Citation
Clark's nutcracker seed use and limber pine metapopulation structure in Rocky Mountain National Park

Material Information

Title:
Clark's nutcracker seed use and limber pine metapopulation structure in Rocky Mountain National Park
Creator:
Williams, Tyler Justin ( author )
Place of Publication:
Denver, Colo.
Publisher:
University of Colorado Denver
Publication Date:
Language:
English
Physical Description:
1 electronic file (76 pages) : ;

Thesis/Dissertation Information

Degree:
Master's ( Master of science)
Degree Grantor:
University of Colorado Denver
Degree Divisions:
Department of Integrative Biology, CU Denver
Degree Disciplines:
Biology

Subjects

Subjects / Keywords:
Clark's nutcracker ( lcsh )
Limber pine -- Colorado -- Rocky Mountain National Park ( lcsh )
Clark's nutcracker ( fast )
Limber pine ( fast )
Colorado -- Rocky Mountain National Park ( fast )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Review:
The Clark's nutcracker (Nucifraga columbiana) is the major seed disperser for limber pine (Pinus flexilis), which in turn is a preferred conifer food source for the nutcrackers. Limber pine exhibits a metapopulation structure: regional populations consist of small, local populations connected by seed dispersal. Current and predicted threats to limber pine include mountain pine beetle (Dendroctonus ponderosae) outbreaks, larger wildfires, and especially exotic white pine blister rust (pathogen Cronartium ribicola). Extensive limber pine mortality may force nutcrackers to rely on alternative conifer seed sources. In Rocky Mountain National Park (RMNP), I investigated the importance of alternative seed sources for nutcrackers, limber pine's metapopulation structure, and how local population connectivity might be influenced by nutcracker spatial use. From mid-June to late October 2014 to 2016, data were collected on: 1) cone production in stands of limber pine, ponderosa pine (Pinus ponderosa), and Douglas-fir (Pseudotsuga menziesii); 2) timing of nutcracker stand visitation; and, 3) nutcracker seed harvest and caching behavior; 4) limber pine geographic occurrence within RMNP using GIS layers; and 5) radio-tracked locations for nutcrackers in 2015-2016. ( , )
Review:
Limber pine component populations (n = 51) ranged from 1 - 400 ha in size with inter-population distances of 1 - 36 km. Nutcrackers traveled distances ranging from 1 - 12 km (n = 7 nutcrackers) and cached seeds 2.2 km away from the foraging site (n = 1 nutcracker), indicating potentially high limber pine metapopulation connectivity, even with a smaller nutcracker population. Annual variation in cone production influenced nutcracker foraging preferences and the use of different seed sources. Each year starting in mid to late August, nutcrackers foraged on limber pine seeds, which ripen in early September. In 2014 and 2015, nutcrackers transitioned to harvesting ponderosa pine seeds, which ripen in early October. In 2016, instead they transitioned to Douglas-fir seeds, which ripen in late September. With potential limber pine losses, I believe that alternative seed sources will support a nutcracker population, although carrying capacity may be lower. I suggest that ponderosa pine will serve as an increasingly critical food resource if limber pine declines.
Bibliography:
Includes bibliographic resource.
System Details:
System requirements: Adobe Reader.
Statement of Responsibility:
Tyler Justin Williams.

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:
on10191 ( NOTIS )
1019158781 ( OCLC )
on1019158781
Classification:
LD1193.L45 2017m W45 ( lcc )

Downloads

This item has the following downloads:


Full Text
CLARKS NUTCRACKER SEED USE
AND LIMBER PINE METAPOPULATION STRUCTURE IN ROCKY MOUNTAIN NATIONAL PARK by
TYLER JUSTIN WILLIAMS B.S., Arkansas Tech University, 2012
A thesis submitted to the Faculty of the Graduate School of the University of Colorado in partial fulfillment of the requirements for the degree of Master of Science Biology Program
2017


This thesis for the Master of Science degree by Tyler Justin Williams has been approved for the Biology Program by
Diana F. Tomback, Chair Michael Wunder Michael Greene
Date: May 13, 2017
n


Williams, Tyler Justin (MS, Biology Program)
Clarks Nutcracker Seed Use and Limber Pine Metapopulation Structure in Rocky Mountain National Park
Thesis directed by Professor Diana F. Tomback
Abstract
The Clarks nutcracker (Nucifraga Columbiana) is the major seed disperser for limber pine (Pinus flexilis), which in turn is a preferred conifer food source for the nutcrackers. Limber pine exhibits a metapopulation structure: regional populations consist of small, local populations connected by seed dispersal. Current and predicted threats to limber pine include mountain pine beetle (Dendroctonusponderosae) outbreaks, larger wildfires, and especially exotic white pine blister rust (pathogen Cronartium ribicola). Extensive limber pine mortality may force nutcrackers to rely on alternative conifer seed sources. In Rocky Mountain National Park (RMNP), I investigated the importance of alternative seed sources for nutcrackers, limber pines metapopulation structure, and how local population connectivity might be influenced by nutcracker spatial use. From mid-June to late October 2014 to 2016, data were collected on: 1) cone production in stands of limber pine, ponderosa pine (Pinus ponderosa), and Douglas-fir (Pseudotsuga menziesii); 2) timing of nutcracker stand visitation; and, 3) nutcracker seed harvest and caching behavior; 4) limber pine geographic occurrence within RMNP using GIS layers; and 5) radio-tracked locations for nutcrackers in 2015-2016.
Limber pine component populations (n = 51) ranged from 1 400 ha in size with inter-population distances of 1 36 km. Nutcrackers traveled distances ranging from 1-12 km (n = 7 nutcrackers) and cached seeds 2.2 km away from the foraging site (n = 1


nutcracker), indicating potentially high limber pine metapopulation connectivity, even with a smaller nutcracker population. Annual variation in cone production influenced nutcracker foraging preferences and the use of different seed sources. Each year starting in mid to late August, nutcrackers foraged on limber pine seeds, which ripen in early September. In 2014 and 2015, nutcrackers transitioned to harvesting ponderosa pine seeds, which ripen in early October. In 2016, instead they transitioned to Douglas-fir seeds, which ripen in late September. With potential limber pine losses, I believe that alternative seed sources will support a nutcracker population, although carrying capacity may be lower. I suggest that ponderosa pine will serve as an increasingly critical food resource if limber pine declines.
The form and content of this abstract are approved. I recommend its publication.
Approved: Diana F. Tomback
iv


Acknowledgments
I would like to thank my advisor Diana F. Tomback for all of her help, direction, and advice along the way. I would also like to thank Michael Wunder and Michael Greene for being on my committee and providing insight on project organization, data collection, and data analysis. I appreciate the work done by Kristin Broms in writing the RStudio code for occupancy analyses. I would also like to extend my gratitude to the Rocky Mountain Conservancy for awarding me the Bailey Research Fellowship in 2015 and the Denver Field Ornithologists for awarding me the Research, Education, and Conservation Grant in 2015 and 2016. These were instrumental in obtaining radio telemetry equipment and an additional Garmin GPSMAP 62stc, in addition to assisting with gas costs for myself and volunteers. I would like to thank Rocky Mountain National Park staff for their help with logistics and permitting, especially Paul McLaughlin and Scott Esser. I would like to acknowledge the University of Colorado Denver IACUC for ensuring that this research was conducted humanely (protocol # 88314(06)lc). I am also very grateful for all of the field assistance offered by the 16 volunteers who helped me collect data. Furthermore, I would like to thank the other graduate students within the Department of Integrative Biology for helpful discussions and workshopping. Finally, I am deeply grateful for all of the support offered by Amber Williams and other loved ones throughout this program.


Table of Contents
CHAPTER
LITERATURE REVIEW.......................................................1
Background: Clarks Nutcracker..........................................1
Background: Limber Pine.................................................5
I. THE IMPORTANCE OF MULTIPLE SEED SOURCES: CLARKS NUTCRACKER SEED USE IN ROCKY MOUNTAIN NATIONAL
PARK...................................................................10
Introduction...........................................................10
Methods................................................................15
Study Area.......................................................15
Field Methods....................................................16
Stand Selection...........................................16
Estimation of Tree Density, Cones/Tree, Cone
Density, and Cone Insect Infestation......................16
Nutcracker Presence/Absence Surveys.......................19
Focal Behavior Surveys....................................19
Statistical Analyses.............................................21
Tree and Cone Density and Landscape Energetics............21
Nutcracker Presence and Foraging Patterns.................23
Results................................................................24
Tree and Cone Density, Cone Infestation, Landscape Energetics,
and Ripening Phenology...........................................24
vi


Nutcracker Stand Visitation
27
Nutcracker Foraging and Caching Behavior.........................27
Discussion.............................................................30
Tables.................................................................38
Figures...............................................................40
IF LIMBER PINE METAPOPULATION STRUCTURE AND
CONNECTIVITY IN ROCKY MOUNTAIN NATIONAL PARK..........................49
Introduction...........................................................49
Methods................................................................52
Study Area.......................................................52
Limber Pine Metapopulation Structure............................53
Nutcracker Flight and Seed-dispersal Distances..................53
Connectivity....................................................54
Results...............................................................54
Limber Pine Metapopulation Structure in Rocky Mountain
National Park...................................................54
Nutcracker Flight and Seed-dispersal Distances..................55
Inter-population Connectivity...................................56
Discussion............................................................57
Tables................................................................60
Figures...............................................................62
REFERENCES...................................................................64
vii


CHAPTER I
LITERATURE REVIEW Background: Limber Pine
Limber pine (Pinus flexilis) is classified in class Pinopsida, order Pinales, family Pinaceae, genus Pinus, section Strobus, and in subsection Strobi. This classification is being challenged by some researchers (Liston et al. 1999, Gernandt et al. 2005, Syring et al. 2007), who suggest that the subsections Strobi and Cembrae (which includes whitebark pine, Pinus albicaulis, and other stone pines) should be merged to form a new subsection Strobus (Gernandt et al. 2005). They also recommend merging this new subsection with the subsections Gerardianae and Krempfianae to form a new section Quinquefoliae (Gernandt et al. 2005).
Limber pine, which occurs in patchy and scattered stands, has a comparatively broad distribution in western North America. The distribution runs from southern Alberta and British Columbia south to northern New Mexico and Arizona; and west from the Sierra Nevada and coastal ranges of California to its easternmost range in the Black Hills, South Dakota (Critchfield and Little 1966). Limber pine also has a broader altitudinal distribution than other white pines, occurring in Colorado from elevations ranging from 1600 m in grassland steppe habitat to 3400 m in the Rocky Mountain treeline where krummholz trees exist (Schoettle and Rochelle 2000).
Limber pine is a pioneer species in that it is one of the first populations to be established at an open site, such as that caused by a burn (Lanner and Vander Wall 1980, Rebertus et al. 1991). At productive sites, limber pine regeneration may occur over the first ~30 years, but this species will be outcompeted by shade-tolerant trees such as spruce and fir (Rebertus et al. 1991, Coop and Schoettle 2009). These shade-tolerant trees are actually able
1


to become established at these sites as a result of limber pine mitigating harsh conditions (Rebertus et al. 1991). In more xeric habitats, limber pine is replaced by shade-tolerant trees at a slower rate and the process of regeneration may continue for 100 years (Veblen 1986, Rebertus et al. 1991). In extremely harsh environments, such as in the upper subalpine, limber pine will form climax communities (Rebertus et al. 1991). The reason that climax stands of limber pine are restricted to extremely harsh sites is that limber pine is a poor competitor (Veblen 1986), which also accounts for some aspect of metapopulation dynamics (Webster and Johnson 2000 and Veblen 1986). In the Rocky Mountains, limber pine sometimes occurs in dwarfed form (krummholz) at treeline, as both solitary trees and within tree islands (Resler and Tomback 2008).
A coevolved mutualist of the Clarks Nutcracker (Nucifraga Columbiana, Family Corvidae, order Passeriformes), limber pine has several adaptations for bird dispersal (Tomback 1982, Tomback 1983). In contrast to the small, winged pine seeds (e.g., ponderosa pine (Pinusponderosa) producing on average 37 mg seeds and lodgepole pine (Pinus contorta) 4 mg seeds) of wind-dispersed pines, which are considered ancestral and characterize the majority of species in Pinaceae, limber pine seeds are large (93 mg) and wingless (Lanner 1980, Critchfield 1986, Tomback and Linhart 1990). Nutcrackers prefer these seeds because they are more efficiently harvested than winged seeds (Tomback 1978, Tomback and Linhart 1990). In addition, the branches are lyrate (upswept) and the cones are horizontally oriented in whorls at branch ends. The orientation of the branches and cones allow for the nutcrackers to assess how productive a tree is and to perch on cones and branch tips for efficient seed harvesting (Lanner 1980, 1982, Tomback and Linhart 1990). In
2


contrast to whitebark pines closed cones, the cones of limber pine open when seeds are ripe, which means that they may become dislodged and drop to the ground.
Limber pine populations are declining in the Rocky Mountains (Tomback and Achuff 2010, Schwandt et al. 2010). The most serious threat is the disease white pine blister rust. Cronartium ribicola, the fungal pathogen that causes the disease, is native to Asia and was accidentally introduced to the Pacific Northwest in the early 1900s (McDonald and Hoff 2001, Geils et al. 2010). Since then, it has spread widely throughout the Western U.S. and into Canada (Schwandt et al. 2010, Tomback and Achuff 2010). Basidiospores from diseased Ribes leaves initiate infection by entering the stomata of pine needles during humid conditions in late summer (McDonald and Hoff 2001, Geils et al. 2010). From the stomata, hyphae eventually spread to the phloem of branches and stems. After two to three years, cankers developfrom which spores are releasedand eventually girdle and kill the branch or stem. If infection occurs in the canopy, the death of branches will result in fewer cones for reproduction and less foliage for photosynthesis. If infection spreads to the bole (trunk), the tree will eventually die (McDonald and Hoff 2001, Geils et al. 2010). White pine blister rust may result in extensive canopy damage, with mortality as high as 100% in some limber pine stands (Kearns and Jacobi 2007). In 1998, the first infected limber pine in Colorado was detected (Johnson and Jacobi 2000). In 2010, blister rust was detected in limber pine in RMNP (Schoettle et al. 2011).
Mountain pine beetles are another major threat that is reducing limber pine populations in many regions (Gibson et al. 2008). They are native to the West and coevolved with the widespread hosts ponderosa and lodgepole pine, but may attack any pine species. The female will enter the bark of a pine, lay her eggs, and deposit fungal spores within the
3


tree. The larvae feed on the phloem and sapwood while the fungus spreads within the tree, resulting in mortality via nutrient transport disruption (Gibson et al. 2009). Mountain pine beetle outbreaks periodically occur in western forests, in response to higher temperatures and drought cycles, sometimes lasting for 10 years or more (Romme et al. 1986, Perkins and Swetnam 1996, Lynch et al. 2006). During severe outbreaks, the pine beetles may invade the stands of higher-elevation white pine species, such as limber pine. During the late 1990s, an outbreak began that appears to be on the downswing. This has resulted in unprecedented mountain pine beetle kill, and has been attributed to mature even-aged stands of lodgepole and ponderosa pine as well as recent higher temperatures and drought (Gibson et al. 2008), which in turn has been attributed to climate change (Logan and Powell 2001, Logan et al. 2003, Raffa et al. 2008).
Fire suppression has also had a negative impact on limber pine, as a result of fire exclusion programs beginning in the 1920s (Arno and Allison-Bunnell 2002, Taylor and Carroll 2004). These programs aimed to eliminate forest fires and were successful in reducing fire-frequency. However, it was not until late in the 20th century that studies were conducted that demonstrated the negative effects of fire suppression (Brown et al. 1994, Van Wagner et al. 2006, Keane et al. 2002). While limber pines range actually expanded into lower elevations as a result of fire suppression in some areas, reduced fire frequency also led to increased community succession (Gruell 1983), which decreased the number of early successional limber pine populations and also decreased landscape diversity.
Climate change may affect limber pine through means other than mountain pine beetle outbreaks. Climate change is also expected to cause plant populations to shift in distribution. Since the populations of different plant species may shift differently or at
4


different rates, this may result in new plant communities with unknown consequences (IPCC 2001 and references therein). Limber pine is predicted to shift northward and to higher elevations while the southern populations and lower elevation populations may die out (Tomback and Achuff 2010). Issues may arise in the shifting of limber pine as a result of metapopulation structure, harsh conditions above treeline, and complications from mountain pine beetle mortality, fire suppression, and white pine blister rust infection (Tomback et al. 2011). If limber pines high elevation and northern ranges are not able to expand as quickly as the low elevation and southern ranges are receding, this will result in an overall decrease in limber pine numbers.
Background: Clarks Nutcracker
The Clarks Nutcracker feeds on conifer seeds when cones ripen, caches seeds during late summer and fall, and retrieves seeds until the next cone crop is available (Tomback 1978). Since the seeds in these caches may germinate if not retrieved, this has resulted in some species of pine coevolving with the nutcracker and becoming adapted for bird dispersal (Tomback and Linhart 1990). In the U.S., these species include whitebark pine, limber pine, southwestern white pine (Pinus strobiformis), Colorado pinyon pine (Pinus edulis), and single-leaf pinyon pine (Pinus monophylla). Because of the large, wingless seeds of these bird-dispersed pines, nutcrackers prefer to harvest them over the seeds of other conifers and serve as major seed dispersers for these species (Tomback 1978, Lanner and Vander Wall 1980, Tomback and Linhart 1990).
Nutcrackers are highly adapted to forage on conifer seeds, but they still retain the ability to eat foods opportunistically, which is characteristic of corvids. They will eat arthropods, carrion, eggs, nestlings, and small mammals (Decker and Bowles 1931, Cottam
5


1945, Bent 1946, Dixon 1956, Giuntoli and Mewaldt 1978). It is not known how much they depend on these alternative food sources and how this changes seasonally and from year to year. An additional food source that they utilize in certain areas is handouts from tourists (Tomback and Taylor 1987). In RMNP, nutcrackers are abundant at scenic turnouts where tourists commonly stop and feed animals, despite rules against this. Some nutcrackers become regulars at these turnouts, even bringing their young. There is also some evidence that the young brought to turnouts will become regulars themselves (Tomback and Taylor 1987). Multiple issues potentially arise from this: if too many nutcrackers become regulars and do not cache seeds as a result of their dependence on handouts, bird dispersal of limber pine seeds (and thus regeneration) may be diminished. On the other hand, this abundance of food may affect juvenile survival or cause a significant rise in the nutcracker population size through increased immigration. This increase may cause more unripe seeds to be harvested, leaving fewer ripe seeds to be cached.
The nutcrackers seasonal activities are closely tied to the timing of conifer seed production. Nutcrackers will begin harvesting unripe seeds from closed cones in late summer and cache ripe seeds from late August to December (Tomback 1998). Which conifer seeds the nutcracker harvests and caches will differ both across time and region due to different conifer communities, yearly cone crop variation, nutcracker preferences, and various ripening phenologies. In the case of limber pine, nutcrackers will begin harvesting unripe seeds in late July and cache seeds from late August or early September until late October or early November (Vander Wall 1988).
Nutcrackers show general caching behaviors regardless of the seed source. These are summarized from Tomback (1978) and Vander Wall (1988). It is indicated that nutcrackers
6


have been observed to cache seeds in both divergent and convergent storage areas (terminology from Tomback 1978). When caching in divergent storage areas, the nutcrackers are alone or occasionally in pairs. They may cache seeds in the ground under leaf litter, near the base of a tree, rock, or log, in gravel or volcanic pumice substrate, in trees under bark or in a hole, and in disturbed areas such as burns. When caching seeds in the substrate, they typically swipe their bill along the ground to make a trench and bury 1-15 seeds (average of 3) before covering the trench with substrate and sometimes with a cone or pebble.
Convergent areas are communal storage areas, with local populations of nutcrackers potentially caching seeds in the same terrain. These are typically located on steep, southfacing slopes near stands of pines.
These caches provide nutcrackers with food during times when little other food is available. During the winter and early spring, nutcrackers will retrieve seeds from caches (although some seeds in cones are still available during winter). Nutcrackers will feed their young seeds from caches. They nest relatively early, with eggs often hatching in March or April (Bent 1946, Mewaldt 1956, Campbell et al. 1997). Early nesting in this species seems counterintuitive given the harsh environments they inhabit. It is thought that this is an adaptation which allows juveniles to reach independence by the time seed caching begins (Vander Wall and Baida 1977, Tomback 1978). Once the young have fledged, they will follow their parents and watch them harvest, cache, and retrieve seeds (Tomback 1978, Vander Wall and Hutchins 1983). By watching their parents, they will be prepared to make caches themselves in the fall.
Observing the Clarks Nutcrackers seasonal activity resulted in two hypotheses regarding their spatiotemporal use of space. One hypothesis was that the nutcrackers would
7


migrate altitudinally within a region to follow the production of cone crops of different pines (Tomback 1978). The other hypothesis proposed that nutcrackers either behaved as emigrants or residents (Vander Wall et al. 1981). Under this hypothesis, emigrants migrated latitudinally in search of productive cone crops while residents did not migrate, but did move altitudinally in search of cone crops, and remained within their home range year round (except during excessively poor cone crop years).
Results from radio telemetry conducted in the Cascade Mountains, Washington, supported the second hypothesis (Lorenz and Sullivan 2009). Nutcrackers occupied a home range for winter, spring, and summer (year round home range). However, nutcrackers expanded their range in the fall to search for cones. More research needs to be conducted on this question of spatial use because nutcrackers may behave differently across their wide range as a result of different climates and conifer communities.
Since nutcrackers forage on multiple conifer species, a poor cone crop in one of their major conifer seed sources can be compensated by seed production in other conifers (e.g., Tomback 1978, Giuntoli and Mewaldt 1978). However, it has also been shown for whitebark pine that nutcrackers are not as likely to visit stands with low cone production as they are to visit stands with greater cone production (McKinney and Tomback 2007, McKinney et al. 2009, Barringer et al. 2012). For example, the probability of observing nutcrackers when no cones are present is 0.22-0.35 (Barringer et al. 2012). The issue is that in stands with high mortality from blister rust and mountain pine beetle, not many cones will be produced even with a good cone crop. If these stands are not visited, little caching occurs, and regeneration by nutcrackers will be less likely. In addition, if mortality is widespread, it is unknown how
8


this will affect the range and movement patterns of nutcrackers. The nutcrackers may emigrate more frequently, but there is speculation that they may leave the area entirely.
Nutcracker visitation is important not only for natural regeneration, but also for local white pine restoration projects. As a result of nutcrackers being either a primary or major seed disperser for several white pines, the Proactive Restoration Strategy in the southern Rocky Mountains depends on natural seed dispersal to spread potentially rust-resistant genotypes (Schoettle et al. 2011). The purpose of this strategy is to protect white pines in advance of serious damage from the threats previously discussed. Providing disturbed areas for nutcracker caching by thinning or placing controlled bums near stands with trees that have known genetic resistance to blister rust is one strategy (Schoettle et al. 2011).
9


CHAPTER II
THE IMPORTANCE OF MULTIPLE SEED SOURCES: CLARKS NUTCRACKER SEED USE IN ROCKY MOUNTAIN NATIONAL PARK
Introduction
Clarks nutcrackers (.Nucifraga Columbiana) are a widely-ranging keystone species of western coniferous forests whose ecosystem services include long distance seed dispersal, local tree establishment after disturbance, and forest regeneration over time for several western pines (Tomback 1998, Tomback 2001, Tomback and Kendall 2001). Across their range, nutcrackers utilize these and additional conifer species as a food source. Several conifer species that are not strongly dependent on nutcrackers for seed dispersal may be more important to nutcracker ecology and the stability of population numbers than previously recognized.
Studies often emphasize that nutcrackers use pines with large, wingless seeds as mainstay or preferred food resources. For example, Tomback (1978) describes nutcrackers using whitebark pine (Pinus albicaulis) seeds in California and Benkman et al. (1984) and Vander Wall (1988) describe nutcrackers using limber pine (Pinus flexilis) seeds in Arizona and Utah, respectively. In late summer and fall, nutcrackers forage on the seeds of these two pine species and transport them within their sublingual pouch to cache them nearby or tens of kilometers away as a winter and spring food source (Vander Wall and Baida 1977, Tomback 1978, Lorenz and Sullivan 2009). These two conifer species are mutualists of nutcrackers, relying on the birds for seed dispersal (Tomback and Linhart 1990). Nutcrackers enable small, isolated stands of limber pine to function together as a metapopulation by connecting them through seed dispersal flights (Webster and Johnson 2000, Williams, Chapter III). Nutcrackers also forage on and disperse the large, wingless seeds of southwestern white pine
10


