Characterizing the Habitat Requirements of Rare and Hard-to-Establish Puget-Trough Prairie Forbs

Item

Identifier

Thesis_MES_2019_MessinaJ

Title

Characterizing the Habitat Requirements of Rare and Hard-to-Establish Puget-Trough Prairie Forbs

Date

October 2019

Creator

Messina, John

extracted text

Characterizing the Habitat Requirements of Rare and
Hard-to-Establish Puget-Trough Prairie Forbs

By
John James Messina

A Thesis
Submitted in partial fulfillment
of the requirements for the degree
Master of Environmental Studies
The Evergreen State College
September 2019

 2019 by John James Messina. All rights reserved.

This Thesis for the Master of Environmental Studies Degree
By
John James Messina

Has been approved for
The Evergreen State College
By

________________________
Thesis Reader

________________________
Date

ABSTRACT

The prairies of the Puget-Trough and Willamette Valley represent some of the most
fragmented landscapes in the Pacific Northwest, while also providing critical
habitat to a wide array of rare and threatened species. This thesis asks three basic
questions: 1.) What is the simplest and most accurate way to describe the
characteristics that differentiate microsites on the prairie landscape? 2.) Which
microsites yield the strongest germination performance of locally rare species,
Balsamorhiza deltoidea and Gaillardia aristata? 3.) How do microsite
characteristics influence germination of Balsamorhiza deltoidea and Gaillardia
aristata? Looking across three different sites, this study finds that Balsamorhiza
deltoidea and Gaillardia aristata diverge from each other in both their preferred
microsite type and how environmental parameters and biotic interactions influence
germination rates. Balsamorhiza deltoidea has a stronger preference for mounds
and highland sites, which can be characterized by lower soil bulk densities and
more vigorous plant growth. Conversely, Gaillardia aristata, has a stronger
preference for intermound and lowland sites, characterized by higher soil bulk
densities and less dense plant growth. Microsites offer land managers and
restoration ecologists a valuable scale by which restoration activities can be carried
out without specialized tools or knowledge.

i

Table of Contents
Chapter One: Literature Review ......................................................................................... 1
Restoration History and Strategy .................................................................................... 1
Invasive Species Theory and Management .................................................................... 2
Fire History and Management ........................................................................................ 6
Soil Disturbance .............................................................................................................. 8
Native Species Restoration and Impacts on Ecosystem Functioning ........................... 10
Species Rarity: Causes and Conservation ..................................................................... 15
Literature Review Works Cited ......................................................................................... 20
Chapter Two: Thesis Research .......................................................................................... 30
Introduction .................................................................................................................. 30
Methods ........................................................................................................................ 38
Site Layout and Description ...................................................................................... 38
Data Collection .......................................................................................................... 39
Statistical Methods and Rationale ............................................................................ 40
Results ........................................................................................................................... 42
Microsite Characteristics .......................................................................................... 42
Impacts on Germination ........................................................................................... 44
Discussion ..................................................................................................................... 47
Principle Conclusions ................................................................................................... 54
Thesis Works Cited ............................................................................................................ 56
Appendix 1: Figures........................................................................................................... 63
Appendix 2: Tables ............................................................................................................ 69

ii

List of Figures
Figure 1: Bulk density………..………………………………………...………...63
Figure 2: Glacial Heritage soil moisture……….……………………….…….......63
Figure 3: JBLM soil moisture……………………………………….……………64
Figure 4: Microsite aboveground plant density…………………….…………….64
Figure 5: Microsite exotic plant cover……………………………………………65
Figure 6: Microsite species richness……………………………………………...65
Figure 7: Microsite native plant cover……………………………………………66
Figure 8: Microsite forb cover……………………………………………………66
Figure 9: Microsite grass cover…………………………………………………...67
Figure 10: Gaillardia germination………………………………………..………67
Figure 11: GH Balsamroot germination……………...…………………………..68
Figure 12: JBLM Balsamroot germination…………….…..……………………..68
List of Tables
Table 1: Descriptive statistics for abiotic parameters…………...…….…………69
Table 2: Post-hoc test for differences in bulk density between JP microsites…….69
Table 3: Post-hoc test for differences in bulk density between TA15 microsites…70
Table 4: Post-hoc test for differences in moisture between microsites at GH…....70
Table 5: Descriptive statistics for biotic parameters…………………….………..71
Table 6: Glacial Heritage count model goodness of fit…………………………...72
Table 7: Balsamroot germination GLM by microsite at GH…..………………….72
Table 8: Gaillardia germination GLM by microsite at GH…………….…………72
Table 9: Johnson Prairie balsamroot germination GLM………………….………73
Table 10: Training Area 15 balsamroot germination GLM………………………73
Table 11: GH balsamroot GLM with abiotic predictors…………………………..73
Table 12: GH gaillardia GLM with abiotic predictors………………………..….74
Table 13: GH balsamroot GLM with biotic predictors………………...………...74
Table 14: GH gaillardia GLM with biotic predictors………………………….….75
iii

Acknowledgements

First and foremost, this thesis would not be possible without the guidance, hard
work, and patience of my reader, Sarah Hamman, PhD, restoration ecologist at the
Center for Natural Lands Management. Thank you, Sarah. I would also like to give
high accolades to the MES program, specifically Kevin Francis and Andrea Martin
– thank you for everything you did to enable me completing this project, I know it
wasn’t always easy. Evergreen faculty John Withey and Erin Martin were gracious
with their time and advice, thank you. Stu, Rachel, and Jean (CNLM) assisted me
with field ID and species counts. Without technicians, the grunts of the science
world, science does not happen - you guys rule. Thank you, Max, for your enduring
friendship and wicked good stats skills. My wonderful wife, Alexandria, and cats,
Mackerel and Hamo, provided me with unconditional love and support throughout
this process. Thank you to my parents, John and Lynn – you have always believed
in me and supported me; I couldn’t ask for anything else. MES alumni Sara Krock
was essential in providing me access to JBLM, without which this thesis would not
be possible. Evergreen State College has a world class science support center, I
could not have completed this thesis or learned as much as I have without the help
of Kaile Adney and Jenna Nelson. I cannot mention every person who helped me
through this process but believe that I haven’t forgotten your help, thank you.

iv

Chapter One: Literature Review
I begin this literature review with a broad survey of grassland restoration
strategies and processes. Next, I will turn to the issue of biodiversity in restoration
and what role biodiversity has in ecosystem functioning. After I establish some
specific and broad roles for increased biodiversity, I will turn to a discussion of
plant rarity and the specific benefits that may come with restoration and
conservation of rare species. This literature review serves the purpose of orienting
the reader to the field of restoration ecology and the challenges of this work in
grasslands. Ecological restoration can be tricky business; this review provides
examples why some environmental disturbances, restoration treatments, and
ecological theories do not always provide clear guidance for restoration of rare
and hard-to-establish grassland species.

Restoration History and Strategy
Native grasslands continue to decline globally due to a myriad of causes,
including cattle grazing, desertification, and intensive agriculture (Ceballos et al.
2010, Isselstein et al. 2005, Millennium Ecosystem Assessment 2005). In one
study researchers found that of 468 articles published in the journal Restoration
Ecology, 16% were concerned with the ongoing restoration of grasslands (RuizJaen and Aide 2005). The widespread loss of native grassland diversity and
ecosystem services thus requires well developed restoration strategies, informed
by research, long-term monitoring, and identification of concrete goals (Kaye
2009).
1

Since the nascent field of restoration ecology grew out of an ecological
restoration project at the University of Wisconsin, our understanding of prairie
restoration has increased an appreciable amount (arboretum.wisc.edu/aboutus/history/). Restoration ecology, like conservation biology, is very action/goaloriented: working with stakeholders is just as important as the ecological theories
that underpin management recommendations. Bridging the gap between ecology
and land management is at the core of the Society for Ecological Restoration’s
mission (http://www.ser.org/ page/MissionandVision). Academic research has
improved our ability to predict restoration outcomes (Laughlin et al. 2017), our
understanding of community assembly processes for both aboveground and
belowground communities (Kraft and Ackerly 2014, Vályi et al. 2016), and,
through these advances, greatly informed the practices of land managers.
Emphasis on exotic species removal, active seeding/installation of native species
and restoration of historic disturbance regimes are key insights following decades
of research (Rowe, 2010).

Invasive Species Theory and Management
The presence of endangered species has driven much of the science that
has informed the restoration of fragmented and degraded prairies. One of the
largest obstacles to reviving some endangered species and associated native
diversity is the constant unrelenting competition of non-natives. The former range
of the South Sound prairies is often by overrun by Scotch broom (Cytisus
scoparius), when left unmaintained. Other non-natives, such as hairy cat’s ear
(Hypochaeris radicata) or oxeye daisy (Leucanthemum vulgare) thrive even when
2

prairies are burned and actively seeded. Key to restoring native diversity and
understanding the habitat needs of hard-to-establish species is the management of
invasive species.

Typical Scotch broom-invaded prairie just a few miles from one of the study
sites.

