Citation
A greenhouse study of the effects of cache size and relatedness on seed germination, seedling survival, robustness, and root grafting in whitebark pine

Material Information

Title:
A greenhouse study of the effects of cache size and relatedness on seed germination, seedling survival, robustness, and root grafting in whitebark pine
Creator:
Powell, Mary L
Publication Date:
Language:
English
Physical Description:
vii, 50 leaves : illustrations ; 29 cm

Subjects

Subjects / Keywords:
Germination ( lcsh )
Seeds -- Physiology ( lcsh )
Trees -- Growth ( lcsh )
Whitebark pine ( lcsh )
Germination ( fast )
Seeds -- Physiology ( fast )
Trees -- Growth ( fast )
Whitebark pine ( fast )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Bibliography:
Includes bibliographical references (leaves 46-50).
General Note:
Submitted in partial fulfillment of the requirements for the degree, Master of Arts, Biology.
Statement of Responsibility:
by Mary L. Powell.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
37306887 ( OCLC )
ocm37306887
Classification:
LD1190.L45 1996m .P69 ( lcc )

Full Text
A GREENHOUSE STUDY OF THE EFFECTS OF CACHE SIZE AND
RELATEDNESS ON SEED GERMINATION, SEEDLING SURVIVAL,
ROBUSTNESS, AND ROOT GRAFTING IN WHITEBARK PINE
by
Mary L. Powell
B. S., University of Colorado at Denver 1985
B. A., University of Colorado at Denver 1990
A thesis submitted to the
University of Colorado at Denver '
in partial fulfillment
of the requirements for the degree of
Master of Arts
Biology
1996


This thesis for the Master of Arts
degree by
Mary L. Powell
has been approved
by
Diana F. Tomback
James Koehler
Leo P. Bruederle
Date


Powell, Mary L. (M. A., Biology)
A Greenhouse Study of the Effects of Cache Size and Relatedness on Seed
Germination, Seedling Survival, Robustness, and Root Grafting in Whitebark Pine
Thesis directed by Professor Diana F. Tomback
ABSTRACT
Whitebark pine (Pinus albicaulis Englm.) relies on the Clarks nutcracker
(Nucifraga columbiana Wilson) for seed dispersal. Nutcrackers place between 1 and
15 or more seeds (x = 3 or 4) in subterranean caches for later retrieval. Seeds in
unretrieved caches may germinate, resulting in whitebark pine regeneration. Because
more than one seedling from a cache may survive to maturity, whitebark pine often
exhibits a clustered growth form of two or more genetically distinct individuals.
Occasionally, cluster mates form root grafts. Because nutcrackers will harvest and
cache together seeds from one stem or several stems in one tree clump, seeds in
caches may be related. Interest in managed whitebark pine regeneration has increased
and so, in an effort to improve whitebark pine silvicultural techniques, I investigated
m


influence of cache size and relatedness among seeds in a cache on: 1) seed
germination success, 2) seedling survival and robustness (using number of adult
fascicles and seedling height as measures), and 3) root grafting.
I collected whitebark pine seeds from northwestern Wyoming. The seeds
were sorted by gravity and stratified twice to maximize percent germination. I then
planted the seeds in a greenhouse and monitored them for 255 days. Seeds were
planted in five treatments: singles, eight-related, eight-unrelated, four-related, and
four-unrelated. At the end of the study, caches with eight seeds had significantly
higher germination success (p < 0.05). Caches with four and eight seeds showed
interaction effects between the factors of size and relatedness (p = 0.004). Caches
with single seeds had a significantly higher percent survival (p < 0.05). There were
no significant differences in the numbers of adult fascicles among the treatments.
However, unrelated seedlings in groups of eight were significantly taller than related
seedlings in groups of eight (p < 0.05). I found no evidence of root grafting.
The results suggest trade-offs: in a greenhouse situation, planting seeds in
large caches (related or unrelated) will lead to a higher percent of germinated seeds.
However, caches with single seeds may have a higher percent of survivors. Whether
these effects carry over into natural environments merits further study.
This abstract accurately represents the content of the candidates thesis. I recommend
its publication.
Signe
IV


For
Marci L. Eads
Though I do not believe
that a plant will spring up
where no seed has been,
I have great faith in a seed.
Convince me that you have a seed there,
and I am prepared to expect wonders.
Henry D. Thoreau


CONTENTS
Acknowledgments......................................vii
CHAPTER
1. INTRODUCTION....................................1
2. METHODS....................................... 7
Experimental Design............................7
Cone Collection................................7
Seed Preparation.............'...............10
Planting......................................11
Data Collection............................. 12
Field Data....................................13
Data Analysis...,.............................16
3. RESULTS........................................18
Germination Success......................... 18
Seedling Survival.............................22
Robustness....................................29
Root Grafting.................................32
Field Study...................................32
4. DISCUSSION.....................................37
BIBLIOGRAPHY...............................................46
vi


ACKNOWLEDGMENTS
The assistance of several people has helped to make this project possible. I
would like to thank the members of my committe: Leo Bruederle, Jim Koehler, and
Diana Tomback, my committe chair. I would also like to thank Lisa Torick for her
help in the field and in the greenhouse, Marci Eads for her help in the greenhouse and
with editing, and Kathy Carsey and Karen Baud for being excellent sounding boards.
Ward McCaughey of the U. S. Forest Service provided advice on greenhouse
techniques. Finally, I would like to thank Jackie Powell and Gary Higgins for their
financial support of my field work, etc.
vii


CHAPTER 1
INTRODUCTION
Whitebark pine (Pinus albicaulis) populations are declining in the northern
Rocky Mountains as a result of past fire suppression practices, mountain pine beetle
{Dendroctonus ponderosae) epidemics, and, more importantly, infestation by an
introduced pathogen, white pine blister rust (Cronartium ribicola) (Kendall and Amo
1990). Because the pine is a keystone species with respect to a variety of wildlife,
including many birds and mammals and the grizzly bear (Ursus arctos) (Kendall and
Amo 1990, Mattson and Reinhart 1994), the possible decline of the whitebark pine
ecosystem in parts of its range has generated serious concern. As a result,
management intervention may be both appropriate and necessary to supplement and
hasten natural regeneration processes. Sowing seeds or planting seedlings may be an
important option, particularly where seed sources have been devastated by blister rust,
clearcutting, or prescribed or natural bums (Hoff and Hagle 1990).
Several facts about the biology of whitebark pine suggest that modifications to
conventional silvicultural techniques may improve the success of managed
regeneration efforts in mimicking the growth form and population structure of
1


whitebark pine. In this study, I explore a method that has potential for improving
germination success and seedling survival.
Wind dispersal of the large, wingless seeds of whitebark pine cannot occur.
Rather, whitebark pine depends on a bird, the Clark's nutcracker (Nucifraga
columbiana), for seed dispersal (Tomback 1978,1981,1982; Lanner 1980; Hutchins
and Lanner 1982). Clark's nutcrackers bury whitebark pine seeds under 2 to 3 cm of
soil or litter in caches consisting of 1 to 15 or more seeds (x = 3 or 4 seeds per cache)
(Tomback 1982). Cache sites are typically on south-facing slopes and in open areas
such as bums or rocky slopes (Tomback 1982, Hutchins and Lanner 1982).
Nutcrackers are known to transport whitebark pine seeds as far as 12 km to cache
sites (Tomback 1978,1982; Hutchins and Lanner 1982). Nutcrackers recover these
caches during winter and spring months as an important food source (Tomback 1978),
but unretrieved caches may germinate and produce clusters of seedlings (Tomback
1982).
In addition to whitebark pine, there are at least seven other species of large,
wingless-seeded pines known to be regularly dispersed by nutcrackers (Nucifraga
ssp.) (Tomback and Linhart 1990). Of those pines studied, several species show
similarities in life history and population structure. For example, in whitebark pine,
limber pine (P. flexilis), and Swiss stone pine (P. cembra), mature trees may have
either a single trunk or may consist of clumps of two or more trunks fused at the base.
Tomback and Linhart (1990) determined that as many as 58% of the mature tree sites
in a Palmer Mountain whitebark pine stand (Absaroka Range) and 90% of the tree
sites on Mammoth Mountain (Sierra Nevada) supported tree clumps.
2


