Citation
Simulating Clark's Nutcracker caching behavior

Material Information

Title:
Simulating Clark's Nutcracker caching behavior germination and predation of seed caches
Creator:
Baud, Karen Sharer
Publication Date:
Language:
English
Physical Description:
viii, 67 leaves : map ; 29 cm

Subjects

Subjects / Keywords:
Corvidae -- Behavior -- Colorado ( lcsh )
Pine -- Seeds ( lcsh )
Germination ( lcsh )
Corvidae -- Behavior ( fast )
Germination ( fast )
Pine -- Seeds ( fast )
Colorado ( fast )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Bibliography:
Includes bibliographical references (leaves 60-67).
General Note:
Submitted in partial fulfillment of the reqirements for the degree, Master of Arts, Biology.
General Note:
Department of Integrative Biology
Statement of Responsibility:
by Karen Sharer Baud.

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:
30834696 ( OCLC )
ocm30834696
Classification:
LD1190.L45 1993m .B38 ( lcc )

Full Text
SIMULATING CLARKS NUTCRACKER CACHING BEHAVIOR:
GERMINATION AND PREDATION OF SEED CACHES
by
Karen Sharer Baud
B.A., Oberlin College, 1987
A thesis submitted to the
Faculty of the Graduate School of the
University of Colorado at Denver
in partial fulfillment
of the requirements for the degree of
Master of Arts
Biology
1993
r
a


This thesis for the Master of Arts
degree by
Karen Sharer Baud
has been approved for the
Department of
Biology
by
Diana F. Tomback

Date


Baud, Karen Sharer (M.A., Biology)
Simulating Clarks Nutcracker Seed Caching Behavior: Germination and
Predation of Nutcracker Seed Caches
Thesis directed by Associate Professor Diana F. Tomback
ABSTRACT
Clarks Nutcracker (Nucifraga Columbiana) is an important seed-disperser
for several pines with large, wingless seeds, and may contribute to the dispersal of
wind-dispersed seeds. This study examines the relationship between the cache site
preferences of nutcrackers and the seed germination success and losses to rodent
seed predators of limber pine (Pinus flexilis), ponderosa pine (Pinus ponderosa),
and bristlecone pine (Pinus aristata) seed caches.
I observed nutcracker foraging and caching behavior in the Front Range of
the Colorado Rocky Mountains. Seed germination studies were conducted in the
Mount Evans area of the Arapaho National Forest, Colorado, at study areas in the
montane forest, subalpine forest, and alpine tundra. I made simulated nutcracker
caches of limber pine seeds at montane forest, subalpine forest and alpine tundra
study areas, ponderosa pine seeds at two montane forest study areas, and
bristlecone pine seeds at the alpine tundra study area. At each study area twenty-
seven caches of each pine seed species were made at microsites in the open, near
rocks, near logs and near trees. Caches were made in the fall of 1991 and
iii


examined for germination or predation the following summer. Rodents were
trapped in the spring and early summer of 1992 at each site to assess their relative
densities.
Nutcrackers cached ponderosa pine seeds in montane forests and
bristlecone and limber pine seeds above timberline. Germination success was
highest for limber pine caches at the alpine study area (70%), and lowest at the
montane study area (28%). Predation of limber pine caches was lowest at the
alpine study area (13%), and highest at the montane study area (60%).
Differences in seed predation among microsites and among seed species were less
apparent. Rodent densities corresponded to seed predation rates at different study
areas.
This study demonstrates that small, winged seeds as well as large, wingless
seeds dispersed by nutcrackers in caches can germinate. Nutcrackers may also
contribute to the expansion of tree-limit if environmental conditions become
suitable. The importance of seed predation seems to vary with respect to
vegetation zone. The density of seed caching rodents predicted relative seed losses
to predation at the different study sites.

IV


This abstract accurately represents the content of the candidates thesis. I
v


DEDICATION
To David, for his continual love, support and encouragement.


ACKNOWLEDGEMENTS
I wish to thank all the friends and family members who so willingly
assisted me in the field and provided me with unending encouragement. My field
study was possible thanks to help from Jon Krauss, Mary Powell, Lisa Torick, Jim
Jenson, Steve Bousquin, Ron Espinoza, Randy Carver, Andrea Adamo, Kirsten,
Dave, and Jean Sharer, and my most faithful field assistant and husband, David
Baud. I am especially grateful to the members of my faculty committee. I am
fortunate to have had the opportunity to learn from Emily Hartman, for whom I
have the deepest respect. Her encouragement and expertise have been invaluable
to me. I thank Jim Halfpenny for teaching me mammal trapping techniques,
making helpful suggestions, and being on my graduate committee, despite the fact
that he didnt know me. My advisor, Diana Tomback, has guided me throughout
my graduate studies and has provided me with a solid background in Ecology. I
thank Jim Koehler, who helped me immensly with my statistical analysis. Lisa
Torick and I worked together on occasion, and I thank her for her cooperation and
companionship. I am grateful for the help and encouragement I received from
Charlie Ferguson and Gerry and Terry Audesirk. Finally, 1 thank the staff of the
Clear Creek and Boulder Ranger Districts, as well as the Colorado Division of
Wildlife, for granting me the necessary permits to conduct my study.
vii


CONTENTS
CHAPTER
1. INTRODUCTION........................................... 1
2. METHODS................................................ 6
Study Areas........................... ............. 6
Nutcracker Foraging and Caching Observations ....... 10
Seed Germination Study ............................. 10
Rodent Community Diversity and Relative Density 15
Statistical Analysis................................ 16
3. RESULTS .............................................. 20
Nutcracker Foraging and Caching..................... 20
Seed Germination Study ............................. 22
Germination Success .......................... 22
Survivorship ................................. 31
Seed Predation................................ 35
Rodent Community Diversity and Relative Density 42
4. DISCUSSION ........................................... 47
5. CONCLUSIONS ......................................... 59
LITERATURE CITED.............................................. 60
viii


CHAPTER I
INTRODUCTION
Substantial research in the past fifteen years indicates that Clarks
Nutcracker (Nucijraga Columbiana) is an important seed disperser for several
western montane pine species with large, wingless seeds that it eats and stores for
winter food (see Tomback 1994 for review). These pines include whitebark (Pirns
albicaulis) (Tomback 1978, 1981, 1982, Lanner 1980, Hutchins and Lanner
1982), limber (Pinus flexilis) (Lanner 1980, Lanner and Vander Wall 1980,
Tomback and Kramer 1980), and the pinyon pines (Pinus edulis and Pinus
monophylla) (Vander Wall and Baida 1977, Ligon 1978, Tomback 1978, Lanner
and Vander Wall 1980, Vander Wall 1988). There is strong evidence that the
relationship between Clarks Nutcracker and whitebark pine is both mutualistic and
coevolved (Tomback 1981, 1982, 1983, Lanner 1982, Tomback and Linhart
1990). However, the nutcracker shares the role of pinyon pine seed disperser with
the pinyon jay (Gymnorhinus cyanocephalus). A fourth pine that may be dispersed
by nutcrackers is the southwestern white pine (Pinus strobiformis) (Benkman et al.
1984).
In addition, nutcrackers eat and may store seeds of conifers with smaller,
winged seeds such as ponderosa pine (Pinus ponderosa) (Vander Wall and Baida
1977, Giuntoli and Mewaldt 1978, Torick unpublished data), Jeffrey pine (Pinus
1


jejfreyi) (Tomback 1978), Rocky Mountain bristlecone pine (Pinus aristata)
(Peattie 1953, Vander Wall and Baida 1977), and Douglas-fir (Pseudotsuga
menziesii) (Vander Wall and Baida 1977, Giuntoli and Mewaldt 1978, Lanner
1988). Although supposedly wind-dispersed, the seeds of these species may also
be disseminated by nutcrackers. In fact, Lanner (1988) speculates that the Great
Basin bristlecone pine (Pinus longaeva) is dependent on Clarks Nutcracker for
regeneration at high elevations. However, no studies to date have documented the
establishment of these species from nutcracker caches.
Nutcrackers begin foraging on pine seeds in mid-summer (see Vander Wall
and Baida 1977 and Tomback 1978 for review of annual cycle). By late August or
early September, they start storing seeds in caches ranging from 1 to 15 seeds or
more, each buried 2 to 3 cm in the ground (see Table 5 in Tomback and Linhart
1990). Caches are later retrieved in the winter or spring when other foods are
scarce. Scatterhoarded seeds provide a nutritious food source that enables the
nutcracker to survive year-round in mountain regions and to nest in late winter and
spring. Nestlings and fledglings are fed from food stores.
When food supplies are abundant, estimates of the numbers of seeds
individual nutcrackers store per year greatly exceed their projected energetic needs
(Vander Wall and Baida 1977, Hutchins and Lanner 1982, Tomback 1982, Vander
Wall 1988). Consequently, in years when there is a good cone crop, some seeds
may not be recovered and may germinate and eventually establish trees (Tomback
2


