facilitation function of whitebark pine in established tree islands Since whitebark pine is often the intiating or most windward conifer wit hin a muilti tr ee island, it may provide important leeward shelter to other conifer species mitigating the harsh wind and particle blast of treeline environments ( Habeck, 1969; R esler, 2004). Blister rust is currently infecting and killing many whitebark pine in some regions We hypothesi ze that the loss of these windward whitebark pines will result in exposure and thus damage to leeward conifers. As

PAGE 32

20 indicated in the #3 position of Figure II. 1, t he the effects of blister rust o n windward whitebark pine and we monitored the growth and vigor of t he non whitebark conifer immediately leeward.

PAGE 33

21 Figures and Tables Figure II. 1 Overall conceptual m odel The role of whitebark pine in tree island formation can be explained by 1 the establishment of a solitary conifer in an exposed area without shelter from other tree islands. T o accomplish this, the species must be hardy and vigorous to withstand the harsh treeline climate. We tes t this hypothesis in the Relative Vigor Study. As the conifer establishes and grows it generates a sheltering leeward microsite (indicated by star) 2 where other conifers can germinate, eventually leading to the formation of a tree island. W e examine this hypothesized process by testing w h ether whitebark pine microsites are associated with the highest conifer germination and survival rates in the Planting and Sowing Study. Blister rust is currently killing whitebark pine at treeline, potentially expos ing leeward conifers to harsh wind and ice particle s The impact o f this windward shelter loss, 3 on establi shed tree islands is unknown. We simulate blister rust on windward whitebark pine (x indicates blister rust simulation) in the Girdling Study and monitor impacts on the newly exposed leeward conifer.

PAGE 34

22 CHAPTER III METHODS Study Areas This research was conducted over three field seasons within a bioclimatically induced krummholz tr eeline at two separate study areas (Figure III.1 ). These study areas were selected because whitebark pine is a major ecosystem component and also because of the accessibility of treeline The northern study area include s Divide and Whitecalf Mountains, Montana. Divide Mountain is located on both Blackfeet Trib al Land, as well as on the east slope (Rocky Mountain eastern front) of Glacier National Park at approximately 48 39' 25" N and 113 23' 45" W. Treeline occurs a t approximately 2200 m elevation. Whitecalf Mountain is located within the east slope of Gla cier National Park at 48 38' 20" N and 113 24' 08" W. Treeline occurs a t approximately 2100 m elevation. Divide and Whitecalf Mountains are characterized by steep slopes ( = 25.7 ) and poorly developed soils with limestone bed rock. Mountain avens ( D ryas octopetala) and bearberry ( Arctostaphylos uva ursi) are the dominant herbaceous understory vegetation in this study area Willows ( Salix sp p. ) and junipers ( Juniperus spp. ) are also distributed in patches. Subalpine fir, Engelmann spruce, and whiteba rk pine are the dominant trees on Divide Mountain, but there is a noticeable absence of Engelmann spruce on Whitecalf Mountain. Found in small numbers are limber pine ( Pinus flexilis ), lodgepole pine ( Pinus contorta ), and Douglas fir ( Pseudotsuga menziesi i ). The southern study area is located 530 kilometers directly southeast of the northern study area (RNA)

PAGE 35

23 MT, in Custer National Forest at 45 01' 47.45" N and 109 24' 09.22" W. Subalpine fir is nearly absent from the ecosystem and is only commonly found in the shelter of willow patches or within dense tree islands Engelmann spr uce and whitebark pine are the two most common species. Lodgepole and limber pine a re found in very small numbers. Krummholz rapidly grades into erect trees on the northeast facing slope. Like the Divide Mountain and White Calf study areas, the Line Creek RNA is characterized by many solitary krummholz whitebark pine The terrain is largely open with gentle slopes (~ 20) S edges ( Carex sp p .), mountain meadow cinquefoil ( Potentilla diversifolia.), American bistort (Polygonum bistortoides ), and silvery lupine ( Lupinus argenteus ) are common among the herbaceous groundcover. Relative Vigor Study Two studies were conduc ted to examine relative vigor in different way s One was an observational study addressing our first hypothesis by comparing shoot lengths, shoot growth rates, and vigor of the three most common treeline species in our study areas : whitebark pine subalpi ne fir and Engelmann spruce For the first study we used measurements representing readily obtainable characteristics of growth and vigor. We originally intended to use s eedlings for each species, but w hen selecting trees for this study we found very f ew seedlings at either study area I nstead we selected small krummholz trees that were already established. These trees ranged roughly from 1 to 40 years of age, based on stem constriction counts (Appendix I, Tables 1 and 2). These small trees were unli kely to die during the short term three year observation period. Our measurements act as a surrogate for survival and are also important clues to any species differences in ability to utilize resources from the environment. Shoot lengths are

PAGE 36

24 particularly relevant to vigor and survival. Ability to increase photosynthetic canopy rapidly is important at treeline, where growing seasons many only last a few months in the summer. The second study was a random sampling effort to determine the relative numbers of solitary krummholz trees of each species in our study area. Field Methods. Th ese studies were conducted both on Divide Mountain and at the Line Creek RNA. In July 2010, 100 smal l solitary krummholz conifers (<30 cm high ) of the three major conifer sp ecies were haphazardly selected based on exposed growing conditions i.e., unsheltered by a tree island or ot her large microsites (Table III. 2 ). We marked conifers with a tagged leeward nailspike and monitored them fro m July 2010 to September 2012. S tem d iameter at ground level length from 3 5 shoots (defined as the total length of the new branch elongation plus extending needles ) canopy area (calculated using longest dimension and the dimension immediately perpendicular to longest dimension) and woody tissue height were measured annually in July. The only exceptions were shoot length measurements. In 2011 and 2012 we measured five shoots if available, from each tree, first in July near the beginning of the growing season and then again in September a fter shoots were fully extended. T he difference between the July and September mean shoot length per tree was used to calculate the shoot growth rate s with the following formula: (September Mean Shoot July Mean Shoot)/No. Days between Measurements for e ach conifer (mm/day) Canopy areas were calculated using the formula for the area of an ellipse ( a b ) All tree geographic locations were marked with a Trimble Geo XT handheld GPS unit ( GeoExplorer 2008 series ) Measuring tapes were used to measure height and canopy area to the nearest half centimeter. Mitutoyo (500 195 20) d igital c alipers w ith a

PAGE 37

25 precision of a hundredth of a millimeter were used to measure shoot lengths. Care was taken to ensure consistency of re measurement from year to year. In 2011, we marked f ive branches on each conifer with zip ties so that the same terminal branch sh oots were re measured Each conifer stem was marked with tree specific paint to ensure consistency in caliper placement for stem diameter measurements. The shoot lengths of subalpine forest conifers which are taller trees with large diameters and upright growth forms, were measured for comparison with krummholz conifers to see how shoots growing in less harsh conditions might differ In both study areas, larger statur e trees occurred in sheltered sites a nd at the lower limit of the alpine treeline ecoton e In 2011, five shoots each of 10 haphazardly selected conifers of each species at each study area were measur ed in September. In 2012, we increased the sample size to 20 conifers of each species at each study area. The measurement procedure was identi cal to that of the small krummholz trees except tree branches were not marked so the same trees were not necessarily revisited from year to year In order to determine the relative abundance and density of solitary krummholz whitebark pine, subalpine f ir, and Engelmann spruce, growing at treeline, transects 50 m long transects with 10 m wide belts ( 500 m 2 ) were established using a subset of ra n domly generated GIS points within each study areas We sampled t wenty transects at each study area. On each tr ansect, solitary k rummholz conifers growing in largely unsheltered conditions were measured for species, height, canopy area, microsite, and qualitative vigor. Qualitative vigor was determined using a four category ranking scale of poor to excellent based on characteristics of windward needle death, health of new annual s hoots, and needle color (Table III.1 ).

PAGE 38

26 Data Analysis All data analyses were completed using R ( GUI statistical software program version 2.11.1 ) We compared krummholz shoot lengths amon g species using a Kruskal Wal lis rank sum test ( because data were not normal ly distributed and sample sizes were unequal) Because of an unequal number of shoots per tree resulting from small tree size and shoot death on some marked branches, one shoot per tree was selec ted for this analysis using randomly generated numbers Wilcox signed r ank post hoc tests with Bonferroni corrections were used for paired comparisons of shoot lengths between species Our shoot measurements do not separately account for branch extension length and needle length, but rather the sum of both. We were interested in seeing if species differences remained when needle length was subtracted from our measured shoo t lengths. This would compensate for any length attributed to the needle extending beyond the branch extension point. We obtained minimum needle lengths for each of the three species from Flora of North America (eFloras, 2008) and subtracted them from the mean shoot lengths of the upright upper subalpine conifers, as follows: Engelmann spruce 16 mm, subalpine fir 18 mm, and whitebark pine 30 mm. The reduced shoot lengths were then compared b y species in a series of Wilcox signed rank tests with Bonfe rroni corrections by year and study area. Nonparametric statistics were used because data were not normally distribut ed Krummholz conifer shoots were not assessed in this manner because their needle lengths were generally shorter than the mini mum needle lengths suggested by Flora of North America. This analysis would have resulted in negative shoot lengths for krummholz individuals

PAGE 39

27 Other measurements analyzed with Kruskal Wallis rank sum test s and Wilcox on signed Rank post hoc test s with Bo nferroni correction s were krummholz shoot growth rates, krummholz canopy area, krummholz height, krummholz stem diameter, and qualitative vigor of trees sampled in the transects Non parametric statistics were used because of unequal sample sizes and data that were not normally distributed. In order to use the mean shoot for each individual upright tree, we verified that the within tree variation in shoot lengths was smaller than among species variation as follows: The variation of the five measured u pright subal pine conifer shoot s was compared within and among species with a Nested ANOVA. Mean shoot length was calcula ted for each tree and compared with a One test to determine species differences. Parametric stat istics were used be cause data were normally distributed and sample sizes were equal. Significance level was set We used a Chi Square test to analyze differences in microsites responsible for establishment for the conifers sampled within the transec ts. In this comparison there were two categories for microsite : t he proportions of conifers found in relatively unsheltered, unknown, or small microtopographic ground depression microsites were compared to the proportions of conifers found established lee ward of more substantial shelter, such as rocks or vegetation. These proportions were also compared between species at each study area. We used a Binomial Distribution test to determine the proportional occurrenc e of whitebark pine in relation to the number of all so li tary conifers sampled for the transects at each study area. This test was performed for each of the 20 transects at each study area. The theoretical probability was considered to be an equal proportion of