(.Pinus strobiformis., Benkman et al. 1984, Samano and Tomback 2003), Colorado pinyon pine {Pinus edulis; Vander Wall and Baida 1977), and singledeaf pinyon pine {Pinus monophylla\ Vander Wall 1988). Each of these conifer species is an important regional food source.
In addition, some authors describe nutcracker use of conifer species with smaller, winged seeds in different parts of the nutcracker range (Tomback et al. 2011). Tomback (1978) reports the harvest and caching of Jeffery pine {Pinus jeffreyi) seed following whitebark pine seed use in the eastern Sierra Nevada. Giuntoli and Mewaldt (1978) report nutcracker stomach contents including ponderosa pine {Pinusponderosa) and Douglas-fir {Pseudotsuga menziesii) seeds in addition to whitebark pine seeds in western Montana; Vander Wall et al. (1981) note nutcracker harvest of Douglas-fir seeds in Utah; Lorenz et al. (2009) reports use of ponderosa pine and Douglas-fir seeds in the Cascade Range, Washington; and, Schaming (2016) observed nutcrackers to forage on Douglas-fir seeds in Wyoming. Nutcrackers are also known to forage on the small, winged, seeds of bristlecone pines {Pinus longaeva and Pinus aristata; Lanner 1988, Torick et al. 1996). In addition to these smaller-seeded conifer species, nutcrackers predictably forage on the larger, winged seeds of sugar pine {Pinus lambertiana\ Murray and Tomback 2010, Turner et al. 2011). While some of these smaller, winged conifer species have limited distributions and use of these seed sources is thus geographically restricted, I propose that several widely-distributed seed sources are ecologically crucial to provide sufficient seed production to sustain nutcracker populations over time.
Conifer species produce variable cone crops each year (Krugman and Jenkinson 1974, McCaughey and Tomback 2001), and nutcrackers are known to emigrate when little
11


food is available (Davis and Williams 1957, Davis and Williams 1961, Vander Wall et al. 1981). During years of low cone production by large, wingless-seeded pines, other seed sources enable nutcrackers to store sufficient food for winter and early spring. These conifer species may also help to stabilize the mutualisms between nutcrackers and five-needle white pines (Tomback and Linhart 1990). While some mutualisms appear specific, they may persist over time only because of the presence other conifer species.
Alternative seed sources may be especially important as populations of whitebark and limber pine (as well as southwestern white, sugar, and bristlecone pinesall five-needle white pines, subgenus Strobus) decline because of multiple anthropogenic stressors (Tomback and Achuff 2010, Tomback et al. 2011). Severe mountain pine beetle (Dendroctonusponderosae) outbreaks, attributed to recent high temperatures and drought, spread into higher-elevation forests and impact pine species other than their historical host lodgepole pine (Pinus contorta) (Gibson et al. 2008, Logan et al. 2010). Fire-suppression has also resulted in advanced succession in many forest communities, leading to declines in five-needle white pine populations (Gruell 1983, Rebertus et al. 1991, Murray et al. 2000). Climate change is expected to shift five-needle white pine populations north and to higher elevations while southern and lower elevation populations decline; ranges may shrink if colonization is slow (Tomback and Achuff 2010). The ongoing spread of white pine blister rust (caused by the non-native pathogen Cronartium ribicola), the most serious threat, is killing five-needle white pines and causing up to 90% mortality in some stands (Kearns and Jacobi 2007, Tomback and Achuff 2010).
Some conifer species whose seeds are harvested by nutcrackers exhibit comparatively small ranges, including Jeffrey pine, foxtail pine, the bristlecone pines, and sugar pine (Little
12


1971); these species, therefore, provide only regional or localized benefits for nutcracker populations. However, ponderosa pine and Douglas-fir exhibit comparatively expansive geographic ranges and locally large forest expanses, potentially serving as extensive food sources for nutcrackers. These conifers are the only two seed sources to occur in all 11 western states, and there is much overlap in range with other pines, with ponderosa pine extending more into the dry southwest and Douglas-fir ranging north into Canada (Little 1971).
In northern Colorado, limber pine is the only widely-distributed large-seeded pine; it is a preferred food source for nutcrackers in this region (e.g., Tomback and Taylor 1987, Torick et al. 1996). In addition to large, wingless seeds, limber pine tree morphology, like whitebark pine, features upswept branches with horizontally-oriented cones in whorls at branch tips that allow for efficient assessment and foraging by nutcrackers (Lanner 1980). Nutcrackers begin foraging on green, unripe limber pine cones during mid to late summer, using their long, sturdy bill to pry open cone scales (Vander Wall and Baida 1977, Tomback and Taylor 1987). They make late summer and autumn caches in the soil using sidesweeping motions of the bill to create a trench, and then insert three to four seeds individually, each a few centimeters deep (Lanner and Vander Wall 1980, Vander Wall 1988). Nutcrackers are prolific seed dispersers, estimated to cache 16,000 limber pine seeds each year per bird (Vander Wall 1988). Seeds within unretrieved nutcracker caches may germinate after spring rains or snowmelt and lead to forest regeneration.
Whereas nutcrackers may rely on limber pine seeds in northern Colorado, ponderosa pine and Douglas-fir are also widely available in this region and may provide supplemental seed sources. However, the extent to which nutcrackers rely on these species in this region is
13


unknown. White pine blister rust was first detected in this region in 1998 (Johnson and Jacobi 2000), with an infected tree in Rocky Mountain National Park (RMNP) in 2010 (Schoettle et al. 2011). Blister rust infecting tree canopies reduces cone production and photosynthetic biomass, weakening trees; blister rust infection in tree stems kills trees (McDonald and Hoff 2001, Geils et al. 2010). Estimates suggest that up to 50% of limber pine habitat in northern Colorado is susceptible to blister rust (Kearns 2005, Howell et al. 2006). A limber pine restoration plan is currently being implemented in RMNP to minimize mortality from pine beetles and blister rust, involving insecticide and verbenone application, seed collections, screening for blister rust resistance, and developing artificial regeneration guidelines (Schoettle et al. 2011). With increasing limber pine mortality in the coming decades as a result of blister rust, however, nutcracker populations will come to rely more on alternative seed sources if nutcracker populations remain in northern Colorado.
I examined nutcracker use of conifer seeds across three field seasons in RMNP, a relatively undisturbed natural area that features diverse mountain topography and representative northern Colorado forest communities. My goal was to determine the relative importance of limber pine, ponderosa pine, and Douglas-fir as food sources and as habitat to nutcrackers in RMNP. To accomplish this goal, my objectives were to: 1) estimate annual cone production of the limber pine, ponderosa pine, and Douglas-fir forest types, and the total energy each conifer species provided across the landscape in RMNP; 2) estimate inter-and intra-annual patterns of nutcracker visitation of each forest type; and 3) examine the relative numbers and timing of nutcrackers foraging on and caching the seeds of each conifer species and the odds of nutcrackers using seeds of different conifer species. Historically, RMNP has supported a robust population of nutcrackers (Tomback and Taylor 1987). By
14


understanding how nutcrackers utilize alternative seed sources in a system not yet impacted by blister rust, we can understand the implications of increased limber pine mortality. Fewer limber pine trees will reduce carrying capacity in the nutcracker population; complete loss of nutcrackers, a keystone species, is expected to have ecological impacts on many other pines that benefit from their seed dispersal services (Tomback and Kendall 2001, Tomback et al. 2011). In addition, because ponderosa pine and Douglas-fir are also the most widespread alternative seed sources across the nutcrackers range, the results of this study could apply broadly.
Methods
Study Area
I studied nutcrackers on the eastern slope of RMNP from June to October, 2014 to 2016 (Fig. 1). Park elevations range from 2,380 to 4,350 m; study stand elevations ranged from 2,400 to 3,400 m. As determined from ArcGIS layers obtained from the National Park Service, within RMNP, there are communities of ponderosa pine (3,396 ha), Douglas-fir (4,327 ha), and limber pine (2,212 ha)the forest types of interest. Because RMNP is a high-elevation park, it only includes the western extent of ponderosa pine and Douglas-fir communities on the northeastern front; these forest types continue outside park boundaries further east at lower altitudes. For example, 12,000 ha of ponderosa pine woodlands and 11,000 ha of Douglas-fir forests are located within approximately 10 km east of the RMNP boundary in the Roosevelt National Forest. Since nutcrackers do not recognize park boundaries, some may have entered and exited RMNP during the study period while using these forests. Other habitat within the park includes lodgepole pine forests, spruce-fir (Picea engelmannii and Abies lasiocarpa) forests, park meadows, riparian zones, alpine tundra, and
15


talus slopes. Average minimum January temperature in nearby Estes Park is -9C and average maximum July temperature is 26C. Estes Park receives 13.10 inches of precipitation per year on average. The wide elevation range within RMNP results in highly variable climatic conditions.
Field Methods Stand Selection
I selected stands of ponderosa pine, Douglas-fir, and limber pine for cone density estimation, nutcracker presence/absence surveys, and focal behavior surveys. Here, a stand is defined as a continuous forested area where one conifer species of interest dominates. I located stands by scouting RMNP and examining ArcGIS vegetation layers obtained from the National Park Service. All three forest types are primarily located on the eastern slope of RMNP, and I methodically selected stands that span this distribution. Stand selection was based on proximity to the parks road or trail system, enabling me to visit two stands per day for presence/absence surveys and focal behavior sampling (not possible with the Ute Trail and Estes Cone stands). Stands also needed to be large enough, approximately five hectares in area, to fit a transect. Once a stand was identified that met these criteria, I selected it as a study stand only after visiting it to ensure that it was accessible and the correct forest type. In 2014,1 selected two stands of Douglas-fir, three stands of ponderosa pine, and five stands of limber pine. In 2015 and 2016, evenly distributing survey time, I opted for four stands of each species by removing one limber pine stand and adding stands of ponderosa pine and Douglas-fir.
I established a virtual transect using GPS points within each stand for all three survey types. Transects were a maximum of 1000 m in length when possible (min = 300 m, mean =
16


817 m) and consisted of points separated by 100 m. Transects consisted of either one continuous line (n = 10 stands), two lines (n = 2 stands), or three lines (n = 1 stand). I placed transects within the stand randomly using ArcGIS when possible, but if constrained I placed them along an accessible ridge, slope, or trail. Constraints on transect length, connectedness, and random placement resulted from small stand area or hazardous terrain.
Estimation of Tree Density, Cones/Tree. Cone Density, and Cone Insect Infestation
I surveyed each stand to estimate cone density for each conifer species once per year between June 13th and September 13th. During this time period, unripe cones are easy to distinguish from old cones and large enough to observe for all three conifer species. I used distance sampling to model cone detection probability (Buckland et al. 2001). While walking a transect, I recorded the following information from each observed tree: distance to tree from observer using a laser rangefinder (Nikon ProStaff 550), direction of tree from observer using a compass (Silva Ranger), and number of cones observed from transect using 10x42 mm binoculars. Information was recorded only for trees within 50 m because observing cones beyond this distance is difficult without a scope. To ensure that all recorded trees were large enough to bear cones, I only included trees that were at least two meters in height. Because the observer cannot leave the transect during distance sampling, I only counted cones on one side of each tree (a partial count), even if the tree was near the transect. The disadvantage here is that Program Distance will not estimate the number of cones present on the unobserved sides of trees, resulting in low estimates of cone density for each forest type. Because I am interested in only comparing the relative cone production between conifer species, this disadvantage does not affect my interpretation of Program Distance results.
17


In 2014,1 only sampled between 100 to 1000 m of each transect (median = 500 m); I walked transects more slowly in stands with higher tree density. Based on Program Distance analyses of 2014 data, I determined that sampling 100 trees per transect would be sufficient for good model fit (based on a chi-square test; described below), cone density estimates without high standard error, and collecting data for one transect within one field day. To obtain 100 trees in 2015,1 sampled only 50 to 350 m of each transect (median = 200 m), randomly selected the starting point of each survey within a transect when possible (n = 7 stands). I also conducted a full cone count for every fifth tree (in addition to a partial count for that tree)by circling the tree and counting cones not observed from the transect. This allowed me to estimate the proportion of cones missed on observed trees while on the transect. The full cone counts were only utilized in the approximate estimation of landscape energetics (described below). I used the same methodology in 2016.
Infestation of ponderosa and limber pine cones by insects, identified by small light brown or reddish-brown cones filled with frass, occurred during my study. These cones do not provide any food energy to nutcrackers because the seeds are inedible. I was unable to identify the species of insects causing infestation, but they were likely either ponderosa pine coneworms (Dioryctria auranticella) or lodgepole pine cone beetles (Conophthorus contortae) (Schoettle and Negron 2001). To estimate the extent of infestation, I sampled 20 ponderosa pine trees in October 2015 and 20 limber pine trees in September 2016 during presence/absence surveysfive trees from each stand, selecting the nearest cone-bearing tree from five points per transect. I circled each tree and counted the number of healthy and infested cones to estimate the percentage of infested cones for each tree and for each conifer
18


species as a whole. These estimates provide further details for the amount of food energy available to nutcrackers.
Nutcracker Presence/Absence Surveys
I surveyed stands to examine trends in nutcracker presence in forest types over time. I walked transects from August 11th to October 31st each year. At the beginning of these surveys, cones in all three forest types were green and unripe, and they ripened throughout the field season; nutcrackers may harvest seeds from ripening or ripe seed sources (Tomback 1978, Tomback and Taylor 1987). I conducted from two to seven surveys per transect in 2014, with 20 total surveys in limber pine stands, 10 surveys in ponderosa pine stands, and six surveys in Douglas-fir stands. I conducted from four to seven surveys per transect in 2015 and 2016, with 20 total surveys within each forest type each year. I attempted to distribute surveys for each transect across the field season so that different stages of cone ripening were included in the surveys. In 2015 and 2016,1 attempted to visit each transect once every other week.
At each point along the transect, I listened and watched for nutcrackers for 10 min and recorded whether I detected nutcrackers within the stand. I walked two transects per day when possible; the number of points included in each survey depended on the amount of time that could be devoted to the survey and any environmental hazards, including extreme weather, moose presence, and rutting elk (minimum percentage of points included in a survey = 18%; mean = 88%).
Focal Behavior Surveys
I surveyed nutcracker foraging and caching behavior each year to estimate trends over time and to compare the odds of nutcracker seed use for each conifer species. Focal behavior
19


surveys were conducted during presence/absence surveys (n = 156 surveys), commencing as I entered the stand, and following a cone count if conducted after August 10th (n = 8 surveys). During a behavior survey, I approached a detected nutcracker if within -100 m. I halted presence/absence surveys during this time. Seed foraging and caching were the behaviors of interest, including the seeds of limber pine, ponderosa pine, and Douglas-fir. Once the bird was within view, I recorded each behavior and the time allocated to each behavior. There were three instances in 2016 where I observed a bird either holding a cone or flying with a cone for only 5 to 10 sec; I included this as foraging behavior. I recorded all other behaviors, including calling, perching, preening, bill wiping, and flying, as other, also recording the associated conifer species. I ceased recording a birds behavior when the bird flew out of view or after 10 min of behavior data were collected. If other birds were present, I began recording data on the next visible bird after data recording had ceased on the previous bird. A behavior survey ceased when I collected a total of 60 min of behavior data or I exited the stand. Survey time was used to normalize the foraging time of nutcrackers, and was calculated as the duration between commencement and cessation of a behavior survey. I did not record stand entrance or exit times (used to calculate survey time) in 2014 and for some of the first surveys in 2015, and so I estimated them based on 2015 and 2016 data of the duration between the times of stand entrance and conducting the first point (for entrance time) and the duration between the times of conducting the last point and exiting the stand (for exit time).
20


Statistical Analyses
Tree and Cone Density and Landscape Energetics
To prepare distance sampling data for analysis, I first calculated sighting angle (9) for each tree by calculating the difference between the tree direction and transect direction. I then used the distance from observer (r) for each tree to calculate perpendicular distance from transect (x) by using the equation:
x = r sin (9) (1)
I used Program Distance (Thomas et al. 1998) for all Distance Sampling analyses. This program estimates the proportion of objects not detected, which allows for a more accurate estimate of density. To calculate object density, Program Distance plots a histogram of the number of detected objects against x. A detection function, g(x), is then plotted along the bin peaks in the histogram. The equation of g(x) may be defined by one of four key functions: uniform, half-normal, hazard rate, and negative exponential. Each key function may be further defined by three series expansions: cosine, simple polynomial, and hermite polynomial. I selected the key function and series expansion based on the lowest AIC values. However, I rejected the negative exponential key function because it was not appropriate for my data sets (I believe that none of the forest types should result in a spiked distribution on the histogram); the remaining key functions generally estimated similar densities, with the 95% confidence intervals overlapping nearly completely. The observed number of objects (n) is divided by transect length (L) and the integral of g(x) to obtain the point estimate of true density.
I calculated tree density for each stand by analyzing each stand individually and including transect data from all years. I also calculated tree density for each forest type by
21


including data from all stands of a forest type across all years, stratifying by stand, and using a weighted average based on stand area. For each year, I calculated cone density for each forest type by combining data from all stands of a forest type within that year, stratifying by stand, and using a weighted average based on stand area. Trees were only included in the cone density analysis if they contained at least one cone. I examined the histogram and left-truncated the data if bins near the transect contained relatively few observations and right-truncated the data if outliers were present. I evaluated the fit of the models, g(x), with a chi-square goodness of fit test, which compares the expected histogram values of the model to the observed histogram values. Low X2 values indicate that the models are significantly different than the observed data and are a poor fit.
Using cone density data, I estimated the approximate amount of energy on the landscape for each conifer species. I obtained estimates of full tree cone counts for each conifer species for use in landscape energetics as follows: I plotted partial counts (from transects) against full counts for each conifer species, combining 2015 and 2016 data. I used linear regression to plot a line of best fit, with a y-intercept of 0; otherwise, the y-intercept can result in a significant inflation of the density estimate. For each conifer species, I multiplied the slope of the line by the number of cones observed from the transect for every tree to obtain a total cone estimate for each tree. I then analyzed these data using the same Program Distance protocol described above. To estimate landscape energetics, I examined the literature for data on seeds per cone and kJ per seed (either directly or I calculated it from calories per gram and grams per seed) for each conifer species. I then multiplied cones per hectare by area of distribution in RMNP (in hectares) and by seeds per cone to obtain the
22


total number of seeds present in RMNP for each conifer species. I then multiplied that estimate by kJ per seed to obtain the total kJ estimate in RMNP for each conifer species. Nutcracker Presence and Foraging Patterns
Each year, I pooled presence/absence data into 10-day bins to estimate habitat use changes over time, beginning with August 11th and ending October 31st. Within each bin, I calculated percentage of surveys with detections for each forest type by summing the number of surveys with detections and dividing by the total number of surveys. I plotted the results on a time-series graph with 10-day bins along the x-axis and the percentage of surveys with detections along the y-axis. Because these graphs do not account for differences in detection probability between forest types, I focus on how visitation changes over time within each forest typemaking the assumption that detection probability stays relatively consistent over time.
I analyzed seed use data with graphic analysis and an odds ratio test. For the time-series graphs, each year I pooled foraging observations into 10-day bins, beginning with August 11th and ending with October 31st. Within each bin, I summed foraging time across all birds for each conifer species, summed the survey time for each conifer species, and divided foraging time by survey time. I plotted the results on a time-series graph with 10-day bins along the x-axis and foraging sec per survey min along the y-axis in order to examine the time sequence for nutcracker foraging on different conifer species, how many days they forage on different conifer species, and the timing of transition between conifer species.
The Odds Ratio test compares the odds of nutcrackers using the seeds (foraging or caching) of a given conifer species to the odds of using the seeds of another conifer species (Rita and Komonen 2008). Each year I summed the number of birds observed using seeds
23


and not using seeds (those marked as other) for each conifer species. Conifer species was not assigned by the forest type being surveyed, but by the conifer species selected by the nutcracker. For example, if a bird foraged on an isolated limber pine tree within a ponderosa pine stand, the assigned conifer species would be limber pine. If behavior was other, I used the forest type being surveyed unless the bird was part of a foraging flock; here I used the conifer species being utilized by the flock. Birds within a foraging flock were always observed to forage on the same conifer species, with one exception in 2016 when one bird briefly foraged on Douglas-fir seeds while the other birds foraged on limber pine seeds. This was within a Douglas-fir stand, and for other birds the associated conifer species would be limber pine here. Conifer species were only included in the analysis if nutcrackers used the seeds of that conifer species within a given year. Birds that removed seeds of an unknown conifer species from caches (one bird in 2016) were excluded. Birds that were caching seeds of an unknown conifer species (three birds in 2016) were excluded. I included all birds regardless of duration of seed use (5 to 600 sec). I used RStudio (version 3.24, R core team 2016) for the Odds Ratio test.
Results
Tree and Cone Density, Cone Infestation, Landscape Energetics, and Ripening Phenology
The geographic data and estimated tree densities of each stand are reported in Table 1. Point estimates for limber pine tree density within stands ranged from 64 to 654 trees per ha (Table 1). Point estimates for ponderosa pine tree density within stands ranged from 120 to 203 trees per ha. Point estimates for Douglas-fir tree density within stands ranged from 152 to 1086 trees per ha. Multiple conifer species were present within each stand, especially
24


Engelmann spruce and subalpine fir within limber pine stands and lodgepole pine within limber pine and Douglas-fir stands; ponderosa pine woodlands often lacked tree diversity. The reported tree density only applies to the focus seed source within each stand.
I observed a unique combination of cone crops among all conifer species each year, producing multiple scenarios of nutcracker foraging behavior. Cone crops were relatively small for all conifer species in 2014, ranging from 180 to 480 cones per ha (Fig. 2). On average, for all conifer species, average cone productivity ranged from 0.4 to 5 cones per tree. In 2015, limber pine and Douglas-fir again produced small cone crops, ranging from 200 to 810 cones per ha (Fig. 2), with average cone productivity of 0.4 to 2 cones per tree. However, ponderosa pine cone density was much higher in 2015an estimated 6,830 cones per ha (Fig. 2), with an average productivity of 51 cones per tree. In 2016, cone density for limber and ponderosa pine ranged from 510 to 850 cones per ha (Fig. 2), with average cone productivity of 2 to 4 cones per tree. However, Douglas-fir cone density was an estimated 48,830 cones per ha (Fig. 2), with an average cone productivity of 97 cones per tree. It is noteworthy that limber pine experienced a comparatively small cone crop in all three years while both ponderosa pine and Douglas-fir each experienced one large cone crop year: one in 2015 and one in 2016.
Cone infestation by insects was estimated to be extensive. Within the 20 ponderosa pine trees I examined for cone infestation in 2015, 53% of the cones showed infestation by cone insects. Cone infestation of individual ponderosa pine trees ranged from 8% to 97%. Within the 20 limber pine trees I examined in 2016, 79% of the cones showed infestation by cone insects. Cone infestation by insects of individual limber pine trees ranged from 0% to 100%. I did not include a portion of these infested limber pine cones in the distance sampling
25


cone counts (for cone density estimates), because some infestation could be detected by June or July. For the cones included in distance sampling cone counts (and therefore program Distance), I estimate that 57% of limber pine cones were infested. It is possible that an unknown small percentage of infested cones were from previous years. Cone infestation occurred in all three years of the study for these two conifer species.
Using the results of regressions based on full cone counts (Figure 3), distance sampling results, and other seed weight and energy values from the literature, I estimated the energy available across the landscape in RMNP for each conifer species (Table 2). These estimates of landscape energetics are approximate and rough; this is an exploratory exercise and inferences are limited. Point estimates of limber pine cone densities across years ranged from 800 to 1,400 cones per ha. Using values reported in Table 2,1 estimated the energy across the RMNP landscape provided by limber pine seeds during this study to range from 4.1 108 to 7.3 108 kJ. Ponderosa pines point estimates of cone density ranged from 700 to 9,600 cones per ha. I estimated the energy across the RMNP landscape provided by ponderosa pine seeds during this study to range from 1.4 108 kJ to 1.9 109 kJ. Point estimates of Douglas-fir cone densities ranged from 300 to 92,900 cones per ha. I estimated the energy across the RMNP landscape provided by Douglas-fir seeds during this study to range from 1.6 107 kJ to 4.3 109 kJ.
Each conifer species studied followed a unique ripening phenology; ripening entails the transition from unripe cones with closed scales, to scales slightly separating from the cone core. Unripe limber pine cones are green; I observed them to begin ripening from late August to early September. Unripe Douglas-fir cones appear green or purple and began ripening from mid to late September. Unripe ponderosa pine cones are green and began
26


ripening from late September to early October. During ripening for all three conifer species, the cones transitioned to a brown color and cone scales separated to reveal seeds.
Nutcracker Stand Visitation
Nutcrackers were observed in all forest types in all three years (Fig. 4). Changes in nutcracker detections were often coincident with cone ripening for each of the conifer species. In 2014,1 detected nutcrackers more frequently in limber and ponderosa pine stands when cones began ripening for each species. Relatively few surveys were conducted in 2014, which may have resulted in more variable estimates of detection over time. In 2015,1 again detected nutcrackers more frequently in limber and ponderosa pine stands after cones ripened; this increase was especially ephemeral in limber pine stands. I also detected nutcrackers more frequently in Douglas-fir stands in 2015 after cones began ripening. In 2016, nutcracker detections in limber pine stands peaked in late August and steadily dropped afterward. Relative to prior years, nutcracker detections in 2016 were comparatively low in ponderosa pine stands and comparatively high in Douglas-fir stands, with detections increasing as cones ripened for each species.
Nutcracker Foraging and Caching Behavior
I observed differences in the timing and behavior of nutcrackers foraging on each conifer species. I observed evidence of nutcracker damage on unripe limber pine cones (torn cones, with the red-brown exposed core contrasting with the unripe green color of the exterior) as early as August 7th. During this time, birds tore away scales and foraged on the developing seeds but were not caching seeds. Nutcrackers either positioned themselves atop an attached cone, or removed cones from branch ends and used a sturdy object as an anvil including rocks, branches, and logs. Once limber pine cones began ripening, nutcrackers
27


removed the ripe seeds directly from cones on trees and began pouching the seeds for transportation to caching sites. Some birds were observed to cache seeds nearby in the ground or in a log, rock, or tree, while others flew off with full throat pouches to cache elsewhere. I observed nutcrackers foraging on unripe ponderosa pine cones as early as September 1st; they foraged on these cones in a manner similar to unripe limber pine cones. Once ponderosa pine cones ripened, nutcrackers removed seeds by loosening cone scales and then removed the seed, holding it in their bill. Nutcrackers then removed the seed wing by either shaking the seed or rubbing it against a tree branch. Birds were observed to pouch seeds, with some birds caching nearby and others flying out of view to cache elsewhere. One nutcracker was observed to forage on unripe Douglas-fir cones on August 12th, but most foraging on these cones was observed after ripening. Within 10 to 15 days of Douglas-fir cones ripening, nutcrackers were observed to carefully remove intact Douglas-fir seeds from between the cone scales. Nutcracker flocks were later observed to forage primarily on the ground for Douglas-fir seeds that had been blown out of cones. Similar to ponderosa pine, birds removed the seed wings of Douglas-fir seeds which appeared to detach more easily. Some birds were observed to pouch Douglas-fir seeds and subsequently cache them approximately 20 meters away or less.
In 2014, when cone crops were relatively small for all conifer species, I observed nutcrackers to forage only on the seeds of limber pine and ponderosa pine (Fig. 5). Nutcrackers in 2014 were first observed to forage on unripe limber pine cones in late August, continuing to forage on ripened cones that September. Beginning in early October, observations of nutcrackers foraging on limber pine seeds declined, and instead I observed them foraging on ponderosa pine seeds, shortly after seed ripening. At the end of October, a
28