1978, 1982, Hutchins and Lanner 1982).
Nutcrackers cache at different altitudes and in different forest types. They
may either cache in areas near parent trees, in communal storage slopes several
kilometers away, or at higher or lower elevations (Vander Wall and Baida 1977,
Tomback 1978, Lanner and Vander Wall 1980, Hutchins and Lanner 1982).
Because some of the caches will be accessible at times when others are not, this
strategy increases the probability that a constant supply of seeds will be available
and also results in more widespread seed dispersal (Vander Wall and Baida 1977,
Tomback 1978, Lanner 1980).
Studies indicate that nutcrackers may prefer cache sites with specific
microsite features (Tomback 1978, Hutchins and Lanner 1982). Tomback (1978)
observed the highest frequency of caches at the base of trees, followed by open
pumice, the base of rocks, under pine needle litter, among tree roots, near fallen
logs or branches, in trees, and among plants (n= 80). Clarks nutcrackers have
also been observed caching in moss, on cliff ledges, in rock crevices, in burns, at
the edges of meadows, near springs and stream banks, and even in a puddle
(Tomback and Kramer 1980, Hutchins and Lanner 1982, Tomback 1986). Certain
microsite features may render caches more accessible or prevent them from
spoiling. Sites exposed to sun and/or wind are not covered by deep snow (Vander
Wall and Baida 1977, Tomback 1978, Hutchins and Lanner 1982, Lanner 1988).
Reradiation of heat may result in faster snowmelt around objects such as tree
3


bases, where Tomback (1978) observed the highest frequency of caches.
However, the seed germination potential at the microsites preferred by nutcrackers
remains to be determined.
By storing more seeds than they actually need, nutcrackers not only
increase the likelihood that food will be available throughout the year, but also
provide a buffer against the effects of post-dispersal seed predators (Tomback
1980). Rodents such as deermice (Peromyscus maniculatus), chipmunks (Tamias
spp.), red-backed voles (Clethrionomys gapperi) and golden-mantled ground
squirrels (Spermophilus lateralis) are notorious post-dispersal seed predators in
western montane forests (Tevis 1952, 1953, Williams 1955, Smith and Baida
1979). Experiments using artificial seed caches have resulted in high rates of seed
loss to rodents (Tomback 1980, Hutchins and Lanner 1982, McCaughey and
Weaver 1990). Hutchins and Lanner (1982) found that seed loss to rodents was
greater in forested areas than in open meadows. Given these findings, in the
absence of strongly counteracting factors, such as cache accessibility, selection
should favor nutcrackers that choose cache sites where the relative density of seed
predators is low.
In this study I will focus on the following questions:
1. What is the relationship between cache site preferences observed for
Clarks Nutcrackers and germination potential for the following pine
species: limber pine, bristlecone pine, and ponderosa pine?
4


2. Do seed predators take seeds from seed caches? How important is seed
predation, and does it vary with respect to elevation, microsite, and pine
species?
a. Nutcrackers cache in montane forest, subalpine forest, and alpine
tundra. Does the relative density of seed predators vary among
these habitats at different elevations?
b. Does seed predator density in a given habitat predict the
proportion of caches lost in that habitat?
5


CHAPTER 2
METHODS
Study Areas
I observed nutcrackers in the Front Range of the Colorado Rocky
Mountains in Rocky Mountain National Park, in the Indian Peaks area of
Roosevelt National Forest in the vicinity of the University of Colorado Mountain
Research Station, in the Arapaho National Forest along Squaw Pass road, and in
the Mount Evans area. Seeds of limber and bristlecone pines were collected for
germination studies from the Mount Evans area and ponderosa pine seeds from
areas near highway 72 in the Roosevelt National Forest.
In the southern Rocky Mountains, which includes the Front Range,
ponderosa pine is the dominant forest tree on south-facing slopes of the montane
zone, ranging from approximately 1830 2440 m in elevation. Limber pine is
found predominantly on dry, rocky ridges, and exposed, south-facing slopes from
about 1800 m to timberline, at about 3350 m, but is most common near elevations
of 2700 m. Rocky Mountain bristlecone pine is primarily a subalpine species
occupying dry, steep sites. In the Mount Evans region, limber and bristlecone
pines often occur together, although bristlecone pine is more common on open,
south-facing slopes with fine-textured soils, and limber pine is more prevalent on
6


rocky ridges (Marr 1967, Benedict 1991).
Seed germination and rodent density studies were conducted at study areas
in three vegetation zones in the Mount Evans area of the Arapaho National Forest
(Fig. 1). Study areas were labeled in accordance with their vegetation zone as
"Montane", "Subalpine" and "Alpine". At the Montane study site the predominate
forest trees were ponderosa pine and lodgepole pine (Pinus contorta) -, at the
Subalpine study area they were Engelmann spruce (Picea engelmannii) and limber
and bristlecone pines. To examine the effects of regional differences in
environmental conditions, a "Comparison" study area was established near
Raymond, Colorado, in the Roosevelt National Forest. The predominate tree
species at the Comparison site was ponderosa pine. I chose study areas that had
the following characteristics: relative flatness, such that the effects of aspect were
minimized, open forest (except at the Alpine study area), and the presence of the
tree species whose seeds were being cached (not including the Alpine study area).
Other study area information is listed in Table 1.
7


Fig. 1. Location of Montane, Subalpine and Alpine study areas in the
vicinity of Mount Evans, Arapaho National Forest, Colorado Front Range,
Rocky Mountains.
8


Table 1. Characteristics of seed germination and predation study sites in the Colorado Front Range of the Rocky Mountains. See
Fig. 1 for locations in the Mount Evans area.
Study area
Comparison
Montane
Subalpine
Alpine
General location Elevation Longitude Latitude Distance to nearest Habitat
(m) site (km)
1 Km North of 2515 40 105" 105 2730" 50 from Montane Ponderosa pine forest with a few
Raymond, CO. scattered limber pines. Rock
outcroppings in vicinity. Scattered
grasses and forbs.
Near Barbour 2741 39 4230
Fork of Soda
Creek
105 3610" 2.5 from Subalpine
Mixed ponderosa pine/lodgepole
pine/Douglas-fir forest. Some Aspen
and a few limber pines. Grasses and
forbs moderately abundant in open
areas.
Near Squaw 3402 39 4030" 105 3605 2 from Alpine
Pass Road
Bristlecone/limber pine forest. Some
Engelmann spruce. Scattered, short
forbs.
Near Mt. 3560 39 3830" 105 3515" 2 from Subalpine Abundant grasses and forbs. Many
Goliath cushion plants. Areas of loose soil
from gopher excavations. Rocks
scattered throughout.


Nutcracker Foraging and Caching Observations
I observed nutcracker foraging and caching behavior from 18 August
through 5 November, 1991, for approximately 200 hours. Information was
recorded regarding which species of pine seeds nutcrackers ate, placed in their
sublingual pouch, or cached. For each caching observation, the number of seeds
cached, as well as cache site features such as elevation and general topography
were noted. Cache microsite was characterized according to which object the
cache was near, if any. In an effort to find communal cache sites, I followed
nutcrackers flying with full throat pouches as far as possible.
Seed Germination Study
Table 2 summarizes the experimental design and sample sizes for the
simulated nutcracker seed cache study of germination and predation. Fifty ripe,
just opening cones each of limber, ponderosa and bristlecone pines were collected
in September and October of 1991. In November, 1991, I made simulated
nutcracker caches at each study area. I cached limber pine seeds at three Mount
Evans study areas, bristlecone pine seeds at the Alpine study area only, and
ponderosa pine seeds at the Montane and Comparison study area. Seed caches
were placed in four microsite features commonly used by nutcrackers (Tomback
1978) as cache sites: 1) in the open (at least 3 m from the nearest tree); 2) near
rocks (over 0.25 m in diameter); 3) near logs (at least 20 cm in diameter and at
10


least 0.5 m in length); and 4) at the base of trees (over 2 rain height). Caches
were only made in the open or near rocks at the Alpine study area.
For each pine species, 27 caches were placed at each type of microsite.
Individual caches consisted of 2 or 3 seeds for limber pine and 5 seeds for
ponderosa and bristlecone pines. One hundred limber pine seeds were weighed
and opened to determine if they were filled. Of the seeds weighing 0.90 g or
greater, 85% were filled. In an effort to eliminate some of the unfilled seeds,
only limber pine seeds over 0.090 g in weight were used for caches. I selected
only the largest and darkest seeds of ponderosa and bristlecone pine for the
germination experiment, since about 80% of the smaller, lighter colored seeds
were empty. The size of limber pine caches was limited by the number of seeds
collected.
11


Table 2. Experimental design and sample sizes for simulated nutcracker cache
study of germination and seed predation rates.
Study area No. of microsites1 No. of caches per microsite Pine species No. of seeds per cache Total caches per pine species
Comparison 4 27 Ponderosa 5 108
Montane 4 27 Limber 2 or 3 108
4 27 Ponderosa 5 108
Subalpine 4 27 Limber 2 or 3 108
Alpine 2 27 Limber 2 or 3 54
Tm A., r Bristlecone 5 54
'Microsites for all but Alpine study areas are: in the open, near rocks, near logs,
and at the base of trees. Only microsites in the open and near rocks were used at
the Alpine study area.
Seed weight, endosperm weight and seed coat thickness were determined
for limber, ponderosa, and bristlecone pines. I obtained measurements of seed
coat thickness using Kanon museum calipers on the flattest area of the seed coat
for 179 seeds of limber pine and 300 seeds each of ponderosa and bristlecone
pines. Other seed characteristics were based on sample sizes of 300 seeds. Seed
characteristics of limber, bristlecone and ponderosa pine are presented in Table 3.
Significant differences among species were found for all measured characteristics
(Table 3).
1 determined caloric content per gram of three samples each of limber,
ponderosa, and bristlecone pine seeds by combusting 1 gram pellets of crushed
endosperm in a Parr Micro Oxygen Bomb Calorimeter, following the techniques of
12


Shoemaker (1981). Caloric content per gram was similar among species:
x = 7197 43 (S.D.) for limber pine, x = 7046 85 (S.D.) for ponderosa
pine, and x = 6852 + 40 (S.D.) for bristlecone pine.
Given evidence that some rodents may use olfaction to locate buried seeds
(Howard and Cole 1967, Howard et al. 1968, Vander Wall 1991), I made efforts
to equalize olfactory cues by using about the same biomass per cache for each
species. Maximum cache size was set at 5 seeds because of seed availability.
Assuming all seeds cached were filled, seed mass per cache was estimated at
0.255, 0.170, 0.210 and 0.080 g for caches of 3 limber pine seeds, 2 limber pine
seeds, 5 ponderosa pine seeds, and 5 bristlecone pine seeds, respectively (based on
mean seed mass values in Table 3).
13


Table 3. Seed characteristics of limber, bristlecone, and ponderosa pines. Only filled seeds were used for
measurements.
Species Seed mass (g) Endosperm mass (g) Hull thickness (mm) Endosperm mass/ seed mass
X SD n X SD n X SD n X SD n
Limber pine (Pinus flexilis) 0.085 0.022 300 0.042 0.012 300 0.32 0.05 179 0.491 0.049 298
Ponderosa pine (P. ponderosa) 0.042 0.007 300 0.020 0.004 300 0.29 0.04 300 0.481 0.061 300
Bristlecone pine (P. aristata) 0.016 0.003 300 0.010 0.003 300 0.10 0.02 300 0.635 0.126 300
P < * .0001 .0001 .0001 .0001
* Results of one-way ANOVA. F2i8 = 999.9 for all comparisons.