PAGE 40

28 each species ( 0.33 ) We determined how many of the 20 transects at each study area had solitary whitebark pine in a significantly greater proportion than expected. The mean density of each species w as also calculated as the number of conifers per square meter for each transect These densities were compared visually in a figure. Planting and Sowing Study This study addresses hypothesis 2 by comparing the survival rates of planted seedlings and germi nation and survival rates of sown seeds leeward of fou r common treeline microsites. Field Methods Study areas were Divide M ountain and the Line Creek RNA. For the Divide M ountain seed sowing and seedling planting study, we collected Engelmann spruce a nd subalpine fir cones at Divi de Mountain in September, 2010. The Colorado State Forest Service Nursery found that the Eng elmann spruce cones contained non viable seeds, so only subalpine fir seeds were available for the study The quantity of subal pine fir seeds was adequate only to produce enough seedlings for the seedling planting component of the study. Engelmann spruce seeds from the same seed transfer zone ( McDonald Pass Helena National Forest 6300 m ele v ation) were provided by the U SDA Forest Se rvice N ursery in Coeur d'Alene, ID. On the entire Beartooth P lateau there was no cone crop for either spruce or fir in 2010. Engelmann spruce seeds collected by the Dubois Ranger District of th e Shoshone National Forest were provided by the USDA Forest Service Nursery in Bessey, South Dakota. These seeds were collected at an elevation of 2712 m and were the highest elevation Engelmann spruce seeds available in the same seed transfer zone as the

PAGE 41

29 Beartooth RNA. These seeds were used for both the direct seed sowing experiment and the seedling planting experiments Seeds for the direct sowing component of the study were chilled for 4 months at approximately 35 before planting. Seedlings were grown by the Colorado State Forest Service Nursery in Fort C ollins, CO. Planting and sowing took place in Ju ly 2011. In each study area, we located 20 replicates each of the four microsites krummholz whitebark pine, krummholz spruce, rock, and exposed site for the seedling planting study, and at an additiona l 20 replicates of each microsite for the seed sowing study Microsite geographic locations were marked with a Trimble Geo XT handheld GPS unit ( GeoExplorer 2008 series ) Microsites ranged in height from 4 66 cm In general, conifer microsites were t aller than rocks. For each study, spruce and whitebark microsites were selected based on similar heights (Appendix III). Once suitable microsi tes were marked with a numbered nailspike, either 5 seeds or 2 seedlings were placed immediately leeward of the microsite object or in the middle of the exposed microsite Leeward direction was determined by observing dominant flagging of krummholz conifers. Seedlings were labeled with colored zip ties pla ced at the base of the stems for 2012 identification and pla nted in 25 cm h oles dug to fit th e container substrate (Figure III.2) Seed s were planted 0.5 cm deep. Seedling sites received 1 liter of water at the time of planting and seed sites received liter of water at the time of planting. Germination and su rvival were assessed in July 2012. At this time, terminal shoot lengths were measured and qualitative vigor on a poor to excellent four level categorical scale assessed for surviving seedlings Germinated seed s were located

PAGE 42

30 and coun ted by microsite type (Figure III.3 ). Germination sites were revisited September 2012 to document germinant survival over the summer months Data Analysis The results of this study were analyzed separately by planting type, microsite type, and stud y area. Data that were not normally distributed were analyzed with non parametric statistics. S ignificance levels were set The resulting analyses included one year seedling survival rates and seedling vigor assessments, and seed germination and new seedling summer survival for both Divide Mountain and Line Creek RNA. One year seedling survival was examined with squared test of independence to assess whether survival differe d among microsite types. Qualitative squared test to determine differences among microsite types. Terminal shoot lengths were com pared among microsite types with a Kruskal Wal lis rank sum test July 2012 seed germination among microsite type s was compared with a E xact p robability test September 2012 summer survival numbers of these initial germinants were also E xact probability test to examine survival associated with t he different microsite types. I results were statistic ally significant, we compared the observed survival values to Chi square expected survival values to determine which microsite type(s) contributed the most towards statistical significance. Lastly, the relative risk of survival or death by microsite type was calculated using an odds ration calculation comparing actual germinant summer survival numbers to expected values A ratio near 1.0 was interpreted as relative risk of survival or death near expected.

PAGE 43

31 Girdling Study T his study addresses the third h ypothesis by simulating whitebark pine mortality caused by blister rus t infection. Field Methods The girdling study was conducted on both Divide and Whitecalf Mountains In July 2010, we selected tree islands with whitebark pine as the windward species and placed them into either control or experimental groups. In order for a site to be classified as experimental, the windward whitebark tree had to be infected with blister rust. This wa s a condition of the research permits issued by both Glacier Natio nal Park and the Blackfeet Indian Reservation. There were a total of 44 sites in this study, with an equal number of control and experimental sites. All site locations were marked with a Trimble Geo XT handheld GPS unit ( GeoExplorer 2008 series) At eac h site, the conifer species immediately leeward of the whitebark pine was in most cases subalpine fir (n = 40) with the remaining sites Engelmann spruce We collected b aseline measurements of height, vigor, shoot lengths and canopy on both the windward w hitebark pine and the immediately leeward conifer. Heights were measured to the nearest half centimeter with a metric tape measure. Shoot lengths were measured using Mitutoyo (500 195 20) digital calipers with a hundredth of a millimeter precision. Afte r baseline measurements were collected, the whitebark pine at all experimental sites was defoliated and girdled (Figure III.4 ) This was accomplished by manually removing all foliage from the tree and sawing deep grooves completely around the trunk to ens ure no future growth. This simulated the effects of blister rust infection, which reduces small trees to branch and stem skeletons in a short time period (Tomback et al 1995). In sites that had extensive layering and tree islands larger than the windwa rd whitebark pine only

PAGE 44

32 the discrete section which was she ltered by the whitebark pine was used for measurements. In 2011 and 2012 we monitored the effects of girdling on the conifer immediately leeward of the whitebark pine. Subsequent measureme nts w ere taken to determine whether exposure on the experimental sites impacted the leeward conifer measurements of length of new shoots, shoot mortality, and qualitative tree vigor (Table III.1), differently than control sites. In July 2011, z ip t ies were plac ed on the branch es immediately leeward of the whitebark in both the control and experimental sites for repeated measurements. At this time, we also marked shoots from fully wind exposed subalpine fir or Engelmann spruce conifers in the same tree island th at were not associated with our experiment al or control conifer (Figure III .5 ). These trees represent a natural measure of exposure to windward conditions in the same tree island and are of the same leeward species we are investigating. Data Analysis. T he difference in 2010 and 2012 vigor of the leeward conifer was determined and categorically ranked as follows: loss of vigor, no change in vigor, or increase in vigor. C ategories were compared between control and experimental site leeward conifers with a Test. If this test was statistically significant, t he categories contributing most to significance were determined using observed values vs. Chi square expected values. The changes in leeward conifer shoot lengths over the 201 0 2012 time period were compared in a Before After Control Impact (BACI) analysis. One measured shoot length per leeward conifer per year was randomly selected for comparison between control and experimental site groups.

PAGE 45

33 We also compared shoot length and mortality by treatment type at a tree island level. We paired the natural, wind exposed shoots to the shoots experimentally exposed by girdling or sheltered by a control whitebark from the same tree island. We determined differences between these pai rings for 2011 and 2012 shoot lengths and also the proportion of dead shoots over the 2011 2012 timeframe. One shoot was randomly selected from the five sampled per tree. The shoot length differences were compared in a BACI analysis, and shoot mortality differences were examined with a Wilcox Signed rank t est with a Bonferroni correction to determine whether experimental leeward conifers had shorter shoot lengths and higher shoot mortality than control conifers

PAGE 46

34 Figures and Tables Figu re III. 1 Research study a reas Divide and Whitecalf Mountai ns MT, in east Glacier NP and Blackfeet Reservation (48 39' 25" N, 113 F orest, MT (113 24' 09" W) Basemap from Montana Government Natural Resources: http://nris.mt.gov/gis/gisdatalib/mtmaps.aspx

PAGE 47

35 Table III 1 Qualitative vigor c ategories This table demonstrates the characteristics responsible for classifying conifers into a particular vigor category. A conifer is assigned to a category based on meeting the majority of specified criteria. Exce llent Good Fair Poor Windward Appearance Tree may be flagged, but no obvious windward damage Minimal windward damage, but only on a few branches Most windward branches are damaged to some extent Tree is extensively flagged with lots of windward die off Needle Health and Color Needles are lon g & numerous ; color is characteristic of a healthy specimen by species (i e dark blue green for whitebark) Needles generally healthy, but may have slightly yellowish color due to drought conditions Some needles have been blasted and/or yellow due to drought Red or brown dying and dead needles are numerous New Shoot Status Numerous new shoots throughout entire tree; shoots are fully developing and healthy Many shoots present, but some may be underdeveloped New shoots developed, but were blasted and/or are underdeveloped New shoots generally absent from branch tips Table III 2 Sample sizes of small krummholz trees in the relative vigor study Whitebark Pine Subalpine Fir Engelmann Spruce Divide Mountain 17 15 15 Line Creek RNA 21 12 20

PAGE 48

36 Figure III 2 Planted seedlings at the Line Creek RNA N ursery grown Engelmann spruce seedlings planted on the leeward side of the four experimental microsites at the Line Cr eek RNA. Microsites are as follows: a. whitebark pine, b. Engelmann spruce, c. rock, and d. open or unprotected. These images were taken in July 2011 at the time of planting. (Photo credits Sarah Blakeslee) a. b. c. d.

PAGE 49

37 Figure III 3 Germinated seeds on Divide Mou ntain Representative Engelmann spruce seedling cluster on Divide Mountain showing germination in three out of five sown seeds; this image was taken in July 2012 shortly after germination and early in the treeline growing season. (Photo credit Sarah Blakesl ee)

PAGE 50

38 a. b. Figure III 4 Example of before and after girdling and defoliation treatment Representative girdling treatment site. In image a, the whitebark pine is sheltering the windward edge of the tree island. Image b shows the subsequent exposur e after sawing through the main stem and defoliating the tree. Leeward conifer measurements were taken in the areas previously sheltered by the whitebark pine. (Photo Credit: Sarah Blakeslee)

PAGE 51

39 Figure III 5 Leeward shoot vs exposed shoot sampling areas This image shows representative sampling locations for leeward shoot and exposed shoot measurements in the gird ling study. The leeward shoots were measured on the subalpine fir immediately leeward of the whitebark pine. Exposed shoots were measured on a conifer of the same species located elsewhere on the same tree island but without windward whitebark protection. The exposed shoots represent natural exposure and serve as a baseline comparison to the initially protected leeward shoots. This procedure w as done for control as wel l as experimental tree islands. (Photo Credit: Sarah Blakeslee)

PAGE 52

40 CHAPTER I V. RESULTS Relative Vigor Study Transects On Divide Mountain, 487 solitary, wind exposed (unsheltered) krummholz conifers were sampled within 20 transect s. Species composition comprised 64% whitebark pine (n = 312), 23% subalpine fir (n = 111), and 13% Engelmann spruce (n = 64). At the Line Creek RNA, 209 solitary exposed krummholz conifers were sampled. We found species composition to be 83% whitebark pine (n = 174), 15% Engelmann spruce (n = 32), and 1.4% subalpine fir (n = 3). Binomial test s of individual transects indicated that sampled whitebark p ine was present at statistically significant higher abundance s than expected at both Divide Mountain (79 % transects with solitary trees n = 15/19) and the Line Creek RNA (80% transects with solitary trees n = 12/15) (Table IV.1) Transects with no solitary trees present were not included in these analyses. Based on transect data for solitary trees, t ree density per square meter was calculated by species for each study area (Figure IV.1). Divide Mountain had greate r densities for all three species than Line Creek RNA. At both study areas whitebark densities were the highest of the three species (Divide = 0.031 0.03 trees/m 2 ; Line Creek RNA = 0.017 0.02 trees/m 2 ) Whitebark pine densities with respect to other species were as follows: Divide Mountain 5 times spruce and 3 times fir ; Line Creek RNA 5 .5 times spruce and 58 times fir.