Invasive plants typically must undergo some management plan, often
requiring intensive efforts in the initial stages (Solecki 1997). Reoccurring manual
and chemical treatment of exotic species is often required to create space for
natives to reestablish. Removal of these non-native plants is necessitated due to
insidious strategies utilized by exotic species that result in loss of native
biodiversity and associated deleterious impacts to ecosystem services (Greipsson
and DiTommaso 2006, Pimentel et al. 2000).
While the challenges non-native species present to restoration practices
may seem obvious, some restoration ecologist argue for a management approach
that focuses on establishing novel communities accompanied with a high degree a

3

functional diversity (Jackson and Hobbs 2009). Embracing a restoration paradigm
that leaves room for constructing novel ecosystems with a mix of native and
exotic species means acknowledging that humans have exerted some type of
influence on many native landscapes (Vale 2002) as well as recognizing that the
future likely brings dramatic shifts in historic climactic regimes (IPCC Climate
Change Synthesis Report, 2014). Further, some have argued that because nonnative species often represent a significant contribution to the total abundance and
richness of plant communities, they should be considered a part of the local
diversity, not something that ‘removes’ diversity (Schlaepfer 2018). A common
and abundant exotic in the South Sound prairies, English plantain (Plantago
lanceolata), is the oviposition plant of choice for Taylor’s Checkerspot butterfly
(Euphydryas editha taylori) when golden paintbrush (Castilleja levisecta) or other
host plants are not present (Kaye et al. 2011).
Further, pollinator network analysis shows that exotic species do provide
some facilitation of pollination for native species that should be taken into
consideration when invasive treatment plans are devised (Waters et al. 2014).
Treatment of exotic species is therefore not always as easy as native vs exotic;
care must be taken to focus efforts on species that reduce biodiversity and provide
few resources for pollinators or other grassland fauna.
For nearly as long as naturalists have been observing native plants, they
have been documenting plant invasions into those native environments. Seminal
works like Charles Elton’s 1933 The Ecology of Invasions by Animals and Plants
and extensive research by invasion ecologist Daniel Simberloff and others lead us
4

to today where we have a well-developed understanding of non-native species
ecology and competitive dynamics (Kuebbing and Meyerson 2018).
While much has been written on the means by which non-native plants
succeed, this review only aims to briefly address a few concepts. Mechanisms for
exotic species success are often explained by the ‘Escape from Enemies’ and
‘Novel Weapons’ hypotheses (Hierro et al. 2005). Escape from Enemies has its
roots in Darwin’s theory of naturalization which states that closely related species
will likely be similar in functionality and thus will compete intensely, resulting in
neither species being dominant (Darwin 1859). When an exotic species is
introduced, lower genetic relatedness to co-occurring plants will allow the
introduced species to outcompete co-evolved natives.
Modern molecular techniques allow ecologists to utilize comparative
phylogenetics to tease out how big of a role the relatedness of species can play
regarding whether a non-native plant’s invasion bid is successful or not. A 2009
study in Australia looked at the phylogenetic clustering of nonnative plants in
national parks to determine if a trend of successful, phylogenetically distinct
invasives, could be identified. Researchers found that successful invasives in
Asteraceae, Caryophyllaceae, Poaceae and Solanaceae all seemed to be
phylogenetically clustered at larger spatial scales (Cadotte et al. 2009). Further,
grass invasion in California can also be partially explained by the phylogenetic
distance of native grasses from nonnative ones (Strauss et al. 2006). It follows
logically then that phylogenetically diverse environments would be more resistant
to invasion, which is what researchers found when examining the invasion
5

success of alien plants on Robben Island, South Africa (Yessoufou et al. 2019).
When a plant community is more phylogenetically diverse, i.e. the species in the
community are less closely related, then the probability that an invasive species
will encounter a plant that it is phylogenetically near to increases, and so too does
the chance that those species will be stronger competitors.
The Novel Weapons Hypothesis purports that exotic and invasive species
use various means of changing the chemical and/or microbial environment to a
state that inhibits the growth and fitness of native species. Often this is done
through allelopathy, the process by which biochemicals are released by one
species to the benefit or detriment of neighboring plant species. These can be
active processes like root exudation or passive ones like decomposition and
leeching. Many exotic species weaken the mycorrhizal mutualism of native
species via allelopathy, including the local exotic Cytisus scoparius (Grove et al.
2011).
In both Escape from Enemies and Novel Weapons hypotheses, exotics
succeed due to anachronistic effects, meaning that invasive species enter a system
without a co-evolved history with native competitors. While nothing about plant
invasion is simple, the challenge presented by nonnative species requires careful
study and deliberate methods.

Fire History and Management
Puget Trough Prairies existed naturally from approximately 10,000 years
ago to around 5,000 years ago during a period of warming that came after the

6

return of glacial conditions characteristic of the Younger Dryas (Barnosky 1985).
During this warming period, conifer encroachment diminished, and a frequent fire
regime was established. Prairies thrived naturally throughout this warm period.
Beginning around 5,000 years ago, however, the brief period of dry and warm
weather succumbed to the current cooler and wetter conditions. These new
conditions were much less conducive to natural maintenance of prairie structure
typified by sparse tree growth and plant communities dominated by grasses and
forbs. However, prairies continued to persist thanks to indigenous burning
practices (Boyd 1999). Prairies were incredibly valuable landscapes for native
peoples, offering excellent hunting grounds and gathering opportunities for food
and medicine (Storm and Shebitz 2006). Thanks to indigenous practices, we still
have intact Puget Trough prairies today.
Fire is a critical component of grassland persistence; species that have coevolved histories with fire have been well documented, especially the iconic
indigenous food plant, Camassia quamash (Storm and Shebitz 2006). Gillespie
and Allen (2006) found that a rare California grassland forb, Erodium
macrophyllum, had higher fecundity and germination/ survival in burned plots.
This was likely due to a decreased abundance of exotic grasses, which inhibited
persistence of E. macrophyllum. Prairie restoration, as examined through
ethnographic methods, is often successful when a low intensity frequent fire
regime is re-introduced (Rowe 2010).
Fire, in addition to promoting the persistence of rare prairie endemics, has
another critical job: to help prevent conifer encroachment, something native
7

peoples understood well. Without an established fire regime, conifers easily
invade prairies (Haugo 2010, South Puget Sound Prairies Site Conservation Plan
2002), leading to altered soil moisture and nutrient regimes which may eliminate
the requirements for many prairie species to persist. Today fire, along with
invasive species treatment, constitute the most important restoration actions in the
South Sound Prairies.
While used as a restoration tool, fire can also be manipulated by some
noxious invasives to their great advantage. Bromus tectorum (cheatgrass), a
systemic invasive in the Great Basin, grows rapidly between the well-spaced
sagebrush - bitterbrush scrub that have historically characterized the Great Basin
(Billings 1992). Once established in dense meadows, the cheatgrass rapidly dries
out and burns annually. Native vegetation reestablishes much less aggressively
than cheatgrass, completing the feedback loop (Billings 1992). Moving to the
South Puget Sound prairies, an invasive forb, H. radicata, often flushes after a
fire (personal observation 2017, Buschmann 2000). Without follow-up chemical
treatments, prescribed burning of the prairies would likely result in large
populations of H. radicata, leading to less space for more desirable species to
establish.

Soil Disturbance
In addition to fire, another important disturbance is found on the prairie:
mechanical disruption of soil. Endemic to many prairies in the South Puget
Sound, especially on Joint Base Lewis McChord, are gophers - including the
endangered Mazama Pocket Gopher (Thomomys mazama). Pocket gophers are
8

prolific burrowers, creating constant soil turnover, both below the soil and on the
surface. This soil disturbance has been shown to create micro-conditions more
conducive to increased plant cover, higher species richness, and greater variation
in species composition among disturbed soils (Jones et al. 2008). Further, in a
1997 study, soil disturbance and mound building by gophers also seemed to
increase the presence of a rare forb, Aster curtis, (Hartway and Steinburg 1997).
While the micro-disturbances caused by pocket gophers have helped
promote a more heterogeneous landscape, they can also increase the number of
exotic forbs if the prairie is already invaded to a moderate degree (Hartway and
Steinburg 1997). The degree to which the undisturbed soils in proximity to gopher
burrow sites are invaded strongly influences how well invasive species
outcompete natives in the disturbed soils (Hartway and Steinburg 1997);
essentially, the gopher disturbed soils magnify the native-exotic dynamics found
in nearby undisturbed soils. The success of the most ruthless Puget Trough prairie
invasive, Scotch broom (Cytisus scoparius), is greatly enhanced by soil
disturbance. A study on Scotch broom in the Northern Californian grasslands,
found that soil disturbance (meant to mimic that of a gopher) resulted in a
significantly higher number of C. scoparius recruits in comparison to reference
plots (Bossard 1991). Indeed, many forms of soil disturbance resulting from
anthropogenic sources such as road building, cattle ranching, and agriculture have
increased presence of alien exotic species (Hobbs and Huenneke 1992).
Restoration of the South Sound prairies may include reintroduction of
historic disturbance regimes including soil tillage via gophers and fire, both
9

coupled with targeted removal of exotic species and problematic native species
(i.e., Douglas fir, Pseudotsuga menziesii). Juggling these tasks can be tricky given
that invasive species can take advantage of many of the same disturbances that
natives rely on. In addition to management surrounding the maintenance of
disturbance, native species need to be actively managed. Active, not passive,
restoration that emphasizes planting and seeding native species, is often needed to
restore high levels of native diversity.