At each study area I recorded compass bearing and distance from a central
stake for each cache. A penny was buried approximately 2 cm from each cache so
that the cache could be relocated using a metal detector. From early June through
mid-September of 1992 caches were checked bi-monthly for germinating seedlings
and signs of predation. In September of 1992 I excavated caches with no
seedlings and recorded the numbers of filled and unfilled seeds. To follow
seedling survivorship, I left caches with seedlings surviving through the first
growing season undisturbed until July, 1993.
Rodent Community Diversity and Relative Density
The relative population densities of overwintering potential post-dispersal
seed predators were determined by trapping animals for three consecutive days at
each study area shortly after the spring melt-off ("Spring" trapping). I also
trapped in early summer, when rodents might forage on seeds from caches just
prior to their germination ("Early Summer" trapping). Heavy, early snowfall in
November, 1991, precluded trapping just after nutcracker caching. Because
rodents emerge earliest in the spring at lower elevations and later at higher
elevations, I began work at the Montane study area first and progressed up the
mountain for both trapping periods (Dunmire 1960, McKeever 1964, Halfpenny
1980). All trapping was conducted in 1992. The Spring trapping was conducted
at the Montane study area between May 8 and 11, at the Comparison site between
15


May 17 and 20, at the Subalpine site between June 5 and 8 and at the Alpine site
between June 12 and 15. Early Summer trapping was conducted from June 20 to
23 at the Montane site, from June 27to 30 at the Subalpine site and from July 20
to 23 at the Alpine site.
A square grid of 100 medium sized Sherman traps (10 cm x 10 cm x 10
cm) placed 10 m apart (total area = 1 hectare) was superimposed on each
experimental caching study area. Using a mixture of peanut butter and oats for
bait, I set traps at dusk, checked them at dawn, reset them, and checked them
again at the end of the day. Captured rodents were identified, toe-clipped,
weighed, sexed, and released.
Statistical Analysis
Germination success (proportion of germinants per cache) was calculated as:
______________no. of germinants_________________
no. germinants + no. ungerminated (filled) seeds
Filled seeds included those with whole, but moldy, endosperm. I did not include
seeds taken by rodents, since their fate, had they been left undisturbed, could not
be determined. In addition, I assumed that if seeds were to germinate, they would
do so within the first growing season.
16


Predation (the proportion of seeds taken per cache) was calculated as:
_______________________________no. seeds taken________________________________
no. of seeds taken + no. of germinants + no. of ungerminated (filled) seeds
Because several studies indicate that rodents discriminate between seed
characteristics such as energy and nutrient content, seed size, seed coat thickness,
palatability, and filled and unfilled seeds (Gashwiler 1967, Howard and Cole
1967, Howard et al. 1968, Lockard and Lockard 1971, Lanner 1980), only filled
seeds were included in this calculation.
For statistical analysis, the proportions were arcsine-transformed to
normalize the data (Sokal and Rohlf 1981). I then made comparisons of
germination success and predation at different study areas and microsites and
among different species using two-way ANOVA (Sokal and Rohlf 1981).
Reported means were backtransformed, which resulted in uneven confidence
intervals for means.
Seedling survivorship per cache was calculated as:
no. surviving germinants
total no. germinants
Sample sizes of surviving seedlings at each microsite were too small to be
statistically useful. Thus, I analyzed differences in survivorship among study areas
and species using one-way ANOVA.
17


Two microsites were not available at the Alpine study area (near logs and
at the base of trees). Thus, for comparisons of germination success of limber pine
seed caches, two methods were initially used to examine possible interactions
between study areas and microsites. For Method I, which excluded caches made
near logs and at the base of trees, germination rates for the microsites in common
for all study areas were compared. For Method II, the Alpine study area was
excluded and germination success was compared among the study areas that had
caches at all four microsites. If interactions were not significant for either
method, germination success was combined for two-way ANOVA. I used the
same procedure to analyze predation of limber pine seed caches. Significance for
all statistical analysis was set at the p = .05 level.
Because only limber pine seeds were cached at different elevations, study
area-species interactions could not be adequately examined. Therefore, in order to
analyze microsite effects for all study areas and species, while still taking possible
study area-species interactions into account, a new combined variable, "study
area/species", was created. Two-way ANOVA was then used to examine
differences between germination success and predation from microsites and study
area/species combinations.
Relative rodent densities at each study area were examined using numbers
of animals captured minus those recaptured. Differences among numbers of
trapped rodents at each study area were evaluated using the G-test for goodness of
18


fit with the Williams correction (Sokal and Rohlf 1981). The Pearson product-
moment correlation was used to evaluate possible correlations between predation
and rodent densities.
Although I found documented cases of Peromyscus maniculatus and Tamias
spp. digging up buried seeds (Howard and Cole 1967, Cole et al. 1968, Johnson
and Jorgensen 1981, Vander Wall 1991), I did not find similar data for other
rodent species trapped in the study areas. Since these two taxa were most likely to
be the principal seed predators, separate Peromyscus maniculatus and Tamias totals
were calculated. There seemed to be no relevant differences in foraging behavior
of the two Tamias species, T. umbrinus and T. minimus, found in the study area
(Vaughan 1974, Telleen 1978); thus, individuals of these species were grouped
into one category .
19


CHAPTER 3
RESULTS
Nutcracker Foraging and Caching
Nutcrackers first harvested seeds from closed, green limber pine cones on
20 August, 1991. I saw them harvesting limber pine seeds and placing them in
their sublingual pouches ("pouching" seeds) for the first time on 28 August. The
most active period of limber pine harvesting occurred after the majority of cones
had opened in mid-September.
All of my caching observations were made in Rocky Mountain National
Park. Communal caching slopes were never located, although I did see
nutcrackers with full pouches flying from a limber pine stand near Beaver Ponds at
the bottom of Trail Ridge Road over and beyond a tundra area about 10 kilometers
away. Between 18 and 31 September, six nutcrackers were observed making
eleven limber pine caches - seven in open areas, three near rocks, and one near
the base of a tree. Cache sizes ranged from 2 to 5 seeds, with an average of 3.5
seeds per cache. I observed one nutcracker make three limber pine seed caches in
a bare, gravelly area just above treeline. Of the caches observed, two were made
at 3150 m, three at 3350 m, and the remainder at 2500 m. Seven caches were
made in areas with loose, gravely, soil, and four caches were in areas where the
20


soil was covered by two to three cm of pine needle litter. All caches were made
on steep slopes (over 25) with aspects of cache areas ranging from 170 to 305.
By mid-October most of the limber pine seeds had been harvested. By 3
November nutcrackers in the vicinity of the University of Colorado Mountain
Research Station started foraging in the open cones of ponderosa pine; the
ponderosa pine cone crop was particularly good this fall in this area. Nutcrackers
took seeds from both the cones and the ground, removing seed wings before
pouching seeds.
Further south, in the Mount Evans Wilderness Area, the bristlecone pines
had an abundance of ripe cones. In mid-October, after most limber pine seeds
were gone, nutcrackers foraged on open bristlecone pine cones. Again, they
removed seed wings before pouching seeds. I observed a group of four nutcrackers
flying toward tundra areas, presumably to make caches, with pouch loads of
bristlecone pine seeds. Although I never saw nutcrackers breaking open unripe
bristlecone pine cones in 1991, 1 did observe one nutcracker doing so in the fall of
1992, a year of poor limber pine cone production. Thus, it appears that
nutcrackers do cache both ponderosa and bristlecone pine seeds, in addition to the
wingless seeds of limber pine.
21


Seed Germination Study
Germination Success
Some seeds germinated from caches made at every study area and treatment
for all three pine species. Of the total caches, 70% had one or more germinants at
the Alpine study area, 34% at the Subalpine study area, 28% at the Montane study
area and 20% at the Comparison study area. I did not relocate three caches at
the Montane study area, including two made in the open and one near a rock; no
metal was detected, and there were no seeds in the vicinity of these caches. These
missing caches were not included in calculations.
As mentioned in the "Methods", I compared germination success of limber
pine seed caches among study areas using two smaller ANOVA approaches with
common microsite treatments. Study area-microsite interactions were not
significant for either model (Two-way ANOVA, F2>16i= .224, P < .799 for
Method I, and F3i2IS= 1.288, P < .799 for Method II). Germination success of
pooled limber pine seed caches was then compared among Montane, Subalpine,
and Alpine study areas, and among all microsites. Descriptive statistics and results
of two-way A NOV As of germination success for different species at different
study areas and microsites are summarized in Table 4. For all analyses,
interactions between factors were not significant.
22


Table 4. Descriptive statistics and results of two-way ANOVAs for germination success (proportion of seeds germinating per
cache) among limber, ponderosa, and bristlecone pine seed caches, Comparison, Montane, Subalpine and Alpine study areas, and
microsites in the open, near rocks, near logs, and at the base of trees.1
Dependent Source of variation no. Mean 95% Confidence d.f. F P
variable caches germination intervals
(germination success
success)
Factors Level + -
Germination Study area- Method I 162 2 0.224 .800
success of microsite
limber pine seed interactions Method II 216 3 1.288 .279
caches
Study Area 2 8.871 <.001
Alpine 54 0.464 0.178 0.173
Subalpine 108 0.166 0.092 0.075
Montane 108 0.073 0.073 0.052
Microsite 3 3.536 .0153
Open 81 0.349 0.129 0.119
Rock 81 0.298 0.126 0.112
Log 54 0.192 0.151 0.115
Tree 54 0.081 0.117 0.068
Germination Species- 216 3 1.764 .155
success of microsite
caches at interactions
Montane study
area


