University of Montana
ScholarWorks at University of Montana
Theses, Dissertations, Professional Papers
Graduate School
2015
The science behind the selection of native plant
materials: local adaptation, response to invasion,
cytotypic variation, and seed transfer zones
Alexis Gibson
The University of Montana
Follow this and additional works at: http://scholarworks.umt.edu/etd
Recommended Citation
Gibson, Alexis, "The science behind the selection of native plant materials: local adaptation, response to invasion, cytotypic variation,
and seed transfer zones" (2015). Theses, Dissertations, Professional Papers. Paper 4588.
This Dissertation is brought to you for free and open access by the Graduate School at ScholarWorks at University of Montana. It has been accepted for
inclusion in Theses, Dissertations, Professional Papers by an authorized administrator of ScholarWorks at University of Montana. For more
information, please contact [email protected].
THE SCIENCE BEHIND THE SELECTION OF NATIVE PLANT MATERIALS:
LOCAL ADAPTATION, RESPONSE TO INVASION, CYTOTYPIC VARIATION,
AND SEED TRANSFER ZONES
By
ALEXIS LEIGH GIBSON
B.A., Biology, University of Montana, Missoula 2007
Dissertation
presented in partial fulfillment of the requirements
for the degree of
PhD
in Forestry
The University of Montana
Missoula, MT
May 2015
To be approved by:
Sandy Ross, Dean of the Graduate School
Graduate School
Cara R. Nelson, Chair
Department of Ecosystem and Conservation Sciences
Ragan M. Callaway
Division of Biological Sciences
Elizabeth Crone
Department of Biology, Tufts University
Lila Fishman
Division of Biological Sciences
Solomon Z. Dobrowski
Department of Forest Management
Dean E. Pearson
Rocky Mountain Research Station, USDA Forest Service
ABSTRACT
Gibson, Alexis L., Ph.D., 2015
Forestry
THE SCIENCE BEHIND THE SELECTION OF NATIVE PLANT MATERIALS:
LOCAL ADAPTATION, RESPONSE TO INVASION, CYTOTYPIC VARIATION,
AND SEED TRANSFER ZONES
Chairperson: Dr. Cara R. Nelson
The use of locally adapted native plant materials is promoted as a mechanism for
improving revegetation and restoration outcomes and increasing the long-term ecological
function of a site. Challenges arise in the use local native plant materials, however, when
the conditions in which they are used differ from those to which they are adapted, like in
heavily disturbed sites undergoing restoration or invaded ecosystems, or when local
adaptation is not the only genetic consideration for fitness (polyploidy). In addition,
managers need to select among multiple guidelines for how far to move seeds, each of
which comes with its own assumptions about the scale of factors that drive genetic
differentiation. In my dissertation, I explore key ecological and genetic considerations for
improving the use of native plant materials.
In the first chapter, I assessed the current body of research on local adaptation with
respect to the use of six methodological variables that could affect application to
management. I found that the majority of local adaptation research occurred in conditions
that were not comparable to field revegetation conditions, and that a limited number of
experiments used direct fitness measures. Additionally, research has focused primarily on
single abiotic selective factors, ran for year or less, and used only a few populations.
Taken as a whole, it would be difficult to apply research findings to native plant materials
choices. In the second chapter, I investigated whether invasion by Centaurea stoebe has
resulted in shifts in traits or competitive ability of a native grass, Pseudoroegneria
spicata, and the relationship between a population’s tolerance and suppression of C.
stoebe. While I found differences between invader-experienced and invader-naïve
populations of P. spicata, the effect of population and site appears to be more important
for trait differences. I also found that a population’s tolerance of C. stoebe was not related
to its ability to suppress the invader. In my third chapter, I explored how trait differences
between cytotypes and polyploidy frequency in the northern range of P. spicata could
complicate seed transfer. I found that polyploidy was common and that cytotypes were
distributed across the landscape. The distribution of cytotypes and the lack of consistent
trait differences between them make it likely that attempts to move P. spicata in this
region will result in cytotypic mixing, which could result in triploid formation and reduce
population fitness. Finally, in my fourth chapter, I explored the applicability of four
provisional seed transfer zone guidelines for P. spicata. Populations were not
significantly different at local scales, but trait variation was not well-explanted at the
ecoregion scale either. Instead, I found that seed transfer for P. spicata in its northern
range should be guided by climate within Level III Ecoregion. Taken as a whole, my
dissertation supports the need for more integration of ecology into ecosystem
management and restoration.
i
ACKNOWLEDGEMENTS
I’d like to thank my committee for their feedback on my research, as well as the
education you gave me in the classroom and lab. I’m a better thinker and scientist
because of you, and this work has been improved through your comments and ideas. I’d
also like to acknowledge Susan Reinhart from the Forest Service Region 1 Native Plant
Materials program. Susan provided the funding for the majority of this work, logistical
support and materials, and was always an enthusiastic audience for my results.
I cannot overstate my gratitude for my advisor, Dr. Cara Nelson, who kept me
going through this long endeavor with encouragement and humor. You were there to
support me when I was unsure, to celebrate my good news, and always encouraged me to
follow my ideas.
The Nelson lab members (current and former) made it fun to come to work, and
were always there when I needed another set of eyes or ears. I’d especially like to thank
Megan Keville and Kali Becher for assistance in and out of the field. Beth Miller also
deserves thanks for her years of long-distance help with data interpretation, statistics
support, editing, and her general bonhomie.
This dissertation would not exist without the love and support of my friends and
family. You kept me sane, distracted me, fed me, poured me drinks when the day was
terrible, and gave me confidence with your unwavering belief that I could finish. My
mom, Anita, deserves special thanks in this regard.
Finally, I’d like to dedicate this work to my husband, Bryan, and my son,
Sullivan. You remind me that good things are always on the way.
ii
TABLE OF CONTENTS
Introduction
iv
Chapter 1. Can local adaptation research in plants inform selection of native plant
materials? An analysis of experimental methodologies
1
Chapter 2. Response of bluebunch wheatgrass to invasion: competitive ability of
invader-experienced and naïve populations
38
Chapter 3. The missing link in the conversation about native plant materials:
polyploidy in a commonly seeded native grass
80
Chapter 4. Comparing provisional seed transfer zone strategies for a commonly
seeded grass, Pseudoroegneria spicata
118
iii
INTRODUCTION
Since Turesson’s (1922) original description of ecotypes as populations showing
heritable morphological, phenological, and physiological differences in response to their
environment, investigators have demonstrated that populations can be adapted to a
variety of site-specific selective pressures (e.g., Linhart and Grant 1996, Joshi et al.
2001). Knowledge of the importance of these local selective pressures has led to
increased interest in and requirement for using local ecotypes for revegetation and
restoration (USDI and USDA 2002, Hufford and Mazer 2003, Johnson et al. 2010).
While these requirements aim to improve restoration outcomes and long-term population
persistence on a site, they may not provide the best outcome in all situations: there is
limited research on the scope and scale of local adaption that is applicable to revegetation
and restoration, invasive plant species can alter the conditions populations are adapted to,
the focus on seed transfer over broad areas may neglect population-level genetic
considerations like polyploidy, and assumptions about “how local is local?” may not
match the scale of population differentiation. In this dissertation, I explore some of the
practical implications of prioritizing the use of locally collected seeds in order to improve
use of native plant materials by better integrating ecology and genetics for land
management.
In chapter 1, I found that the current body of local adaptation research (LAR) has
methodological limitations that could limit its application to revegetation and other types
of management. While previous reviews have assessed the frequency of local adaption
(Leimu and Fischer 2008, Hereford 2009) or which methods should be used to research
its occurrence (Kawecki and Ebert 2004, Kawecki et al. 2012, Blanquart et al. 2013), I
iv
assessed the use of six key experimental methodologies in LAR. My results indicate that
LAR results should be cautiously applied to land management, due to methodological
choices that make research conditions dissimilar or potentially uninformative to field
conditions and the long-term goal of viable restored populations. These findings suggest
the need for scientists and managers to collaborate on LAR that more directly answers
questions about landscape-scale use of native plant materials.
In chapter 2, I explored whether invasion can cause shifts in population traits or
competitive ability by studying the effect of invasion by Centaurea stoebe (spotted
knapweed) on Pseudoroegneria spicata (bluebunch wheatgrass, a native grass). I found
evidence that populations from the same invader-experience type show significant
variation in traits across the Missoula Valley, supporting previous research on variation
in response to novel community members within and across landscapes (Thompson et al.
2002, Goergen et al. 2011). Additionally, I found support for the ability of native
populations to respond to invasion through shifts in traits and competitive ability
(Callaway et al. 2005, Lau et al. 2008, Leger 2008). Although adult plants from invaded
areas were better at tolerating competition from C. stoebe than those from uninvaded
areas, a population’s tolerance was unrelated to its ability to suppress the invader.
Variation among ecotypes in the ability to suppress or tolerate competition could have
long-term impacts on ecotypic richness in plant communities (Atwater 2012), and my
results highlight the importance of considering response to novel selective pressures
when determining the best seed sources for restoration.
In chapter 3, I used P. spicata as a case study for why ploidy should be directly
considered when selecting seed sources for revegetation and land management. The
v
presence of multiple cytotypes in a region presents challenges for the use of local seed
(Jones 2003, McKay et al. 2005), and yet polyploidy is rarely included in revegetation or
restoration plans (Delaney 2012, Severns et al. 2013, Mutegi et al. 2014). I found that
polyploidy is common in the northern range of P. spicata, with tetraploids occurring in
56% of surveyed populations. Because ecoregion was a more important driver of traits
than was cytotype, it would be difficult or impossible to identify polyploids through
phenotype and cytotypic mixing is likely to occur unless ploidy is an explicit revegetation
consideration. Mixing cytotypes can result in reduced population fertility (Ramsey and
Schemske 1998, Burton and Husband 2001) and consequently decreased population
fitness (Burton and Husband 2000). Level III Ecoregions are commonly suggested to
guide seed movement; I found that tetraploid and mixed-cytotype populations were not
segregated by EPA Level III Ecoregions, which would limit their potential use as
provisional seed transfer zones. Overall, my results indicate that conventional methods
for determining what constitutes “local” for revegetation will not prevent cytotypic
mixing in P. spicata, and supports the increased use of flow cytometry as a simple
method for managers to determine the ploidy of native plant materials.
In chapter 4, I assessed which options for provisional seed transfer zones best
predict trait variation and performance of P. spicata in the Northern Rockies. Research
on local adaptation cautions against combining dissimilar populations (Hufford and
Mazer 2003), but revegetation usually requires moving and combining seeds from
multiple populations. Managers rely on generalized provisional seed transfer guidelines
to determine how far to move seeds. Transfer guidelines need to balance the risk of
adding maladapted ecotypes (Waser and Price 1994, Montalvo and Ellstrand 2001) with
vi
the positive population- and community-level impacts of genetic diversity (Edmands
2007, Broadhurst et al. 2008, Havens et al. 2015). Currently, provisional transfer
guidelines range from limiting seed collection to populations immediately around the
restoration site to moving seeds within broad ecological or climatic areas. I found that
variation in P. spicata traits is driven by large-scale environmental differences suggesting
that seed transfer should occur at landscape, rather than population, scales. Of the four
models for provisional seed transfer zone that I tested, models that included climate
within ecoregion were most predictive of phenotypic traits, mortality and seed set. The
ecoregion-only model ranked low in model comparisons, suggesting that Level III
Ecoregions by themselves are likely too broad to use for provisional seed transfer zones.
Results from my common garden study also support this conclusion: seeds from the
home ecoregion did not perform better than seeds from foreign ecoregions when grown
together. Although provisional seed transfer zones are intended to be used in the absence
of species-specific data, my findings highlight the risk of using untested transfer
guidelines.
As a whole, my dissertation supports the need for more integration of ecology,
genetics and land management. While requirements for the use of locally adapted seeds
stem from a desire to maintain local genetics and increase revegetation and restoration
success, my results highlight the need to continue to redefine what we think of as
genetically appropriate for managing ecosystems.
LITERATURE CITED
vii
Atwater, D. 2012. Interplay between competition and evolution in invaded and native
plant communities. Theses, Dissertations, Professional Papers. Paper 4156.
University of Montana, Missoula, MT.
Blanquart, F., O. Kaltz, S. L. Nuismer, and S. Gandon. 2013. A practical guide to
measuring local adaptation. Ecology Letters 16:1195-1205.
Broadhurst, L. M., A. Lowe, D. J. Coates, S. A. Cunningham, M. McDonald, P. A. Vesk,
and C. Yates. 2008. Seed supply for broadscale restoration: maximizing
evolutionary potential. Evolutionary Applications 1:587-597.
Burton, T. L., and B. C. Husband. 2000. Fitness differences among diploids, tetraploids,
and their triploid progeny in Chamerion angustifolium: Mechanisms of inviability
and implications for polyploid evolution. Evolution 54:1182-1191.
Burton, T. L., and B. C. Husband. 2001. Fecundity and offspring ploidy in matings
among diploid, triploid and tetraploid Chamerion angustifolium (Onagraceae):
consequences for tetraploid establishment. Heredity 87:573-582.
Callaway, R. M., W. M. Ridenour, T. Laboski, T. Weir, and J. M. Vivanco. 2005. Natural
selection for resistance to the allelopathic effects of invasive plants. Journal of
Ecology 93:576-583.
Delaney, J. T. 2012. Intraspecific chromosome number variation and prairie restoration—
a case study in northeast Iowa, U.S.A. Restoration Ecology 20:576-583.
Edmands, S. 2007. Between a rock and a hard place: evaluating the relative risks of
inbreeding and outbreeding for conservation and management. Molecular
Ecology 16:463-475.
viii
Goergen, E. M., E. A. Leger, and E. K. Espeland. 2011. Native perennial grasses show
evolutionary response to Bromus tectorum (cheatgrass) invasion. Plos One 6:
e18145. doi: 10.1371/journal.pone.0018145.
Havens, K., P. V. Vitt, S. Still, A. T. Kramer, J. B. Fant, and K. Schatz. 2015. Seed
sourcing for restoration in an era of climate change. Natural Areas Journal
35:122-133.
Hereford, J. 2009. A quantitative survey of local adaptation and fitness trade-offs.
American Naturalist 173:579-588.
Hufford, K. M., and S. J. Mazer. 2003. Plant ecotypes: genetic differentiation in the age
of ecological restoration. Trends in Ecology & Evolution 18:147-155.
Johnson, R., L. Stritch, P. Olwell, S. Lambert, M. E. Horning, and R. Cronn. 2010. What
are the best seed sources for ecosystem restoration on BLM and USFS lands?
Native Plants Journal 11:117-131.
Jones, T. A. 2003. The restoration gene pool concept: beyond the native versus nonnative debate. Restoration Ecology 11:281-290.
Joshi, J., B. Schmid, M. C. Caldeira, P. G. Dimitrakopoulos, J. Good, R. Harris, A.
Hector, K. Huss-Danell, A. Jumpponen, A. Minns, C. P. H. Mulder, J. S. Pereira,
A. Prinz, M. Scherer-Lorenzen, A. S. D. Siamantziouras, A. C. Terry, A. Y.
Troumbis, and J. H. Lawton. 2001. Local adaptation enhances performance of
common plant species. Ecology Letters 4:536-544.
Kawecki, T. J., and D. Ebert. 2004. Conceptual issues in local adaptation. Ecology
Letters 7:1225-1241.
ix
Kawecki, T. J., R. E. Lenski, D. Ebert, B. Hollis, I. Olivieri, and M. C. Whitlock. 2012.
Experimental evolution. Trends in Ecology & Evolution 27:547-560.
Larson, S. R., T. A. Jones, and K. B. Jensen. 2004. Population structure in
Pseudoroegneria spicata (Poaceae : Triticeae) modeled by Bayesian clustering of
AFLP genotypes. American Journal of Botany 91:1789-1801.
Lau, J. A., A. C. McCall, K. F. Davies, J. K. McKay, and J. W. Wright. 2008. Herbivores
and edaphic factors constrain the realized niche of a native plant. Ecology 89:754762.
Leger, E. A. 2008. The adaptive value of remnant native plants in invaded communities:
an example from the Great Basin. Ecological Applications 18:1226-1235.
Leimu, R., and M. Fischer. 2008. A meta-analysis of local adaptation in plants. Plos One
3: e4010. doi: 10.1371/journal.pone.0004010.
Linhart, Y. B., and M. C. Grant. 1996. Evolutionary significance of local genetic
differentiation in plants. Annual Review of Ecology and Systematics 27:237-277.
McKay, J. K., C. E. Christian, S. Harrison, and K. J. Rice. 2005. "How local is local?" A review of practical and conceptual issues in the genetics of restoration.
Restoration Ecology 13:432-440.
Montalvo, A. M., and N. C. Ellstrand. 2001. Nonlocal transplantation and outbreeding
depression in the subshrub Lotus scoparius (Fabaceae). American Journal of
Botany 88:258-269.
Mutegi, E., A. L. Stottlemyer, A. A. Snow, and P. M. Sweeney. 2014. Genetic structure
of remnant populations and cultivars of switchgrass (Panicum virgatum) in the
context of prairie conservation and restoration. Restoration Ecology 22:223-231.
x
Ramsey, J., and D. W. Schemske. 1998. Pathways, mechanisms, and rates of polyploid
formation in flowering plants. Annual Review of Ecology and Systematics
29:467-501.
Severns, P. M., E. Bradford, and A. Liston. 2013. Whole genome duplication in a
threatened grassland plant and the efficacy of seed transfer zones. Diversity and
Distributions 19:455-464.
Thompson, J. N., S. L. Nuismer, and R. Gomulkiewicz. 2002. Coevolution and
maladaptation. Integrative and Comparative Biology 42:381-387.
USDI, and USDA. 2002. Report to the Congress: interagency program to supply and
manage native plant materials for restoration and rehabilitation on federal lands.
Washington, DC, USA.
Waser, N. M., and M. V. Price. 1994. Crossing distance effects in Delphinium nelsonii outbreeding and inbreeding depression in progeny fitness. Evolution 48:842-852.
xi
Gibson
PhD Dissertation: Chapter 1
2015
CHAPTER 1
CAN LOCAL ADAPTATION RESEARCH IN PLANTS INFORM SELECTION OF
NATIVE PLANT MATERIALS? AN ANALYSIS OF EXPERIMENTAL
METHODOLOGIES
1
Gibson
PhD Dissertation: Chapter 1
2015
ABSTRACT
Local adaptation is a key process in ecology and evolution. It is increasingly used in
ecological restoration and land management as a criterion to select plant materials that
will display highest fitness in the new environment. A large body of research literature
has explored local adaptation in plants; however, to what extent its findings can inform
management decisions has not been formally evaluated. I assessed local adaptation
literature (308 experiments) for six key experimental methodologies that have the
greatest effect on the application of research to native plant materials choices:
experimental environment, response variables, maternal effects, intraspecific variation,
selective agents, and spatial and temporal variability. I found that less than half of
experiments used reciprocal transplants (39%) or natural field conditions (41%), both of
which are most informative for revegetation and restoration. In addition, population
growth rate, which would determine the persistence and expansion of populations, was
rarely (5%) assessed, and most studies measured only a single generation (96%) and ran
for less than a year. Emergence and establishment are key limiting factors in successful
revegetation and restoration, but the majority of studies measured later life history stages
(66%). In addition, most studies included limited replication at the populations and
habitat level, and tested response to single abiotic selective factors (66%; climatic
variables were most common). Consequently, local adaptation research should be
cautiously applied to management and future research could use alternative
methodologies to increase the ability of managers to directly apply findings.
Keywords: ecological experiments, experimental design, experimental methodology,
local adaptation, lifetime fitness, plants
2
Gibson
PhD Dissertation: Chapter 1
2015
INTRODUCTION
Local adaptation is the process by which resident genotypes exhibit higher fitness
in their home environment compared to non-local genotypes due to divergent selection as
a consequence of variation in environment (Kawecki and Ebert 2004). Over the course of
the 20th century, research on local adaptation has expanded from a primary focus on longterm evolutionary processes, such as speciation (Jordan 1905), to a broader set of issues
including rapid evolutionary processes and responses to changing environmental
conditions (Barrett, Colautti, and Eckert 2008; Leger and Espeland 2010; Hoffmann and
Sgro 2011). Scientists and managers are increasingly using results of local adaptation
research (LAR) to inform complex management decisions (Hufford and Mazer 2003),
such as assisted migration for climate change mitigation (Vitt et al. 2010), and choice of
native plant materials for revegetation and restoration (McKay et al. 2005). For example,
positive findings of adaptation to local selective pressures (Joshi et al. 2001; Leimu and
Fischer 2008; Hereford 2009) have been used as an argument in favor of primarily using
local ecotypes in restoration (USDI and USDA 2002; Johnson, Stritch, et al. 2010;
Vander Mijnsbrugge, Bischoff, and Smith 2010). Native plant materials choices impact
the viability and adaptive potential of restored populations (Williams 2001; Broadhurst,
North, and Young 2006; Aavik et al. 2012), as well as the feasibility of using locally
collected seeds in large-scale restoration (Merritt and Dixon 2011). Because of this, it is
critical to understand the extent to which local adaptation research can be broadly applied
to land management.
Findings of LAR have substantially advanced our understanding of local
adaptation in plants, yet it remains unclear to what extent findings can be extrapolated to
3
Gibson
PhD Dissertation: Chapter 1
2015
land management decisions. Local adaptation research aims to understand evolution at
the population level, while restoration as a whole is conducted at ecosystem and
landscape scale. The ability to apply findings to landscape-scale management depends in
large part on whether experimental methodologies capture selective pressures at relevant
time and spatial scales. In addition, the experimental environment, response variables
selected, and maternal effects all affect the extent to which one can apply LAR to native
plant materials choices.
Experimental environment – The most conclusive method for detecting local
adaptation is through replicated reciprocal transplant experiments that compare fitness in
multiple home and foreign sites (Kawecki and Ebert 2004; Blanquart et al. 2013). Local
adaptation research will be informative for land management if experiments use whole
environments (Nuismer and Gandon 2008), and occur at multiple sites and in
experimental conditions that are similar to those found during revegetation. In contrast,
experiments conducted at single sites, such as common-garden studies, can only show
phenotypic variation among populations, not whether fitness is higher for local versus
non-local populations. Conducting experiments using in situ environments is especially
important when local adaptation is only observed under specific environmental
conditions (cryptic adaptation), such as the presence or absence of native plant
community competitors (Knight and Miller 2004; Bischoff et al. 2006; Rice and Knapp
2008).
Measures of response – From a restoration perspective, lambda (λ), or population
growth, is the most relevant direct fitness measure because it represents long-term
population viability (Menges 1990, Rice and Emery 2003). Simply put, λ must be greater
4
Gibson
PhD Dissertation: Chapter 1
2015
than or equal to one for a population to persist. Plant traits that respond to selection in the
populations’ home sites can be also used to detect evidence of local adaptation, but they
are less likely to be directly related to fitness and may not show a signal for response to
selection. There is also variation among species in which traits are most important for
population fitness, and important traits may not be known for certain species.
Furthermore, ecological restoration benefits from research conducted across multiple life
history stages and generations, as fitness responses can vary across these scales (Donovan
and Ehleringer 1992; Kelly 1992; Rice and Knapp 2008). Given that the majority of
revegetation projects rely on seeds to establish native plants (Koch 2007; Broadhurst et
al. 2008), research that focuses the expression and magnitude of local adaptation during
germination and establishment may provide especially important information for land
management
Maternal effects – Observed phenotypic differences among populations can result
from differences among genotypes (local adaptation) or maternal effects (Roach and
Wulff 1987). Adaptive maternal effects have been found to increase performance of the
progeny of maternal plants exposed to drought (Sultan, Barton, and Wilczek 2009),
herbivory (Agrawal 2001, 2002), herbicide (Bozorgipour and Snape 1997), and shading
(Donohue and Schmitt 1999; Galloway and Etterson 2007; Bell and Galloway 2008) in
these environments. In addition, the effects on phenotype of progeny can persistent for
multiple generations (Miao, Bazzaz, and Primack 1991). For populations that remain in
place in the landscape, maternal effects may strengthen links between phenotypes and
response to selection (Espeland and Rice 2012). In the case of land management,
however, seeds are moved away from the maternal plant environment and expected to
5
Gibson
PhD Dissertation: Chapter 1
2015
show the same traits and performance. When maternal effects drive adaptive plant traits
and when maternal environments (i.e. seed production farms) differ from target
environments, determining whether traits are the result of maternal effects or local
adaption will be critical for predicting seed and plant performance in revegetation.
Among population variability – Just as populations differ in the selective
pressures they experience, they also differ in the magnitude and direction of response to
those pressures (Thompson, Nuismer, and Gomulkiewicz 2002; Leger and Espeland
2010). When planning a restoration, the practitioner calculates the likelihood of
differential genotypic success in the environment. At the coarse scale, this is done by
planting varieties developed for specific environments (i.e. ‘Aztec’ Maxmilian’s
sunflower for wet areas and ‘Prairie Gold’ Maxmilian’s sunflower for dry areas in the
southern Great Plains of the United States; USDA-NRCS 2011, 2013). When differences
between sites are ambiguous, however, the practitioner is confronted with the question of
“how local is local?” (McKay et al. 2005). Assessing the relationship between
environment and inter-population differences requires sampling from many individuals
and populations (Manel et al. 2003), especially if there is significant variation among
populations. In addition, populations may show fitness differences unrelated to local
adaptation due to habitat quality or genetic factors such as inbreeding (Blanquart et al.
2013). When multiple populations collected from many habitats are used, research
findings could enhance the ability of practitioners to make these difficult decisions. The
popularity of genecological studies that measure hundreds of field-collected populations
in common gardens to generate geographic limits of appropriate seed transfer (e.g.,
6
Gibson
PhD Dissertation: Chapter 1
2015
Johnson, Erickson, et al. 2010; St Clair et al. 2013) is evidence that this magnitude of
population sampling may be necessary to assist practitioners in seed selection.
Selective agents – Understanding the factors that drive population differentiation
is important in choosing native plant materials. Plant species can be adapted to both
abiotic conditions (e.g., soil and climate; Macel et al. 2007; Goransson, Andersson, and
Falkengren-Grerup 2009) and biotic factors (e.g., pollinators and soil pathogens;
Svenning, Junttila, and Solheim 1991; Thrall, Burdon, and Bever 2002; Streisfeld and
Kohn 2007), and interactions between factors can alter the observance or strength of local
adaptation (Hufford, Mazer, and Camara 2008; Lau et al. 2008). Understanding the
impact of multiple selective factors on population fitness will not only help managers
identify which factors define “local”, but also provides information about the field
conditions under which higher home-site fitness is observed.