small flock of nutcrackers was observed to raid red squirrel (Tamiasciurus hudsonicus) middens for closed limber pine cones in the Estes Cone stand. Squirrels often chattered at and chased the nutcrackers during this time, especially if a nutcracker perched nearby after removing a cone. Nutcrackers mostly used tree branches as anvils when foraging on these cones. In 2015, when ponderosa pine experienced a large cone crop, nutcrackers again foraged only on limber and ponderosa pine seeds (Fig. 5). While nutcrackers again began foraging on unripe limber pine cones in late August, they were observed to begin foraging on ponderosa pine cones one month earlier in 2015. This was the only year that nutcrackers were observed to forage on unripe ponderosa pine cones. In 2016, when Douglas-fir experienced a large cone crop, nutcrackers were observed to forage first on limber pine seeds (with one bird briefly observed to forage on unripe Douglas-fir seeds) until late August, and then they focused only on ripe Douglas-fir seeds (Fig. 5). This is the only year that I observed nutcrackers to forage on Douglas-fir seeds and also the only year I did not observe nutcrackers to use ponderosa pine seeds. This year was also unique in that I observed the majority of nutcracker foraging on limber pine prior to cone ripening of this conifer.
I have observed nutcrackers to utilize two conifer species each year, but the odds of use were not always equal. In 2014,1 observed that about half of the nutcrackers near ponderosa pine trees to use ponderosa pine seeds while the majority of birds near limber pine trees used limber pine seeds (Table 3). The odds of nutcrackers using ponderosa pine seeds in 2014 were from 0.05 to 0.77 times lower than the odds of nutcrackers using limber pine seeds. In 2015, ponderosa pines large cone crop apparently influenced nutcracker foraging behavior, because the odds of nutcrackers using ponderosa pine seeds were now from 1.34 to 7.34 times greater than the odds of nutcrackers using limber pine seeds. In 2016, when I
29


observed nutcrackers foraging on Douglas-fir seeds for the first time in this study, the odds of nutcrackers using these seeds were from 0.81 to 6.98 times the odds of nutcrackers using limber pine seeds. The observed numbers of birds using and not using seeds for each conifer species are reported in Table 3.
Discussion
My goal for this study was to determine the relative importance of limber pine and alternative seed sources with respect to their potential value as a food source and as habitat to nutcrackers in RMNP. I first discuss my reasoning for assigning the relative importance of each food source followed by the relative importance of these habitats.
Across years, conifer species provided markedly different estimates of food energy across the landscape in RMNP. I found most cone density estimates for conifer species to be below 1,000 cones per ha, which I am defining as a relatively small cone crop, although I do not know the maximum that limber pine can produce (Fig. 2). I am defining limber pine cone production in relation to the greater cone density estimates of ponderosa pine in 2015 and Douglas-fir in 2016, which are large cone crops or mast years. Comparing the total kJ point estimates from Table 2, a low cone crop of limber pine results in 3.3 to 5.9 times more kJ of potential food energy in RMNP than a low cone crop of ponderosa pine and 27.7 to 49.2 times more kJ than a low cone crop of Douglas-fir, because of differences in seed energy per cone. In contrast, high cone production in ponderosa pine results in 2.3 to 4.1 times more kJ in RMNP than a low cone crop of limber pine. Mast cone production in Douglas-fir results in 5.4 to 9.6 times more kJ in RMNP than a low cone crop of limber pine. These comparisons do not account for insect infestation. If my estimates of cone infestation
30


by insects are accurate, total energy across the RMNP landscape provided by limber and ponderosa pine during my study would be approximately half of what I report here.
Nutcrackers are considered energy-sensitive foragers, and landscape energetics likely impact the foraging decisions of nutcrackers and their choices of conifer species as seed sources each year (Tomback 1978, Vander Wall 1988). In Table 2, we see that limber pine shows stable production of energy every year (even with low cone crops), Douglas-fir production is the most variable (ranging from relatively minimal energy to the maximum observed), and ponderosa pine production is relatively intermediate (with small crops producing more energy than a small crop of Douglas-fir and a large crop producing less energy than a large crop of Douglas-fir). Considering the changing landscape energetics each year, as well as the constants of seed size, cone morphology, and ripening phenology of each conifer species, I now discuss the nutcrackers use of seeds in RMNP across years.
Limber pine was the only seed source that nutcrackers used intensively every year. It was also the first seed source to be used each year, despite the fact that limber pine produced a relatively small cone crop each year, its stands are relatively small and scattered, and it comprises the smallest forested area of the three species on the landscape in RMNP. However, its large seed size, high landscape energetics during a low cone crop year, and early ripening date support the importance of this seed source. In addition, limber pine seeds ripen at what appears to be an important time in the summerafter most cached seeds from the previous fall were retrieved (e.g., Tomback 1978, Vander Wall and Hutchins 1993) but prior to seed ripening of other conifer species. The seeds of the heavily armored ponderosa pine, like the closely-related Jeffrey pine (Tomback 1978), may be energetically costly to remove from unripe cones. Nutcrackers appear to search widely for limber pine seeds as I
31


observed them to forage on isolated limber pine trees in stands of other forest types on multiple occasions. If some nutcrackers are residents within RMNP and remember the locations of productive limber pine trees within their home ranges, they may utilize seeds from these trees if energetic rewards are sufficient.
This suggests a landscape-level approach to seed utilization for some birds, in contrast to stand-based decisions. For the closely-related whitebark pine, observations have suggested that the probability of visitation or seed harvest by nutcrackers in stands increases as cone production increases (McKinney et al. 2009, Barringer et al. 2012), but Barringer et al. suggested that foraging decisions may occur at a landscape rather than stand level. In this study, however, cone production levels for limber pine and ponderosa pine were higher generally than many whitebark pine stands assessed by McKinney et al. (2009) and Barringer et al. (2012).
Nutcrackers were observed to forage intensively on ponderosa pine seeds for two out of three years, including a small cone crop year. The odds of nutcracker foraging on ponderosa pine seeds were lower compared to limber pine seeds during the low ponderosa pine cone crop year. This is unsurprising, considering ponderosa pines lower energetic reward during a small cone crop compared to limber pine. However, during ponderosa pines mast year, the odds of a nutcracker using ponderosa pine seeds were greater compared to limber pine, indicating that ponderosa pine importance to nutcracker diet may have been greater in 2015 compared to limber pine.
Nutcrackers foraged on Douglas-fir seeds only during the mast year. Even though the energy available for Douglas-fir during a mast year was estimated to be considerably greater than that for limber pine during a low cone crop year, the odds of nutcracker seed use were
32


not observed to be greater for Douglas-fir compared to limber pine that year. This likely results from limber pines earlier ripening phenology, seed size, and foraging energetics.
While the importance of each conifer species to nutcrackers as a food source appears to vary each year, I believe that over time each conifer species plays a role in maintaining nutcracker populations in RMNP. I argue that limber pine is the most important source of food because its seeds were used every year by nutcrackers and it fills a gap in time when other conifer seeds are not available. The odds of nutcracker foraging on limber pine seeds were higher compared to other conifer species experiencing a similarly sized cone crop they were even comparable to the odds of nutcrackers foraging on Douglas-fir seeds during a mast year. Comparing the alternative seed sources, I suggest that ponderosa pine has a greater role in maintaining nutcracker numbers over time than Douglas-fir, based on several reasons. The first is that Douglas-fir seeds were not observed to be used by nutcrackers during low cone production while ponderosa pine seeds were. The second involves how nutcrackers utilized seeds during a mast year for each species: (1) the odds of nutcracker foraging were higher for ponderosa pine compared to limber pine, the stable food source for nutcrackers, while the odds of foraging were comparable between Douglas-fir and limber pine (the odds of foraging were higher for ponderosa pine compared to Douglas-fir as well); and (2) nutcrackers were observed to forage on ponderosa pine seeds for a greater period of time in the fall compared to Douglas-fir. Finally, ponderosa pine generally experiences moderate to high cone production years more frequently than Douglas-fir (Krugman and Jenkinson 1974, Hermann and Lavender 1990).
Each forest type of interest appears to provide important habitat for nutcrackers because I observed them to visit each forest type across all years and occupy every study
33


stand. The production of seeds is an obvious incentive to nutcrackers to visit these forest stands, since I observed nutcrackers forage on the seeds of all three conifer species. However, during time periods and even years when I did not observe nutcrackers to forage on the seeds of a specific conifer species, nutcrackers were still present in those forest types. This indicates that there may be benefits to this habitat other than just food.
The importance of these two alternative seed sources also has been noted elsewhere in other studies. For example, Giuntoli and Mewaldt (1978) observed that nutcrackers had foraged on the seeds of both species in Montana. More recently, Lorenz and Sullivan (2009) observed nutcrackers to utilize both species in the Cascade Range of Washington State. Schaming (2016) observed nutcrackers in Wyoming to preferentially select Douglas-fir forests as habitat during the breeding season and she emphasized the importance of this species as a food source to nutcrackers in this region. As five-needle white pines experience mortality from blister rust and other disturbances, ponderosa pine and Douglas-fir may provide the most support as a backup food source across the entire range of Clarks nutcrackers.
Given that nutcrackers are energetically-sensitive foragers (Tomback 1978, Vander Wall 1988, Barringer et al. 2012), and I believe optimal foraging theory may help to explain conifer seed importance. While landscape energetics may detail how much energy is available to nutcrackers, we must also consider the rate of energy acquisition. Larger seeds allow for greater reward per extracted seed, but the time for seed extraction (finding and removing seeds from cones or on the ground) and processing time (removing seed wing, if present, and seed coat as well as assessment of seed soundness through bill clicking) is also likely to differ among conifer species. Nutcrackers were mostly observed to remove limber
34


and ponderosa pine seeds from ripening cones with opening cone scaleswhich reduces foraging time and effort. Tomback (1978) observed nutcrackers in California to forage on the seeds of whitebark pine (closely related to limber pine) at a faster rate compared to the seeds of Jeffrey pine (closely related to ponderosa pine); a similar relationship may exist between limber and ponderosa pine. Even if we assume that seed extraction rate is approximately equal between limber and ponderosa pine, but seed processing time is longer for each ponderosa pine seed because of seed wing removal, then limber pine leads to a faster rate of energy acquisition during low cone crops. However, a large ponderosa pine cone crop may lead to a faster rate of energy acquisition if the abundance of cones appreciably decreases seed extraction time. Douglas-fir seeds are considerably smaller than limber and ponderosa pine seeds (Table 2) and also have seed wings; nutcrackers foraging on Douglas-fir seeds would thus experience a very low rate of energy acquisition per seed. Nutcrackers may gain energetically only by foraging on these seeds during a large cone crop, with many cones within the same canopy. When nutcrackers removed Douglas-fir seeds from cone scales shortly after cones ripened, they appeared to do so carefully and slowly, which may result in lengthy seed extraction times. The mature cones also release seeds quickly, unlike ponderosa pine and limber pine, which are less synchronous. Later in the fall, nutcrackers foraged for Douglas-fir seeds only on the ground; I am unsure how this impacts seed extraction time. Considering similarities between nutcrackers foraging on limber pine and Douglas-fir in 2016, in terms of odds of seed use and length of time observed foraging, the rate of energy acquisition appeared to be similar between limber pine and Douglas-firprobably the result of the small Douglas-fir seed size and perhaps that nutcrackers are better adapted for foraging on limber pine. If this is the case, the greater energy across the landscape during Douglas-
35


firs mast year may not translate to greater use or preference by nutcrackers in comparison to limber pine.
My observations of nutcracker seed use are under conditions of comparatively low limber pine mortality from the mountain pine beetle outbreaks during the preceding 15 years and no observed mortality from white pine blister rust in RMNP. Increasing limber pine mortality will likely impact nutcracker foraging decisions as cone production is reduced from the disease. A major aspect of my study goal was to determine the importance of multiple seed sources to nutcracker foraging, as nutcrackers will need to rely on these species if limber pine seed production decreases in RMNP. I believe that the alternative seed sources in RMNP may support a nutcracker population, although the carrying capacity may drop and frequency of nutcracker emigration may increase (e.g., Vander Wall et al. 1981, Lorenz and Sullivan 2009). Based on my observations, ponderosa pine will be more important in supporting the nutcracker population over time. It is also likely that nutcrackers will use the seeds of both ponderosa pine and Douglas-fir to a greater extent than I observed.
It should be noted that RMNP is located at the western edge of the ponderosa pine and Douglas-fir forest zones, with additional forest area for these habitat types in adjacent national forest. If the entire extent of these zones are considered, they may serve an even greater role as food sources. Indeed, nutcrackers are present east of the park, where limber pine is much less common. If some of these birds are residents, they could be sustaining themselves primarily on ponderosa pine and Douglas-fir. If nutcrackers do not have access to limber pine seeds during late summer, they may instead sustain themselves on insects and occasionally on small vertebrates and any remaining cached seeds from the previous fall;
36


whether these food sources are adequate prior to cone ripening by ponderosa pine and Douglas-fir requires further study.
With the blister rust pathogen potentially spreading, as it as for the last century (Geils et al. 2010, Tomback and Achuff 2010), nutcrackers may not be limited to using only alternative seeds. It is estimated that insects infested 79% of all limber pine cones in RMNP in 2016. If we assume that there was no infestation but blister rust instead damaged 79% of limber pine cones in RMNP, we could theoretically expect to see similar foraging behavior by nutcrackers as was observed during this study, but, unfortunately, the effects from cone insects and blister rust are likely to be additive. In addition, I have detected nutcrackers in all stands across all years (including stands with low cone crops and/or no foraging), observed them to travel up to 12 km straight-line distance (Williams, Chapter III), and observed them to forage on isolated limber pine trees within stands of other forest types. This means that nutcrackers travel widely across the park, with individuals moving long distances to find seeds. Even with high mortality from blister rust, nutcrackers may be able to search for, forage on, and cache a portion of the limber pine seeds produced each year.
37


Tables
Table 1. Geographic data for each of the 13 study stands. The reported tree density is the point estimate as calculated by Program Distance; 95% confidence intervals are also provided for tree density. Forest type tree density was calculated by Program Distance using data from all stands.
Forest type Stand location Size (ha) Slope () Aspect Tree density (trees per ha)1 95% Confidence interval
Limber pine Beaver Ponds 62.2 11 194 96 63 145
Limber pine Rainbow Curve 7.4 24 152 169 93 306
Limber pine Estes Cone 89.1 16 129 654 318 1345
Limber pine Windy Gulch 398.9 20 187 530 163 1716
Limber pine Black Canyon 18.4 15 211 64 57 73
Forest type tree density: 361 176 737
Ponderosa pine North Lateral Moraine 404.7 15 201 120 57 255
Ponderosa pine Deer Ridge, south slope 167.8 14 167 128 43 384
Ponderosa pine Bighorn Mountain 154.43 20 167 203 46 891
Ponderosa pine Sheep Mountain 72.6 23 170 121 26 550
Forest type tree density: 133 101 175
Douglas-fir South Lateral Moraine 149.4 38 356 152 35 653
Douglas-fir Deer Ridge, north slope 45.5 23 335 182 89 368
Douglas-fir Eagle Cliff Mountain 104.14 23 320 1086 268 4401
Douglas-fir Cow Creek 29.96 17 17 712 501 1011
Forest type tree density: 503 266 952


Table 2. Landscape energetics for each conifer species within RMNP. For each conifer species, the minimum and maximum point estimate of cone density across all years are examined. Because limber pine cone productivity was relatively consistent during this study, the minimum and maximum point estimates are similar. In contrast, the maximum point estimates for ponderosa pine and Douglas-fir occurred during high cone crop years, in 2015 and 2016, respectively.
Conifer species Cones per ha Forest type area (ha) Seeds per cone Total seeds kJ per seed Total kJ
Limber pine 800 2,200 901 1.6 108 2.82 4.5 108
Limber pine 1,400 2.9 108 8.0 108
Ponderosa pine 700 3,400 503 1.2 108 l.l4 1.4 108
Ponderosa pine 9,600 1.7 109 1.9 109
Douglas-fir 300 4,300 505 6.8 107 0.25 1.6 107
Douglas-fir 92,900 1.8 1010 4.3 109
1 VanderWall and Baida 1977
2 Lanner 1996 (calories/gram) and Krugman and Jenkinson 1974 (grams/seed)
3 Value is averaged between point estimates of Curtis and Lynch 1965 and Schubert 1974
4 Vander Wall and Baida 1977 (calories/gram) and Krugman and Jenkinson 1974 (grams/seed)
5 Smith 1970
Table 3. Odds ratios calculated for a) 2014, b) 2015, and c) 2016. The numbers for each conifer species reflect the numbers of nutcrackers observed and not observed using seeds (foraging and/or caching). Odds ratios indicate the odds of nutcracker seed use of the alternative conifer seed source (top row) compared to the odds of nutcracker seed use of limber pine (bottom row).
a.
________2014 Foraging and caching counts_____
Conifer species No seed use Seed use Ponderosa pine 7 6
Limber pine____________8____________34
Odds ratio of alternative seed use = 0.20 95% Confidence interval = 0.05 0.77
39


b.
2015 Foraging and caching counts_____
Conifer species No seed use Seed use Ponderosa pine 16 59
Limber pine__________L7____________20
Odds ratio of alternative seed use = 3.13 95% Confidence interval = 1.34 7.34
______2016 Foraging and caching counts______
Conifer species No seed use Seed use Douglas-fir 10 19
Limber pine____________15_____________12
Odds ratio of alternative seed use = 2.38 95% Confidence interval = 0.81 6.98
a.
Idaho
Utah
Figures
Wyoming
Nebraska
Kansas
Oklahoma
Arizona
New Mexico
80 160 _l___i___I__l_
Texas
40


b.
Ponderosa Pine Douglas-fir Lodgepole Pine
I Limber Pine Spruce-Fir
Stand Locations
'
14 km
i. J
Figure 1. The map in a) shows the location of the study area, RMNP, in Colorado. The study stand locations in RMNP are shown in b), in addition to the distribution of forest types within RMNP.
41


a.
2014 Cone Density
8000- CD x: k. d) Q. tp 6000- o c o o


4UUU" '35 c a> o c 2000-O a
E i *
u Douglas-fir Limber pine Ponderosa pine Forest Type
b.
2015 Cone Density
-8000- CO -C k. a> Q. in 6000-(1) C o 5*4000- c a> T3 0> onnn
<


C xuuu O X n *
U
Douglas-fir Limber pine Ponderosa pine Forest Type
42


c.
2016 Cone Density
ar\r\nr\.
UUVJVJU -C < a.45000- (/> o c o o

' c a> D g15000- o O n + A

u Douglas-fir Limber pine Ponderosa pine Forest Type
Figure 2. Estimated cone densities for a) 2014, b) 2015, and c) 2016. Open circles indicate point estimates; errors bars indicate 95% confidence intervals. All estimates were calculated by program Distance (Thomas et al. 1998). The same scale is used for 2014 and 2015; the scale for 2016 is larger.
43


a.
Limber Pine Cone Count Regression
0) u +* 200 180
c o -o 160
> u 140
0) VI pfi o 120
c 100
o CJ 80
o 0) pfi 60
s 0 40
c "5 20
eS 0
0 20 40 60 80 100 120
Number of cones on tree observed from transect
Ponderosa Pine Cone Count Regression
44


c.
Douglas-fir Cone Count Regression
Number of cones on tree observed from transect
Figure 3. The relationship between the number of cones observed from the transect and the total number of cones observed on the tree for a) limber pine, b) ponderosa pine, and c) Douglas-fir. The slope of the regression line was used to estimate the total number of cones on each tree for each conifer species for Program Distance. The resulting cone density estimates were used for the estimation of landscape energetics of for each conifer species.
45


a,
Nutcracker visitation 2014
b.
Nutcracker visitation 2015
46


c.
Figure 4. Nutcracker visitation in a) 2014, b) 2015, and c) 2016. For each 10-day bin, the total number of surveys with detections is divided by the total number of surveys; both visual and auditory detections are included. Detections need to be within the stand being surveyed to be included. Arrows indicate the average time of cone ripening for each conifer species.
a.
2014 Nutcracker foraging timeline
47


b.
2015 Nutcracker foraging timeline
C.
2016 Nutcracker foraging timeline
Figure 5. Nutcracker foraging timeline in a) 2014, b) 2015, and c) 2016. Time observed foraging (in sec) is summed across all birds and divided by survey time (in min) within forest types for each 10-day bin; reported units are observed foraging sec per survey min. A different scale is used in 2015, compared to 2014 and 2016.
48


CHAPTER III
LIMBER PINE METAPOPULATION STRUCTURE AND CONNECTIVITY IN ROCKY MOUNTAIN NATIONAL PARK Introduction
Metapopulations are defined as large, regional populations consisting of smaller, local populations connected by migration (dispersal). The constituent local populations are susceptible to extinction, colonization, and recolonization (Hanski and Gilpin 1997, Hanski 1999). There are many metapopulation models that been devised and explored, some theoretical and some simulating natural metapopulation systems, with the goal of examining persistence overtime (Levins 1969, Levins 1970, Hanski 1999). Metapopulations are of increasing importance to managers as continuous populations become fragmented or reduced in size by habitat alteration or invasive pests and pathogens (Opdam 1991, Fahrig 2002). However, many natural populations exist as metapopulations, and these, too, are now vulnerable to disturbance.
Examples of natural metapopulations include the population structure for several butterfly species, where local populations often experience limited migration and relatively high extinction risk. For example, Granville fritillary (Melitaea cinxia) metapopulations in Finland have been extensively studied by Hanksi (e.g., Hanski 1994) and bay checkerspot butterfly (Euphydryas editha bayensis) metapopulations have been studied in California (Harrison et al. 1998). The population structure of mammals and other mobile vertebrates is often less suitable for this theory, but some species form metapopulations with lower extinction risk for local populations compared to insects. For example, the local populations of bighorn sheep (Ovis canadensis) metapopulations in western North America inhabit steep
49


canyons in mountains and deserts (Bleich et al. 1990, Singer et al. 2000). Small populations of the American pika (iOchotona princeps) inhabit scattered talus slopes throughout the high mountains of western North America (Moilanen et al. 1998, Beever et al. 2003). Metapopulation theory has also been applied to the management of spotted owls (Strix occidentalis), which inhabit isolated stands of old growth forest along coastal ranges of southern California (Lahaye et al. 1994, Gutierrez and Harrison 1996).
The application of metapopulation theory to plant species is less common. However, some authors suggest that plant species in general exist as metapopulations (Husband and Barrett 1996). This theory has been applied to the prickly lettuce (Lactuca serriola) in Europe, for example (Petrzelova and Lebeda 2008). However, the regional distributions of most forest tree species are less suitable for metapopulation theory, because the trees tend to occur either as large, continuous forest stands or may be one component of a widely-distributed forest habitat type. Limber pine (Pinus flexilis), a five-needle white pine (Subgenus Strobus) that ranges throughout the mountainous regions of the western U.S. and southwestern Canada, and occurs as regional metapopulations, may be one exception (Webster and Johnson 2000).
Limber pine is widely distributed both regionally and in elevationfrom lower to upper treeline (Steele 1990, Tomback et al. 2005). It is also a pioneer species, establishing in sites after a disturbance (Rebertus et al. 1991). Limber pine often forms serai stands in mesic sites where more competitive conifer species replace limber pine through succession, but may form climax stands in harsh areas with dry soil, intense radiation, and high wind velocity. Limber pine stands are relatively small, patchy, and isolated, forming regional metapopulations. Migration for limber pine stands is accomplished by the dispersal of
50


seeds, or propagules. Wind is the primary (moving seeds from tree to ground) seed-dispersal agent for most tree species, but limber pine seeds are dispersed long distances by the Clarks nutcracker (Nucifraga Columbiana) (Lanner and Vander Wall 1980, Tomback and Kramer 1980, Benkman et al. 1984). The nutcracker and limber pine are coadapted mutualists: nutcrackers transport the large, wingless seeds of limber pine and bury them in ground caches as winter and spring food sources. After snowmelt or summer rains, seeds within unretrieved caches may germinate and lead to forest regeneration (Tomback and Linhart 1990). Whereas nutcrackers forage on the seeds of several regional conifer species (Tomback 1998), limber pine primarily relies on nutcrackers for long-distance seed dispersal; rodents may disperse seeds over short distances (Tomback et al. 2005). Nutcrackers have been estimated to cache tens of thousands of pine seeds per bird each year (Vander Wall and Baida 1977, Tomback 1982). In addition, they may cache seeds near the foraging site or several kilometers away (Tomback 1978).
Webster and Johnson (2000) sampled 13 stands of a limber pine metapopulation within the Kananaskis Valley in Canada to determine whether limber pine exists as a classic metapopulationwhere local populations experience high extinction risk and limited seed dispersalor a patchy population metapopulationwith low extinction risk for local populations and extensive (non-limiting) seed dispersal. They determined that extinction events were infrequent; only two stands were consumed by a stand-replacing fire within the past century (there was no evidence for other causes of extinction). Stands were recolonized within approximately five years. Recolonization of stands likely resulted from nutcrackers dispersing seeds long distances from extant limber pine stands, because stands were isolated by at least a few kilometerstoo far for wind- or rodent-mediated seed dispersal. With
51


infrequent extinction events and immediate recolonization by nutcracker caching, Webster and Johnson (2000) concluded that limber pine existed as a patchy population metapopulation.
I examined the limber pine metapopulation in Rocky Mountain National Park (RMNP) with two goals: (1) determine limber pines metapopulation spatial configuration within RMNP; and, (2) examine how constituent populations within the park may be connected by nutcrackers, as determined by nutcracker flight and seed-dispersal distances. In other words, what is the average number of stands a nutcracker may potentially connect through seed-dispersal and caching flights? If the nutcracker population is reduced, how will this affect limber pine population connectivity within the RMNP metapopulation?
Methods
Study Area
I studied the limber pine metapopulation structure and Clarks nutcracker flight and seed-dispersal distances in RMNP (Fig. 1) while also investigating nutcracker foraging ecology (see Chapter 2 for details and map) from mid-June through late October in 2014, 2015, and 2016. Radio telemetry was conducted from June to October, 2015 and 2016. I determined that limber pine woodlands cover 2,212 ha of landscape within the park (see Chapter II). Additional habitat types within RMNP include ponderosa pine (Pinus ponderosa) parklands, Douglas-fir (Pseudotsuga menziesii) forests, lodgepole pine (.Pinus contorta) forests, spruce-fir (Picea engelmannii and Abies lasiocarpa) forests, alpine tundra, dry and moist meadows and grasslands, riparian corridors, and talus slopes. These habitats cover a wide elevation range within this high altitude park, 2380-4350 m.
52