Table 4. continued
Dependent Source of variation no. Mean 95% Confidence d.f. F P
variable caches germination intervals
(germination success
success)
Factors Levels + -
Germination Pine Species 1 1.360 .245
success of Limber 108 0.084 0.069 0.050
caches at Montane study Ponderosa 108 0.141 0.084 0.067
area Microsite Open 54 0.250 0.143 0.119 3 3.093 .028
Rock 54 0.112 0.114 0.077
Log 54 0.084 0.105 0.065
Tree 54 0.039 0.079 0.037
Germination Species- 108 1 0.684 .419
success of microsite
caches at Alpine interactions
study area Pine species Limber 54 0.584 0.166 0.176 1 0.015 .904
Bristlecone 54 0.599 0.164 0.176
Microsite Open 54 0.541 0.169 0.174 1 0.634 .436
Rock 54 0.640 0.159 0.177


Table 4. continued
to
ui
Dependent Source of variation no. Mean 95% Confidence d.f. F P
variable caches germination intervals
(germination success
success)
Factors Levels +
Germination Study area-
success of microsite
ponderosa pine interactions
seed caches
Study area
213 3 0.269 .848
1 3.711 .056
Montane 108 0.140 0.078 0.063
Comparison 105 0.057 0.058 0.039
Open 52 0.263 0.137 0.117
Rock 53 0.083 0.099 0.062
Log 54 0.030 0.070 0.029
Tree 54 0.059 0.084 0.048
Microsite
3 4.540
.004


Table 4. continued
Dependent Source of variation no. Mean 95% Confidence d.f. F P
variable caches germination intervals
(germination success
success)
Factors Levels + -
Germination Microsite-study Method I 429 9 0.942 .488
success of area/species
limber, ponderosa and interactions
bristlecone
pine seed
caches Microsite Open 160 0.342 0.089 0.083 3 5.538 .001
Rock 161 0.275 0.085 0.077
Log 108 0.147 0.095 0.074
Tree 108 0.114 0.086 0.064
Study area/ species Comp.-ponderosa 105 0.058 0.065 0.042 5 8.862 <.001
variables Mont.-limber 108 0.084 0.071 0.051
Mont.-ponderosa 108 0.141 0.085 0.068
Sub.-limber 108 0.166 0.090 0.073
Alp.-limber 54 0.471 0.165 0.161
Alp.-bristlecone 54 0.490 0.164 0.163
1 "Germination success" was calculated as the number of germinants divided by the number of germinants plus the number of
ungerminated, filled seeds. For statistical analysis, proportions were arcsine transformed to normalize the data (Sokal and Rohlf
1981). Reported means were backtransformed. Differences between groups were tested by the Least Significant Difference
(LSD) multiple-range test. P values with bold numbering are significant at < .05.


Germination success of limber pine seed caches was significantly different
among both study areas and microsites. Mean germination success increased as
elevation of study areas increased, with the greatest germination success at the
Alpine study area. When pooled for Montane, Subalpine, and Alpine study areas,
germination success of limber pine seed caches at different microsites decreased in
the order of open, rock, log, base of tree microsites.
Caches of ponderosa and limber pine seeds at the Montane study area did
not differ in germination success from each other. However, germination success
for caches of limber and ponderosa seeds combined were significantly different
among microsites, and followed the same trend as described above for limber pine.
At the Alpine study area, germination success was not significantly
different between either bristlecone and limber pine seed caches, or, when
germination success of both species were pooled, microsites in the open and near
rocks.
Germination success for ponderosa pine seed caches at the Montane and
Comparison study areas combined was significantly different among microsites.
Mean germination success was lowest in caches near logs and highest in caches in
the open. There was nearly a significant difference between germination success
of ponderosa pine seed caches at Montane and Comparison study areas, with
caches at the Montane study area having greater mean germination success.
27


When all germination data were pooled, significant differences were found
among microsites (Fig. 2). Mean germination success decreased in the following
microsite order; in the open, near rocks, near logs, and at the base of trees. There
were also significant differences among germination success among study
area/species, which reiterated the aforementioned study area differences (Fig. 3).
28


0.5
0.4 -
in
in
L*J
o
^ 0.3 -
in
O
i
<
z 0.2 -
5
C£
U
O
0.1 -
Open Rock Log Tree
MICROSITE
Fig. 2. Germination success (mean proportion of seeds germinating per cache)
with respect to microsites in the open, near rocks, near logs or at the base of trees
for all study areas (Comparison, Montane, Subalpine and Alpine) and all species
(Limber, Ponderosa and Bristlecone pines). Central bar marks the mean value,
bounded by 95% confidence intervals.
29


Fig. 3. Germination success (mean proportion of seeds germinating per cache)
with respect to study area/species variables for all microsites (in the open, near
rocks, near logs, or at the base of trees). First letters C, M, S and A represent
Comparison, Montane, Subalpine and Alpine study sites, in that order. Second
letters L, P and B denote limber, ponderosa and bristlecone pine seed caches.
Central bar marks the mean value, bounded by 95% confidence intervals.
30


Survivorship
Descriptive statistics and results of one-way ANOVAs of survivorship
through the first summer and the first year for limber, ponderosa, and bristlecone
pine seed caches at Comparison, Montane, Subalpine, and Alpine study areas are
reported in Table 5. Survivorship of limber pine seedlings through the first
summer was significantly different among study areas and increased with elevation
of study areas (Fig. 4). Study area differences were nearly significant for first
year survivorship of limber pine seedlings. Mean survivorship of limber pine
seedlings for the first year was greatest at the Subalpine study area, followed by
the Alpine, and then Montane study areas.
Comparisons of summer survivorship at the Montane study area showed
that mean survivorship of limber pine seed caches was significantly greater for
limber pine seed caches than for ponderosa pine seed caches. Though not
significantly different, yearly mean survivorship was also greater for limber pine
seed caches than for ponderosa pine seed caches.
When summer and first year survivorship were compared between caches
of limber and bristlecone pine seeds, and between caches of ponderosa pine seeds
at Montane and Comparison study areas, no significant results were found.
Comparisons were not made among microsites because sample sizes of germinants
were too small to be statistically useful.
31


Table 5. Descriptive statistics and results of one-way ANOVAs of seedling survivorship through the first summer and the first
year among limber, ponderosa, and bristlecone pine seedlings in caches at Comparison, Montane, Subalpine and Alpine study
areas.1
Seedling survivorship Source of variation no. Mean 95% Confidence d.f. F P
seedlings survivorship intervals
Factor Levels + -
Summer survivorship Study area 2 4.665 .012
of limber pine Alpine 39 0.677 0.193 0.231
seedlings Subalpine 27 0.555 0.224 0.237
Montane 24 0.146 0.261 0.138
First year survivorship Study area 2 2.718 .071
of limber pine Alpine 39 0.131 0.162 0.103
seedlings Subalpine 37 0.320 0.203 0.173
Montane 24 0.052 0.167 0.014
Summer seedling survivorship at the Pine species Limber 24 0.146 0.168 0.112 1 5.597 .021
Montane study area Ponderosa 37 0.007 0.053 0.001
First year seedling survivorship at the Pine species Limber 24 0.052 0.096 0.048 1 2.424 .125
Montane study area Ponderosa 37 0.004 0.034 0.000
Summer seedling survivorship at the Pine species Limber 39 0.677 0.187 0.223 1 0.007 .936
Alpine study area Bristlecone 37 0.665 0.195 0.230
First year seedling survivorship at the Pine species Limber 39 0.131 0.165 0.104 1 1.658 .203
Alpine study area Bristlecone 37 0.249 0.199 0.155


Table 5. continued
Seedling survivorship Source of variation no. seedlings Mean survivorship 95% Confidence intervals d.f. F P
Factor Levels + -
Summer survivorship Study area 1 1.658 .203
of ponderosa pine Montane 25 0.010 0.053 0.007
seedlings Comparison 32 0.060 0.107 0.055
First year survivorship Study area 1 1.438 .236
of ponderosa pine Montane 25 0.005 0.019 0.005
seedlings Comparison 32 0.000 0.009 0.000 :rrmrrrrr_
statistical analysis, proportions were arcsine transformed to normalize the data (Sokal and Rohlf 1981). Reported means were
backtransformed. Differences between groups were tested by the Least Significant Difference (LSD) multiple-range test. P values
with bold numbering are significant at < .05.