Spatial and temporal variability – Beyond biotic and abiotic factors that are
largely consistent across years, temporal and environmental variation can also be agents
of selection. For example, selective agents that drive local adaptation may only act on
some generations of the target species (Rice and Mack 1991; Geber and Griffen 2003;
Thompson et al. 2007), and impacts on non-local sources may not be apparent for
decades (Millar and Libby 1989). Space is often used as a proxy for time in ecological
experiments (Haubensak and Parker 2004) because temporally rare events required for
the expression of local adaptation – such as disease or drought – are more likely to be
captured when multiple sites are used. Therefore, the number of environments and the
type of variation encompassed within LAR (either by conducting an experiment over
multiple experiment years or using many sites) is important for assessing the constancy
7
Gibson
PhD Dissertation: Chapter 1
2015
of the expression of local adaptation and the comparative risk of using non-local
genotypes.
To date, reviews of LAR have focused on identifying the overall frequency and
drivers of local adaptation (Leimu and Fischer 2008; Hereford 2009, 2010) or on best
practices for researching local adaptation (Kawecki and Ebert 2004; Kawecki et al. 2012;
Blanquart et al. 2013). There is an additional need to assess the extent to which existing
LAR can inform decisions regarding genetically-appropriate plant materials for land
management; these decisions require an understanding of how selection across the
landscape shapes plant traits that are most important for restoration establishment and
long-term success. I conducted a literature review in order to quantify to what extent
LAR has integrated six key methodological considerations and can guide choices of
native plant materials for management.
MATERIALS AND METHODS
I performed a literature search in ISI Web of Science using the search terms “local
adapt*” and “plant*”, for the period of 1965 to February 2013. A total of 1,046 studies
were identified. I reviewed titles, abstracts, and keywords of each article to determine
suitability for inclusion and excluded studies that: did not focus on local adaptation in
vascular plants (439 studies), had primary species of interest that were nonnative invasive
species (113 studies), used only molecular analysis (93 studies), focused on crop plant(s)
(42 studies), or were not experimental (e.g., theoretical, modeling, and review papers;
124 studies). If a study was comprised of multiple experiments, I recorded data on each
experiment individually. The final analysis comprised 234 articles describing 308
experiments. The experiments tested for local adaptation in 278 different plant species,
8
Gibson
PhD Dissertation: Chapter 1
2015
mostly forbs (69%) and graminoids (20%) and, to a lesser degree, trees (9%) and shrubs
(2%). Of the non-tree species, 74% were perennial and 26% were annual.
For each experiment, I assessed six methodological variables that are relevant for
ecological restoration: experimental environment, measures of response, maternal effects,
among population variability, selective agents, and spatial and temporal variability. I
recorded components of the experimental environment (type of experiment, site type,
inclusion of the home plant community) as well as the response variables analyzed (the
life stages studied, whether data was collected over the plant’s entire lifespan, and
whether multiple generations were studied). To classify the extent to which experiments
controlled maternal effects, I recorded whether plant materials used in each study were
the result of collections from a controlled environment, or if authors accounted for
maternal effects by using early stage measurements (initial seed weight or initial plant
size) as covariates in statistical analysis. I also recorded the number of different habitat
types that populations were collected from as reported by authors (e.g., grassland and
dune sites, inland and coastal sites) and the number of different populations from which
plant material was collected (defined by authors). I identified the type and number of
agents of natural selection that were tested within each experiment (biotic interactions
and abiotic factors). To determine the spatial and temporal variability captured in
experimental design, I recorded the duration of each experiment (rounded to the nearest
year), the number of environments that were used in studies that were done in unmanipulated field conditions, or the number of experimental conditions tested if
investigators used treatments to create multiple experimental environments.
RESULTS
9
Gibson
PhD Dissertation: Chapter 1
2015
Experimental environment
Thirty-nine percent of experiments used reciprocal field transplants among the
populations’ home sites, whereas 33% used common garden designs (Table 1). Roughly
half (N = 55) of the common garden experiments were conducted at a single site.
Greenhouse and growth chamber experiments were the least frequently used (28%, N =
87). Approximately equal numbers of experiments were performed in natural sites (41%,
N = 125) as in artificial settings (pots, greenhouses and growth chambers; Table 1).
Sixty-eight percent of experiments (N = 208; Table 1) removed local vegetation from the
experimental environment.
Response variables
Although 82% of experiments calculated a measure of fitness, only 5% (N = 14)
included λ as a response variable (Table 1). Biomass was the most frequently used
measure of fitness (59%), followed by reproductive success (44%). The most common
life history stage assessed was non-reproductive, followed by reproductive adult (Table
1); germination was the least commonly tracked (26%; Table 1). Forty-one percent of
experiments tracked two life stages, and 13% tracked plants across all three life stages
(Figure 1). The majority of studies did not follow plants until death (77%, N = 244; Table
1) or track multiple generations (96%, N = 296; Table 1).
Maternal effects
Approximately three quarters of experiments controlled for maternal effects in
some way. However, most of these (45%, N = 138) used initial plant size or seed weight
as a statistical covariate, or kept maternal families separate in statistical analysis (Table
10
Gibson
PhD Dissertation: Chapter 1
2015
1). Only a third (29%) included plant material that had been grown in a controlled
maternal environment
Among-population variability
I found wide variation in the number of collection populations and habitats
(Figure 2). On average, experiments used plant materials collected from eight populations
and three different habitat types.
Selective agents
The majority of experiments tested adaptation to abiotic factors (89%, N = 271).
Biotic factors were rarely considered (25%, N = 76), and less than one percent (N = 7)
assessed adaptation in the presence of multiple biotic factors (Table 1). The majority of
studies that tested abiotic factors focused on climate (Table 1). Additional factors were
overall ecological and geographic differences between populations, salt-spray tolerance,
and inundation gradients (Table 1). Ten percent (N = 27) of studies tested adaptation to
multiple abiotic factors or abiotic and biotic factors in combination (14%, N = 42).
Spatial and temporal variability
On average, experiments ran for two years, with the median being less than one
year (Figure 3). The longest running experiment lasted 45 years (Gomory et al. 2012).
There was a wide range of variability in the number of environments experiments
occurred in (sites or environmental conditions if a greenhouse or common garden study;
mean = 4, median = 3; Table 1).
DISCUSSION
11
Gibson
PhD Dissertation: Chapter 1
2015
Practitioners have increasingly used results from LAR to guide management
decisions (e.g., Vander Mijnsbrugge, Bischoff, and Smith 2010). However, my results
suggest that findings from LAR are not easily transferable to land management due to
experimental constraints. In particular, LAR primarily used experimental environments
that did not mimic natural conditions, chose response variables that did not reflect
lifetime fitness, excluded biotic and multiple selective factors, and used limited
replication and experimental duration. While these methodological choices do not reflect
the quality or findings of individual experiments designed to test specific factors of
interest, practitioners should interpret such results from LAR with caution.
As with previous reviews (Leimu and Fischer 2008; Hereford 2009), I found that
only a portion of LAR directly addresses local adaptation through the use of reciprocal
transplant studies (39%). Not only are reciprocal transplant studies required to test for
local adaptation, but observed differences among populations can differ from common
garden studies (Table 2). Perhaps more importantly for applying LAR, less than half of
studies occurred in natural environments (41%), or retained the native plant community
(25%). Although removing confounding factors like natural site variation and the home
plant community can make it easier to study factors of interest, it impacts both
probability of detection and whether findings are relevant in situ (McCarragher,
Goldblum, and Rigg 2011; Ehlers et al. 2012; Pankova, Raabova, and Munzbergova
2014; Table 2). Furthermore, the choice of traits or inclusion of λ in LAR is relevant for
whether findings of higher fitness translate to increased population persistence, and the
two may give contradictory results that alter whether local adaptation is observed (Table
2). I found that only 40% of LAR used direct fitness measures (either survival or
12
Gibson
PhD Dissertation: Chapter 1
2015
reproductive success) and very few (5%) used λ. Incorporating multiple life stages
increases the applicability of LAR to management since the use of local native plant
materials is often predicated on the assumption that local adaptation will increase
population fitness at critical life stages, yet local populations may not show consistent
trends of higher fitness across their entire life cycle (Table 2). Germination and
emergence are often the limiting factor in revegetation success (Khurana and Singh 2001;
Pywell et al. 2003) and critical to population regulation (Horvitz and Schemske 1995;
Freville and Silvertown 2005), however less than a quarter of experiments incorporated
these life stages as a measure of fitness.
Reciprocal transplants and direct fitness measures are just two of the important
experimental considerations for applying LAR to restoration; given the expense of using
local seeds, managers need to be confident that local sources will result in long-term
increased fitness in restored populations. Replication over space and time, and the
inclusion of relevant selective agents are equally important, but rarely adequately
addressed. Thus, it is unknown whether findings of local adaptation are due to fitness
differences in response to selective agents or simple population differentiation, and it
could be additionally difficult to determine if local seeds will be consistent in showing
higher fitness under altered site conditions (Table 2). The limited number of habitats
plant materials were collected from increases the risk that LAR has selectively used
populations from a few highly contrasting environments thereby increasing the chance of
finding fitness differences (Hereford and Winn 2008; Hereford 2009; Table 2), directly
limiting the application of LAR to decisions regarding the scale and importance of local
adaptation in choosing plant materials. In addition, the strength, direction, and sources of
13
Gibson
PhD Dissertation: Chapter 1
2015
selection frequently change among years (Siepielski, DiBattista, and Carlson 2009) – the
short duration and limited testing conditions of most LAR indicates that even normal
variation at experimental sites is unlikely to be captured. The magnitude of fitness
differences due to local adaptation can change over decades (Table 2), leaving the
question of whether short duration research accurately represent the population dynamics
that will occur post-revegetation.
One essential consideration that was frequently addressed in LAR was maternal
effects. Although Hereford (2009) anecdotally noted that most LAR experiments did not
account for maternal effects, I found that 74% of experiments controlled for maternal
effects in some way, although only 29% used plant materials from common
environments. Maternal effects can increase the observed differences among populations
(Table 2) and could alter the interpretation of higher fitness. The frequent use of
measures to control for maternal effects suggests that most LAR does not confound
transgenerational plasticity and genetic differentiation. In this aspect, LAR can be
appropriately applied to problems of moving genotypes from one environment to another.
Future direction - The difficulty of conducting LAR that can be applied to
management may in part stem from logistical obstacles in research and dependence on
short-term funding. For instance, the inclusion of λ as a response variable is complicated
by that fact that: 1) extended periods of data collection are required to accurately estimate
it for long-lived species (Che-Castaldo and Inouye 2011); and 2) that estimates of λ in
plants requires accounting for factors like seed banks (Adams, Marsh, and Knox 2005),
dormancy (Miller, Antos, and Allen 2012), and non-seed reproduction (Nault and
14
Gibson
PhD Dissertation: Chapter 1
2015
Gagnon 1993). It can also be difficult to study multiple selective factors in concert or to
determine which selective agents are important in natural field settings.
Even though ideal experimental considerations are likely unattainable,
investigators interested in research for restoration application could address a greater set
of considerations in their designs (Figure 4). First, they could increase the number of
populations and the sites and life-history stages assessed, and increase study duration.
Second, if utilizing λ is not feasible, researchers could test for fitness differences in
response to selective agents at specific life-history stages concurrently, rather than
sequentially. Third, performing local adaptation research over environmental gradients or
clines (Etterson 2004, Fant et al. 2008) has the advantage of determining the importance
of landscape variability over multiple scales on the expression of local adaptation.
Researchers could increase their participation in inter-regional or inter-continental
collaboration to allow the inclusion of more populations and habitats in local adaptation
experiments. Finally, combining reciprocal transplants in natural conditions with
controlled common garden experiments could provide greater information about the
drivers and magnitude of local adaptation (Raabova, Munzbergova, and Fischer 2011).
Results from experiments that included these six factors illustrate their importance in
assessing local adaptation, and managers should consider how directly LAR could inform
policy.
ACKNOWLEDGMENTS
This research was supported by agreement #09-CS-11015600-031 from the USDA Forest
Service NFN program and the National Science Foundation EPSCoR program (# EPS1101342) at the University of Montana.
15
Gibson
PhD Dissertation: Chapter 1
2015
LITERATURE CITED
Aavik, T., P. J. Edwards, R. Holderegger, R. Graf, and R. Billeter. 2012. Genetic
consequences of using seed mixtures in restoration: A case study of a wetland
plant Lychnis flos-cuculi. Biological Conservation 145 (1):195-204.
Adams, V. M., D. M. Marsh, and J. S. Knox. 2005. Importance of the seed bank for
population viability and population monitoring in a threatened wetland herb.
Biological Conservation 124 (3):425-436.
Agrawal, A. A. 2001. Transgenerational consequences of plant responses to herbivory: an
adaptive maternal effect? American Naturalist 157 (5):555-569.
———. 2002. Herbivory and maternal effects: mechanisms and consequences of
transgenerational induced plant resistance. Ecology 83 (12):3408-3415.
Barrett, S. C. H., R. I. Colautti, and C. G. Eckert. 2008. Plant reproductive systems and
evolution during biological invasion. Molecular Ecology 17 (1):373-383.
Becker, U., G. Colling, P. Dostal, A. Jakobsson, and D. Matthies. 2006. Local adaptation
in the monocarpic perennial Carlina vulgaris at different spatial scales across
Europe. Oecologia 150 (3):506-518.
Bell, D. L. and L. F. Galloway. 2008. Population differentiation for plasticity to light in
an annual herb: adaptation and cost. American Journal of Botany 95 (1):59-65.
Bennington, C. C., N. Fetcher, M. C. Vavrek, G. R. Shaver, K. J. Cummings, and J. B.
McGraw. 2012. Home site advantage in two long-lived arctic plant species:
results from two 30-year reciprocal transplant studies. Journal of Ecology 100
(4):841-851.
16
Gibson
PhD Dissertation: Chapter 1
2015
Bischoff, A., L. Cremieux, M. Smilauerova, C. S. Lawson, S. R. Mortimer, J. Dolezal, V.
Lanta, A. R. Edwards, A. J. Brook, M. Macel, J. Leps, T. Steinger, and H. MullerScharer. 2006. Detecting local adaptation in widespread grassland species - the
importance of scale and local plant community. Journal of Ecology 94 (6):11301142.
Bischoff, A. and H. Muller-Scharer. 2010. Testing population differentiation in plant
species - how important are environmental maternal effects. Oikos 119 (3):445454.
Blanquart, F., O. Kaltz, S. L. Nuismer, and S. Gandon. 2013. A practical guide to
measuring local adaptation. Ecology Letters 16 (9):1195-1205.
Bozorgipour, R. and J. W. Snape. 1997. An assessment of somaclonal variation as a
breeding tool for generating herbicide tolerant genotypes in wheat (Triticum
aestivum L). Euphytica 94 (3):335-340.
Broadhurst, L. M., A. Lowe, D. J. Coates, S. A. Cunningham, M. McDonald, P. A. Vesk,
and C. Yates. 2008. Seed supply for broadscale restoration: maximizing
evolutionary potential. Evolutionary Applications 1 (4):587-597.
Broadhurst, Linda M., Tom North, and Andrew G. Young. 2006. Should we be more
critical of remnant seed sources being used for revegetation? Ecological
Management & Restoration 7 (3):211-217.
Che-Castaldo, J. P. and D. W. Inouye. 2011. The effects of dataset length and mast
seeding on the demography of Frasera speciosa, a long-lived monocarpic plant.
Ecosphere 2 (11):18.
17
Gibson
PhD Dissertation: Chapter 1
2015
Donohue, K. and J. Schmitt. 1999. The genetic architecture of plasticity to density in
Impatiens capensis. Evolution 53 (5):1377-1386.
Donovan, L. A. and J. R. Ehleringer. 1992. Contrasting water-use patterns among size
and life-history classes of a semiarid shrub. Functional Ecology 6 (4):482-488.
Ehlers, B. K., E. Grondahl, J. Ronfort, and T. Bataillon. 2012. "Menage a trois": the
presence/absence of thyme shapes the mutualistic interaction between the host
plant Medicago truncatula (Fabaceae) and its symbiotic bacterium Sinorhizobium
meliloti. Ecology and Evolution 2 (7):1676-1681.
Espeland, E. K. and K. J. Rice. 2012. Within- and trans-generational plasticity affects the
opportunity for selection in barbed goatgrass (Aegilops triuncials). American
Journal of Botany 99 (12):2058-2062.
Freville, H. and J. Silvertown. 2005. Analysis of interspecific competition in perennial
plants using life table response experiments. Plant Ecology 176 (1):69-78.
Galloway, L. F. and J. R. Etterson. 2007. Transgenerational plasticity is adaptive in the
wild. Science 318 (5853):1134-1136.
Geber, M. A. and L. R. Griffen. 2003. Inheritance and natural selection on functional
traits. International Journal of Plant Sciences 164 (3):S21-S42.
Goransson, P., S. Andersson, and U. Falkengren-Grerup. 2009. Genetic adaptation to soil
acidification: experimental evidence from four grass species. Evolutionary
Ecology 23 (6):963-978.
Haubensak, K. A. and I. M. Parker. 2004. Soil changes accompanying invasion of the
exotic shrub Cytisus scoparius in glacial outwash prairies of western Washington
USA. Plant Ecology 175 (1):71-79.
18
Gibson
PhD Dissertation: Chapter 1
2015
Hereford, J. 2009. A quantitative survey of local adaptation and fitness trade-offs.
American Naturalist 173 (5):579-588.
———. 2010. Does selfing or outcrossing promote local adaptation? American Journal
of Botany 97 (2):298-302.
Hereford, J. and A. A. Winn. 2008. Limits to local adaptation in six populations of the
annual plant Diodia teres. New Phytologist 178 (4):888-896.
Hoffmann, A. A. and C. M. Sgro. 2011. Climate change and evolutionary adaptation.
Nature 470 (7335):479-485.
Horvitz, C. C. and D. W. Schemske. 1995. Spatiotemporal variation in demographic
transitions of a tropical understory herb - projection matrix analysis. Ecological
Monographs 65 (2):155-192.
Hufford, K. M. and S. J. Mazer. 2003. Plant ecotypes: genetic differentiation in the age of
ecological restoration. Trends in Ecology & Evolution 18 (3):147-155.
Hufford, K. M., S. J. Mazer, and M. D. Camara. 2008. Local adaptation and effects of
grazing among seedlings of two native california bunchgrass species: implications
for restoration. Restoration Ecology 16 (1):59-69.
Johnson, R. C., V. J. Erickson, N. L. Mandel, J. B. St Clair, and K. W. Vance-Borland.
2010. Mapping genetic variation and seed zones for Bromus carinatus in the Blue
Mountains of eastern Oregon, USA. Botany-Botanique 88 (8):725-736.
Johnson, R., L. Stritch, P. Olwell, S. Lambert, M. E. Horning, and R. Cronn. 2010. What
are the best seed sources for ecosystem restoration on BLM and USFS lands?
Native Plants 11 (2):117-131.
Jordan, D. S. 1905. The origin of species through isolation. Science 22:545-562.
19
Gibson
PhD Dissertation: Chapter 1
2015
Joshi, J., B. Schmid, M. C. Caldeira, P. G. Dimitrakopoulos, J. Good, R. Harris, A.
Hector, K. Huss-Danell, A. Jumpponen, A. Minns, C. P. H. Mulder, J. S. Pereira,
A. Prinz, M. Scherer-Lorenzen, A. S. D. Siamantziouras, A. C. Terry, A. Y.
Troumbis, and J. H. Lawton. 2001. Local adaptation enhances performance of
common plant species. Ecology Letters 4 (6):536-544.
Kawecki, T. J. and D. Ebert. 2004. Conceptual issues in local adaptation. Ecology Letters
7 (12):1225-1241.
Kawecki, T. J., R. E. Lenski, D. Ebert, B. Hollis, I. Olivieri, and M. C. Whitlock. 2012.
Experimental evolution. Trends in Ecology & Evolution 27 (10):547-560.
Kelly, C. A. 1992. Spatial and temporal variation in selection on correlated life-history
traits and plant size in Chamaecrista fasciculata. Evolution 46 (6):1658-1673.
Khurana, E. and J. S. Singh. 2001. Ecology of seed and seedling growth for conservation
and restoration of tropical dry forest : a review. Environmental Conservation 28
(1):39-52.
Knight, T. M. and T. E. Miller. 2004. Local adaptation within a population of
Hydrocotyle bonariensis. Evolutionary Ecology Research 6 (1):103-114.
Koch, John M. 2007. Restoring a jarrah forest understorey vegetation after bauxite
mining in Western Australia. Restoration Ecology 15 (4):S26-S39.
Lau, J. A. 2006. Evolutionary responses of native plants to novel community members.
Evolution 60 (1):56-63.
Lau, J. A., A. C. McCall, K. F. Davies, J. K. McKay, and J. W. Wright. 2008. Herbivores
and edaphic factors constrain the realized niche of a native plant. Ecology 89
(3):754-762.
20
Gibson
PhD Dissertation: Chapter 1
2015
Leger, E. A. and E. K. Espeland. 2010. Coevolution between native and invasive plant
competitors: implications for invasive species management. Evolutionary
Applications 3 (2):169-178.
Leimu, R. and M. Fischer. 2008. A meta-analysis of local adaptation in plants. Plos One
3 (12).
Macel, M., C. S. Lawson, S. R. Mortimer, M. Smilauerova, A. Bischoff, L. Cremieux, J.
Dolezal, A. R. Edwards, V. Lanta, T. M. Bezemer, W. H. van der Putten, J. M.
Igual, C. Rodriguez-Barrueco, H. Muller-Scharer, and T. Steinger. 2007. Climate
vs. soil factors in local adaptation of two common plant species. Ecology 88
(2):424-433.
Manel, S., M. K. Schwartz, G. Luikart, and P. Taberlet. 2003. Landscape genetics:
combining landscape ecology and population genetics. Trends in Ecology &
Evolution 18 (4):189-197.
McCarragher, S. R., D. Goldblum, and L. S. Rigg. 2011. Geographic variation of
germination, growth, and mortality in sugar maple (Acer saccharum): common
garden and reciprocal dispersal experiments. Physical Geography 32 (1):1-21.
McKay, J. K., C. E. Christian, S. Harrison, and K. J. Rice. 2005. "How local is local?" A review of practical and conceptual issues in the genetics of restoration.
Restoration Ecology 13 (3):432-440.
Merritt, D. J. and K. W. Dixon. 2011. Restoration seed banks - a matter of scale. Science
332 (6028):424-425.
Miao, S. L., F. A. Bazzaz, and R. B. Primack. 1991. Persistence of maternal nutrient
effects in Plantago major - the 3rd generation. Ecology 72 (5):1634-1642.
21
Gibson
PhD Dissertation: Chapter 1
2015
Miller, M. T., J. A. Antos, and G. A. Allen. 2012. Demography of a dormancy-prone
geophyte: influence of spatial scale on interpretation of dynamics. Plant Ecology
213 (4):569-579.
Nault, A. and D. Gagnon. 1993. Ramet demography of Allium tricoccum, a spring
ephemeral, perennial forest herb. Journal of Ecology 81 (1):101-119.
Pankova, H., J. Raabova, and Z. Munzbergova. 2014. Mycorrhizal symbiosis and local
adaptation in Aster amellus: a field transplant experiment. Plos One 9 (4):7.
Pywell, R. F., J. M. Bullock, D. B. Roy, L. I. Z. Warman, K. J. Walker, and P. Rothery.
2003. Plant traits as predictors of performance in ecological restoration. Journal
of Applied Ecology 40 (1):65-77.
Raabova, J., Z. Muenzbergova, and M. Fischer. 2007. Ecological rather than geographic
or genetic distance affects local adaptation of the rare perennial herb, Aster
amellus. Biological Conservation 139 (3-4):348-357.
Raabova, J., Z. Munzbergova, and M. Fischer. 2011. The role of spatial scale and soil for
local adaptation in Inula hirta. Basic and Applied Ecology 12 (2):152-160.
Rice, K. J. and E. E. Knapp. 2008. Effects of competition and life history stage on the
expression of local adaptation in two native bunchgrasses. Restoration Ecology 16
(1):12-23.
Rice, K. J. and R. N. Mack. 1991. Ecological genetics of Bromus tectorum. 3. The
demography of reciprocally sown populations. Oecologia 88 (1):91-101.
Roach, D. A. and R. D. Wulff. 1987. Maternal effects in plants. Annual Review of
Ecology and Systematics 18:209-235.
22
Gibson
PhD Dissertation: Chapter 1
2015
Siepielski, A. M., J. D. DiBattista, and S. M. Carlson. 2009. It's about time: the temporal
dynamics of phenotypic selection in the wild. Ecology Letters 12 (11):1261-1276.
St Clair, J. B., F. F. Kilkenny, R. C. Johnson, N. L. Shaw, and G. Weaver. 2013. Genetic
variation in adaptive traits and seed transfer zones for Pseudoroegneria spicata
(bluebunch wheatgrass) in the northwestern United States. Evolutionary
Applications 6 (6):933-948.
Streisfeld, M. A. and J. R. Kohn. 2007. Environment and pollinator-mediated selection
on parapatric floral races of Mimulus aurantiacus. Journal of Evolutionary
Biology 20 (1):122-132.
Sultan, S. E., K. Barton, and A. M. Wilczek. 2009. Contrasting patterns of
transgenerational plasticity in ecologically distinct congeners. Ecology 90
(7):1831-1839.
Svenning, M. M., O. Junttila, and B. Solheim. 1991. Symbiotic growth of indigenous
white clover (Trifolium repens) with local rhizobium Leguminosarum biovar
trifolii. Physiologia Plantarum 83 (3):381-389.
Thompson, J. D., P. Gauthier, J. Amiot, B. K. Ehlers, C. Collin, J. Fossat, V. Barrios, F.
Arnaud-Miramont, K. Keefover-Ring, and Y. B. Linhart. 2007. Ongoing
adaptation to mediterranean climate extremes in a chemically polymorphic plant.
Ecological Monographs 77 (3):421-439.
Thompson, J. N., S. L. Nuismer, and R. Gomulkiewicz. 2002. Coevolution and
maladaptation. Integrative and Comparative Biology 42 (2):381-387.
23
Gibson
PhD Dissertation: Chapter 1
2015
Thrall, P. H., J. J. Burdon, and J. D. Bever. 2002. Local adaptation in the Linum
marginale-Melampsora lini host-pathogen interaction. Evolution 56 (7):13401351.
USDA-NRCS. 2011. Release brochure for Aztec Maximilian Sunflower (Helianthus
maximiliani). Knox City, TX James E. “Bud” Smith Plant Materials Center.
———. 2013. Release brochure for Prairie Gold Maximilian sunflower (Helianthus
maximiliani). Manhattan, Kansas: Manhattan Plant Materials Center
USDI and USDA. 2002. Report to the Congress: interagency program to supply and
manage native plant materials for restoration and rehabilitation on federal lands.
Washington, DC, USA.
Vander Mijnsbrugge, K., A. Bischoff, and B. Smith. 2010. A question of origin: where
and how to collect seed for ecological restoration. Basic and Applied Ecology 11
(4):300-311.
Vitt, P., K. Havens, A. T. Kramer, D. Sollenberger, and E. Yates. 2010. Assisted
migration of plants: changes in latitudes, changes in attitudes. Biological
Conservation 143 (1):18-27.