Limber Pine Metapopulation Structure
To examine the spatial configuration of the RMNP limber pine metapopulation, I first obtained ArcGIS vegetation layers for RMNP from the National Park Service. Within the limber pine layer, I selected polygons whose borders were within 500 m of each other and identified each group of selected polygons as an individual limber pine stand; stands were thus separated from each other by at least 500 m. Each identified stand was exported as an individual layer. I only identified stands that were at least 1 ha in size and whose centroid was within the RMNP boundary. I obtained information on area (number of hectares), elevation (m), slope (), and aspect () from the attribute tables of these stands. I also calculated the distance (km) from each stand to every other stand, measured from the centroid of each stand.
Nutcracker Flight and Seed-Dispersal Distances
To trap nutcrackers for radio-tagging, I set up trapping stations concealed in rocky, forested sites near Rainbow Curve, 3300 m, from July through August in 2015 and 2016. Nutcrackers routinely fly to Rainbow Curve throughout the summer and fall to obtain tourist handouts (signs are posted prohibiting wildlife feeding). I baited the walk-in traps with peanuts, both with and without shells. I attached glue-on dorsal (A2480) or tailmount (A4560) model radiotransmitters (Advanced Telemetry Systems) to 10 nutcrackers in 2015 and 4 nutcrackers in 2016. (The use of backpack harnesses was not compatible with management policy.) I tracked tagged birds from August to October, 2015 and 2016, using a Yagi antenna and R410 scanning receiver (both from Advanced Telemetry Systems). I tracked each tagged bird at least once every two weeks. Upon finding a nutcracker, I recorded its position using a Garmin GPSMAP 62stc. The nutcracker was then followed for
53


two hours and I recorded a new point whenever the nutcracker flew a distance of at least 15 meters. I attempted to track two birds during each tracking day. If a bird was located in hazardous terrain, I used triangulation for point locations.
I uploaded the point locations for all birds to ArcGIS. For each bird that retained its tag long enough to record point locations more than one kilometer from the trapping site (n = 7 birds), the maximum flight distance was calculated as the maximum distance between point locations. If a tagged bird was observed to cache seeds (n = 1 bird), I calculated seed-dispersal distance as the distance between the foraging site and caching site.
Connectivity
With the computed distances between all limber pine stands (populations), I determined average distances for the nth nearest neighbor for each population. I compared this information to flight distances obtained from radio-tagged birds in order to estimate how many stands an individual nutcracker could connect. I also included nutcracker flight distances from the literature for this comparison.
Results
Limber Pine Metapopulation Structure in Rocky Mountain National Park
I identified 51 stands of limber pine within RMNP (Table 1, Fig. 1). These stands are primarily located east of the continental divide, with only 5 stands located on the west slope. Stands are most frequently located on or near ridges within the park, especially ridgelines with exposed granite. Consequently, stands are often associated with steep slopes, especially the drier south-facing aspects (Table 1). Limber pine is distributed across a wide elevation range in RMNP, from 2,702 to 3,447 m. The median stand size is 12 ha, but stand size is highly variable and ranges from 1.0 ha to 400.5 ha in size (Table 1). Size distribution appears
54


to vary across mountain ranges in RMNPfor example, limber pine occurs in several places and in small stands (ranging from 1.0 to 31.5 ha) in the northern Mummy Range, and the Longs Peak massif contains mostly larger stands that range from 113.3 to 292 ha in size. Nutcracker Flight and Seed-Dispersal Distances
I attached radio transmitters to 10 birds in 2015 and to four birds in 2016; one 2015 bird that had lost its tag was recaptured and a second tag applied. After transmitter attachment, tagged nutcrackers at first were present either at Rainbow Curve or at a nearby location in the subalpine forest, within a kilometer of this turnout. Seven birds either removed their tags or their transmitter signal was lost before point locations were ever recorded outside of this local range. The maximum flight distances recorded from the remaining six birds ranged from 1.8 to 12 km, with a median of 3.8 km (Table 2). The shortest distance was recorded from a bird searching a northwest-facing slope of Hidden Valley, possibly for ripe limber pine seeds. The maximum distance was recorded from a bird on Prospect Mountain. The remaining birds were detected either along Deer Ridge, the Beaver Ponds moraine, or the southern stretch of the Mummy Range, and may have been foraging on ripening ponderosa pine or Douglas-fir seeds (see Chapter 2). There is likely a strong bias towards short flights in the reported maximum flight distances for these nutcrackers due to the limited duration of transmitter attachment.
The only nutcracker to retain its tag for the duration of a field season was also the only bird observed to disperse seeds. After being tagged, this bird stayed near Rainbow Curve; at the end of September, it flew down to the south-facing slope of Deer Ridge to forage on ripening ponderosa pine seeds. For the remainder of the field season, it was observed to forage within an area of about 2.5 ha (the foraging site), along with one to five
55


other nutcrackers. It briefly cached seeds within this area on some occasions, flying short distances (see Chapter 2), but once every 35 to 50 min it would fly north or northwest with a full sublingual pouch, indicating that it was also caching in another area (the caching site) (e.g., Vander Wall and Baida 1977, Tomback 1978, Lanner and Vander Wall 1980). I noted its direction and scanned from Horseshoe Park, where the signal directed me to the northfacing slope of the Beaver Ponds moraine. I identified this as the caching site, which was about 2.2 km from the foraging sitemy only observed long-distance seed-dispersal distance (Table 2). I should note that while on the moraine, I detected that the bird was close but never observed it. The signal also appeared to move further northwest on some occasions, indicating the bird may have cached further away, and the seed-dispersal distance is underestimated. The bird was observed to roost on this moraine as well.
Additional nutcracker flight distances in RMNP are reported in Tomback and Taylor (1987), with a maximum of 14.5 km. Seed-dispersal distances are rarely reported for limber pine seeds in the literature. Vander Wall and Baida (1977) observed nutcrackers to disperse limber pine seeds 4 to 5 km and Vander Wall (1988) observed them to disperse limber pine seeds approximately 1 km. The furthest observed nutcracker seed-dispersal distances (regardless of conifer species) include 12.5 km (Tomback 1978), 22 km (Vander Wall and Baida 1977), and 32.6 km (Lorenz and Sullivan 2011).
Inter-Population Connectivity
Fig. 2 displays the distribution of nearest distances for limber pine stands in RMNP. For any given limber pine stand, the nearest stand is 1.6 km away on average and the furthest stand is 32.0 km away on average. Based on the seed-dispersal distance of 2.2 km observed here, a nutcracker could potentially connect nearest stands on average. Based on the
56


maximum observed flight distance of 12 km, a nutcracker could potentially connect approximately 39% of the limber pine metapopulation in RMNP. Based on the maximum seed-dispersal distance in the literature, 32.6 km, a nutcracker could potentially connect the entire metapopulation if it were highly mobile in its foraging and caching behavior.
Discussion
The spatial configuration of the RMNP limber pine metapopulation has implications for how the constituent populations could function over time. The east side of RMNP appears more suitable for the limber pine metapopulation, because it contains more than 90% of the stands in the park. It is possible that the western slope is generally more mesic or experienced less disturbance in the past. The isolation and small size of the five western stands may lead to lower connectivity relative to eastern stands; nutcrackers may visit western stands less frequently and recolonization following extinction could take longer than it might on the east slope.
Stand size is highly variable and larger stands may contribute more to seed dispersal within the metapopulation. The largest stand, at 400.5 ha and located on Trail Ridge, may be so large that loss of all trees is unlikely, except in the case of a severe, stand-replacing fire. In the case of a portion of trees within the stand experiencing mortality, for example from mountain pine beetle or lightning-caused ground fire, these openings could quickly be regenerated by seeds cached by nutcrackers from nearby limber pine trees. I have observed nutcrackers to frequently visit this stand, and they may cache seeds harvested here in other limber pine stands located elsewhere in the park. If these assumptions are true, this stand would act as a core or source population within the metapopulation by reliably providing
57


seeds for recruitment within extant stands and recolonization within extinct stands. The four
larger stands located on the Longs Peak massif (totaling 718 ha) may act in a similar manner.
I experienced challenges in tagging nutcrackers leading to limited data for nutcracker flight and seed-dispersal distances. I am confident that 9 of 10 birds removed tags prior to the end of the field season in 2015. In 2016 I applied more epoxy and tagged two birds in August to ensure that I was late in the molt cycle, so that feathers along the spinal tract would not molt shortly after tag attachment. However, I am confident that all birds in 2016 also removed their tags. Attaching radio transmitters with epoxy directly to bird skin and plumage is not workable for nutcrackers, because they persist at tag removal. Tag removal was confirmed by recapturing two previously tagged birds and finding the dropped tag via radio telemetry for seven birds. I cannot confirm tag removal for all birds in 2015 and 2016, because the transmitters from these other nutcrackers either: (1) were no longer detected in the park (n = 2 birds), likely a result of the removed tag falling into a crevice or other structure that blocked the signal, although the birds may have left the park with tags still attached; or (2) remained in one location inaccessible to me until the end of the field season (n = 3 birds)likely a result of tag removal, given the observed mobility of these birds.
Based on my limited data and data from the literature on flight distances, one nutcracker has the potential to connect a large percentage of the metapopulation in RMNP through seed transportation and caching; even a small population of nutcrackers could potentially maintain the metapopulation structure over time. This is assuming that longdistance seed-dispersal distances observed elsewhere are routine for nutcrackers in RMNP.
Unfortunately, limber pine is presently facing more threats than just wildfire. A combination of warmer temperatures, drought, and the abundance of lodgepole pine has
58


resulted in severe mountain pine beetle outbreaks (Dendroctonusponderosae; a native species), impacting limber pine stands in RMNP and throughout the Rocky Mountains (Romme et al. 1986, Lynch et al. 2006, Gibson et al. 2008). In addition, fire suppression has resulted in less limber pine across its range by facilitating the succession of more competitive species (Gruell 1983, Rebertus et al. 1991). The most serious threat is the invasion of the nonnative fungal pathogen Cronartium ribicola, which causes the disease white pine blister rust in five-needle white pines (McDonald and Hoff 2001). Blister rust has killed 90% of limber pines in some stands in Wyoming and Montana (Kearns and Jacobi 2007, Tomback and Achuff 2010, Schwandt et al. 2010). Blister rust was recently detected in northern Colorado and RMNP, with minimal to no mortality at this early stage (Johnson and Jacobi 2000, Schoettle et al. 2011). It is expected that blister rust will cause extensive loss of limber pine across the landscape in RMNP in the coming decades. In addition, climate change is likely to complicate some or all of these threats, including more severe fire and major outbreaks of mountain pine beetle (Tomback and Achuff 2010, Tomback et al. 2011).
Nutcrackers have been shown to visit stands with low cone production with a lower probability (McKinney and Tomback 2007, McKinney et al. 2009, Barringer et al. 2012). With a future likelihood of declining limber pine across RMNP, it is possible that the associated drop in cone productivity could result in reduced visitation and seed dispersal by nutcrackers, or even the periodic absence of nutcrackers in RMNP. Without nutcrackers, the limber pine metapopulations connectivity would be severely impacted. Seed dispersal by seed drop and rodents (e.g., Tomback et al. 2005) would completely isolate most, if not all, stands from each other. However, nutcrackers may still visit stands, but to a smaller extent, even with increased limber pine mortality. I have observed nutcrackers to forage on the seeds
59


of isolated limber pine trees within stands of other forest types, indicating that the relationship between cone productivity and nutcracker visitation may operate on a finer scale than previously recognized. Resident nutcrackers may learn their environment and include isolated trees in their foraging excursions. Seed caching by nutcrackers in RMNP would be more unpredictable across time and space in this scenario, but would likely maintain the metapopulation over time given the high connecting potential of an individual nutcracker.
Tables
Table 1. Limber pine local population characteristics (n=51)
Area (ha) Elevation (m) Aspect Slope ()
Median 12.3 3194 178.4 22
Min 1.0 2702 NA 12
Max 400.5 3447 NA 42
Table 2. Clarks nutcracker use of space in a) RMNP and b) literature
a.
Clarks Nutcracker use of space in RMNP (km)
Seed-dispersal distance Maximum flight
(n = 1 bird) distance (n = 7 birds)
Median 2.2 3.8
Min NA 1.8
Max NA 12
60


b.
Clarks Nutcracker use of space in literature (km)
Maximum seed-dispersal distance Maximum flight distance
Limber pine (Vander Wall and Baida 1977) 4-5 km NA
Liber pine (Vander Wall 1988) ~1 km NA
RMNP (Tomback and Taylor 1987) NA 14.5 km
Pinyon pine (Vander Wall and Baida 1977) 22 km NA
Whitebark pine (Tomback 1978) 12.5 km NA
Whitebark pine (Lorenz et al. 2011) 32.6 km NA
61


Figures
Figure 1. Map of individual limber pine stands in RMNP. Area of circle is equal to area of stand. The nearby towns of Estes Park and Grand Lake are also shown, in addition to US Highways 34 and 36.
62


Figure 2. Distribution of nearest distances of limber pine local populations.
63


REFERENCES
Barringer, L. E.; Tomback, D. F.; Wunder, M. B.; McKinney, S. T. 2012. Whitebark pine stand condition, tree abundance, and cone production as predictors of visitation by ClarksNutcracker. PLoS ONE. 7(5): e37663. doi: 10.1371/joumal.pone.0037663
Benkman, C. W.; Baida, R. P.; Smith, C. C. 1984. Adaptations for seed dispersal and the compromises due to seed predation in limber pine. Ecology. 65: 632-642.
Bent, A. C. 1946. Life histories of North American jays, crows, and titmice. Part 2. U.S. Nat. Mus. Bull. 191.
Buckland, S. T.; Anderson, D. R.; Burnham, K. P.; Laake, J. L.; Borchers, D. L.; Thomas, L. 2001. Introduction to distance sampling: estimating abundance of biological populations. Oxford University Press, New York.
Campbell, R. W.; Dawe, N. K.; McTaggart-Cowan, I.; Cooper, J. M.; Kaiser, G. W.; et al. 1997. The birds of British Columbia. Vol. 3. Passerines: flycatchers through vireos.
R. Br. Columbia Mus., Victoria.
Critchfield, W. B. 1986. Hybridization and classification of the white pines (Pinus section Strobus). Taxon. 35: 647-56.
Critchfield, W. B.; Little, Jr., E. L. 1966. Geographic distribution of the pines of the world. Miscellaneous Publication 991. Washington, DC: U.S. Department of Agriculture Forest Service. 97 p.
Coop, J. D.; Schoettle, A. W. 2009. Regeneration of Rocky Mountain bristlecone pine (Pinus aristata) and limber pine (Pinus flexilis) three decades after stand-replacing fire. Forest Ecology and Management. 257: 893-903.
Cottam, C. 1945. Feeding habits of the Clarks Nutcracker. The Condor. 47: 168.
Davis, J.; Williams, L. 1957. Irruptions of the Clarks Nutcracker in California. The Condor. 59: 297-307.
Davis, J.; Williams, L. 1964. The 1961 irruption of the Clarks Nutcracker in California. Wilson Bulletin. 76: 10-18.
Decker, F. R.; Bowles, J. H. 1931. Summer birds of the Blue Mountains, Washington. Murrelet. 12: 12-14.
Dixon, J. B. 1956. Clarks Nutcrackers preying on ground squirrels and chipmunks. The Condor. 58: 386.
Geils, B. W.; Hummer, K. E.; Hunt, R. S. 2010. White pines, Ribes, and blister rust: a review and synthesis. Forest Pathology. 40: 147-185.
Gernandt, D. S.; Geada Lopez, G. G.; Ortiz Garcia, S.; Liston, A. 2005. Phylogeny and classification of Pinus. Taxon. 54(1): 29-42.
64


Gibson, K.; Skov, K.; Kegley, S.; Jorgensen, C.; Smith, S.; Witcosky, J. 2008. Mountain pine beetle impacts in high elevation five needle pines: Current trends and challenges. Forest Health Protection Report Rl-08-020. U.S. Department of Agriculture, Forest Service. 32 p.
Gibson, K.; Kegley, S.; Bentz, B. 2009. Mountain pine beetle. Forest Insect and Disease Leaflet 2. Portland, OR: U.S. Department of Agriculture, Forest Service, Pacific Northwest Region (R6): 12 p.
Giuntoli, M.; Mewaldt, L. R. 1978. Stomach contents of Clarks Nutcracker collected in western Montana. The Auk. 95: 595-598.
Gruell, G. E. 1983. Fire and vegetative trends in the Northern Rockies: Interpretations from 1871-1882 photographs. USDA Forest Service General Technical Reporter, INT-252. Ogden, Utah: Intermountain Forest and Range Experiment Station. 117 p.
Hanski, T 1994. A practical model of metapopulation dynamics. Journal of Animal Ecology. 63(1): 151-162.
Hanski, I. 1999. Metapopulation ecology. Oxford University Press, New York.
Husband, B. C.; Barrett, S. C. H. 1996. A metapopulation perspective in plant population biology. Journal of Ecology. 84(3): 461-469.
IPCC. 2001. Climate change 2001: Impacts, adaptation and vulnerability. In: McCarthy, J. J.; Canziani, O. F.; Leary, N. A.; Dokken, D. J.; White. K. S.; eds. Contribution of working group II to the third assessment report of the IPCC. Cambridge, UK: Cambridge University Press. 1000 pp.
Johnson, D. W.; Jacobi, W. R. 2000. First report of white pine blister rust in Colorado. Plant Disease. 84: 595.
Kearns, H. S. J.; Jacobi, W. R. 2007. The distribution and incidence of white pine blister rust in central and southeastern Wyoming and northern Colorado. Canadian Journal of Forest Research. 37: 462-472.
Krugman, S. L.; Jenkinson, J. L. 1974. Pinus L. Pine. In: Schopmeyer, S. C., tech, coord.
Seeds of woody plants in the United States. Agricultural Handbook 450. Washington, DC: USDA Forest Service. 598-638.
Lanner, R. M. 1980. Avian seed dispersal as a factor in the ecology and evolution of limber and whitebark pines. In Proceedings of Sixth North American Forest Biology Workshop, pp. 15-48. University of Alberta, Edmonton, Alberta.
Lanner, R. M. 1982. Adaptations of whitebark pine for seed dispersal by Clarks Nutcracker. Canadian Journal of Forest Research. 12: 391-402.
65


Lanner, R. M.; Vander Wall, S. B. 1980. Dispersal of limber pine seed by Clarks Nutcracker. Journal of Forestry. 78 (October): 637-639.
Liston, A.; Robinson, W. A.; Pinero, D.; Alvarez-Buylla, E. R. 1999. Phylogenetics of Finns (Pinaceae) based on nuclear ribosomal DNA internal transcribed spacer region sequences. Molecular Phylogenetics and Evolution. 11(1): 95-109.
Logan, J. A.; Powell, J. A. 2001. Ghost forests, global warming, and the mountain pine beetle (Coleoptera: Scolytidae). American Entomologist. 47(3): 160-172.
Logan, J. A.; Regniere, J.; Powell, J. A. 2003. Assessing the impacts of global warming on forest pest dynamics. Frontiers in Ecology and the Environment. 1(3): 130-137.
Lorenz, T. J.; Sullivan, K. A. 2009. Seasonal differences in space use by Clarks Nutcrackers in the Cascade Range. The Condor. 11(2): 326-340.
Lorenz, T. J.; Sullivan, K. A.; Bakian, A. V.; Aubry, C. A. 2011. Cache-site selection in Clarks Nutcracker (Nucifraga Columbiana). The Auk. 128(2): 237-247.
Lynch, H. J.; Renkin, R. A.; Crabtree, R. L.; Moorcraft, P. R. 2006. The influence of previous mountain pine beetle (Dendroctonusponderosae) activity on the 1988 Yellowstone fires. Ecosystems. 9: 1318-1327.
MacKenzie, D. I.; Nichols, J. D.; Lachman, G. B.; Droege, S.; Royle, J. A.; Langtimm, C. A. 2002. Estimating site occupancy rates when detection probabilities are less than one. Ecology. 83(8): 2248-2255.
McCaughey, W. W.; Tomback, D. F. 2001. The natural regeneration process. In: Tomback,
D. F.; Amo, S. F.; Keane, R. E.; eds. Whitebark pine communities: ecology and restoration. Washington, DC: Island Press. 105-120.
McDonald, G. T; Hoff, R. J. 2001. Blister rust: an introduced plague. In: Tomback, D. F.;
Arno, S. F.; Keane, R. E.; eds. Whitebark pine communities: ecology and restoration. Washington, DC: Island Press. 263-284.
McKinney, S. T.; Tomback, D. F. 2007. The influence of white pine blister rust on seed dispersal in whitebark pine. Canadian Journal of Forest Research. 37: 1044-1057.
McKinney, S. T.; Fiedler, C. E.; Tomback, D. F. 2009. Invasive pathogen threatens bird-pine mutualism: implications for sustaining a high-elevation ecosystem. Ecological Applications. 19(3): 597-607.
Mewaldt, L. R. 1956. Nesting behavior of the Clarks Nutcracker. The Condor. 58: 3-23.
Moilanen, A.; Hanksi, I.; Smith, A. T. 1998. Long-term dynamics in a metapopulation of the American pika. The American Naturalist. 152(4): 530-542.
Murray, M. P.; Tomback, D. F. 2010. Clarks Nutcrackers harvest sugar pine seeds from cones. Western North American Naturalist. 70(3): 413-414.
66


Perkins, D. L.; Swetnam, T. W. 1996. A dendroecological assessment of whitebark pine in the Sawtooth-Salmon River region Idaho. Canadian Journal of Forest Research. 26: 2123-2133.
Petrzelova, I.; Lebeda, A. 2008. Distribution of race-specific resistance against Bremia lactucae in natural populations of Lactuca serriolci. European Journal of Plant Pathology. 129(2): 233-253.
Raffa, K. F.; Aukema, B. H.; Bentz, B. J.; Carroll, A. L.; Hi eke, J. A.; Turner, M. G.;
Romme, W. H. 2008. Cross-scale drivers of natural disturbances prone to anthropogenic amplification: the dynamics of bark beetle eruptions. BioScience.
58(6): 501-517.
Rebertus, A. J.; Bums, B. R.; Veblen, T. T. 1991. Stand dynamics of Pinus /Ux/V/.s-dominated subalpine forests in the Colorado Front Range. Journal of Vegetation Science. 2: 445-458.
Romme, W. H.; Knight, D. H.; Yavitt, J. B. 1986. Mountain pine beetle outbreaks in the
Rocky Mountains: regulators of primary productivity? American Naturalist. 127(4): 484-494.
Samano, S.; Tomback, D. F. 2003. Cone opening phenology, seed dispersal, and seed predation in southwestern white pine {Pinus strobiformis) in southern Colorado. Ecoscience 10:319-326.
Schoettle, A. W.; Rochelle, S. G. 2000. Morphological variation of Pinus flexilis (Pinaceae), a bird-dispersed pine, across a range of elevations. American Journal of botany. 87: 1797-1806.
Schoettle, A. W.; Goodrich, B. A.; Klutsch, J. G.; Burns, K. S.; Costello, S.; Sniezko, R. A.; Connor, J. 2011. The proactive strategy for sustaining five-needle pine populations: An example of its implementation in the southern Rocky Mountains. In: Keane, R. E.; Tomback, D. F.; Murray, M. P.; Smith, C. M., eds. 2011. The future of high-elevation, five-needle white pines in Western North America: Proceedings of the High Five Symposium. 2010 28-30 June, Missoula, MT. Proceedings RMRS-P-63. Fort Collins, CO. U.S. Department of Agriculture, Forest Service, Rocky Mountain Research Station: 376. http://www.fs.fed.us/rm/pubs/rmrs p063.html
Schwandt, J. W.; Lockman, I. B.; Kliejunas, J. T.; Muir, J. A. 2010. Current health issues and management strategies for white pines in the western United States and Canada.
Forest Pathology. 40: 226-250.
Singer, F. J.; Bleich, V. C.; Gudorf, M. A. Restoration of bighorn sheep metapopulations in and near western national parks. Restoration Ecology. 8(4S): 14-24.
Syring, J.; Farrell, K.; Businsky, R.; Cronn, R.; Liston, A. 2007. Widespread genealogical
nonmonophyly in species of Pinus subgenus Strobus. Systematic Biology. 56(2): 163-181.
67


Thomas, L.; Laake, J. L.; Derry, J. F.; Buckland, S. T.; Borchers, D. L.; Anderson, D. R.; Burnham, K. P.; Strindberg, S.; Hedley, S. L.; Marques, F. F. C.; Pollard, J. FL; Fewster, R. M. 1998. Distance 3.5. Research Unit for Wildlife Population Assessment, University of St. Andrews, St. Andrews, UK.
Tomback, D. F. 1978. Foraging strategies of Clarks Nutcracker. Living Bird. 16: 123-161.
Tomback, D. F. 1983. Nutcrackers and pines: coevolution or coadaptation? Coevolution. Univ. of Chicago Press, Chicago, Illinois.
Tomback, D. F. 1998. Clarks nutcracker (Nucifraga Columbiana). In: A. Poole, ed. The
Birds of North America Online, Species Account Number 331. Cornell Laboratory of Ornithology, Ithaca, NY.
Tomback, D. F. 2001. Clarks nutcracker: agents of regeneration. In: Tomback, D. F.; Arno, S. F.; Keane, R. E. eds. Whitebark Pine Communities: Ecology and Restoration. Washington, DC: Island Press. 89-104.
Tomback, D. F.; Kramer, K. A. 1980. Limber pine seed harvest by Clarks Nutcracker in the Sierra Nevada: Timing and foraging behavior. The Condor. 82: 467-468.
Tomback, D. F.; Taylor, C. L. 1987. Tourist impact on Clarks Nutcracker foraging activities in Rocky Mountain National Park. In: Singer, F. J., ed. Toward the year 2000-Conference on science in the National Parks; Vol. 2 Wildlife management and habitats. Fort Collins, CO: George Wright Society and the U.S. National Park Service: 158-172.
Tomback, D. F.; Linhart, Y. B. 1990. The evolution of bird-dispersed pines. Evolutionary Ecology. 4: 185-219.
Tomback, D. F.; Kendall, K. C. 2001. Biodiversity losses: The downward spiral. In: Tomback, D. F.; Amo, S. F.; Keane, R. E. eds. Whitebark Pine Communities:
Ecology and Restoration. Washington, DC: Island Press. 243-262.
Tomback, D. F.; Achuff, P. 2010. Blister rust and western forest biodiversity: Ecology, values and outlook for white pines. Forest Pathology. 40: 186-225.
Tomback, D. F.; Schoettle, A. S.; Chevalier, K. E.; Jones, C. A. 2005. Life on the edge for limber pine seed dispersal within a peripheral population. Ecoscience. 12: 519-529.
Tomback, D. F.; Achuff, P.; Schoettle, A. W.; Schwandt, J. W.; Mastrogiuseppe, R. J. 2011. The magnificent high-elevation five-needle white pines: Ecological roles and future outlook. In: Keane, R. E.; Tomback, D. F.; Murray, M. P.; Smith, C. M., eds. 2011. The future of high-elevation, five-needle white pines in Western North America: Proceedings of the High Five Symposium. 2010 28-30 June, Missoula, MT. Proceedings RMRS P-63. Fort Collins, CO. U.S. Department of Agriculture, Forest Service, Rocky Mountain Research Station: 376. http://www.fs.fed.us/rm/pubs/rmrs p06 3.html
68