0.9
Fig. 4. Limber pine survivorship (mean proportion of germinants surviving per
cache) through the first summer and the first year at Montane, Subalpine and
Alpine study areas. Open bars represent summer survivorship rates and diagonally
hatched bars represent first-year survivorship rates. Error bars denote 95%
confidence intervals.
34


Seed Predation
There was some predation at 13% of all caches at the Alpine study area,
57% of all caches at the Subalpine study area, 60% of all caches at the Montane
study area and 73% of all caches at the Comparison study area. In seven caches
both predation and germination occurred. For all other caches, if there was seed
theft, all filled seeds were taken. As for analysis of germination success,
predation of limber pine seed caches was first examined using the two methods
mentioned previously. Study area-species interactions were not significant for
either method (two-way ANOVA, Method I: F2> !62 = 0.224, P < .800; Method
II: F3 216 = 1.271, P < .284 for Method II). I then compared predation of
limber pine seed caches among Montane, Subalpine, and Alpine study areas, and
among all microsites. Table 6 provides descriptive statistics and results of two-
way ANOVAs of predation among limber, ponderosa and bristlecone pine seed
caches, Comparison, Montane, Subalpine and Alpine study areas, and microsites
in the open, near rocks, near logs, and at the base of trees.
Predation for limber pine seed caches was significantly different among
both study areas and microsites. Mean predation decreased as study area elevation
increased. Caches near tree bases had the greatest mean predation, and caches
near rocks had the lowest mean predation (Table 6).
35


Table 6. Descriptive statistics and results of two-way ANOVAs of predation (proportion of seeds taken per cache) among limber,
ponderosa, and bristlecone pine seed caches, Comparison, Montane, Subalpine and Alpine study areas, and microsites in the open,
near rocks, near logs, and at the base of trees.1
Dependent Source of variation no. Mean 95% Confidence d.f. F P
variable (predation) caches predation intervals
Factors Level + -
Predation of Study area- Method I 162 2 .224 .800
limber pine microsite
seed caches interactions Method II 216 3 1.27 .284
Study Area Alpine 54 0.146 0.177 0.114 2 13.081 <.001
Subalpine 108 0.583 0.128 0.133
Montane 108 0.748 0.106 0.124
Microsite Open 81 0.348 0.153 0.318 3 3.591 .014
Rock 81 0.318 0.151 0.135
Log 54 0.587 0.188 0.202
Tree 54 0.680 0.170 0.120
Predation of Species- 216 3 1.243 .295
caches at microsite
Montane study area interactions


Table 6. continued
Dependent Source of variation no. Mean 95% Confidence d.f. F P
variable (predation) caches predation intervals
Factors Levels + -
Predation of Pine Species 1 3.063 .082
caches at Limber 108 0.748 0.112 0.132
Montane study Ponderosa 108 0.580 0.136 0.142
area Microsite Open 54 0.500 0.197 0.197 3 2.54 .057
Rock 54 0.587 0.187 0.201
Log 54 0.714 0.162 0.197
Tree 54 0.838 0.116 0.169
Predation of Species- 108 1 0.252 .622
caches at microsite
Alpine interactions
study area Pine species Limber 54 0.056 0.079 0.046 1 0.915 .351
Bristlecone 54 0.021 0.057 0.021
Microsite Open 54 0.035 0.079 0.046 1 0.008 .929
Rock 54 0.038 0.057 0.021


Table 6. continued
Dependent Source of variation no. Mean 95% Confidence d.f. F P
variable (predation) caches predation intervals
Factors Levels + -
Predation of Study area- 213 3 2.307 .078
ponderosa microsite
pine seed interactions
caches Study area Montane 108 0.584 0.133 0.140 1 6.155 .014
Comparison 105 0.812 0.098 0.125
Microsite Open 52 0.544 0.192 0.199 3 2.338 .075
Rock 53 0.775 0.143 0.188
Log 54 0.850 0.113 0.170
Tree 54 0.619 0.177 0.195


Table 6. continued
OJ
Dependent Source of variation no. Mean 95% Confidence
variable (predation) caches predation intervals
Factors Levels + -
Predation of Microsite- Method I 429
limber, study area/
ponderosa and species Open
bristlecone interactions Comp.-ponderosa 25 0.784 0.179 0.271
pine seed Mont.-limber 27 0.698 0.213 0.273
caches Mont.-ponderosa 27 0.302 0.273 0.213
Sub.-limber 27 0.358 0.284 0.236
Rock Comp.-ponderosa 26 0.923 0.077 0.220
Mont.-limber 27 0.587 0.247 0.274
Mont .-ponderosa 27 0.587 0.247 0.274
Sub.-limber 27 0.422 0.269 0.246
Log Comp.-ponderosa 27 0.938 0.060 0.203
Mont.-limber 27 0.698 0.213 0.273
Mont.-ponderosa 27 0.732 0.202 0.275
Sub.-limber 27 0.817 0.155 0.248
Tree Comp.-ponderosa 27 0.529 0.260 0.269
Mont.-limber 27 0.941 0.059 0.188
Mont.-ponderosa 27 0.697 0.211 0.268
Sub.-limber 27 0.698 0.213 0.273
1 "Predation" was calculated as the number of seeds taken divided by the number of seeds taken plus the number of germinants
plus the number of ungerminated, filled seeds. For statistical analysis, proportions were arcsine transformed to normalize the data
(Sokal and Rohlf 1981). Reported means were backtransformed. Differences between groups were tested by the Least Significant
Difference (LSD) multiple-range test. P values with bold numbering are significant at < .05.


Although not significantly different, mean predation was greater at limber
pine seed caches than at ponderosa pine seed caches at the Montane study area.
Similarly, predation was almost significantly different among microsites for limber
and ponderosa pine seed caches combined. The greatest mean predation occurred
at caches near tree bases, followed by those near logs, near rocks, and in the open
(Table 6).
Predation was not significantly different at limber and bristlecone pine seed
caches at the Alpine study area. When data from limber and bristlecone pine seed
caches was pooled predation was nearly the same between microsites. Ponderosa
pine seed caches had significantly higher mean predation at the Comparison study
area than the Montane study area. When predation from both Comparison and
Montane study areas was combined differences among microsites were not
significant.
I pooled all predation data except for caches made at the Alpine study area,
where only two of the microsites were used. Interactions between microsite and
study area/species variables were significant (P = .0347); thus comparisons were
restricted to the Comparison, Montane, and Subalpine study areas. Fig. 5
illustrates mean predation at each microsite for different study area/species
variables.
40


T
T
o s i
0.7 -
2 0 6
O :
UJ : u
CZ 1 JL -L 1
CL 0 4 -j
0 3 -j
0.2 \
0 i -j
0 0 ---------------------------------------------------------------------------------------
CP ML MP SL CP ML MP SL CP ML MP SL CP ML MP SL
Open Rock Log Tree
MICROSITE AND STUDY AREA/SPECIES VARIABLES
Fig. 5. Predation (mean proportion of seeds taken per cache) of caches at microsites in the open, near rocks,
near logs and at the base of trees for different study area/species variables. First letters C, M and S
represent Comparison, Montane, and Subalpine study areas, in that order. Second letters L and P denote
limber or ponderosa pine seed caches. Central bar marks the mean value, bounded by 95% confidence
intervals.


Rodent Community Diversity and Relative Density
Spring and Early Summer trapping results are shown in Tables 7 and 8.
Peromyscus maniculatus and Tamias densities were significantly different among
study areas for the Early summer trapping period (G-test for goodness of fit with
Williams correction, = 8.160, df = 3, P < .0175), and were nearly so for
the Spring period (G-test for goodness of fit with Williams correction, G^ =
6.55, df = 3, P < .0901). Densities of Peromyscus maniculatus and Tamias
were nearly significantly correlated with predation rates (Pearson correlation, v =
1, rSpring = .903, fo._= .994) and increased from the highest to the lowest study
areas, as did predation. Fig. 6 illustrates limber pine seed cache predation vs. the
density of Peromyscus maniculatus and Tamias at each study area.
Because the Montane and Comparison study areas were at similar
elevations, I arbitrarily chose to trap at the Montane study area first. However, a
large portion of the animals trapped at the Comparison study area were pregnant
females or juveniles, indicating that spring breeding had begun a few weeks
earlier. Thus, based on literature values for weights, gestation periods, and
changes in pelage (Dice and Bradley 1942, Dunmire 1960, McKeever 1964, Layne
1968, Linzey 1970, Skiyja 1974, Millar et al. 1979, Halfpenny 1980), I excluded
all captured deermice weighing less than 18 g., since it was likely that these
individuals were bom in the spring and were not part of the overwintering
population. I then recorded the number of animals actually trapped during the
42


Spring period in the Early Summer results, since germination had already begun at
this study area.
43


Table 7. Rodents caught at Comparison, Montane, Subalpine, and Alpine study
areas during the Spring trapping period.
Species No. of individuals
Study area
Comparison1 Montane Subalpine Alpine
Peromyscus maniculatus 14 15 4 5
Tamias spp. 2 0 5 1
Spermophilus lateralis 1 0 3 0
Clethrionomys gapperi 0 0 1 0
Phenacomys intermedius 0 1 0 0
Total rodents trapped 17 16 13 6
Total P. maniculatus and Tamias spp. To m rrrrrn 16 15 9 6
Spring numbers at the Comparison study area were extrapolated from Early
Summer numbers. Based on literature values for gestation periods, pelage and
weights (Dice and Bradley 1942, Dunmire 1960, McKeever 1964, Layne 1968,
Linzey 1970, Skryja 1974, Millar et al. 1979, Halfpenny 1980), all captured P.
maniculatus weighing less than 18 grams were excluded from the Spring trapping
data.
44


Table 8. Rodents caught at Comparison, Montane, Subalpine, and Alpine study
areas during the Early Summer trapping period.
Species No. of individuals
Study area
Comparison Montane Subalpine Alpine
Peromyscus maniculatus 23 13 8 7
Tamias spp. 2 4 6 2
Spermophilus lateralis 2 4 3 0
Clethrionomys gapperi 0 0 7 0
Tamiasciurus hudsonicus 0 0 1 0
Phenacomys intermedius 0 0 0 1
Total rodents trapped 27 21 25 10
Total P. maniculatus and Tamias spp. 25 17 14 9
45


Fig. 6. Number of Peromyscus maniculatus and Tamias spp. trapped vs. predation
(the proportion of seeds taken per cache) of limber pine seed caches at Montane,
Subalpine and Alpine study areas (represented by the letters M, S, and A,
respectively) Spring trapping numbers are represented by the dotted line and
Early Summer numbers by the solid line.
46