Williams, S. L. 2001. Reduced genetic diversity in eelgrass transplantations affects both
population growth and individual fitness. Ecological Applications 11 (5):14721488.
24
Gibson
PhD Dissertation: Chapter 2
2015
Table 1: Frequency (number and %) of use of use of six key experimental methodologies in local
adaptation experiments (N = 308).
Frequency
Variable
#
%
Experimental environment
Experiment type
Reciprocal transplant
120
39
Common garden
101
33
87
28
Natural site
125
41
Artificial conditions
133
59
Only target plant species present
208
68
Native vegetation intact or added
78
25
14
5
137
44
63
20
126
41
22
7
3
1
Greenhouse
Site type
Other vegetation included
Measure of response
Fitness
Population growth rate (λ)
Reproductive success
Germination/emergence
Survival/mortality
Damage by herbivores/pathogens
Visitation from mutualists
25
Gibson
PhD Dissertation: Chapter 1
2015
Frequency
Variable
#
%
Size (e.g. biomass, number of leaves,
182
59
46
15
79
26
Juvenile
258
84
Reproduction
173
56
2 stages
124
40
41
13
Yes
12
4
No
296
96
Yes
64
21
No
244
79
Number of populations (mean)
8
-
Number of habitats plant material collected from (mean)
3
-
Plant material from controlled environment
89
29
Weighed seeds
37
12
circumference)
Other
Lifestages
Germination
All 3 stages
Multiple generations
Entire lifecycle
Among population variability
Maternal effects
26
Gibson
PhD Dissertation: Chapter 1
2015
Frequency
Variable
#
%
Kept maternal families separate
50
16
Initial plant size used as covariate
51
17
Plant
40
13
Herbivore
20
6
Pathogen
3
1
Mutualist
7
2
Soil biota
13
4
5
2
42
14
144
47
Soil
65
21
Light
10
3
Disturbance
31
10
3
1
Other
40
13
Multiple abiotic factors
27
9
2
-
Selective agents
Biotic factors
Multiple biotic factors
Biotic and abiotic factor
Abiotic factors
Climate
Distance
Spatial and temporal variability
Length of experiment (years; mean)
27
Gibson
PhD Dissertation: Chapter 1
2015
Frequency
Variable
#
%
Number of transplant sites or created environments
4
-
(mean)
28
Gibson
PhD Dissertation: Chapter 1
2015
Table 2: Examples of local adaptation experiments that incorporated variables that are informative to ecological restoration, and a
brief summary of the impact of the variable on the findings of local adaptation or population differentiation.
Variable
Authors
Summary
Reciprocal vs common garden
Raabova, Munzbergova,
Results from reciprocal transplant and common garden experiments
and Fischer 2011
differed in the observed level of population differentiation. While
both types of experiments showed greater height of local versus
foreign plants, the magnitude of difference was smaller in the
reciprocal transplant compared to the common garden experiment.
Inclusion of native vegetation
Bischoff et al. 2006
Inclusion or exclusion of the local plant community altered the
detection and magnitude of local adaptation in two species. Fitness
was higher for Plantago lanceolata when the native plant
community was present, while Holcus lanatus showed lower homesite fitness with the local plant community present.
29
Gibson
PhD Dissertation: Chapter 1
2015
Variable
Authors
Summary
Population growth rate (λ)
Becker et al. 2006
Findings about population fitness were different when fitness in
traits and lifetime fitness (λ) were assessed. Four of six life-history
traits studied showed non-significant differences between home
versus away populations; however, λ showed a significant home-site
advantage.
Multiple life stages
Raabova, Muenzbergova,
Findings of local adaptation depended on life-stage assessed.
and Fischer 2007
Evidence of local adaptation was seen in the number of germinates
(up to 68% higher in local versus foreign populations), but no
consistent evidence of local adaptation was found in adults.
Multiple populations/habitats
Hereford and Winn 2008
Evidence of home-site advantage was rare and depended on the
degree of habitat similarity. Local adaptation was not found when
populations were from the same habitat type, but was significantly
likely to be found when populations were from different habitats.
30
Gibson
PhD Dissertation: Chapter 1
2015
Variable
Authors
Summary
Plant materials from controlled
Bischoff and Muller-
Maternal effects impacted level of population differentiation
environment
Scharer 2010
detected and observed traits. Populations showed less differentiation
when using plants from controlled crosses than parent plants. The
ranking of populations in the F1 generation also changed for some
traits. Maternal effects were independent of seed mass.
Multiple factors
Lau 2006
Findings of adaptation varied when multiple biotic factors vs a
single factor were studied. When grown only with the invasive
Medicago polymorpha, Lotus wrangelianus plants from invaded
sites showed adaptation to invasion. There was no evidence of
adaptation to the invader when the insect herbivore Hypera
brunneipennis was included.
31
Gibson
PhD Dissertation: Chapter 1
2015
Variable
Authors
Summary
Experimental length
Bennington et al. 2012
Experimental length was important for the observation and
magnitude of local adaptation. For Dryas octopetala, the strength of
local adaptation increased over a decade. For Eriophorum
vaginatum, there was no evidence of local adaptation until 17 years
after transplant.
32
Gibson
PhD Dissertation: Chapter 2
2015
FIGURE CAPTIONS
Figure 1: Frequency of local adaptation experiments (proportion; N = 308) that tracked
plants during germination (G), non-reproductive juvenile or adult (NR), and reproductive
(R) life stages, or combinations thereof.
Figure 2: Frequency of local adaptation experiments (proportion, N = 308) by level of
replication (none to greater than 10) for populations (black bars) and habitats (white
bars). Population was defined by authors as a single source of plant materials. Habitat
refers to areas from which populations were collected.
Figure 3: Frequency of local adaptation experiments (proportion; N = 308) by
experimental duration in years.
Fig. 4: Schematic graph of the (a) common approach to local adaptation experiments and
(b) a suggested approach that could make experiments more relevant to land
management. Small letters (a-e) indicate plant populations; capital letters (A-E) indicate
sites or selective agents; and t indicates time after the beginning of the experiment. In (b),
grey shaded colors underlying boxes (a-e, A-E) represent an environmental or geographic
gradient. Row I and II indicate two contrasting types of transplantation designs that are
used: row I, material from multiple plant populations is crossed with one site or selective
agent; and row II, material from one plant population is crossed with multiple sites or
selective agents. Fully crossed experiments are not shown in the graph.
33
Gibson
PhD Dissertation: Chapter 2
2015
Figure 1
Frequency (proportion of papers)
0.5
G
NR
R
G/NR
G/R
NR/R
G/NR/R
0.4
0.3
0.2
0.1
0.0
1
2
3 (All)
Number of lifestages tracked
34
Gibson
PhD Dissertation: Chapter 2
2015
Figure 2
Frequency (proportion of papers)
0.8
Populations
Habitats
0.6
0.4
0.2
0.0
2-3
4-6
7-9
>10
Number of replicates
35
Gibson
PhD Dissertation: Chapter 2
2015
Figure 3
Frequency (proportion of papers)
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
1
2
3
4
5
6
7
8
9
>10
Study length (years)
36
Gibson
PhD Dissertation: Chapter 2
2015
Figure 4
37
Gibson
PhD Dissertation: Chapter 2
2015
CHAPTER 2
RESPONSE OF BLUEBUNCH WHEATGRASS TO INVASION: COMPETITIVE
ABILITY OF INVADER-EXPERIENCED AND NAÏVE POPULATIONS
ABSTRACT
38
Gibson
PhD Dissertation: Chapter 2
2015
Invasive species are altering selective pressures on native plant populations; competition
with novel invaders appear to lead to competitive differences among populations that
have experienced invasive weed competition and those that have not. I conducted a
greenhouse experiment to identify trait and competitive ability differences among
populations within experience type, among sites, and between invader-experience groups
of Pseudoroegneria spicata. Seeds collected from 14 wild populations of P. spicata and
the cultivar Anatone were grown alone or in pots with Centaurea stoebe. Invaderexperienced plants had greater shoot biomass at both life stages, and were less affected
by competition with C. stoebe than invader-naïve plants. Populations within experience
type differed in traits at both the seedling and adult life stages. Sites differed
significantly, but paired invader-experienced and naïve populations from the same site
did not, suggesting that site is an important factor in determining traits. Adult invaderexperienced plants showed a marginally significant trend towards greater tolerance for
competition with C. stoebe than did adult invader-naïve plants. Number of leaves
predicted seedlings tolerance to competition, and leaf length predicted the ability of
adults to suppress knapweed. There was not a significant relationship between a
population’s suppression of C. stoebe and its tolerance for competition. My results
suggest that population, site and invader- experience type should be considered when
choosing seed sources that are competitive with invasive species. Since tolerance and
suppression do not appear to be related in these populations, managers would need to
consider which element of competitive ability is most important for their goals.
Keywords: Centaurea stoebe, Pseudoroegneria spicata, competitive ability, invasive
species, native plant materials, relative interaction index, traits
39
Gibson
PhD Dissertation: Chapter 2
2015
INTRODUCTION
Although evolution is considered to be a slow process, rapid evolution in response
to changing environmental conditions has been found in plants from a variety of systems
(Rice and Emery 2003). The need to understand rapid evolutionary responses is
becoming more pressing as invasion by non-native species and anthropogenic land-use
create novel ecosystems – areas with previously unseen biotic assemblages (Hobbs et al.
2006). Invasions provide an interesting system for studying rapid evolution in native
plants; over relatively short time frames invaders can substantially alter ecosystems
through direct or indirect competition with natives and affect food webs, resource
cycling, and hydrology (Ramakrishnan and Vitousek 1989). While there is increasing
evidence that native plants are capable of adapting to invaders (e.g., Callaway et al. 2005,
Leger 2008, Rowe and Leger 2011, Oduor 2013), there are gaps in our understanding of
how adaptation to invasion interacts with population-level variation in traits and
competitive ability.
Because exotic invaders can exert strong selective pressures, they have the
potential to cause heritable trait shifts in local populations. Evidence of trait shifts in
native plants and animals has been seen in as little as 20 years after invasion (Strauss et
al. 2006). For example, Impatiens capensis plants that had been growing with the invader
for a period of time (invader experienced, IE) showed no reduction in biomass or
reproduction when grown with the nonnative invader Alliaria petiolata, while
competition with A. petiolata competition reduced the growth and reproduction of plants
that had have previous invader experience (invader-naïve, IN; Cipollini and Hurley
2008). Similarly, Lau (2006) found significant differences in seed production and
40
Gibson
PhD Dissertation: Chapter 2
2015
flowering time in Lotus wrangelianus populations with different histories of exposure to
the nonnative invader Medicago polymorpha, and Leger (2008) found that IE Elymus
multisetus plants collected from Bromus tectorum-invaded areas initiated growth earlier
and were less impacted by competition with B. tectorum than IN plants. The observed
ability to adapt to competition from invaders, however, does not appear to be consistent
across species (Callaway et al. 2005, Dostal et al. 2012) or maternal families (Rowe and
Leger 2011).
In some cases, an observed lack of significant adaptation to invaders could be
driven by among-population variation in competitive ability. Traits may differ among
populations in different sites, gene flow can swamp coevolved traits, and there are
differences among populations in selection intensity during coevolution (Thompson
1999). For instance, although studies on competitiveness of the native Pseudoroegneria
spicata after invasion by Centaurea stoebe found no overall differences between paired
populations from invaded and uninvaded sites, P. spicata showed high among-site
variability in percent germination in the presence of catachin produced by the invader
(Callaway et al. 2005), as well as in individual biomass, fertile tiller production and
competitive effect when grown in competition with C. stoebe (Lesica and Atthowe 2007).
Similarly, Goergen, Leger & Espeland (2011) found that the effect of invasion experience
varied significantly among populations of natives at different sites in response to the
invader B. tectorum. This population and site variation is important to consider both for
its impacts on signals of adaptation, as well as for understanding the most competitive
native plant materials to use in invaded ecosystems.
41
Gibson
PhD Dissertation: Chapter 2
2015
Competitive ability can be broken into two components – competitive effect
(suppression) and competitive response (Aarssen 1989, Bengtsson et al. 1994).
Competitive effect measures the ability of a plant to suppress its competitors’ growth or
reproduction by denying them resources, while competitive response is the target plant’s
tolerance of competition. These two components can be driven by different sets of traits
(Goldberg and Fleetwood 1987), or the same traits, such as biomass or height (Miller and
Werner 1987, Willis et al. 2010). Novel invaders may exert selective pressures that are
different from those that historically influenced a population’s traits and thus change the
selective environment of a site. Using seed from plants that have high competitive ability
with invaders, rather than seed adapted to prior selective pressures, may be the best
method to facilitate population establishment and to increase genetic diversity, allowing
for future adaptation to invasion (Jones and Monaco 2009).
Cultivars of native plant species are frequently used for revegetation because they
provide consistency in traits, such as faster growth rates or high seed production
compared to wild populations, which may enhance restoration success in highly degraded
ecosystems (Lesica and Allendorf 1999). Although research is limited, there is some
evidence that cultivars may be more competitive against invasives than local ecotypes,
largely due to their increased biomass production (Lesica and Atthowe 2007). However,
this may not be universally true for all cultivars, since competitiveness against invaders is
not an explicit selection criteria during breeding (Shwartz 2011). Understanding the
potential benefits of cultivars in terms of increased competitiveness in highly invaded
areas is critical in determining when their use is appropriate, given concerns that they
42
Gibson
PhD Dissertation: Chapter 2
2015
reduce genetic diversity and result in loss of locally adapted genotypes in their wild
progenitors (Schroder and Prasse 2013a, b).
I conducted an experiment on the response of a native grass, Pseudoroegneria
spicata (Pursh), to invasion by a non-native forb, Centaurea stoebe L. (Gugler) to
determine the nature of the interaction between two common species that each
substantially affect plant community structure, land use and economic value, and
restoration success. This study expands on previous research by explicitly exploring the
importance of populations and sites in response to invasion and adding to knowledge of
the competitive ability of cultivars. Using a greenhouse experiment, I addressed the
following questions:
1. Is there variation in traits of P. spicata between invader-experience and
invader-naïve plants, among populations within the same experience type, or
among sites?
2.
Are there between-experience-type or among-population differences in
competitive ability?
3.
Are traits predictive of competitive ability?
4. Is there a relationship between the two aspects of a population’s competitive
ability, suppression and tolerance?
MATERIALS AND METHODS
Study Species
Centaurea stoebe (spotted knapweed) was first introduced to North America from
central Europe in the late 1800s, and now infests over 1.9 million hectares. A perennial,
tap-rooted plant, C. stoebe lives between five and nine years (Boggs and Story 1987), and
43
Gibson
PhD Dissertation: Chapter 2
2015
plants produce 400 to 25,000 wind-dispersed seeds that can remain viable in the seed
bank for up to five years (Davis and Fay 1989). Centaurea stoebe appears to be
allelopathic in ways that inhibit the germination and growth of surrounding plants
(Thelen et al. 2005, Thorpe et al. 2009). Invasion by C. stoebe alters the demographics of
animal populations (Ortega et al. 2014), community trophic interactions (Ortega et al.
2006, Pearson 2009), plant community (Tyser and Key 1988) and arbuscular mycorrhizal
fungi community composition (Mummey and Rillig 2006), and increases sediment yield
(Lacey et al. 1989).
Pseudoroegneria spicata (bluebunch wheatgrass) is a drought-tolerant, longlived, native perennial bunchgrass found throughout the Intermountain West and Great
Plains. Plants are outcrossing and wind-pollinated (Zlatnik 1999). Although P. spicata is
diploid through most of its range, tetraploid and triploid populations exist throughout the
northwest and British Columbia (Larson et al. 2000, Gibson 2015 – chapter 3). Several
cultivars of P. spicata are currently available, including Anatone, a diploid, nearly
awnless cultivar first collected from the Umatilla National Forest near Anatone,
Washington (Larson et al. 2000). Anatone was developed for rapid establishment, high
leaf production, and ability to survive drought conditions.
Seed Collection
In July 2011, I collected P. spicata seeds from fourteen populations around
western Montana: eight from un-invaded sites (invader-naïve populations, IN) and six
from sites with a dense invasion of C. stoebe (invader-experienced populations, IE; Table
1). Invasion history for specific sites within the invaded areas was not available, but sites
were chosen in dense stands of C. stoebe; C. stoebe population growth in Montana is
44
Gibson
PhD Dissertation: Chapter 2
2015
around 3% per year (Jacobs and Sheley 1998), suggesting that these sites have been
invaded for at least 40 years. For each P. spicata population, I collected seeds from
fifteen individual plants, bulking seeds during collection.
Centaurea stoebe seeds were collected from three invaded sites in summer 2011
and bulked across all populations after collection. Seed of the cultivar Anatone were
ordered from Western Native Seed (Coaldale, CO).
Greenhouse
In fall 2011, I weighed seeds to determine average seed mass for each population.
Seeds of IE and IN P. spicata populations and of Anatone were seeded with and without
C. stoebe in a greenhouse using deep pots (Stuewe and Sons D40; 6.4 cm x 25 cm;
0.65L). Pots were filled with Super Soil Potting Mix (compost, peat moss, and fertilizer).
Plants were grown in an unlit greenhouse (minimum 10 hours natural daylight) that
ranged between 10 and 32° C. For each of the fourteen populations, I planted 44 control
pots with three P. spicata seeds (N = 616) and an additional 44 invaded pots with three P.
spicata and two C. stoebe seeds (invaded; N = 616). For Anatone, I sowed 20 control
pots with three seeds each and 20 pots with three Anatone and two C. stoebe seeds. In
addition, I sowed ten control pots with three C. stoebe seeds each. After seeding, pots
were randomly placed in the greenhouse and moved once a month. Pots were misted until
germination and then hand watered twice a week. I thinned pots with both C. stoebe and
P. spicata to one plant per species per pot after germination.
Trait Variation
45
Gibson
PhD Dissertation: Chapter 2
2015
Seedlings (25 days): For P. spicata, I determined seedling emergence date by
population starting seven days after planting by recording the number of pots with new
germinates weekly for one month and then twice weekly for the remainder of the
experiment; the majority of pots had germinates after one week, so only week 1
germination was used in analysis. I recorded phenological and growth variables 25 days
after the emergence of the first seedling for five plants per population and treatment
(control and invaded): length of longest leaf, number of leaves, total biomass, shoot
biomass, root biomass and root-shoot ratio. To measure total and root biomass, I
destructively sampled plants in up to five pots per population from each treatment at 25
days. Due to low or slow germination in both P. spicata and C. stoebe, I was unable to
harvest all five plants per treatment for some populations at the 25-day sampling (Table
1). Samples were washed, air dried, and weighed to determine biomass components
(total, root, and shoot biomass) and root-shoot biomass ratio.
To quantify the effect of competition of C. stoebe on P. spicata, I calculated the
relative interaction index (Armas et al. 2004), a measure of interaction intensity on a
scale of -1 to +1, with plant biomass (B) as the variable of interest:
RII 
( Bw  Bo )
( Bw  Bo )
where Bw is the mass of the plant when grown with C. stoebe and Bo is the mass when
grown alone. Relative interaction index values close to -1 (low biomass when grown with
competitor) indicate that the plant was highly harmed by competition, while values close
to +1 (high biomass with competitor) indicate positive effects of competition. Four
populations were excluded from the seedling 25-day analyses due to lack of germination
46
Gibson
PhD Dissertation: Chapter 2
2015
in the competition treatment pots (Table1). Seedling suppression of C. stoebe was not
calculated due to lack of invader seedlings.
Adult (100 days): Pots were watered bi-weekly and kept in the greenhouse under
the same growing conditions maintained during the seedling phase. After three months, I
recorded data on the same growth variables used in the seedling phase. Biomass was
recorded for up to five of the remaining plants per treatment per population (Table 1). In
addition, five of the C. stoebe plants were harvested and weighed for biomass. Data
collection methods for each of these variables were the same as used in the seedling
phase. After data had been collected, I calculated adult P. spicata tolerance of C. stoebe
(RII-Adult) and C. stoebe suppression by P. spicata (RII-Knapweed) using adult
biomass.
Statistical Analysis
Trait variation: Most response variables were normally distributed; the exceptions
were seedling length of longest leaf, seedling root biomass, and adult root-shoot ratio,
each of which were log transformed for normalcy. All statistical analysis were performed
in R (R Core Development Team 2012).
First, I assessed variation between invasion-experience types in average seed
weight and week-one germination using ANOVA models, with separate tests for each
variable. Next, I assessed trait variation among populations of the same experience type,
and between invader-experience groups (IE and IN) using general linear models with
population nested in experience type, and with separate models for each response
variable (length of longest leaf, number of leaves, total biomass, shoot biomass, root
biomass, and root-shoot ratio) at each phase (seedling and adult). I also assessed the
47
Gibson
PhD Dissertation: Chapter 2
2015
effects of treatment (competition) and invader-experience (IE or IN) using separate
general linear models that allowed for unequal variance by population (Pinheiro et al.
2012). For significant interactions between treatment and experience type, post-hoc
Tukey HSD tests were used to determine significant differences between experience
types at the p = 0.05 level. I excluded four populations from seedling trait analysis and
three from adult trait analysis because there were not enough plants available from
treatment pots (Table 1).
I assessed variation between among sites, invasion experience types, among
paired IE and IN populations nested within site for all response variables using
MANOVA models, with separate tests for seedlings and adults. Due to a limited number
of germinates, the effect of site was assessed in a subset of paired IE-IN populations
(seedling stage sites MM, MCL, and PT; adult stage sites LH and MM). Biomass was
excluded from the adult MANOVA model due to correlated data and small sample size.
If an effect was significant in MANOVA, I assessed which individual traits showed
significant differences using ANOVA models. For significant individual traits I assessed
which populations varied at the p = 0.05 level using post-hoc Tukey HSD tests.
Although I did not control for maternal effects, I did assess whether maternal seed
provisioning was predictive of traits. I used linear regression models with seedling and
adult traits and population seed weight with separate models for each trait at each life
stage; individual seed weight could not be used since I sowed multiple seeds per pot.
Seed weight did not significantly predict any measured trait.
48
Gibson
PhD Dissertation: Chapter 2
2015
Competitive response: I tested for differences in P. spicata tolerance (RII-Seedling,
RII-Adult) between invasion-experience types using ANOVA, with separate models for
seedlings and adults. Anatone was included as an IN population. I excluded four
populations from seedling RII analysis and six from adult RII analyses because there
were less than three plants per population (Table 1).
In addition to assessing effects of the invader on P. spicata populations, I also tested
for competitive effects of P. spicata, and of P. spicata invasion experience, on C. stoebe
using ANOVA models with C. stoebe suppression (RII-Knapweed) as the response
variable.
Traits and competitive response: Multiple regression was used to assess whether
traits from plants grown in competition with C. stoebe predicated a population’s tolerance
(RII-Seedling, RII-Adult) or suppression (RII-Knapweed). Seedling total biomass and
adult total biomass were removed from regression models because of significant
correlations between seedling root biomass and leaf number, and between shoot and root
biomass, respectively.
Relationship between suppression and tolerance: I used linear regression to test the
relationship between a population’s tolerance (RII-Adult) and suppression of C. stoebe
(RII-Knapweed).
RESULTS
Trait Variation
Seedling: Experience type was not significant for seed weight or week-1 percent
germination, although there was a trend toward slower germination in Anatone regardless
49
Gibson
PhD Dissertation: Chapter 2
2015
of treatment. There was a significant effect of competition treatment for three of six
measured traits (Table 2). Plants grown in pots with C. stoebe had shorter and fewer
leaves (Figures 1a, b) and higher root-shoot ratios (more allocation to shoot biomass;
Figure 1f). Experience type was significant for only one trait: average shoot biomass was
higher for IE populations (mean = 0.17, SD = 0.12) than for IN populations (mean = 0.15,
SD = 0.08; Figure 1d). Populations within experience type varied for four traits (Table 2):
length of longest leaf, leaf number, shoot biomass, and root shoot ratio.
There were significant interactions of treatment and experience type for number of
leaves and root-shoot ratio (Table 2). There were no significant differences, however,
between within-treatment means (Figure 1).
Adult: Experience type was significant for two traits: length of longest leaf and
shoot biomass (Table 2). Compared to IN plants, IE plants had shorter leaves
(mean=36.1, SD = 10.3; versus mean=36.6, SD=12.4), and greater average shoot biomass
(mean = 1.23, SD=0.71; versus mean=1.43, SD = 0.68). For a third variable, leaf number,
differences between groups was marginally significant (Table 2), with higher leaf
numbers for IE than IN plants (mean = 52.3, SD = 22.4; versus mean = 46.5; SD = 26.9).
Competition treatment was significant for four of six measured traits (Table 2). Plants
growing with C. stoebe had fewer leaves (Figure 2b) and lower total, root and shoot
biomass (Figures 2c – e). Population within experience type was significant for five out
of six variables (Table 2): length of longest leaf, number of leaves, biomass, root
biomass, and shoot biomass.
There was a significant interaction between treatment and experience type for
length of longest leaf, biomass, and root biomass (Table 2). For the C. stoebe-invaded
50
Gibson
PhD Dissertation: Chapter 2
2015
treatments, post-hoc tests showed significant differences between IE and IN groups for
total biomass, as well as marginally significant differences between groups for leaf length
(P = 0.08) and root biomass (P = 0.14). When grown with C. stoebe, IE plants had
greater total biomass, longer leaves, and greater root biomass than did IN plants (Figures
2a, c, e). There were no differences between experience types in control pots.
For the subset of populations used for the site analysis, there was a significant effect
of site on seedling traits, but no effect of experience type or population within site (Table
3). Site was statistically significant for differences in number of leaves (F(2,48)=7.791, P
= 0.001), biomass (F(2, 48)=6.232, P = 0.004), and root-shoot ratio (F=5.361, P = 0.008).
Post-hoc tests showed variation in which sites differed from each other (Figure 3). There
was a significant effect of site on adult traits, but no effect of experience type or
population within site (Table 3). There were significant differences between the two sites
for number of leaves (F(1, 33)=16.3, P < 0.001; Figure 4b) and root biomass (F=3.247, P
= 0.01; Figure 4e).
Competitive Response
Seedlings from IE and IN populations did not differ significantly in their tolerance of
C. stoebe (RII-Seedling; F (1,9)=0.118, P = 0.739; Figure 5a). Adult plants from IE
populations, however, tended to be less harmed by competition with C. stoebe compared
to plants from IN populations, although the difference was only marginally significant
(RII-Adult; F(1,9) = 4.82, P = 0.056; Figure 5b). There was no difference between
experience types in their suppression of knapweed (RII-Knapweed; F=0.418, P = 0.534,
Figure 5c).
Traits and Competitive Response
51
Gibson
PhD Dissertation: Chapter 2
2015
Only one trait, number of leaves, was predictive of seedlings ability to tolerate
competition (β=0.04, P = 0.01), although it explained a large amount of observed
variation (RII-Seedling; R2 = 0.53, F(1,9)=10.35, P = 0.01). No traits significantly
predicted adult tolerance of knapweed (RII-Adult).