Torick, L. L.; Tomback, D. F.; Espinoza, R. 1996. Occurrence of multi-genet tree clusters in wind-dispersed pines. American Midland Naturalist. 136(2): 262.
Turner, T. J.; Tomback, D. F.; Van Anderson, B.; Murray, M. 2011. Sugar pine seed harvest by Clarks nutcracker: Annual use of a transient resource in Crater Lake National Park, Oregon. In: Keane, Robert E.; Tomback, Diana F.; Murray, Michael P.; Smith, Cyndi M., eds. 2011. The future of high-elevation, five-needle white pines in Western North America: Proceedings of the High Five Symposium. 2010 28-30 June, Missoula, MT. Proceedings RMRS P-63. Fort Collins, CO. U.S. Department of Agriculture, Forest Service, Rocky Mountain Research Station: 376. http://www.fs.fed.us/rm/pubs/rmrs p06 3.html
Vander Wall, S. B. 1988. Foraging of Clarks Nutcracker on rapidly changing pine seed resources. The Condor. 90: 621-631.
Vander Wall, S. B.; Baida, R. B. 1977. Coadaptations of the Clarks Nutcracker and the pinyon pine for efficient seed harvest and dispersal. Ecol. Monogr. 47: 89-111.
Vander Wall, S. B.; Hutchins, H. E. 1983. Dependence of Clarks Nutcracker, Nucifrciga
columbiana, on conifer seeds during postfledging period. Canadian Field-Naturalist. 97: 208-214.
Vander Wall, S. B.; Hoffman, S. W.; Potts, W. K. 1981. Emigration behavior of Clarks Nutcracker. The Condor. 83: 162-170.
Veblen, T. T. 1986. Age and size structure of subalpine forests in the Colorado Front Range. Bulletin of the Torrey Botanical Club. 113: 225-240.
Webster, K. L.; Johnson, E. A. 2000. The importance of regional dynamics in local populations of limber pine (Pimis flexilis). Ecoscience. 7(2): 175-182.
69


Full Text

PAGE 1

AND LIMBER PINE METAPOPULATION STRUCTURE IN ROCKY MOUNTAIN NATIONAL PARK by TYLER JUSTIN WILLIAMS B.S., Arkansas Tech University, 2012 A thesis submitted to the Faculty of the Graduate School of the University of Colorado in partial fulfillment of the requirements for the degree of Master of Science Biology Program 2017

PAGE 2

ii This thesis for the Master of Science degree by Tyler Justin Williams has been approved for the Biology Program by Diana F. Tombac k, C hair Michael Wunder Michael Greene Date: May 13, 2017

PAGE 3

iii Williams, Tyler Justin (MS, Biology Program ) Structure in Rocky Mountain National Park Thesis directed by Professor Diana F. Tomback Abstract utcracker ( Nucifraga columbiana ) is the major seed disperser for limber pine ( Pinus flexilis ), which in turn is a preferred conifer food source for the nutcrackers. Limber pine exhibits a metapopulation structure: regional populations consist of small, local populations connected by seed dispersal. Current and predicted threats to limber pine include mountain pine beetle ( Dendroctonus ponderosae ) outbreaks, larger wildfires, and especially exotic white pine blister rust (pathogen Cronartium ribicola ). Extensive limber pine mortality may force nutcrackers to rely on alternative conifer seed sources. In Rocky M ountain National Park (RMNP) I investigated the importance of alternative seed sources connectivity might be influenced by nutcracker spatial use. From mid June to late October 2014 to 2016, data were collected on: 1) cone production in stands of limber pine, ponderosa pine ( Pinus ponderosa ), and Douglas fir ( Pseudotsuga menziesii ); 2) timing of nutcracker stand visitation; and, 3) nutcracker se ed harvest and caching behavior; 4) limber pine geographic occurrence within RMNP using G IS layers; and 5) radio tracked locations for nutcrackers in 2015 2016 Limber pine component populations (n = 51) ranged from 1 400 ha in size with inter population distances of 1 36 km. Nutcrackers traveled distances r anging from 1 12 km (n = 7 nutcrackers) and cache d seeds 2.2 km away from the foraging site (n = 1

PAGE 4

iv nutcracker) indicating potentially high limber pine metapopulation connectivity, even with a smaller nutcracker population. Annual variation in cone produ ction influenced nutcracker foraging preferences and the use of different seed source s Each year starting in mid to late August, nutcrackers foraged on limber pine seeds, which ripen in early September. In 2014 and 2015, nutcrackers transitioned to harves ting ponderosa pine seeds, which ripen in early October. In 2016, instead they transitioned to Douglas fir seeds, which ripen in late September. With potential limber pine losses, I believe that alternative seed sources will support a nutcracker population although carrying capacity may be lower. I suggest that ponderosa pine will serve as an increasingly critical food resource if limber pine declines. The form and content of this abstract are approved. I recommend its publication. Approved: Diana F. Tomback

PAGE 5

v Acknowledgments I would like to thank my advisor Diana F. Tomback for all of her help direction, and advice along the way. I would also like to thank Michael Wunder and Michael Greene for being on my committee and providing insight on project organization, data collection, and data analysis. I appreciate the work done by Kristin Broms in writing the RStudio code for occupancy analyses I would also like to extend my gratitude to the Rocky Mountain Conservancy for awarding me the Bailey Research Fellowship in 2015 and the Denver Field Ornithologists for awarding me the Research, Education, and Conservation Grant in 2015 and 2016. These were instrumental in obtaining radio telemetry equipment and an additional Garmin GPSMAP 62stc in addition to a ssisting with gas costs for myself and volunteers I would like to thank R ocky M ountain N ational P ark staff for their help with logistics and permitting, especially Paul McLaughlin and Scott Esser. I would like to acknowledge the University of Colorado Denver IACUC for ensuring that this research wa s conducted humanely (protocol # 88314(06)1c). I am also very grateful for all of the field assistance offered by the 16 volunteers who helped me collect data. Furthermore, I would like to thank the other graduate students within the Department of Integrative Biology for helpful discussions and workshopping. Finally, I am deeply grateful for all of the support offered by Amber Williams and other loved ones throughout this program.

PAGE 6

v i Table of Contents CHAPTER LITERATURE REVIEW 1 1 Background: Limber ..5 I. THE IMPORTANCE OF MULTIPLE SEED SOURCES: CLARK NUTCRACKER SEED USE IN ROCKY MOUNTAIN NATIONAL PARK 10 Introduction 10 Methods 15 Study Area ... 15 Field Methods .. .1 6 Estimation of Tree Density, Cones/Tree, Cone Statistical .21 Results .24 Tree and Cone Density Cone Infestation, Landscape Energetics, ..

PAGE 7

vii .. Discussion .38 Figures II. LIMBER PINE METAPOPULA TION STRUCTURE AND 49 49 Methods 52 ... .52 .53 Nutcracker Flight and Seed .53 54 54 Limber Pine Metapopulation Structure in Rocky Mountain National Park ..54 Nutcracker Flight and Seed 55 Inter 56 .57 62

PAGE 8

1 CHAPTER I LITERATURE REVIEW Background: Limber Pine Limber pine ( Pinus flexilis ) is classified in class Pinopsida, order Pinales, family Pinaceae, genus Pinus section Strobus and in subsection Strobi This classification is being challenged by some researchers (Liston et al. 1999, Gernandt et al. 2005, Syring et al. 2007), who suggest that the subsections Strobi and Cembrae (which includes whitebark pine, Pinus albicaulis and other stone pines) should be merged to form a new subsection Strobus (Gernandt et al. 2005). They also recommend merging this new subsection with the subsections Gerardianae and Krempfianae to form a new section Quinquefoliae (Gernandt et al. 2005). Limber pine, which occurs in patchy and scattered stands, has a comparativ ely broad distribution in western North America. The distribution runs from southern Alberta and British Columbia south to northern New Mexico and Arizona; and west from the Sierra Nevada and coastal ranges of California to its easternmost range in the Bla ck Hills, South Dakota (Critchfield and Little 1966). Limber pine also has a broader altitudinal distribution than other white pines, occurring in Colorado from elevations ranging from 1600 m in grassland steppe habitat to 3400 m in the Rocky Mountain tree line where krummholz trees exist (Schoettle and Rochelle 2000). Limber pine is a pioneer species in that it is one of the first populations to be established at an open site, such as that caused by a burn (Lanner and Vander Wall 1980, Rebertus et al. 1991 ). At productive sites, limber pine regeneration may occur over the first ~30 years, but this species will be outcompeted by shade tolerant trees such as spruce and fir (Rebertus et al. 1991, Coop and Schoettle 2009). These shade tolerant trees are actuall y able

PAGE 9

2 to become established at these sites as a result of limber pine mitigating harsh conditions (Rebertus et al. 1991). In more xeric habitats, limber pine is replaced by shade tolerant trees at a slower rate and the process of regeneration may continue for 100 years (Veblen 1986, Rebertus et al. 1991). In extremely harsh environments, such as in the upper subalpine, limber pine will form climax communities (Rebertus et al. 1991). The reason that climax stands of limber pine are restricted to extremely h arsh sites is that limber pine is a poor competitor (Veblen 1986), which also accounts for some aspect of metapopulation dynamics (Webster and Johnson 2000 and Veblen 1986). In the Rocky Mountains, limber pine sometimes occurs in dwarfed form (krummholz) a t treeline, as both solitary trees and within tree islands (Resler and Tomback 2008). ( Nucifraga columbiana Family Corvidae order Passeriformes ) limber pine has several adaptations for bird dispersal (Tom back 1982, Tomback 1983). In contrast to the small, winged pine seeds (e.g., ponderosa pine ( Pinus ponderosa ) producing on average 37 mg seeds and lodgepole pine ( Pinus contorta ) 4 mg seeds) of wind dispersed pines, which are considered ancestral and chara cterize the majority of species in Pinaceae, limber pine seeds are large (93 mg) and wingless (Lanner 1980, Critchfield 1986, Tomback and Linhart 1990). Nutcrackers prefer these seeds because they are more efficiently harvested than winged seeds (Tomback 1 978, Tomback and Linhart 1990). In addition, the branches are lyrate (upswept) and the cones are horizontally oriented in whorls at branch ends. The orientation of the branches and cones allow for the nutcrackers to assess how productive a tree is and to p erch on cones and branch tips for efficient seed harves ting ( Lanner 1 980, 1982 Tomback and Linhart 1990 ). In

PAGE 10

3 which means that they may become dislodged and drop to the ground. Limber pine populations are declining in the Rocky Mountains (Tomback and Achuff 2010, Schwandt et al. 2010). The most serious threat is the disease white pine blister rust. Cronartium ribicola the fungal pathogen that causes the disease, is native to Asia and was accidentally introduced to the Pacific Northwest in the early 1900s (McDonald and Hoff 2001, Geils et al. 2010). Since then, it has spread widely throughout the Western U.S. and into Canada (Schwandt et al. 2010, Tomback and Achuf f 2010). Basidiospores from diseased Ribes leaves initiate infection by entering the stomata of pine needles during humid conditions in late summer (McDonald and Hoff 2001, Geils et al. 2010). From the stomata, hyphae eventually spread to the phloe m of branches and stems. After two to three years, cankers develop from which spores are released and eventually girdle and kill the branch or stem. If infection occurs in the canopy, the death of branches will result in fewer cones for reproduction and less fo liage for photosynthesis. If infection spreads to the bole (trunk), the tree will eventually die (McDonald and Hoff 2001, Geils et al. 2010). White pine blister rust may result in extensive canopy damage, with mortality as high as 100% in some limber pine stands (Kearns and Jacobi 2007). In 1998, the first infected limber pine in Colorado was detected (J ohnson and Jacobi 2000). In 2010 blister rust was detected in limber pine in RMNP (Schoettle et al. 2011). Mountain pine beetles are another major threat that is reducing limber pine populations in many regions (Gibson et al. 2008). They are native to the West and coevolved with the widespread hosts ponderosa and lodgepole pine, but may attack any pine species. The female will enter the bark of a pine, lay her eggs, and deposit fungal spores within the

PAGE 11

4 tree. The larvae feed on the phloem and sapwood while the fungus spreads within the tree, resulting in mortality via nutrient transport disruption (Gibson et al. 2009). Mountain pine beetle outbreaks periodica lly occur in western forests, in response to higher temperatures and drought cycles, sometimes lasting for 10 years or more (Romme et al. 1986, Perkins and Swetnam 1996, Lynch et al. 2006). During severe outbreaks, the pine beetles may invade the stands of higher elevation white pine species such as limber pine During the late 1990s, an outbreak began that appears to be on the downswing. This has resulted in unprecedented mountain pine beetle kill, and has been attributed to mature even aged stands of lod gepole and ponderosa pine as well as recent higher temperatures and drought (Gibson et al. 2008), which in turn has been attributed to climate change (Logan and Powell 2001, Logan et al. 2003, Raffa et al. 2008). Fire suppression has also had a negative i mpact on limber pine, as a result of fire exclusion programs beginning in the 1920s (Arno and Allison Bunnell 2002, Taylor and Carroll 2004). These programs aimed to eliminate forest fires and were successful in reducing fire frequency. However, it was not until late in the 20 th century that studies were conducted that demonstrated the negative effects of fire suppression (Brown et al. 1994, Van lower elevations as a re sult of fire suppression in some areas, reduced fire frequency also led to increased community succession (Gruell 1983), which decreased the number of early successional limber pine populations and also decreased landscape diversity. Climate change may a ffect limber pine through means other than mountain pine beetle outbreaks. Climate change is also expected to cause plant populations to shift in distribution. Since the populations of different plant species may shift differently or at

PAGE 12

5 different rates, th is may result in new plant communities with unknown consequences (IPCC 2001 and references therein). Limber pine is predicted to shift northward and to higher elevations while the southern populations and lower elevation populations may die out (Tomback an d Achuff 2010). Issues may arise in the shifting of limber pine as a result of metapopulation structure, harsh conditions above treeline, and complications from mountain pine beetle mortality, fire suppression, and white pine blister rust infection (Tombac k et al. as the low elevation and southern ranges are receding, this will result in an overall decrease in limber pine numbers. The late summer and fall, and retrieves seeds until the next cone crop is available (Tomback 1978). Since the seeds in these caches may germinate if not retrieved, this has result ed in some species of pine coevolving with the nutcracker and becoming adapted for bird dispersal (Tomback and Linhart 1990). In the U.S., these species include whitebark pine, limber pine, southwestern white pine ( Pinus strobiformis ), Colorado pinyon pine ( Pinus edulis ), and single leaf pinyon pine ( Pinus monophylla ). Because of the large, wingless seeds of these bird dispersed pines, nutcrackers prefer to harvest them over the seeds of other conifers and serve as major seed dispersers for these species (T omback 1978, Lanner and Vander Wall 1980, Tomback and Linhart 1990). Nutcr ackers are highly adapted to forage on conifer seeds, but they still retain the ability to eat foods opportunistic ally, which is characteristic of corvids. They will eat arthropods carrion, eggs, nestlings, and small mammals (Decker and Bowles 1931, Cottam

PAGE 13

6 1945, Bent 1946, Dixon 1956, Giuntoli and Mewaldt 1978). It is not known how much they depend on these alternative food sources and how this changes seasonally and from year to y ear. An additional food source that they utilize in certain areas is handouts from tourists (Tomback and Taylor 1987). In RMNP, nutcrackers are abundant at scenic turnouts where tourists commonly stop and feed animals, despite rules against this. Some nutc rackers become regulars at these turnouts, even bringing their young. There is also some evidence that the young brought to turnouts will become regulars themselves (Tomback and Taylor 1987). Multiple issues potentially arise from this: if too many nutcrac kers become regulars and do not cache seeds as a result of their dependence on handouts, bird dispersal of limber pine seeds (and thus regeneration) may be diminished. On the other hand, this abundance of food may affect juvenile survival or cause a signif icant rise in the nutcracker population size through increased immigration. This increase may cause more unripe seeds to be harvested, leaving fewer ripe seeds to be cached. eed production. Nutcrackers will begin harvesting unripe seeds from closed cones in late summer and cache ripe seeds from late August to December (Tomback 1998). Which conifer seeds the nutcracker harvests and caches will differ both across time and region due to different conifer communities, yearly cone crop variation, nutcracker preferences, and various ripening phenologies. In the case of limber pine, nutcrackers will begin harvesting unripe seeds in late July and cache seeds from late August or early S eptember until late October or early November (Vander Wall 1988). Nutcrackers show general caching behaviors regardless of the seed source. These are summarized from Tomback (1978) and Vander Wall (1988). It is indicated that nutcrackers

PAGE 14

7 have been observ ed to cache seeds in both divergent and convergent storage areas (terminology from Tomback 1978). When caching in divergent storage areas, the nutcrackers are alone or occasionally in pairs. They may cache seeds in the ground under leaf litter, near the ba se of a tree, rock, or log, in gravel or volcanic pumice substrate, in trees under bark or in a hole, and in disturbed areas such as burns. When caching seeds in the substrate, they typically swipe their bill along the ground to make a trench and bury 1 15 seeds (average of 3) before covering the trench with substrate and sometimes with a cone or pebble. Convergent areas are communal storage areas, with local populations of nutcrackers potentially caching seeds in the same terrain. These are typically locat ed on steep, south facing slopes near stands of pines. These caches provide nutcrackers with food during times when little other food is available. During the winter and early spring, nutcrackers will retrieve seeds from caches (although some seeds in co nes are still available during winter). Nutcrackers will feed their young seeds from caches. They nest relatively early, with eggs often hatching in March or April (Bent 1946, Mewaldt 1956, Campbell et al. 1997). Early nesting in this species seems counter intuitive given the harsh environments they inhabit. It is thought that this is an adaptation which allows juveniles to reach independence by the time seed caching begins (Vander Wall and Balda 1977, Tomback 1978). Once the young have fledged, they will fo llow their parents and watch them harvest, cache, and retrieve seeds (Tomback 1978, Vander Wall and Hutchins 1983). By watching their parents, they will be prepared to make caches themselves in the fall. ty resulted in two hypotheses regarding their spatiotemporal use of space. One hypothesis was that the nutcrackers would

PAGE 15

8 migrate altitudinally within a region to follow the production of cone crops of different pines (Tomback 1978). The other hypothesis pr oposed that nutcrackers either behaved as emigrants or residents (Vander Wall et al. 1981). Under this hypothesis, emigrants migrated latitudinally in search of productive cone crops while residents did not migrate, but did move altitudinally in search of cone crops, and remained within their home range year round (except during excessively poor cone crop years). Results from radio telemetry conducted in the Cascade Mountains, Washington, supported the second hypothesis (Lorenz and Sullivan 2009). Nutcrackers occupied a home range for winter, spring, and summer (year round home range). However, nutcrackers expande d their range in the fall to search for cones. More research needs to be conducted on this question of spatial use because nutcrackers may behave differently across their wide range as a result of different climates and conifer communities. Since nutcrac kers forage on multiple conifer species, a poor cone crop in one of their major conifer seed sources can be compensated by seed production in other conifers (e.g., Tomback 1978, Giuntoli and Mewaldt 1978). However, it has also been shown for whitebark pine that nutcrackers are not as likely to visit stands with low cone production as they are to visit stands with greater cone production (McKinney and Tomback 2007, McKinney et al. 2009, Barringer et al. 2012). For example, the probability of observing nutcra ckers when no cones are present is 0.22 0.35 (Barringer et al. 2012). The issue is that in stands with high mortality from blister rust and mountain pine beetle, not many cones will be produced even with a good cone crop. If these stands are not visited, l ittle caching occurs, and regeneration by nutcrackers will be less likely. In addition, if mortality is widespread, it is unknown how

PAGE 16

9 this will affect the range and movement patterns of nutcrackers. The nutcrackers may emigrate more frequently, but there i s speculation that they may leave the area entirely. Nutcracker visitation is important not only for natural regeneration, but also for local white pine restoration projects. As a result of nutcrackers being either a primary or major seed disperser for s everal white pines, the Proactive Restoration Strategy in the southern Rocky Mountains depends on natural seed dispersal to spread potentially rust resistant genotypes (Schoettle et al. 2011). The purpose of this strategy is to protect white pines in advan ce of serious damage from the threats previously discussed. Providing disturbed areas for nutcracker caching by thinning or placing controlled burns near stands with trees that have known genetic resistance to blister rust is one strategy (Schoettle et al. 2011).

PAGE 17

10 CHAPTER II SEED USE IN ROCKY MOUNTAIN NATIONAL PARK Introduction utcrackers ( Nucifraga columbiana ) are a widely ranging keystone species of western coniferous forests whose ecosystem services include long distance seed dispersal, local tree establishment after disturbance and forest regeneration over time for several western pines (Tomback 1998, Tomback 2001, Tomback and Kendall 2001) Across their range, nutcrackers u tilize these and additional conifer spec ies as a food source. S everal conifer species that are not strongly dependent on nutcrackers for seed dispersal may be more important to nutcracker ecology and the stability of population numbers than previously reco gnized. Studies often emphasize that nutcrackers use pines with large, wingless seed s as mainstay or preferred food resources For example Tomback (1978) describes nutcrackers using whitebark pine ( Pinus albicaulis ) seeds in California and Benkman et al. (1984) and Vander Wall (1988) describe nutcrackers using limber pine ( Pinus flexilis ) seeds in Arizona and Utah, respectively. In late summer and fall, nutcrackers forage on the seeds of these two pine species and transport them within their sublingual po uch to cache them nearby or tens of kilometers away as a winter and spring food source ( Vander Wall and Balda 1977, Tomback 1978, Lorenz and Sullivan 2009). These two conifer species are mutualists of nutcrackers, relying on the birds for seed dispersal (T omback and Linhart 1990). Nutcrackers enable small isolated stands of limber pine to function together as a metapopulation by connecting them through seed dispersal flights (Webster and Johnson 2000, Williams Chapter III ). Nutcrackers also forage on and disperse the large, wingless seeds of southwestern white pine

PAGE 18

11 ( Pinus strobiformis ; Benkman et al. 1984, Samano and Tomback 2003), Colorado pinyon pine ( Pinus edulis ; Vander Wall and Balda 1977), and single leaf pinyon pine ( Pinus monophylla ; Vander Wall 1988). Each of these co nifer species is an important regional food source In addition, some authors describe nutcracker use of conifer species with smaller, winged seeds in different parts of the nutcracker range (Tomback et al. 2011). Tomback (1978) rep orts the harvest and caching of Jeffery pine ( Pinus jeffreyi ) seed following whitebark pine seed use in the eastern Sierra Nevada. Giuntoli and Mewaldt (1978) report nutcracker stomach contents including ponderosa pine ( Pinus ponderosa ) and Douglas fir ( Ps eudotsuga menziesii ) seeds in addition to whitebark pine seeds in western Montana; Vander Wall et al. (1981) note nutcracker harvest of Douglas fir seeds in Utah; Lorenz et al. (2009) reports use of ponderosa pine and Douglas fir seeds in the Cascade Range Washington; and, Schaming (2016) observed nutcrackers to forage on Douglas fir seeds in Wyoming. Nutcrackers are also known to forage on the small, winged, seeds of bristlecone pines ( Pinus longaeva and Pinus aristata ; Lanner 1988, Torick et al. 1996). I n addition to these smaller seeded conifer species, nutcrackers predictably forage on the larger, winged seeds of sugar pine ( Pinus lambertiana ; Murray and Tomback 2010, Turner et al. 2011). While some of these smaller, winged conifer species have limited distributions and use of these seed sources is thus geographically restricted, I propose that several widely distributed seed sources are ecologically crucial to provide sufficient seed p roduction to sustain nutcracker populations over time Conifer speci es produce variable cone crops each year (Krugman and Jenkinson 1974, McCaughey and Tomback 2001), and nutcrackers are known to emigrate when little

PAGE 19

12 food is available (Davis and Williams 1957, Davis and Williams 1961, Vander Wall et al. 1981). During years of low cone production by large, wingless seeded pine s, other seed sources enable nutcrackers to store sufficient food for winter and early spring. These conifer species may also help to stabilize the mutualisms between nutcrackers and five needle white p ines (Tomback and Linhart 1990). While some mutualisms appear specific, they may persist over time only because of the presence other conifer species. Alternative seed sources may be especially important as populations of whitebark and limber pine (as well as southwestern white, sugar, and bristlecone pines needle white pines subgenus Strobus ) decline because of multiple anthropogenic stressors (Tomback and Achuff 2010, Tomback et al. 2011). Severe mountain pine beetle ( Dendroctonus ponderosae ) outbreaks, attributed to recent high temperatures and drought, spread into higher elevation forests and impact pine species other than their historical host lodgepole pine ( Pinus contorta ) ( Gibson et al. 2008, Logan et al. 2010). Fire suppression has also resulted in advanced succession in many forest communities, leading to declines in five needle white pine populations (Gruell 1983, Rebertus et al. 1991, Murray et al. 2000). Climate change is expected to shift five needle white pine populations north and to higher elevations while southern and lower elevation populations decline; ranges may shrink if colonization is slow (Tomback and Ach uff 2010). The ongoing spread of white pine blister rust (caused by the non native pathogen Cronartium ribicola ), the mos t serious threat, is killing five needle white pines and causing up to 90% mortality in some stands (Kearns and Jacobi 2007, Tomback and Achuff 2010). Some conifer species whose seeds are harvested by nutcrackers exhibit comparatively small ranges, includ ing Jeffrey pine, foxtail pine, the bristlecone pines, and sugar pine (Little