CHAPTER 4
DISCUSSION
My observations confirm speculations that Clarks Nutcrackers not only eat but
also cache seeds of ponderosa and bristlecone pine, two pine species with small,
winged seeds. In addition, Torick {unpublished data), also working in the Front
Range, observed nutcrackers cache ponderosa pine seeds in 1991. In the Front
Range, nutcrackers foraged first on seeds of limber pine, then ponderosa or
bristlecone pine, depending on which was regionally abundant. As observed in
previous studies (see Tomback 1994 for review), timing of nutcracker foraging
suggests that nutcrackers may have pine seed preferences but that species use is
also determined by ripening phenology.
Although the Eurasian Nutcracker {N. caryocatactes) has been observed
caching regularly above treeline (Reimers 1953, Mattes 1978), this study is the
first to document Clarks Nutcrackers caching at such high elevations. In a study
on Niwot Ridge, Colorado, Daly and Shankman (1985) found limber pine
seedlings above tree-limit on more than an occasional basis. As they concluded, if
environmental conditions have changed such that seedling survival at high
elevations is improved, nutcracker caches may provide opportunities for limber
pine to colonize tundra areas. Mean germination success for limber pine was
highest at the Alpine study area. These results, though unexpected, support the
47


suggestion that nutcrackers may play an important role in the advancement of
timberline.
According to Harper (1977), germination success is determined in part by
the number of "safe sites" offered by the environment. Harper uses the term "safe
site" to describe a particular "zone" that provides conditions necessary for seed
germination. Such conditions include adequate oxygen, water, stimuli needed to
break seed dormancy, and other environmental conditions required for the process
of germination, and absence of predators, competitors, toxic soil constituents, and
pre-emergence pathogens.
It is likely that differences in soil moisture among study areas could have a
considerable effect on germination rates. The east slope of the Rocky Mountains
lies in a rain shadow, and water is usually a limiting factor for vegetation of this
region (Barry 1981). Although no climatic data are available for the specific study
areas, precipitation is generally greater in alpine zones than in either subalpine or
montane zones, although most precipitation occurs in winter in the form of snow
(Halfpenny 1980, Barry 1981, Weber 1992).
Barry (1981) characterizes alpine tundra as having low moisture and high
rates of evaporation. However, conditions vary depending on topographic
location. The distribution of snow across the alpine landscape is not uniform.
Snow may accumulate and be retained longer in some areas than in others,
depending on relief and orientation with respect to sun and wind (Buttrick 1977,
48


Walker et al. 1993). The water supplied from snow beds is a major source of soil
moisture for alpine plants, influencing community composition (Billings and Bliss
1959, Johnson and Billings 1962, Hrapko and La Roi 1978, Komarkova and
Webber 1978, Bell and Bliss 1979, Greenland et al. 1984). Part of the Alpine
study area has a slightly leeward orientation, creating a snow layer that did not
recede until early June of 1991. Moisture from this lingering snow could have
created conditions favorable to germination at the Alpine study area. Also,
because winter temperatures are generally lower at alpine elevations (Barry 1981),
it is possible that a longer cold period at the Alpine study area played some role in
breaking dormancy more often than at lower study areas.
Differences in substrate characteristics among study areas may also affect
germination. Soil that is too compact can impede an emerging radicle, and can
limit oxygen flow to seeds (Harper 1977). Chemical properties of the soils at
different study areas may have influenced germination. An allelopathic species
could prevent others from germinating and becoming established (Geier-Hayes
1987, Harrington 1987). Daniel and Schmidt (1972) tested the effects of various
organic constituents on the germination of several conifers. They found that soil
from beneath the trees of subalpine fir (Abies lasiocarpa), lodgepole pine,
Engelmann spruce and Douglas-fir could inhibit germination of the seeds of other
species, as well as its own. In addition, the presence of needle litter increases soil
acidity, which can affect germination (Rorison 1967). Results from McCaughey
49


and Weavers (1990) study showed that seedling emergence was significantly
higher on mineral than on litter seedbeds, though differences in emergence rates
were insignificant under open conditions. Perhaps the absence of trees and needle
litter enhanced conditions for germination at the Alpine study area.
At the microsite level, I found significantly lower mean germination
success for caches planted near trees for limber pine at all Mount Evans study
areas, ponderosa pine at the Montane study area and pooled data; these results may
also be caused by effects of increased needle litter or allelopathic substances. In
her work in the Sierra Nevadas, Tomback (1982) found whitebark pine seedling
occurrence at tree bases to be below expectation. However, in Nevada Everett et
al. (1986) found that most pinyon pine seedlings occurred in microsites with
needle litter. The effects of microsites on germination probably vary, depending
on environmental conditions and seed species, as evidenced by the different
microsite trends found in this study for germination success at different study
areas.
Overall germination success was similar between species at Alpine and
Montane study areas. From these results, there is no indication that the
germination of any one species is particularly promoted by nutcracker type caches.
Many of the filled seeds that did not germinate had moldy endosperm. In a
germination study on ponderosa pine, Vander Wall (1992) also found large
numbers of moldy seeds. Many other authors mention the adverse affects of
50


pathogens on germination (Baker 1950, Janzen 1977, Rosochacka and Giywacz
1980, Tomback 1983).
Differences between the results of our study and those of others examining
germination and seedling establishment indicate that the effect of various
environmental conditions on germination varies widely among regions, habitat
types, and species (Everett et al. 1986, Geier-Hayes 1987). For instance, the
previously mentioned study, Everett et al. (1986) found the highest percentage of
singleleaf pinyon pine seedlings under tree crowns and under needle litter. They
suggested that needle litter might protect an emerging root tip from desiccation,
allowing it contact with the soil. However, this explanation concerns only surface-
lying seeds: buried or cached seeds would not benefit ffom needle litter in this
manner. In addition, Geier-Hayes (1987) found most ponderosa pine seedlings in
microsites with moderate shade, though the percentage of seedlings found in
association with other microsite features such as bare soil, moss mats, and rotten
wood varied among habitat types.
In a burned forest in Montana, Tomback et al. (1993) found that whitebark
pine regeneration was more frequently associated with Vaccinium scoparium and
less frequently associated with Xerophyllum tenax. Everett et al. (1986) found
evidence that pinyon pine seedling establishment may require the presence of a
"nurse" plant. In addition, Tomback (1982) found that whitebark pine seedling
clusters near sulphur flower plants (Eriogonum umbellatum) occurred beyond
51


expectation. Although these studies suggest that germination success of some
nutcracker-dispersed pines could be affected by the presence of specific plants, we
did not investigate this possibility for limber, ponderosa and bristlecone pines.
Mean summer survivorship of limber pine seedlings was greatest at the
Alpine study area. Most of the dead seedlings at the lower study areas were
shriveled and dry, having apparently succumbed to the effects of heat and summer
drought. A smaller portion of seedlings had their cotyledons eaten or were ripped
out of the ground. Higher mean summer seedling survivorship at the Alpine study
area is probably a result of both lower temperatures and fewer rodents.
The Subalpine study area had the highest mean first-year seedling
survivorship. High mortality at the Alpine study area was probably a result of
extremely harsh winter conditions, indicating that severe climatic conditions may
prevent seedling establishment.
Mean summer seedling survivorship was significantly greater for limber
pine seed caches than ponderosa pine seed caches at the Montane study area.
However, summer seedling survivorship was similar between limber and
bristlecone pine seed caches at the Alpine study area. Although limber pine stands
are more typical of subalpine than montane elevations, it is possible that the larger
limber pine seedlings are more robust and can withstand dry, hot summer
conditions at montane elevations better than ponderosa pine seedlings.
The impact of seed predation varied with respect to elevation, being
52


significantly lower for limber pine seed caches at the Alpine study area than at the
Montane study area. Not surprisingly, rodent numbers were positively related to
predation at the different elevations. These results suggest that high densities of
seed predators probably lower the number of seed caches available to nutcrackers,
and may decrease successful germination of nutcracker caches. Although rodent
populations surely vary from one alpine tundra community to another, nutcrackers
may place some caches in tundra because they have a higher "return" rate at this
elevation.
Predation of ponderosa pine seed caches was significantly higher at the
Comparison study area than at the Montane study area. In fact, predation at the
Comparison study area may actually have been depressed by a local abundance of
seeds. Rodents there may have had little motivation to search for buried seeds. I
found many seedling clumps there in the spring that werent from my caches. The
differences in predation between these two study areas, despite their similar
elevations and relatively close proximity, demonstrates the effect that local
conditions might have on rodent populations.
Predation of limber pine seed caches at all study areas and ponderosa pine
seed caches at the Montane study area were greatest near trees and logs. Fallen
branches and logs are often used as runways by mice, and trees provide cover
from predators (Merritt and Merritt 1978, Halfpenny 1980, Vander Wall 1991).
These microsite features are probably more likely to be encountered by rodents
53


than those in more open areas. As in this experiment, Hutchins and Lanner (1982)
found that fewer caches were taken by rodents in open sites than in than forested
areas.
If seed predation influences nutcracker caching behavior, then nutcrackers
might be expected to avoid caching near logs and trees. However, according to
Tombacks (1978) observations, the highest frequency of caches were made at the
base of trees. The effects of predation at certain microsites may be counteracted
by benefits that these microsites may offer to nutcrackers. Earlier snow melt-off
at the base of trees may provide nutcrackers earlier access to seed caches. In
addition, predation of caches at different microsites seem to vary from one study
area to another.
Obviously, predation and germination cannot be considered entirely
separately, since those caches raided by rodents are removed from the pool of
caches with potential germinants. Arno and Hoff (1989) stated that microsites
favoring seedling establishment might not coincide with nutcracker cache site
preferences. For example, Tomback (1982) found that seedling clusters near trees
occurred less often than expected, which could be a result of either increased
predation, decreased germination, or a combination of both at this microsite. As
demonstrated by this study, germination success of caches is a function of the
likelihood of both predation and germination.
Howard et al. (1968), found that percent detection of buried conifer seeds
54


was positively related to seed weight. Unfortunately, comparisons of seed mass
per cache among seed caches of different species are difficult to make for this
study, since 18% of the cached seeds were unfilled. Howard et al. (1968) found
that attractiveness of seeds to mice was increased by the presence of oils.
Differences in fat content as well as seed coat thickness may also influence seed
preferences in rodents (Lockard and Lockard 1971, Smith and Follmer 1972).
Ponderosa pine seeds have slightly thinner seed coats than limber pine seeds.
However, more limber than ponderosa pine seed caches were taken by rodents at
the Montane study area. Given these results, relative attractiveness of seed caches
of different species is probably determined by a combination of factors, and may
depend on how efficiently rodents forage on certain pine seed species.
My results for seed characteristics differ somewhat from literature values
(Smith 1968, Krugman and Jenkinson 1974, Vander Wall and Baida 1977,
Benkman et al. 1984). Regional variation in seed characteristics add to the
difficulty in generalizing relative attractiveness of seed caches of different species
to rodents. Seed weights in particular are much lower in this study than in others.
There are not enough data available to speculate as to the causes of such regional
variation in seed characteristics.
Rodent densities among forest communities have been compared in
previous work. In a study in the Arkansas River Watershed, Colorado, Armstrong
et al. (1978) found that rodent species composition and densities varied with
55