Adult length of longest leaf, however, significantly predicted suppression of
knapweed (RII-Knapweed; β = -0.02, p = 0.04; R2 = 0.29, F(1,12) = 5.01, P = 0.04).
Relationship between Suppression and Tolerance
A population’s tolerance (RII-Adult) was not predictive of its suppression of C.
stoebe (RII-Knapweed; Figure 6). Anatone was the least competitive with C. stoebe,
showing both low tolerance and suppression. Anatone had high leverage values in the
regression and, when it was dropped from the model, the relationship between tolerance
and suppression was not significant (R2 = 0.17, F(1,11) = 2.20, P = 0.16).
DISCUSSION
Ongoing invasion of habitats by non-native species is changing the environment and
consequently community dynamics (e.g., Pearson 2009; Thorpe & Callaway 2011;
Malick, Belant & Bruggink 2012; Ortega, Benson & Greene 2014). Although there is
limited data on trait changes in response to invasion, an increasing number of
investigations are addressing this topic — not only because it is ecologically interesting,
but also because of its relevance for making informed management decisions (Leger and
Espeland 2010). My results add to the body of work on the effects of invasion on native
species’ traits and competitive ability and demonstrate the importance of explicit
52
Gibson
PhD Dissertation: Chapter 2
2015
consideration of variation among populations and sites when researching response to
invasion (Goergen, Leger & Espeland 2011).
Population, rather than experience type, was the most important driver of trait
differences, and populations within the same experience type showed significant
variation for the majority of traits and competitive ability. This is likely due to the fact
that a population’s traits and genetics are the result of historic selective pressures; just as
populations currently show variation in response to historic local conditions, the selective
pressures of a new invader have the potential to interact with a population’s historic
selective regime leading to variation among populations spread across the landscape
(Thompson 1999, Leger and Espeland 2010). These results suggest that population
differences should be explicitly considered, and not just relegated to background
variation, when assessing overall species response to invasion. They also suggest that
inference at the species level may require data from a relatively large number of
populations.
Despite widespread invasion of nonnative species in many terrestrial ecosystems,
there are still only a handful of studies on native species response to invasion, and results
from these experiments show mixed results both within and among species (Oduor 2013).
I found evidence that C. stoebe can impact P. spicata – seedlings from IE populations
were significantly larger than those from IN populations, and adults had higher shoot
biomass. Previous research found that size is positively correlated with competitive
ability against invaders in general (Gaudet and Keddy 1988) and, specifically, against C.
stoebe (Lesica and Atthowe 2007), and the larger biomass and leaf number observed in
IE populations used in this study could be a response to competition with C. stoebe.
53
Gibson
PhD Dissertation: Chapter 2
2015
My findings, however, differ from previous studies of P. spicata and C. stoebe
(Callaway et al. 2005, Lesica and Atthowe 2007) which did not find significant trait
differences due to invasion experience. Differences in findings could be due to the small
number of populations used in each study; given my observation of high amongpopulation variability in response, a large number of populations would need to be
assessed to determine a mean response for the species. Furthermore, one IN population in
my study (Anatone, the cultivar) had a large effect on observed trends. Of all the
populations, it had the lowest total and shoot biomass and leaf length – three traits that
were significantly different between adult plants from the two invader-experience groups.
My analyses also reveals that site significantly influenced traits. Other studies of
response to invasion have also found that site is an important driver of traits for both P.
spicata and other species (e.g., Callaway et al. 2005, Leger 2008, Atwater 2012). I did
not find between-experience-type differences in traits in populations from the same site;
given the limited number of paired populations included in analysis, however, it is
difficult to say if site is more important than experience in determining population-level
traits. Researchers generally rely on paired IE-IN populations to prevent the conflation of
site traits with response to invasion. If I had observed consistent differences in traits
between unpaired IE and IN populations across all populations (regardless of C. stoebe
competition) this would be a strong indication that site was the primary driver of traits
(i.e., IE populations are always show higher total biomass because they come from sites
with larger plants than IN populations). While I did find some traits that showed these
consistent differences, I primarily found evidence that IE plants responded to competition
differently than those from IN population. When C. stoebe was present, there were
54
Gibson
PhD Dissertation: Chapter 2
2015
between experience-type differences in adult total biomass and leaf number; overall,
plants from IE populations were larger and less affected by competition than those from
IN populations even though there was no difference in the absence of C. stoebe.
Plants can compete by reducing resources in their environment (suppression), by
tolerating reduced resources (Goldberg 1996), or by being both suppressive and tolerant.
I found a trend towards greater tolerance of C. stoebe in adult plants from IE populations,
but no differences in suppressive ability. Competition with C. stoebe may be the driver of
between-experience-type differences in tolerance; there is some evidence that increased
tolerance is selected for by intraspecific competition (Uriarte et al. 2002), and tolerance is
likely an important predictor of P. spicata competitive success in invaded ecosystems
(Atwater 2012). Not only did tolerance vary between experience types, but competitive
ability (both suppression and tolerance) varied among populations. These competitive
differences among populations could be due to length of exposure to C. stoebe, or
populations may exhibit these differences due to unrelated site selective factors.
Traits can be useful in selecting species that will be successful for revegetation in
restoration (Fischer et al. 2013). I found two traits that were predictive of greater
competitive ability: at the seedling stage, plants with more leaves tolerated competition
with C. stoebe better than did smaller plants, while at the adult stage plants with longer
leaves suppressed C. stoebe more than plants with shorter leaves. Larger plants – either in
height or leaf number – are better at obtaining resources and competing at high densities
(Weiner 1990, Keddy et al. 2002, Craine 2005), and both higher leaf number and longer
leaves were associated with plants from IE populations. Leaf-related traits may have
55
Gibson
PhD Dissertation: Chapter 2
2015
responded to invasion by C. stoebe, or plants that respond to competition by increasing
leaf number and length may have persisted after invasion.
I did not find a relationship between a population’s tolerance and suppression, which
supports previous findings that the two are unrelated in P. spicata (Atwater 2012).
Considering that C. stoebe uses the allelopathic compound (±)-catechin to suppress its
competitor’s growth (Bais et al. 2003, Thorpe et al. 2009), the traits that allow P. spicata
to tolerate suppression by catechin are likely different than the traits that allow the
species to preempt resources (suppression) from C. stoebe. Understanding a population’s
tolerance and suppressive abilities could provide important information for ecosystem
managers because an ecotype’s tolerance and suppression ability can impact future
community structure. Not only does competitive ability determine species’ abundance
(MacDougall and Turkington 2004), but the two components of competitive ability can
also influence genetic and species diversity within an invaded ecosystems depending on
whether a population is tolerant of an invader (helping only itself) or suppresses the
invader, thus helping the entire native community (Atwater 2012). Selecting highly
tolerant seed sources for restoration could increase survival of individual ecotypes, but
selecting for highly suppressive ecotypes could increase overall genetic diversity within a
population, which can positively impact ecosystem function (Hughes et al. 2008, Bolnick
et al. 2011).
Cultivars are considered to perform well in stressful environments and are
recommended for use in highly disturbed sites (Lesica and Allendorf 1999), yet I found
that the cultivar Anatone performed worse than wild populations when grown in
competition with C. stoebe. This was unexpected given Lesica and Atthowe’s (2007)
56
Gibson
PhD Dissertation: Chapter 2
2015
finding that the cultivar Goldar was more competitive against C. stoebe than were wild
populations. The difference between the performances of these two cultivars could be due
to growth rates. Compared to Goldar, Anatone takes longer to initiate shoot growth (RayMukherjee et al. 2011) and has lower specific leaf area (Ray-Mukherjee 2010), both of
which could impact its competitive ability. For instance, specific leaf area has been
linked to competitive ability in a variety of species (e.g. Maron and Marler 2008, Rosch
et al. 1997), and differences in shoot growth between IE and IN plants has been found in
Elymus multisetus plants from sites invaded by Bromus tectorum (Leger 2008). In
addition to these traits, germination speed is important for competitive success (Dyer et
al. 2000, Orrock and Christopher 2010), but Anatone germinated later than other
populations in the greenhouse. Pseudoroegneria spicata plants that germinated early
were likely better able to compete with C. stoebe for the limited space and resources in
the pots.
Selective pressures need to be strong to cause population differentiation over small
spatial scales (Sambatti and Rice 2006), and gene flow between populations can prevent
local adaptation in the population receiving genes or migrants (Lenormand 2002, Rice
and Emery 2003, Garant et al. 2007 Leger and Espeland 2010). A lack of difference
between IE and IN populations from the same site is not surprising given that P. spicata
is outcrossing and wind-pollinated. Previous results of significant differences between IE
and IN populations have largely been found in species with pollination or reproductive
methods that limit gene flow (apomictic, self-fertile, or animal-pollinated species; e.g.,
Cipollini and Hurley 2008, Leger 2008, Lau 2008, Atwater 2012). Anatone represents the
most naïve P. spicata population, since does not compete with C. stoebe during
57
Gibson
PhD Dissertation: Chapter 2
2015
cultivation on seed farms and is likely isolated from gene flow from invader-experienced
populations. Anatone’s low tolerance and suppression of C. stobe and its inclusion as an
IN population could reconcile my findings of differences between IE and IN plants in the
presence of C. stoebe with the lack of observed differences between IE and IN
populations at individual sites. Future research on response to invasion could focus more
on the relationship between isolation and trait shifts as a way to better understand the
mosaic of coevolution on the landscape (Thompson et al. 2002).
One potential limitation of my findings is that I cannot fully rule out maternal effects
as the driver of observed competitive or trait differences, given that maternal influences
affect plant size, competitive ability, and fitness (Roach and Wulff 1987, Galloway 2005,
Galloway and Etterson 2007). However, that fact that seed weight, which is frequently
used as a proxy for controlling maternal effects (e.g., Rowe and Leger 2012, Gibson 2015
– chapter 2), did not vary among populations or between invader-experience groups
suggests that there were not differences in maternal provisioning, but maternal effects can
influence traits beyond seed weight (Dyer et al. 2010).
Management and Policy Implications - Results from this study support previous
research on the potential for invasive species to impact population traits of a native
species, but I found that the frequency and magnitude of response varies dramatically
among populations and sites. While plants from invaded areas were observed to be more
competitive with C. stoebe than those from uninvaded areas, some invader-naïve
populations showed high tolerance or suppression of C. stoebe. Screening for potential
competitive ability with invaders should be based on traits, not just invader experience. I
found that leaf length and leaf number were related to a population’s competitive ability,
58
Gibson
PhD Dissertation: Chapter 2
2015
and these traits could be important for selecting seed sources for plant material in areas
where competition with invasive weeds is the primary land management goal. My
findings also challenge the assumption that cultivars are better competitors with invasive
species than are local ecotypes. The large variation between our findings of low
competitive ability of Anatone and Lesica and Atthowe’s (2007) findings of high
competitive ability of Goldar suggests that cultivar varieties should be individually tested.
Ideally, future research on this topic will include a greater number of wild populations
and cultivars in long-term field experiments. Selection of plant materials for revegetation
in invaded sites may be more about identifying the most competitive populations from
specific sites rather than the most competitive species.
ACKNOWLEDGEMENTS
This research was supported by agreement #09-CS-11015600-031 from the USDA Forest
Service NFN program and the National Science Foundation EPSCoR program (# EPS1101342) at the University of Montana.
LITERATURE CITED
Aarssen, L. W. 1989. Competitive ability and species coexistence - a plant's eye view.
Oikos 56:386-401.
Armas, C., R. Ordiales, and F. I. Pugnaire. 2004. Measuring plant interactions: a new
comparative index. Ecology 85:2682-2686.
Atwater, D. 2012. Interplay between competition and evolution in invaded and native
plant communities. Theses, Dissertations, Professional Papers. Paper 4156.
University of Montana, Missoula, MT.
59
Gibson
PhD Dissertation: Chapter 2
2015
Bengtsson, J., T. Fagerstrom, and H. Rydin. 1994. Competition and coexistence in plant
communtities. Trends in Ecology & Evolution 9:246-250.
Boggs, K. W., and J. M. Story. 1987. The population age structure of spotted knapweed
(Centaurea maculosa) in Montana. Weed Science 35:194-198.
Callaway, R. M., W. M. Ridenour, T. Laboski, T. Weir, and J. M. Vivanco. 2005. Natural
selection for resistance to the allelopathic effects of invasive plants. Journal of
Ecology 93:576-583.
Cipollini, K. A., and S. L. Hurley. 2008. Variation in resistance of experienced and naive
seedlings of jewelweed (Impatiens capensis) to invasive garlic mustard (Alliaria
petiolata). Ohio Journal of Science 108:47-49.
Craine, J. M. 2005. Reconciling plant strategy theories of Grime and Tilman. Journal of
Ecology 93:1041-1052.
Davis, E., and P. Fay. 1989. The longevity of spotted knapweed seed in Montana. Pages
67 - 72 in P. Fay and J. Lacey, editors. Proceedings: knapweed symposium; 1989
April 4-5; Bozeman, MT, Bozeman, MT: Montana State University.
Dostal, P., M. Weiser, and T. Koubek. 2012. Native jewelweed, but not other native
species, displays post-invasion trait divergence. Oikos 121:1849-1859.
Dyer, A. R., C. S. Brown, E. K. Espeland, J. K. McKay, H. Meimberg, and K. J. Rice.
2010. The role of adaptive trans-generational plasticity in biological invasions of
plants. Evolutionary Applications 3:179-192.
Dyer, A. R., A. Fenech, and K. J. Rice. 2000. Accelerated seedling emergence in
interspecific competitive neighbourhoods. Ecology Letters 3:523-529.
60
Gibson
PhD Dissertation: Chapter 2
2015
Fischer, L. K., M. von der Lippe, and I. Kowarik. 2013. Urban grassland restoration:
which plant traits make desired species successful colonizers? Applied Vegetation
Science 16:272-285.
Galloway, L. F. 2005. Maternal effects provide phenotypic adaptation to local
environmental conditions. New Phytologist 166:93-99.
Galloway, L. F., and J. R. Etterson. 2007. Transgenerational plasticity is adaptive in the
wild. Science 318:1134-1136.
Garant, D., S. E. Forde, and A. P. Hendry. 2007. The multifarious effects of dispersal and
gene flow on contemporary adaptation. Functional Ecology 21:434-443.
Gaudet, C. L., and P. A. Keddy. 1988. A comparative approach to predicting competitive
ability from plant traits. Nature 334:242-243.
Gibson, A. 2014. Polyploidy: a missing link in the conversation about restoration of a
commonly seeded native grass in western North America. University of Montana,
Missoula, MT.
Goergen, E. M., E. A. Leger, and E. K. Espeland. 2011. Native perennial grasses show
evolutionary response to Bromus tectorum (cheatgrass) invasion. Plos One 6:
e18145. doi: 10.1371/journal.pone.0018145.
Goldberg, D. E. 1996. Competitive ability: definitions, contingency and correlated traits.
Philosophical Transactions of the Royal Society of London Series B-Biological
Sciences 351:1377-1385.
Goldberg, D. E., and L. Fleetwood. 1987. Competitive effect and response in four annual
plants. Journal of Ecology 75:1131-1143.
61
Gibson
PhD Dissertation: Chapter 2
2015
Hobbs, R. J., S. Arico, J. Aronson, J. S. Baron, P. Bridgewater, V. A. Cramer, P. R.
Epstein, J. J. Ewel, C. A. Klink, A. E. Lugo, D. Norton, D. Ojima, D. M.
Richardson, E. W. Sanderson, F. Valladares, M. Vila, R. Zamora, and M. Zobel.
2006. Novel ecosystems: theoretical and management aspects of the new
ecological world order. Global Ecology and Biogeography 15:1-7.
Jacobs, J. S., and R. L. Sheley. 1998. Observation: life history of spotted knapweed.
Journal of Range Management 51:665-673.
Jones, T. A., and T. A. Monaco. 2009. A role for assisted evolution in designing native
plant materials for domesticated landscapes. Frontiers in Ecology and the
Environment 7:541-547.
Keddy, P., K. Nielsen, E. Weiher, and R. Lawson. 2002. Relative competitive
performance of 63 species of terrestrial herbaceous plants. Journal of Vegetation
Science 13:5-16.
Lacey, J. R., C. B. Marlow, and J. R. Lane. 1989. Influence of spotted knapweed
(Centaurea maculosa) on surface runoff and sediment yield. Weed Technology
3:627-631.
Larson, S. R., T. A. Jones, Z. M. Hu, C. L. McCracken, and A. Palazzo. 2000. Genetic
diversity of bluebunch wheatgrass cultivars and a multiple-origin polycross. Crop
Science 40:1142-1147.
Lau, J. A. 2006. Evolutionary responses of native plants to novel community members.
Evolution 60:56-63.
Leger, E. A. 2008. The adaptive value of remnant native plants in invaded communities:
an example from the Great Basin. Ecological Applications 18:1226-1235.
62
Gibson
PhD Dissertation: Chapter 2
2015
Leger, E. A., and E. K. Espeland. 2010. Coevolution between native and invasive plant
competitors: implications for invasive species management. Evolutionary
Applications 3:169-178.
Lenormand, T. 2002. Gene flow and the limits to natural selection. Trends in Ecology &
Evolution 17:183-189.
Lesica, P., and F. W. Allendorf. 1999. Ecological genetics and the restoration of plant
communities: mix or match? Restoration Ecology 7:42-50.
Lesica, P., and H. E. Atthowe. 2007. Identifying weed-resistant bluebunch wheatgrass for
restoration in western Montana. Ecological Restoration 25:191-198.
MacDougall, A. S., and R. Turkington. 2004. Relative importance of suppression-based
and tolerance-based competition in an invaded oak savanna. Journal of Ecology
92:422-434.
Miller, T. E., and P. A. Werner. 1987. Competitive effects and responses between plantspecies in a 1st-year old-field community. Ecology 68:1201-1210.
Mummey, D. L., and M. C. Rillig. 2006. The invasive plant species Centaurea maculosa
alters arbuscular mycorrhizal fungal communities in the field. Plant and Soil
288:81-90.
Oduor, A. M. O. 2013. Evolutionary responses of native plant species to invasive plants:
a review. New Phytologist 200:986-992.
Orrock, J. L., and C. C. Christopher. 2010. Density of intraspecific competitors
determines the occurrence and benefits of accelerated germination. American
Journal of Botany 97:694-699.
63
Gibson
PhD Dissertation: Chapter 2
2015
Ortega, Y. K., A. Benson, and E. Greene. 2014. Invasive plant erodes local song diversity
in a migratory passerine. Ecology 95:458-465.
Ortega, Y. K., K. S. McKelvey, and D. L. Six. 2006. Invasion of an exotic forb impacts
reproductive success and site fidelity of a migratory songbird. Oecologia 149:340351.
Pearson, D. E. 2009. Invasive plant architecture alters trophic interactions by changing
predator abundance and behavior. Oecologia 159:549-558.
Ramakrishnan, P. S., and P. M. Vitousek. 1989. Ecosystem-level processes and the
consequences of biological invasions. in J. Drake, F. diCastri, R. Groves, F.
Kruger, H. Mooney, M. Rejmanek, and M. Williamson, editors. Biological
invasions: a global perspective. Wiley, New York, NY, USA.
Ray-Mukherjee, J. 2010. Evaluating native wheatgrass for restoration of Sagebrush
Steppes. Utah State University.
Ray-Mukherjee, J., T. A. Jones, P. B. Adler, and T. A. Monaco. 2011. Immature seedling
growth of two North American native perennial bunchgrasses and the invasive
grass Bromus tectorum. Rangeland Ecology & Management 64:358-365.
Rice, K. J., and N. C. Emery. 2003. Managing microevolution: restoration in the face of
global change. Frontiers in Ecology and the Environment 1:469-478.
Roach, D. A., and R. D. Wulff. 1987. Maternal effects in plants. Annual Review of
Ecology and Systematics 18:209-235.
Rowe, C. L. J., and E. A. Leger. 2011. Competitive seedlings and inherited traits: a test of
rapid evolution of Elymus multisetus (big squirreltail) in response to cheatgrass
invasion. Evolutionary Applications 4:485-498.
64
Gibson
PhD Dissertation: Chapter 2
2015
Sambatti, J. B. M., and K. J. Rice. 2006. Local adaptation, patterns of selection, and gene
flow in the Californian serpentine sunflower (Helianthus exilis). Evolution
60:696-710.
Schroder, R., and R. Prasse. 2013a. Do cultivated varieties of native plants have the
ability to outperform their wild relatives? Plos One 8.
Schroder, R., and R. Prasse. 2013b. From nursery into nature: a study on performance of
cultivated varieties of native plants used in re-vegetation, their wild relatives and
evolving wild x cultivar hybrids. Ecological Engineering 60:428-437.
Shwartz, L. 2011. The competitive response of Panicum virgatum cultivars to non-native
invasive species in southern Illinois. Southern Illinois University, Carbondale.
Strauss, S. Y., J. A. Lau, and S. P. Carroll. 2006. Evolutionary responses of natives to
introduced species: what do introductions tell us about natural communities?
Ecology letters 9:357-374.
Thelen, G. C., J. M. Vivanco, B. Newingham, W. Good, H. P. Bais, P. Landres, A.
Caesar, and R. M. Callaway. 2005. Insect herbivory stimulates allelopathic
exudation by an invasive plant and the suppression of natives. Ecology Letters
8:209-217.
Thompson, J. N. 1999. Specific hypotheses on the geographic mosaic of coevolution.
American Naturalist 153:S1-S14.
Thompson, J. N., S. L. Nuismer, and R. Gomulkiewicz. 2002. Coevolution and
maladaptation. Integrative and Comparative Biology 42:381-387.
65
Gibson
PhD Dissertation: Chapter 2
2015
Thorpe, A. S., G. C. Thelen, A. Diaconu, and R. M. Callaway. 2009. Root exudate is
allelopathic in invaded community but not in native community: field evidence
for the novel weapons hypothesis. Journal of Ecology 97:641-645.
Tyser, R. W., and C. H. Key. 1988. Spotted knapweed in natural area fescue grasslands an ecological assessment. Northwest Science 62:151-160.
Uriarte, M., C.D. Canham, and R.B. Root. 2002. A model of simultaneous evolution of
competitive ability and herbivore resistance in a perennial plant. Ecology
83:2649-2663.
Weiner, J. 1990. Asymmetric competition in plant populations. Trends in Ecology &
Evolution 5:360-364.
Willis, C. G., M. T. Brock, and C. Weinig. 2010. Genetic variation in tolerance of
competition and neighbour suppression in Arabidopsis thaliana. Journal of
Evolutionary Biology 23:1412-1424.
Zlatnik, E. 1999. Pseudoroegneria spicata. In: Fire Effects Information System [Online].
U.S. Department of Agriculture, Forest Service, Rocky Mountain Research
Station, Fire Sciences Laboratory, http://www.fs.fed.us/database/feis/.
66
Gibson
PhD Dissertation: Chapter 2
2015
Table 1: For 14 wild populations and a cultivar of Pseudoroegneria spicata, collection location, geographic coordinates, and number
of plants used for analysis in the seedling and adult phase in uninvaded control pots and invaded treatment pots. Populations were
either invader experienced (IE) or invader naïve (IN).
Seedling (N = 119)
Coordinates
Control
Adult (N = 119)
Population code
Collection location
Treatment Control
Treatment
IE-BL
Blue Mountain
46.8243
114.0878
5
0
5
3
IN-BL
Blue Mountain
46.8225
114.0904
5
0
5
1
IN-H
Huson
47.0229
114.3661
5
5
5
5
IE-LH
Lincoln Hills
46.8893
113.9476
5
0
5
2
IN-LH
Lincoln Hills
46.8850
113.9501
5
5
5
5
IN-LH#2
Lincoln Hills
46.8995
113.9514
5
3
5
1
IE-LNF
Lolo National Forest
46.8362
113.9702
5
5
5
5
IE-MC
Lower Miller Creek
46.8136
114.0456
5
5
5
3
IN-MC
Lower Miller Creek
46.8136
114.0456
5
3
5
0
IE-MM
Marshall Mountain
46.9002
113.9246
5
5
5
5
IN-MM
Marshall Mountain
46.8965
113.9232
5
5
5
5
67
Gibson
PhD Dissertation: Chapter 2
2015
Seedling (N = 119)
Coordinates
Population code
Collection location
IN-MJ
Mount Jumbo
46.8695
113.9741
5
5
5
2
IE-PT
Pengelly trail
46.8372
113.9784
5
3
5
3
IN-PT
Pengelly trail
46.8367
113.9775
5
5
5
1
IN-Anatone*
Cultivar
117.1322
0
0
5
3
46.1348
Control
Adult (N = 119)
Treatment Control
Treatment
* Due to slow germination, no Anatone plants were available for analysis during the seedling phase. Anatone
coordinates for the the town of Anatone, WA.
68
Gibson
PhD Dissertation: Chapter 2
2015
Table 2: Results of two-way nested ANOVAs testing differences in traits between treatments
(Centaurea stoebe competition), between invader-experience types (experienced vs naïve), and
among 14 Pseudoroegneria spicata populations.
Seedlings (N= 102)
Source
Adults (N= 97)
df
F
P
df
F
P
Treatment
1
7.78
0.021
1
0.93
0.341
Experience
1
0.85
0.660
1
3.86
0.053
Population
12
2.69
0.004
13
4.19
<0.001
Treatment x Experience
1
0.01
0.593
1
5.06
0.027
Treatment
1
12.39
0.001
1
49.88
<0.001
Experience
1
0.91
0.342
1
2.98
0.067
Population
12
3.79
< 0.001
13
5.67
<0.001
Treatment x Experience
1
4.55
0.035
1
0.0001
0.992
Length of longest leaf (cm)
Leaf number
Total biomass (g)
Treatment
1
0.08
0.934
1
5.10
0.026
Experience
1
0.02
0.888
1
1.84
0.175
Population
12
1.52
0.131
13
3.07
0.001
Treatment x Experience
1
0.09
0.766
1
5.19
0.025
Root biomass (g)
Treatment
1
0.51
0.478
1
4.54
0.036
Experience
1
0.13
0.719
1
0.45
0.503
69
Gibson
PhD Dissertation: Chapter 2
2015
Seedlings (N= 102)
Source
df
F
Population
12
Treatment x Experience
Adults (N= 97)
P
df
F
P
1.70
0.077
13
3.13
0.001
1
0.74
0.392
1
5.21
0.025
Treatment
1
1.90
0.171
1
10.09
0.003
Experience
1
5.22
0.024
1
13.10
0.001
Population
12
1.96
0.035
13
5.89
<0.001
Treatment x Experience
1
0.60
0.441
1
2.63
0.108
Treatment
1
100.97
<0.001
1
0.30
0.587
Experience
1
0.74
0.170
1
2.16
0.145
Population
12
14.29
<0.001
13
1.60
0.099
Treatment x Experience
1
5.65
0.019
1
0.07
0.788
Week-1 germination
12
0.15
0.706
Seed weight (g)
12
0.02
0.882
Shoot biomass (g)
Root shoot ratio
70
Gibson
PhD Dissertation: Chapter 2
2015
Table 3: MANOVA results for trait differences between sites, invader-experience type, and
paired invader experienced/naïve populations within sites. Three sites were compared at the
seedling stage (MM, MCL, PT), and two were compared at the adult stage (LH, MM).