PAGE 20

13 1971); these species therefore provide only regional or localized benefits for nutcracker populations. However, ponderosa p ine and Douglas fir exhibit comparatively expansive g eographic ranges and locally large forest expanses, potentially serving as extensive food sources for nutcrackers. These conifers are the only two seed sources to occur in all 11 western states, and there is much overlap in range with other pines, with pon derosa pine extending more into the dry southwest and Douglas fir ranging north into Canada (Little 1971). In northern Colorado, limber pine is the only widely distributed large seeded pine; it is a preferred food source for nutcrackers in this region (e.g ., Tomback and Taylor 1987, Torick et al. 1996). In addition to large, wingless seeds, limber pine tree morphology, like whitebark pine, features upswept branches with horizontally oriented cones in whorls at branch tips that allow for efficient assessment and foraging by nutcrackers (Lanner 1980). Nutcrackers begin foraging on green, unripe limber pine cones during mid to late summer, using their long, sturdy bill to pry open cone scales (Vander Wall and Balda 1977, Tomback and Taylor 1987). They make late summer and autumn caches in the soil using side sweeping motions of the bill to create a trench, and then insert three to four seeds individually each a few centimeters deep (Lanner and Vander Wall 1980, Vander Wall 1988). Nutcrackers are prolific seed d ispersers, estimated to cache 16,000 limber pine seeds each year per bird (Vander Wall 1988). Seeds within unretrieved nutcracker caches may germinate after spring rains or snowmelt and lead to forest regeneration. Whereas nutcrackers may rely on limber p ine seeds in northern Colorado, ponderosa pine and Douglas fir are also widely available in this region and may provide supplemental seed sources. However, the extent to which nutcrackers rely on these species in this region is

PAGE 21

14 unknown. White pine blister rust was first detected in this region in 1998 (Johnson and Jacobi 2000), with an infected tree in Rocky Mountain National Park (RMNP) in 2010 (Schoettle et al. 2011). Blister rust infecting tree canopies reduces cone production and photosynthetic biomass, weakening trees; blister rust infection in tree stems kills trees (McDonald and Hoff 2001, Geils et al. 2010). Estimates suggest that up to 50% of limber pine habitat in northern Colorado is susceptible to blister rust (Kearns 2005, Howell et al. 2006). A limber pine restoration plan is currently being implemented in RMNP to minimize mortality from pine beetles and blister rust, involving insecticide and verbenone application, seed collections, screening for blister rust resistance, and developing artifici al regeneration guidelines (Schoettle et al. 2011). With increasing limber pine mortality in the coming decades as a result of blister rust, however, nutcracker populations will come to rely more on alternative seed sources i f nutcracker populations remain in northern Colorado. I examined nutcracker use of conifer seeds across three field seasons in RMNP, a relatively undisturbed natural area that features diverse mountain topography and representative northern Colorado forest communities. My goal was to de termine the relative importance of limber pine, ponderosa pine, and Douglas fir as food sources and as habitat to nutcrackers in RMNP. To accomplish this goal, my objectives were to: 1) estimate annual cone production of the limber pine, ponderosa pine, an d Douglas fir forest types, and the total energy each conifer species provided across the landscape in RMNP; 2) estimate inter and intra annual patterns of nutcracke r visitation of each forest type ; and 3) examine the relative numbers and timing of nutcra ckers foraging on and caching the seeds of each conifer species and the odds of nutcrackers using seeds of different conifer species. Historically, RMNP has supported a robust population of nutcrackers (Tomback and Taylor 1987). By

PAGE 22

15 understanding how nutcra ckers utilize alternative seed sources in a system not yet impacted by blister rust, we can understand the implications of increased limber pine mortality. Fewer limber pine trees will reduce carrying capacity in the nutcracker population; complete loss of nutcrackers, a keystone species, is expected to have ecological impacts on many other pines that benefit from their seed dispersal services (Tomback and Kendall 2001, Tomback et al. 2011). In addition, because ponderosa pine and Douglas fir are also the m ost widespread alternative seed sources across the nutcracker range, the results of this study could apply broadly. Methods Study Area I studied nutcrackers on the eastern slope of RMNP from June to October, 2014 to 2016 (Fig. 1) Park elevations range from 2,380 to 4, 350 m; study stand elevations ranged from 2,400 to 3 400 m. As determined from ArcGIS layers obtained from the National Park Service, w ithin RMNP there are communities of ponderosa pine (3,396 ha), Douglas fir (4,327 ha), and limber pine ( 2,212 ha) the forest types of interest. Because RMNP is a high elevation park, it only includes the western extent of ponderosa pine and Douglas fir communities on the northeastern front; these forest types continue outside park boundaries further east at lower altitudes For example, 12,000 ha of ponderosa pine woodlands and 11,000 ha of Douglas fir forests are located within approximately 10 km east of the RMNP boundary in the Roosevelt National Forest. Since nutcrackers do not recognize park boundaries, some may have entered and exited RMNP during the study period while using these forests. Other habitat within th e park includes lodgepole pine forests, spruce fir ( Picea engelmannii and Abies lasiocarpa ) forests, park meadows, riparian zones, alpine tundra and

PAGE 23

16 talus slopes. Average minimum January temperature in nearby Estes Park is 9 C and average maximum Jul y temperature is 26 C. Estes Park receives 13.10 inches of precipitation per year on average The wide elevation range within RMNP results in highly variable climatic conditions Field Methods Stand Selection I selected stands of ponderosa pine, Douglas fir, and limber pine for cone density estimation, nutcracker presence/absence surveys, and focal behavior surveys Here, a stand is defined as a continuous forested area where one conifer species of interest dominates. I located stands by scouting RMNP and examining ArcGIS vegetation layers obtained from the National Park S ervice. All three forest types are primarily located on the eastern slope of RMNP, and I methodically selected stands that span this distribution. Stand selection was for presence/absence surveys and focal behavior sampling (not possible with the Ute Trail and Estes Cone stands). Stands also needed to be large enough, approximately five hectares in area, to fit a transect. Once a stand was identified that met these criteria, I selected it as a study stand o nly after visiting it to ensure that it was accessible and the correct forest type. In 2014, I selected two stands of Douglas fir, three stands of ponderosa pine, and five stands of limber pine In 2015 and 2016, evenly distributing survey time, I opted fo r four stands of each species by removing one limber pine stand and adding stands of ponderosa pine and Douglas fir. I established a virtual transect using GPS points within each stand for all three survey types. Transects were a maximum of 1000 m in leng th when possible (min = 300 m, mean =

PAGE 24

17 817 m) and consisted of points separated by 100 m. Transects consisted of either one continuous line (n = 10 stands), two lines (n = 2 stands), or three lines (n = 1 stand). I placed t r ansects within the stand randomly using ArcGIS when possible, but if constrained I placed them along a n accessible ridge, slope, or trail. Constraints on transect length, connectedness, and random placement resulted from small stand area or hazardous terrain Estimation o f Tree D ensity, Cones/T ree, Co ne Density and Cone Insect Infestation I surveyed each stand to e stimate cone density for each conifer species once per year between June 13 th and September 13 th During this time period unripe cones are easy to distinguish from o ld cones and large enough to observe for all three conifer species I used distance s ampling to model cone detection probability ( Buckland et al. 2001 ). While walking a transect, I recorded the following information from each observed tree: distance to tre e from observer using a laser rangefinder (Nikon ProStaff 550) direction of tree fro m observer using a compass (Silva Ranger) and number of con es observed from transect using 10x42 mm binoculars. Information was recorded only for trees within 50 m because observing cones beyond this distance is difficult without a scope To ensure that all recorded trees were large enough to bear cones, I only included trees that were at least two meters in height Because the observer cannot leave the transect during dista nce sampling, I only counted cones on one side of each tree (a partial count) even if the tree was near the transect The disadvantage here is that Program Distance will not estimate the number of cones present on the unobserved sides of trees, resulting in low estimates of cone density for each forest type. Because I am interested in only comparing the relative cone production between conifer species, this disadvantage do es not affect my interpretation of Program Distance results.

PAGE 25

18 In 2014, I only sampled between 100 to 1000 m of each transect (median = 500 m) ; I walked transects more slowly in stands with higher tree density B ase d on Program Distance analyses of 2014 data I determined that sampling 100 trees per transect would be sufficient for good model fit (based on a chi sq uare test; described below), cone density estimates without high standard error and collecting data for one transect within one field day To obta in 100 trees in 2015 I sampled only 50 to 350 m of each transect (median = 200 m) randomly selected the starting point of each survey within a transect when possible (n = 7 stands) I also conducted a full cone count for every fifth tree (in addition to a partial count for that tree) by circling the tree and counting cones not observed from the transect. This allowed me to estimate the proportion of cones missed on observed trees while on the transect. The full cone counts were only utili zed in the approximate estimation of landscape energetics (described below). I used the same methodology in 2016. Infestation of ponderosa and limber pin e cones by insects, identified by small light brown or reddish brown cones filled with frass, occurred during my study. These cones do not provide any food energy to nutcrackers because the seeds are inedible. I was unable to identify the species of insects causing infestation, but they were likely either ponderosa pine coneworms ( Dioryctria auranticella ) or lodgepole pine cone beetles ( Conophthorus contortae ) (Schoettle and Negron 2001) To estimate the extent of infestation, I sampled 20 ponderosa pine trees in October 2015 and 20 limber pine trees in September 2016 during presence/absence surveys five tr ees from each stand, selecting the nearest cone bearing tree from five points per transect. I circled each tree and counted the number of healthy and infested cones to estimate the percentage of infested cones for each tree and for each conifer

PAGE 26

19 species as a whole. These estimates provide further details for the amount of food energy available to nutcrackers. Nutcracker Presence/A bsence Surveys I surveyed stands to examine trends in nutcracker presence in forest types over time. I walked t ransects from August 11 th to October 31 st each year. At the beginning of these surveys cones in all three forest types were green and unripe, and they ripened throughout the field season; nutcrackers may harvest seeds from ripening or ripe seed sources (Tomback 1978, T omback and Taylor 1987) I conducted from two to seven surveys per transect in 2014, with 20 total surveys in limber pine stands, 10 surveys in ponderosa pine stands, and six surveys in Douglas fir stands. I conducted from four to seven surveys per transec t in 2015 and 2016, with 20 total surveys within each forest type each year I attempted to distribute surveys for each transect across the field season so that different stages of cone ripening were included in the surveys. In 2015 and 2016, I attempted t o visit each transect once every other week. At each point along t he transect, I listened and watched for nutcrackers for 10 min and recorded whether I detected nutcrackers within the stand. I walked t wo transects per day when possible; the number of points included in each survey depended on the amount of time that could be devoted to the survey and any environmental hazards, including extreme weather, mo ose presence, and rutting elk (minimum percentage of points incl uded in a survey = 18%; mean = 88%) Focal Behavior Surveys I surveyed nutcracker foraging and caching behavior each year to estimate trends over time and to compare the odds of nutcracker seed use for each conifer species F ocal behavior

PAGE 27

20 surveys were conducted during presence/absence surveys (n = 156 surveys), commencing as I entered the stand, and following a cone count if conducted after August 10 th ( n = 8 surveys). During a behavior survey, I approached a detected nutcracker if within ~100 m. I hal ted presenc e/absence surveys during this time Seed foraging and caching were the behaviors of interest, including the seeds of limber pine, ponderosa pine, and Douglas fir. Onc e the bird was within view, I recorded each behavior and the time allocated to each behavior. There were three instances in 2016 where I observed a bird either holding a cone or flying with a cone for only 5 to 10 sec; I included this as foraging behavior. I recorded all other behaviors, including calling, perching, preening, bill wi associated conifer species. I the bird flew out of view or after 10 min of behavior data were collected. I f other birds were present, I began recording data on the nex t visible bird after data recording had ceased on the previous bird A behavior survey ceased when I collected a total of 60 min of behavior data or I exited the stand. Survey time was used to normalize the foraging time of nutcrackers, and was calculated as the duration between commencement and cessation of a behavior survey I did not record stand entrance or exit times (used to calculate survey time) in 2014 and for some of the first surveys in 2015, and so I estimated th em based on 2015 and 2016 data of the duration between the times of stand entrance and conducting the first point (for entrance time) and the duration between the times of conducting the last point and exiting the stand (for exit time)

PAGE 28

21 Statistical Analyses Tree and Cone Density and Landscape Energetics To prepare distance s ampling data for analysis, I each tree by calculating the difference between the tree direction and transect direction I then used the distance from observer (r) for ea ch tree to calculate perpendicular distance from transect (x) by using the equation: (1) I used Program Distance (Thomas et al. 1998) for all Distance Sampling analyses. This program estimates the proportion of objects not de tected, which allows for a more accurate estimate of density To calculate object density, Program Distance plots a histogram of the number of detected objects against x. A detection function, g(x), is then plotted along the bin peaks in the histogram. The equation of g(x) may b e defined by one of four key functions: uniform, half normal, hazard rate, and negative exponential. Each key func tion may be further defined by three series expansions: cosine, simple polynomial, and hermite polynomial. I selected t he key function and series expansion based on the lowest AIC values. However, I rejected the negative exponential key function because it was not appropriate for my data sets (I believe that none of the forest types should result in a spiked distribution on the histogram) ; the remaining key functions generally estimated similar densities with the 95% confidence intervals overlapping nearly completely The observed number of objects (n) is divided by transect length (L) and the integral of g( x) to obtain the point estimate of true density. I calculated tree density for each stand by analyzing each stand individually and including transect data from all years. I also calculated tree density for each forest type by

PAGE 29

22 including data from all stand s of a forest type across all years, stratifying by stand, and using a weighted average based on stand area. For each year, I calculated cone density for each forest type by combining data from all stands of a forest type within that year, stratifying by stand, and using a weighted average based on stand area. Trees were only included in the cone density analysis if they contained at least one cone. I examined the histogram and left truncated the data if bins near the transect contained relatively few obse rvations and right truncated the data if outliers were present I evaluated the fit of the models, g(x), with a chi square goodness of fit test, which compares the expected histogram values of the model to the observed histogram values Low X 2 values indic ate that the models are significantly different than the observed data and are a poor fit. Using cone density data, I estimated the approximate amount of energy on the landscape for each conifer species. I obtained estimates of full tree cone counts for each conifer species for use in landscape energetics as follows: I plotted partial counts (from transects) against full counts for each conifer species, combining 2015 and 2016 data. I used linear regressi on to plot a line of best fit, with a y intercept of 0; otherwise, the y intercept can result in a significant inflation of the density estimate. For each conifer species, I multiplied the slope of the line by the number of cones observed from the transect for every tree to obtain a total cone estimate for each tree. I then analyzed these data using the same Program Distance protocol described above. To estimate landscape energetics, I examined the literature for data on seeds per cone and kJ per seed (eith er directly or I calculated it from calories per gram and grams per seed) for each conifer species. I then multiplied cones per hectare by area of distribution in RMNP (in hectares) and by seeds per cone to obtain the

PAGE 30

23 total number of seeds present in RMNP for each conifer species. I then multiplied that estimate by kJ per seed to obtain the total kJ estimate in RMNP for each conifer species. Nutcracker Presence and Foraging Patterns Each year, I pooled presence/absence data into 10 day bins to estimate ha bitat use changes over time, beginning with August 11 th and ending October 31 st Within each bin, I calculated percentage of surveys with detections for each forest type by summing the number of surveys with detections and dividing by the total number of surveys. I plotted the results on a time series graph with 10 day bins along the x axis and the perce ntage of surveys with detections along the y axis. Because these graphs do not account for differences in detection probability between forest types, I focus on how visitation changes over time within each forest type making the assumption that detection p robability stays relatively consistent over time. I analyzed seed use data with graphic analysis and an odds ratio test. For the time series graphs, each year I pooled foraging observations into 10 day bins, beginning with August 11 th and ending with Octob er 31 st Within each bin, I summed foraging time across all birds for each conifer species, summed the survey time for each conifer species, and divided foraging time by survey time. I plotted the results on a time series graph with 10 day bins along the x axis and foraging sec per survey min along the y axis in order to examine the time sequence for nutcracker foraging on different conifer species, how many days they forage on different conifer species, and the timing of transition between conifer species. The Odds Ratio test compares the odds of nutcrackers using the seeds (foraging or caching) of a given conifer species to the odds of using the seeds of another conifer species ( Rita and Komonen 2008 ). Each year I summed the number of birds observed using seeds

PAGE 31

24 and not using seeds for each conifer species. C onifer species was not assigned by the forest type being surveyed but by the conifer species selected by the nutcracker. For example, if a bird foraged on an isolated limber p ine tree within a ponderosa pine stand, the assigned conifer species would be limber pine. I used the forest type being surveyed unless the bird was part of a foraging flock; here I used the conifer species being utilized by the fl ock. Birds within a foraging flock were always observed to f orage on the same conifer species, with one exception in 2016 when one bird briefly foraged on Douglas fir seeds while the other birds foraged on limber pine seeds. This was within a Douglas fir s tand, and for birds the associated conifer species would be Conifer species were only included in the analysis if nutcrackers used the seeds of that coni fer species within a given year Birds that removed seeds of an unknown conifer species from caches (one bird in 2016) were excluded. Birds that were caching seeds of an unknown conifer species (three birds in 2016) were excluded I included all birds regardless of duration of seed use (5 to 600 sec). I used RStudio (version 3.24, R core team 2016) for the Odds Ratio test. Results Tree and Cone D ensity Cone Infestation, Landscape Energetics, and Ripening Phenology The geographic data and estimated tree densities of each stand are report ed in Table 1. Point estimates for limber pine tree density within stands ranged from 64 to 654 trees per ha (Table 1). Point estimates for ponderosa pine tree density within stands ranged from 120 to 203 trees per ha. Point estimates for Douglas fir tree density within stands ranged from 152 to 1086 trees per ha. Multiple conifer species were present within each stand, especially

PAGE 32

25 Engelmann spruce and subalpine fir within limber pine stands and lodgepole pine within limber pine and Douglas fir stands; ponde rosa pine woodlands often lacked tree diversity. The reported tree density only applies to the focus seed source within each stand. I observed a unique combination of cone crops among all conifer species each year, producing multiple scenarios of nutcrack er foraging behavior. Cone crops were relatively small for all conifer species in 2014, ranging from 180 to 48 0 cones per ha (Fig. 2). On average, for all conifer species, average cone productivity ranged from 0.4 to 5 cones per tree. In 2015, limber pine and Douglas fir again produced small cone crops, ranging from 200 to 810 cones per ha (Fig. 2), wit h average cone productivity of 0.4 to 2 cones per tree. However, ponderosa pine cone density was much higher in 2015 an esti mated 6,83 0 cones per ha (Fig. 2), wi th an average productivity of 51 cones per tree. In 2016, cone density for limber an d ponderosa pine ranged from 510 to 850 cones per ha (Fig. 2), wit h average cone productivity of 2 to 4 cones per tree. However, Douglas fir cone density was an estim ated 48,83 0 cones per ha (Fig. 2), with an average cone productivity of 97 cones per tree. It is noteworthy that limber pine experienced a comparatively small cone crop in all three years while both ponderosa pine and Douglas fir each experienced one large cone crop year: one in 2015 and one in 2016. Cone infestation by insects was estimated to be extensive. Within the 20 ponderosa pine trees I examined for cone infestation in 2015, 53% of the cones showed infestation by cone insects. Cone infestation of individual ponderosa pine trees ranged from 8% to 97%. Within the 20 limber pine trees I examined in 2016, 79% of the cones showed infestation b y cone insects. Cone infestation by insects of individual limber pine trees ranged from 0% to 100%. I did not include a portion of these infested limber pine cones in the distance sampling

PAGE 33

26 cone counts (for cone density estimates), because some infestation could be detected by June or July. For the cones included in distance sampling cone counts (and therefore program Distance), I estimate th at 57% of limber pine cones were infested. It is possible that an unknown small percentage of infested cones were from previous years. Cone infestation occurred in all three years of the study for these two conifer species. Using the results of regressions based on full cone counts (Figure 3), distance sampling results and other seed weight and energy values from the li terature, I estimated the energy available across the landscape in RMNP for each conifer species (Table 2 ). These estimates of landscape energetics are approximate and rough; this is an exploratory exercise and inferences are limited. Point estimates of limber pine cone densit ies across years ranged from 800 to 1,400 cones per ha. Using values reported in Table 2 I estimated the energy across the RMNP landscape provided by limber pine seeds duri ng this study to range from 4.1 10 8 to 7.3 10 8 kJ. Ponderosa p point e stimates of cone density ranged from 700 to 9,600 cones per ha. I estimated the energy across the RMNP landscape provided by ponderosa pine seeds duri ng this study to range from 1.4 10 8 kJ to 1.9 10 9 kJ. Point est imates of Douglas fir cone dens ities ranged from 300 to 92,900 cones per ha. I estimated the energy across the RMNP landscape provided by Douglas fir seeds duri ng this study to range from 1.6 10 7 kJ to 4.3 10 9 kJ. Each conifer species studied followed a unique ripening phenology; ripening entails the transition from unripe cones with closed scales, to scales slightly sepa rating from the cone core. Unripe limber pine cones are green; I observed them to begin ripening from late August to early September. Unripe Douglas fir cones appear green or purple and began ripening from mid to late September. Unripe ponderosa pine cones are green and began

PAGE 34

27 ripening from late September to early October. During ripening for all three conifer species, the cones transiti oned to a brown color and cone scales separated to reveal seeds. Nutcracker Stand Visitation Nutcrackers were observed in all forest types in all three years (Fig. 4 ). Changes in nutcracker detections were often coincident with cone ripening for each of th e conifer species. In 2014, I detected nutcrackers more frequently in limber and ponderosa pine stands when cones began ripening for each species. Relatively few surveys were conducted in 2014, which may have resulted in more variable estimates of detectio n over time. In 2015, I again detected nutcrackers more frequently in limber and ponderosa pine stands after cones ripened; this increase was especially ephemeral in limber pine stands. I also detected nutcrackers more frequently in Douglas fir stands in 2 015 after cones began ripening. In 2016, nutcracker detections in limber pine stands peaked in late August and steadily dropped afterward. Relative to prior years, nutcracker detections in 2016 were comparatively low in ponderosa pine stands and comparativ ely high in Douglas fir stands, with detections increasing as cones ripened for each species. Nutcracker Foraging and Caching Behavior I observed differences in the timing and behavior of nutcrackers foraging on each conifer species. I observed evidence of nutcracker damage on unripe limber pine cones (torn cones, with the red brown exposed core contrasting with the unripe green color of the exterior) as early as August 7 th During this time, birds tore away scales and foraged on the developing seeds but we re not caching seeds. Nutcrackers either positioned themselves atop an attached cone, or removed cones from branch ends and used a sturdy object as an anvil including rocks, branches, and logs. Once limber pine cones began ripening, nutcrackers

PAGE 35

28 removed the ripe seeds directly from cones on trees and began pouching the seeds for transportation to caching sites. Some birds were observed to cache seeds nearby in the ground or in a log, rock, or tree, while others flew off with full throat pouches to cache elsewhere. I observed nutcrackers foraging on unripe ponderosa pine cones as early as September 1 st ; they for aged on these cones in a manner similar to unripe limber pine cones. Once ponderosa pine cones ripened, nutcrackers removed seeds by loosening cone scales and then removed the seed, holding it in their bill. Nutcrackers then removed the seed wing by either shaking the seed or rubbing it against a tree branch. Birds we re observed to pouch seeds, with some bi rds caching nearby and others flying out of view to cache elsewhere One nutcracker was observed to forage on unripe Douglas fir cones on August 12 th but most foraging on these cones was observed after ripening. Within 10 to 15 days of Douglas fir cones r ipening, nutcrackers were observed to carefully remove intact Douglas fir seeds from between the cone scales. Nutcracker flocks were later observed to forage primarily on the ground for Douglas fir seeds that had been blown out of cones. Similar to pondero sa pine, birds removed the seed wings of Douglas fir seeds which appeared to detach more easily. Some birds were observed to pouch Douglas fir seeds an d subsequently cache them approximately 20 meters away or less In 2014, when cone crops were relatively small for all conifer species, I observed nutcrackers to forage only on the seeds of limber pine and ponderosa pine (Fig. 5 ). Nutcrackers in 2014 were first observed to forage on unripe limber pine cones in late August, continuing to forage on ripened cone s that September. Beginning in early October, observations of nutcrackers foraging on limber pine seeds declined, and instead I observed them foraging on ponderosa pine seeds, shortly after seed ripening. At the end of October, a

PAGE 36

29 small flock of nutcrackers was observed to raid red squirrel ( Tamiasciurus hudsonicus ) middens for closed limber pine cones in the Estes Cone stand. Squirrels often chattered at and chased the nutcrackers during this time, especially if a nutcracker perched nearby after removing a cone. Nutcrackers mostly used tree branches as anvils when foraging on these cones. In 2015, when ponderosa pine experienced a large cone crop, nutcrackers again foraged only on limber and ponderosa pine seeds (Fig. 5 ). While nutcrackers again began foraging on unripe limber pine cones in late August they were ob served to begin foraging on ponderosa pine cones one month earlier in 2015. This was the only year that nutcrackers were observed to forage on unripe ponderosa pine cones. In 2016, when Douglas fir experienced a large cone crop, nutcrackers were observed t o forage first on limber pine seeds ( with one bird briefly observed to forage on unripe Douglas fir seeds) until late August, and then they focused only o n ripe Douglas fir seeds (Fig. 5 ). This is the only year that I observed nutcrackers to forage on Doug las fir seeds and also the only year I did not observe nutcrackers to use ponderosa pine seeds. This year was also unique in that I observed the majority of nutcracker foraging on limber pine prior to cone ripening of this conifer. I have observed nutcrac kers to utilize two conifer species each year, but the odds of use were not always equal. In 2014, I observed that about half of the nutcrackers near ponderosa pine trees to use ponderosa pine seeds while the majority of birds near limber pine trees used l imber pine seeds (Table 3 ). The odds of nutcrackers using ponderosa pine seeds in 2014 were from 0.05 to 0.77 times lower than the odds of nutcrackers using limber pine ing behavior, because the odds of nutcrackers using ponderosa pine seeds were now from 1.34 to 7.34 times greater than the odds of nutcrackers using limber pine seeds. In 2016, when I