elevation. Dunmire (1960) studied deermice populations at four elevations in the
White Mountains of California and found that they were least abundant at the
highest elevation site (3800 m). Halfpenny (1980) found that densities of deermice
were lowest at the "Mountain" study site in the subalpine zone of the Colorado
Front Range and highest at the "Plains" study site. There have been no studies to
date on the relative densities of potential seed predators in areas where nutcrackers
cache. Furthermore, the species identity of potential "cache-robbers" has not been
examined.
Smith and Baida (1979) listed potential conifer seed predators; however,
they did not distinguish between rodents that foraged only on surface-lying seeds
and those that were known to dig up buried seeds. Several studies indicate that
golden-mantled ground squirrels, red squirrels (Tamiasciurus hudsonicus),
deermice, and chipmunks all cache seeds (Klugh 1927, Aldous 1941, Tevis 1953,
Kendall 1983, Vander Wall 1992).
Whether red-backed voles (Clethrionomys gapperi) cache or dig up seeds is
not known. Abbott (1961) and Abbott and Quink (1970) studied seed consumption
of white pine (Pinus strobus) by rodents. They placed seeds in trays accessible
only to deermice and red-backed voles, the only small mammals caught during
trapping sessions in both studies. The number of seeds taken from trays far
exceeded daily intake of the two species under laboratory conditions. Abbott and
Quink (1970) also found caches of radio-tagged seeds taken from feeders. In these
56


studies the authors concluded that both species cache. However, in neither study
was the relative seed consumption of each species from the feeding trays
determined. Thus, it is possible that red-backed voles do not cache seeds. In a
subalpine forest in Colorado, Merritt and Merritt (1978) examined the stomach
contents of winter-trapped red-backed voles and found that their diet consisted
primarily of seeds. They suggested that red-backed voles might store conifer seeds
but did not have any direct evidence.
Ground squirrels usually store seeds in large "larders" and do not seem to
exhibit behavior that would lead to discovery of buried seeds (Tevis 1953, Vander
Wall 1992). Red squirrels are known to cache seeds (Kendall 1983), although
they depend primarily on cones stored in middens for food (Klugh 1927, Hatt
1943, Tevis 1953, Smith 1968). However, I trapped only one red squirrel in this
study; thus it is likely that the impact of red squirrels on the buried seed caches in
this study was relatively minimal. Deermice and chipmunks have been observed
caching and digging up buried seeds (Howard and Cole 1967, Howard et al. 1968,
Johnson and Jorgensen 1981, Vander Wall 1991). Numbers of Peromyscus
maniculatus and Tamias spp. were positively related to predation rates, and so it is
possible that these two species are the principal seed predators of nutcracker
caches in the Mount Evans area.
The results of this study raise interesting questions concerning the evolution
of large, wingless seeds in some taxa of the subgenus Strobus and their
57


dependence on birds for seed dispersal (see Tomback and Linhart 1990 for a
review of the problem). In simulated nutcracker caches, small, winged seeds
germinated as readily as the large, wingless limber pine seeds and showed
comparable first-year survivorship. Lanner (1980) and Tomback (1983) present
models, based on directional selection by seed-storing nutcrackers and jays
(Corvidae) and the environment, for the evolution of large, wingless seeds from
ancestral pines with small, winged seeds. According to Tombacks model (1983
and Fig. 8 therein), large, wingless seeds will not evolve if the ecological
requirements of the ancestral species are incompatible with corvid cache sites.
However, Models 2 and 3 of Tomback and Linhart (1990) require genetic drift
initially, and not selection alone, to propel ancestral populations towards
winglessness.
My data lend more support to the latter viewpoint. Assuming that
survivorship of winged-seed species germinated in seed caches is similar to or
better than that resulting from dispersal by wind (and similar to that of large,
wingless seeds), we see that directional selection alone has not resulted in large,
wingless seeds. It is possible that seeds must first attain a large enough size so
that nutcrackers routinely harvest and cache a major proportion of the seed crop.
Clearly, there are many other complicating variables involved, such as climate
variation, within-seedling cluster dynamics, and taxon-based genetic tendencies.
Without more information, we can only speculate about processes.
58


CHAPTER 5
CONCLUSIONS
In the Colorado Front Range, nutcrackers cache the winged seeds of
ponderosa and bristlecone pines, in addition to the wingless limber pine seeds.
Seed caches of all three pines germinated at all study areas and microsites. Thus,
small, winged seeds as well as large, wingless seeds dispersed by nutcrackers can
germinate in caches. Nutcrackers store some seeds above timberline and thus may
contribute to the expansion of the tree-limit, if environmental conditions become
suitable.
This study demonstrates that seed predators routinely deplete cached seeds.
The Alpine study area had the lowest predation rates, suggesting that the
importance of seed predation, and thus rodent densities, varies with respect to
vegetation zone. Nutcrackers may, in fact, cache in tundra to insure that some
seed stores escape predation. Seed predation rates differed among microsites and
among species and are probably influenced by a number of factors, including
rodent behavior and seed attractiveness.
Finally, The density of seed caching rodents predicted relative seed
predation rates at the different study areas.
59


LITERATURE CITED
Abbott, H. G. 1961. White pine seed consumption by small mammals. Journal
of Forestry 59:197-210.
Abbott, H. G., and T. F. Quink. 1970. Ecology of eastern white pine caches
made by small forest mammals. Ecology 51:271-278.
Aldous, S. E. 1941. Food habits of chipmunks. Journal of Mammalogy 22:18-
24.
Armstrong, D. M., B. H. Banta, and E. J. Pokropus. 1973. Altitudinal
distribution of small mammals along a cross-sectional transect through the
Arkansas River Watershed, Colorado. Southwestern Naturalist 17:315-317.
Arno, S., and R. J. Hoff. 1989. Silvics of whitebark pine {Pirns albicaulis).
U.S.D.A., Forest Service, Gen. Tech. Rep. INT-253. Intermountain Research
Station, Ogden, Utah.
Baker, F. S. 1950. Principles of silviculture. McGraw-Hill Book Company,
Inc., New York, New York, U.S.A.
Barry, R. G. 1981. Mountain weather and climate. Methuen and Company,
New York, New York, U.S.A.
Bell, K. L., and L. C. Bliss. 1979. Autecology of Kobresia bellardii: why
winter snow accumulation limits local distribution. Ecological Monographs
49:377-402.
Benedict, A. 1991. A Sierra Club naturalists guide to the southern Rockies: the
Rocky Mountain Regions of southern Wyoming, Colorado, and northern New
Mexico. Sierra Club Books, San Francisco, California, U.S.A.
Benkman, C. W., R. P. Baida, and C. C. Smith. 1984; Adaptations for seed
dispersal and the compromises due to seed predation in limber pine. Ecology
65:632-642.
Billings, W. D., and L. C. Bliss. 1959. An alpine snowbank environment and its
effects on vegetation, plant development, and productivity. Ecology 40:388-
397.
60


Buttrick, S. C. 1977. The alpine flora of Teresa Island, Atlin Lake, British
Columbia, with notes on its distribution. Canadian Journal of Botany 55:1399-
1409.
Daly, C., and D. Shankman. 1985. Seedling establishment by conifers above
tree limit on Niwot Ridge, Front Range, Colorado, U.S.A. Arctic and Alpine
Research 17:389-400.
Daniel, T. W., and J. Schmidt. 1972. Lethal and nonlethal effects of the organic
horizons of forested soils on germination of seeds from several associated-
conifer species of the Rocky Mountains. Canadian Journal of Forestry
Research 2:179-184.
Dice, L. R., and R. M. Bradley. 1942. Growth in the deer mouse, Peromyscus
maniculatus. Journal of Mammalogy 23:416-427.
Dunmire, W. W. 1960. An altitudinal survey of reproduction in Peromyscus
maniculatus. Ecology 41:174-182.
Everett, R. L., S. Koniak, and J. D. Budy. 1986. Pinyon seedling distribution
among soil surface microsites. U.S.D.A., Forest Service, Research Paper INT-
363. Intermountain Forest and Range Experiment Station, Ogden, Utah,
U.S.A.
Gashwiler, J. S. 1967. Conifer seed survival in a western Oregon clearcut.
Ecology 51:849-854.
Geier-Hayes, K. 1987. Occurrence of conifer seedlings and their micro-
environments on disturbed areas in central Idaho. U.S.D.A., Forest Service,
Research Paper INT-383. Intermountain Forest and Range Experiment Station,
Ogden, Utah, U.S.A.
Giuntoli, M., and L. R. Mewaldt. 1978. Stomach contents of Clarks Nutcracker
collected in western Montana. Auk 95:595-598.
Greenland, D., N. Caine, and O. Poliak. 1984. The summer water budget and
its importance in the alpine tundra of Colorado. Physical Geography 5:221-
239.
61