Genetic analyses of the three species indicate that trees in clumps consist of
either single-genet, multi-trunk trees or multi-genet tree clusters (Linhart and
Tomback 1985; Fumier et al. 1987; see Tomback et al. 1990 for growth form
terminology; Schuster and Mitton 1991; Tomback et al. 1993a; Carsey and Tomback
1994). The multi-genet cluster growth form is the result of germination and survival
of more than one seedling in a nutcracker cache. Estimates indicate that tree clusters
may account for about 20% of the tree sites in stands of limber pine (Carsey and
Tomback 1994) and 13% to 21% of tree sites in Swiss stone pine (Tomback et al.
1993a). Tomback et al. (1990) and Tomback et al. (1993b), when examining
regeneration in two large subalpine bums, determined that whitebark pine seedlings
occurred in clusters at 48% and 43%, respectively, of the total whitebark pine
regeneration sites encountered on study plots. Linhart and Tomback (1985) found that
five of six (83%) mature whitebark pine clumps consisted of multiple genets, while
Fumier et al. (1987) found that 23 of 35 (66%) clumps consisted of multiple genets.
In addition, since nutcrackers tend to harvest a number of seeds from the same
cone and from several cones on the same tree, it follows that seeds within a
nutcracker cache may be related as half or full siblings or as selfed seeds (Tomback
1988; Tomback personal communication). Thus, seedlings and mature trees in
clusters may be closely related. Two studies of limber pine indicate that members of
tree clusters are closely related, on average as half-siblings, with a coefficient of
relatedness of r = 0.19 (Schuster and Mitton 1991), and as between half and full
siblings, with a coefficient of relatedness of r = 0.43 (Carsey and Tomback 1994). In
addition, Schuster and Mitton (1991) found that the more closely related the cluster
members, the greater the tendency to root graft.
3


Root grafting occurs when secondarily produced tissues such as periderm,
secondary xylem, and secondary phloem, fuse (Raven et al. 1992). Secondary growth
typically begins within the first or second growing season of a tree, although grafting
may not occur until years later when the roots of adjacent trees become tightly
crowded. Schuster and Mitton (1991) found that a limber pine cluster cross-sectioned
for aging had experienced eight years of growth before grafting. Root grafting has
been documented in over 150 species of woody trees, including gymnosperms and
angiosperms (Graham and Bormann 1966). Root grafting may provide advantages to
grafted individuals through the translocation of nutrients and strengthened root
structure (Keeley 1988 and Bormann 1966, but see Loehle and Jones 1990).
Additionally, if grafting between related individuals commonly results in an
individual surviving and reproducing that otherwise would have died, it is possible for
kin selection to occur (Hamilton 1963). Schuster and Mitton (1991) present evidence
that the requirements for kin selection to occur may be present in limber pine. If
grafting increases inclusive fitness and has a heritable component, then the trait could
be spread in populations where related seeds are cached together and the resulting
individuals have direct contact, since grafting occurs in limber pine more frequently
between more closely related individuals.
In addition to affecting growth form, dispersal of seeds by nutcrackers also
affects pine population structure. Because nutcrackers may harvest and cache
together seeds from the same cone or tree, individuals in tree clusters are genetically
more similar to cluster mates than to individuals in adjacent tree sites (Fumier et al.
1987, Schuster and Mitton 1991, Carsey and Tomback 1994). This population
structure differs from that of wind-dispersed pines which tends to show a family
4


structure in which trees in close proximity are more closely related to each other than
to trees at some greater distance (Linhart et al. 1981). This family structure results
from small patches of disturbance, which are suitable for conifer establishment, short
dispersal distances, and seed shadows resulting from prevailing wind patterns
(Linhart 1989, Linhart et al. 1981).
Lanner (1980), Tomback (1983), and Tomback and Linhart (1990) have
discussed the evolutionary origins and implications of the nutcracker-pine interaction.
An important possible result of that interaction is the ability of whitebark pine to
grow in tree clusters. Tomback and Linhart (1990) suggest that in the stress-tolerant
whitebark pine, tree clusters may be more common in the harsh, xeric conditions of
the subalpine and treeline. Slow growth under these conditions may prevent one
seedling from out-competing others in the same cache. They note that whitebark pine
is found growing in the multi-trunk growth form more frequently in upper subalpine
stands than in lower-subalpine stands. They additionally suggest that tolerance to
crowding could have been selected for over time as a result of higher reproductive
fitness of crowd-tolerant genotypes in nutcracker caches.
Tolerance to crowding would seem to be advantageous, since the majority of
nutcracker caches contain more than one seed (Tomback 1982, Hutchins and Lanner
1982) and bird-dispersed pines frequently grow in multi-seedling clusters (Tomback
et al. 1990, Tomback et al. 1993b). So, has selection resulted in genotypes that are
favored during germination and growth under these conditions? Does changing this
common condition (related seeds in multi-seed caches) change the likelihood of
germination, survival, or root grafting? No study to date has focused on these
questions.
5


In this study, I investigate whether relatedness of seeds and number of seeds in
a cache affect percent seed germination, seedling survival, seedling robustness, and
root grafting. I conducted a greenhouse study from July, 1993, to June, 1994 to
examine these variables. I created five treatments: caches with single seeds, with four
unrelated seeds, with four related seeds, with eight unrelated seeds, and with eight
related seeds. My study was designed to investigate the following questions: Are
both germination success and seedling survival greater in multi-seed caches than in
single seed caches, and are both greatest in larger multi-seed caches? Are
germination success and seedling survival higher in caches containing related seeds
than in caches containing unrelated seeds? Is seedling robustness (as measured by
height and number of adult fascicles) greater in caches with related seeds than in
caches with unrelated seeds? And, finally, does root grafting occur more readily
among related seedlings than among unrelated seedlings?
Additionally, if seed germination and seedling survival in natural conditions
are higher in multi-seed caches, I would expect the average frequency of regeneration
sites with clusters of two or more seedlings to be higher than the average frequency of
caches in which nutcrackers place two or more seeds. Therefore, in addition to the
greenhouse study, I collected data in the field on the frequency of the different
numbers of seedlings per regeneration site for seedlings from one to four years old. I
then compared these data to data on the distribution of the number of seeds per cache
stored by nutcrackers (Hutchins and Lanner 1982).
6