Native Species Restoration and Impacts on Ecosystem Functioning
Restoration ecology informs us of the methods needed to achieve more
native diversity, however, why go through all the trouble and money to strive for
near historic levels of native diversity, especially when passive regeneration has
been fruitful in other systems (Crouzeilles et al. 2017, Prach and Moral 2015). For
experimental and applied ecologists, many would argue that increased native
biodiversity can be linked to an increase in ecosystem functioning (Balvanera et
al. 2006, Cardinale et al. 2006, Frainer et al. 2013, Tilman and Downy 1994);
although some theoreticians are less than convinced on the positive link between
species diversity and ecosystem function and stability (Schwartz et al. 2000).
Ecosystem functioning itself can be complex to measure experimentally
but can be thought of as the cumulative biotic and abiotic processes of an
ecosystem that contribute to its inherent sustainability, resilience, and resource
transfer dynamics. Ecosystem functions are basically mechanisms that help
deliver ecosystem services. Ecosystem services encompass every ecosystem
product that promotes a sustainable system for humanity. Clean air and water are
10

classic examples of ecosystem services. Ecosystem functions can be as large scale
as the spawning salmon’s role in riparian nutrient cycling to something as small
as a single vole digging for a truffle, which turns organic matter into the soil,
resulting in higher soil fertility.
Native plant diversity, when examined through various diversity indices
such as richness, evenness, Shannon’s diversity, etc., contributes strongly to
ecosystem functioning by creating redundancy, an important component of a
system’s resilience (Meadows 2009). Redundancy can improve the strategic use
of resources and, hence, ecosystem resiliency. For instance, diverse plant
assemblages may exploit water more efficiently than lower diversity assemblages
when water is a limiting resource, which occurs regularly in grassland ecosystems
(Harpole, Potts, & Suding, 2007).
Guderle et al. (2017) carried out a large field study where water uptake
was measured by several biophysical and soil abiotic methods. Species richness
and functional richness were manipulated to low and high levels while depth of
soil penetrated by roots and aboveground biomass were controlled for. Leaf area
size was found to be related to increased water uptake and high diversity
assemblages maintained greater total leaf area than low diversity areas. Diverse
plant assemblages would be more resilient and desirable if increases in water
uptake efficiency are correlated with an increase in species richness. Many
regions in North America can expect longer and hotter summers resulting from
changing climatic regimes (IPCC Climate Change Synthesis Report 2014),
making strategic use of water resources paramount.
11

In addition to facilitation of water uptake, increases in biodiversity also
lead to greater overall primary production (Hector et al. 1999). Primary
production is essential for many ecosystem processes as it supplies the core of
terrestrial food webs (Hui 2012). Diverse suites of plants may lead to greater
water uptake and greater overall primary productivity by taking advantage of
positive plant to plant facilitation processes – especially as environmental
conditions harshen (Brooker et al. 2008). A useful example of this interaction can
be found in alpine plant communities; as elevation increases, interspecific species
competition tends to lessen as harsher conditions are buffered by mutual plant
facilitation. Plant to plant facilitation allows for more specialized resource
utilization as well as more amenable microhabitat abiotic conditions, such as
temperature buffering and enhanced soil moisture (Anthelme et al. 2014, Brooker
et al. 2008). Maximizing resource utilization and primary productivity are yet
more ways in which increased biodiversity builds resiliency.
Another oft cited role for biodiversity of native plant species is found in
the biotic resistance hypothesis. This hypothesis addresses the question of why
invasion succeeds or fails at a given site. It proposes two outcomes dependent on
diversity: communities with high levels of biodiversity have very little niche
space that can be exploited, resulting in competitive exclusion of an introduced
invasive species. On the other hand, a community depauperate of species will
have more unexploited niche space and be vulnerable to alien species gaining a
foothold (Elton 1958). In a meta-analysis by Levine et al. (2004), the notion that
communities with high levels of diversity will expel invaders requires a more

12

nuanced understanding. It seems that invasion is rarely completely repelled, rather
diversity is negatively correlated with the abundance of invader species (i.e.
higher diversity leads to lower invader abundance). In an examination of four
possible mechanisms underpinning biotic resistance (competition, species
diversity, herbivory, and soil fungal composition) Levine et al. (2004) found that
all factors except soil fungal communities contributed to biotic resistance.
Pollinator services have been shown to increase when grasslands retain
high floristic diversity (Collinge et al. 2003, Potts et al. 2009). Pollinator species,
like many groups of vulnerable species, are experiencing global declines (Potts et
al. 2010). Drivers of these declines include the strong effects of habitat loss and
changing environmental conditions, including changes in seasonal weather
patterns. The presence of harmful chemical classes used in pesticides in the
environment is also threatening to pollinator health (Potts et al. 2010) and
persistent non-lethal effects to pollinators remain a cause for concern (Morandin
et al. 2005). Restoring native grasslands is a necessary action to help mitigate
some of the difficult challenges pollinators face in the Anthropocene.
While most of the conversation here has revolved around the effect of
increasing biodiversity, there is some debate as to the utility of emphasizing
species diversity (often measured as species richness) over functional diversity.
Functional diversity is defined as follows in the excellent review by Diaz and
Cabido (2001):

13

“(Functional diversity) is the value and range of functional traits of
the organisms present in a given ecosystem. The value of traits refers
to the presence and relative abundance of certain values (or kinds)
of leaf size, nitrogen content, canopy heights, seed dispersal and
dormancy characteristics, vegetative and reproductive phenology,
etc. The range of traits refers to the difference between extreme
values of functional traits, for example, the range of leaf sizes,
canopy heights, or rooting depths deployed by different plants in an
ecosystem.”
Using this definition, it is possible that adding plants with very redundant
traits may increase species diversity, but not functional diversity. While in most
circumstances, species richness may be a good surrogate for functional diversity,
functional diversity and species diversity do have subtle differences. Short-term
fluxes in energy and primary production are more strongly influenced by
functional diversity while species diversity within functional traits helps to
increase more long-term sustainability (Diaz and Cabido 2001). Researchers may
find value in using species diversity as a broad measurement of ‘biodiversity,’
however, considering both functional and species diversity may paint a more
holistic picture of ecosystem dynamics.
At its core, restoration ecology is the restoration of ecosystem functions.
Understanding what functions are lost and/or desired is a critical step in the
restoration process. Insights from experimentation and theory can help to inform
how native species improve functioning of water usage, resistance to invasion,
pollination, etc. While different methods to measure and evaluate the effects of
diversity on ecosystem functioning are available, the broader conceptual ties
between diversity and ecosystem functioning are key to the underpinnings of any
successful restoration project.

14

Species Rarity: Causes and Conservation
There are two types of rare plants in a restoration setting. First, there are
plants that have become rare due to disturbances they have not co-evolved with
(excessive habitat fragmentation, conversion to agriculture, invasive species, etc.),
and second, plants that are naturally rare across the landscape. This thesis is
concerned with restoring plants whose rarity is natural and whose persistence is
more fragile. Abundant species are habitat generalists, whereas rare species tend
to be more habitat specialists (Pärtel 2002). Rare species conservation poses
unique challenges due to the nature of habitat specialists. When a species is a
habitat specialist, degradation of that habitat disproportionately harms that species
compared to a generalist that will thrive in several different habitat types
(Reinartz 1997). Before delving into the conservation and ecology of rare plants,
understanding what causes plant rarity needs be addressed.
One way to determine the cause of rarity is to ask if a rare plant is seed or
habitat limited (Candeias and Warren 2016). A species may not be releasing
enough propagules into ideal habitat to be able to compete interspecifically (seed
limitation), or conversely a species may be dispersing in inappropriate habitat,
resulting in loss of local populations (habitat limitation). In a study on rare plant
persistence in gravelly glacial outwash soils in New York State, Candeias and
Warren (2016) found three prairie forbs whose rarity on the landscape was better
explained by habitat limitation than seed limitation. Further, the researchers
discussed how competition from abundant species was a leading cause of failure
to survive after germination.
15

As mentioned before, plant species can be split into two groups:
generalists and habitat specialists. The Douglas Fir (Pseudotsuga menziesii)
illustrates the generalist species well; this impressive conifer is one of the most
iconic and recognizable species in the Pacific Northwest. This partially has to do
with the cultural values attached to this tree, as well as the fact that it is highly
abundant. In every ecosystem there are species like the Douglas Fir, those that are
numerous and successful. Now consider the Matsutake mushroom (Trichloma
matsutake); while it is difficult to throw a rock in a PNW forest and not hit a
Douglas Fir, chances are most people living in the PNW have not found a
Matsutake mushroom in a wild setting. These fungi form a symbiotic relationship
with only a handful of tree species (Yamanaka et al. 2014). The lifecycle and
symbiotic requirements of Matsutake mushrooms makes them rare in the
landscape. Species abundance curves (whether plants, animals, or fungi) in ideal
circumstances are comprised of a few very abundant and successful species, a
group of moderately abundant species, and an abundance of rare species (McGill
et al. 2007, Mouillot et al. 2013, Preston 1948). This has implications for the
conservation of biodiversity; when restoration seeks to restore historic, or near
historic, levels of biodiversity, this often means restoring rare species.
Care should also be taken to not conflate sites with high species richness
with sites where rare and/or threatened species are found. Species rich sites and
the presence of rare species do not always go together; Prendergast et al. (1993),
found that more often species-rich sites do not contain rare species. To further
complicate the issues surrounding the conservation of rare plants is the fact that

16

rare plants are often poor competitors against native habitat generalists (Candeias
and Warren 2016, Lloyd et al. 2002).
As discussed above, rare plant restoration has the potential to boost overall
biodiversity, which theoretically improves ecosystem functioning. In addition to
the increase in biodiversity, there is experimental evidence for unique
contributions made by rare plant restoration. Researchers removing plants from
experimental plots can control for consistent biomass and level of soil
disturbance, while manipulating the species richness by pulling rare species out of
experimental plots and pulling the equivalent biomass of abundant species out of
control plot. Using this approach, Hector et al. (1999) found that plots with rare
species left intact had increased survival of several native species that were sown
as a restoration treatment.
While rare plants are often thought to be poor competitors, evidence
shows this isn’t always the case. In a study looking at plant competition between
rare and common plants of the families Rosaceae and Poaceae, the common
Poaceae species outcompeted the rare ones, while four of the five rare Rosaceae
species were highly competitive both in monoculture with the common plants and
in a mixed plot invaded by the grass Agrostis capillaris (Lloyd et al. 2002).
Competitive outcomes between plants thus can’t be boiled down to the
commonness or rarity of a given species.
One useful exercise for evaluating the importance of rare species is
grounded in quantifying species functional traits and the redundancy of those
traits (Mouillot et al. 2013). First, we can ask if a rare species goes extinct, are the
17

functional traits of that species restricted to just that individual or may those traits
be found in other, more common, species? Another way to ask this: Is the loss of
rare specie’s functional traits (given local extinction) insured against by more
common species also maintaining similar functional traits? When examining three
species-rich ecosystems (marine coral reefs, alpine plant communities, and
tropical forests), researchers found that species with the lowest levels of
functional redundancy tend to be the rarest species in the species pool (Mouillot et
al. 2013). One third of alpine plants, for example, that exhibited the low
functional redundancy, were locally rare. Scaling up, over three quarters of alpine
plants with low levels of functional redundancy are regionally rare (Mouillot et al.
2013). Rare plants are thus shown to possess rare functional traits (for example
lifecycle, leaf area, vegetative height, leaf persistence, dispersion mode, flowering
phenology, etc.) that are not common in their more abundant neighbors. Mouillot
et al. (2013) give an example of the cliff dwelling Pyramidal Saxifrage (Saxifraga
cotyledon), a slow growing rare forb with uniquely long stems which help to
make the flowers detectable to pollinators.
Extant grassland species work both as individual plants and as a
community to help build ecosystem resiliency and deliver ecosystem services.
Attention to rare species and their conservation is an important component of
these processes. A diverse grassland with few specialists will likely lack resilience
and functioning when compared to a diverse grassland supportive of rare species
and habitat specialists (Mouillot et al. 2013). While intensifying climate change
has been shown to be extremely disruptive of plant species distributions,

18

abundances, and interactions – often to the increased benefit of generalists and
detriment to specialists (Van der Putten et al. 2010), conservation of rare species
should remain a goal of restoration ecologists and managers, especially given
their unique contributions. This means that active and persistent restoration work
will be needed in the future to ensure the health and viability of our rare species
and the environments to which they belong.