Halfpenny, J. C. 1980. Reproductive strategies: intra and interspecific
comparisons within the genus of Peromyscus. Ph.D. dissertation, University of
Colorado, Boulder, Colorado.
Harper, J. L. 1977. Population biology of plants. Academic Press, New York,
New York, U.S.A.
Harrington, M. G. 1987. Phytotoxic potential of gambel oak on ponderosa pine
seed germination and initial growth. U.S.D.A., Forest Service, Research Paper
RM-277. Rocky Mountain Forest and Range Experiment Station, Fort Collins,
Colorado, U.S.A.
Hatt, R. T. 1943. The pine squirrel in Colorado. Journal of Mammalogy
24:311-345.
Howard, W.E., and R. E. Cole. 1967. Olfaction in seed detection by deer mice.
Journal of Mammalogy 48:147-150.
Howard, W. E., R. E. Marsh, and R. E. Cole. 1968. Food detection by deer
mice using olfactory rather than visual cues. Animal Behavior 16:13-17.
Hrapko, J. O., and G. A. La Roi. 1978. The alpine tundra vegetation of Signal
Mountain, Jasper National Park. Canadian Journal of Botany 56:309-332.
Hutchins, H. E., and R. M. Lanner. 1982. The central role of Clarks
Nutcracker in the dispersal and establishment of whitebark pine. Oecologia
55:192-201.
Janzen, D. H. 1977. Why fruits rot, seeds mold, and meat spoils. American
Naturalist 111:691-713.
Johnson, N. M., and W. D. Billings. 1962. The alpine vegetation of the
Beartooth Plateau in relation to cryopedogenic processes and patterns.
Ecological Monographs 32:105-135.
Johnson, T. K., and C. D. Jorgensen. 1981. Ability of desert rodents to find
buried seeds. Journal of Range Management 34:312-314.
Kendall, K. C. 1983. Use of pine nuts by grizzly and black bears in the
Yellowstone area. International Conference on Bear Research and Management
5:166-173.
62


Klugh, B. A. 1927. Ecology of the red squirrel. Journal of Mammalogy 8:1-32.
Komarkova, V., and P. J. Webber. 1978. An alpine vegetation map of Niwot
Ridge, Colorado. Arctic and Alpine Research 10:1-30.
Krugman, S. L., and J. L. Jenkinson. 1974. Pirns. Pages 598-638 in C. S.
Schopmeyer, compiler. Seeds of woody plants in the United States. U.S.
Agricultural Handbook 450.
Lanner, R. M. 1980. Avian seed dispersal as a factor in the ecology and
evolution of limber and whitebark pines. Pages 15-48 in Proceedings of the
Sixth North American Forest Biology Workshop, University of Alberta,
Edmonton, Alberta, Canada.
Lanner, R. M. 1982. Adaptations of whitebark pine for seed dispersal by Clarks
Nutcracker. Canadian Journal of Forestry Research 12:391-402.
Lanner, R. M. 1988. Dependence of Great Basin bristlecone pine on Clarks
Nutcracker for regeneration at high elevation. Arctic and Alpine Research
20:358-362.
Lanner, R. M., and S. B. Vander Wall. 1980. Dispersal of limber pine seed by
Clarks Nutcracker. Journal of Forestry 78:637-639.
Layne, H. H. 1968. Ontogeny. Pages 148-253 in J. A. King, editor. Biology
of Peromyscus (Rodentia). The American Society of Mammalogists, Special
Publication 2:1-593.
Ligon, J. D. 1978. Reproductive interdependence of pinon jays and pinon pines.
Ecological Monographs 48:111-126.
Linzey, A. V. 1970. Postnatal growth and development of Peromyscus
maniculatus. Journal of Mammalogy 51:152-155.
Lockard, R. B., and J. S. Lockard. 1971. Seed preference and buried seed
retrieval of Dipodomys deserti. Journal of Mammalogy 52:219-221.
Marr, J. W. 1967. Ecosystems of the East slope of the Front Range in Colorado.
University of Colorado Studies: Series in Biology 8:1-134.
63


Mattes, H. 1978. Der Tannenhaher (Nucifraga caryocatactes L.) im
engadinistudien zu seiner okologie und function im arvenwald. (Pinus cembra
L.). Munsterische Geographische Arbeiten, Heft 2. Ferdinand Schoningh,
Paderbom.
McCaughey, W. W., and T. Weaver. 1990. Biotic and microsite factors
affecting whitebark pine establishment. Pages 140-150 in W. C. Schmidt and
K. J. McDonald, compilers. Proceedings of the Symposium on Whitebark Pine
Ecosystems: Ecology and Management of a High-Mountain Resource.
U.S.D.A., Forest Service, General Technical Report INT-270. Intermountain
Forest and Range Experiment Station, Ogden, Utah, U.S.A.
McKeever, S. 1964. The biology of the golden-mantled ground squirrel, Citellis
lateralis. Ecological Monographs 34:383-401.
Merritt, J. F., and J. M. Merritt. 1978. Population ecology and energy
relationships of Cleithrionomys gapperi in a Colorado subalpine forest. Journal
of Mammalogy 59:576-598.
Millar, J. S., F. B. Wille, and S. L. Iverson. 1979. Breeding by Peromyscus in
seasonal environments. Canadian Journal of Zoology 57:719-727.
Peattie, D. C. 1953. A natural history of western trees. Houghton-Miffin
Company, New York, New York, U.S.A.
Reimers, N. F. 1953. The food of the nutcracker and its role in the dispersal of
the cedar-pine in the mountains of Khamar-Daban. Lesnoe Khozyaistro 1, 63-
64. (translated from Russian by L. Kelso).
Rorison, I. H. 1967. A seedling bioassay on some soils in the Sheffield area.
Journal of Ecology 55:725-741.
Rosochacka, J., and A. P. Grywacz. 1980. The color of Pinus sylvestris seeds
and their susceptibility to damping off: color of seed coats and their chemical
composition. European Journal of Forest Pathology 10:193-201.
Shoemaker, D. P, C. W. Garland, J. I. Steinfeld, and J. W. Nibler. 1981.
Experiments in physical chemistry, 4th edition. McGraw-Hill, New York, New
York, U.S.A.
64


Skryja, D. D. 1974. Reproductive biology of the least chipmunk (Eutamias
minimus operias) in Southeastern Wyoming. Journal of Mammalogy 55:221-
223.
Smith, C. C. 1968. The adaptive nature of social organization of three squirrels
in the genus Tamiasdurus. Ecological Monographs 38: 31-63.
Smith, C. C., and R. P. Baida. 1979. Competition among insects, birds, and
mammals for conifer seeds. American Zoologist 19:1065-1083.
Smith, C. C., and D. Follmer. 1972. Food preferences of squirrels. Ecology
53:82-91.
Sokal, R. R., and F. J. Rohlf. 1981. Biometry: the principles and practice of
statistics in biological research, 2nd edition. W. H. Freeman and Company,
New York, New York, U.S.A.
Telleen, S. L. 1978. Structural niches of Eutamias minimus and E. umbrinus in
Rocky Mountain National Park. Ph.D. dissertation. University of Colorado,
Boulder, Colorado.
Tevis, L., Jr. 1952. Autumn foods of chipmunks and golden-mantled ground
squirrels in the northern Sierra Nevada. Journal of Mammalogy 33:198-205.
Tevis, L, Jr. 1953. Effect of vertebrate animals on seed crop of sugar pine.
Journal of Wildlife Management 17:128-131.
Tomback, D. F. 1978. Foraging strategies of Clarks Nutcracker. Living Bird
16:123-161.
Tomback, D. F. 1980. How nutcrackers find their seeds stores. Condor 82:
10-19.
Tomback, D. F. 1981. Notes on cones and vertebrate-mediated dispersal of
Pinus albicaulis. 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.
65


Tomback, D. F. 1983. Nutcrackers and pines: coevolution or coadaptation?
Pages 179-223 in M. H. Nitecki, editor. Coevolution. University of Chicago
Press, Chicago, Illinois, U.S.A.
Tomback, D. F. 1986. Post-fire regeneration of krummholz whitebark pine: a
consequence of nutcracker seed caching. Madrono 33:100-110.
Tomback, D. F. 1994. The ecological relationship between Clarks Nutcracker
and four wingless-seed Strobus pines of western North America. In W. C.
Schmidt, and F. K. Holtmeier, eds. Subalpine stone pines and their
environment: the status of our knowledge. Proceedings: 1992 September 5-11;
St. Moritz, Switzerland. U.S.D.A., Forest Service, Intermountain Forest and
Range Experiment Station, Ogden, Utah, U.S.A, in press.
Tomback, D. F., and K. A. Kramer. 1980. Limber pine seed harvest by Clarks
Nutcracker in the Sierra Nevada: timing and foraging behavior. Condor
82:467-468.
Tomback, D. F., and Y. B. Linhart. 1990. The evolution of bird-dispersed
pines. Evolutionary Ecology 4:185-219.
Tomback, D. F., S. K. Sund, L. A. Hoffman. 1993. Post-fire regeneration of
Pinus albicaulis: height-age relationships, age structure, and microsite
characteristics. Canadian Journal of Forestry Research 23:113-118.
Vander Wall, S. B. 1988. Foraging of Clarks Nutcrackers on rapidly changing
pine seed resources. Condor 90:621-631.
Vander Wall, S. B. 1991. Mechanisms of cache recovery by yellow pine
chipmunks. Animal Behavior 41:851 -863.
Vander Wall, S. B. 1992. The role of animals in dispersing a "wind-dispesed"
pine. Ecology 73:614-621.
Vander Wall, S. B., and R. P. Baida. 1977. Coadaptations of the Clarks
Nutcracker and the pinon pine tor efficient seed harvest and dispersal.
Ecological Monographs 47:89-111.
Vaughan, T. A. 1974. Resource allocation in some sympatric subalpine rodents.
Journal of Mammalogy 55:764-795.
66


Walker, D. A., J. C. Halfpenny, M. D. Walker, and C. A. Wessman. 1993.
Long-term studies of snow-vegetation interactions. Bioscience 43:287-301.
Weber, J. C. 1992. Geographic variation in speed of seed germination in central
Oregon ponderosa pine. U.S.D.A., Forest Service, Research Paper PNW-RP-
444. Pacific Northwest Research Station, Portland, Oregon, U.S.A.
Williams, O. 1955. The food of mice and shrews in a Colorado montane forest.
University of Colorado Studies: Series in Biology 3:109-114.
67