CHAPTER 2
METHODS
Experimental Design
The greenhouse experimental design was based on two factors: cache size and
relatedness. The first factor, cache size, consisted of caches of 1,4, or 8 seeds per
cache. Cache sizes of 4 and 8 had the additional factor of relatedness, for a total of
five treatments. In related caches, all seeds came from cones from a single trunk (one
genet). Thus, these seeds were related as half or full siblings or, in the case of self-
pollination, were more related. Each seed in the unrelated caches came from a cone
of a different tree clump or a single-stemmed tree. Each treatment was replicated 50
times for a total of 250 caches.
Cone Collection
Whitebark pine cones were collected September 6 to 8,1992.1 collected as
late in the season as possible to allow for maximum embryo development while
7


avoiding excessive seed depletion by nutcracker foraging. The collection site was in
a roadside stand of whitebark pine at an elevation of about 2798 meters on Union
Pass in the Wind River Range, Shoshone National Forest, Wyoming (Fig. 2.1).
I selected 50 clumps of whitebark pine, i.e. trees with multiple stems, as
candidates for cone collection. The number of stems per clump ranged from two to
eight. I removed as many cones as possible from each stem with a 3-meter pruning
pole. A numbering system was used to keep track of cones from each tree clump and,
within a clump, each stem. The cones from each stem were put into a paper bag that
was marked with the stem and clump numbers. After cones had been collected from
all the stems in a clump, the separate stem bags were placed in a larger paper bag
designated with the individual clump number.
8


Moran 82 kMomotara
CONE COLLECTION SITE

0 .25 .5 1
KILOMETER
Figure 2.1. Map of the whitebark pine cone collection site on Union Pass in the
Wind River Range, Shoshone National Forest, Wyoming, U.S.A.
9
Dubois 11 kllometora


Seed Preparation
I used the following seed handling techniques in order to maximize
germination success. Following cone collection, I removed seeds from cones. Using
the cyclohexane method of sink or float, I discarded the floaters, which have been
shown to have a high percent of empty seed coats and underdeveloped embryos (Pitel
and Wang 1990). A subsample of 303 seeds was weighed to determine the overall
distribution of seed weights and to document the weight distribution cut-off for
discarded seeds. The seed weights were normally distributed, with the floating seeds
weighing significantly less than those that sank (two sample t-test, t = 3.75, p
=0.0002, df =' 239). The floaters were cracked open to determine the condition of the
discarded seeds. Fifty-six percent of the embryos were classified as shriveled, 32%
were partially shriveled, and 11% were full.
I divided the seeds that sank into planting treatments after the cyclohexane
separation process. In order to ensure adequate numbers of related seeds, I first
created caches for the treatment with eight related seeds (8R). Using a computer-
generated random numbers table, I selected one stem from fifty different clumps. For
each cache, eight seeds from one stem were put into cheese cloth bundles and labeled.
The next treatment I created consisted of four related seeds per cache (4R). Again,
50 unrelated stems were randomly chosen and four seeds were taken from each. The
seeds were bundled and labeled. The next treatment I created consisted of eight
unrelated seeds per cache (8U). Because of the possibility that different stems within
a clump could be genetically identical, or one genet (Schuster and Mitton 1991,
Carsey and Tomback 1994), the seeds forming the unrelated caches were selected
10


from different clumps. For each cache, I randomly chose eight different clumps from
that to take seeds. I repeated this 50 times. I created the caches with four unrelated
seeds (4U) in the same manner. Finally, I randomly chose 50 single seeds (S) from
50 different clumps.
Before cold stratification, I washed the seeds in a 10% bleach solution to
minimize fungal growth which may adversely affect embryo development or
germination (McCaughey, personal communication). I then rinsed the seeds in
running water for 24 hours.
Caches were cold stratified (0 C) for 45 days in separate small, plastic vials
filled with moist, aerated sand (Feldman 1991; references therein). Water was
periodically added to the vials to ensure adequate moisture. After 45 days, I removed
the caches from the sand and air dried them for 30 days under room temperature
conditions. Finally, in order to maximize embryo development by mimicking a
second winter, the caches underwent a second round of 45-day cold stratification
(McCaughey, personal communication).
Planting
The seeds were removed from the second round of cold stratification and
planted on July 18,1993. Each cache or single seed was planted in a 6 plastic pot
with a substrate of peatmoss and cleaned playground sand (McCaughey, personal
communication). Prior to planting, the peat and sand were mixed thoroughly and
11


placed in pots to within 2 cm of the rim. Water was added to the pots and the
peat/sand mixture was stirred in the pot by hand to allow the peat to be moistened.
The pots were labeled according to treatment and replicate number. Prior to planting,
the radical end of each seed was clipped to aid in germination (Pitel and Wang 1990).
The seeds were then placed into a depression approximately 4 cm deep and covered.
After the caches or single seeds had been planted, the pots were thoroughly watered.
The pots were lightly watered every day for the first two weeks and were watered
thoroughly twice a week thereafter. Because rewetting the peat is difficult, the
peat/sand mixture was kept moist to prevent the mixture from shrinking and pulling
away from the pot surface. Miracid fertilizer was applied according to manufactures
instructions after the first month (McCaughey, personal communication).
The pots were arranged in the greenhouse on two benches oriented north to
south. In order to minimize north to south temperature differences and their attendant
position effects, the pots were lined up in parallel, north to south columns, with one
column for each treatment, for a total of five columns on each bench. Each treatment
had an equal number of pots across the north to south gradient.
Data Collection
Data were collected on germination success and seedling survival once or
twice a week for the first 50 days and then once a month over the remainder of 255
days. Initially, observations included only the number of seedlings in each pot.
12


Beginning 164 days after planting, additional data were collected once a month for
three months on seedling height and number of adult fascicles. In whitebark pine,
adult fascicles are bundles of five leaves that are typically first produced during the
second growing season. Additionally, the seedlings in each pot were mapped so they
could be tracked individually. Height was measured from the soil surface to the tip of
epicotyl growth. Robustness was measured by the number of adult fascicles on each
seedling and by seedling height at the end of 171 days.
After 255 days I examined surviving seedling groups for evidence of root
grafting. The sand and peat (with the seedling group in place) were carefully
removed from the plastic pot. The sand and peat were then gently removed by hand
from the root masses of the seedlings. I made every effort to not pull or shake the
root masses, which might have destroyed any root grafts. The root masses were then
placed under a dissecting microscope. Remaining peat and sand were carefully
removed with a needle probe and tweezers. The roots were sprayed with a mist of
water which made them pliable enough to untangle. Seedlings were then untangled
from the root mass one at a time and set aside.
Field Data
I collected data on the number of seedlings in regeneration sites in two
subalpine zone areas that burned during the Yellowstone Fires of 1988. The first area
was a severely burned forest northeast of Cooke City, Montana, in Gallatin National
13


Forest at the base of Henderson Mountain. The second area was on a severely burned
moist slope on Mount Washburn in north-central Yellowstone National Park,
Wyoming (Fig. 2.2). At each area, I ran a 50 meter tape down the slope through the
burned forest of whitebark pine. I then carefully examined the forest floor for one
meter on each side of the tape. When I located a seedling site, I recorded the number
and age of the seedlings present. After examining a transect, I moved the tape over
five meters and repeated the procedure. I repeated this until I had moved through the
entire stand of trees in each area.
14


Figure 2.2. Maps of areas used for data collection on number of seedlings per
regeneration site. The Mount Washburn area is a severely burned moist forest in
Yellowstone National Park, Wyoming, U.S.A. The Cooke City area is a severely
burned forest in Gallatin National Forest, Montana, U.S.A. Both areas were burned
during the Yellowstone Fires of 1988.
15