19

Literature Review Works Cited
Anthelme, F., Cavieres, L. A., & Dangles, O. (2014). Facilitation among plants in
alpine environments in the face of climate change. Frontiers in Plant Science, 5.
Balvanera, P., Pfisterer, A. B., Buchmann, N., He, J.-S., Nakashizuka, T., Raffaelli, D.,
& Schmid, B. (2006). Quantifying the evidence for biodiversity effects on
ecosystem functioning and services. Ecology Letters, 9(10), 1146–1156.
Barnosky, C. W. (1985). Late Quaternary vegetation near Battle Ground Lake,
southern Puget Trough, Washington. Geological Society of America Bulletin,
96,263–271.
Billings, W. D. (1992). Ecological Impacts of Cheatgrass and Resultant Fire on
Ecosystems in the Western Great Basin. Paper presented at the Symposium on
Ecology, Management, and Restoration of lnter-mountain Annual Rangelands, 9.
Bossard, C. C. (1991). The Role of Habitat Disturbance, Seed Predation and Ant
Dispersal on Establishment of the Exotic Shrub Cytisus scoparius in California.
American Midland Naturalist, 126(1), 1.
Boyd, R. (1999). Indians, Fire, and the Land in the Pacific Northwest. Oregon State
University Press, Corvallis.
Brooker, R. W., Maestre, F. T., Callaway, R. M., Lortie, C. L., Cavieres, L. A.,
Kunstler, G., … Michalet, R. (2008). Facilitation in plant communities: the past,
the present, and the future. Journal of Ecology, 96(1), 18–34.
Buschmann, G. (2000). Weeds of South Puget Sound. Cascadia Prairie Oak
Partnership.

20

Cadotte, M. W., Hamilton, M. A., & Murray, B. R. (2009). Phylogenetic relatedness
and plant invader success across two spatial scales. Diversity and Distributions,
15(3), 481–488.
Candeias, M., & Warren, R. J. (2016). Rareness starts early for disturbance-dependent
grassland plant species. Biodiversity and Conservation, 25(13), 2771–2785.
Cardinale, B. J., Srivastava, D. S., Emmett Duffy, J., Wright, J. P., Downing, A. L.,
Sankaran, M., & Jouseau, C. (2006). Effects of biodiversity on the functioning of
trophic groups and ecosystems. Nature, 443(7114), 989–992.
Ceballos, G., Davidson, A., List, R., Pacheco, J., Manzano-Fischer, P., Santos-Barrera,
G., & Cruzado, J. (2010). Rapid Decline of a Grassland System and Its Ecological
and Conservation Implications. PLoS ONE, 5(1).
Collinge, S. K., Prudic, K. L., & Oliver, J. C. (2003). Effects of local habitat
characteristics and landscape context on grassland butterfly diversity.
Conservation Biology, 17(1), 178–187.
Crouzeilles, R., Ferreira, M. S., Chazdon, R. L., Lindenmayer, D. B., Sansevero, J. B.
B., Monteiro, L., … Strassburg, B. B. N. (2017). Ecological restoration success is
higher for natural regeneration than for active restoration in tropical forests.
Science Advances, 3(11), e1701345.
Darwin, C., (1859). On the origin of species by means of natural selection, or
preservation of favored races in the struggle for life. London: John Murray.
Dıá z, S., & Cabido, M. (2001). Vive la difference: Plant functional diversity matters to
ecosystem processes. Trends in Ecology & Evolution, 16(11), 646–655.

21

Dunwiddie, P. W., & Martin, R. A. (2016). Microsites Matter: Improving the Success
of Rare Species Reintroductions. PLoS ONE, 11(3), e0150417.
Elton, C. S., (1958). The Ecology of Invasions by Animals and Plants. Springer
Netherlands.
Frainer, A., McKie, B. G., & Malmqvist, B. (2013). When does diversity matter?
Species functional diversity and ecosystem functioning across habitats and
seasons in a field experiment. Journal of Animal Ecology, 83(2), 460–469.
Gillespie, I. G., & Allen, E. B. (2006). Fire and competition in a southern California
grassland: impacts on the rare forb Erodium macrophyllum. Journal of Applied
Ecology, 41(4), 643–652.
Greipsson, S., & DiTommaso, A. (2006). Invasive non-native plants alter the
occurrence of arbuscular mycorrhizal fungi and benefit from this association.
Ecological Restoration, 24(4), 236–241.
Grove, S., Parker, I., Haubensak, K. (2011). Soil legacy effects of Scotch broom
invasion on Douglas-fir mycorrhizae. Paper Presented at 96th ESA Annual
Convention. Austin, TX.
Guderle, M., Bachmann, D., Milcu, A., Gockele, A., Bechmann, M., Fischer, C., …
Hildebrandt, A. (2017). Dynamic niche partitioning in root water uptake
facilitates efficient water use in more diverse grassland plant communities.
Functional Ecology. 32(1), 214-227.
Harpole, W. S., Potts, D. L. and Suding, K. N. (2007), Ecosystem responses to water
and nitrogen amendment in a California grassland. Global Change Biology, 13:
2341-2348.

22

Hartway, C and E.K. Steinberg. (1997). The influence of pocket gopher disturbance on
the distribution and diversity of plants in western Washington prairies, p. 131–
139. In: P. Dunn and K. Ewing (eds.). Ecology and conservation of the south
Puget Sound prairie landscape. The Nature Conservancy of Washington, Seattle,
U.S.A
Haugo, R. D. (2010). Causes and Consequences of Conifer Invasion into Pacific
Northwest Grasslands. University of Washington.
Haveren, B. P. V. (1983). Soil Bulk Density as Influenced by Grazing Intensity and
Soil Type on a Shortgrass Prairie Site. Journal of Range Management, 36(5), 586.
Hector, A., Schmid, B., Beierkuhnlein, C., Caldeira, M., Diemer, M., Dimitrakopoulos,
P., … Lawton, J. (1999). Plant Diversity and Productivity Experiments in
European Grasslands. Science, 286.
Hierro, J. L., Maron, J. L., & Callaway, R. M. (2005). A biogeographical approach to
plant invasions: The importance of studying exotics in their introduced and native
range. Journal of Ecology, 93(1), 5–15.
Hobbs, R. J., & Huenneke, L. F. (1992). Disturbance, Diversity, and Invasion:
Implications for Conservation. Conservation Biology, 6(3), 324–337.
Hobbs and Harris (2001) Restoration Ecology: Repairing the Earth’s Ecosystems in the
New Millennium. Restoration Ecology, 9(2), 239–246
Hui, D. (2012) Food Web: Concept and Applications. Nature Education
Knowledge 3(12):6
IPCC. – (2014): Climate Change 2014: Synthesis Report. Contribution of Working
Groups I, II and III to the Fifth Assessment Report of the Intergovernmental Panel
23

on Climate Change [Core Writing Team, R.K. Pachauri and L.A. Meyer (eds.)].
IPCC, Geneva, Switzerland, 151pp.
Isselstein, J., Jeangros, B., & Pavlu, V. (2005). Agronomic aspects of biodiversity
targeted management of temperate grasslands in Europe – A review. Agronomy
Research. 3(2).
Jackson, S. T., & Hobbs, R. J. (2009). Ecological Restoration in the Light of
Ecological History. Science, 325(5940), 567–569.
Jones, C. C., Halpern, C. B., & Niederer, J. (2008). Plant Succession on Gopher
Mounds in Western Cascade Meadows: Consequences for Species Diversity and
Heterogeneity. The American Midland Naturalist, 159(2), 275–286.
Kaye, T. N., Stanley, A. G., & Ross, D. (2011). Dispersal behavior and habitat
variation of Taylor’s checkerspot butterfly (p. 27). Institute for Applied Ecology.
Kaye, T. N. (2009). Toward successful reintroductions: The combined importance of
species traits, site quality, and restoration technique. Proceedings of the CNPS
Conservation Conference, 17, 99–106
Kraft, N. J. B., & Ackerly, D. D. (2014). Assembly of Plant Communities. In R. K.
Monson (Ed.), Ecology and the Environment (pp. 67–88). New York, NY:
Springer New York.
Kuebbing, S. E., & Meyerson, L. A. (2018). Honoring Daniel Simberloff: an
unwavering champion of invasion biology. Biological Invasions, 20(12), 3379–
3383.