Wilks λ
df
F
P
Site
0.446
10,88
4.380
< 0.001
Experience
0.921
5,44
0.751
0.59
Population in site
0.795
10,88
1.067
0.396
Site
0.507
5,29
5.634
< 0.001
Experience
0.913
5,29
0.554
0.734
Population in site
0.927
5,29
0.450
0.802
Factor
Seedling
Adult
71
Gibson
PhD Dissertation: Chapter 2
2015
FIGURE CAPTIONS
Figure 1: Response of Pseudoroegneria spicata seedlings from invader-experienced (IE;
black symbols) and invader-naïve (IN; white symbols) populations when grown alone
(control) or in competition with Centaurea stoebe (invaded) with respect to (a) length of
longest leaf, (b) number of leaves, (c) total biomass, (d) shoot biomass, (e) root biomass,
and (f) root shoot ratio. Error bars represent standard error of the mean.
Figure 2: Response of Pseudoroegneria spicata adult plants from invader-experienced
(IE; black symbols) and invader-naïve (IN; white symbols) populations when grown
alone (control) or in competition with Centaurea stoebe (invaded), with respect to (a)
length of longest leaf, (b) number of leaves, (c) total biomass, (d) shoot biomass, (e) root
biomass, and (f) root shoot ratio. Error bars represent standard error of the mean. *
indicates marginally significant relationships (p < 0.1), and ** indicates significant
relationships (p <0.05) between invader-experience groups within treatment.
Figure 3: Trait differences among populations of Pseudoroegneria spicata at the seedling
stage. Sites are collection locations, and populations were either invader experienced (IE,
black bars) or invader naïve (IN, white bars). Letters represent significant differences (p
= 0.05) among sites and are based on comparisons of sites (MC, MM, PT) with paired IEIN populations. * represent marginally significant differences (p = 0.1). Error bars
represent standard error of the mean.
72
Gibson
PhD Dissertation: Chapter 2
2015
Figure 4: Trait differences among populations of Pseudoroegneria spicata at the adult
stage. Sites are collection locations, and populations were either invader experienced (IE,
black bars) or invader naïve (IN, white bars). Letters represent significant differences (p
= 0.05) among sites (LH, MM) with paired IE-IN populations. Error bars represent
standard error of the mean.
Figure 5: Tolerance of Pseudoroegneria spicata (a) seedlings and (b) adults to Centaurea
stoebe, and (c) ability of adult Pseudoroegneria spicata to suppress C. stoebe by invader
experience type (invader-experienced, IE, and invader-naïve, IN) for populations at 14
sites and the cultivar, Anatone.“Overall” is pooled data by weed experience type; see
Table 1 for site codes. Error bars represent standard error of the mean. Higher tolerance
values indicate less impact of C. stoebe on P. spicata biomass. Higher suppression values
indicate a greater impact of P. spicata on C. stoebe.
Figure 6: Relationship between Pseudoroegneria spicata population tolerance (RIIAdult) and suppression (RII-Knapweed) of Centaurea stoebe for invader-experienced
(IE) and invader-naïve (IN) populations. The symbol for Anatone is labelled. Values
below zero for tolerance indicate that P. spicata biomass was reduced by competition
with C. stoebe, while values above zero indicate higher biomass in the presence of
competition. Suppression values below zero indicate C. stobe biomass had higher
biomass with competition, while values above zero indicate C. stoebe biomass was
reduced by competition with P. spicata. R2=0.421, and p = 0.01.
73
Gibson
PhD Dissertation: Chapter 2
2015
Figure 1
30
(a)
20
19
18
26
24
22
20
18
16
0.70
14
0.24
(d)
(c)
0.22
Shoot biomass (g)
0.65
Biomass (g)
IE
IN
17
16
0.60
0.55
0.50
0.18
0.16
0.12
0.40
2.0
0.20
0.14
0.45
3.0
(e)
1.5
(f)
2.8
1.0
Root shoot ratio
Root biomass (g)
(b)
28
Number of leaves
Length of longest leaf (cm)
21
0.5
0.0
-0.5
2.6
2.4
2.2
2.0
-1.0
-1.5
1.8
Control
Invaded
Treatment
Control
Invaded
Treatment
74
Gibson
PhD Dissertation: Chapter 2
2015
Figure 2
42
(a)
65
*
Number of leaves
Longest leaf (cm)
40
38
36
34
32
45
40
35
1.8
(c)
**
Treatment vs N_Number
Shoot biomass (g)
4.0
3.5
3.0
1.4
1.2
1.0
2.0
0.8
2.8
2.6
*
(e)
(d)
1.6
2.5
2.6
(f)
Treatment vs E_Biomass
Treatment vs N_Biomass
2.4
2.4
Root shoot ratio
Biomass (g)
50
25
Treatment vs E_Number
Root biomass (g)
55
30
30
4.5
IE
IN
(b)
60
2.2
2.0
1.8
2.2
2.0
1.8
1.6
1.4
1.6
Control
Invaded
Treatment
Control
Invaded
Treatment
75
Gibson
PhD Dissertation: Chapter 2
2015
Figure 3
60
50
(b)
(a)
Number of leaves
Longest leaf (cm)
30
20
10
a
40
30
20
0
H
LH
LNF
MC
MJ
MM
PT
b
1.6
H
LH
LNF
MC
MJ
MM
PT
H
LH
LNF
MC
MJ
MM
PT
a
*
MM
PT
0.6
(c)
(d)
1.4
a
0.5
a
Shoot biomass (g)
1.2
1.0
0.8
0.6
0.4
0.4
0.3
0.2
0.1
0.2
0.0
0.0
H
LH
LNF
MC
MJ
MM
PT
1.2
8
(e)
(f)
a
Root shoot ratio
1.0
Root biomass (g)
b
10
0
Biomass (g)
a
50
40
0.8
0.6
0.4
6
4
2
0.2
0.0
0
H
LH
LNF
MC
Site
MJ
MM
PT
H
LH
LNF
MC
MJ
Site
76
Gibson
PhD Dissertation: Chapter 2
2015
Figure 4
50
90
(b)
(a)
a
b
Number of leaves
Longest leaf (cm)
40
30
20
60
30
10
0
0
Ana
BL
H
LH LNF MCL MJ
MM
PT
Ana
6
Shoot biomass (g)
Biomass (g)
H
LH LNF MCL MJ
MM
PT
BL
H
LH LNF MCL MJ
MM
PT
(d)
(c)
5
4
3
2
1.6
1.2
0.8
0.4
1
0
0.0
Ana
BL
H
LH LNF MCL MJ
MM
PT
Ana
4
4.5
(e)
a
(f)
b
3
Root shoot ratio
Root biomass (g)
BL
2.0
2
1
0
3.0
1.5
0.0
Ana
BL
H
LH LNF MCL MJ
Site
MM
PT
Ana BL
H
LH LNF MCL MJ MM PT
Site
77
Gibson
PhD Dissertation: Chapter 2
2015
Figure 5
0.4
Tolerance (RII-Seedling)
(a)
IE
IN
0.2
0.0
-0.2
-0.4
Anatone BL
H
LH
LNF MCL
MJ
MM
PT Overall
H
LH
LNF MCL
MJ
MM
PT Overall
H
LH
LNF MCL
MJ
MM
PT Overall
0.3
(b)
Tolerance (RII-Adult)
0.2
0.1
0.0
-0.1
-0.2
-0.3
-0.4
-0.5
Anatone BL
0.2
Suppression (RII-Knap)
(c)
0.0
-0.2
-0.4
-0.6
-0.8
Anatone BL
Site
78
Gibson
PhD Dissertation: Chapter 2
2015
Figure 6
0.3
IE
IN
0.2
Tolerance
0.1
0.0
-0.1
-0.2
-0.3
-0.4
-0.5
0.2
Anatone
0.0
-0.2
-0.4
-0.6
-0.8
-1.0
Suppression
79
Gibson
PhD Dissertation: Chapter 3
2015
CHAPTER 3
POLYPLOIDY: A MISSING LINK IN THE CONVERSATION ABOUT
RESTORATION OF A COMMONLY SEEDED NATIVE GRASS IN WESTERN
NORTH AMERICA
80
Gibson
PhD Dissertation: Chapter 3
2015
ABSTRACT
The use of local, native plant materials is now common in restoration, but testing for
polyploidy in seed sources is not. Diversity in cytotypes across a landscape can pose
special management challenges, because the methods used to determine genetically
appropriate materials for seed transfer do not account for cytotypic variation. This lack of
consideration may result in mixing cytotypes through revegetation, which could reduce
long-term population viability. I surveyed nine populations of a native bunchgrass,
Pseudoroegneria spicata, in three EPA Level III Ecoregions in the western United States
to determine the frequency of polyploidy, whether there are differences in traits
(phenotype, fecundity, mortality) among plants of different cytotypes or from different
ecoregions, and whether cytotype frequency varies among ecoregions. I assessed trait
variation over two years in a common garden and determined ploidy using flow
cytometry. Four of the nine populations were diploid. The other five had tetraploids
present: three had only tetraploid individuals, while two had mixed diploid/tetraploid
cytotypes. All ecoregions had both cytotypes present. There was significant variation in
traits among cytotypes: plants from tetraploid populations were larger than diploid or
mixed populations. Height, basal circumference, and final biomass also varied by
ecoregion, but diploid, tetraploid, and mixed populations were not consistently different
within ecoregion. The frequency and distribution of cytotypes make it likely that seeding
in the study area will inadvertently introduce new cytotypes into home populations. The
increasing availability of flow cytometry may allow ploidy to be incorporated into native
plant materials sourcing and seed transfer.
81
Gibson
PhD Dissertation: Chapter 3
2015
Keywords: polyploidy, Pseudoroegneria spicata, revegetation, local ecotype, native
plant materials
82
Gibson
PhD Dissertation: Chapter 3
2015
INTRODUCTION
Knowledge of the benefits of native seed sources has led to increased interest in,
and mandates for, using genetically appropriate seeds for revegetation and restoration. To
date, genetic considerations have focused largely on the use of locally adapted native
plant materials (Johnson et al. 2010a). What is genetically appropriate for restoration,
however, may depend on more than local adaptation (Jones 2003; McKay et al. 2005;
Jones 2013). One important genetic factor that has been left out of the conversation is
polyploidy (when an individual possesses extra chromosomes above the diploid level),
despite the fact that it is estimated to occur in up to 80% of angiosperms (Masterson
1994). Polyploidy becomes an especially important genetic consideration for restoration
when species exhibit multiple cytotypes (when individuals of the same species have
different ploidy levels) due to the potential for cross breeding and sterility. There is only
limited information on ploidy level and, even more importantly, cytotypic variation for
most species (Severns & Liston 2008). Furthermore, cytotypic variation is generally not
included in development of seed transfer guidelines, even though restoration choices
regarding seed source have the potential to significantly alter population genetics and
ploidy (Delaney and Baack 2012; Mutegi et al. 2014). Choosing the best plant materials
for restoration requires that managers not only understand which environmental factors
limit seed transfer, but also how cytotypic differences among populations could impact
choices about which populations to combine during revegetation.
Ignoring ploidy level when planning revegetation could have important
consequences for population fitness. For example, crosses between parents that differ in
83
Gibson
PhD Dissertation: Chapter 3
2015
ploidy level can reduce seed set (Burton & Husband 2000). Cross-pollination between
diploid (2x) and tetraploid (4x) plants has been observed to reduce seed set up to 66%
compared to same-cytotype crosses in Ranunculus adoneus (Baack 2005). In addition,
even if seeds are produced, offspring from interploidy crosses may be non-viable or have
reduced fitness (Otto 2007). For instance, triploid offspring (3x) from diploid (2x) and
tetraploid (4x) crosses have been observed to have abnormalities in their seed endosperm
(triploid block; Hakansson & Ellerstrom 1950; Petit, Bretagnolle & Felber 1999; Kohler,
Scheid & Erilova 2010); Ramsey and Schemske (1998) estimated that triploid block
reduces seed set by 34 to 100%. Triploid adults also may be sterile or have limited seed
set (Ramsey & Schemske 1998; Burton & Husband 2000; Baack 2005). Observed
reductions in seed set has been found in interploidy triploid crosses (3x-3x), as well as
when triploids cross with parental cytotypes (Burton & Husband 2001), indicating that
their presence can reduce population viability beyond their initial formation.
Much of the focus on population genetics for revegetation and restoration has
been on developing seed-transfer zones based on the relationship between phenotypic
traits and environmental variables (Johnson et al. 2004; Johnson et al. 2010b; St Clair et
al. 2013). In the absence of species-specific seed transfer zones, factors such as
geographic proximity or the degree of ecological or climatic similarity are used as to
construct provisional transfer boundaries (Jones 2003; Miller et al. 2011; Bower, St Clair
& Erickson 2014). Multiple cytotypes complicate seed transfer because populations
within the same ecological region may vary in ploidy and, therefore, may not be
genetically compatible (McKay et al. 2005). The relationship between seed-transfer
zones and ploidy levels, however, remains unclear. While polyploid cytotypes may be
84
Gibson
PhD Dissertation: Chapter 3
2015
segregated from diploids due to the same kind of ecological constraints that are used to
define seed transfer zones (Raabova, Fischer & Munzbergova 2008; Ramsey 2011),
multiple cytotypes also can occur within the same habitat type or over small spatial scales
(Baack 2004; Duchoslav, Safarova & Krahulec 2010; Sonnleitner et al. 2010; Krejcikova
et al. 2013). Ployploids and diploids that co-occur within the same habitats may not differ
phenotypically (Keeler 2004), which would reduce the chance of polyploid populations
being identified in common garden experiments used to develop seed transfer zones.
Restoration and revegetation have the potential to significantly impact population
genetics through seed source choices. While the frequency of multiple cytotypes has not
been well documented for most species, one recent study conservatively estimated that
13% of California species have multiple ploidy levels (Soltis et al. 2007). In addition,
even if the existence of multiple cytotypes is known, their distribution and frequency may
not be. I assessed ploidal variants of a commonly seeded native graminoid,
Pseudoroegneria spicata (bluebunch wheatgrass), to explore the importance of
considering ploidy when selecting genetically appropriate native plant materials.
Specifically, I assessed the scale of population differentiation and the frequency of
polyploidy in this species, by measuring patterns of genetic and trait differentiation in the
northern part of its range in the western US. Because of its deep and extensive root
system, high seedling vigor, high forage quality and high tolerance of drought, P. spicata
is commonly used for ecological restoration and to stabilize roadsides (Smoliak et al.
2006; Ogle, St John & Jones 2010). The frequency of polyploidy and the spatial
distribution of cytotypes, however, are not known for this species.
85
Gibson
PhD Dissertation: Chapter 3
2015
I addressed the following questions:
1. What is the frequency of polyploidy within P. spicata populations in the
northern part of its range?
2. Are there among-ecoregion differences in cytotype occurrence?
3. Are there phenotypic, fecundity, or mortality differences among
populations of different ploidy levels or from different ecoregions?
Although this study was limited to a single plant species, it serves as a case study for
scientists and managers who are transferring seeds for revegetation and restoration.
MATERIALS AND METHODS
Study Species
Pseudoroegneria spicata (Poaceae) is a native, perennial bunchgrass that is
common throughout the Intermountain West and Great Plains of North America, with a
range stretching from Texas to Alaska. Most populations of P. spicata are diploid (2n =
14), although autotetraploid (4n = 28) populations have been found in the northern part of
the species’ range, and triploid (3n = 21) populations have been found in Washington and
Saskatchewan (Carlson 1986; Larson, Jones & Jensen 2004). Plants can grow in shallow,
stony or sandy soils, as well as south-facing and other dry sites. The species is
outcrossing and wind-pollinated, and the plants often reproduce through tillers (Ogle, St
John & Jones 2010).
Seed Collection
86
Gibson
PhD Dissertation: Chapter 3
2015
In 2009, the Forest Service bulk-collected seeds from 10 P. spicata parent plants
in each nine populations as part of a larger collection (Figure 1), with parent plants
spaced at least 9.3 m apart. A “population” for seed-collection purposes was defined as a
contiguous distribution of individuals (same field or valley). Populations were located in
three EPA Level III Ecoregions: 15 (Northern Rockies; mean elevation of populations=
1077 m, range = 792 – 1417 m), 16 (Idaho Batholith; mean = 1488 m, range = 1463 –
1524 m) and 17 (Middle Rockies; mean = 1769 m, range = 1684 – 1854 m). Ecoregions
are areas that share a common geography, climate, soil type and vegetation type
(Omernik 1987); the contiguous US is divided into 104 of these Level III Ecoregions
(hereafter, “ecoregion”). After collection, seeds were cleaned and stored at the FS nursery
in Coeur d’Alene, Idaho. In 2010, I obtained P. spicata seeds from the USDA Forest
Service (FS) Region 1 Native Plant Materials Program.
Ploidy Determination
Estimation of nuclear DNA content was performed using flow cytometry. I
germinated seeds from the nine source populations in 96-well flats in a heated, unlit
greenhouse on the University of Montana campus (Missoula, Montana). I collected fresh
leaf tissue from nine individuals per population, except for population 1578, for which
only five seeds germinated (N = 77). Leaf samples were wrapped in damp tissue, stored
in plastic bags, and shipped overnight to the Flow Cytometry Core Lab at Benaroya
Research Institute at Virgina Mason Medical Center (Seattle, WA). Flow cytometry was
carried out using 50-mg fresh leaf tissue sliced with a sharp scalpel, using an EPICS
PROFILE flow cytometer 4 (Coulter Electronics, Hialeah, FL) with an argon-ion laser
operating at wavelengths of 488 nm. The protocol used for isolating, staining, and
87
Gibson
PhD Dissertation: Chapter 3
2015
estimating DNA content followed Arumuganathan and Earle (1991). Individuals were
identified as diploid (2x), triploid (3x), or tetraploid (4x).
Common Garden – Trait Variation
Before seeding, I estimated each population’s seed weight by averaging the
weight of 100 seeds. On March 11, 2011, I sowed seeds in a heated, unlit greenhouse on
the University of Montana campus (Missoula, Montana), using Ray Leach conetainers to
facilitate root growth and potting mix (compost, sphagnum peat moss, and perlite). I
sowed four seeds per container, with 100 containers for each of the nine populations. I
randomly placed containers in the greenhouse and moved them every month. After
germination, I randomly thinned seedlings to one per pot. After eight weeks in the
greenhouse, I moved the juvenile plants to a shade house for one week to harden off and
then transplanted seedlings into 30 0.5-m2 plots at the University’s common garden on
May 17, 2011. After planting, I watered all plots daily for two weeks to minimize
mortality and removed weeds from the plots throughout the experiment.
During the 2011 and 2012 growing seasons, I recorded data on phenotypic traits
growth, fecundity, and mortality. Specifically, I recorded plant height (length of longest
leaf) and basal circumference at 15 weeks (June 2011) and at 67 weeks (June 2012) after
germination. I recorded plant mortality at 15, 67, and 75 weeks (September 2012). I also
collected seeds throughout the 2012-growing season. At the end of the experiment (75
weeks; September 2012), I harvested all plants at ground level; plants were then dried
and weighed to determine aboveground biomass.
Data Analysis
88
Gibson
PhD Dissertation: Chapter 3
2015
Since I performed genetic testing on different individuals than those measured for
traits, I analyzed both ploidy level and traits at the population level. First, each population
was categorized as 2x, 4x, or mixed (2x/4x) ploidy. Then, I assessed among-ecoregion
differences in cytotypes by tallying the number of populations within Ecoregions 15, 16,
and 17 that were 2x, 4x, or mixed. I tested for among-ploidy group and -ecoregion
differences in initial seed weight, phenotype (plant height, basal circumference, final
biomass), seed set, and percent mortality using ANOVA models, with separate tests for
each response variable. I was not able to test for interactions between ploidy group and
ecoregion because the two factors were not fully crossed; therefore, to determine if
variation in traits was due to ploidy or ecoregion, I tested for among-ploidy group
differences within each ecoregion using ANOVA models with separate tests for each
response variable. There were not enough replicates to assess differences among ploidy
group within ecoregion for mortality. For traits that had significant differences among
ploidy group, I further explored differences using Tukey’s HSD tests at the p = 0.05
significance level with Bonferroni corrections for multiple comparisons.
All statistical analysis were performed in R (R Core Development Team 2012).
Prior to analyses, final biomass was square root transformed, and initial seed weight and
seed set were log transformed for normalcy. Levene’s test was used to check for
homogeneity of variance among groups. Because plants were organized in a 3 x 5 grid,
the number of neighbors for a plant ranged from three to eight; the significance of
neighbor on the measured traits was analyzed using ANOVA models. Neighbor number
was non-significant for all traits.
89
Gibson
PhD Dissertation: Chapter 3
2015
RESULTS
Ploidy Determination
The overall frequency of polyploidy among study populations was 56%. Three of
the nine populations tested were tetraploid (4x), four were diploid (2x), and two had
mixed 4x and 2x cytotypes (Table 1). Of the mixed populations, one (2013) was 14% 4x
and the other (1766) was 44% 4x. No triploids were identified in any population.
Ecoregion Differences
All three ecoregions had both 2x and 4x cytotypes present, although strictly 4x
populations and mixed populations were each only present in two of the three regions.
Ecoregions had different frequencies of cytoptypes (Figure 2): Ecoregion 15 populations
were predominantly 2x, with an equal number of 4x and mixed populations. Ecoregion
16 populations were predominately 4x, with a lower frequency of 2x and no mixed
populations; and Ecoregion 17 had no 4x populations, although mixed populations were
as prevalent as 2x ones.
Trait Variation
Although there were no significant differences among ploidy groups for initial
seed weight (Table 2), at 15 weeks trait variation was detected among ploidy groups in
both of the phenotypic traits tested (height and basal circumference; Table 2). Tetraploid
(4x) populations were significantly taller than mixed populations (Figure 3A) and had
significantly larger basal circumferences than 2x or mixed populations (Figure 3B). The
trends observed at 15 weeks continued through the next year, and at 67 weeks, all
phenotypic traits differed among ploidy groups (Table 3). Individuals from 2x
90
Gibson
PhD Dissertation: Chapter 3
2015
populations were shorter than those from mixed or 4x populations (Figure 3C), and the
4x populations had the largest basal circumference (Figure 4D).
Both phenotypic traits (height and basal circumference) also varied among
ecoregions (Table 2). At 15 weeks, plants from Ecoregion 15 were the tallest while those
from 17 were the shortest (Figure 4A), and plants from Ecoregion 17 had significantly
smaller basal circumferences than those from 15 or 16 (Figure 4B). At 67 weeks, plants
from Ecoregion 15 continued to be the tallest and have the largest basal circumference,
while those from Ecoregion 17 were the smallest. Plants from Ecoregion 16 showed
intermediate traits to those from the other two regions (Figures 4C, D).
Ploidy groups within ecoregion were not significantly different (Table 3), and
ploidy group ranking in height and circumference changed depending on the trait and
ecoregion. Plants from 4x populations were taller than other ploidy groups from the same
region at both 15 and 67 weeks, although this pattern depended on ecoregion (Figures
4A, C). The mixed population plants were also taller than those from the diploid
population in region 17 at 67 weeks (Figure 4C). There were no differences among
ploidy groups within ecoregion for basal circumference (Figures 4B, D).
There were significant differences among ploidy groups and ecoregion in final
biomass (Table 2). Mixed populations had lower biomass than did 2x or 4x populations
(Figure 5A). There were no significant differences among ploidy groups for seed set
(Figure 5B) or mortality, although there was a non-significant trend towards higher seed
set in 4x populations. Ecoregion 17 plants had lower final biomass than those from
Ecoregions 15 or 16 (Figure 5C). There were no significant differences among
91
Gibson
PhD Dissertation: Chapter 3
2015
ecoregions in seed set (Figure 5D) or mortality. While there were no differences among
ploidy group within ecoregion for biomass (Figure 5C), plants from 4x populations set
more seeds than plants from diploid or mixed populations in Ecoregion 15 (Figure 5D).
DISCUSSION
Adaptive and geographic factors are now commonly considered in the selection of
native plant materials for revegetation and restoration (Hufford & Mazer 2003), but
polyploidy is less frequently examined (Severns, Bradford & Liston 2013) despite the
fact that cytotypes can differ in fitness, and between-cytotype crosses can reduce
population fertility (Burton & Husband 2000; Baack 2005; Gross & Schiestl 2015).
Besides potential issues with population fertility, ploidy can impact community structure
via interactions with pollinators (Segraves & Thompson 1999; Thompson & Merg 2008),
environmental stress tolerance (Bretagnolle & Thompson 1996; Manzaneda et al. 2012),
and even invasiveness (Treier et al. 2009; Pandit, Pocock & Kunin 2011). All of these
factors make polyploidy an important issue to consider when determining which
populations to combine in revegetation and restoration.
The presence of tetraploids of P. spicata in the northern range of the species has
been documented, yet ploidy level has not previously been addressed in recent research
on adaptation and genetically appropriate native plant materials for this species (e.g.,
Herron et al. 2001; Page & Bork 2005; Lesica & Atthowe 2007) and is not mentioned in
commonly available species information sources (Zlatnik 1999; Tilley & St. John 2013).
Despite this lack of attention, I found that polyploidy was common, ranging from 30%
(only 4x) to nearly 60% (4x and mixed) of studied populations. These results highlight
92
Gibson
PhD Dissertation: Chapter 3
2015
the importance of considering cytotype in both the source and target populations before
selecting plant material for revegetation.
There is only limited information on cytotype co-occurrence for most species. For
species that have been studied, the frequency of mixed populations has been found to
vary from typical across the range (Husband 2004; Keeler 2004), to common in contact
zones (Husband & Schemske 1998; Halverson et al. 2008), to rare (Rothera & Davy
1986; Lumaret et al. 1987; Delaney 2012). The frequency of mixed populations, and
where they occur on the landscape, has implications for P. spicata genecology and
management efficacy. Mixed populations for P. spicata in this region appear to be
uncommon, but they occurred across a wide geographic range. It is possible that mixed
populations are more common than my results indicate, given that I only analyzed 10
individuals per population. While this level of replication is not uncommon in cytotype
studies (e.g., Rothera & Davey 1986, Husband & Schemske 1998, Delaney & Baack
2012), increasing sample size could provide a better estimate. It is unclear how the
mixing occurred. It may be the result of natural gene flow; however, it is also possible
that it is a consequence of outplanting seeds, a practice performed on federal and public
lands after fires and road removal, in mine reclamation, and along roadsides (Richards,
Chambers & Ross 1998; Peppin et al. 2011). Using native species in broad-scale seeding
could have substantial benefits for native community establishment (Grant et al. 2011;
Hoelzle, Jonas & Paschke 2012; Hagen et al. 2014), but unless the ploidy of native
materials is considered, it could lead to unintentional mixing of cytotypes and consequent
issues with fertility.