PAGE 53

41 Differences in q ualitative vigor trends between species were found on Divide Mountain (Kruskal 2 = 18.9, df = 2, P = 7.8e 5). Whitebark pine had higher vigor than both fir (W = 19557, P = 0.037) and spruce ( W = 13026.5, P = 2.08e 5). Fir vigor was highe r than spruce (W = 2981.5, P = 0.047). Statistical differences in species vigor were not observed at the Line Creek RNA (Kruskal 2 = 0.82, df = 2, P = 0.67 ). Trends in microsites associated with initial tree establishment were characte rized by species. On Divide Mountain we found a statistical difference among proportions of species found in unknown or minimally protecting microsites (i e small ground terraces) 2 = 9 769, df = 2, P = 0.008) This difference was due to a larger than expected number of whitebark pine that established with no clear protective microsite or minimal protection and a greater than expected number of subalpine fir associated with more shelteri ng microsites. At the Line Creek RNA, similar statistical differences were also found 2 = 11.3217, df = 2, P = 0.003). Statistical s ignificance largely derived from a proportionally greater than expected number of spruce and fir i n more sheltering microsites. Shoot Lengths and Shoot Growth Rate Comparisons K rummholz shoot lengths wer e compared by year between the species. Trends in length were similar regar dless of year and study area. Krummholz w hitebark pine shoots were roughly 2 to 3 times longer than both spruce and fir shoots ; and spruce and fir shoot lengths were not statistic ally different from e ach other (Table I V.2 ). from July to September w ere the highest of all comparisons representing growth rates on

PAGE 54

42 average 2 to 10 times faster than fir and spruce. Fir and spruce growth rates were not sta tistically different (Table I V.3 ). The one exception to this trend w as similar growth rates for subalpine fir and whitebark pi n e on the Line Creek RNA in 2011. For upright, subalpine conifers, in all c omparisons of shoot lengths by year and study area, nested ANOVA results demonstrated that within tree variance contributed only minimally to the overall variance (0.25% +/ 0.19). The large st portion of variance was described by differences between means of different species (97.6% +/ 1.98%). Because within tree variance was low, we used the mean of the five measured shoots per individual tree for all upright shoot comparisons for all subsequent analyses Similar trends were observed at both study areas (Table I V.4 ). On Divide Mountain, the same shoot length trend s occurred in both years: w hitebark shoots were 1.5 to 3 times longer than both spruce and fir shoots. In 2011, spruce and fir shoots were not statistically different, but in 2012 spruce shoots were longer than fir shoots. At the Line Creek RNA, the same trend was observed in both years: w hitebark pine shoots were 2 to 3 times longer than both fir and spruce shoots, and the latter two species were not statistically different. Means of u pright conifer shoot lengths after the minimum needle length was subtracted were also compared in a One by ye ar and study area to determine if species differences still existed. One Way ANOVA analyses were significant (Divide, 2 011: F = 7.2835 df = 2, P = 0.003 2012: F = 54.3 df = 2, P = 6.297e 14 ; Line Creek, 2011: F = 10.861 df = 2, P = 0.0003, 2012: F = 197.23 df = 2, P < 2.2e 16 ). species mean shoot length trends to remain mostly unc hanged in most comparisons (Table IV.5) Whitebark pine

PAGE 55

43 shoots were longer than subalpine fir shoots on Divide and Line Creek RNA. Whitebark pine shoots were not longer than Engelmann spruce shoots on Divide in 2011 but they were longer in 2012. At the Line Creek RNA, whitebark pine shoots were longer than Engelmann spruce shoots in both years. Subalpine fir shoots were found to be equal in length to Engelmann spruce in all comparisons With only one exception, upright shoots were longer than krummholz shoots of the same species at each study area. The one exception was subalpine fir shoots on the Line Creek RNA in 2012, where krummholz shoots did not differ from the upright shoot counterparts. Proportions of upright to krummholz shoots were generally similar between species for both years at both study areas (Table IV.6 ). Other Small Tree Measurements 2010 2012 increases in small tree st em diameter, canopy area, and height measurements did not produce consistent statistical differences in terms of species trends. Kruskal Wallis rank sum test for differences in median stem diameter resulted in no statistical differences among species on D ivide Mountain ( 2 = 0.83, df = 2, P = 0.66). At the Line Creek RNA, this test was significant ( 2 = 11.9 df = 2, P = 0. 003 ) Median w hitebark pine stem diameter increases were smaller than both spruce ( W = 294.5, P = 0.03 ) and fir ( W = 34, P = 6.1e 4). Spruce and fir median stem diameter increases were not significantly different from each other ( W = 92.5, P = 0.29) ( Appendix II, Table 1 ). Kruskal Wallis rank sum analysi s of median canopy area increase indicated no statistical differences among species at the Line C reek RNA ( 2 = 12.029, df = 2, P = 0. 21 ). On Divide Moun tain, whitebark had a greater median increase in canopy area than Engelmann spruce ( W = 140, P = 0.02), but not subalpine fir ( W = 103, P = 0.95).

PAGE 56

44 Subalpine fir and Engelmann spruce were not statistically dif ferent ( W = 20, P = 3.6e 4 ) ( Appendix II, Table 2 ). Analyses of median height increases resulted in no significant difference s at either study area ( Appendix II, Table 3 ). Results Summary. We observed higher proportions of whitebark pine growing in a solitary exposed state as compared to subalpine fir and Engelmann spruce. Whitebark pine was also more common in minimally sheltering microsites than other species. These results indicate that whitebark pine may be able to survive in harsh treeline cond itions better than spruce and fir. P roportions of krummholz shoot lengths to upright tree shoot lengths were similar between species, with krummholz shoot lengths being generally shorter than upright tree shoot lengths. This indicates that there is a r educed ability to grow at treeline for all three species. In terms of growth, whitebark pine produced the longest shoot lengths, both as a krummholz and upright tree. Whitebark shoot growth rates were also generally faster than spruce and fir. These res ults indicate the whitebark pine is capable of vigorous growth during short growing seasons. Planting and Sowing Study Seedling Survival. One year after planting on Divide Mountain the nursery grown subalpine fir seedlings experienced very high overall mortality (90% ). Of the 40 total seedlings planted per microsite type, survival was as follows: whitebark microsites 12.5% (n = 5) spruce microsites 7.5% (n = 3) rock microsites 5% (n=2) and open microsites 15% (n=6 ) Chi square goodness of fi t analysis showed no significant differences in survival among microsite types ( 2 = 2.5, df = 3, P = 0.47).

PAGE 57

45 Survival of the planted Engelmann spruce seedlings was generally higher at the Line Creek RNA, but overall mortality was high at 63.1%. Of the 40 seedlings planted per microsite type, survival was as follows: whitebark microsites 32.5% (n=13) spruce microsites 3 5% (n=14), rock microsites 42.5% (n=18) and open microsites 37.5% (n=14). There were no significant differences in survival among microsite types ( 2 = 0.59, df = 3, P = 0.9) (Figure IV.2). A One Way ANOVA test for differences in mean apical terminal shoot lengths of surviving seedlings did not show any statistical differences among microsite type at either study area ( Line Creek: F = 0.26, df = 3, P = 0. 85 ; Divide: F = 2.3, df = 3, P = 0.81). We found qualitative measurement of seedling vigor to differ statistically among microsite types P = 7.3e 4; Line Creek P = 9.6e 3). On Divide, whitebark microsites had a greate r than expected number of seedlings classified Spruce and rock microsite vigor trends were distributed across vigor classes as expected. Open microsites had a greater than expected number of poor and fair vigor seedlings and no seedl ings in the good and excellent vigor classes. At the Line Creek RNA, whitebark pine microsites had a greater than expected number of excellent vigor seedlings. Spruce microsites had fewer than expected good vigor class seedlings and a greater than expecte d number of excellent vigor s eedlings. Vigor of rock microsite seedlings was distributed among classes as expected Open microsites had a greater than expected number of good vigor seedlings and a fewer than expected number of excellent vigor seedlings.

PAGE 58

46 Seed Germination and Summer Survival Sown Engelmann spruce seed germination totals were counted in July 2012, and revisited in September 2012 to assess proportion of summer survival by microsite type Seed germination numbers at the Line Creek RNA we re small. Of the 400 seeds planted, only 7 (1.8%) germinated. The germinant distribution was as follows: Whitebark microsites 2, spruce microsites 3, rock microsites none, and open microsites 2. There were no significant differences in germinati on among the four microsite types ( P = 0.44). On Divide Mountain, 80 ou t of 400 (20%) seeds germinated In July, whitebark microsites had 12 germinants, spruce microsites had 17 germinants, rock microsites had 32 germinants, and open microsites had 19 germinants. Dif ferences in germination among the different microsite types was statistically significant ( P = 0.01). This is largely due to more germinations than expected in rock microsites and fewer than expected in whitebark microsites. When sites were revisited in September 2012, we observed that both study areas had experienced substantial germinant mortality over the summer months. At the Line Creek RNA, only 3 out of 7 germinants survived: 2 at whitebark mic rosites and 1 at a spruce microsite Rock and open microsites had no living germinants. Mortality was not significantly different a est P = 0.31). On Divide Mountain, 42 out of 80 germinants survived. The numbers of surviving germinants are as follows: whitebark microsites 11 seedlings, spruce microsites 8 seedlings, rock microsites 18 seedlings, and open microsites 5 seedlings (Figure IV.3 ). Difference in microsite type survival was statistically significan t

PAGE 59

47 est P = 0.004). A comparison of observed survival numbers vs Chi square expected survival numbers revealed that whitebark microsites were associated with a 5.7 times greater than expected survival advantage and a very low comparative risk of death (0.18 times expected). Expected values were at or near 1.0. Open microsites had the lowest chance of survival after germination at 0.64 times greater than expected and were associated with a 1.56 times greater than expected relative expecte d risk of death. Spruce and rock microsites had relative survival advantages very close to expected (0.89 and 1.08 times expected, respectively). Similarly, these microsites had relative death risks close to expected (1.12 and 0.93 respectively) (Table IV.7). Results Summary. The seedling planting experiment did not have significant results in terms of survival or terminal shoot lengths by microsite type. However, conifer microsites were generally associated with higher overall health and vigor of th e leeward seedlings. For the seed sowing experiment only the results at the Divide Mountain study area were instructive. Very few seeds germinated at the Line Creek RNA, and thus differences among microsite types were not observed. On Divide Mountain, rock microsites initially favored germination. However, whitebark pine microsites had the highest seedling summer month survival, indicating better shelter and perhaps more favorable growing conditions than the other microsites examined. Whitebark micro sites also had the highest summer survival at the Line Creek RNA, but small sample sizes did not yield statistical significance.