Data Analysis
I used one-way analysis of variance to compare germination success, seedling
survival, and both robustness measures among the five treatments. I used two-way
analysis of variance to determine the effect of cache size and relatedness on
germination success and seedling survival in the multi-seed caches. For analysis of
variance tests for germination success and seedling survival, I performed an arcsine
transformation of the square root of percent germination and percent survival to
stabilize variance. I set alpha at 0.05 for all statistical tests.
I defined germination success to be the percent of seeds in each pot that had
germinated by a given data collection date. I calculated descriptive statistics for each
treatment, including mean, range, and standard deviation. I calculated percent
survival of seedlings by dividing the number of seedlings alive in a pot by a given
data collection date by the total number of seedlings that had germinated in the pot. I
again calculated descriptive statistics for each treatment.
When examining germination success, I performed a one-way analysis of
variance on all five treatments and then, for significant findings, performed pairwise
comparisons using Tukeys method of multiple-comparison (Kleinbaum et al. 1988).
I then performed a balanced two-way analysis of variance on the multi-seed caches to
determine if interaction effects were present.
When examining seedling survival, I again performed a one-way analysis of
variance on all five treatments. Because not every pot had germinants, the sample
sizes were unequal. Therefore I used Scheffes method of multiple comparisons to
16


identify significant differences between treatments (Kleinbaum et al. 1988). I then
performed an imbalanced two-way analysis of variance on the multi-seed caches.
I log 10 transformed the data on number of seedlings per regeneration site
collected at the two field study areas to give a normal distribution; I then compared
them with a 2-sample t-test. The data from the two collection sites were combined
and compared to results from a study on the number of seeds cached by nutcrackers
(see results in Hutchins and Lanner 1982). Using Hutchins and Lanners sample
mean and standard deviation, I created a normally distributed data set. A Welchs
two-sample t-test was used to determine whether significant differences existed
between the means of the Hutchins and Lanner study and the means of the field data.
Finally, a goodness of fit test was done to determine if the data sets were comparable.
17


CHAPTER 3
RESULTS
Germination Success
On the third day after planting (the first day of data collection), mean
germination percentages ranged from 0% for the caches with single seeds, to 15.75%
for caches with eight unrelated seeds (Fig. 3.1). Of the total number of seeds that
germinated during the study, ninety-eight percent had germinated by day 77.
Between day 77 and day 171, a total of five seeds germinated, with no germination
from day 171 to the end of the study at day 255.
On day 17, there was already a significant difference in germination success
among treatments (one-way ANOVA, F = 5.04, p = 0.001, df = 4). On day 255,
significant differences remained (one-way ANOVA, F = 11.74, p = 0.000, df = 4).
Table 3.1 provides descriptive statistics for germination percentages on days 17 and
255.
I performed pairwise comparisons to identify specific differences between
treatments in germination after 17 and 255 days (Tukeys method, day 17: Mean
Standard Error (MSE) = 0.206, q = 3.86, p < 0.05; day 255: MSE = 0.270, q = 3.86,
p < 0.05). The results of the pairwise comparisons are summarized graphically in
Figure 3.2.
18


Percent Germination
Days Since Seed Planting
Figure 3.1. Average percent germination over time. Average percent germination
for each treatment was calculated as the cumulative average percent of seeds in each
pot which had germinated by a given data collection point.
19


Table 3.1. Descriptive statistics for average percent germination at day 17 and at day
255. Average percent germination for each treatment was calculated as the
cumulative average percent of seeds in each pot that had germinated by a given
data collection point. Sample size N represents the number of replicates in each
treatment.
Day 17
Treatment N Mean Median Std. Dev. Min. Max.
4R 50 0.2300 0.0000 0.3306 0.0000 1.0000
4U 50 0.0700 0.0000 0.1519 0.0000 0.5000
8R 50 0.2300 0.1250 0.2555 0.0000 1.0000
8U 50 0.3075 0.2500 0.2980 0.0000 1.0000
Single 50 0.1800 0.0000 0.3881 0.0000 1.0000
Day 255
Treatment N Mean Median Std. Dev. Min. Max.
4R 50 0.3450 0.2500 0.3776 0.0000 1.0000
4U 50 0.1550 0.0000 0.2852 0.0000 1.0000
8R 50 0.5025 0.5000 0.3249 0.0000 1.0000
8U 50 0.5575 0.6250 0.2503 0.0000 1.0000
Single 50 0.2800 0.0000 0.4536 0.0000 1.0000
20


1$ it 20 £!f 30 35 40 45 50
AVERAGE PERCENT GERMINATION AFTER 17 DAYS
!§"
"4!"
"55"
5
60
Figure 3.2. Graphical depiction of Tukeys pairwise comparisons of germination
success after 17 (top) and 255 (bottom) days. The circles represent each treatments
average percent germination at the given number of days. The lines around the
treatments indicate treatments that did not differ significantly. For example, after 255
days, 4R, with 34% average germination, did not differ significantly from Singles nor
8R, but did differ significantly from both 4U and 8U.
21


There was a significant difference in germination success at day 255 when the
caches were grouped according to size alone (one-way ANOVA, F = 18.39, p =
0.000, df = 2). The caches with eight seeds had the highest average germination
success, which differed significantly from the single- and four-seeded caches
(Scheffes multiple-comparison technique, n = 250, MSE = 0.278, F = 3.035, k = 3, p
< 0.05). The two way analysis of variance based on caches with four and eight seeds
showed significant interaction between the factors of relatedness and size (two-way
ANOVA, F = 8.68, p = 0.004, df = 1) (Table 3.2).
Seedling Survival
Of seeds that had germinated, the single-seeded caches had 71.43% of
seedlings surviving at the end of 255 days, the highest of all treatments. The percent
of seedlings surviving in each treatment for 8U, 8R, 4U, and 4R are given in Table
3.3.
One-way analysis of variance showed a significant difference among
treatments in percentages of seedlings surviving to day 255 (F = 10.94, df = 4, p =
0.000). Using Scheffes technique of multiple comparisons, I determined that the
seedling survival of the single-seeded caches was significantly higher than all other
treatments (n = 148, k = 5, MSE = 0.204, F = 2.43, p < 0.05). There were no other
significant differences in percent survival between treatments.
22


For the multi-seeded treatments alone, a two-way analysis of variance showed
no significant differences in survival based on the factors of cache size and
relatedness, but did show a significant interaction effect (two-way ANOVA, F = 5.17,
p = 0.0247, df = 1) (Table 3.4).
To understand overall trends in germination, survival and mortality in each
treatment, I divided the number of seedlings that were alive at each data collection
point, by the total number of seeds planted. Figure 3.3 is a graph of the resulting data.
The graph shows a dramatic drop in the percent of seedlings alive after day 171.
Because so many of the seedlings died between days 171 and 255,1 performed a one-
way analysis of variance on the ratio of the number of seedlings alive at day 171 to
the number of seeds which had germinated to determine if significant differences in
survival were present before the high-mortality period. Summary statistics for the
percent of seedling survival by day 171 are contained in Table 3.5. There was no
significant difference in seedling survival among treatments at day 171 (F = 1.97, p =
0.101, df= 4).
23


Table 3.2. Results of two-way analysis of variance of germination success used to
determine the effect of cache size and relatedness on multi-seed caches.
Source DF SS MS F P
Relatedness 1 0.6317 0.6317 3.00 0.085
Cache Size 1 9.0505 9.0505 42.95 0.000
Relate* size (interaction) 1 1.8296 1.8296 8.68 0.004
Error 196 41.3003 0.2107
Total 199 52.8121
24