93
Gibson
PhD Dissertation: Chapter 3
2015
The differences in occurrence among species and locations can depend on
variation in fitness between cytotypes – mixed ploidy populations are unstable when
there are fitness differences between cytotypes (Baack 2004; Baack & Stanton 2005;
Buggs & Pannell 2007) leading populations to eventually fix at one cytotype (e.g.,
Ramsey and Schemske 1998). The low occurrence of mixed-cytotype populations of P.
spicata in this region could be an indication that there are fitness differences between
tetraploids and diploids that eventually lead to single-cytotype populations. One
management concern related to mixing cytotypes is that triploid formation in diploidtetraploid crosses can reduce population fertility through triploid block and low fitness
adults. The fact that I did not find any triploid individuals may be an indication of a
triploid block effect. These results are consistent with research in other species, which
found that triploid adults occur at low frequencies in wild populations (Petit & Thompson
1997; Burton & Husband 2001; Sonnleitner et al. 2010), and that triploid formation in
diploid-tetraploid crosses is a rare event. Triploid block can significantly reduce seed set
(Ramsey & Schemske 1998), which could impact long-term population viability; the
greater the probability of reduction in seed set caused by interploidy crosses, the more
cautious restorationists should be about combining P. spicata cytotypes.
The commonality of tetraploids and the presence of mixed-ploidy populations
across the three ecoregions surveyed indicate that P. spicata cytotypes are not segregated
by the types of broad environmental differences used to develop seed transfer zones. Both
provisional and species-specific seed transfer guidelines link genetic differences to broad
environmental drivers, and then use environmental boundaries to delineate which
populations can be mixed in seed transfer. The co-occurrence of diploid, tetraploid and
94
Gibson
PhD Dissertation: Chapter 3
2015
mixed-cytotype populations over small geographic scales would make it unlikely that
either provisional or species-specific seed transfer zones would prevent mixing cytotypes
during restoration (Severns, Bradford & Liston 2013). While I did not find support for
large-scale environmental differences in cytotype occurrence, variation in cytotype
occurrence can be driven by small-scale habitat differences like anthropogenic
disturbance (Mraz et al. 2012), vegetative cover (Sonnleitner et al. 2010), and plant
community composition (Johnson, Husband & Burton 2003); in addition, polyploidy can
allow cytotypes to colonize novel habitats (Ramsey 2011; Martin & Husband 2013). I did
not find that P. spicata cytotype occurrence differed due to broad ecological differences,
but the observation of tetraploid-only populations suggests that there could be small-scale
differences in the habitat occupied by P. spicata cytotypes. Furthermore, the high
occurrence of single-cytotype populations could indicate that tetraploids have higher
fitness than diploids in certain habitats. If both cytotypes showed equal fitness in all
habitats, then mixed populations would be more common across the landscape.
Increases in chromosomal number can lead to phenotypic differences among
cytotypes within a species (Balao, Herrera & Talavera 2011). While previous
investigators did not find phenotypic differences in traits (Carlson 1986; Jones & Larson
2005), I observed phenotypic variation among ploidy groups at all assessed life stages.
Polyploid cytotypes are often larger (Vichiato et al. 2014) and more competitive than
diploids (Thebault et al. 2011), and P. spicata plants from tetraploid populations were
larger than those from diploid or mixed populations; although it was not significant, there
was also trend towards them having higher final biomass and seed set than diploid
populations. In this experiment, mixed populations were not consistently different from
95
Gibson
PhD Dissertation: Chapter 3
2015
either diploid or tetraploid populations, and instead tended to be closer to one or the other
depending on the trait. Since there were only two mixed populations, however, this could
be due to chance rather than an indication of mixed population traits.
Traits varied among-cytotypes, but Level III Ecoregion appears to play a more
important role in influencing traits in P. spicata than cytotype. Plants from Ecoregions 15
and 16 were larger than those from Ecoregion 17, and those two ecoregions contained all
three of the observed tetraploid populations. There were fewer differences among ploidygroups when I compared them within ecoregion, although plants from the tetraploid
population in Ecoregion 15 were taller and had higher seed set than those from the
diploid or mixed populations. Having so few replicates of each population type (diploid,
tetraploid, or mixed) within an ecoregion, however, means observed traits could be due to
population differences and not cytotype. Future research should assess ploidy and traits in
more populations per ecoregion to better compare the impact of cytotype and ecoregion
on traits.
Although I did not find significant between-cytoype differences in seed set, this
experiment did not explicitly test for fitness differences. Other measures of fecundity –
including seed viability and number of emerging seedlings – may be impacted by ploidy
(Burton & Husband 2001; Keeler 2004; Rieseberg & Willis 2007). In addition,
polyploids often show differences in traits related to fitness like stomatal size and water
use efficiency (Khazaei et al. 2010; Manzaneda et al. 2012; Mraz, Tarbush & MullerScharer 2014). Furthermore, not only can there be fitness differences between cytotypes,
but fitness and morphological trait differences among cytotypes may be so great that they
96
Gibson
PhD Dissertation: Chapter 3
2015
lead to to invasiveness (Prentis et al. 2008; Speek et al. 2011; Suda et al. 2015).
Managers should consider the potential negative population fitness impacts of combining
cytotypes, as well as the possibility that P. spicata tetraploids could be more competitive
than diploids, when choosing a seed source for restoration.
Conclusion: For land managers, flow cytometry offers a quick, easy way to assess
the ploidy of target populations and seed source materials. While common gardens have
traditionally been used to develop seed transfer guidelines, it could be difficult to
impossible to determine cytotype differences from morphological traits alone. Given the
co-occurrence of P. spicata cytotypes in northern ecoregions, it is unlikely that using
Level III Ecoregions as provisional transfer zones would prevent mixing during
revegetation, and species-specific transfer zones would likely be too complex for easy
management decisions. The ploidy level of target populations and seed sources should
receive as much attention as local adaptation when deciding what constitutes genetically
appropriate plant materials. Due to reproductive barriers between diploids and tetraploids,
mixing cytotypes through revegetation may be less desirable than bringing in non-local
seeds of the same ploidy (Jones 2003).
ACKNOWLEDGEMENTS
This research was supported by agreement #09-CS-11015600-031 from the USDA Forest
Service NFN program and the National Science Foundation EPSCoR program (# EPS1101342) at the University of Montana.
LITERATURE CITED
97
Gibson
PhD Dissertation: Chapter 3
2015
Arumuganathan, K. & Earle, E. (1991) Nuclear DNA content of some important plant
species. Plant Molecular Biology Reporter, 9, 208 - 213.
Baack, E.J. (2004) Cytotype segregation on regional and microgeographic scales in snow
buttercups (Ranunculus adoneus: Ranunculaceae). American Journal of Botany,
91, 1783-1788.
Baack, E.J. (2005) Ecological factors influencing tetraploid, establishment in snow
buttercups (Ranunculus adoneus, Ranunculaceae): minority cytotype exclusion
and barriers to triploid formation. American Journal of Botany, 92, 1827-1835.
Baack, E.J. & Stanton, M.L. (2005) Ecological factors influencing tetraploid speciation in
snow buttercups (Ranunculus adoneus): niche differentiation and tetraploid
establishment. Evolution, 59, 1936-1944.
Balao, F., Herrera, J., & Talavera, S. (2011) Phenotypic consequences of polyploidy and
genome size at the microevolutionary scale: a multivariate morphological
approach. New Phytologist, 192, 256-265.
Bower, A.D., St Clair, B. & Erickson, V. (2014) Generalized provisional seed zones for
native plants. Ecological Applications, 24, 913-919.
Bretagnolle, F. & Thompson, J.D. (1996) An experimental study of ecological
differences in winter growth between sympatric diploid and autotetraploid
Dactylis glomerata. Journal of Ecology, 84, 343-351.
Buggs, R.J.A. & Pannell, J.R. (2007) Ecological differentiation and diploid superiority
across a moving ploidy contact zone. Evolution, 61, 125-140.
98
Gibson
PhD Dissertation: Chapter 3
2015
Burton, T.L. & Husband, B.C. (2000) Fitness differences among diploids, tetraploids, and
their triploid progeny in Chamerion angustifolium: mechanisms of inviability and
implications for polyploid evolution. Evolution, 54, 1182-1191.
Burton, T.L. & Husband, B.C. (2001) Fecundity and offspring ploidy in matings among
diploid, triploid and tetraploid Chamerion angustifolium (Onagraceae):
consequences for tetraploid establishment. Heredity, 87, 573-582.
Carlson, J. (1986) A study of morphological variation within Pseudoroegneria spicata
(Pursh) A. Love (Poaceae: Triticeae). Masters, Oregon State University.
Delaney, J.T. & Baack, E.J. (2012) Intraspecific chromosome number variation and
prairie restoration—a case study in northeast Iowa, U.S.A. Restoration Ecology,
20, 576-583.
Duchoslav, M., Safarova, L. & Krahulec, F. (2010) Complex distribution patterns,
ecology and coexistence of ploidy levels of Allium oleraceum (Alliaceae) in the
Czech Republic. Annals of Botany, 105, 719-735.
Grant, A.S., Nelson, C.R., Switalski, T.A. & Rinehart, S.M. (2011) Restoration of native
plant communities after road decommissioning in the Rocky Mountains: effect of
seed-mix composition on vegetative establishment. Restoration Ecology, 19, 160169.
Gross, K. & Schiestl, F. (2015) Are tetraploids more successful? Floral signals,
reproductive success and floral isolation in mixed-ploidy populations of a
terrestrial orchid. Annals of Botany, 115, 263 - 273.
99
Gibson
PhD Dissertation: Chapter 3
2015
Hagen, D., Hansen, T.I., Graae, B.J. & Rydgren, K. (2014) To seed or not to seed in
alpine restoration: introduced grass species outcompete rather than facilitate
native species. Ecological Engineering, 64, 255-261.
Hakansson, A. & Ellerstrom, S. (1950) Seed development after reciprocal crosses
between diploid and tetraploid rye. Hereditas, 36, 256-296.
Halverson, K., Heard, S.B., Nason, J.D. & Stireman, J.O. (2008) Origins, distribution,
and local co-occurrence of polyploid cytotypes in Solidago altissima
(Asteraceae). American Journal of Botany, 95, 50-58.
Herron, G.J., Sheley, R.L., Maxwell, B.D. & Jacobsen, J.S. (2001) Influence of nutrient
availability on the interaction between spotted knapweed and bluebunch
wheatgrass. Restoration Ecology, 9, 326-331.
Hoelzle, T.B., Jonas, J.L. & Paschke, M.W. (2012) Twenty-five years of sagebrush
steppe plant community development following seed addition. Journal of Applied
Ecology, 49, 911-918.
Hufford, K.M. & Mazer, S.J. (2003) Plant ecotypes: genetic differentiation in the age of
ecological restoration. Trends in Ecology & Evolution, 18, 147-155.
Husband, B.C. (2004) The role of triploid hybrids in the evolutionary dynamics of mixedploidy populations. Biological Journal of the Linnean Society, 82, 537-546.
Husband, B.C. & Schemske, D.W. (1998) Cytotype distribution at a diploid-tetraploid
contact zone in Chamerion (Epilobium) angustifolium (Onagraceae). American
Journal of Botany, 85, 1688-1694.
Johnson, G., Sorensen, F., St Clair, J. & Cronn, R. (2004) Pacific Northwest forest tree
seed zones: a template for native plants? Native Plants, 5, 131 - 140.
100
Gibson
PhD Dissertation: Chapter 3
2015
Johnson, M.T.J., Husband, B.C. & Burton, T.L. (2003) Habitat differentiation between
diploid and tetraploid Galax urceolata (Diapensiaceae). International Journal of
Plant Sciences, 164, 703-710.
Johnson, R., Stritch, L., Olwell, P., Lambert, S., Horning, M.E. & Cronn, R. (2010a)
What are the best seed sources for ecosystem restoration on BLM and USFS
lands? Native Plants, 11, 117-131.
Johnson, R.C., Erickson, V.J., Mandel, N.L., St Clair, J.B. & Vance-Borland, K.W.
(2010b) Mapping genetic variation and seed zones for Bromus carinatus in the
Blue Mountains of eastern Oregon, USA. Botany-Botanique, 88, 725-736.
Jones, T. & Larson, S. (2005) Status and use of important native grasses adapted to
sagebrush communities. USDA Forest Service Proceedings RMRS-P-38. (ed.
U.D.o.A.F. Service), pp. 49 - 55. Fort Collins, CO.
Jones, T.A. (2003) The restoration gene pool concept: beyond the native versus nonnative debate. Restoration Ecology, 11, 281-290.
Jones, T.A. (2013) When local isn't best. Evolutionary Applications, 6, 1109-1118.
Keeler, K.H. (2004) Impact of intraspecific polyploidy in Andropogon gerardii (Poaceae)
populations. American Midland Naturalist, 152, 63-74.
Khazaei, H., Monneveux, P., Shao, H.B. & Mohammady, S. (2010) Variation for
stomatal characteristics and water use efficiency among diploid, tetraploid and
hexaploid Iranian wheat landraces. Genetic Resources and Crop Evolution, 57,
307-314.
Kohler, C., Scheid, O.M. & Erilova, A. (2010) The impact of the triploid block on the
origin and evolution of polyploid plants. Trends in Genetics, 26, 142-148.
101
Gibson
PhD Dissertation: Chapter 3
2015
Krejcikova, J., Sudova, R., Oberlander, K., Dreyer, L.L. & Suda, J. (2013) The spatioecological segregation of different cytotypes of Oxalis obtusa (Oxalidaceae) in
contact zones. South African Journal of Botany, 88, 62-68.
Larson, S.R., Jones, T.A. & Jensen, K.B. (2004) Population structure in Pseudoroegneria
spicata (Poaceae : Triticeae) modeled by Bayesian clustering of AFLP genotypes.
American Journal of Botany, 91, 1789-1801.
Lesica, P. & Atthowe, H.E. (2007) Identifying weed-resistant bluebunch wheatgrass for
restoration in western Montana. Ecological Restoration 25, 191-198.
Lumaret, R., Guillerm, J.L., Delay, J., Loutfi, A.A.L., Izco, J. & Jay, M. (1987)
Polyploidy and habitat differentiation in Dactylis glomerata from Galicia (Spain).
Oecologia, 73, 436-446.
Manzaneda, A.J., Rey, P.J., Bastida, J.M., Weiss-Lehman, C., Raskin, E. & MitchellOlds, T. (2012) Environmental aridity is associated with cytotype segregation and
polyploidy occurrence in Brachypodium distachyon (Poaceae). New Phytologist,
193, 797-805.
Martin, S.L. & Husband, B.C. (2013) Adaptation of diploid and tetraploid Chamerion
angustifolium to elevation but not local environment. Evolution, 67, 1780-1791.
Masterson, J. (1994) Stomatal size in fossil plants - evidence for polyploidy in majority
of angiosperms. Science, 264, 421-424.
McKay, J.K., Christian, C.E., Harrison, S. & Rice, K.J. (2005) "How local is local?" - A
review of practical and conceptual issues in the genetics of restoration.
Restoration Ecology, 13, 432-440.
102
Gibson
PhD Dissertation: Chapter 3
2015
Miller, S.A., Bartow, A., Gisler, M., Ward, K., Young, A.S. & Kaye, T.N. (2011) Can an
ecoregion serve as a seed transfer zone? Evidence from a common garden study
with five native species. Restoration Ecology, 19, 268-276.
Mraz, P., Spaniel, S., Keller, A., Bowmann, G., Farkas, A., Singliarova, B., Rohr, R.P.,
Broennimann, O. & Muller-Scharer, H. (2012) Anthropogenic disturbance as a
driver of microspatial and microhabitat segregation of cytotypes of Centaurea
stoebe and cytotype interactions in secondary contact zones. Annals of Botany,
110, 615-627.
Mraz, P., Tarbush, E. & Muller-Scharer, H. (2014) Drought tolerance and plasticity in the
invasive knapweed Centaurea stoebe s.l. (Asteraceae): effect of populations
stronger than those of cytotype and range. Annals of Botany, 114, 289-299.
Mutegi, E., Stottlemyer, A.L., Snow, A.A. & Sweeney, P.M. (2014) Genetic structure of
remnant populations and cultivars of switchgrass (Panicum virgatum) in the
context of prairie conservation and restoration. Restoration Ecology, 22, 223-231.
Ogle, D., St John, L. & Jones, T. (2010) Plant guide for bluebunch wheatgrass
(Pseudoroegneria spicata). USDA-Natural Resource Conservation Service. Idaho
and Washington Plant Materials Program, http://www.plantmaterials.nrcs.usda.gov/pubs/idpmspg9765.pdf.
Omernik, J.M. (1987) Ecoregions of the conterminous United States. Annals of the
Association of American Geographers, 77, 118-125.
Otto, S.P. (2007) The evolutionary consequences of polyploidy. Cell, 131, 452-462.
103
Gibson
PhD Dissertation: Chapter 3
2015
Page, H.N. & Bork, E.W. (2005) Effect of planting season, bunchgrass species, and
neighbor control on the success of transplants for grassland restoration.
Restoration Ecology, 13, 651-658.
Pandit, M.K., Pocock, M.J.O. & Kunin, W.E. (2011) Ploidy influences rarity and
invasiveness in plants. Journal of Ecology, 99, 1108-1115.
Peppin, D.L., Fule, P.Z., Sieg, C.H., Beyers, J.L., Hunter, M.E. & Robichaud, P.R.
(2011) Recent trends in post-wildfire seeding in western US forests: costs and
seed mixes. International Journal of Wildland Fire, 20, 702-708.
Petit, C., Bretagnolle, F. & Felber, F. (1999) Evolutionary consequences of diploidpolyploid hybrid zones in wild species. Trends in Ecology & Evolution, 14, 306311.
Petit, C. & Thompson, J.D. (1997) Variation in phenotypic response to light availability
between diploid and tetraploid populations of the perennial grass Arrhenatherum
elatius from open and woodland sites. Journal of Ecology, 85, 657-667.
Prentis, P.J., Wilson, J.R.U., Dormontt, E.E., Richardson, D.M. & Lowe, A.J. (2008)
Adaptive evolution in invasive species. Trends in Plant Science, 13, 288-294.
Raabova, J., Fischer, M. & Munzbergova, Z. (2008) Niche differentiation between
diploid and hexaploid Aster amellus. Oecologia, 158, 463-472.
Ramsey, J. (2011) Polyploidy and ecological adaptation in wild yarrow. Proceedings of
the National Academy of Sciences of the United States of America, 108, 70967101.
104
Gibson
PhD Dissertation: Chapter 3
2015
Ramsey, J. & Schemske, D.W. (1998) Pathways, mechanisms, and rates of polyploid
formation in flowering plants. Annual Review of Ecology and Systematics, 29,
467-501.
Richards, R.T., Chambers, J.C. & Ross, C. (1998) Use of native plants on federal lands:
policy and practice. Journal of Range Management, 51, 625-632.
Rieseberg, L.H. & Willis, J.H. (2007) Plant speciation. Science, 317, 910-914.
Rothera, S.L. & Davy, A.J. (1986) Polyploidy and habitat differentiation in Deschampsia
cespitosa. New Phytologist, 102, 449-467.
Segraves, K.A. & Thompson, J.N. (1999) Plant polyploidy and pollination: floral traits
and insect visits to diploid and tetraploid Heuchera grossulariifolia. Evolution,
53, 1114-1127.
Severns, P.M., Bradford, E. & Liston, A. (2013) Whole genome duplication in a
threatened grassland plant and the efficacy of seed transfer zones. Diversity and
Distributions, 19, 455-464.
Severns, P.M. & Liston, A. (2008) Intraspecific chromosome number variation: a
neglected threat to the conservation of rare plants. Conservation Biology, 22,
1641-1647.
Smoliak, S., Ditterline, R., Scheetz, J., Holzworth, L., Sims, J., Wiesner, L., Balderidge,
D. & Tiblie, G. (2006) Bluebunch wheatgrass. Montana Interagency Plant
Materials Handbook.
http://animalrangeextension.montana.edu/articles/forage/species/grasses/Bluebunc
h-wheatgrass.htm.
105
Gibson
PhD Dissertation: Chapter 3
2015
Soltis, D.E., Soltis, P.S., Schemske, D.W., Hancock, J.F., Thompson, J.N., Husband,
B.C. & Judd, W.S. (2007) Autopolyploidy in angiosperms: have we grossly
underestimated the number of species? Taxon, 56, 13-30.
Sonnleitner, M., Flatscher, R., Garcia, P.E., Rauchova, J., Suda, J., Schneeweiss, G.M.,
Hulber, K. & Schonswetter, P. (2010) Distribution and habitat segregation on
different spatial scales among diploid, tetraploid and hexaploid cytotypes of
Senecio carniolicus (Asteraceae) in the Eastern Alps. Annals of Botany, 106, 967977.
Speek, T.A.A., Lotz, L.A.P., Ozinga, W.A., Tamis, W.L.M., Schaminee, J.H.J. & van der
Putten, W.H. (2011) Factors relating to regional and local success of exotic plant
species in their new range. Diversity and Distributions, 17, 542-551.
St Clair, J.B., Kilkenny, F.F., Johnson, R.C., Shaw, N.L. & Weaver, G. (2013) Genetic
variation in adaptive traits and seed transfer zones for Pseudoroegneria spicata
(bluebunch wheatgrass) in the northwestern United States. Evolutionary
Applications, 6, 933-948.
Suda, J., Meyerson, L.A., Leitch, I.J. & Pysek, P. (2015) The hidden side of plant
invasions: the role of genome size. New Phytologist, 205, 994-1007.
R Core Development Team (2012) R: A language and environment for statistical
computing. R Foundation for Statistical Computing, Vienna, Austria.
Thebault, A., Gillet, F., Muller-Scharer, H. & Buttler, A. (2011) Polyploidy and invasion
success: trait trade-offs in native and introduced cytotypes of two Asteraceae
species. Plant Ecology, 212, 315-325.
106
Gibson
PhD Dissertation: Chapter 3
2015
Thompson, J.N. & Merg, K.F. (2008) Evolution of polyploidy and the diversification of
plant-pollinator interactions. Ecology, 89, 2197-2206.
Tilley, D. & St. John, L. (2013) Plant fact sheet for bluebunch wheatgrass
(Pseudoroegneria spicata). USDA-Natural Resources Conservation Service,
Aberdeen Plant Materials Center,
http://plants.usda.gov/factsheet/pdf/fs_pssp6.pdf.
Treier, U.A., Broennimann, O., Normand, S., Guisan, A., Schaffner, U., Steinger, T. &
Mueller-Schaerer, H. (2009) Shift in cytotype frequency and niche space in the
invasive plant Centaurea maculosa. Ecology, 90, 1366-1377.
Vichiato, M.R.D., Vichiato, M., Pasqual, M., Rodrigues, F.A. & de Castro, D.M. (2014)
Morphological effects of induced polyploidy in Dendrobium nobile Lindl.
(Orchidaceae). Crop Breeding and Applied Biotechnology, 14, 154-159.
Zlatnik, E. (1999) Pseudoroegneria spicata. In: Fire Effects Information System
[Online]. U.S. Department of Agriculture, Forest Service, Rocky Mountain
Research Station, Fire Sciences Laboratory, http://www.fs.fed.us/database/feis/.
107
Gibson
PhD Dissertation: Chapter 3
2015
Table 1: Nuclear DNA content (picograms) for nine populations of P. spicata collected in
Level III Ecoregion 15, 16, and 17. Population ploidy is either diploid (2x), tetraploid
(4x), or mixed with individuals of both cytotype (2x/4x).
No. plants
DNA content (pg/2C)
Mean
St. Dev.
Ecoregion
Population
Ploidy
15
1764
9
14.6
0.377
4x
15
1766
9
11.59
0.281
2x/4x
15
2028
9
8.35
0.16
2x
15
2029
9
7.76
0.236
2x
16
1578
5
13.97
0.258
4x
16
1886
9
14.84
0.501
4x
16
1926
9
7.79
0.092
2x
17
2010
9
7.59
0.196
2x
17
2013
9
8.95
0.254
2x/4x
108
Gibson
PhD Dissertation: Chapter 3
2015
Table 2: Statistics (df, F, and p) for tests of difference among ploidy groups (4x, 2x, and
2x/4x) and Ecoregions (15, 16, and 17) for initial seed weight, 15- and 67-week
phenotypic traits (height, basal circumference, final biomass), seed set, and percent
mortality in Pseudoroegneria spicata.
Ploidy
Ecoregion
Trait
df
F
p
df
F
p
Initial seed weight
2
0.438
0.438
2
1.891
0.218
15 week height
2
4.229
0.016
2
29.685 < 0.001
15 week circumference
2
9.752 < 0.001
2
28.845 < 0.001
67 week height
2
7.495
<0.001
2
5.013
67 week circumference
2
9.752 < 0.001
2
18.095 < 0.001
Final biomass
2
5.913
0.003
2
12.35
<0.001
Seed set
2
0.44
0.645
2
1.309
0.274
Mortality
2
0.297
0.758
2
0.41
0.689
Phenotypic traits
0.007
109
Gibson
PhD Dissertation: Chapter 3
2015
Table 3: Statistics (df, F, and p) for tests for differences among ploidy groups (2x, 4x and
2x/4x) within ecoregions 15, 16, and 17 for 15- and 67-week phenotypic traits (height,
basal circumference, final biomass) and seed set.
15
16
Trait
df
F
p
df
15-week height
2
2.598
0.079
1
Residual
99
15-week circumference
2
Residual
99
67-week height
2
Residual
87
67-week circumference
2
Residual
87
Final biomass
2
Residual
86
Seed set
2
Residual
55
F
17
p
7.031 0.009
87
2.137
0.123
1
0.0482
1
1.107 0.296
0.807
1
0.663 0.418
0.386
1
0.322 0.572
1
50
0.436
0.512
0.706
0.405
6.61
0.014
0.269
0.607
0.288
0.595
0.02
0.89
1
1
1
40
0.606 0.439
74
3.579 0.0346
1
40
74
0.963
p
47
74
0.215
F
47
87
3.14
df
1
38
1.092 0.301
1
25
110
Gibson
PhD Dissertation: Chapter 3
2015
FIGURE CAPTIONS
Figure 1: Location of the nine Pseudoroegneria spicata study populations. Shaded areas
represent approximate locations of Level III Ecoregions 15 (Northern Rockies), 16 (Idaho
Batholith) and 17 (Middle Rockies).
Figure 2: Frequency (percent) of diploid (2x), tetraploid (4x) and mixed (4x/2x)
populations of Pseudoroegneria spicata in Ecoregions 15, 16, and 17.