PAGE 60

48 Girdling Study Th ree ungirdled control whitebark trees (1 on Divide Mountain and 2 on White Calf) were infected by blister rus t and died over the course of the study. Their death resulted in canopy defoliation. As a result, they no longer provided windward shelter to the leeward conifer. These three sites were rem oved from analyses Leeward Conifer Vigor There was a statis tical ly significant difference between control and experimental groups in the change in leeward conifer vigor over the 2010 to 2012 time period ( n = 44, Exact Test P = 0.002). This difference was attributed to the number of experimental sites th at lost vigor (77%; n = 17 of 22), and the number of control sites that remained the same or increased vigor over the course of the study (79%; n = 15 of 19). Shoot Lengths A BACI analysis comparing leeward conifer shoot lengths for experimental or con trol treatment type indicated statistical differences over the course of the study for treatment in terms of year ( F = 8.17 df = 2 P = 0.00 5 ) Initially in 2010, there were no differences in shoot length by treatment site type (Wilcox Signed Rank W = 2 36, P = 0.89). Overall, shoot lengths generally decreased over time, with experimental sites experiencing greater decline in length than control sites (Figure I V.4). The greatest difference in shoot lengths occurred over the 2011 to 2012 time period, wher e shoot mortality was a lso highest ( Wilcox Signed Rank W = 74.5, P = 0.0003). A partial explanation for this decline in overall sample mean shoot lengths is the mortality of some marked shoots over the course of the study W e compared naturally exposed and leeward conifer shoot lengths for 2011 vs. 2012. We analyzed these shoots lengths by computing the difference between leeward

PAGE 61

49 and exposed shoots for each tree island in the study. The differences were then compared in a BACI analysis as groups of tre e islands with a either control or girdled whitebark pine for 2011 and 2012 (Fig ure III. 4) Results did not show different trends of mean shoot length changes in terms of treatment type from year to year ( F = 1.4, df = 2 P = 0.24). However, there was a difference in shoot lengths based solely on treatment type. C ontrol and experimental tree island shoot length differences were significant ( F = 26.2, df = 1, P = 2.3e 6). There was a significantly greater difference in exp osed vs leeward shoot lengths f or tree islands with a control whiteb ark than a girdled whitebark: leeward conifers in control tree islands had on average of 10 cm longer shoot length compared to exposed shoots. In experimental tree islands, this difference was roughly 1 cm with exposed shoots slightly longer than leeward shoots This indicated that tree islands with a dead windward whitebark pine will have reduced shoot length grow th similar to areas of the tree island with no windward protection. Results also indicated that the prese nce of a windward whitebark pine is associated with longer annual shoot growth Shoot Mortality The proportional differences in mortality from 2011 to 2012 between naturally exposed shoots and experimental or control leeward shoots from the same tree isl and w ere statistically significant ( W = 129.5, P = 0.05 ). There were similar differences in mortality between naturally exposed shoots with no windward protection and the shoots leeward of a girdled whitebark pine Leeward conifers in control sites had a n overall lower proportion of mortality than experimental leeward conifers. Mean shoo t mortality was 18.9% (SE = 5%) for control leeward conifers and 59% (SE = 7.7%) for experimental leeward conifers, indicating that presence of windward whitebark pine re duced shoot mortality in leeward conifers.

PAGE 62

50 Results Summary. After losing windward whitebark pine shelter, leeward conifers in experimentally girdled sites lost health and vigor over the course of the study. These conifers also experienced shorter shoot lengths and higher shoot mortality than leeward conifers in control sites with shelter from a windward whitebark pine, wh ich supports the hypothesis that loss of a windward whitebark from blister rust will be lead to decreased health of the immediately le eward conifer(s).

PAGE 63

51 Figures and Tables Figure I V. 1 Solitary krummholz tree density by species on Divide Mountain and Line Creek RNA The number of conifers per square meter was calculated using the me an from 20 transects at each study area. We found that Divide Mountain had the highest solitary krummholz conifer densities. Whitebark pine had the highest density at both study areas.

PAGE 64

52 Table IV 1 Species abundances of so litary conifers in transects The number of solitary krummholz conifers is shown by transect per study area. P values are the result of binomial distribution tests comparing whitebark pine to an expected equal distribution of 33%. The total number of tr ansects that had a higher than expected number of solitary whitebark pine is shown at the bottom of the P value column. Bolded P values are transects that had a higher abundance of solitary whitebark pine than expected based on an equal distribution of the three species. Divide Mountain Transect ID # WB # SF # ES # Total P Value 1 18 6 8 32 0.0037 2 26 6 6 38 6.71e 10 3 4 0 0 4 0.012 4 25 3 6 34 1.31e 6 5 39 10 19 68 2.073e 5 6 14 1 6 21 0.001 7 10 0 0 10 1.53e 5 8 15 0 0 15 5.99e 8 9 19 1 5 1 35 0.0048 10 3 16 4 23 0.02 11 3 0 0 3 0.036 12 0 0 0 0 n/a 13 2 8 1 11 0.16 14 17 9 0 26 0.006 15 22 5 1 28 8.71e 7 16 1 0 0 1 0.33 17 5 8 1 14 0.21 18 10 1 2 13 0.0013 19 75 23 9 107 3.95e 15 20 4 0 0 4 0.012 TOTAL: 312 111 64 487 15/19 Line Creek RNA Transect ID # WB # SF # ES # Total P Value 1 26 0 0 26 3.03e 13 2 3 0 0 3 0.036 3 1 0 0 1 0.33 4 11 0 0 11 5.05e 6 5 1 0 0 1 0.33 6 4 0 0 4 0.012 7 25 0 0 25 9.18e 13 8 6 1 1 8 0.02 9 49 8 0 57 1.71e 16 10 0 0 0 0 n/a 11 12 1 2 2 26 0.059 12 4 1 0 5 0.04 13 12 9 0 21 0.013 14 0 0 0 0 n/a 15 0 0 0 0 n/a 16 0 0 0 0 n/a 17 7 0 0 7 0.00043 18 8 1 0 9 0.00085 19 0 0 0 0 n/a 20 5 0 0 5 0.0039 TOTAL: 174 32 3 209 12/15 a. b.

PAGE 65

53 Table I V. 2 Krummholz sho ot le ngths 2011 and 2012 s hoot lengths ( mm ) were compared using Kruskal Wallis Analysis Rank Sum Test and Wilcox Post Hoc with a Bonferroni correction at a, Divide Mountain and b, Line Creek RNA All P values and test statistics shown are from the Wil cox Post Hoc. Divide Mountain Year Length Spp., Mean (SE) n Species Comparisons W Statistic P Value 2011 W P 22.0 ( 2.93) ES 8.8 (0.85) SF 11.3 (1.56) 17 15 15 WB > ES WB > SF SF = ES 205 190 149 5.3e 4 0.006 0.14 2012 WP 28.7 (3.35) ES 9.3 ( 0.92) SF 11.0 ( 1.15) 17 15 15 WB > ES WB > SF ES = SF 349 192 82 6.1e 6 8.4e 5 0.36 Line Creek RNA Year Length Spp., Mean (SE) n Species Comparisons W Statistic P Value 2011 WP 48.1 (4.25) ES 22.9 ( 2.45) SF 27.6 ( 3.34) 21 12 20 WB > ES WB > SF SF = ES 36 36 149 2.8e 5 4.0e 4 0.36 2012 WP 70.16 ( 3.67) ES 26.24 ( 3.74) SF 23.99 ( 4.3) 21 12 20 WB > ES WB > SF ES = SF 16 220 90 1.0e 8 0.0004 0.35 a. b.

PAGE 66

54 Table IV 3 Krummholz tree shoot growth r ates 2011 2012 s hoot growth rates u test p values are shown for a, Divide Mountain and b, Line Creek RNA. All P values and test statistics Sample sizes are shown for trees that did not display a negative growth rate (i.e., loss of shoot growth from wind blasts or experimental error in caliper placement) a. b. Line Creek RNA Year Growth Rate Spp., Mean (SE) n Species Comparisons t S tatistic df P Value 2011 WP 0.52 ( 0.06) ES 0.25 (0.41) SF 0.42 ( 0.03) 20 20 10 WB > ES WB = SF SF > ES 9.22 1.57 13.16 38 28 28 0.0021 0.80 0 .069 2012 WP 0.21 (0.03) ES 0.033 ( 0.01) SF 0.022 (0.007) 21 20 11 WB > ES WB > SF ES = SF 4.93 5.37 0.87 39 30 29 6.1e 6 4.3e 5 0. 96 Divide Mountain Year Growth Rate Spp., Mean (SE) n Species Comparisons t S tatistic df P Value 2011 W P 0.21 ( 0.02 ) ES 0.069 ( 0.01) SF 0.093 ( 0.06) 15 11 15 WB > ES WB > SF SF = ES 5.50 4.04 1.20 24 24 20 2.8e 5 3.8e 4 0 .69 2012 WP 0.15 (0.03) ES 0.03 ( 0.007) SF 0.05 (0.007) 17 15 15 WB > ES WB > SF ES = SF 4.59 3.88 2.01 30 30 28 1.5e 5 2.1e 4 0. 73 a.

PAGE 67

55 Table IV 4 Upright u pper subalpine conifer shoot lengths Data are based on 2011 and 2012 measur ements of f ive shoot lengths (mm) from each of 10 trees per species in 2011 and 20 trees per species in 2012. Results are shown for a, Divide Mountain and b, Line Creek RNA. P values and test statistics shown are the following significance from a One Way ANOVA. a. b. Divide Mountain Year Length Spp., Mean (SE) n Species Comparisons t statistic df P Value 2011 W P 54.64 ( 1.6) ES 33.55 ( 3.5) SF 27.06 ( 3.2) 20 20 20 WB > ES WB > SF SF = ES 5.44 7.73 1.37 38 38 38 5.8e 5 9.0e 7 0.27 2012 WP 67.12 ( 3.0) ES 32.66 ( 1.9) SF 23.21 ( 1.4) 20 20 20 WB > ES WB > SF ES = SF 9.76 13.42 3.96 38 38 38 <0.0001 <0.0001 0.01 Line Creek RNA Year Length Spp., Mean (SE) n Species Comparisons t statistic df P Value 2011 WP 69.92 ( 4.1) ES 34.64 ( 3.2) SF 38.4 ( 3.3) 20 20 20 WB > ES WB = SF SF > ES 6.72 5.95 0.81 38 38 38 5.0e 7 0.74 3.5e 6 2012 WP 70.16 ( 1.4) ES 26.24 ( 1.3 ) SF 23.99 ( 1.3) 20 20 20 WB > ES WB > SF ES = SF 23.55 24.10 1.22 38 38 38 <0.0001 <0.0001 0.46 b.