Table 3.3. Descriptive statistics for average percent of seedlings surviving after 255
days. Percent seedling survival was calculated by dividing the number of
seedlings alive in the pot by a given data collection date by the total number of
seedlings that had germinated in the pot. Sample size N represents the number of
replicates that had seed germination over the course of 255 days.
Treatment N Mean Median Std. Dev. Min. Max.
4R 28 0.2232 0.0000 0.3102 0.0000 1.0000
4U 14 0.1071 0.0000 0.2895 0.0000 1.0000
8R 43 0.1481 0.0000 0.1972 0.0000 0.7500
8U 49 0.2756 0.2500 0.2778 0.0000 1.0000
Single 14 0.7140 1.0000 0.4690 0.0000 1.0000
25


Table 3.4. Results of two-way analysis of variance of seedling survival after 255
days used to determine the effect of cache size and relatedness on multi-seed
caches.
Source DF SS MS F P
Relatedness 1 0.544821 0.544821 3.40 0.0895
Cache Size 1 0.469194 0.469194 2.93 0.0895
Relate* size (interaction) 1 0.828008 0.828008 5.17 0.0247
Error 128 20.517300 0.160292
Total 131 22.033000
26


Ratio of Live Seeds
Figure 3.3. Ratio of live seedlings to seeds planted over time. Percentages were
generated for each treatment by dividing the number of seedlings which were alive at
each data collection point, by the total number of seeds planted.
27


Table 3.5. Descriptive statistics for average percent survival after 171 days. Percent
survival of seedlings was calculated by dividing the number of seedlings alive in
the pot by a given data collection date by the total number of seedlings that had
germinated in the pot. Sample size N represents the number of replicates that had
seed germination over the course of 171 days.
Teatment N Mean Median Std. Dev. Min. Max.
4R 28 0.8539 1.0000 0.2938 0.0000 1.0000
4U 14 0.8329 1.0000 0.3069 0.0000 1.0000
8R 43 0.8015 1.0000 0.2499 0.1666 1.0000
8U 49 0.7454 0.8000 0.2734 0.0000 1.0000
Single 14 1.0000 1.0000 0.0000 1.0000 1.0000
28


Robustness
There were no significant differences in the mean number of adult fascicles
per seedling per replicate among the five treatments (one-way ANOVA, F = 0.45, p
0.771, df = 4) (Table 3.6). However, there was a significant difference in height
among treatments (one-way ANOVA, F = 2.73, p = 0.031, df = 4) (Table 3.7). The
significant difference was between caches with eight seeds; caches with eight
unrelated seeds were significantly taller than caches with eight related seeds
(Scheffes pairwise comparisons, n = 145, k = 5, MSE = 24.0, F = 2.40, p < 0.05).
29


Table 3.6. Summary statistics for mean number of fascicles at day 255. The average
number of fascicles was calculated for each replicate and each replicate was
treated as a unit during summary statistics calculations. Sample size N represents
the number of replicates that had seed germination over the course of 255 days.
Treatment N Mean Median Std. Dev. Min. Max.
4R 26 0.2820 0.0000 0.7720 0
4U 13 0.2120 0.0000 0.5580 0 2
8R 43 0.2216 0.0000 0.3424 0 1
8U 49 0.3358 0.0000 0.6398 0 J
Single 14 0.1429 0.0000 0.3631 0 1
30


Table 3.7. Summary statistics for mean per replicate height after 255 days. Height
in millimeters was measured from the soil surface to the tip of epicotyl growth.
An average height was calculated for each replicate and each replicate was treated
as a unit during summary statistics calculations. Sample size N represents the
number of replicates that had seed germination over the course of 255 days.
Treatment N Mean Median Std. Dev. Min. Max.
4R 26 29.179 29.750 4.609 18.333 37.500
4U 13 29.400 29.000 4.380 23.000 38.500
8R 43 27.888 29.000 5.634 14.500 40.375
8U 49 31.238 31.000 4.542 19.500 42.000
Single 14 29.640 30.500 4.580 22.000 37.000
31


Root Grafting
Many roots were observed to be growing parallel to one another. In some
cases, the adjacent surfaces of two parallel roots were flattened to accommodate each
other. The flattened surfaces were tightly pressed together giving the appearance of a
single root. Root nodules and root hairs were closely entwined, adding to the
appearance of being connected. Despite the close proximity, no root grafting was
observed.
Field Study
I found no significant difference between the moist and dry forest areas in the
number of seedlings per regeneration site (two sample t-test, t = -0.08, p = 0.94, df=
68). I also found no significant difference in number of seedlings per cluster when
grouped by seedling the ages of 1,2,3, and 4 years (one-way analysis of variance, F
= 1.67, p = 0.181, df = 3). The average number of seedlings per seedling site for both
sites combined was 4.816 (SD = 4.068, n = 76) (Table 3.8 and Fig. 3.4).
Hutchins and Lanner (1982) observed nutcrackers caching a mean of 3.2 (SD
= 2.8, n = 157) seeds per cache. The average number of seedlings that I found per
seedling site is significantly higher than the average number of seeds per cache
32


observed by Hutchins and Lanner (Welchs two-sided t-test, t = -3.093, df = 110, p =
0.0025). However, the results of the goodness of fit test showed that while both
distributions were geometric ( for seedling data: = 5.28, df = 7, p = 0.626; for
caching data: x^ = 3.088, df = 2, p = 0.214), the seedling data were shifted too far to
the right to be comparable with the caching data (Fig. 3.5). More seedlings per cache
had germinated at my field data collection sites than Hutchins and Lanner had
observed being cached at their study site.
33


Table 3.8. Descriptive statistics for seedlings per regeneration site for two collection
sites. Cooke City is a severely burned dry site. Mount Washburn is a severely
burned moist site. Sample size N represents the number of regeneration sites
encountered.
Site N Mean Median Std. Dev. Min. Max.
Cooke City 35 4.971 4.000 4.469 1 20
Mt. Washburn 41 4.683 3.000 3.745 1 15
34


Number of Regeneration Sites
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Number of Seedlings
Figure 3.4. Distribution of number of seedlings per regeneration site (n = 76). The
data resulted from the combined samples from the Cooke City and Mount Washburn
study areas.
35


Percent of Observations
Figure 3.5. Comparison of percent of observations for four categories of data on
number of seedlings per regeneration site and on number of seeds placed by
nutcrackers per cache. Data for caching observations are from Hutchins and Lanner
(1982).
36


CHAPTER 4
DISCUSSION
In the interest of whitebark pine conservation, researchers and forest managers
have begun to study not only the life history of whitebark pine but also silvicultural
techniques (e.g. Proceedings Symposium on Whitebark Pine Ecosystems: Ecology
and Management of a High-Mountain Resource 1990, and Proceedings -
International Workshop on Subalpine Stone Pines and Their Environment: the Status
of Our Knowledge, 1994). Since seeds collected from whitebark pine cones are
generally hard to germinate because of immature embryos and moisture impervious
seed coats (Pitel and Wang 1990), there have been several recent studies concerning
techniques to improve whitebark pine seed germination success and seedling growth.
Pitel and Wang (1990) examined seed stratification and scarification techniques.
Additionally, they tested the effect of gibberelic acid on seed germination. Jacobs
and Weaver (1990) investigated temperature effects on germination and seedling
performance. McCaughey and Weaver (1990) performed a field study, which
examined site preparation, percent shading, and predator exclusion on seed
germination, patterns of emergence, and seedling mortality. Only single seed caches
were sown.
None of the previous studies provided information on the effects of conditions
in which whitebark pine seeds are typically cached, i.e. in multi-seed caches, often
37