Figure 3: Differences in growth traits among ploidy groups (2x =diploid; M = mixed;
4x=tetraploid) for 15-weeks (A and B) and 67-weeks (C and D) after germination. Error
bars represent one standard error. Lower case letters represent significant differences
among groups at p = 0.05.
Figure 4: Differences in growth traits among-ploidy group (2x =diploid [black bars]; M =
mixed [gray bars]; 4x=tetraploid [white bars]), and among-ploidy group within ecoregion
15, 16, and 17 at 15-weeks (A and B) and 67-weeks (C and D) after germination. Error
bars represent one standard error. Lower case letters represent significant differences
among groups at p = 0.05. Letters above the lines represent differences among
ecoregions; letters below bars represent differences among ploidy groups within
ecoregion.
Figure 5: Differences in final biomass (A and B) and seed set (C and D) among-ploidy
group (2x =diploid [black bars]; M = mixed [gray bars]; 4x=tetraploid [white bars]), and
among-ploidy group within ecoregion 15, 16, and 17. Error bars represent one standard
error. Lower case letters represent significant differences among groups at p = 0.05.
111
Gibson
PhD Dissertation: Chapter 3
2015
Letters above the lines represent differences among ecoregions; letters below bars
represent differences among ploidy groups within ecoregion.
112
Gibson
PhD Dissertation: Chapter 3
2015
Figure 1
113
Gibson
PhD Dissertation: Chapter 3
2015
Figure 2
n=4
n=3
n=2
15
16
17
100
2x
4x
Mixed
Frequency
80
60
40
20
0
Ecoregion
114
Gibson
PhD Dissertation: Chapter 3
2015
Figure 3
15 weeks
20
67 weeks
60
b
A)
ab
C)
a
b
b
a
50
Height (cm)
Height (cm)
15
10
40
30
20
5
10
0
0
25
B)
b
25
a
a
20
15
10
5
Basal circumference (cm)
Basal circumference (cm)
30
20
D)
b
a
a
15
10
5
0
0
2x
M
Ploidy group
4x
2x
M
4x
Ploidy group
115
Gibson
PhD Dissertation: Chapter 3
2015
Figure 4
67 weeks
15 weeks
30
A)
C)
b
a
c
25
a
a
20
c
b
b
b
a
a
a
a
a
b
a
15
a
a
10
Height (cm)
Height (cm)
a
60
a
40
20
5
0
0
30
30
a
a
b
20
10
0
Basal circumference (cm)
Basal circumference (cm)
B)
D)
a
b
15
16
c
20
10
0
15
16
Ecoregion
17
17
Ecoregion
116
Gibson
PhD Dissertation: Chapter 3
2015
Figure 5
20
20
C)
a
15
a
b
10
5
Final biomass (g)
Final biomass (g)
A)
0
300
b
15
10
5
0
600
B)
D)
250
b
500
Seed set (number)
Seed set (number)
a
a
200
150
100
50
400
a
300
200
a
a
a
a
a
100
0
0
2x
M
Ploidy group
4x
15
16
17
Ecoregion
117
Gibson
PhD Dissertation: Chapter 4
2015
CHAPTER 4
COMPARING PROVISIONAL SEED TRANSFER ZONE STRATEGIES FOR A
COMMONLY SEEDED GRASS, PSEUDOROEGNERIA SPICATA
118
Gibson
PhD Dissertation: Chapter 4
2015
ABSTRACT
Restoration practitioners need to balance the desire to use locally adapted plant materials
with the uncertainty of what constitutes local. Provisional seed transfer zones are
intended to guide managers on how far plant materials can be moved during revegetation
in the absence of species-specific seed transfer zones, with the assumption that all
populations within a zone will show similarly adapted traits. There are multiple
provisional transfer guidelines, including population-specific, within Level III Ecoregion
or climate boundary, or climate within ecoregion. There is little information about which
are best, and which best reflects population differentiation could vary by region. I used
Pseudoroegneria spicata as a test species to assess (1) whether EPA Level III Ecoregion
or population explained more variation in traits; and (2) which of four common
approaches to developing provisional seed transfer zones (Population-only, Ecoregiononly, Climate-only, and Ecoregion + climate) best explains trait variation. Ecoregion
explained more variation than population in most growth traits and mortality; population
explained more variation only in seed set. There were not strong differences in traits
between two of the three ecoregions. Plants from the home ecoregion showed the
smallest size and seed set, and the highest mortality. Out of the four models, the
Ecoregion + climate model best predicted growth traits, while the Climate-only model
best predicted seed set and mortality. Findings suggest that Level III Ecoregions may be
too broad for seed transfer in this region and that managers should consider climate
within ecoregion when making seed transfer decisions.
Keywords: seed transfer, Pseudoroegneria spicata, ecoregion, population differentiation
119
Gibson
PhD Dissertation: Chapter 4
2015
INTRODUCTION
Knowledge of the importance of local selective pressures has led to increased
interest in using local ecotypes for revegetation and restoration to ensure that native plant
materials will be site-adapted (Joshi et al. 2001, Vander Mijnsbrugge et al. 2010) and to
reduce outbreeding depression and genetic swamping in natural populations (Hufford and
Mazer 2003). Utilizing local seed in restoration requires answering “how local is local?”
(McKay et al. 2005). The answer to this question varies by species and, therefore, there is
need for species-specific seed transfer zones (i.e. areas within which plant materials can
be moved with limited risk of maladaptation). Although the number of species-specific
seed transfer zones has increased (Wilson et al. 2001, Wilson et al. 2008, Johnson et al.
2010b, Johnson et al. 2013, St Clair et al. 2013), seed transfer zones have not been
developed for most species used for revegetation and restoration because of the time and
effort required to develop them. Thus, managers often rely on provisional transfer zones.
As opposed to species-specific zones, which are the result of empirical genecological
studies (Johnson et al. 2004), provisional zones are based on assumptions about the
environmental and ecological factors that are generally important across all species.
Provisional seed transfer zones guide managers in determining the extent to which
populations are combined and, therefore, have a substantive effect on revegetation
success and population viability, as well as the economic feasibility of using locally
collected seeds.
There are several suggested guidelines for developing provisional seed transfer
zones that range from very small spatial scales (population-specific) to very broad scales
(across entire ecological regions), each with a different set of assumptions about the
120
Gibson
PhD Dissertation: Chapter 4
2015
factors that drive population differentiation. The most conservative option is to use seeds
collected from the nearest sources to the site (Saari and Glisson 2012), and populationspecific seed collection has been recommended to reduce the chance of outbreeding
depression (Knapp and Rice 1994). While moving seeds over short distances can protect
locally adapted genotypes, using such limited areas for seed transfer can reduce genetic
diversity in restored populations (Broadhurst et al. 2008, Bischoff et al. 2010, Godefroid
et al. 2011, Lankau et al. 2011), especially if plant materials are from isolated or small
populations that suffer from inbreeding depression or founder effects (Ellstrand and Elam
1993, Keller and Waller 2002). In addition, this strategy is likely to be overly cautious
given evidence that broad-scale factors often drive population differentiation (Becker et
al. 2006, Macel et al. 2007, Weisshuhn et al. 2012). Finally, the use of strictly local seeds
is difficult to implement in large management activities (Merritt and Dixon 2011)
because seed collection from local sites may be limited by the population size or the
genetic quality of remnant populations (Broadhurst et al. 2006, Broadhurst et al. 2008,
Borders et al. 2011).
While population-specific guidelines are used by some managers, the most
commonly recommended provisional transfer zones are the much larger EPA Level III
Ecoregions (Jones and Larson 2005, Shaw et al. 2005, Johnson et al. 2010a). Ecoregions
range from very broad (level I) to fine scales (level IV) and were created by dividing the
US into areas of similar land-use, climate, geography, soils, and vegetation (Omernik
1987); the importance of each factor in determining ecoregion size varies by ecoregion,
and factors show additional variation within ecoregion. There is only limited information
about the extent to which Level III Ecoregions accurately reflect genetic variation in most
121
Gibson
PhD Dissertation: Chapter 4
2015
understory plants, although there is support for their use from a limited number of forb
(Miller et al. 2011) and grass species (Wilson et al. 2001, Erickson et al. 2004, Parsons et
al. 2011). Using Level III Ecoregions as provisional transfer zones could promote genetic
diversity by mixing multiple, differently adapted populations, and thereby improve
restored population function (Kettenring et al. 2014), fitness (Williams 2001), and
response to changing conditions (Lesica and Allendorf 1999, Havens et al. 2015). Given
the size of Level III Ecoregions, however, environmental and climatic conditions are
likely to vary significantly within their boundaries, and conditions on one edge may be
more similar to those in an adjacent region than in the ecoregion center or opposite edge
(Johnson et al. 2004). This suggests that Level III Ecoregions may be too broad as
transfer zones for many species and lead to planting maladapted genotypes.
An alternate approach to using ecoregions for provisional transfer zones is the use of
climate envelopes and associate geophysiological attributes, like elevation and latitude.
Climate has been found to drive local adaptation in a variety of species (McKay et al.
2001, Etterson 2004, Becker et al. 2006) and has been used for decades to define speciesspecific transfer zones in forestry tree species (Randall and Berrang 2002, Ying and
Yanchuk 2006). Climate-only provisional transfer zones are developed based on climatic
factors that are believed to be important for adaptation, like winter minimum
temperatures and aridity. Like Ecoregion-only seed transfer guidelines though, using
climate alone may miss other significant ecological drivers of population differentiation
that occur at the regional scale, such as common habitat and regional soil type, or land
use. Combining climate isoclines with ecoregion boundaries could assist in separating
areas that are similar climatically but dissimilar for other key ecological drivers, and the
122
Gibson
PhD Dissertation: Chapter 4
2015
combination of climatic factors and Level III Ecoregions is becoming an accepted
addition to ecoregions for provisional seed transfer guidelines (Vogel et al. 2005, Bower
et al. 2014).
Restoration and revegetation often requires moving seeds over large distances and
combining multiple seed sources. Not only does the use of a variety of seed sources
increase available seeds (Millar et al. 2008), but it also increases genetic diversity and
adaptive potential in restored populations (Rice and Emery 2003). In the absence of
species-specific data, managers use provisional seed transfer zones to balance the desire
to use locally adapted seed sources with the need to use seeds from multiple sources. To
date only two studies have compared whether climate or climate within ecoregion better
explains morphological variation among populations (Bower et al. 2014, Kramer et al.
2015). Data from multiple areas and species will ultimately determine how generally any
one approach can be applied (Bower et al. 2014), and multiple strategies for provisional
transfer zones should be compared to see which environmental or geographic factors are
most important for a species (Figure 1). Toward that end, I established a common garden
in Missoula, MT to assess which options for provisional seed transfer zones best predict
trait differences and performance of a commonly seeded grass (Pseudoroegneria
spicata). I ask (1) is population or ecoregion the better predictor of traits? and (2) which
approach to provisional seed transfer zones (Population-only, Ecoregion-only, Climateonly, Ecoregion + climate) best explains trait variation in P. spicata?
METHODS AND MATERIALS
Study Species and Seed Collection
123
Gibson
PhD Dissertation: Chapter 4
2015
Pseudoroegneria spicata (Poaceae) is a cool-season perennial bunchgrass native
to the Intermountain West and Great Plains. The species is ideal for us in revegetation
projects since it can grow in a variety of soil conditions, as well as on south-facing slopes
and other dry sites. Plants are outcrossing and wind-pollinated (Ogle et al. 2010), and the
species shows high natural variation in growth form and awn length (Passey and Hugie
1963). Most populations of P. spicata are diploid, but tetraploid populations occur in the
northern part of the species’ range (Carlson 1986, Larson et al. 2004), and several
populations included in this study contain tetraploid individuals (Gibson 2015 - chapter
3).
Seeds were bulk collected by the Forest Service Region 1 Native Plant Materials
Program in 2009 from populations in EPA Level III Ecoregion 15 (Northern Rockies), 16
(Idaho Batholith), and 17 (Middle Rockies; Figure 2). After collection, seeds were
cleaned and stored at the Forest Service Coeur d’Alene, ID Nursery. I acquired seeds
from the Forest Service in 2010.
Common Garden Site
On March 11, 2011, I germinated seeds in pots filled with Eko Compost potting
mix (Eko compost, sphagnum peat moss, and perlite); four seeds were sown per
container, with 100 containers for each of the 14 populations. Pots were randomly placed
in an unlit, heated greenhouse on the University of Montana campus, and were
reorganized monthly. Pots were misted to maintain moisture until seeds germinated, and
then watered four times until they were transplanted. After four weeks, I thinned
germinates to one per pot. After nine weeks (May 17, 2011), I transplanted plants into 30
124
Gibson
PhD Dissertation: Chapter 4
2015
0.5-m2 plots in the University’s common garden (Missoula, MT; elevation = 975 m).
Although the original designed called for planting one plant from each of the 14
populations in each plot, because of poor germination in some populations plots
contained between seven and 14 plants placed in a 3 x 5 grid (N = 331). I watered plants
were watered daily for two weeks to minimize mortality and removed weeds throughout
the project.
Data Collection
During the 2011 and 2012 growing seasons, I recorded data on growth (plant
height and plant basal circumference) in mid-June, and censused plants weekly for
mortality from May through September. Plant height was determined by measuring the
length of the longest leaf from the base of the plant; basal circumference was the
circumference of the plant at the ground. During the 2012 growing season, I determined
individual seed set by collecting and counting all seeds from individual plants. At the end
of the experiment in fall 2012, plants were harvested at ground level, dried, and weighed
to determine above-ground biomass for each individual.
Climate and Geographic Data
For each of the 14 populations, I obtained elevation and latitude of the collection
site from Forest Service seed-collection records. Climate data was extracted from the
Parameter Elevation Regression on Independent Slopes Model (PRISM Climate Group
2004), a statistical geographic model that estimates monthly and annual data for 800 x
800 m grid cells. I extracted maximum and mean temperature in the warmest month,
minimum and mean temperature in the coldest month, and annual precipitation over the
125
Gibson
PhD Dissertation: Chapter 4
2015
last thirty years (1981 – 2010). I also used PRISM data to calculate average precipitation
during the growing season (precipitation between April – September; see Table 1 for all
population environmental factors).
Statistical Methods
I tested for among ecoregion and among population variation in all 2012 traits
(height, basal circumference, final biomass, and seed set) using permutational
multivariate analysis of variance (PERMANOVA); PERMANOVA was performed using
the R vegan package (Oksanen et al. 2014) with Bray-Curtis distance similarity matrices
to test the effect of population nested in ecoregion (999 permutations). Final biomass and
seed set were log transformed for normalcy.
To determine whether ecoregion or population was a stronger predictor of traits, I
used two different mixed models. For phenotypic variables (2011 and 2012 height, basal
circumference), final biomass, and seed set, I used linear mixed-effect models (Pinheiro
and Bates 2000). For mortality, I used a generalized linear mixed model with a binomial
link function. Biomass and seed set were log transformed for normalcy. Ecoregion was
included in the models as a fixed effect, and population as a random effect. Differences
among ecoregions were tested using post-hoc Tukey tests with significance at the p =
0.05 level. All analyses were done using R (R Development Core Team 2012) with
combined packages lme4 (Bates et al. 2014), MuMIn (Barton 2013) and multcomp
(Hothorn et al. 2008). The proportion of variance attributable to ecoregion and population
was estimated by comparing the marginal and conditional R2 of the models (Nakagawa
and Schielzeth 2013); marginal R2 (R2m) is the variance explained by model fixed effects
126
Gibson
PhD Dissertation: Chapter 4
2015
(ecoregion), and the difference between the conditional R2 (model variance, R2c) and
R2m is the variation explained by random effects (population).
To compare the relative support for potential approaches to provisional seed
transfer zones, I developed five a priori candidate models with combinations of
ecoregion, population, and climate that I hypothesized could affect traits. The models
were:
1. Population-only: Y ~ (1|Pop)
2. Ecoregion-only: Y ~ Ecoregion
3. Climate-only: Y ~ Climate
4. Ecoregion and climate: Y ~ Ecoregion + climate
5. Null model: Y ~ 1
To reduce the number of climatic variables included in the models, I performed a
principle component analysis (PCA) of population climatic and geographic variables
(Table 1) using R. Elevation and latitude were included since they are related to climate.
The first two principle components (PC1 and 2) explained 81% of the variation. The
scores from PC1 and PC2 were then used as a composite measure of environment in
model testing. Pearson product-moment correlation coefficients were calculated to assess
the relationship between PC1 and 2 scores and environmental variables using Bonferroni
corrections for multiple comparisons in the R psych package (Revelle 2015).
I used several approaches to test the five provisional seed transfer zone models. For
growth traits and seed set, I used a linear mixed effect model to test the Population-only
strategy, with population included as a random effect, and linear regression to test the
Ecoregion-only, Climate-only, and Null models. Models for mortality were constructed
127
Gibson
PhD Dissertation: Chapter 4
2015
using either generalized linear mixed models (Population-only) or generalized linear
models with a binomial family (Ecoregion-only, Climate-only, Ecoregion + climate,
Null). For each response variable (traits, seed set and mortality), I compared the five
models using Akaike Information Criterion (AIC). To determine the best model for each
response variable, I considered (1) Akaike weights (wi), which range from 1 to 0 and
represent the “probability” that the candidate model is the best given the set of models,
and (2) the difference (=ΔAIC) between the lowest AIC model and a specific model;
models with Δi ≤ 2 have strong support, although models with Δi ≤ 7 should also be
considered (Burnham et al. 2011). Model comparison was performed in R with the
package bbmle (Bolker and R Core Development Team 2014).
If a model containing climate (either Climate-only or Ecoregion+climate) was
ranked highest, I conducted multiple linear regression to determine if the PCs
significantly predicted traits. McFadden’s pseudo r-squared was calculated for the
mortality regression model in the R pscl pacakage (Long 1997).
RESULTS
Ecoregion and Population
For all traits combined, multivariate PERMANOVA analyses showed significant
variation among ecoregion (F(2, 274) = 9.98, p < 0.001) and population (F(13,274) =
3.98, p = 0.001), in other words P. spicata from different populations and ecoregions
were significantly different.
Except for mortality, models containing ecoregion and population explained
around 20% of variation in phenotypic traits and seed set (Table 2); for mortality, the
model explained only 1% of observed variation. For six out of eight traits, ecoregion
128
Gibson
PhD Dissertation: Chapter 4
2015
explained more model variation than population; for the other two (2012 height and seed
set) population explained more variance (Table 2).
There were significant differences among ecoregions in phenotypic traits and seed
set. Plants from Ecoregion 17 were shorter (Figure 3A), had smaller basal circumferences
(Figures 3B, D) and lower final biomass (Figure 4A) than did those from the other two
ecoregions. Ecoregion 15 and 16 had similar basal circumferences (Figures 3B, D) and
final biomass (Figure 4A). Plants from the three ecoregions did not differ significantly in
seed set or mortality (Figures 4C, D).
Model Selection
Principle component 1 (PC1) was significantly correlated with source population
annual precipitation and temperatures in the warmest month (Table 3). Principle
component 2 (PC2) was significantly correlated with source population elevation and
temperatures in the coldest month (Table 3). Ecoregion 15 divided from 16 and 17 along
PC1 (Figure 5), with cooler summer and higher precipitation home sites in Ecoregion 15.
Populations from Ecoregion 17 had higher PC2 scores than the populations from other
ecoregions (Figure 5) and came from higher elevation sites with colder temperatures.
There was not strong support for either the Population-only or Ecoregion-only
model (Table 4). Each had the highest model weight for only one trait: the Populationonly model ranked highest for 2012 height, and the Ecoregion-only model ranked highest
2011 basal circumference (Table 4). Both had Δi > 2 for all other traits.
The Climate-only model had the highest weight for seed set and mortality (Table
4), but otherwise had Δi > 2 for the remaining traits.
129
Gibson
PhD Dissertation: Chapter 4
2015
The Ecoregion + climate model had the highest weight for three of the seven traits
(2011 height, 2012 basal circumference, and final biomass; Table 4).
Environment
Principle component 1 (correlated with annual precipitation and warm
temperatures) significantly predicted only one of the six traits (Table 5). There was a
positive relationship between 2011 height and PC1 scores (Table 5). Principle component
2 (correlated with elevation and cold temperatures) significantly predicted 2011 height,
seed set, and mortality (Table 5). There were positive relationships between PC2 and
plant height and mortality (mortality rates increased with PC2 scores), and a negative
relationship between PC2 and seed set (Table 5). The relationship between seed set and
predictor variables and between mortality and predictor variables were weak (Table 5).
DISCUSSION
Selecting the right spatial and environmental scales for seed transfer is important
for maintaining locally adapted genotypes and preventing outbreeding depression
(Cremieux et al. 2010, Vander Mijnsbrugge et al. 2010). Provisional seed transfer zones
are created based on assumptions about the environmental factors that drive adaptive
variation in multiple species, and provisional seed transfer zone policies can impact both
the feasibility of using locally collected seeds and the genetic composition and adaptive
potential of the revegetated population (Kaye 2001, Breed et al. 2013, Jones 2013).
Locally collected seeds are presumed to increase the rates of population growth and
persistence (Montalvo et al. 1997), but identifying which factors determine local is
important in choosing adapted and ecologically appropriate seed sources. Although the
goal of generalized provisional seed transfer zones is to guide seed transfer in species
130
Gibson
PhD Dissertation: Chapter 4
2015
without species-specific information, provisional guidelines have tradeoffs and their
applicability will likely need some degree of testing in multiple species.
My results indicate that that “local” for P. spicata in these traits lies somewhere
between very local and relatively broad. The use of highly local populations of P. spicata
in revegetation is likely unwarranted, given that population explained little variation in
growth traits (height, circumference, or final biomass). Pseudoroegneria spicata is
outcrossing and wind-pollinated, which could explain the lack of strong population-level
differences in traits (Loveless and Hamrick 1984, St Clair et al. 2013). The current focus
of native plant materials research is primarily to determine which populations are
genetically similar enough to be combined in revegetation; for restoration success,
however, population-specific traits, like higher seed set, could affect the fitness of
restored populations in significant ways (Broadhurst et al. 2008).
While P. spicata population explained relatively little phenotypic trait variation
compared to ecoregion, population was more important in explaining variation in seed
set. High among-population variation in seed set could have impacts on native plant
materials selection: when one or a few populations within a species contain the most fit
individuals for any environment, these are clearly the populations that will perform best
within restorations across the whole of the species range (Jones 2003, Carter and Blair
2013). In addition, polyploidy adds a layer of difficulty in defining appropriate scales of
genotype movement for this species. Polyploid cytotypes of P. spicata are common in
Montana and Idaho, and several of these populations have tetraploids present (Gibson
2015 – chapter 3). Negative reproductive consequences from combining diploids and
tetraploids (for example, reduced seed viability and fitness in triploids; Petit et al. 1999,
131
Gibson
PhD Dissertation: Chapter 4
2015
Burton and Husband 2000, Otto 2007) means population ploidy should also be
considered when selecting a seed source for restoration.
Ecoregion explained more variation in growth traits than population, but this
explanatory power is due to differences between Ecoregion 17 and the other two regions.
The observed lack of differences between Ecoregions 15 and 16 support previous
findings of lack of among-ecoregion differences in P. spicata (St Clair et al. 2013) and
other species (Horning et al 2008), suggesting that for some species, seeds could be
combined across ecoregion boundaries. One limitation of this experiment was that
populations from Ecoregion 16 and 17 were not well dispersed within their respective
ecoregions. If within-ecoregion variation among populations is related to geographic
distance or distance from the ecoregion center, then the trends I observed may have been
different if populations were more broadly dispersed.
The best provisional seed transfer model based on plant growth (height,
circumference, and final biomass) was Climate + ecoregion. Although Level III
Ecoregions have been recommended for provisional seed transfer zones (Jones 2005), my
findings suggest that for P. spicata they may be too broad to accurately predict
population differentiation when there are significant climatic or topographic differences
within their boundaries (Miller 2010). Climate and elevation are important drivers of
variation (St Clair et al 2005, Horning et al. 2008, Wilson et al 2001), and ecoregions
may need to be further subdivided by climate to best track population differences (Wilson
et al. 2008, Bower 2014, Kramer et al. 2015). Trait variation in common plant species has
been associated with within-ecoregion variation in winter temperatures (Price et al. 2012,
Bower et al. 2014), and the high performance of the Ecoregion + climate model in this
132
Gibson
PhD Dissertation: Chapter 4
2015
study indicates that provisional seed transfer zones for P. spicata in this region should
include both factors.
Managers may want to consider moving from the use of ecoregion-based zones to
using a climate model. The Climate-only model (which included elevation) best predicted
seed set and mortality, and both traits have important implications for revegetation
success and viability; it is worth noting, however, that even though the Climate-only
model ranked highest out of the five models for mortality it explained very little of the
variation, indicating that there are other significant factors driving mortality. Temperature
has been observed to be an important climatic driver of morphological variation in P.
spicata in other areas (St Clair et al 2013). My results support including winter minimum
temperatures in provisional seed zones for P. spicata (Vogel et al 2005, Bower et al.
2014). Summer temperatures and precipitation, both of which have been included in other
provisional transfer zones (Bower et al 2014) were not significant drivers of variation in
my study. This is likely because summer temperatures and precipitation primarily
differed between Ecoregions 15 and 16, and these two regions did not show differences
in traits. Finally, although previous genetic analysis of P. spicata in British Columbia did
not find a relationship between elevation and genetic differentiation (Fu and Thompson
2006), elevation was a significant factor in my study; the more extreme elevation
gradients in the Intermountain west are one possible explanation for differences in results
between these studies. Elevation can be simple to include in seed transfer zone
development, but since elevation is correlated with a variety of other climatic and
environmental factors, it is likely that factors beyond temperature and precipitation are
important for P. spicata population differentiation in this region.
133
Gibson
PhD Dissertation: Chapter 4
2015
The further subdivision of ecoregions by climate for seed transfer is supported by
the results from the common garden. Plants from Ecoregion 17 (the home ecoregion)
performed the worst in the common garden, likely due to ecological differences between
home site and the location of the common garden. The home sites for the three Ecoregion
17 populations were at higher elevations than Missoula and experience colder winters,
and both of these factors were significant in predicting mortality and seed set at the study
site. The goal of provisional seed transfer zones is to identify areas where seeds can be
moved and be the least likely to show maladaptation. Survival and seed set are both
strong indicators of “restoration success” – that is, which populations do the best at a site
and will create a lasting population. The small size of plants from Ecoregion 17 could
confer an adaptive advantage at higher elevation sites (e.g., Gonzalo-Turpin and Hazard
2009) and may not be related home-site fitness (Rowe and Leger 2012). Low seed set and
reduced survival, however, indicate reduced fitness in the Ecoregion 17 plants when they
are moved to contrasting habitats within the same ecoregion It appears that Level III
Ecoregions may be too broad for P. spicata seed transfer when there are significant
gradients in elevation and differences in winter temperatures. The populations included
from Ecoregions 15 and 16 were closer to the common garden site in Missoula than those
from Ecoregion 17, making it likely that they were adapted to local conditions.