24

Laughlin, D. C., Strahan, R. T., Moore, M. M., Fulé, P. Z., Huffman, D. W., &
Covington, W. W. (2017). The hierarchy of predictability in ecological
restoration: are vegetation structure and functional diversity more predictable than
community composition? Journal of Applied Ecology, 54(4), 1058–1069.
Levine, J. M., Adler, P. B., & Yelenik, S. G. (2004). A meta-analysis of biotic
resistance to exotic plant invasions. Ecology Letters, 7(10), 975–989.
Lloyd, K., Lee, W., & Wilson, J. (2002). Competitive Abilities of Rare and Common
Plants: Comparisons Using Acaena (Rosaceae) and Chionochloa (Poaceae) from
New Zealand. Conservation Biology, 16(4), 975-985.
McGill, B. J., Etienne, R. S., Gray, J. S., Alonso, D., Anderson, M. J., Benecha, H. K.,
Dornelas, M., Enquist, B. J., Green, J. L., He, F., Hurlbert, A. H., Magurran, A.
E., Marquet, P. A., Maurer, B. A., Ostling, A., Soykan, C. U., Ugland, K. I. and
White, E. P. (2007),- Species abundance distributions: moving beyond single
prediction theories to integration within an ecological framework. Ecology
Letters, 10: 995–1015.
Meadows, Donella H. (2009). Thinking in systems: a primer. London; Sterling, VA:
Earthscan
Millennium Ecosystem Assessment. Ecosystems and human well-being: desertification
synthesis. Washington, D. C.: World Resources Institute; 2005.
Morandin, L. A., Winston, M. L., Franklin, M. T., & Abbott, V. A. (2005). Lethal and
sub-lethal effects of spinosad on bumble bees (Bombus impatiens Cresson). Pest
Management Science, 61(7), 619–626.

25

Mouillot, D., Bellwood, D. R., Baraloto, C., Chave, J., Galzin, R., Harmelin-Vivien,
M., … Thuiller, W. (2013). Rare species support vulnerable functions in highdiversity ecosystems. PLoS biology, 11(5).
Newsham, K. K., Fitter, A. H., & Watkinson, A. R. (1995). Arbuscular Mycorrhiza
Protect an Annual Grass from Root Pathogenic Fungi in the Field. Journal of
Ecology, 83(6), 991–1000.
Öpik, M., Moora, M., Liira, J. and Zobel, M. (2006), Composition of root-colonizing
arbuscular mycorrhizal fungal communities in different ecosystems around the
globe. Journal of Ecology, 94: 778–790
Parniske, M. (2008). Arbuscular mycorrhiza: the mother of plant root endosymbioses.
Nature Reviews Microbiology, 6(10), 763–775.
Pärtel, M. (2002), Local Plant Diversity Patterns and Evolutionary History at the
Regional Scale. Ecology, 83: 2361–2366.
Pimentel, D., Lach, L., Zuniga, R., & Morrison, D. (2000). Environmental and
Economic Costs of Nonindigenous Species in the United States. BioScience,
50(1), 53.
Potts, S. G., Biesmeijer, J. C., Kremen, C., Neumann, P., Schweiger, O., & Kunin, W.
E. (2010). Global pollinator declines: trends, impacts and drivers. Trends in
Ecology & Evolution, 25(6), 345–353.
Potts, S. G., Woodcock, B. A., Roberts, S. P. M., Tscheulin, T., Pilgrim, E. S., Brown,
V. K., & Tallowin, J. R. (2009). Enhancing pollinator biodiversity in intensive
grasslands. Journal of Applied Ecology, 46(2), 369–379.

26

Prach, K., & Moral, R. del. (2015). Passive restoration is often quite effective:
Response to Zahawi et al. (2014). Restoration Ecology, 23(4), 344–346.
Prendergast, J. R., Quinn, R. M., Lawton, J. H., Eversham, B. C., & Gibbons, D. W.
(1993). Rare species, the coincidence of diversity hotspots and conservation
strategies. Nature, 365(6444), 335–337.
Preston, F. W. (1948), The Commonness, And Rarity, of Species. Ecology, 29, 254–
283.
Schlaepfer, M. A. (2018). Do non-native species contribute to biodiversity? PLOS
Biology, 16(4), e2005568.
Schwartz, M., Brigham, C., J. D. Hoeksema, Lyons, K., Mills, M., & P. J. van
Mantgem. (2000). Linking Biodiversity to Ecosystem Function: Implications for
Conservation Ecology. Oecologia, 122(3), 297-305.
Solecki, M. K. (1997). Controlling Invasive Plants. In The Tallgrass Restoration
Handbook (pp. 251–278). Island Press.
Stinson, D. W. (2005). Washington State Status Report for the Mazama Pocket
Gopher, Streaked Horned Lark, and Taylor’s Checkerspot. Washington
Department of Fish and Wildlife, Olympia. 129+ xii pp.
Storm, L., & Shebitz, D. (2006). Evaluating the Purpose, Extent, and Ecological
Restoration Applications of Indigenous Burning Practices in Southwestern
Washington. Ecological Restoration, 14(4), 256-268.
Reinartz, J. (1997). Restoring Populations of Rare Plants. In The Tallgrass Restoration
Handbook (pp. 89–95).

27

Rowe, H. I. (2010). Tricks of the Trade: Techniques and Opinions from 38 Experts in
Tallgrass Prairie Restoration. Restoration Ecology, 18, 253–262.
Ruiz‐Jaen, M. C., & Aide, T. M. (2005). Restoration Success: How Is It Being
Measured? Restoration Ecology, 13(3), 569-577.
Schultz, C. B. (2001). Restoring resources for an endangered butterfly. Journal of
Applied Ecology, 38(5), 1007–1019.
Strauss, S. Y., Webb, C. O., & Salamin, N. (2006). Exotic taxa less related to native
species are more invasive. Proceedings of the National Academy of Sciences,
103(15), 5841–5845.
Vale, T. (2002). Fire, native peoples, and the natural landscape. Washington, DC:
Island Press.
Vályi, K., Mardhiah, U., Rillig, M. C., & Hempel, S. (2016). Community assembly and
coexistence in communities of arbuscular mycorrhizal fungi. The ISME Journal,
10(10), 2341–2351.
Van der Putten, W. H., Macel, M., & Visser, M. E. (2010). Predicting species
distribution and abundance responses to climate change: why it is essential to
include biotic interactions across trophic levels. Philosophical transactions of the
Royal Society of London. Series B, Biological sciences, 365(1549), 2025–2034.
Waters, S., Fisher, S., and Hille Ris Lambers, J. (2014). Neighborhood-contingent
indirect interactions between native and exotic plants: multiple shared pollinators
mediate plant reproductive success during invasions. Oikos 123: 433-440
Washburn, A. L. (1988). Mima Mounds: An Evaluation of Proposed Origins with
special Reference to the Puget Lowlands (p. 59). Washington DNR.

28

Yamanaka, T., Ota, Y., Konno, M., Kawai, M., Ohta, A., Neda, H., … Yamada, A.
(2014). The host ranges of conifer-associated Tricholoma matsutake, Fagaceaeassociated T. bakamatsutake and T. fulvocastaneum are wider in vitro than in
nature. Mycologia, 106(3), 397–406.
Yessoufou, K., Bezeng, B. S., Gaoue, O. G., Bengu, T., & van der Bank, M. (2019).
Phylogenetically diverse native systems are more resistant to invasive plant
species on Robben Island, South Africa. Genome, 62(3), 217–228.

29

Chapter Two: Thesis Research
Introduction
Loss of grassland biodiversity may not tug on the heartstrings as much as
loss of pristine tropical rainforest, however grasslands provide niche habitats for
many rare and threatened species and, like tropical rain forests, are consistently
threatened by a myriad of anthropogenic forces (Bond and Parr 2010). Our very
own Puget Trough prairies, along with their cousins to the south - the Willamette
Valley prairies, are situated in areas where the human population is forecast to
grow dramatically in the next decade, which will inevitably carry with it more
intensive land use modification (Thurston Regional Planning Council 2017).
Further, our prairies are highly fragmented, which reduces dispersal of many rare
and hard-to-establish native prairie plants (Soons and Heil 2002). This fragmented
landscape with reduced opportunity for species dispersal is subject to the forces of
extinction as determined by the theory of island biogeography (MacArthur and
Wilson 1967). This ultimately means that land managers must actively manage
these landscapes by removing invasive species, reintroducing historic fire
regimes, and augmenting both extant and locally extinct native plant populations
through seeding and plug installation (Dunwiddie and Bakker 2011).
Understanding the complex ecology of our Puget Trough prairies will allow for
these management techniques to be fine-tuned so that the most cost-effective and
impactful methods can be used.

30

Puget Balsamroot (Balsamorhiza deltoidea, hereafter Balsamroot) is a
perennial forb in the aster family that is listed on the Washington Natural Heritage
Rare Species list for 2018 (www.dnr.wa.gov/nhplists). In Sarah Krock’s thesis,
which looked at the effect of sowing time and site diversity on native germination
and establishment, Krock (2016) found that Balsamroot had extremely low
establishment rates, often less than 1%. While Balsamroot is rare on the landscape
and has experienced low field establishment, it does grow very well at the Center
for Natural Management’s (CNLM) seed farm, suggesting some field-linked
parameter limits Balsamroot establishment (S. Hamman, personal
communication). Balsamroot yields large, sunflower-like flowers that bloom
throughout the late spring and early summer, providing a high-quality nectar
source to species such as the Taylor’s checkerspot butterfly (Linders et al. 2015).
Blanket flower (Gaillardia aristata), another perennial forb in the aster
family, is not listed as endangered or threatened in the state of Washington and is
common east of the cascades in dry and open locales (www.pnwherbaria.org).
West of the Cascades, however, it is rare or absent in many of the prairies in the
South Sound (personal observation, 2017). Like Balsamroot, Blanket flower
yields large sunflower-like flowers and is an excellent late-season nectar source
for many pollinator species (Lee-Mäder et al. 2016). Given its preference for
well-drained soils, Blanket flower represents a good candidate to help increase
native diversity, thus improving habitat for many pollinators in the South Sound
prairies.