with more than one related seed. My study investigated the effects of relatedness and
cache size on germination, seedling survival and robustness, and root grafting. The
results of this study provide somewhat mixed answers to the questions asked in
Chapter 1: 1) are both germination success and seedling survival greater in multi-seed
caches than in single-seed caches, and are both greatest in larger multi-seed caches;
2) are germination success and seedling survival higher in caches containing related
seeds than in caches containing unrelated seeds; 3) is seedling robustness (as
measured by height and number of adult fascicles) greater in caches with related seeds
than in caches with unrelated seeds; and 4) does root grafting more readily occur
among related seedlings than among unrelated seedlings?
The germination results provided the most mixed answers. Caches with eight
seeds had significantly higher germination percentages compared to single-seed and
four-seed caches. The germination percent for the eight-seed caches was greater by a
factor of two than the germination percent for the single-seed caches. There was no
significant difference in germination percentages between the single-seed and four-
seed caches. However, the two-way analysis of variance on the multi-seed caches
showed interaction between the factors of relatedness and cache size. Relatedness in
the eight-seed caches was not a significant factor in percent germination but was
significant for the four-seed caches. Related four-seed caches had significantly higher
germination percentages than caches with four unrelated seeds.
Germination rates for whitebark pine in greenhouse and laboratory situations
are quite poor, ranging from 0% to 90%, but typically around 30-50% (Amo and Hoff
1989, Pitel and Wang 1990, McCaughey, personal communication). Pitel and Wang
(1990) attribute the poor germination rates to immature embryos and to the presence
38


of a hard seed coat. Immature embryos may be encountered when seeds are collected
by researchers and forest managers. Germination in natural conditions may be higher
because nutcrackers are apparently able to discriminate amongst seeds with empty,
diseased, or insect-infested embryos (Tomback 1978). A hard seed coat may cause
seed dormancy by being impermeable to water and/or gases or by mechanically
constraining the embryo (Mayer and Poljakoff-Mayber 1982). Additionally, prior to
germination, whitebark pine seeds require a period of cold temperatures. This very
common requirement in species in temperate climates effectively breaks dormancy
caused by the embryo and by the seed coat (Bewley and Black 1985).
The difference in germination between the eights and the fours and singles
may reflect a hormone threshold effect. For example, a speculative mechanism for
this effect could involve hormone secretion by newly germinated seedlings that
stimulates the remaining seeds to germinate. Growth regulators such as gibberelins
have been shown to stimulate germination in dormant seeds (Kozlowski 1971a,
Bewley 1979, Fenner 1985). While gibberelins have been implicated in metabolic
changes in the embryo during germination, the details of the mechanism are unclear
(Bewley and Black 1985, Raven et al. 1992). However, application of gibberelic acid
to intact whitebark pine seeds did not significantly improve germination success in a
laboratory study (Pitel and Wang 1990). Because Pitel and Wang treated only single,
unclipped seeds, the effect of gibberelic acid on multi-seed caches in natural
conditions remains untested. Alternatively, larger caches may create micro-
environments more favorable to germination. For instance, if larger caches somehow
retained higher levels of moisture by inhibiting water loss, imbibition of water may be
improved and so increase germination. Higher moisture levels might also aid in
39


softening the seed coat and lessening the mechanical constraint on the embryo. It is
unclear why seed relatedness affected seed germination and seedling survival in the
treatments with four seeds but did not affect the treatments with eight seeds.
Although it may be possible that a differential response to stimuli by unrelated
genotypes is overcome by a cache size threshold, I am unable to speculate on a
mechnism.
Results on survival were also mixed. After 255 days, the single-seeded caches
had a significantly higher percent of survivors over the course of the study. In fact,
the single-seed treatment had a percent survival greater by more than a factor of two
than the percent survival of the eight-unrelated treatment. However, at 171 days
(before seedling mortality dramatically increased) there was no significant difference
in seedling survival among the treatments. While this study cannot determine
whether the decline in seedling survival for all but the singles was comparable to
nature or an artifact of the greenhouse (i.e. a poor growing medium or a greenhouse-
specific pathogen), there are several possible causes of the high mortality at the end of
the study. As in natural conditions, if nutrients in the growing medium of sand and
peat moss were lacking or inadequate, seedlings in the multi-seed caches would
exhaust the nutrient supply before single seedlings. Because this possibility was
identified prior to the initiation of the study, I used a commercial fertilizer (Miracid)
as a supplement to watering. It is possible that application rates were too low or that
other nutrients not provided by the fertilizer were limiting. Alternatively, the
conditions in the greenhouse might not have been supportive of the clustered growth
form, but rather favored solitary seedlings. Tomback and Linhart (1990) suggest that
the clustered growth form is most common in stressful environments with their
40


attendant extremely slow growth rates. The greenhouse conditions, which typically
allow for high rates of growth, may have led to increased competition between cluster
members. Finally, seedling survival may have been adversely affected by a lack of
mycorrhizal colonization. Mycorrhizae improve seedling survival and growth by
enhancing uptake of nutrients and water (Bold 1987, Perry et al. 1987). Mycorrhizae
typically colonize soon after seed germination, although differences in growth and
suvival may not be apparent for several weeks (Perry et al. 1987, Trappe 1988, Kropp
and Langlois 1990). In an attempt to inoculate my pots with mycorrhizal fungi, I
added a small amount of duff, gathered from under a stand of whitebark pine, to each
pot on approximately day 30. The inoculation may have been inadequate or poorly
timed.
While the results of the field portion of the study were not quantitatively
comparable to the caching observations (Hutchins and Lahner 1982), several points
can be made. Because of mortality in multi-seedling regeneration sites, I would
expect the percent of regeneration sites with single seedlings to be higher than the
percent of cache sites with single seeds. However, while single-seedling regeneration
sites were most numerous (18 out of 76), they were a smaller percent of the overall
seedling distribution than single-seed caches were of the overall seeds-cached
distribution. The disparity between the regeneration data and caching data has several
possible causes. Nutcracker caching behavior may vary geographically, resulting in
different seed and seedling distributions. Or, it may be that cache sizes are skewed
highly to the right and that observers have seen only part of this distribution.
Alternatively, the larger caches may germinate and survive in relatively higher
numbers than smaller caches.
41


The results of the greenhouse portion of this study suggest that above a certain
size, larger caches have higher percent germination. This tendency may have been
selected over time, because the occurrence of clusters of whitebark pine is a direct
result of nutcrackers dispersing seeds in multi-seed caches. While clusters of
whitebark pine are less common in older stands and in stands with a thicker canopy
(Weaver and Jacobs 1990), they occur with significantly greater frequency than in
non-bird dispersed pines (Carsey and Tomback 1994). The attrition of cluster
members may be partially attributed to intra-cluster competition.
This widespread occurrence of tree clusters raises the question of whether
germination in clusters is advantageous or disadvantageous for seedling and tree
growth, survival, and cone production. Several authors have speculated about the
advantages and disadvantages of the tree cluster growth form (e.g. Tomback and
Linhart 1990, Schuster and Mitton 1991, and references therein). Possible advantages
to growing in a cluster include the development of a sturdy root and shoot structure
(especially if root grafting occurs), which may be better able to resist high winds and
unstable soil (Schuster and Mitton 1991). Kozlowski (1971b) provides numerous
examples of root grafting in pines such as eastern white pine (Pinus strobus L.) and
lodgepole pine (Pinus contorta D.), which allows for various levels of nutrient
transfer between individuals. This transfer may increase the efficacy of nutrient
acquisition by weaker cluster members. By growing in clusters, cross-pollination
may occur readily, reducing levels of self-pollination, particularly in isolated clusters
(Tomback and Linhart 1990).
Despite the possible advantages to a clustered growth form, disadvantages
have been identified as well. Competition for resources among cluster members may
42