Conclusion: While provisional transfer zones come with their own assumptions
about which factors drive genetic differentiation, my results suggest that differences in
revegetation outcomes could depend on the strategy used for selecting native plant
materials. Using Climate-only transfer zones would result in increased seed set and
decreased mortality, while the Ecoregion + climate model would result in a better match
134
Gibson
PhD Dissertation: Chapter 4
2015
in phenotypic traits in native plant materials. There is currently a focus in research on
developing and testing the applicability of provisional transfer zones (e.g., Miller 2010,
Bower 2014, Kramer et al. 2015); the use of the Ecoregion + climate transfer zones is
likely more conservative for P. spicata in this region, but future research on the
applicability of provisional seed transfer zones may need to include a greater focus on
whether transfer zone boundaries are more accurately capturing variation in a suite of
phenotypic traits, or seed set and mortality.
ACKNOWLEDGEMENTS
This research was supported by agreement #09-CS-11015600-031 from the USDA Forest
Service NFN program and the National Science Foundation EPSCoR program (# EPS1101342) at the University of Montana.
LITERATURE CITED
Barton, K. 2013. MuMIn: multi-model inference. R package version 1.9. 5.
Bates, D., M. Maechler, B. Bolker, and S. Walker. 2014. lme4: linear mixed-effects
models using Eigen and S4. R package version 1.1-7.
Becker, U., G. Colling, P. Dostal, A. Jakobsson, and D. Matthies. 2006. Local adaptation
in the monocarpic perennial Carlina vulgaris at different spatial scales across
Europe. Oecologia 150:506-518.
Bischoff, A., T. Steinger, and H. Muller-Scharer. 2010. The importance of plant
provenance and genotypic diversity of seed material used for ecological
restoration. Restoration Ecology 18:338-348.
135
Gibson
PhD Dissertation: Chapter 4
2015
Borders, B. D., B. L. Cypher, N. P. Ritter, and P. A. Kelly. 2011. The challenge of
locating seed sources for restoration in the San Joaquin Valley, California.
Natural Areas Journal 31:190-199.
Bower, A. D., B. St Clair, and V. Erickson. 2014. Generalized provisional seed zones for
native plants. Ecological Applications 24:913-919.
Breed, M. F., M. G. Stead, K. M. Ottewell, M. G. Gardner, and A. J. Lowe. 2013. Which
provenance and where? Seed sourcing strategies for revegetation in a changing
environment. Conservation Genetics 14:1-10.
Broadhurst, L. M., A. Lowe, D. J. Coates, S. A. Cunningham, M. McDonald, P. A. Vesk,
and C. Yates. 2008. Seed supply for broadscale restoration: maximizing
evolutionary potential. Evolutionary Applications 1:587-597.
Broadhurst, L. M., A. G. Young, P. H. Thrall, and B. G. Murray. 2006. Sourcing seed for
Acacia acinacea, a key revegetation species in south eastern Australia.
Conservation Genetics 7:49-63.
Burnham, K. P., D. R. Anderson, and K. P. Huyvaert. 2011. AIC model selection and
multimodel inference in behavioral ecology: some background, observations, and
comparisons. Behavioral Ecology and Sociobiology 65:23-35.
Burton, T. L., and B. C. Husband. 2000. Fitness differences among diploids, tetraploids,
and their triploid progeny in Chamerion angustifolium: mechanisms of inviability
and implications for polyploid evolution. Evolution 54:1182-1191.
Carlson, J. 1986. A study of morphological variation within Pseudoroegneria spicata
(Pursh) A. Love (Poaceae: Triticeae). Oregon State University.
136
Gibson
PhD Dissertation: Chapter 4
2015
Carter, D. L., and J. M. Blair. 2013. Seed source has variable effects on species,
communities, and ecosystem properties in grassland restorations. Ecosphere 4:16.
Cremieux, L., A. Bischoff, H. Muller-Scharer, and T. Steinger. 2010. Gene flow from
foreign provenances into local plant populations: fitness consequences and
implication for biodiversity restoration. American Journal of Botany 97:94-100.
Ellstrand, N. C., and D. R. Elam. 1993. Population genetic consequences of small
population-size - implications for plant conservation. Annual Review of Ecology
and Systematics 24:217-242.
Erickson, V. J., N. L. Mandel, and F. C. Sorensen. 2004. Landscape patterns of
phenotypic variation and population structuring in a selfing grass, Elymus glaucus
(blue wildrye). Canadian Journal of Botany-Revue Canadienne De Botanique
82:1776-1789.
Etterson, J. R. 2004. Evolutionary potential of Chamaecrista fasciculata in relation to
climate change. 1. Clinal patterns of selection along an environmental gradient in
the Great Plains. Evolution 58:1446-1458.
Fu, Y. B., and D. Thompson. 2006. Genetic diversity of bluebunch wheatgrass
(Pseudoroegneria spicata) in the Thompson River valley of British Columbia.
Canadian Journal of Botany-Revue Canadienne De Botanique 84:1122-1128.
Godefroid, S., C. Piazza, G. Rossi, S. Buord, A. D. Stevens, R. Aguraiuja, C. Cowell, C.
W. Weekley, G. Vogg, J. M. Iriondo, I. Johnson, B. Dixon, D. Gordon, S.
Magnanon, B. Valentin, K. Bjureke, R. Koopman, M. Vicens, M. Virevaire, and
T. Vanderborght. 2011. How successful are plant species reintroductions?
Biological Conservation 144:672-682.
137
Gibson
PhD Dissertation: Chapter 4
2015
Gonzalo-Turpin, H., and L. Hazard. 2009. Local adaptation occurs along altitudinal
gradient despite the existence of gene flow in the alpine plant species Festuca
eskia. Journal of Ecology 97:742-751.
Havens, K., P. V. Vitt, S. Still, A. T. Kramer, J. B. Fant, and K. Schatz. 2015. Seed
sourcing for restoration in an era of climate change. Natural Areas Journal
35:122-133.
Horning, M. E., T. R. McGovern, D. C. Darris, N. L. Mandel, and R. Johnson. 2008.
Genecological of Holodiscus discolor (Rosaceae) in the Pacific Northwest,
U.S.A. Restoration Ecology 20:1–9.
Hothorn, T., F. Bretz, and P. Westfall. 2008. Simultaneous inference in general
parametric models. Biometrical Journal 50:346--363.
Hufford, K. M., and S. J. Mazer. 2003. Plant ecotypes: genetic differentiation in the age
of ecological restoration. Trends in Ecology & Evolution 18:147-155.
Johnson, G., F. Sorensen, J. St Clair, and R. Cronn. 2004. Pacific Northwest forest tree
seed zones: a template for native plants? Native Plants 5:131 - 140.
Johnson, R., L. Stritch, P. Olwell, S. Lambert, M. E. Horning, and R. Cronn. 2010a. What
are the best seed sources for ecosystem restoration on BLM and USFS lands?
Native Plants 11:117-131.
Johnson, R. C., V. J. Erickson, N. L. Mandel, J. B. St Clair, and K. W. Vance-Borland.
2010b. Mapping genetic variation and seed zones for Bromus carinatus in the
Blue Mountains of eastern Oregon, USA. Botany-Botanique 88:725-736.
Johnson, R. C., B. C. Hellier, and K. W. Vance-Borland. 2013. Genecology and seed
zones for tapertip onion in the US Great Basin. Botany-Botanique 91:686-694.
138
Gibson
PhD Dissertation: Chapter 4
2015
Jones, T., and S. Larson. 2005. Status and use of important native grasses adapted to
sagebrush communities. USDA Forest Service Proceedings RMRS-P-38., Fort
Collins, CO.
Jones, T. A. 2003. The restoration gene pool concept: beyond the native versus nonnative debate. Restoration Ecology 11:281-290.
Jones, T. A. 2013. When local isn't best. Evolutionary Applications 6:1109-1118.
Joshi, J., B. Schmid, M. C. Caldeira, P. G. Dimitrakopoulos, J. Good, R. Harris, A.
Hector, K. Huss-Danell, A. Jumpponen, A. Minns, C. P. H. Mulder, J. S. Pereira,
A. Prinz, M. Scherer-Lorenzen, A. S. D. Siamantziouras, A. C. Terry, A. Y.
Troumbis, and J. H. Lawton. 2001. Local adaptation enhances performance of
common plant species. Ecology Letters 4:536-544.
Kaye, T. 2001. Common ground and controversy in native plant restoration: the SOMS
debate, source distance, plant selections, and a restoration-oriented definition of
native. Pages 5 - 12 in R. Rose and D. Haase, editors. Nursery Technology
Cooperative and Western Forestry and Conservation Association, Corvallis, OR.
Keller, L. F., and D. M. Waller. 2002. Inbreeding effects in wild populations. Trends in
Ecology & Evolution 17:230-241.
Kettenring, K. M., K. L. Mercer, C. R. Adams, and J. Hines. 2014. Application of genetic
diversity- ecosystem function research to ecological restoration. Journal of
Applied Ecology 51:339-348.
Knapp, E., and K. Rice. 1994. Starting from seed: genetic issues in using native grasses
for restoration. Restoration and Management Notes 12:40 - 45.
139
Gibson
PhD Dissertation: Chapter 4
2015
Kramer, A., D. Larkin, and J. Fant. 2015. Assessing potential seed transfer zones for five
forb species from the Great Basin Floristic Region, USA. Natural Areas Journal
35:174-188.
Lankau, R., P. S. Jorgensen, D. J. Harris, and A. Sih. 2011. Incorporating evolutionary
principles into environmental management and policy. Evolutionary Applications
4:315-325.
Larson, S. R., T. A. Jones, and K. B. Jensen. 2004. Population structure in
Pseudoroegneria spicata (Poaceae: Triticeae) modeled by Bayesian clustering of
AFLP genotypes. American Journal of Botany 91:1789-1801.
Lesica, P., and F. W. Allendorf. 1999. Ecological genetics and the restoration of plant
communities: mix or match? Restoration Ecology 7:42-50.
Loveless, M. D., and J. L. Hamrick. 1984. Ecological determinants of genetic structure in
plant populations. Annual Review of Ecology and Systematics 15:65-95.
Macel, M., C. S. Lawson, S. R. Mortimer, M. Smilauerova, A. Bischoff, L. Cremieux, J.
Dolezal, A. R. Edwards, V. Lanta, T. M. Bezemer, W. H. van der Putten, J. M.
Igual, C. Rodriguez-Barrueco, H. Muller-Scharer, and T. Steinger. 2007. Climate
vs. soil factors in local adaptation of two common plant species. Ecology 88:424433.
McKay, J. K., J. G. Bishop, J. Z. Lin, J. H. Richards, A. Sala, and T. Mitchell-Olds. 2001.
Local adaptation across a climatic gradient despite small effective population size
in the rare sapphire rockcress. Proceedings of the Royal Society B-Biological
Sciences 268:1715-1721.
140
Gibson
PhD Dissertation: Chapter 4
2015
McKay, J. K., C. E. Christian, S. Harrison, and K. J. Rice. 2005. "How local is local?" A review of practical and conceptual issues in the genetics of restoration.
Restoration Ecology 13:432-440.
Merritt, D. J., and K. W. Dixon. 2011. Restoration seed banks - a matter of scale. Science
332:424-425.
Millar, M.A., M. Byrne, and D.J. Coates. 2008. Seed collection for revegetation:
guidelines for Western Australia flora. Journal of the Royal Society of Western
Australia 91: 293-299.
Miller, S. A., A. Bartow, M. Gisler, K. Ward, A. S. Young, and T. N. Kaye. 2011. Can an
ecoregion serve as a seed transfer zone? Evidence from a common garden study
with five native species. Restoration Ecology 19:268-276.
Montalvo, A. M., S. L. Williams, K. J. Rice, S. L. Buchmann, C. Cory, S. N. Handel, G.
P. Nabhan, R. Primack, and R. H. Robichaux. 1997. Restoration biology: a
population biology perspective. Restoration Ecology 5:277-290.
Nakagawa, S., and H. Schielzeth. 2013. A general and simple method for obtaining R2
from generalized linear mixed-effects models. Methods in Ecology and Evolution
4:133-142.
Ogle, D., L. St John, and T. Jones. 2010. Plant guide for bluebunch wheatgrass
(Pseudoroegneria spicata). USDA-Natural Resource Conservation Service. Idaho
and Washington Plant Materials Program, http://www.plantmaterials.nrcs.usda.gov/pubs/idpmspg9765.pdf.
141
Gibson
PhD Dissertation: Chapter 4
2015
Oksanen, J., F. Blanchet, R. Kindt, P. Legendre, P. Minchin, R. O'Hara, G. Simpson, P.
Solymos, H. Stevens, and H. Wagner. 2014. vegan: community ecology package.
R package version 2.2-0.
Omernik, J. M. 1987. Ecoregions of the conterminous United States. Annals of the
Association of American Geographers 77:118-125.
Otto, S. P. 2007. The evolutionary consequences of polyploidy. Cell 131:452-462.
Parsons, M. C., T. A. Jones, S. R. Larson, I. W. Mott, and T. A. Monaco. 2011. Ecotypic
variation in Elymus elymoides subsp brevifolius in the Northern Intermountain
West. Rangeland Ecology & Management 64:649-658.
Passey, H. B., and V. K. Hugie. 1963. Variation in bluebunch wheatgrass in relation to
environment and geographic location. Ecology 44:158-&.
Petit, C., F. Bretagnolle, and F. Felber. 1999. Evolutionary consequences of diploidpolyploid hybrid zones in wild species. Trends in Ecology & Evolution 14:306311.
Pinheiro, J. C., and D. Bates. 2000. Mixed-Effects Models in S and S-PLUS. Springer.
Price, D. L., P. R. Salon, and M. D. Casler. 2012. Big bluestem gene pools in the Central
and Northeastern United States. Crop Science 52:189-200.
PRISM Climate Group, Oregon State University. 2004.
R Development Core Team. 2012. R: A language and environment for statistical
computing. R Foundation for Statistical Computing, Vienna, Austria.
Raabova, J., Z. Munzbergova, and M. Fischer. 2007. Ecological rather than geographic or
genetic distance affects local adaptation of the rare perennial herb, Aster amellus.
Biological Conservation 139:348-357.
142
Gibson
PhD Dissertation: Chapter 4
2015
Randall, W., and P. Berrang. 2002. Washington tree seed transfer zones. Washington
State Department of Natural Resources (DNR).
Revelle, W. 2015. psych: procedures for personality and psychological research,
Northwestern University, Evanston, Illinois, USA. Version = 1.5.1.
Rice, K.J., and N.C. Emery. 2003. Managing microevolution: restoration in the face of
global climate change. Frontiers in Ecology and the Environment 1:469-478.
Saari, C., and W. Glisson. 2012. Survey of Chicago region restoration seed source
policies. Ecological Restoration 30:162–165.
Shaw, N., S. Lambert, A. DeBolt, and M. Pellant. 2005. Increasing native forb seed
supplies for the Great Basin. USDA Forest Service Proceedings RMRS-P-35.
2005 in R. K. Dumroese, L. E. Riley, and T. D. Landis, editors. National
proceedings: Forest and Conservation Nursery Associations—2004; 2004 July
12–15; Charleston, NC; and 2004 July 26–29; Medford, OR. Proc. RMRS-P-35.
U.S. Department of Agriculture, Forest Service, Rocky Mountain Research
Station, Fort Collins, CO.
St Clair, J. B., F. F. Kilkenny, R. C. Johnson, N. L. Shaw, and G. Weaver. 2013. Genetic
variation in adaptive traits and seed transfer zones for Pseudoroegneria spicata
(bluebunch wheatgrass) in the northwestern United States. Evolutionary
Applications 6:933-948.
Vander Mijnsbrugge, K., A. Bischoff, and B. Smith. 2010. A question of origin: where
and how to collect seed for ecological restoration. Basic and Applied Ecology
11:300-311.
143
Gibson
PhD Dissertation: Chapter 4
2015
Vogel, K. P., M. R. Schmer, and R. B. Mitchell. 2005. Plant adaptation regions:
ecological and climatic classification of plant materials. Rangeland Ecology &
Management 58:315-319.
Weisshuhn, K., D. Prati, M. Fischer, and H. Auge. 2012. Regional adaptation improves
the performance of grassland plant communities. Basic and Applied Ecology
13:551-559.
Williams, S. L. 2001. Reduced genetic diversity in eelgrass transplantations affects both
population growth and individual fitness. Ecological Applications 11:1472-1488.
Wilson, B., D. Darris, R. Fiegener, R. Johnson, M. Horning, and K. Kuykendall. 2008.
Seed transfer zones for a native grass: Festuca roemeri Native Plants Journal
9:287 - 302.
Wilson, B. K., J. Kitzmiller, W. Rolle, and V. D. Hipkins. 2001. Isozyme variation and
its environmental correlates in Elymus glaucus from the California Floristic
Province. Canadian Journal of Botany-Revue Canadienne De Botanique 79:139153.
Ying, C. C., and A. D. Yanchuk. 2006. The development of British Columbia's tree seed
transfer guidelines: Purpose, concept, methodology, and implementation. Forest
Ecology and Management 227:1-13.
144
Gibson
PhD Dissertation: Chapter 4
2015
Table 1: EPA Level III Ecoregion, elevation, geographic location, and climate for 14 populations of P. spicata collected in Montana
and Idaho. Growing season precipitation was the total precipitation from April to September; annual precipitation includes snow and
rainfall. Cold temperatures are based on the coldest month for each site, and warm temperatures on the warmest month for each site.
Ecoregion
15
15
15
15
15
15
16
16
16
16
16
17
17
17
Elevation
Population
(m)
15-1
15-2
15-3
15-4
15-5
15-6
16-1
16-2
16-3
16-4
16-5
17-1
17-2
17-3
1006
1067
1097
1417
792
916
1478
1524
1341
1463
1634
1694
1854
1684
Lat.
48.245
48.84
47.912
47.568
47.322
46.986
45.865
45.69
45.547
45.619
45.547
45.216
45.007
45.235
Growing
season precip.
(cm)
Annual
precip.
(cm)
30.2
33.8
37.4
31.7
22.8
27.1
26.7
27.9
26.6
29.1
27.7
29.7
31.8
38.5
85.4
96.6
121.9
94.4
47.1
81.1
55.8
64.0
50.5
56.8
67.7
43.8
49.9
62.6
Cold temperatures
(oC)
Minimum
-8.2
-7.7
-5.9
-7.3
-6.3
-6.8
-9.6
-9.6
-6
-5.7
-6.2
-9.9
-11
-10.5
Mean
-4.5
-5.3
-3.5
-4.7
-2.8
-1.6
-5
-5.2
-2.2
-2.7
-3.1
-4
-5
-5
Warm temperatures
(oC)
Maximum
26.2
21.4
23
23.3
29.9
28.2
28.6
27.9
28.8
27
25.7
28.6
25.3
26.2
Mean
17
16.1
17.2
16.5
20.1
18
18
17.5
19.6
18.9
17.8
19.9
17.2
17.6
145
Gibson
PhD Dissertation: Chapter 4
2015
Table 2: Marginal (R2m) and conditional (R2c) variation explained by model Y ~
Ecoregion + (1|Population) for phenotypic traits, seed set, and mortality in
Pseudoroegneria spicata.
Trait
2011 height
R2 m
0.21
R2 c
0.25
2012 height
0.06
0.22
2011 basal circumference
0.18
0.19
2012 basal circumference
0.21
0.23
Biomass
0.14
0.16
Seed set
0.09
0.21
Survival
0.01
0.01
146
Gibson
PhD Dissertation: Chapter 4
2015
Table 3: Pearson correlations coefficients between principle components (PC1 and PC2)
scores and elevation, geographic location and climate variables. Correlations in bold are
significant at the p = 0.05 level.
Variable
PC1
PC2
Elevation
0.216
0.873
Latitude
-0.701
-0.596
Growing season precipitation
-0.717
0.406
Annual precipitation
-0.897
-0.336
Mean warm temperature
0.891
-0.231
Mean cold temperature
0.470
-0.739
Maximum warm temperature
0.926
-0.095
Minimum cold temperature
-0.027
-0.881
147
Gibson
PhD Dissertation: Chapter 4
2015
Table 4: Model selection for phenotypic traits, seed set, and mortality in Pseudoroegneria
spicata. AIC = Aikake’s information criterion; ΔAIC = AIC of model – AIC of best
model; R2 is model adjusted R2.
AIC weight
df
AIC
ΔAIC
(wi)
R2
Y ~ Ecoregion + PC1 + PC2
6
1971.1
0
0.998
0.24
Y ~ Ecoregion
4
1984
12.8
0.002
Y ~ (1|Pop)
3
2000.6
29.5
<0.001
Y ~ PC1 + PC2
4
2016.8
45.7
<0.001
Y~1
2
2054.5
83.4
<0.001
Y ~ Ecoregion
4
2202.5
0
0.830
Y ~ Ecoregion + PC1 + PC2
6
2205.7
3.2
0.170
Y ~ (1|Pop)
3
2224.2
21.7
<0.001
Y ~ PC1 + PC2
4
2224.5
22.0
<0.001
Y~1
2
2260.4
57.0
<0.001
Y ~ (1|Pop)
3
2100.2
0
0.999
Y ~ Ecoregion
4
2114
13.8
0.001
Y ~ Ecoregion + PC1 + PC2
6
2117.2
17.0
<0.001
Y ~ PC1 + PC2
4
2120.4
20.2
<0.001
Y~1
2
2121.9
21.7
<0.001
Model
2011 Height
2011 Basal circumference
0.17
2012 Height
0.16
148
Gibson
PhD Dissertation: Chapter 4
2015
AIC weight
df
AIC
ΔAIC
(wi)
R2
Y ~ Ecoregion + PC1 + PC2
6
1942.8
0
0.955
0.15
Y ~ Ecoregion
4
1949
6.2
0.050
Y ~ PC1 + PC2
4
1954.6
11.8
0.003
Y ~ (1|Pop)
3
1960.2
17.4
<0.001
Y~1
2
1984.8
42.0
<0.001
Y ~ Ecoregion + PC1 + PC2
6
831.2
0
0.950
Y ~ Ecoregion
4
837.1
5.9
0.077
Y ~ PC1 + PC2
4
899.4
68.2
0.046
Y ~ (1|Pop)
3
919.4
88.2
<0.001
Y~1
2
939.2
108.0
<0.001
Y ~ PC1 + PC2
4
578.1
0
0.810
Y ~ Ecoregion + PC1 + PC2
6
581.0
2.9
0.190
Y ~ (1|Pop)
3
653.1
75.0
<0.001
Y ~ Ecoregion
4
667.7
88.7
<0.001
Y~1
2
919.4
341.3
<0.001
Y ~ PC1 + PC2
3
308.0
0
0.810
Y ~ Ecoregion + PC1 + PC2
1
310.8
2.9
0.190
Model
2012 Basal circumference
Final biomass
0.11
Seed set
0.10
Mortality
0.08
149
Gibson
PhD Dissertation: Chapter 4
2015
AIC weight
Model
df
AIC
ΔAIC
(wi)
Y~1
3
330.0
22.0
<0.001
Y ~ (1|Pop)
2
331.8
23.8
<0.001
Y ~ Ecoregion
6
333.1
25.1
<0.001
R2
150
Gibson
PhD Dissertation: Chapter 4
2015
Table 5: Multiple regression of principle components (PC) related to trait variation in
Pseudoroegneria spicata in Montana and Idaho. Model R2 are adjusted or McFadden’s
pseudo R-square (mortality).
PC1
PC2
Model
β
p
β
p
R2
p
2011 Height*
1.048
<0.001
1.560
< 0.001
0.240
<0.001
2011 Basal circumference*
0.189
0.630
-0.161
0.798
0.169
<0.001
2012 Basal circumference*
0.326
0.426
-1.245
0.065
0.151
<0.001
Final biomass*
0.097
0.090
-0.072
0.447
0.113
<0.001
Seed set
0.031
0.529
-0.178
<0.001
0.052
0.003
Mortality
-0.003
0.974
0.174
0.046
0.013
-
Trait
* Models contained level III ecoregions but values are not reported.
151
Gibson
PhD Dissertation: Chapter 4
2015
FIGURE CAPTIONS
Figure 1: Conceptual diagram of provisional seed transfer zone options. Closed circles
represent populations (A – D); dashed line represents a Level III Ecoregion boundary; the
shaded area is a different climate region than the unshaded area. (1) Population-only
provisional seed transfer would occur in the open circle around each population. (2)
Ecoregion-only seed transfer would allow populations A, B, and C to be mixed, but not
population D. (3) Climate-only seed transfer would allow populations B, C, and D to be
mixed, but not population A. (4) Ecoregion + climate seed transfer would allow only
populations B and C to mix.
Figure 2: Locations of the fourteen study populations of Pseudoroegneria spicata and the
common garden (Missoula, MT). Populations were located in EPA Level III Ecoregions
15 (Northern Rockies), 16 (Idaho Batholith), and 17 (Middle Rockies).
Figure 3: Differences in growth traits among Level III Ecoregions (15 [black], 16 [white],
and 17 [grey]) for 2011 (A and B) and 2012 (C and D) plant height and basal
circumference, respectively. Letters represent significant differences among groups of
bars at the p = 0.05 level. Error bars represent one standard error.
Figure 4: Differences among Level III Ecoregions (15 [black], 16 [white], and 17 [grey])
in (A) final above-ground biomass, (B) seed set, and (C) mortality. Bars with different
letters represent significant differences at the p = 0.05 level. Error bars represent one
standard error.
152
Gibson
PhD Dissertation: Chapter 4
2015
Figure 5: Principle component (PC) scores for 14 populations of P. spicata collected
from EPA Level III Ecoregions 15, 16, and 17. Principle component 1 was associated
with home site annual precipitation and mean and maximum warm temperatures; PC2
was associated with home site elevation and mean and minimum cold temperatures
153
Gibson
PhD Dissertation: Chapter 4
2015
Figure 1
154
Gibson
PhD Dissertation: Chapter 4
2015
Figure 2
155
Gibson
PhD Dissertation: Chapter 4
2015
Figure 3
a
20
Height (cm)
60
A)
b)
b
15
b
10
5
30
20
0
25
B)
a
a
20
b
15
10
5
Basal circumference (cm)
Basal circumference (cm)
25
40
10
0
30
C)
50
Height (cm)
25
20
a
D)
a
b
15
10
5
0
0
15
16
Ecoregion
17
15
16
17
Ecoregion
156
Gibson
PhD Dissertation: Chapter 4
2015
Figure 4
14
A)
a
12
a
Biomass (g)
10
8
6
b
4
2
0
Seed set (number/plant)
160
B)
140
120
100
80
60
40
20
0
0.30
C)
Mortality (%)
0.25
0.20
0.15
0.10
0.05
0.00
15
16
17
Ecoregion
157
Gibson
PhD Dissertation: Chapter 4
2015
Figure 5
4
15
16
3
17
2
PC2
1
0
-1
-2
-3
-5
-4
-3
-2
-1
0
1
2
3
PC1
158