PAGE 37

30 observed nutcrackers foraging on Douglas fir seeds for the first time in this study, the odds of nutcrackers using these seeds were from 0.81 to 6.98 times the odds of nutcrackers using limber pine seeds. The observed numbers of birds using and not using seeds for each conifer species are reported in Table 3 Discussion My goal for this study was to determine the relative importance of limber pine and alternative seed sources with respect to their potential value as a food source and as habitat to nutcrackers in RMNP. I first discuss my reasoning for assigning the relative i mportance of each food source followed by the relative importance of these habitats. Across years, conifer species provided markedly different estimates of food energy across the landscape in RMNP. I found most cone density estimates for conifer specie s t o be below 1 ,000 cones per ha, which I am defining as a rel I (Fig. 2). I am defining limber pine cone production in relation to the greater cone density estimates of ponderosa pine in 2015 and Dou gl as more kJ of potential food energy in RMNP than a low cone crop of ponderosa pine and 27.7 to 49.2 times more kJ than a low cone crop of Douglas fir, because of differences in seed energy per cone. In contrast, high cone production in ponderosa pine results in 2.3 to 4.1 times more kJ in RMNP than a low cone crop of limber pine. Mast cone production in Douglas fir results in 5.4 to 9.6 times more kJ in RMNP than a low cone crop of limber pine. These comparisons do not account for insect infestation. If my estimates of cone infestation

PAGE 38

31 by insects are accur ate, total energy across the RMNP landscape provided by limber and ponderosa pine during my study would be approximately half of what I report here. Nutcrackers are considered energy sensitive foragers, and landscape energetics likely impact the foraging d ecisions of nutcrackers and their choices of conifer species as seed sources each year (Tomback 1978, Vander Wall 1988). In Table 2 we see that limber pine shows stable production of energy every year (even with low cone crops), Douglas fir production is the most variable (ranging from relatively minimal energy to the maximum observed), and ponderosa pine production is relatively intermediate (with small crops producing more energy than a small crop of Douglas fir and a large crop producing less energy tha n a large crop of Douglas fir). Considering the changing landscape energetics each year, as well as the constants of seed size, cone morphology, and ripening phenology of each conifer species, I s. Limber pine was the only seed source that nutcrackers used intensively every year. It was also the first seed source to be used each year, despite the fact that limber pine produced a relatively small cone crop each year, its stands are relatively small and scattered, and it comprises the smallest forested area of the three species on the landscape in RMNP. However, its large seed size, high landscape energetics during a low cone crop year, and early ripening date support the importance of this seed sour ce. In addition, limber pine seeds ripen at what appears to be a n important time in the summer after most cached seeds from the previous fall were retrieved (e.g., Tomback 1978, Vander Wall and Hutchins 1993) but prior to seed ripening of other conifer spe cies. The seeds of the heavily armored ponderosa pine, like the closely related Jeffrey pine (Tomback 1978), may be energetically costly to remove from unripe cones. Nutcrackers appear to search widely for limber pine seeds as I

PAGE 39

32 observed them to forage on isolated limber pine trees in stands of other forest types on multiple occasions. If some nutcrackers are residents within RMNP and remember the locations of productive limber pine trees within their home ranges, they may utilize seeds from these trees if energetic rewards are sufficient. This suggests a landscape level approach to seed utilization for some birds, in contrast to stand based decisions. For the closely related whitebark pine, observations have suggested that the probability of visitation or seed harvest by nutcrackers in stan ds increases as cone production increases (McKinney et al. 2009, Barringer et al. 2012), but Barringer et al. suggested that foraging decisions may occur at a landscape rather than stand level. In this study, however, cone production levels for limber pine and ponderosa pine were higher generally than many whitebark pine stands assessed by McKinney et al. (2009) and Barringer et al. (2012). Nutcrackers were observed to forage intensively on ponderosa pine seeds for two out of three years, including a small cone crop year. The odds of nutcracker foraging on ponderosa pine seeds were lower compared to limber pine seeds during the low ponderosa pine reward during a small cone cr mast year, the odds of a nutcracker using pon derosa pine seeds were greater compared to limber pine, indicating that ponderosa pine importance to nutcracker diet may have been greater in 2015 com pared to limber pine. Nutcrackers foraged on Douglas fir seeds only during the mast year. Even though the energy available for Douglas fir during a mast year was estimated to be considerably greater than that for limber pine during a low cone crop year, t he odds of nutcracker seed use were

PAGE 40

33 not observed to be greater for Douglas fir compared to limber pine that year. This likely While the importance of each conifer sp ecies to nutcrackers as a food source appears to vary each year, I believe that over time each conifer species plays a role in maintaining nutcracker populations in RMNP. I argue that limber pine is the most important source of food because its seeds were used every year by nutcrackers and it fills a gap in time when other conifer seeds are not available. The odds of nutcracker foraging on limber pine seeds were higher compared to other conifer species experiencing a similarly sized cone crop they were even comparable to the odds of nutcrackers foraging on Douglas fir seeds during a mast year. Comparing the alternative seed sources, I suggest that ponderosa pine has a greater role in maintaining nutcracker numbers over time than Douglas fir, based on several reasons. The first is that Douglas fir seeds were not observed to be used by nutcrackers during low cone production while ponderosa pine seeds were. The second involves how nutcrackers utilized seeds during a mast year for each species: (1) the odds of nu tcracker foraging were higher for ponderosa pine compared to limber pine, the stable food source for nutcrackers, while the odds of foraging were comparable between Douglas fir and limber pine (the odds of foraging were higher for ponderosa pine compared t o Douglas fir as well); and (2) nutcrackers were observed to forage on ponderosa pine seeds for a greater period of time in the fall compared to Douglas fir. Finally, ponderosa pine generally experiences moderate to high cone production years more frequent ly than Douglas fir (Krugman and Jenkinson 1974, Hermann and Lavender 1990) Each forest type of interest appears to provide important habitat for nutcrackers because I observed them to visit each forest type across all years and occupy every study

PAGE 41

34 stand. The production of seeds is an obvious incentive to nutcrackers to visit these forest stands, since I observed nutcrackers forage on the seeds of all three conifer species. However, during time periods and even years when I did not observe nutcrackers to f orage on the seeds of a specific conifer species, nutcrackers were still present in those forest types. This indicates that there may be benefits to this habitat other than just food. The importance of these two alternative seed sources also has been note d elsewhere in other studies. For example, Giuntoli and Mewaldt (1978) observed that nutcrackers had foraged on the seeds of both species in Montana. More recently, Lorenz and Sullivan (2009) observed nutcrackers to utilize both species in the Cascade Rang e of Washington State. Schaming (2016) observed nutcrackers in Wyoming to preferentially select Douglas fir forests as habitat during the breeding season and she emphasized the importance of this species as a food source to nutcrackers in this region. As f ive needle white pines experience mortality from blister rust and other disturbances, ponderosa pine and Douglas fir may nutcrackers. Given that nutcrackers are energetica lly sensitive foragers (Tomback 1978, Vander Wall 1988, Barringer et al. 2012), and I believe optimal foraging theory may help to explain conifer seed importance. While landscape energetics may detail how much energy is available to nutcrackers, we must al so consider the rate of energy acquisition. Larger seeds allow for greater reward per extracted seed, but the time for seed extraction (finding and removing seeds from cones or on the ground) and processing time (removing seed wing, if present, and seed co at as well as assessment of seed soundness through bill clicking) is also likely to differ among conifer species. Nutcrackers were mostly observed to remove limber

PAGE 42

35 and ponderosa pine seeds from ripening cones with opening cone scales which reduces foraging time and effort. Tomback (1978) observed nutcrackers in California to forage on the seeds of whitebark pine (closely related to limber pine) at a faster rate compared to the seeds of Jeffrey pine (closely related to ponderosa pine); a similar relationship may exist between limber and ponderosa pine. Even if we assume that seed extraction rate is approximately equal between limber and ponderosa pine, but seed processing time is longer for each ponderosa pine seed because of seed wing removal, then limber pi ne leads to a faster rate of energy acquisition during low cone crops. However, a large ponderosa pine cone crop may lead to a faster rate of energy acquisition if the abundance of cones appreciably decreases seed extraction time. Douglas fir seeds are con siderably smaller than limber and ponderosa pine seeds (Table 2) and also have seed wings; nutcrackers foraging on Douglas fir seeds would thus experience a very low rate of energy acquisition per seed. Nutcrackers may gain energetically only by foraging o n these seeds during a large cone crop, with many cones within the same canopy. When nutcrackers removed Douglas fir seeds from cone scales shortly after cones ripened, they appeared to do so carefully and slowly, which may result in lengthy seed extractio n times. The mature cones also release seeds quickly, unlike ponderosa pine and limber pine, which are less synchronous. Later in the fall, nutcrackers foraged for Douglas fir seeds only on the ground ; I am unsure how this impacts seed extraction time. Con sidering similarities between nutcrackers foraging on limber pine and Douglas fir in 2016, in terms of odds of seed use and length of time observed foraging, the rate of energy acquisition appeared to be similar between limber pine and Douglas fir probably the result of the small Douglas fir seed size and perhaps that nutcrackers are better adapted for foraging on limber pine. If this is the case, the greater energy across the landscape during Douglas

PAGE 43

36 erence by nutcrackers in comparison to limber pine. My observations of nutcracker seed use are under conditions of comparatively low limber pine mortality from the mountain pine beetle outbreaks during the preceding 15 years and no observed mortality from white pine blister rust in RMNP. Increasing limber pine mortality will likely impact nutcracker foraging decisions as cone production is reduced from the disease. A major aspect of my study goal was to determine the importance of multiple seed sources to nutcracker foraging, as nutcrackers will need to rely on these species if limber pine seed production decreases in RMNP. I believe that the alternative seed sources in RMNP may support a nutcracker population, although the carrying capacity may drop and frequency of nutcracker emigration may increase (e.g., Van der Wall et al. 1981, Lorenz and Sullivan 2009). Based on my observations, ponderosa pine will be more important in supporting the nutcracker population over time. It is also likely that nutcrackers will use the seeds of both ponderosa pine and Douglas fir to a greater extent than I observed. It should be noted that RMNP is located at the western edge of the ponderosa pine and Douglas fir forest zones, with additional forest area for these habitat types in adjacent national forest. If the entire extent of these zones are considered, they may serve an even greater role as food sources. Indeed, nutcrackers are present east of the park, where limber pine is much less common. If some of these birds are residents, they could be sustaining themselves primarily on ponderosa pine and Douglas fir. If nutcrackers do not have access to limber pine seeds during late summer, they may instead sustain themselves on insects and occasionally on small vertebrates and any remaining cached seeds from the previous fall;

PAGE 44

37 whether these food sources are adequate prior to cone ripening by ponderosa pine and Douglas fir requires further study. With the blister rust pathogen potentially spreading, as it as for the last century (Geils et al. 2010, Tomback and Achuff 2010), nutcrackers may not be limited to using only alternative seeds. It is estimated that insects infested 79% of all limber pine cones in RMNP in 2016. If we assume that there was no infestation but blister rust instead damaged 79% of limber pine cones in RMNP, we could t heoretically expect to see similar foraging behavior by nutcrackers as was observed during this study, but unfortunately, the effects from cone insects and blister rust are likely to be additive. In addition, I have detected nutcrackers in all stands acro ss all years (including stands with low cone crops and/or no foraging), observed them to travel up to 12 km straight line distance (W illiams, Chapter III ), and observed them to forage on isolated limber pine trees within stands of other forest types. This means that nutcrackers travel widely across the park, with individuals moving long distances to find seeds. Even with high mortality from blister rust, nutcrackers may be able to search for, forage on, and cache a portion of the limber pine seeds produced each year.

PAGE 45

38 Tables Table 1. Geographic data for each of the 13 study stands. The reported tree density is the point estimate as calculated by Program Distance; 95% confidence intervals are also provided for tree density. Forest type tree density was calculated by Program Distance using data from all stands. F orest type S tand location S ize (ha) S lope () A spect T ree density (trees per ha) 1 95% C onfidence interval L imber pine Beaver Ponds 62.2 11 194 96 63 145 L imber pine Rainbow Curve 7.4 24 152 169 93 306 Li mber pine Estes Cone 89.1 16 129 654 318 1345 L imber pine Windy Gulch 398.9 20 187 530 163 1716 L imber pine Black Canyon 18.4 15 211 64 57 73 Forest type tree density: 361 176 737 P onderosa pine North Lateral Moraine 404.7 15 201 120 57 255 P onderosa pine Deer Ridge, south slope 167.8 14 167 128 43 384 P onderosa pine Bighorn Mountain 154.43 20 167 203 46 891 P onderosa pine Sheep Mountain 72.6 23 170 121 26 550 Forest type tree density: 133 101 175 Douglas fir South Lateral Moraine 149.4 38 356 152 35 653 Douglas fir Deer Ridge, north slope 45.5 23 335 182 89 368 Douglas fir Eagle Cliff Mountain 104.14 23 320 1086 268 4401 Douglas fir Cow Creek 29.96 17 17 712 501 1011 Forest type tree density: 503 266 952

PAGE 46

39 Table 2. Landscape energetics for each conifer species within RMNP. For each conifer species, the minimum and maximum point estimate of cone density across all years are examined. Because limber pine cone productivity was relatively consistent during this study, th e minimum and maximum point estimates are similar. In contrast, the maximum point estimates for ponderosa pine and Douglas fir occurred during high cone crop years, in 2015 and 2016, respectively. C onifer species Cones per ha Forest type area (ha) Seeds per cone T otal seeds kJ per seed T otal kJ L imber pine 800 2 200 90 1 1.6 10 8 2.8 2 4.5 10 8 L imber pine 1 400 --2.9 10 8 -8.0 10 8 P onderosa pine 700 3 400 50 3 1.2 10 8 1.1 4 1.4 10 8 P onderosa pine 9 600 --1.7 10 9 -1.9 10 9 Douglas fir 300 4 300 50 5 6. 8 10 7 0.2 5 1.6 10 7 Douglas fir 92 90 0 --1.8 10 10 -4.3 10 9 1 Vander Wall and Balda 1977 2 Lanner 1996 (calories/gram) and Krugman and Jenkinson 1974 (grams/seed) 3 Value is averaged between point estimates of Curtis and Lynch 1965 and Schubert 1974 4 Vander Wall and Balda 1977 (calories/gram) and Krugman and Jenkinson 1974 (grams/seed) 5 Smith 1970 Table 3 Odds ratios calculated for a) 2014, b) 2015, and c) 2016. The numbers for each conifer species reflect the numbers of nutcrackers observed and not observed using seeds (foraging and/or caching). Odds ratios indicate the odds of nutcracker seed use of the alternative conifer seed source (top row) compared to the o dds of nutcracker seed use of limber pine (bottom row). a. 2014 Foraging and caching counts Conifer species No seed use Seed use Ponderosa pine 7 6 Limber pine 8 34 Odds ratio of alternative seed use = 0.20 95% Confidence interval = 0.05 0.77

PAGE 47

40 b. 2015 Foraging and caching counts Conifer species No seed use Seed use Ponderosa pine 16 59 Limber pine 17 20 Odds ratio of alternative seed use = 3.13 95% Confidence interval = 1.34 7.34 c 2016 Foraging and caching counts Conifer species No seed use Seed use Douglas fir 10 19 Limber pine 15 12 Odds ratio of alternative seed use = 2.38 95% Confidence interval = 0.81 6.98 Figures a.

PAGE 48

41 b. Figure 1. The map in a) shows the location of the study area, RMNP, in Colorado. The study stand locations in RMNP are shown in b), in addition to the distribution of forest types within RMNP.

PAGE 49

42 a. b.

PAGE 50

43 c. Figure 2 Estimated cone densities for a) 2014, b) 2015, and c) 2016. Open circles indicate point estimates; errors bars indicate 95% confidence intervals. All estimates were calculated by program Distance (Thomas et al. 1998). The same scale is used for 2014 and 2 015; the scale for 2016 is larger.

PAGE 51

44 a. b. y = 1.6509x R = 0.945 0 20 40 60 80 100 120 140 160 180 200 0 20 40 60 80 100 120 Total number of cones observed on tree Number of cones on tree observed from transect Limber Pine Cone Count Regression y = 1.4073x R = 0.8787 0 50 100 150 200 250 300 350 400 450 0 50 100 150 200 250 300 Total number of cones observed on tree Number of cones on tree observed from transect Ponderosa Pine Cone Count Regression

PAGE 52

45 c. Figure 3. The relationship between the number of cones observed from the transect and the total number of cones observed on the tree for a) limber pine, b) ponderosa pine, and c) Douglas fir. The slope of the regression line was used to estimate the total number of cones on each tree for each conifer species for Program Distance. The resulting cone density estimates were used for the estimation of landscape energetics of for each c onifer species. y = 1.9032x R = 0.9159 0 200 400 600 800 1000 1200 1400 0 100 200 300 400 500 600 700 Total number of cones observed on tree Number of cones on tree observed from transect Douglas fir Cone Count Regression

PAGE 53

46 a. b.

PAGE 54

47 c. Figure 4 Nutcracker visitation in a) 2014, b) 2015, and c) 2016. For each 10 day bin, the total number of surveys with detections is divided by the total number of surveys; both visual and auditory detections are included. Detections need to be within the stand being surveyed to be included. Arrows indicate the average time of cone ripening for each conifer species. a.

PAGE 55

48 b. c. Figure 5 Nutcracker foraging timeline in a) 2014, b) 2015, and c) 2016. Time observed foraging (in sec) is summed across all birds and divided by survey time (in min) within forest types for each 10 day bin; reported units are observed foraging sec per survey min. A different scale is used in 2015, compared to 2014 and 2016.

PAGE 56

49 CHAPTE R III LIMBER PINE METAPOPULATION STRUCTURE AND CONNECTIVITY IN ROCKY MOUNTAIN NATIONAL PARK Introduction Metapopulations are defined as large, regional populations consisting of smaller, local populations connected by migration (dispersal). The constituent local populations are susceptible to extinction, colonization, and recolonization (Hanski and Gilpin 1997, Hanski 1999). There are many metapopulation models that been devised and explored, some theoretical and some simulating natural metapopulation syste ms, with the goal of examining persistence over time (Levins 1969, Levins 1970, Hanski 1999). Metapopulations are of increasing importance to managers as continuous populations become fragmented or reduced in size by habitat alteration or invasive pests an d pathogens (Opdam 1991, Fahrig 2002). However, many natural populations exist as metapopulations, and these, too, are now vulnerable to disturbance. Examples of natural metapopulations include the population structure for several butterfly species, where local populations often experience limited migration and relatively high extinction risk. For example, Granville fritillary ( Melitaea cinxia ) metapopulations in Finland have been extensively studied by Hanksi (e.g., Hanski 1994) and bay che ckerspot butterfly ( Euphydryas editha bayensis ) metapopulations have been studied in California (Harrison et al. 1998). The population structure of mammals and other mobile vertebrates is often less suitable for this theory, but some species form metapopul ations with lower extinction risk for local populations compared to insects. For example, the local populations of bighorn sheep ( Ovis canadensis ) metapopulations in western North America inhabit steep

PAGE 57

50 canyons in mountains and deserts (Bleich et al. 1990, Singer et al. 2000). Small populations of the American pika ( Ochotona princeps ) inhabit scattered talus slopes throughout the high mountains of western North America (Moilanen et al. 1998, Beever et al. 2003). Metapopulation theory has also been applied to the management of spotted owls ( Strix occidentalis ), which inhabit isolated stands of old growth forest along coastal ranges of southern California (Lahaye et al. 1994, Guti r rez and Harrison 1996). The application of metapopulation theory to plant speci es is less common. However, some authors suggest that plant species in general exist as metapopulations (Husband and Barrett 1996). This theory has been applied to the prickly lettuce ( Lactuca serriola ) in Europe, for example (Petrzelova and Lebeda 2008). However, the regional distributions of most forest tree species are less suitable for metapopulation theory, because the trees tend to occur either as large, continuous forest stands or may be one component of a widely distributed forest habitat type. Limb er pine ( Pinus flexilis ), a five needle white pine (Subgenus Strobus ) that ranges throughout the mountainous regions of the western U.S. and southwestern Canada, and occurs as regional metapopulations, may be one exception (Webster and Johnson 2000). Limber pine is widely distributed both regionally and in elevation from lower to upper treeline ( Steele 1990, Tomback et al. 2005). It is also a pioneer species, establishing in sites after a disturbance (Rebertus et al. 1991). Limber pine often forms ser al stands in mesic sites where more competitive conife r species replace limber pine through succession, but may form climax stands in harsh areas with dry soil, intense radiation, and high wind velocity. Limber pine stands are relatively small, patchy, and isolated, forming regional

PAGE 58

51 seeds, or propagules. Wind is the primary (moving seeds from tree to ground) seed dispersal agent for most tree species, but limber pine see nutcracker ( Nucifraga columbiana ) (Lanner and Vander Wall 1980, Tomback and Kramer 1980, Benkman et al. 1984). The nutcracker and limber pine are coadapted mutualists: nutcrackers transport the large, wingless seeds of limber pine and bury them in ground caches as winter and spring food sources. After snowmelt or summer rains, seeds within unretrieved caches may germinate and lead to forest regeneration ( Tomback and Linhart 1990). Whereas nutcrackers forage on the seeds of several regional conifer species (Tomback 1998), limber pine primarily relies on nutcrackers for long distance seed dispersal; rodents may disperse seeds over short distances (Tomback et al. 2005). Nutcrackers have been estimated to cache tens of thousands of pine seeds per bird each year (Vander Wall and Balda 1977, Tomback 1982). In addition, they may cache seeds near the foraging site or several kilometers away (Tomback 1978). Webster and Johnson (2000) sampled 13 stands of a limber pine met apopulation within the Kananaskis Valley in Canada to determine whether limber pine exists as a classic metapopulation where local populations experience high extinction risk and limited seed dispersal or a patchy population metapopulation with low extinct ion risk for local populations and extensive (non limiting) seed dispersal. They determined that extinction events were infrequent; only two stands were consumed by a stand replacing fire within the past century (there was no evidence for other causes of e xtinction). Stands were recolonized within approximately five years. Recolonization of stands likely resulted from nutcrackers dispersing seeds long distances from extant limber pine stands, because stands were isolated by at least a few kilometers too far for wind or rodent mediated seed dispersal. With

PAGE 59

52 infrequent extinction events and immediate recolonization by nutcracker caching, Webster and Johnson (2000) concluded that limber pine existed as a patchy population metapopulation. I examined the limber p ine metapopulation in Rocky Mountain National Park within RMNP; and, (2) examine how constituent populations within the park may be connected by nutcrackers, as determi ned by nutcracker flight and seed dispersal distances. In other words, what is the average number of stands a nutcracker may potentially connect through seed dispersal and caching flights? If the nutcracker population is reduced, how will this affect limbe r pine population connectivity within the RMNP metapopulation? Methods Study Area I seed dispersal distances in RMNP (Fig. 1) while also investigating nutcracker foraging ecology (see Chapter 2 for details and map ) from mid June through late October in 2014, 2015, and 2016. Radio telemetry was conducted from June to October, 2015 and 2016. I determined that l imber pine woodlands cover 2,212 ha of landscape within the park ( see Chapter II) Additional habitat types within RMNP include ponderosa pine ( Pinus ponderosa ) parklands, Douglas fir ( Pseudotsuga menziesii ) forests, lodgepole pine ( Pinus contorta ) forests, spruce fir ( Picea engelmannii and Abies lasiocarpa ) forests, alp ine tundra, dry and moist meadows and grasslands, riparian corridors, and talus slopes. These habitats cover a wide elevation range within this high altitude park, 2380 4350 m.