PAGE 68

56 Table IV 5 Upright shoots with minimum needle lengths subtracted Data are based on 2011 and 2012 measurements upright tree shoot lengths with the minimum needle l engths (as described in Flora of North America) subtracted. All P way ANOVA at a. Divide Mountain and b. Line Creek RNA. a. b. Line Creek RNA Year Length Spp., Mea n (SE) n Species Comparisons t statistic df P Value 2011 WP 54.64 ( 1.6) ES 33.55 ( 3.5) SF 27.06 ( 3.2) 20 20 20 WB > ES WB > SF SF = ES 4.05 3.68 0.38 38 38 38 <0.001 0.001 0.935 2012 WP 67.12 ( 3.0) ES 32.66 ( 1.9) SF 23.21 ( 1.4) 20 20 20 WB > ES WB > SF ES = SF 16.04 17.84 2.30 38 38 38 <0.001 <0.001 0.069 Divide Mountain Year Length Spp., Mean (SE) n Species Comparisons t statistic df P Value 2011 WP 54.64 ( 1.6) ES 33.55 ( 3.5) SF 27.06 ( 3 .2) 20 20 20 WB > ES WB > SF SF = ES 1.83 4.37 1.80 38 38 38 0.211 0.002 0.113 2012 WP 67.12 ( 3.0) ES 32.66 ( 1.9) SF 23.21 ( 1.4) 20 20 20 WB > ES WB > SF ES = SF 5.79 9.76 4.80 38 38 38 <0.001 <0.001 0.001 b.

PAGE 69

57 Table IV 6 Small s hoot lengths vs. upright shoot lengths : proportions Data are based on 2011 and 2012 measurements of small trees and upright trees. Mean (SE) is shown for each sample popul ation. The proportion of small shoots to upright shoots is shown for a. D ivide Mountain and b. Line Creek RNA. 2011 Small Tree Mean (SE) Upright Tree Mean (SE) Proportion Small to Up WP 22.02 (2.9) 54.64 (1.6) 0.403 ES 8.8 (0.85) 33.55 (3.5) 0.262 SF 11.3 (1.6) 27.06 (3.2) 0.419 2012 WP 28.66 (3.4) 67.12 (3.0) 0.42 7 ES 9.29 (0.9) 32.66 (1.9) 0.284 SF 10.97 (1.1) 23.21 (1.4) 0.473 2011 Small Tree Mean (SE) Upright Tree Mean (SE) Proportion Small to Up WP 48.08 (4.2) 69.92 (4.1) 0.688 ES 22.87 (2.4) 34.63 (3.2) 0.660 SF 27.56 (3.8) 38.40 (3.3) 0.718 201 2 WP 48.18 (3.7) 70.16 (1.4) 0.687 ES 21.65 (3.7) 26.24 (1.3) 0.825 SF 25.56 (4.3) 23.99 (1.3) 1.065 a. b.

PAGE 70

58 Figure IV 2 One year post planting seedling survival per microsite The total number of seedlings that survived for each microsite type is shown for the Line Creek RNA (dark grey), and Divide Mountain (light grey).

PAGE 71

59 Figure IV 3 2012 Divide Mountain seed germination c ounts Seedling microsites were visited in July t o observe initial germination and again in September 2012 to observe summer drought mortality. While all microsite types experienced some mortality, conifer germinants leeward of whitebark microsites tended to have a lower chance of mortality while those in open exposed conditions experienced the highest mortality This indicates whitebark microsites may provide more favorable conditions for seedling establishment. Table IV 7 Summer 2012 survival advantage and relative death risk of seed germinants on Divide Mountain Using Chi square expected microsite survival totals compared to actual survival totals per microsite we calculated the relative survival advantage and death risk for each microsite compared to each other. Of the four types, whitebark microsites are associated with the highest survival advantage and lowest risk of death. Open microsites have the lowest s urvival advantage and highest risk of death. Microsite Relative Survival Advantage Relative Risk of Death Whitebark 5.70 0.18 Spruce 0.89 1.12 Rock 1.08 0.93 Open 0.64 1.56

PAGE 72

60 Figure IV 4 Girdling Study leeward conifer shoot length trends over time Three ye ar leeward conifer shoot length means for both treatment ( girdled) and control (non girdled) sites at Whitecalf Mountain and Divide Mountain. Trends generally reflect a decrease in mean shoot lengths over time, with treatment sites experiencing shorter shoot lengths than control sites. Note: C ontrol sites where t he windward whitebark died during the study were removed from this comparison.

PAGE 73

61 CHAPTER V SYNTHESIS AND DISCUSSION Study Conclusions. The overall objectives of this research were to determine experimentally and empirically the attributes and ecological interactions that enable whitebark pine to facilitate tree island development and to address how the mortality of whitebark pine from blister rust may impact t hese ecosystem functions We tested three hypotheses focused on learning about white facilitative functions: 1 ) W hitebark pin e is hardier than other alpine treeline ecotone conifer species as demonstrated by more rapid shoot growth and higher survival at treeline; 2) whitebark pine provides a more favorable microsite for tree island recruitme nt than other common alpine treeline ecotone microsites; and 3) blister rust mortality of whitebark pine in established tree islands will lead to loss of vigor of leeward conifers. Each of these hypotheses relates to different aspects of wh tree islands and the maintenance of e stablished tree islands (Figure II. 1). The results foundation species at treeline. F irst our results clarify the issues of hardiness concerning the prevalence, and shoot growth rates of whitebark pine in our two study area s as illustrated by position #1 in the conceptual model (Figure II. 1) This first finding is extremely important, b ecause whitebark pine in our study area was previously found to be the most common species initiating multi tree tree islands (Resler and Tomback 2008). Our alpine treeline ecotone study areas are characterized by harsh climatic conditions consisting of high winds, cold

PAGE 74

62 temperatures, and direct solar radiation ( Marr, 1977; Arno and Hammerly, 1984; Finklin, 1986 ; Maher et al., 2005) With respect to the first hypothesis, un der harsh, treeline conditions, whitebark pine is the most prevalent conifer growin g in exposed sites as a seed distribution in open areas at treeline, whitebark pine appears better able to germinate and establish under challenging conditions than both subalpine fir and E nglemann spruce. This is also demonstrated by the proportionally greater number of solitary whitebark pines found in minimally sheltering niches or non sheltering microsites compared to Engelmann spruce or subalpine fir. Also indica tive of survival and vigor is the ability of whitebark pine to grow longer shoot lengths, thus potentially expanding canopy biomass more rapidly, than the other common treeline conifers. Conifer shoots are responsive to a variety of environmental factors, including length of growing season, soil texture, moisture, and nutrient levels, temperature, photoperiod, tree vigor, and tree species (Kozlowski, 1964). Identical trends were found in analyses of shoot lengths both for krummholz conifers and upper suba lpine whitebark, spruce, and fir, with whitebark pine shoots longer than spruce and fir shoots in both growth forms. This suggests that whitebark may have a species growth advantage in general in the upper subalpine but this is also the case at treeline, although all implications are not completely clear. The higher growth rates of krummholz whitebark shoots during summer months in comparison with spruce and fir also supports this finding. In order to produce longer shoots, whitebark pine must be able to capitalize on the scarce environmental resources found at treeline and allocate them into annual growth. There could well be trade offs in growth that we are not aware of,

PAGE 75

63 such as differential shoot to root ratios among the conifer species ( Tilman 1988 ). Our measurements of increases in stem diameter, canopy area, and height did not show species differences. This is likely due to the short duration of the study. There may be differences among species growth strategies (i.e., height vs. canopy volume). Longer shoot lengths of whitebark pine suggest a species strategy for increasing canopy volume. This trend might be advantageous in harsh treeline environments where upright growth is often lost by wind and snow blasts and ground level canopy growth is favored (Arno and Hammerly, 1984). Given the short growing season at treeline, the ability to increase the volume of photosynthetic biomass appears to support the premise that whitebark pine is hardier. Whether this can happen may depend on exposure, f lagging, water and nutrient availability, and annual snowpack depth, which provides protection. Because whitebark pine appears better able to survive and grow in the alpine treeline ecotone than other conifer species, and is thus more prevalent, it is more likely to initiate tree islands by acting as a nurse object for less hardy species, such as Engelmann spruce as stated in hypothesis #2 Spruce seedling survival is facilitated with the presence of overhead branches and windward protection ( Hattensch wiler and Smith, 1999; Germino et al., 2002). H ypothesis #2 predicts that whitebark pine is a better f acilitator or nurse object than Engelmann spruce, rocks, or no object. If it is, than whitebark pine will be more likely to be a tree island initiator, as depicted in position 2 and the leeward red star in the conceptual model. Our studies provide some evidence that whitebark microsites facilitate survival of other conifers better than the other microsites examined. On Divide Mountain, rock

PAGE 76

64 microsites better facilitated the initial germination of sown seeds, most likely due to g reater radiant heat. However, seeds germinating in whitebark microsites were associated with greater survival during the summer months than all other tested microsit es, includi ng Engelmann spruce. While not significant due to small sample sizes, the same trend was observed on the Line Creek RNA. The summer months represent the most critical stage for a newly germinated seedling, which must endure periodic drought and UV radiat ion exposure. Studies have found minimal seedling mortality over winter months (Day, 1964; Cui and Smith, 1991), likely because seedlings are covered by snowpack at that time and have reduced exposure to harsh solar radiation, temperature extremes or chil ling winds. Differences in survival for the nursery grown seedling s among the various microsite types did not show statistical differences but seedling health and vigor w ere greatest when associated with conifer microsites. Of the conifer microsites, w hitebark was the most likely to be associated with excellent seedling vigor, demonstrating that growing conditions may be more moderate leeward of whitebark as compared to rocks, Engelmann spruce, and open microsites. Hence, if a seed germinat es leeward o f a whitebark pine at treeline, chances of summer mortality are lower and it will likely have greater growth vigor should it become established. Every established solitary tree potentially could facilitate the establishment of a leeward c onifer, thus starting a tree island. Whitebark pine in our study areas is the most common tree island initiator, and may create a more favorable microsite for leeward tree survival. Once a whitebark pine facilitates the establishment of a leeward conifer, other conifers may continuously establish on the leeward side of the developing tree island.