lead to mortality of weaker individuals over time. In their study of Pinus flexilis,
Schuster and Mitton (1991) found a significant negative correlation between cluster
age and tree number. Inbreeding may be a widespread problem because of the high
mean relatedness among cluster members (Tomback and Linhart 1990). Feldman
(1991) found that tree radial growth rates and male cone production in three stands
and female cone production in one stand of Pinus flexilis was lower for members of
clusters compared to solitary trees. This corresponds to constrained canopies of
cluster members as noted by Tomback and Linhart (1990) and Schuster and Mitton
(1991). Constrained canopies, and attendant lower cone production, may be the result
of increased competition for resources. Finally, transmission of diseases may be
greater among clustermates (L. Bruederle, personal communication).
Despite roots growing parallel and even molded to one another, I found no
evidence for root grafting. In order for root grafting to occur, the plants must be
producing secondary growth. Trees typically begin producing secondary tissues,
including periderm and secondary phloem and xylem, during their first year of growth
(Kozlowski 1971b and Raven et al. 1992). It is possible that this study did not last
long enough to allow the seedlings to produce secondary tissue. Alternatively, if
secondary tissue was produced, the conditions required for grafting to occur may have
been absent. For example, a threshold of growth pressure or stress may be required
before tissues graft. Future greenhouse studies with field components may show that
soil conditions in the field contribute to grafting (soil flora and fauna, mycorrhizae,
root compaction, etc.).
There was no evidence in this study that seedling relatedness or cache size
consistently affected seedling robustness. However, in natural conditions it seems
43


likely that microsite variables and intra-cluster competition related to cluster size play
a larger role in affecting canopy size or height than the relatedness of cluster
members.
Because of an increased interest in conservation and silvicultural management
of whitebark pine (Kendall 1994, Eggers 1990), there is a need for information on
techniques that will improve the efficacy of managed establishment efforts. Studies
to date include investigations of the effects on whitebark pine of microsite and
predator factors on seed germination (McCaughey and Weaver 1990), of standard
reforestation techniques (Kracht and McCaughey 1990), and of logging on crown
characteristics (Kipfer et al. 1994). Sources of whitebark pine for these studies have
included planting greenhouse-grown seedlings and sowing collected seeds. While my
study focused on greenhouse conditions, I believe the results provide a rationale for
future investigation of the effect of cache size and seed relatedness on the germination
and survival of seeds sown in natural conditions.
The results of this study suggest that, for growing whitebark pine seedlings in
greenhouses, there is a trade-off: higher germination success with large caches or
higher survival with single seed caches. In addition to other commonly accepted
practices such as cold stratification and seed coat clipping, germination percentages
may be improved by planting seeds in large caches. Additional testing of various
cache sizes may determine when or if a threshold effect exists. However, increases in
germination may be offset by lower survival over time in multi-seedling clusters.
Given that large greenhouse operations typically have germination rates between 30-
50%, planting caches of seeds may not significantly improve overall greenhouse
productivity. But, survival of larger caches may be increased if the precipitous
44


mortality in this study was caused by greenhouse conditions and is not typical of
natural conditions.
While my results may have limited application to increasing greenhouse
production of whitebark pine seedlings, they further support that the clustered growth
form may be an important aspect of whitebark pine biology. As discussed in Chapter
1, seed dispersal by nutcrackers results in patterns of genetic population structure and
a clustered growth form that differ from those of most wind-dispersed conifers.
Efforts aimed at augmenting natural regeneration of whitebark pine should take these
factors into account as a means to more accurately emulate natural stands of
whitebark pine.
In operations involving outplanting of greenhouse seedlings, a percent should
be planted as clusters. Ideally, in order to minimize disturbance to root systems and
to maximize acclimation to a clustered situation, the seedlings should be grown as a
cluster from seeds. The percent of clustered seedlings and the number of seedlings
per cluster should follow general patterns of any existing stand of whitebark pine near
the establishment site. For seed-sowing operations, seeds that have been screened to
improve germination success should be sown in caches that reflect known
distributions of the number of seeds in nutcracker caches.
While the relatedness of the seeds was not a clear factor in this study, related
seeds should be used for at least a portion of the caches planted in the greenhouse and
in the field, because outplanting clusters of related seedlings or sowing caches of
related seeds more closely mimics the population structure of naturally occurring
stands of whitebark pine.
45


BIBLIOGRAPHY
Amo, S. F.; and Hoff, R. J. 1989. Silvics of whitebark pine (Firms albicaulis).
General Technical Report. INT-253. Ogden, UT: U. S. Department of
Agriculture, Forest Service, Intermountain Research Station. 1 lp.
Bewley, J. D. 1979. Dormancy breaking by hormones and other chemicals action at
the molecular level. Pages 219-239 in I. Rubenstein, R. L. Phillips, C. I.
Green, and B. G. Gengenbach editors. The plant seed: development,
preservation, and germination. Academic Press, New York, New York, USA.
Bewley, J. D., and Black, M. 1985. Seeds: physiology of development and
germination. Plenum Press. New York. 367 p.
Bold H. C. 1987. Morphology of plants and fungi. Harper and Row, New York, 912
P-
Bormann, F. H. 1966. The structure, function, and ecological significance of root
grafts in Pinus strobus L. Ecological Monographs 36:1-36.
Carsey K.S., and Tomback D.F. 1994. Growth form distribution and genetic
relationships in tree clusters of Pinus flexilis, a bird-dispersed pine. Oecologia
98 (3/4): 402-413.
Eggers, D. E. 1990. Silvicultural management alternatives for whitebark pine. Pages
324-328 in W. C. Schmidt (compiler), Proceedings Symposium on
whitebark pine ecosystems: ecology and management of a high-mountain
resource.
Feldman, R. 1991. Growth form and reproductive output in limber pine: the cost of
mutualism. Denver, Colorado: University of Colorado at Denver, Department
of Biology. 46 p. Thesis.
46


Fenner, M. 1985. Seed ecology. Chapman and Hall, New York, New York, 151 p.
Fumier G. R., Knowles P., Clyde M. A., Dancik B. P. 1987. Effects of avian seed
dispersal on the genetic structure of whitebark pine populations. Evolution
41:607-612.
Graham, B. F., and Bormann, F. H. 1966. Natural root grafts. Botanical Review
32:255-292.
Hamilton, W. D. 1963. The evolution of altruistic behavior. American Naturalist
97:354-356.
Hoff R., Hagle, S. 1990. Diseases of whitebark pine with special emphasis on white
pine blister rust. Pages 179-190 in W. C. Schmidt (compiler), Proceedings -
Symposium on whitebark pine ecosystems: ecology and management of a
high-mountain resource.
Hutchins H. E., Lanner R. M. 1982. The central role of Clarks nutcracker in the
dispersal and establishment of whitebark pine. Oecologia 55:192-201.
Jacobs J., Weaver, T. 1990. Effects of temperature and temperature preconditioning
on seedling performance of whitebark pine. Pages 134-139 in W. C. Schmidt
(compiler), Proceedings Symposium on whitebark pine ecosystems: ecology
and management of a high-mountain resource.
Keeley, J. E. 1988. Population variation in root grafting and a hypothesis. Oikos
52:364-366.
Kendall, K. C., Amo, S. F. 1990. Whitebark pine-An important but endangered
wildlife resource. In Schmidt, W. C. (compiler), Proceedings Symposium on
whitebark pine ecosystems: Ecology and management of a high-mountain
resource. USDA For Serv Intermountain Forest and Range Experimental
Station, Bozeman, Montana
Kendall, K. C. 1994. Whitebark pine conservation in North American national parks.
In Schmidt, W. C.; Holtmeier, Friedrich-Karl, comps. 1994. Proceedings -
international workshop on subalpine stone pines and their environment: the
status of our knowledge; 1992 September 5-11; St. Moritz, Switzerland. Gen
47