PAGE 60

53 Limber Pine Metapopulation S tructure To examine the spatial configuration of the RMNP limber pine metapopulation, I first obtained ArcGIS vegetation layers for RMNP from the National Park Service. Within the limber pine layer, I selected polygons whose borders were within 500 m of each other and identified each group of selected polygons as an individual limber pine stand; stands were thus separated from each other by at least 500 m. Each identified stand was exported as an indivi dual layer. I only identified stands that were at least 1 ha in size and whose centroid was within the RMNP boundary. I obtained information on area (number of hectares), elevation (m), slope (), and aspect () from the attribute tables of these stands. I also calculated the distance (km) from each stand to every other stand, measured from the centroid of each stand. Nutcracker F light and S eed D ispersal D istances To trap nutcrackers for radio tagging, I set up trapping stations concealed in rocky, foreste d sites near Rainbow Curve, 3300 m, from July through August in 2015 and 2016. Nutcrackers routinely fly to Rainbow Curve throughout the summer and fall to obtain tourist handouts (signs are posted prohibiting wildlife feeding). I baited the walk in traps with peanuts, both with and without shells. I attached glue on dorsal (A2480) or tailmount (A4560) model radiotransmitters (Advanced Telemetry Systems) to 10 nutcrackers in 2015 and 4 nutcrackers in 2016. (The use of backpack harnesses was not compatible w ith management policy.) I tracked tagged birds from August to October, 2015 and 2016, using a Yagi antenna and R410 scanning receiver (both from Advanced Telemetry Systems). I tracked each tagged bird at least once every two weeks. Upon finding a nutcracke r, I recorded its position using a Garmin GPSMAP 62stc. The nutcracker was then followed for

PAGE 61

54 two hours and I recorded a new point whenever the nutcracker flew a distance of at least 15 meters. I attempted to track two birds during each tracking day. If a b ird was located in hazardous terrain, I used triangulation for point locations. I uploaded the point locations for all birds to ArcGIS. For each bird that retained its tag long enough to record point locations more than one kilomet er from the trapping sit e (n = 7 birds), the maximum flight distance was calculated as the maximum distance between point locations. If a tagged bird was observed to cache seeds (n = 1 bird), I calculated seed dispersal distance as the distance between the foraging site and cachi ng site. Connectivity With the computed distances between all limber pine stands (populations), I determined average distances for the nth nearest neighbor for each population. I compared this information to flight distances obtained from radio tagged birds in order to estimate how many stands an individual nutcracker could connect. I also included nutcracker flight distances from the literature for this comparison. Results Limber Pine Metapopulation S tructure in Rocky Mountain National Park I identified 51 stands of limber pine within RMNP ( Table 1, Fig. 1 ). These stands are primarily located east of the continental divide, with only 5 stands located on the west slope. Stands are most frequently located on or near ridges within the park, especially ridgelines with exposed granite. Consequently, stands are often associated with steep slopes, especially the drier south facing aspects (Table 1). Limber pine is distributed across a wi de elevation range in RMNP, from 2,702 to 3,447 m. The median stand size is 12 ha, but stand size is highly variable and ranges from 1.0 ha to 400.5 ha in size (Table 1) Size distribution appears

PAGE 62

55 to vary across mountain ranges in RMNP for example, limber pine occurs in several places and in small stands (ranging from 1.0 to 31.5 ha) in the northern Mummy Range, and the Longs Peak massif contains mostly larger stands that range from 113.3 to 292 ha in size Nutcracker Flight and See d D ispersal D istances I attached radio transmitters to 10 birds in 2015 and to four birds in 2016; one 2015 bird that had lost its tag was recaptured and a second tag applied. After transmitter attachment, tagged nutcrackers at first were present either at Rainbow Curve or at a nearby location in the subalpine forest, within a kilometer of this turnout. Seven birds either removed their tags or their transmitter signal was lost before point locations were ever recorded outside of this local range. The maximum flight distances rec orded from the remaining six birds ranged from 1. 8 to 12 km, with a median of 3.8 km (Tab le 2 ). The shortest distance was recorded from a bird searching a northwest facing slope of Hidden Valley, possibly for ripe limber pine seeds. The maximum distance was recorded from a bird on Prospect Mountain. The remaining birds were detected either along D eer Ridge, the Beaver Ponds moraine, or the southern stretch of the Mummy Range, and may have been foraging on ripening ponderosa pine or Douglas fir seeds (see Chapter 2). There is likely a strong bias towards short flights in the reported maximum flight distances for these nutcrackers due to the limited duration of transmitter attachment. The only nutcracker to retain its tag for the duration of a field season was also the only bird observed to disperse seeds. After being tagged, this bird stayed near Ra inbow Curve; at the end of September, it flew down to the south facing slope of Deer Ridge to forage on ripening ponderosa pine seeds. For the remainder of the field season, it was ong with one to five

PAGE 63

56 other nutcrackers. It briefly cached seeds within this area on some occasions, flying short distances (see Chapter 2), but once every 35 to 50 min it would fly north or northwest with a full sublingual pouch, indicating that it was als (e.g., Vander Wall and Balda 1977, Tomback 1978, Lanner and Vander Wall 1980). I noted its direction and scanned from Horseshoe Pa rk, where the signal directed me to the north facing slope of the Beaver Ponds moraine. I identified this as the caching site, which was about 2.2 km from the foraging site my only observed long distance seed dispersal distance (Table 2). I should note that while on the moraine, I detected that the bird was close but never observed i t. The signal also appeared to move further northwest on some occasions, indicating the bird may have cached further away, and the seed dispersal distance is underestimated. The bird was observed to roost on this moraine as well. Additional nutcracker fli ght distances in RMNP are reported in Tomback and Taylor (1987), with a maximum of 14.5 km. Seed dispersal distances are rarely reported for limber pine seeds in the literature. Vander Wall and Balda (1977) observed nutcrackers to disperse limber pine seed s 4 to 5 km and Vander Wall (1988) observed them to disperse limber pine seeds approximately 1 km. The furthest observed nutcracker seed dispersal distances (regardless of conifer species) include 12.5 km (Tomback 1978), 22 km (Vander Wall and Balda 1977), and 32.6 km (Lorenz and Sullivan 2011). Inter P opulation Connectivity Fig. 2 displays the distribution of nearest distances for limber pine stands in RMNP. For any given limber pine stand, the nearest stand is 1.6 km away on average and the furthest stand is 32.0 km away on average. Based on the seed dispersal distance of 2.2 km observed here, a nutcracker could potentially connect nearest stands on average. Based on the

PAGE 64

57 maximum observed flight distance of 12 km, a nutcracker could potentially connect appr oximately 39% of the limber pine metapopulation in RMNP. Based on the maximum seed dispersal distance in the literature, 32.6 km, a nutcracker could potentially connect the entire metapopulation if it were highly mobile in its foraging and caching behavior Discussion The spatial configuration of the RMNP limber pine metapopulation has implications for how the constituent populations could function over time. The east side of RMNP appears more suitable for the limber pine metapopulation, because it contain s more than 90% of the stands in the park. It is possible that the western slope is generally more mesic or experienced less disturbance in the past. The isolation and small size of the five western stands may lead to lower connectivity relative to eastern stands; nutcrackers may visit western stands less frequently and recolonization following extinction could take longer than it might on the east slope Stand size is highly variable and larger stands may contribute more to seed dispersal within the metap opulation. The largest stand, at 400.5 ha and located on Trail Ridge, may be so large that loss of all trees is unlikely, except in the case of a severe, stand replacing fire. In the case of a portion of trees within the s tand experiencing mortality, for example from mountain pine beetle or lightning caused ground fire, these openings could quickly be regenerated by seeds cached by nutcrackers from nearby limber pine trees. I have observed nutcrackers to frequently visit this stand, and they m ay cache seed s harvested here in other limber pine stands located elsewhere in the park If these assumptions are true, this stand

PAGE 65

58 seeds for recruitment within extant stands and recolonization within extinct stands. The four larger stands located on the Longs Peak massif (totaling 718 ha) may act in a similar manner. I experienced challenges in tagging nutcrackers leading to limited data for nutcracker flight and see d dispersal d istances I am confident that 9 of 10 birds removed tags prior to the end of the field season in 2015. In 2016 I applied more epoxy and tagged two birds in August to ensure that I was late in the molt cycle, so that feathers along the spinal tract would no t molt shortly after tag attachment. However, I am confident that all birds in 2016 also removed their tags. Attaching radio transmitters with epoxy directly to bird skin and plumage is not workable for nutcrackers, because they persist at tag removal. Tag removal was confirmed by recapturing two previously tagged birds and finding the dropped tag via radio telemetry for seven birds. I cannot confirm tag removal for all birds in 2015 and 2016, because the transmitters from these other nutcrackers either: (1 ) were no longer detected in the park (n = 2 birds), likely a result of the removed tag falling into a crevice or other structure that blocked the signal, although the birds may have left the park with tags still attached; or (2) remained in one location i naccessible to me until the end of the field season (n = 3 birds) likely a result of tag removal, given the obse rved mobility of these birds. Based on my limited data and data from the literature on flight distances one nutcracker has the potential to con nect a large percentage of the metapopulation in RMNP through seed transportation and caching; even a small population of nutcrackers could potentially maintain the metapopulation structure over time. This is assuming that long distance seed dispersal dist ances observe d elsewhere are routine for nutcrackers in RMNP. Unfortunately, limber pine is presently facing more threats than just wildfire. A combination of warmer temperatures, drought, and the abundance of lodgepole pine has

PAGE 66

59 resulted in severe mountai n pine beetle outbreaks ( Dendroctonus ponderosae ; a native species), impacting limber pine stands in RMNP and throughout the Rocky Mountains (Romme et al. 1986, Lynch et al. 2006, Gibson et al. 2008). In addition, fire suppression has resulted in less limb er pine across its range by facilitating the succession of more competitive species (Gruell 1983, Rebertus et al. 1991). The most serious threat is the invasion of the nonnative fungal pathogen Cronartium ribicola which causes the disease white pine blist er rust in five needle white pines (McDonald and Hoff 2001). Blister rust has killed 90% of limber pines in some stands in Wyoming and Montana (Kearns and Jacobi 2007, Tomback and Achuff 2010, Schwandt et al. 2010). Blister rust was recently detected in no rthern Colorado and RMNP, with minimal to no mortality at this early stage (Johnson and Jacobi 2000, Schoettle et al. 2011). It is expect ed that blister rust will cause extensive loss of limber pine across the landscape in RMNP in the coming decades. In ad dition, climate change is likely to complicate some or all of these threats, including more severe fire and major outbreaks of mountain pine beetle ( Tomback and Achuff 2010, Tomback et al. 2011). Nutcrackers have been shown to visit stands with low cone p roduction with a lower probability (McKinney and Tomback 2007, McKinney et al. 2009, Barringer et al. 2012). With a future likelihood of declining limber pine across RMNP, it is possible that the associated drop in cone productivity could result in reduced visitation and seed dispersal by nutcrackers, or even the periodic absence of nutcrackers in RMNP. Without nutcrackers, the seed drop and rodents (e.g., Tomback et al. 2005) would completely isolate most, if not all, stands from each other. However, nutcrackers may still visit stands, but to a smaller extent, even with increased limber pine mortality. I have observed nutcrackers to forage on the seeds

PAGE 67

60 of isolated limber pine trees within stands of other forest types, indicating that the relationship between cone productivity and nutcracker visitation may operate on a finer scale than previously recognized Resident nutcrackers may learn their environment and include isolated trees in their foraging excursions. Seed caching by nutcrackers in RMNP would be more unpredictable across time and space in this scenario, but would likely maintain the metapopulation over time given the high connecting potential of an individual nutcracker. Tables Table 1. Limber pine local population characteristics (n=51 ) Area (ha) Elevation (m) Aspect Slope () Median 12.3 3194 178.4 22 Min 1.0 2702 NA 12 Max 400.5 3447 NA 42 Table 2. literature a. Clark s Nutcracker use of space in RMNP (km) Seed dispersal distance (n = 1 bird) Maximum flight distance (n = 7 birds) Median 2.2 3.8 Min NA 1.8 Max NA 12

PAGE 68

61 b. Clark s Nutcracker use of space in literature (km) Maximum s eed dispersal distance Maximum flight distance Limber pine (Vander Wall and Balda 1977) 4 5 km NA Liber pine (Vander Wall 1988) ~1 km NA RMNP (Tomback and Taylor 1987) NA 14.5 km Pinyon pine (Vander Wall and Balda 1977) 22 km NA Whitebark pine (Tomback 1978) 12.5 km NA Whitebark pine (Lorenz et al. 2011) 32.6 km NA

PAGE 69

62 Figures Figure 1. Map of individual limber pine stands in RMNP. Area of circle is equal to area of stand. The nearby towns of Estes Park and Grand Lake are also shown, in addition to US Highways 34 and 36.

PAGE 70

63 Figure 2. Distribution of nearest distances of limber pine local populations. 0 5 10 15 20 25 30 35 1 5 9 13 17 21 25 29 33 37 41 45 49 Median distance from given local population (km) nth closest population

PAGE 71

64 R EFERENCES Barringer, L. E.; Tomback, D F.; Wunder, M. B.; McKinney, S T. 2012. Whitebark pine stand condition, tree ab undance, and cone production as predictors of visitation by doi: 10.1371/journal.pone.0037663 Benkman, C. W.; Balda, R. P.; Smith, C. C. 1984. Adaptations for seed dispersal and the compromises due to seed predation in limber pine. Ecology. 65: 632 642. Bent, A. C. 1946. L ife histories of North American jays, cr ows, and titmice. Part 2. U.S. Nat. Mus. Bull. 191. Buckland, S. T.; Anderson, D. R. ; Burnham, K. P. ; Laake, J. L. ; Borchers, D. L. ; Thomas, L. 2001. Introduction to distance sam pling: estimating abundance of biologi cal populations. Oxford University Press, New York. Campbell, R. W.; Dawe, N. K.; McTaggart Cowan, I.; Cooper J. M.; Kaiser, G. W.; et al. 1997. The birds of British Columbia. Vol. 3. P asserines: flycatchers through vireos. R. Br. Columbia Mus., Victoria. Critchfield, W. B. 1986. Hybridization and classification of the white pines ( Pinus section Strobus ). Taxon. 35: 647 56. Critchfield, W B.; Little, Jr., E. L. 1966. Geographi c distribution of the pines of the world. Miscellaneous Publication 991. Washington, DC: U.S. D epartment of Agriculture Forest Service. 97 p. Coop, J. D.; Schoettle, A W. 2009. Regeneration of Rocky Mountain bristlecone pine ( Pinus aristata ) and limber pine ( Pinus flexilis ) three decades after stand replacing fire. Forest Ecol ogy and Management. 257: 893 903. Nutcracker in California. The Condor. 59: 297 307. Davis, J.; Williams, L. 1964. The 1961 irruption of the Cla Wilson Bulletin. 76: 10 18. Decker, F. R.; Bowles, J. H. 1931. Summer birds of t he Blue Mountains, Washington. Murrelet. 12: 12 14. and c hipmunks. The Condor. 58: 386. Geils, B W.; Hummer, K E.; Hunt, R S. 2010. White pines, Ribes and blister rust: a review and synthesis. Forest Pathology. 40: 147 185. Gernandt, D S.; Geada Lopez, G G.; Ortiz Gar cia, S. ; Liston, A. 2005. Phylogeny and classification of Pinus Taxon. 54(1): 29 42.

PAGE 72

65 Gibson, K.; Skov, K. ; Kegley, S ; Jorgensen, C ; Smith, S ; Witcosky, J. 2008. Mountain pine beetle impacts in high elevation five needle pines: Current trends and challenges. Forest Health Prot ection Report R1 08 020. U.S. Department of Agriculture, Forest Service. 32 p. Gibson, K. ; Kegley, S ; Bentz, B. 2009. Mount ain pine beetle. Forest Insect and Disease Leaflet 2. Portland, OR: U.S. Dep artment of Agriculture, Forest Service, Pacific Northwest Region (R6): 12 p. Giuntoli, M.; Mewaldt, L. R. 1978. Stomach contents of C western Montana. The Auk. 95: 595 598. Gruell, G. E. 1983. Fire and vegetative trends in the Northern Rockies: Interpretations from 1871 188 2 photographs. USDA Forest Service General Technical Reporter, INT 252. Ogden, Utah: Intermountain Forest and Range Experiment Station. 117 p. Hanski, I. 1994. A practical model of metapopulation dynami cs. Journal of Animal Ecology. 63(1): 151 162. Hanski, I. 1999. Metapopulation ecology. Oxford University Press, New York. Husband, B. C.; Barrett, S. C. H. 1996. A metapopulation perspective in plant population biology. Journal of Ecology. 84(3): 461 469. IPCC. 2001. Climate change 2001: Impacts, adaptation and vulnerability. In: McCarthy, J. J.; Canziani, O. F.; Leary, N. A.; Dokken, D. J.; White. K. S.; eds. Contribution of working group II to the third assessment report of the IPCC. Cambridge, UK: Cambridge University Press. 1000 pp. Johnson, D. W.; Jacob i, W. R. 2000. First report of white pine blister rust in Colorado. Plant Disease. 84: 595. Kearns, H. S. J.; Jacobi, W. R. 2007. The distributi on and incidence of white pine blister rust in central and southeastern Wyoming and northern Colorado. Canadian Journal of Forest Research. 37: 462 472. Krugman, S. L.; Jenkinson, J. L. 1974. Pinus L. Pine. In: Schopmeyer, S. C., tech. coord. Seeds of woody plants in the United States. Agricultural Handbook 450. Washington, DC: USDA Forest Service. 598 638. Lanner, R. M. 1980. Avian seed dispersal as a facto r in the ecology and evolution of limber and whitebark pines. In Proceedings of Sixth North American Forest Biology Workshop pp. 15 48. University of Alberta, Edmonton, Alberta. Lanner, R. M. 1982. Adaptations of whitebark pine Nutcracker. Canadian Journal of Forest Research. 12: 391 402.

PAGE 73

66 Lanner, R. M.; Vander Wall, S. B. 1980. Di spersal of limber pine seed by Nutcracker. Journal of Forestry. 78 (October): 637 639. Liston, A.; Robinson, W. A.; Pinero, D. ; Alvarez Buylla, E. R. 1999. Phylogenetics of Pinus (Pinaceae) based on nuclear ribosomal DNA internal transcribed spacer region sequences. Molecula r Phylogenetics and Evolution. 11(1): 95 109. Logan, J. A.; Powell, J. A. 2 001. Ghost fo rests, global warming, and the mountain pine beetle (Coleoptera: Scolytidae). American Entomologist. 47(3): 160 172. Logan, J. A.; Regniere, J.; Powell, J. A. 2003. Assessing the impacts of global warming on forest pest dynamics Frontiers in Ecology and the Environment. 1(3): 130 137. Lorenz, T J.; Sullivan, K. in the Cascade Range. The Condor. 11(2): 326 340. Lorenz, T. J.; Sullivan, K. A.; Bakian, A. V.; Aubry, C. A. 2011. Cache site selection in Nucifraga columbiana ). The Auk. 128(2): 237 247. Lynch, H. J.; Renkin, R. A.; Crabtree, R. L. ; Moorcraft, P. R. 2006. The influence of previous mountain pine beetle ( Dendroctonus ponderosae ) activity on the 1988 Yellowstone fires. Ecosystems. 9: 1318 1327. MacKenzie, D. I.; Nichols, J. D.; Lachman, G. B.; Droege, S.; Royle, J. A.; Langtimm, C. A. 2002. Estimating site occupancy rates when de tection probabilities are less than one. Ecology. 83(8): 2248 2255. McCaug hey, W. W.; Tomback, D. F. 2001. The natural regeneration process. In: Tomback, D. F.; Arno, S. F.; Keane, R. E.; eds. Whitebark pine communities: ecology and restoration. Washi ngton, DC: Island Press. 105 120 McDonald, G. I.; Hoff, R. J. 2001. Blister r ust: an introduced plague. In: Tomback, D. F.; Arno, S. F.; Keane, R. E.; eds. Whitebark pine communities: ecology and restoratio n. Washington, DC: Island Press. 263 284. McKinney, S. T.; Tomback, D. F. 2007. The influence of white pine blister rust on see d dispersal in whitebark pine. Canadian Journal of Forest Research. 37: 1044 1057. McKinney, S. T.; Fiedler, C. E.; Tomback, D. F. 2009. Invasive pathogen threatens bird pine mutualism: implications for sustaining a high elevation ecosystem. Ecological App lications. 19(3): 597 607. 23. Moilanen, A.; Hanksi, I.; Smith, A. T. 1998. Long term dynamics in a metapopulation of the American pika. The American Naturalist. 152(4): 530 542. cones. Western North American Naturalist. 70(3): 413 414.

PAGE 74

67 Perkins, D. L.; Swetnam, T. W. 1996. A dendroecological assessment of whitebark pine in the Sawtooth Salmon River reg ion Idaho. Canadian Journal of Forest Research. 26: 2123 2133. Petrzelova, I.; Lebeda, A. 2008. Distribution of race specific resistance against Bremia lactucae in natural populations of Lactuca serriola European Journal of Plant Pathology. 129(2): 233 253. Raffa, K. F.; Aukema, B. H.; Bentz, B. J.; Carroll, A. L.; Hic ke, J. A.; Turner, M. G.; Romme, W. H. 2008. Cross scale drivers of natural disturbances prone to anthropogenic amplificatio n: the dynamics of bark beetle eruptions. BioScience. 58(6): 501 517. Rebertus, A. J.; Burns, B. R.; Veblen, T. T. 1991. Stand dynamics of Pinus flexilis dominated subalpine forests in the Colorado Front Range. Journal of Vegetation Science. 2: 445 458. Romme, W. H.; Knight, D. H.; Yavitt, J. B. 1986. Mountain pine beetle outbreaks in the Rocky Mountain s : regulators of primary productivity? American Naturalist. 127( 4 ) : 484 494. Samano, S.; Tomback, D. F. 2003. Cone opening phenology, seed dispersal, and seed predation in southwestern white pi ne ( Pinus strobiformis ) in southern Colorado. Ecoscience 10:319 326. Schoettle, A. W.; Rochelle, S. G. 2000. Morphological variation of Pinus flexilis (Pinaceae), a bird dispersed pine, across a range of elevations. American Journal of botany. 87: 1797 180 6. Schoettle, A. W.; Goodrich, B. A.; Klutsch, J. G.; Burns, K. S.; Costello, S.; Sniezko, R. A.; Connor, J. 2011. The proactive strategy for sustaining five needle pine populations: An example of its implementation in the southern Rocky Mountains. In: Kea ne, R. E.; Tomback, D. F.; Murray, M. P.; Smith, C. M., eds. 2011. The future of high elevation, five needle white pines in Western North America: Proceedings of the High Five Symposium. 2010 28 30 June, Missoula, MT. Proceedings RMRS P 63. Fort Collins, CO. U.S. Department of Agriculture, Forest Service, Rocky Mountain Research Station: 376. http://www.fs.fed.us/rm/pubs/rmrs_p063.html Schwandt, J. W.; Lockman, I. B.; Kliejunas, J. T.; Muir, J. A. 2010. Current health issues and management strategies for white pines in the western United States and Canada. Forest Pathology. 40: 226 250. Singer, F. J.; Bleich, V. C.; Gudorf, M. A. Restor ation of bighorn sheep metapopulations in and near western national parks. Restoration Ecology. 8(4S): 14 24. Syring, J.; Farrell, K.; Businsky, R. ; Cronn, R. ; Liston A 2007. Widespread genealogical nonmonop h yly in species of Pinus subgenus Strobus Sys tematic Biology. 56(2): 163 181.

PAGE 75

68 Thomas, L.; Laake, J. L.; Derry, J. F.; Buckland, S. T.; Borc hers, D. L.; Anderson, D. R.; Burnham, K. P.; Strindberg, S.; Hedley, S. L.; Marques, F. F. C.; Pollard, J. H.; Fewster, R. M. 1998. Distance 3.5. Research Unit for Wildlife Population Assessment, University of St. Andrews, St. Andrews, UK. Tomback, D. 161. Tomback, D. F. 1983. Nutcrackers and pines: coevolution or coadaptation? Coevolution. Univ. of Chicago Press, Chicago, Illinois. Tomback, D. Nucifraga columbiana ). In: A. Poole, ed. The Birds of North America Online, Species Account Number 331. Cornell Laboratory of Ornithology, Ithaca, NY. Tombac S. F.; Keane, R. E. eds. Whitebark Pine Communities: Ecology and Restoration. Washington, DC : Island Press. 89 104. Tomback, D. F.; Kramer, K. A. 1980. Limbe r pine seed ha Nutcracker in the Sierra Nevada: Timing and foraging behavior. The Condor. 82: 467 468. Tomback, D. F.; Taylor, C. in Rocky Mountain Na tional Park. In: Singer, F. J., ed. To ward the year 2000 Conference on science in the National Parks; Vol. 2 Wildlife management and habitats. Fort Collins, CO: George Wright Society and the U.S. National Park Service: 158 172. Tomback, D. F.; Linhart Y. B. 1990. The evolution of bird dispersed pines. Evolutionary Ecology. 4: 185 219. Tomback, D. F.; Kendall, K. C. 2001. Biodiversity losses: The downward spiral. In: Tomback, D. F.; Arno, S. F.; Keane, R. E. eds. Whitebark Pine Communities: Ecology and Res toration. Washington, DC : Island Press. 243 262. Tomback, D. F.; Achuff, P. 2010. Blister rust and western forest biodiversity: Ecology, values and outlook for white pines. Forest Pathology. 40: 186 225. Tomback, D. F.; Schoettle, A. S.; Chevalier, K. E.; Jones C. A. 2005. Life on the edge for limber pine seed dispersal within a peripher al population. Ecoscience. 12: 519 529. Tomback, D. F.; Achuff, P.; Schoettle, A. W.; Schwa ndt, J. W.; Mastrogiuseppe, R. J. 2011. The magnificent high elevation five needl e white pines: Ecological roles and f uture outlook. In: Keane, R. E.; Tomback, D. F.; Murray, M. P.; Smith, C. M., eds. 2011. The future of high elevation, five needle white pines in Western North America: Proceedings of the High Five Symposium. 2010 28 30 June, Missoula, MT. Proceedings RMRS P 63. Fort Collins, CO. U.S. Department of Agriculture, Forest Service, Rocky Mountain Research Station: 376. http://www.fs.fed.us/rm/pubs/rmrs_p06 3.html

PAGE 76

69 Torick, L. L.; Tomback, D. F.; Espinoza, R. 1996. Occurrence of multi genet tree clusters in pines. American Midland Naturalist. 136(2): 262. Turner, T. J.; Tomback, D. F.; Van Anderson, B.; Mu rray, M. 2011. Sugar pine seed harvest n Crater Lake National Park, Oregon. In: Keane, Robert E.; Tomback, Diana F.; Murray, Michael P.; Smith, Cyndi M., eds. 2011. The future of high elevation, five needle white pines in Western North America: Proceedings of the High Five Symposium. 2010 28 30 June, Missoula, MT. Proceedings RMRS P 63. Fort Collin s, CO. U.S. Department of Agriculture, Forest Service, Rocky Mountain Research Station: 376. http://www.fs.fed.us/rm/pubs/rmrs_p06 3.html Vander Wall, S. acker on rapidly changing pine seed resources. The Condor. 90: 621 631. Vander Wall, S. B.; Balda, R. B. 1977. Coadaptat and the pinyon pine for efficient seed harvest an d dispersal. Ecol. Monogr. 47: 89 111. Vander Wall, S. B.; H utchins, H. E. 1983. Dep Nucifraga columbiana on conifer seeds during postfledging period. Canadian Field Naturalist. 97: 208 214. Vander Wall, S B.; Hoffman, S. W.; Potts, W. K 1981. Emigration behavior of Nutcracker. The Condor. 83: 162 170. Veblen, T T. 1986. Age and size structure of subalpine forests in the Colorado Front Range. Bulletin of the Torrey Botanical Club. 113: 225 240. Webster, K. L.; Johnson, E. A. 2000. The importance of regional dynamics in local populations of limber pine ( Pinus flexilis ). Ecoscience. 7(2): 175 182.