PAGE 77

65 This cycle of establishment often continues until a large multi tree island has formed. Even after a tree island becomes established, the most windward conifer still provides a sheltering role to the leeward individual(s). However, the importance of the windward whitebark pine in offering protection to established conifers needs to be demonstrated in order to fully understand the potential effects of mortality f rom blister rust, as stated in hypothesis #3 and represented in the conceptual model in position #3. With blister rust rapidly killing whitebark pine in the alpine treeline ecotone (3 of 22 control whitebark died from blister rust over the course of our s tudy) these previously sheltered subalpine fir and Engelmann spruce individuals experience new exposure t o wind, snow, and ice blasts. By simulating blister rust through girdling and defoliating the windward whitebark pine, we found that exposed leeward conifers were more likely to experience decreased qualitative vigor, shorter shoot lengths, and have higher terminal shoot mortality than control sites with a healthy windward whitebark pine. This windward shelter may be most important in years with harsh winter conditions, low snowpack, and cold, high winds. In summary, we have found evidence for the importance of whitebark pine at every investigated stage of the tree island development: the hardiness of whitebark pine as an exposed, solitary tree and i ts role as a facilitator in providing a sheltering microsite, the protection of an establishing seedling, and the windward role in established tree islands. With whitebark pine declining at treeline from canopy damage and mortality caused by blister rust infection, the ecosystem functions that we have studied here may well be diminished. Potential Implications for Whitebark Pine Decline at Treeline

PAGE 78

66 Several interactive factors currently threaten the regeneration of whitebark pine and the health of establis h ed whitebark pine within the alpine treeline ecotone At the same time these krummholz conifers are primarily experiencing blister rust mortality, upper subalpine whitebark pine are declining due to blister rust infection, mountain pine beetle outbreaks, and fire suppression, leading to the successional advancement of predominantly subalpine fir and Engelmann spruce stands (Tomback and Achuff 2010; Tomback et al. 2011 ). The loss of these reproductively viable upper subalpine individuals has a direct imp act on the alpine treeline ecotone Since treeline whitebark pine rarely produce cones with viable seeds, krummholz whitebark pine are regenerated by upper subalpine elevation whitebark pine seed sources ( Malanson et al,. 2007; Tomback and Resler, 2007 ). Upper subalpine whitebark pine mortality leads to lower seed production, and consequently, fewer seeds available for dispersal at treeline by (Tomback and Resler, 2007) Alpine treeline ecotone vegetation dynamics may potentially be i mpacted by mortality of treeline whitebark pine and the reduced regeneration due to loss of cone bearing upper subalpine individuals (Tomback and Resler, 2007; Resler and Tomback 2008). Since w hitebark pine is a dominant alpine treeline ecotone tree islan d initiator, fewer tree islands will be facilitated by whitebark (Figure V.1). Potential impacts of fewer tree islands include increased soil erosion and snowpack melt off (Smith et al., 2009; Tomback et al. 2011), possibly leading to changes in downstrea m hydrology. Another is sue likely to further impact alpine treeline ecotone dynamics is global climate change. Warmer temperatures are expected to cause an upward shift in treeline (Millar et al, 2004 ; Schrag et al., 2008 ), with an estimated elevation ga in of 140 to 700 m

PAGE 79

67 (Grace et al, 2002). As treeline moves upward, new tree islands will form when solitary conifer s are able to establish in areas that were previously solely alpine tundra. These conifers will act as nurse objects by providing windward s helter for leeward conifers, ultimately leading to the formation of new tree islands at higher elevations With declining numbers of whitebark pine from b l ister rust at treeline and loss of seed production there will be less effective facilitation at the uppermost elevations of suitable condition s to initiate tree islands. This is because whitebark pine is a dominant tree island initiator through parts of its range. Thus, the decline of whitebark pine may lead to a reduced ability of treeline as a whole to respond, possibly leading to the perception that treeline is not moving up or moving more slowly than suitable temperature zones (Tomback and Resle r 2007). Potential implications for this issue are a loss of treeline biodiversity as species compositio ns change (i.e., greater proportions of fir and spruce), and a reduced range of the alpine treeline ecotone community if tree islands do not form at higher elevations and the subalpine forest moves upwards. The re is a growing r ecognition of the importan ce of plant plant facilitative interactions in stressful environmen ts, such as treeline ecosystems ( Calloway et al., 2002; Lortie et al 2004; Brooker et al 2008 ). Plant facilitators often germinate in harsh environments with minimal protection. Onc e established, they provide favorable growing conditions such as windward shelter, shade, and moisture retention (i e snowpack) for other species ( Calloway, 1998; Lortie et al., 2004; Baumeister and Calloway, 2006 ; Brooker et al., 2008; Batllori et al., 2009 ). T he facilitation offered by one species often has the potential to benefit more than one other species ( Lortie et al., 2004; Baumeister and Calloway, 2006). Here, we have identified the importance of

PAGE 80

68 whitebark pine for the establishment of spruce and fir leading to development of tree island s and have thus demonstra ted the crucial facilitation role this species plays at treeline in parts of its range. With whitebark pine currently declining, and climate change potentially moving treelines upward, these community changes are already taking place in treeline ecosystems in the Rocky Mountains.

PAGE 81

69 Figures and Tables Figure V. 1 Potential consequences of blister rust to alpine treel ine dynamics (modified from Tomback and Resler 2007). Fewer seeds dispersed to treeline by nutcrackers: blister rust in subalpine whitebark pine Blister rust damages and kills whitebark pine at treeline Decline in treelin e whitebark pine Fewer tree islands initiated by whitebark pine (less facilitation) Whitebark pine shows little or no response to global warming in upper treeline boundary Reduced ability of treeline to respond (or lag in response time) to global w arming at the upper boundary

PAGE 82

70 REFERENCES Arno, S. F. and R. P. Hammerly, 1984. Timberline: Mountain and Arctic Forest Frontiers. Seattle, WA: The Mountaineers Arno, S. F. 1986. Whitebark pine cone crops a dim inishing source of wildlife food? Western Journal of Applied Forestry. 1: 92 94 Arno, S. F., and R. J. Hoff. 1990. Pinus albicaulis Egelm. Whitebark pine. Pages 268 279 in R.M. Burns and B. H. Honkala, technical coordinators. Silvics of North America, Vol 1, Conifers. USDA Forest Service, Agriculture Handbook 654, Washington D.C. Bar tos, D. L., and K. E. Gibson. 1990. Insects of whitebark pine with emphasis on mountain pine beetle. Pages 171 178 in W. C. Schmidt and K. J. McDonald, compilers. Proceedings Symposium on whitebark pine ecosystems: Ecology and Management of a high mountain resource. USDA Forest Service Intermountain Research Station, General Technical Report INT 270, Ogden, Utah Batllori, E., J. J. Camarero, J. M. Ninot, and E. Gutirrez. 20 09. Seedling recruitment, survival and facilitation in alpine Pinus uncinata tree line ecotones. Implications and potential responses to climate warming. Global Ecology and Biogeography. 18:460 472. Baumeister, D., and R. M. Callaway. 2006. Facilitation b y Pinus flexilis during succession: a hierarchy of mechanisms benefits other plant species. Ecology. 87(7):1860 1830. B enedict, J.B. 1984. Rates of Tree Island Migration, Colorado Ro cky Mountains, USA. Ecology, 3: 820 823 Brasier, C. M. and K. W. Buck. 2001. Rapid evolutionary changes in a globally invading fungal pathogen (Dutch elm di sease). Biological Invasions 3: 223 233 Brooker, R.W., F. T. Maestre R. M. Cal laway, and 21 others. (2008) Facilitation in plant communities: the past, the present and the future. Journal of Ecology 96 : 18 34. Callaway, R. M.1998. Competition and facilitation on elevation gradients in subalpine forests of the northern Roc ky Mountains, USA. Oikos, 82: 561 573 Callaway, R.M., R.W. Brooker, P. Choler, Z. Kikvidze, C.J. Lor tie, R. Michalet, L. Paolini, F.I. Pugnaire, B. Newingham, E.T. Aschehoug, C. Arm as, D. Kikodze, and B.J. Cook. 2002. Positive interactions among alpine plants increase with stress. Nature 417: 844 848.

PAGE 83

71 Campbell, E. M. and J. A. Antos. 2000. Distribution and severity of white pine blister rust and mountain pine beetle in whitebark pine in British Columbia. Canadian Journal of Forest Research, 30: 1051 1059 Cavieres, L.A., C. L. Quir oz, M. A. Molina Montenegro A. A. Muoz, and A. Pauchard. 2005. Nurse ef fect of the native cushion plant Azorella monantha on the invasive non native Taraxacum officinale in the high Andes of central Chile. Perspectives in Plant Ecology, Evolution and Systematics, 7 : 217 226. Cavieres, L.A., E. I. Badano, A. Sierra Almeida, an d M. A. Molina Montenegro 2007. Microclimatic modifications of cushion plants and their consequence s for seedlings survival of native and non native plants in the high Andes of central Chile. Arctic, Antarctic and Alpine Research, 39 : 229 236. Chornesky E A, et al. 2005. Science priorities for reducing the threat of invasive species to susta inable forestry. BioScience 55: 335 348. Cole, Walter E., and G. D. Amman. 1969 Mountain pine beetle infestations in relation to lodgepole pine diameters. Res. Note INT 95. Ogden, UT: U.S. Department of Agriculture, Forest Service, Intermountain Forest and Range Experiment Station, 7 Cui, M. and W. K. Smith. 199 1. Photosynthesis, water relations and mortality in Abies lasiocarpa seedlings during natural establishment. Tree Physiology 8: 37 46 Day, R.J. 1964. The microenvironments occupied by spruce and fir regeneration in the Rocky Mountains.Can. Dept. Forestry. For. Res. Publ. No. 1037. Diskin, M., K.C. Steiner and F.V. Hebard. 2006. Recovery of American chestnut ch aracteristics following hybridization and backcross breeding to restore blight ravaged Castanea dentata, For. Ecol. Manage. 223 : 439 447 Easterling, W. 2005. Assessing the consequences of climate change for food and forest resources: a view from the PCC. Climate Change 70: 165 189 eFloras. 2008. Published on the Internet http://www.efloras.org [accessed 7 November 2012]*' Missouri Botanical Garden, St. Louis, MO & Harvard University Herbaria, Cambridge, MA. Ellison, A. M. M. Bank B. Clinton, E. Colburn K. Elliott, C. Ford, D. Foster, B. Kloeppel, J. Knoepp, G. Lovett, J. Mohan, D. Orwig, N. Rodenhouse, W. Sobczak, K. Stinson, J. Stone, C. Swan, V. Thompson, B. Von Holle, and J. Webster. 2005 Loss of foundation species: Consequences for the structure a nd dynamics of forested ecosystems Frontiers in Ecology and the Environment 3 : 479 486

PAGE 84

72 Finklin, A. I. 1986. A Climatic Handbook for Glacier National Park with Data for Waterton Lakes National Park Ogden, UT U.S.D.A. Forest Service, Intermountain Research Station. Germino, M. J., W. K. Smith, and A. C. Resor 2002. Conifer seedling distribution and survival in an alpine treel ine ecotone. Plant Ecology 162: 157 168 Gernandt, D. S., G. G. Lopez, O.S. Garcia, and A. Liston. 2005. Phylogeny and cla ssificatio n of Pinus. Taxon 54: 29 42 Grace, J., F. Berninger and N. Laszlo. 2002. Impacts of Climate Change on the Tree Line. Annals of Botany 90: 537 544 Habeck, J. R. 1969. A gradient analysis if a timberline zo ne at Logan Pass, Glacier Park, Montana. Northwes t Science. 43(2): 65 73 Hamann, A., and T. Wang. 2006. Potential effects of climate change on ecosystem and tree species distribution i n British Columbia. Ecology 87: 2773 2786 Hattenschwiler, S., and W. K. Smith 1999. Seedling occurrence in alpine tre eline conifers: a case study from the central Rocky Mountains, USA. Acta Oecologica Inter national Journal of Ecology 20: 219 224 Holtmeier, F., and G. Broll. 1992. The influence of tree islands and microtopography on pedoecological conditions in the fores ts alpine tundra ecotone on niwot ridge, Colorado front range, USA. Arctic and Alpine Research 24: 216 228 Hutch ins, H. E., and R. M. Lanner. 1982. dispersal and establishment o f whitebark pine. Oecologia 55 : 192 201 Jenkins, J., J. Powell, J. Logan, and B. Bentz. 2001. Low seasonal temperatures promote life cycle synchronization. Bulletin of mathematica l biology (0092 8240), 63(3): 573 595. Keane, R. E. 2001. Successional Dynamics: Modeling an Anthropogenic Threat. In D. F. Tomback, S. F. Arno, and R. E. Keane, eds., Whitebark Pine Communities: Ecology and Restoration. Washington, DC: Island Press, 159 192. Kovacs, Kent F. F., R. G. Haight, D. G. McCullough, R. J. Mercader, N. W. Siegert, A. M. Liebhold. 2 010. Cost of potential emerald ash borer damage in U.S. communities, 2009 2019. Ecological Economics, 69: 569 578. Kozlowski, T. 1964. Shoot growth in woody plan ts. The Botanical Review 30(30): 335 396.