Tech. Rep. INT-GTR-309. Ogden, UT: U. S. Department of Agriculture,
Forest Service, Intermountain Research Station, 321 p.
Kipfer, T., Hansen, K., and McCaughey, W. 1994. Competition and crown
characteristics of whitebark pine following logging in Montana, U.S.A. In
Schmidt, W. C.; Holtmeier, Friedrich-Karl, comps. 1994. Proceedings -
international workshop on subalpine stone pines and their environment: the
status of our knowledge; 1992 September 5-11; St. Moritz, Switzerland. Gen
Tech. Rep. INT-GTR-309. Ogden, UT: U. S. Department of Agriculture,
Forest Service, Intermountain Research Station, 321 p.
Kleinbaum, D. G., Kupper, L. L., Muller, K. E. 1988. Applied regression analysis
and other multivariable methods. PWS-Kent Publishing Company, Boston,
Massachusetts, 718 p.
Kozlowski, T. T. 1971a. Growth and development of trees. Volume I: seed
germination, ontogeny, and shoot growth. Academic Press, New York, New
York. 443 p.
Kozlowski, T. T. 1971b. Growth and development of trees. Volume II: cambial
growth, root growth, and reproductive growth. Academic Press, New York,
New York. 514 p.
Kracht, R. L., and McCaughey, W. 1990. Artificial reforestation of whitebark pine.
Pages 369-370 in Schmidt, W. C. (compiler), Proceedings Symposium on
whitebark pine ecosystems: Ecology and management of a high-mountain
resource. USDA For Serv Intermountain Forest and Range Experimental
Station, Bozeman, Montana
Kropp, B. R., and C. G. Langlois. 1990. Ectomycorrhizae in reforestation.
Canadian Journal of Forest Research 20:438-451.
Lanner, R. M. 1980. Avian seed dispersal as a factor in the ecology and evolution of
limber and whitebark pines. Pages 14-48 in Proceedings of sixth North
American forest biology workshop; Edmonton, AB: University of Alberta.
Linhart, Y. B. 1989. Interactions between genetic and ecological patchiness in forest
trees and their dependent species. Pages 1-31 in Bock, J. E. and Linhart, Y. B.
editors. Evolutionary ecology of plants. Westview Press, Boulder, Cplorado.
48


Linhart, Y. B. and Tomback, D. F. 1985. Seed dispersal by nutcrackers causes multi-
trunk growth form in pines. Oecologia 67:107-110.
Linhart, Y. B., Mitton, J. B., Sturgeon, K. B., and Davis, M. L. 1981. Genetic
variation in space and time in a population of ponderosa pine. Heredity
46:407-426.
Loehle, C. and Jones, R. H. 1990. Adaptive significance of root grafting in trees.
Functional Ecology 4:268-271.
Mattson, D. J. and Reinhart, D. P. 1994. Bear use of whitebark pine seeds in North
America. Pages 212-220 in Schmidt, Wyman C and Holtmeier, Friedrich-
Karl, comps. 1994. Proceedings international workshop on subalpine stone
pines and their environment: the status of our knowledge; 1992 September 5-
11; St. Moritz, Switzerland. Gen Tech. Rep. INT-GTR-309. Ogden, UT: U.
S. Department of Agriculture, Forest Service, Intermountain Research Station,
321 p.
Mayer, A. M., and Poljakoff-Mayber, A. 1982. The germination of seeds. Pergamon
Press, New York, New York, 211 p.
McCaughey, W. W. and Weaver, T. 1990. Biotic and microsite factors affecting
whitebark pine establishment. Pages 140-150iw W. C. Schmidt (compiler),
Proceedings Symposium on whitebark pine ecosystems: ecology and
management of a high-mountain resource.
Perry, D. A., Molina, R., and Amaranthus, M. P. 1987. Mycorrhizae,
mycorrihizospheres, and reforestation: current knowledge and research needs.
Canadian Journal of Forest Research 17:929-940.
Pitel, J. A. and Wang, B. S. P. 1990. Physical and chemical treatments to improve
germination of whitebark pine seeds. Pages 130-133 in W. C. Schmidt
(compiler), Proceedings Symposium on whitebark pine ecosystems: ecology
and management of a high-mountain resource.
Raven, P. H., Evert, R. F., Eichom, S. E. 1992. Biology of plants, 5th ed. Worth
Publishers, New York, New York. 791pp.
Schuster, W. S. F. and Mitton, J. B. 1991. Relatedness within clusters of a bird-
dispersed pine and the potential for kin interactions. Heredity 67:41 -48.
49


Tomback, D. F. 1978. Foraging strategies of Clarks nutcracker. The Living Bird
16:123-161.
Tomback, D. F. 1981. Notes on cones and vertebrate-mediated seed dispersal of Pinus
albicaulis (Pinaceae). Madrono 28:91-94.
Tomback, D. F. 1982. Dispersal of whitebark pine seeds by Clarks nutcracker: a
mutualism hypothesis. Journal of Animal Ecology 51:451 -467.
Tomback, D. F. 1983. Nutcrackers and pines: coevolution or coadaptation? M. H.
Nitecki, editor. Coevolution. University of Chicago Press, Chicago, Illinois.
Tomback, D. F. 1988. Nutcracker-pine mutualisms: multi-trunk trees and seed size.
Pages 518-527 in H. Ouellet (editor), Acta XIX congressus internationalis
ornithologici, Volume 1. University of Ottawa Press, Ottawa.
Tomback, D. F., Hoffman, L. A., Sund, S. K. 1990. Coevolution of whitebark pine
and nutcrackers: Implications for forest regeneration. Pages 118-129 in W.
C. Schmidt (compiler), Proceedings Symposium on whitebark pine
ecosystems: ecology and management of a high-mountain resource.
Tomback, D. F. and Linhart, Y. B. 1990. The evolution of bird-dispresed pines.
Evolutionary Ecology 4: 185-219.
Tomback, D. F., Holtmeier, F.-K., Mattes, H., Carsey, K. S., and Powell, M. L.
1993a. Tree clusters and growth form distribution in Pinus cembra, a bird-
dispersed pine. Arctic and Alpine Research 25:374-381.
Tomback, D. F., Sund, S. K., and Hoffman, L. A. 1993b. Post-fire regeneration of
Pinus albicaulis: height-age relationships, age structure, and microsite
characteristics. Canadian Journal of Forest Research 23:113-119.
Trappe, J. M. 1988. Lessons from alpine fungi. Mycologia 80:1-10.
Weaver, T., and Jacobs, J. 1990. Occurrence of multiple stems in whitebark pine.
Pages 156-159 in W. C. Schmidt (compiler), Proceedings Symposium on
whitebark pine ecosystems: ecology and management of a high-mountain
resource.
50