31

Both Balsamroot and Blanket flower represent species that have not been
successfully restored on a large scale and provide quality native nectar sources for
rare and endangered invertebrates. The Taylor’s checkerspot butterfly
(Euphydryas editha taylori), federally listed as endangered in 2013, is at extreme
risk of extinction, primarily due to the loss of habitat (Stinson 2005, U.S. Fish and
Wildlife Service 2013).
Taylor’s checkerspot butterfly populations vary significantly year by year
due to changing weather patterns. During the larval stages of the butterfly, a host
plant provides vital shelter and nutrition. Larvae must reach an adequate level of
development before plants become desiccated with the arrival of summer and a
long diapause period lasting throughout winter (Stinson 2005). This means many
sites become locally extinct if local conditions aren’t amenable to the fragile
butterfly life-cycle. Maintenance of Taylor’s checkerspot populations is thus
highly dependent on dispersing adults who may recolonize sites that were
unsuitable in previous seasons. Loss of habitat represents a huge challenge to the
persistence of Taylor’s checkerspot and other sensitive pollinators. The story of
the Taylor’s checkerspot is not too dissimilar to that of the spotted owl; both
species are quite rare with their threatened status partially due to the loss of
sensitive habitats (old-growth forest and diverse grasslands able to support
numerous metapopulations to cope with stochastic pressures).
Given the frequent stochastic pressures butterflies face, along with the loss
of habitat due to anthropogenic forces, it follows that augmenting current habitat
and creating new habitat are critical strategies for the recovery of the Taylor’s
32

checkerspot butterfly. This strategy means that seeding/ plug installation of native
forbs to establish populations of both ideal host plants and spring nectar sources
that are available through the various stages of the pollinator’s lifecycle are
essential. Available densities of both host plants and adult nectar sources have
been shown to be strongly linked to population densities of the Fender’s blue
butterfly, a sensitive butterfly species in Oregon prairies (Schultz and Dlugosch
1999).
Unfortunately, restoring many native species to the prairies is no easy task
due to many of the challenges facing restoration of native forbs to the South
Sound prairies. However, one key issue warrants mentioning here: low
germination rates. While this is typical for many restoration projects, low
germination presents a major challenge, especially when seed stock is limited or
expensive to procure. Wilson et al. (2004) monitored seeding establishment in the
Willamette Valley prairies for two years looking at one forb from the Borage
family and three forbs from the Aster family. The researchers found that all the
study forbs were marred by extremely low germination and none of the forbs had
a cover over 1.6% after two years. Another study by Applestein et al. (2018)
monitored seeding establishment, as affected by seeding method (seed-drill,
hydro-mulch, and broadcast) and seeding rate, in the South Sound prairies over
three years. Intuitively, the most predictive measure of establishment for year one
of monitoring was seeding rate, while the most predictive measure for
establishment in year three was year one establishment. Even at the highest
seeding rate, plant density (measured by cover) never topped 10% and was often

33

less than 5% (Applestein et al. 2018). This suggests that establishment of three
study species (families Rosaceae, Asteraceae, and Poaceae) is very limited by
seed availability in the field.
Compounding the issue of low germination is that after years of invasion
and fire suppression, prairie soils are impoverished of native seed. An expansive
study of the Puget Trough – Willamette Valley prairies by Stanley et al. (2011)
found that no level of invasive removal treatment increased native plant cover.
The only treatment that increased plant cover even moderately was seed addition,
emphasizing how seed limited prairies in this ecoregion are and the importance of
native seed additions as part of the restoration process. One way to address the
issue of low germination is identifying microsites where seeds have the highest
probability of germination and survival.
The Puget Trough prairies historically contained a varied and diverse mix
of soil types, but most of the deeper and more productive soils have been
converted to agriculture. Today many of the remnant prairies consist of ‘Mima
Mounds,’ named for a prominent mounded South Sound prairie. The origin of
these mounds has been contested in the past; however most recent research has
coalesced around the opinion that these mounds are a result of glacial outwash
processes. This idea was first formulated as the Sun cup hypothesis by J. Bretz in
1913. Bretz proposed that coarse sediments accumulating in the melting glaciers
may have collected in pits called sun cups. As these cups melted the ice-captured
soil and gravel would have settled in a mounded shape (Bretz 1913).

34

The various prairie microsite types are defined by two things: elevation
and soil class. Elevation affects the movement of water to and from a microsite.
Low-lying swales, for example, are composed of more mesic soils as water tends
to pool in low lying areas. Pedogenesis at these sites can be attributed to glacial
outwash by the Vashon ice stade, which reached the southernmost point
approximately 17,000 BP, afterwards retreating an average of 340 meters per year
(Porter and Swanson 1998). Mounded and swale microsites typically consist of a
Nisqually-Spanaway soil complex that is differentiated by deeper horizons of
more fine sediments than surrounding shallow and sandy/ gravelly intermounded
and upland microsites consisting of Spanaway soils (Bretz 1913, Dunwiddie and
Martin 2016).
While microsites are inherently defined by abiotic processes (soil
sediment composition, glacial legacies), there is also a biotic understanding
regarding the multiple ways that microsites differ. Looking out on a mounded
prairie there is a general sense that the mounds foster more vigorous plant growth
characterized by denser growing forbs, an abundance of non-native blackberry,
and more bracken ferns. Researching any microsite effect on the restoration of
rare plants requires an examination of both the aboveground and belowground
biota that characterize a given microsite.
Understanding the micro-site preference of rare prairie forbs will allow for
research to help inform management and to get their most bang-for-buck in
restoration work (Dunwiddie and Martin 2016, Falk et al. 1996). Microsites offer
distinct, somewhat undefined, niches that plants seem to be selected for. Past
35

research shows that mounded topographies foster greater plant diversity and
higher survival of Castilleja levisecta, (i.e. golden paintbrush) than either the
intermounded areas or low-lying swales (Dunwiddie and Martin 2016). Further,
this finding has stronger predictive power than both functional group richness and
indicator species– both very relevant due to the hemi-parasitic nature of C.
levisecta (Dunwiddie and Martin 2016). Work by Guerrant and Kaye (2007) in
the Willamette Valley prairies showed that microtopographic position strongly
influenced native forb survival; Lomatium sp. preferred lower topographies with
higher soil moisture while Sericocarpus, Erigeron, and Horkelia spp. found
greater survival in higher, drier microsites.
The microsite effect is noticeable simple by eye, as well. After a wildfire
burned the south parcel of the Scatter Creek Prairie in summer 2017, a noticeable
pattern in the fall regrowth could be seen. Much more vigorous regrowth of
prairie vegetation occurred on the mounds as opposed to swales or intermounded
areas that were all equally burned (personal observation, 2017). Indeed, in
mounded prairies that haven’t been burned, one can still pick out the darker
shaded mounds across the landscape due fewer grasses and more ferns/shrubs.

Mound microsites are prominent to the eye due to dense vegetation (Image taken at Mima Mounds Prairie).

36

Evaluating prairie microsites as a method for identifying appropriate
habitat for species of conservation concern may yield promising results for rare
plant restoration. I hope to address both the practical implication of this strategy
(identifying suitable microsites is a cost-effective means to improve
germination/survival and add more pollinator habitat) and potential mechanistic
controls on native plant establishment in a restoration context (the underlying
reasons as to why a microsite may yield better restoration outcomes). Through
this thesis, I address the following questions:
1.) What is the simplest and most accurate way to describe the
characteristics that differentiate microsites on the prairie landscape?
2.) Which microsites yield the strongest germination performance of
locally rare species, Balsamorhiza deltoidea and Gaillardia aristata?
3.) How do microsite characteristics influence germination of
Balsamorhiza deltoidea and Gaillardia aristata?
Preservation of biodiversity is often seen as an ‘insurance policy’ in the
face of constant change, characteristic of the current Anthropocene (Diaz and
Cabido 2001). While biodiversity is generally thought to increase ecosystem
functioning/ services, this research selected the study forbs due to their status as
late season and high-quality nectar sources for endangered species, not ‘diversity
simply for diversity’s sake.’ A common critique of restoration is that it often lacks
concrete goals that link restored habitat to specific species (Kaye 2009). By
linking restoration of specific plant species with endangered invertebrates this

37

study engages a discipline-wide effort to have concrete restoration goals that
promote the viability of target endangered species (Kaye 2009).

Methods
Site Layout and Description
Sites for this study were chosen to minimize differences in disturbance
regimes and to capture a representative diversity of microsites typical to the South
Sound prairies. Glacial Heritage Preserve [46.865128, -123.040876] and two sites
at Joint Base Lewis McChord (Johnson Prairie [46.927283, -122.734468] and
Training Area 15 [47.012644, -122.440316]) all received a prescribed burn the
year before seeding with B. deltoidea and, in the case of Glacial Heritage, B.
deltoidea and G. aristata. Plot arrangement differed between sites, as the Joint
Base Lewis McChord (JBLM) plots had been established as part of a larger study.
Plot layout at Glacial Heritage consisted of three transects with between 7 and 8
1-m2 quadrats on each transect for a total of 24 quadrats, evenly split between two
microsite types. JBLM sites each consisted of three 150 meter long transects.
Along each transect 10 4 m2 quadrats were placed, generating a total of 30
quadrats. The transects at JBLM and Glacial Heritage were all on soils of the
Spanaway series (Washburn 1998, Dunwiddie and Martin 2016). All three sites
have been actively managed for high quality prairie habitat with ongoing burning
and invasive removal treatments over the past 15-20 years.
Seeding strategy also differed between sites, with the goal of maximizing
seed to ground contact. Plots at Glacial Heritage were mowed with string trimmer,
with the cuttings being raked off before seeds were hand scattered. Each plot at
38