PAGE 85

73 Liston, A., W. A. Robinson, D. Pinero, and E. R. Alva rez Buylla, 1999. Phylogenetics of Pinus (Pinaceae) based on nuclear ribosomal DNA internal transcribed spacer region sequences. Mole cular Phylogenetic Evolution 11: 95 109 Logan, J. A., and J. Powell. 2001. Ghost Forests, Global Warming and the Mountain Pine Beetle. American Entomologist, 47(3):160 172. Logan, J. A., W. W. Macfarlane, and L. Willcox. 2010. Whitebark pine vulnerability to climate driven mountain pine beetle disturbance in the Greater Yellowstone Ecosystem. Ecological Applications 20:895 9 02. Lortie, C.J., R.W. Brooker, P. Choler, Z. Kikvidze, R. Michalet, F.I. Pug naire, and R.M. Callaway. 2004. Reth inking plant community theory. Oikos 107: 433 438. Lloy d, A. H. and L. J. Graumlich., 1997. Holocene dynamics of treeline forests i n the Sierr a Nevada. Ecology 78: 119 210. Loeh, C., 1996. Forest response to climate change: do simulations predict unrealistic di eback? Journal of Forestry 94: 13 15. Lorenz Teresa J. J., and K. A. Sullivan. 2009 Seasonal Differences in Space Use by Clark's Nutcra ckers in the Cascade Range. The Condor (Los Angeles, Calif.) (0010 5422), 111 (2), p. 326. Maher, E. L., M. J. Germino and N. J. Hasslequist. 2005. Interactive Effects of Tree and Herb Cover on Survivorship, Physiology and Microclimate of Conifer Seedling s at the Alpine Tree line Acadian. Canadian Jo urnal of Forest Research. 35(3): 567 574. Malanson, G. P. D. R. Butler, D. B. Fagre, S. J. Walsh D. F.Tomback, L. D. Daniels, L. M. Resler, W. K. Smith, D. J. Weiss, D. L. Peterson, A. G. Bunn, C. A. Hiemstra D. Liptzin, P. S. Bourgeron, Z. Shen, and C. Millar 2007 Alpine treeline of western North America: linking organism to landscape dynamics Physical Geography 28 : 378 396 Marr, J. W. 1977. The development and movement of tree islands near the upper li mit of tree growth in the southern Rocky Mountains. Ecolog y 58: 1159 1164. Matts on, D. J., and D. P. Reinhart. 1994. Bear use of whitebar k pine seeds in North America. In W. C. Schmidt and F. K. Holtmeier. Proceedings International workshop on subalpine stone pines and their environment: The status of our knowledge. USDA Forest Service, Intermountain Research Station, General Technical Report INT GTR 309, Ogden, Utah McCaughey W W and W. C. Schmidt 1990. Autecology of whitebark pine. In: Schmidt WC, McDonald KJ (eds) Proceedings Symposium on whitebark pine

PAGE 86

74 ecosystems: Ecology and management of a high mountain resource. Gen Tech Rep INT 270, USDA For Serv Intermountain Res Sta, Ogden, UT, pp 85 96 McDonald, G.I., and R. J. Hoff. 2001. Blist er rust : An introduced plague. In D. F. Tomback, S. F. Arno, and R. E. Keane, eds., Whitebark Pine Communities: Ecology and Restoration. Washington, DC: Island Press, 193 220. McDonald, G I. B. A. Richardson, P. J. Zambino, N. B. Klopfenstein, and M. S. Kim. 2006 Pedicularis and Castilleja are natural hosts of Cronartium ribicola in North America: a first report Forest Pathology 3 6: 73 82 Mc Gregor, M. D., and D. M. Cole. 1985 Integrating management strategies for the mountain pine beetle with multiple res ource management of lodgepole pine forests. USDA Forest Service Intermountain Research Station, General Technical Report INT 174, Ogden, Utah McKenney, D. W., J. H. Pedlar K. Lawrence, K. Campbell, and M. F. Hutchinson 2007. Potential impacts of climat e change on the distribution of North American trees. Bioscience 57: 939 948 McKinney, S. T. and C. E. Fiedler 2010 Tree squirrel habitat selection and predispersal seed predation in a declining su balpine conifer. Oecologia 162: 697 707. Millar, C. I., R. D. Westfall, D. L. Delany, J. C. King, and L. J. Graumlich 2004 Response of subalpine conifers in the Sierra Nevada, California, U.S.A., to 20th century warming and decadal climate variability. Arctic, Antarctic and Alpine Research 36: 181 200. Pric e, R. A., A. Liston, and S. H. Strauss. 1998 Phylogeny and systematic of Pinus. Pages 49 68 in D. M. Richardson, editor. Ecology and biogeography of Pinus. Cambridge University Press, New York. R Development Core Team 2010 R: A language and environ m ent for statistical computing. R Foundation for Statistical C omputing, Vienna, Austria. ISBN 3 900051 07 0, URL http://www.R project.org. Raffa, K. F., B. H. Aukema, B. J. Bentz, A. L. Carroll, J. A. Hicke, M. G. Turner, and W. H. Romme 2008. Cross scale drivers of natural disturbances prone to anthropogenic amplification: the dynamics of bark b eetle eruptions. Bioscience 58: 501 517 Resler, L. M. 2004 Conifer Establishment Sites on a Periglacial Landscape, Glacier National Park, Montana. Unpublished Ph. D. dissertation, Department of Geography, Texas State University San Marcos

PAGE 87

75 Resler, L. M., D. R. Butler, and G. P. Malanson. 2005 Topographic shelter and conifer establishment and mortality in an alpine environment, Glacier National Park, Montana. Phys ical Geography 26: 112 125. Resler, L. M. and D. F. Tomback 2008 Blister rust prevalence in krummholz whitebark pine: Implications for treeline dynamics. Arctic, Antarcti c and Alpine Research, 40: 161 170. Schrag, A. M., A. G. Bunn, L. J. Graumlich. 20 08. Influence of bioclimatic variables on tree line conifer distribution in the Greater Yellowstone Ecosystem : implications for species of conservation concern. Journal of Biogeography 35(4):698 710. Smith, E.K., L. M. Resler, E. Vance, W. Carste nsen, and K Kolivras. 2011 Modeling the Incidence of White Pine Blister Rust Infection in Whitebark Pine at Alpine Treeline in the Northern Rocky Mountains using GIS Arctic, Antarctic and Alpine Research 43(1): 107 117. Smith, William K. K., M. Germino, D. Johns o n and K. Reinhart. 2009 The Altitude of Alpine Treeline: A Bellwether of Climate Change Ef fects. The Botanical review 75: 163 190. Soule, M., J. Estes, J. Berger, and C. Martinez del Rio. 2003 Ecological effectiveness: Conservation goals for interactiv e sp ecies. Conservation Biology 17: 1238 1250. Spaulding, P. 1909 European currant rust on the white pine in America. Circ. 36. Washington, DC: U.S. Department of Agricu lture, Bureau of Plant Industry 4 p. Spaulding, P. 1911 The blister rust of white p ine. Bull. 206. Washingto n, DC: Bureau of Plant Industry 88 p. Spaulding, P. 1922 Investigations of the white pine blister rust. Bull. 957. Washington, DC: U.S. Department of Agriculture, Forest Service 100 p. Tilman, D. 1988. Plant Strategies and the D ynamics and Structure of Plant Communities. (Monographs in Population Biology, No. 26) Princeton University Press; Princeton, New Jersey; First Edition. Tomback, D.F. 1978 Foraging strategies of Clar 123 161 Tomback, D. F. 1982 mutualism hypothesis J. Anim. Ecol. 51 : 451 467 Tomback, D. F., and Y. B. Linhart. 1990. The evolution of bird dispersed pines Evolutionary Ecology 4: 185 219.

PAGE 88

76 Tomback, D. F., J. K. Clar y, J. Koehle r, R. J. Hoff, and S. F. Arno. 1995 The effects of blister rust on post fire regeneration of whitebark pine: The Sundance burn of northern Idaho (U.S.A.). Con servation Biology 9: 654 664. Tomback, D. F. 1998 Clark's nutcracker (Nucifraga colu mbiana). In Poole, A. and F. Gill, editors. The Birds of North America Philadelphia The Birds of North America, Inc., No. 331. Tomback, D. F., A. J. Anderies, K. S. Carsey, M. L. Powell, and S Mellman Brown. 2001. Delayed seed germination in whitebark pi ne and regeneration patterns following the Yellowstone fires. Ecology. 82(9):2587 2600. Tomback, D. F. 200 1. F. Arno, and R. E. Keane, eds., Whitebark Pine Communities: Ecology and Restoratio n. Washington, DC: Island Press, 88 104 Tomback, D. F. and K. C. Kendall 2001. Biodiversity Losses: the downward spiral. In D. F. Tomback, S. F. Arno, and R. E. Keane, eds., Whitebark Pine Communities: Ecology and Restoration. Washington, DC: Island Pre ss, 243 262. Tomback, D.F., and L. M. Resler 2007. Invasive pathogens at alpine treeline: Consequences for treeline dy namics. Physical Geography. 28: 397 418. Tomback, D.F ; and P. Achuff. 2010 Blister rust and western forest biodiversity: Ecology, value s, and outlook for white pines Forest Pathology. 40 : 186 225 Tomback, D.F., P. Achuff, A. W. Schoettle, J. W. Schwandt, and R. J. Mastrogiuseppe. 2011. The magnificent high elevation five needle white pines: ecological roles and future outlook. In: The f uture of high elevation, five need le white pines in Western North. USDA Forest Service Proceedings RMRS P 63. 2011. America: Proceedings of the High Five Symposium. 28 30 June 2010; Missoula, MT. Proceedings RMRS P 63. Fort Collins, CO: U.S. Department of Agriculture, Forest Service, Rocky Mountain Research Station. 376 p. U .S. Fish and Wildlife Service. 2011. Endangered and threatened wildlife and plants; 12 month finding on a petition to list Pinus albicaulis as endangered or threatened with critical hab itat. Federal Register. 76(138):42631 42654. Warwell, M. V., G. E. Rehfeldt and N. L. Crookston 2007. Modeling contemporary climate profiles and predicting their response to global warming for whitebark pine ( Pinus albicaulis ). In: Whitebark Pine: A Pac ific Coast Perspective, Ashland, OR, 2006 August 27 31. Ed. By Goheen, E. M., Sniezko, R. A. Portland, OR: US Department of Agriculture, Forest Service, Pacific Northwest Region, 139 142.

PAGE 89

77 APPENDIX I. Small Tree Measurements This appendix shows summary dat a for the Relative Vigor Study (hypothesis 1) small tree measurements. Table AI 1 Small Tree Measurement Summaries The 2010 baseline measurements taken for the small trees in the relative vigor study are shown in the table b elow. Summaries are presented as means (SE) for each measurement by species at a. Divide Mountain and b. Line Creek RNA. Divide Mountain Species Age (years) Height (cm) Stem Diameter (mm) Canopy Area (cm 2 ) WP ES SF 14.76 (2.8) 12.73 (1.1) 16.73 (2.5) 6.04 (0.85) 6.67 (0.98) 10.80 (2.34) 5.10 (0.92) 6.57 (1.46) 4.79 (0.86) 155.5 (58.7) 102.6 (32.7) 209.7 (88.7) Line Creek RNA Species Age (years) Height (cm) Stem Diameter (mm) Canopy Area (cm 2 ) WP ES SF 10.55 (1.2) 15.80 (1.1) 18.4 (0.72) 9.3 1 (1.13) 11.56 (1.43) 15.42 (1.49) 5.58 (0.73) 7.96 (0.75) 7.76 (0.84) 157.6 (40.1) 192.3 (34.6) 234.6 (45.5) a. b.

PAGE 90

78 Table AI. 2 Divide Mountain Small Tree Measurements by Site The 2010 baseline measurements taken for the small trees in the relative vigor study are shown for each individual site at a. Divide Mountain and b. Line Creek RNA (next page). Site Name Stem Diameter (mm) Height (cm) Canopy Area (cm 2 ) Age (years) Engelmann Spruce 1 2.70 5.7 42 8 Eng elmann Spruce 2 4.33 5.5 172 9 Engelmann Spruce 3 4.81 5.6 36 11 Engelmann Spruce 4 2.06 4.0 26 8 Engelmann Spruce 5 3.40 4.0 26 9 Engelmann Spruce 6 2.03 3.0 8 8 Engelmann Spruce 7 2.33 4.2 10 12 Engelmann Spruce 8 2.81 3.0 43 11 Engelmann Spruce 9 3.36 8.0 105 18 Engelmann Spruce 10 1.86 3.0 31 17 Engelmann Spruce 11 3.29 5.0 22 11 Engelmann Spruce 12 9.78 11 217 15 Engelmann Spruce 13 12.70 11 452 19 Engelmann Spruce 14 8.70 15 280 15 Engelmann Spruce 15 7.76 12 70 20 Subalpine Fir 1 1.68 1 .7 3 4 Subalpine Fir 2 1.81 2.4 3 8 Subalpine Fir 3 2.60 4.7 16 8 Subalpine Fir 4 2.18 5.2 19 9 Subalpine Fir 5 4.09 8.5 53 14 Subalpine Fir 6 4.13 9.0 7 13 Subalpine Fir 7 4.56 7.0 54 16 Subalpine Fir 8 5.54 8.0 240 19 Subalpine Fir 9 4.85 10 223 15 Subalpine Fir 10 2.77 5.0 12 17 Subalpine Fir 11 18.36 17.5 410 36 Subalpine Fir 12 13.45 26.0 105 27 Subalpine Fir 13 16.26 29.0 1140 28 Subalpine Fir 14 3.28 3.0 14 6 Subalpine Fir 15 13.06 25.0 848 31 Whitebark Pine 1 4.81 4.0 54 10 Whitebark Pine 2 2.35 6.0 3 8.5 Whitebark Pine 3 1.71 4.0 2 2.5 Whitebark Pine 4 2.07 3.4 5 5 Whitebark Pine 5 3.73 4.7 2 3 9 Whitebark Pine 6 2.43 4.5 35 8 Whitebark Pine 7 1.76 2.0 8 1 Whitebark Pine 8 4.66 5.5 87 15 Whitebark Pine 9 4.30 7.0 86 13 Whiteba rk Pine 10 2.65 3.5 14 8 Whitebark Pine 11 4.82 6.5 101 29 Whitebark Pine 12 2.82 3.0 36 7 Whitebark Pine 13 10.63 9.5 707 25 Whitebark Pine 14 12.81 7.0 531 26 Whitebark Pine 15 10.50 13.0 678 33 Whitebark Pine 16 2.82 4.0 35 11 Whitebark Pine 17 11.82 15.0 275 40 a.

PAGE 91

79 Site Name Stem Diameter (mm) Height (cm) Canopy Area (cm2) Age (years) Engelmann Spruce 1 6.27 8 99 17 Engelmann Spruce 2 8.50 15 258 26 Engelmann Spruce 3 7.36 10 94 21 Engelmann Spruce 4 6.31 4 56 10 Engelmann Spruce 5 9.33 10 165 14 Engelmann Spruce 6 5.48 7.5 118 21 Engelmann Spruce 7 3.35 4.5 82 10 Engelmann Spruce 8 10.90 14 506 19 Engelmann Spruce 9 12.03 13 212 13 Engelmann Spruce 10 8.05 14 159 16 Engelmann Spruce 11 10.19 23 415 18 Engelmann Spruce 12 3.80 8 90 9 Engelmann Spruce 13 4.79 4 64 15 Engelmann Spruce 14 9.42 13 237 23 Engelmann Spruce 15 9.62 12 163 18 Engelmann Spruce 16 12.19 25.5 560 13 Engelmann Spruce 17 14.03 17 189 15 Engelmann Spruce 18 2.39 3.5 8 7 Engelmann Spruce 19 3.64 5 33 12 Enge lmann Spruce 20 11.45 20.5 339 19 Subalpine Fir 1 9.40 20 467 17 Subalpine Fir 2 5.91 15 119 20 Subalpine Fir 3 7.96 11.5 82 18 Subalpine Fir 4 15.03 24 490 23 Subalpine Fir 5 10.56 24 440 21 Subalpine Fir 6 6.55 12.5 363 18 Subalpine Fir 7 7.64 15. 5 204 20 Subalpine Fir 8 4.98 11.5 121 20 Subalpine Fir 9 5.74 10 121 16 Subalpine Fir 10 3.96 12.5 94 16 Subalpine Fir 11 7.49 9.5 138 14 Subalpine Fir 12 7.89 19 176 18 Whitebark Pine 1 4.90 6 39 9 Whitebark Pine 2 5.97 7.5 113 14 Whitebark Pine 3 2.42 3 21 6 Whitebark Pine 4 2.63 2.5 22 2 Whitebark Pine 5 3.20 8.5 57 12 Whitebark Pine 6 8.42 12 165 12 Whitebark Pine 7 5.02 5 132 6 Whitebark Pine 8 4.22 6 49 5 Whitebark Pine 9 4.63 9 176 9 Whitebark Pine 10 8.65 12 346 15 Wh itebark Pine 11 2.83 7 57 7 Whitebark Pine 12 2.92 6 57 8 Whitebark Pine 13 9.09 16 247 14 Whitebark Pine 14 3.37 11 42 9 Whitebark Pine 15 6.51 10 147 18.5 Whitebark Pine 16 5.72 13 247 13.5 Whitebark Pine 17 7.05 13.5 72 11 Whitebark Pine 18 2.02 6.5 61 5 Whitebark Pine 19 13.52 21.5 785 20 Whitebark Pine 20 1.37 1.5 28 2 Whitebark Pine 21 12.65 18 451 23 b.

PAGE 92

80 II. Small Tree Analyses This appendix shows results from Relative Vigor Study (hypothesis 1) analyses that did not show statistical differe nces among species. Table AII 1 Change in krummholz tree stem diameters The increase in stem diameters (mm) ov er the 2010 to 2012 time period is shown for a, Divide Mountain and b, Line Creek RNA. Analyses were completed w ith a Kruskal Wallis Rank sum test and Wilcox Signed Rank test with a Bonferroni correction. All P values and test statistics shown are from the Wilcox Post Hoc. a. b. Line Creek RNA Increase Spp., Mean (SE) n Species Comparisons W Statistic P Value WP 0.99 ( 0.17) ES 2.31 (0.46) SF 2.58 ( 0.43) 21 20 12 WB < ES WB < SF SF = ES 294.5 34 92.5 0.03 0.0006 0.29 Divide Mountain Increase Spp., Mean (SE) n Species Comparisons W Statistic P Value WP 0.80 ( 0.29) ES 0.98 (0.28) SF 0.77 ( 0.37) 17 15 15 WB = ES WB = SF SF = ES 122.5 149 95 0.86 0.44 0.47 a

PAGE 93

81 Table AII. 2 Krummholz t ree canopy areas Canopy area increases (cm 2 ) from 2010 to 2012 were calculated with a Kruskal Wallis Rank sum test and Wilcox Signed Rank Test with a Bonferroni correction. Results are shown for a, Divide Mo untain and b, Line Creek RNA. All P values a nd test statistics shown are from the Wilcox Post Hoc. Table AII 3 Krummholz tree heights Difference between 2012 and 2010 heights (mm) were calculated with a Kruskal Wallis Rank sum test and Wilcox Sign ed Rank Test with a Bonferroni correction. Results are shown for a, Divide Mountain and b, Line Creek RNA. All P values and test statistics shown are from the Wilcox Post Hoc. a. b. a. b. Divide Mountain Area Spp., Mean (SE) n S pecies Comparisons W Statistic P Value WP 56.3 ( 32.8) ES 1.65 ( 0.76) SF 2.3 ( 63.5) 17 15 15 WB > ES WB = SF SF = ES 140 103 20 0.02 0.95 3.6e 4 Line Creek RNA Area Spp., Mean (SE) n Species Comparisons W Statistic P Value WP 196.6 (29.9) ES 121.3 (18.5) SF 172.6 (38.4) 21 20 12 WB = ES WB = SF SF = ES 272 138 153 0.11 0.67 0.21 Line Creek RNA Height Spp., Mean (SE) n Species Comparisons W Statistic P Value WP 3.66 ( 0.54) ES 4.3 ( 0.69) SF 1.88 ( 0.77) 21 20 12 WB = ES WB = SF SF = ES 204.5 168 163 0.90 0.12 0.10 Divide Mountain Height Spp., Mean (SE) n Species Comparisons W Statistic P Value WP 1.06 (0.74) ES 0.52 ( 0.37) SF 1.40 ( 0.52) 17 1 5 15 WB = ES WB = SF SF = ES 140 103 20 0.78 0.91 0.56 a.

PAGE 94

82 III. Planting Study Microsite Heights This appendix shows the heights for the seedling planting and seed sowing sites by microsite type Table AIII 1 Planting and sowing study m icrosite heights The mean and standard deviation (cm) are reported for each microsite type by seed ling and seed sites at a. Divide Mountain and b. Line Creek RNA All sample sizes are 20 sites per microsite for both seeds and seedlings at each study area. Divide Mountain Site Type Whitebark Spruce Rock Open Seedlings Seeds 19.0 (6.4) 14.1 (5.0) 20. 9 (6.9) 13.4 (4.3) 15.6 (5.8) 9.9 (2.7) n/a n/a Line Creek RNA Site Type Whitebark Spruce Rock Open Seedlings Seeds 44.3 (10.7) 27.2 (10.8) 44.8 (12.5) 28.6 (8.2) 10.6 (3.8) 7.2 (2.4) n/a n/a a. b.