GH received 50 B. deltoidea and 72 G. aristata seeds. The plots at JBLM had 20
‘scratches’ in each plot. In the corners and middle of each quadrat a hand
cultivator was used to scratch the soil surface, after which two seeds of B.
deltoidea were dropped into each scratch by hand.
Microsite identification was primarily based on visual criteria, as easy
identification of the microsite is the most useful to management based on
microsite preferences. Glacial Heritage microsites were placed into two bins:
mounds and the intermounds. JBLM microsites were initially categorized into
four bins: mounds, uplands, slopes, and swales. Visually speaking the differences
between these four microsites was much less stark than the mounds and
intermounds of Glacial Heritage. For the purposes of this thesis, the original four
microsites have been condensed into two categories based on the topographic
position of each microsite for some of the analyses. The ‘highland’ microsites, i.e.
the two highest microsites, consist of the mounds and uplands. The ‘lowland’
microsites, i.e. the two lowest microsites, consist of the slopes and swales. This
was done after finding no significant difference in microsite metrics between the
mounds and the uplands and no significant difference between the slopes and
swales.
Data Collection
Data on seed germination and soil moisture were collected for each site:
once in March, early April, late April, and May. Total germinants were recorded
at each visit, without marking individual plants. Moisture was collected in the
center of each quadrat with a moisture probe at the same time germination was
39

monitored for GH and within one week of monitoring JBLM. Bulk density was
collected in March from all plots at all sites. Bulk density was measured as fineearth bulk density, i.e. the amount of fine (<2 mm) sediments for a given volume.
In addition to collecting data on abiotic parameters, biotic parameters were
also considered in so far as they influenced germination rates. Within each plot
species richness, percent coverage of functional types (grass, forb, bryophyte),
and species status as native or exotic was recorded. For this, a point-intercept
method was used by constructing a meter-squared quadrat with legs and 16
equidistant intersections using twine strung around the frame. A pin was then
dropped at each intersect, perpendicular to the ground. Each plant part that
touched the pin was recorded as a ‘hit.’ The total number of ‘hits’ per plot was
used as a surrogate for plant density. Each hit was also documented as either a
native or non-native and as a grass, forb, or bryophyte. For example, to calculate
the native percent, the number of native hits was divided by the total number of
hits for that plot, then multiplied by 100.
Statistical Methods and Rationale
A variety of different tests were used to explore the central questions of
this thesis. Non-parametric tests were used, as both the count data and data
characterizing the microsites were non-normal, determined most often by a
Shapiro-Wilk test (Ghasemi and Zahediasl 2012). When a simple difference in
data was sought, such as for differences in plant density between microsites,
either a Kruskal Wallace test or Wilcoxon Rank Sum test was used, depending on
the number of microsite-types being considered. Both tests are appropriate for the
40

non-normal nature of these datasets. A negative binomial regression was chosen
to model the influence that microsites themselves and the belowground and
aboveground parameters exert on the count data.
The GLM with a link function designating the negative binomial family
was used here because the negative binomial distribution handles overdispersion
(an issue of higher than expected variance) well, which was the case with zeroinflated count data. A model was considered a good fit when the deviance residual
(based on deviance of the model’s residuals) was less than the five percent critical
chi-squared value (based on the residual for the model’s degrees of freedom)
(Table 6). To determine differences between microsites or sampling periods, an
Estimated Marginal Means (EMM) was used as a post hoc test. It is important to
recognize that an EMM is a prediction based on a model of weighted averages
and not the raw data. Lastly, a negative binomial regression does not provide a R²
as in traditional ordinary least squares (OLS) regression, however the negative
binomial does provide coefficient estimates. Coefficient estimates are equivalent
to one unit of change in the independent variable. In this case, the difference in
the logs of expected values of the dependent variable is expected to change by the
respective coefficient, given other dependent variables in the model are held
constant (stats.idre.ucla.edu).
P-values will be referred to ‘weakly significant’ if p < 0.1 or simply
‘significant’ if p < 0.05. All data were analyzed, and figures created using R
Studio version 3.5.1. Tables reporting the test statistic and other statistical test
information are found in the Tables Appendix.
41

Results
Microsite Characteristics
Bulk Density

Results at Glacial Heritage offer some of the clearest microsite differences
in terms of both germination rates and microsite characteristics. This is likely the
case due to an experimental design that defined only two microsite types for GH.
When considered in abiotic terms, mounds (M) and intermounds (IM) at GH
differ by both soil moisture and bulk density, although the disparity was much
larger for bulk density (Figure 1). Average bulk density for mounds (0.25 g/cm³,
sd = 0.25), was significantly lower than intermounds (0.36 g/cm³, sd = 0.36)
(F(191,190) = -0.106, p < 0.01) (Table 1).
Results at JBLM show less of clear difference in microsite bulk density,
although one trend stayed consistent throughout the two sites. In both Training
Area 15 (TA15) and Johnson Prairie (JP) mounds had a lower bulk density than
either of the other three microsite types. Using a GLM, the disparity in bulk
density for JP was significant between mounds and slopes (p < 0.05), while
weakly significant between mounds and swales (p < 0.1), and mounds and upland
microsites (p < 0.1) (Table 2). At TA15 mound bulk density exhibited a weakly
significant difference from the upland microsites (p < 0.1) (Table 3).
Moisture

Disparities in moisture between the two microsites at GH were less
pronounced than differences in bulk density, although a clear trend throughout the
spring showed that intermounded areas retained a higher soil moisture (Figure 2).
42

Mounds at Glacial Heritage throughout the season on average maintained a soil
moisture of 19.14% (sd = 11.92), while intermounded areas maintained an
average soil moisture of 24.96% (sd = 14.87) (Table 1). High standard deviations
for these values may be due to near zero percent moisture in May once soils had
largely dried up. A Wilcoxon rank sum test confirmed that moisture differed by
microsite (w = 0.6148, p < 0.05). Further, an EMM post-hoc test found there was
a significant difference in moisture between mounds and intermounds during the
March and early April sampling periods (Table 4).
Both TA15 and JP showed almost no difference in moisture between
microsites; moisture stayed constant through the season until soils dried out in
May (Figure 3). In TA15 upland sites did have a slightly higher soil moisture
content than other microsites, however, not enough to register as significant using
a Kruskal Wallace test (h(2) = 3.4037, p > 0.1). Moisture values in May for TA15
were much higher for JP, however this is likely due to a rain event that happened
between sampling the two sites.
Aboveground Biotic Parameters

Considering all the aboveground biotic responses, the strongest difference
between GH mounds and intermounds was that of plant density: mounds were
characterized by a higher number of hits (avg: 35.17, sd: 7.41) than intermounds
(avg: 22.83, sd: 4.62) due to more vigorous plant growth (Wilcoxin rank sum test,
w = 2.5, p < 0.05) (Figure 4, Table 5). The values for nonnative cover also
showed mild, but not statistically significant, differences between GH microsites
with exotic cover on mounds averaging 53.03% (sd = 29.09), compared to that of
43

intermounds, which averaged 44.8% (sd = 7.04) cover (Figure 5, Table 5). These
differences in exotic cover, however, were not detectable by a Wilcoxon rank sum
test (w = 15, p > 0.1). Species richness (w = 12.5, p > 0.1, Figure 6), native cover
(w = 21, p > 0.1, Figure 7), forb cover (w = 17, p > 0.1, Figure 8), and grass
percent cover (w = 22 , p > 0.1, Figure 9), also did not significantly differ by
microsite.
While differences did exist between the JBLM highland (upland and
mound sites) and lowland (slope and swale) microsites for functional and native
vs. nonnative cover, these differences were not consistent between JP and TA15.
At JP a Wilcoxon rank sum test showed a significant difference in species
richness between highland and lowland sites where the lowlands harbored a
greater number of species (w = 4, p = 0.0115) (Figure 6). This difference,
however, was not found at TA15 (Figure 6). No significant differences were
found between microsites for plant density (Figure 4), nonnative cover (Figure 5),
or either forb or grass cover (Figures 8 and 9).
Impacts on Germination
Germination Influenced by Microsite

Glacial Heritage was seeded with two perennial forbs, B. deltoidea and G.
aristata. B. deltoidea emerged earlier in the season, however both G. aristata and
B. deltoidea hit peak germination in late April (Figures 10 & 11). The negative
binomial model (count data treated as the dependent variable and microsite
category as a factored independent variable) was found to be a good fit for both B.
deltoidea and G. aristate (Table 6).
44

While a pattern showing a consistent difference in germination between
GH mounds and intermounds was found for B. deltoidea (Figure 11), the
difference was only predicted by the model to be statistically significant in May
(negative binomial GLM, EMM post-hoc, p < 0.05) (Table 7). This difference and
the overall pattern demonstrate a clear B. deltoidea germination preference for
mounds. Interestingly a stronger preference for intermounds was found for G.
aristata (Figure 10). While the general pattern shows a G. aristata preference for
the intermounds, the difference was weakly significant only in late April (negative
binomial GLM, EMM post-hoc, p < 0.1) (Table 8).
Peak B. deltoidea germination for JP occurred in late April, while
germination at TA15 was surprisingly consistent throughout the season (Figure
12). Average percent germination of B. deltoidea differed throughout the season
between the highland microsites (mounds and uplands) and the lowland
microsites (slopes and swales) for both JBLM sites. At JP the differences in
germination between highland and lowland sites, was weakly significant during
the late April sampling period (negative binomial GLM, EMM post-hoc, p < 0.1)
(Table 9). For TA15, on the other hand, the model predicted significant
differences in germination between highland and lowland sites for early April
(negative binomial GLM, estimated marginal means post-hoc, p < 0.1), late April
(negative binomial GLM, EMM post-hoc, p < 0.05), and May (negative binomial
GLM, EMM post-hoc, p < 0.05) (Table 10).

45

Germination Influenced by Abiotic and Biotic Parameters

Negative binomial models were also used to elucidate any influences on
germination from the abiotic parameters used to help characterize the microsites.
Evaluating bulk density and moisture impacts on GH B. deltoidea germination,
the model predicted a significant negative effect of bulk density on germination
(negative binomial, p < 0.05) (Table 11), however no significant influences
exerted by moisture were found (Table 11).
Using the same approach for the G. aristata counts, the model predicted a
significant positive influence of bulk density on germination (negative binomial, p
< 0.05), while no effect of moisture on germination was found (Table 12). Soil
moisture was not a significant influence on B. deltoidea germination at either JP
or TA15. Bulk density, however, did have a significant effect on germination at
JP (negative binomial GLM, p < 0.05), but not at TA15. Tracking with the GH
results, lower bulk density values at JP yielded a positive influence on B.
deltoidea germination.
Aboveground parameters (functional group cover, species richness, plant
density, and native vs nonnative cover) were also evaluated using a negative
binomial model to find whether an influence on germination exists. Plant density,
grass cover and forb cover all positively impacted GH B. deltoidea germination (p
< 0.05 for all) (Table 13) while G. aristata germination was only weakly
positively affected by forb cover (p < 0.1) (Table 14). The influence of the biotic
parameters (functional cover, native vs nonnative cover) on germination yielded
no significant influence on the germination of B. deltoidea at either JBLM site.
46

Discussion
In 1992, famed ecologist E. O. Wilson declared that the coming century
will be the era of restoration in ecology (E. O. Wilson 1992). Indeed, as the field
of ecology has developed new and powerful methods to demonstrate the
importance of diverse and resilient ecosystems, the need to restore landscapes that
have been heavily degraded has increased greatly. In western Washington state,
native prairies have been subject to fragmentation, species invasion, and a
changing climate. In order to preserve the rich diversity and associated services of
our prairies, active and persistent restoration is needed (Bakker and Dunwiddie
2011). One of the key goals of the restoration of these prairies is establishing and
boosting populations of native forbs.
Native perennial forbs offer some of the strongest habitat for endangered
invertebrates due to the high-quality nectar resources and a long blooming period
– characteristics important to facilitating more resilience in a prairie where
warmer climates threaten the timing of critical pollinator-host interactions
(Hegland et al. 2009, Memmott et al. 2007, Potts et al. 2010, Schweiger et al.
2010). Restoration of several high-quality prairie forbs in the Puget-Trough
prairies is often hindered by extremely low germination. To address this issue the
three critical questions driving this thesis work were: 1.) What is the simplest and
most accurate way to describe the characteristics that differentiate microsites on
the prairie landscape? 2.) Which microsites yield the strongest germination
performance of B. deltoidea and G. aristata? 3.) How do microsite characteristics
influence germination of B. deltoidea and G. aristata?
47

Research on the microsites in the South Sound has largely been driven by
efforts to understand the habitat requirements of rare species. Reestablishing this
habitat for Taylor’s checkerspot butterfly and other pollinators whose populations
and associated ecosystem services are threatened by habitat fragmentation and
degradation is a critical step. Ultimately, the microsite is a convenient way to
define ‘habitat.’ Microsites have long been an important conceptual tool for
ecologists to understand how species-specific recruitment might be stifled,
especially when seed limitation is not the only culprit (Eriksson and Ehrlén 1992).
Even when seed limitation is shown to not be an issue, many species seem to be
microsite limited (Ingersoll and Wilson 1993).
The nature of the microsite and how it becomes defined influences the
microsite’s measurable characteristics, which is ultimately how microsites
manifest effects on germination. Characteristics like light (Severns 2008, Tang et
al. 1992), temperature (Rice 1985), co-occurring vegetation (Donath et al. 2007)
or, in the case of this study, bulk density and moisture (Thill et al. 1979) and
vegetation structure (Ryser 1993) are all variables known to influence plant
germination. Further, all these variables are likely to be differentiated by different
microsite types whether it be patches in grasslands (Rose and Frampton 2010,
Tang et al. 1992), stature of grasses (Rose and Frampton 2010, Severns 2008) or
sites with gopher mounds (Rice 1985).
The role of microsites in a restoration setting is also important to
acknowledge relative to the scale of restoration work. Conducting research on a
scale that is too small or too skewed as to not represent the greater landscape will
48

not yield helpful data to inform restoration actions just as research on too large a
scale can lack in action driven restoration prescriptions. In addition to issues
surrounding scale, microsites are easily recognizable and offer a valuable unit of
observation for land managers, as they do not require specialized knowledge or
instruments to identify them. Mounds are prominent and abundant at Glacial
Heritage while Johnson Prairie and Training Area 15 have fewer mounds but
more dramatic slopes transitioning into upland sites.
While all three sites differed in study design, there were a few parameters
that stayed consistent across GH, JP, and TA15. In particular, the mounds at all
three sites were characterized by a bulk density lower than either the intermounds
at GH or the slopes, swales, and uplands at the JBLM sites. This lower bulk
density is often indicative of less compacted soil (Haveren 1983), and more
organic matter (USDA/NRCS – Soil Quality Indicators). The lowest average bulk
density was found at GH, which was not surprising since GH boasts the most
notable mounds of the three sites. The characterization of mounds across sites as
hosting deeper, finer soils than surrounding soil types conforms with other
descriptions of the South Sound prairie microsites and their soil makeup
(Dunwiddie and Martin 2016).
The ‘mima mounds’ found in all three sites are found in prairies outside of
the Puget-Trough – Willamette Valley prairie complex as well. ‘Mima mound’
structures have been described in the literature as occurring in San Diego county
(Cox 1984), Argentina (Cox and Roig 1986), Minnesota (Ross et al. 1968), Kenya
(Cox and Gakahu 1985), and Missouri (Horwath and Johnson 2006).
49

Interestingly the differences in soil moisture by microsite were not found
to be as strong as the bulk density signal. While microsites were not found to
harbor notable different moisture regimes throughout the spring, they illuminate
another aspect of the South Sound Prairies. Native prairie soils are often thought
to be relatively harsh due to low nutrients and low water holding potential
(Ugolini and Schlichte 1997). The fact that JP microsites in March had an average
moisture of 20.8%, which subsequently dropped to zero percent the next month is
a testament to how excessively drained these soils are. This trend was also found
at GH, and, to a lesser extent due to localized rain right before sampling, TA15.
While the data do demonstrate the harshness of prairie soils, they do not
show how an elongated period of moisture would improve germination of desired
natives. Field observations seemed to suggest that hot and dry conditions in late
May were a major component of germinant stress and death; it is possible more
moist conditions may improve performance. Soil moisture does have clear links to
seedling mortality in grassland systems (Morgan 1995).
While high seedling mortality may be attributed to May’s hot and dry
conditions, native and non-native plants may not respond to soil moisture in the
same way. For example, drought conditions in California grasslands have been
demonstrated to favor native perennial grasses over non-native annual grasses
(Hamilton et al. 1999). Soil moisture, like many other variables, is difficult to see
as a simple good or bad effect. When augmenting soil moisture through irrigation
or other means, restoration practitioners should take note of the level of plant

50

invasion; increasing soil moisture may have unintended consequences on native
vs. non-native competition.
Microsites can easily be differentiated by eye due to another factor aside
from topography, which is that of plant density. Glacial Heritage showed the
clearest difference here where plant growth was noticeably denser on the mounds
than the surrounding intermounded areas. The cause of the denser growth could
possibly be attributed to higher total nutrients in the mounded areas, as higher
nutrients are often associated with lower bulk densities. While this study did not
look at soil nutrients, given the differences found in bulk density between
microsites, analysis of soil nutrients in future studies could provide promising
insights into the conditions that characterize microsites and their associated plant
communities.
Higher aboveground plant density in mounded areas may also be
providing thermal refugia during the late spring and early summer when a lack of
forest cover leads to hot and dry conditions on the prairie. For Johnson Prairie and
Glacial Heritage, a higher plant density on the highland and mound microsites
resulted in greater germination of B. deltoidea. Towards the end of monitoring for
germination, most of the seedling die-off seemed to be related to heat stress as
many of the germinants that survived through the final sampling period had
denser growth surrounding them. Not all forbs responded in the same way; for
example, G. aristata did better in areas with less dense growth. For plants that are
more easily stressed out by late season heat, seeding into areas that have allowed

51

some growth after a burn would be preferred compared to seeding immediately
after a burn when there are few plants to provide a thermal refugia.
Recognizing the connection between differing forb phenologies and their
preferred habitat is an important component of the conservation of rare and
threatened species. The later emergence of G. aristata in the more exposed
microsites follows what would be expected; species that emerge in hotter
conditions are likely to prefer habitats that are more exposed to heat stress but
offer the benefit of reduced plant competition. Conversely the earlier emergence
of B. deltoidea conforms to a preference for microsites that provide more cover
and thermal refuge. While B. deltoidea avoids the environmental stress of a later
season emergence, it must contend with increased plant competition in the early
spring. Exploring the link between phenological traits of individual species and
their preferred microhabitats, such as was done in Galen and Stanton (1991), is an
important avenue for future research that focuses on the restoration of rare and
ecologically important species.
Vigorous growth of non-native plants is generally thought of as being
deleterious to establishment of desirable native species, however this study found
no effect of differences in native vs. non-native cover on the germination of the
study forbs. As far as germination goes it seems that both native and non-native
plants provide the same quality thermal refugia. The study sites are actively
managed with herbicide, so no site had an overwhelming presence of non-native
species. Past a certain threshold it would be expected that non-native growth
would suppress germination of natives (Fabbro et al. 2014, Mangla and Callaway
52

2008) but given the current makeup of the prairie there is no clear evidence that
the current level of invasion is strongly inhibiting the germination of desirable
native forbs, at least when considered on the microsite scale.
Understanding the habitat requirements of rare and hard-to-establish
grassland species is tricky business, as what is ideal habitat for one species may
be detrimental to another. In the context of the South Sound prairies, microsites
offer an easily identifiable and feasible way to manage for features that support
the re-establishment of stubborn species. Further, prairie microsites are
differentiated by multiple biotic and abiotic features, providing niche space for
which target species are selected. Understanding both which microsites yield
strongest growth and what mechanisms underpin that pattern are valuable insights
for researchers and managers alike.

53

Principle Conclusions


Microsites can be differentiated by several parameters, most noticeably
soil bulk density. While many parameters measured by this study were not
consistent across all three sites, mounds consistently had lower bulk
densities than the other microsites.



While the different microsites showed no detectable difference in moisture
regime, this study agrees with previous assessments of Puget-Trough
prairie soils being harsh and excessively drained.



Germination of both G. aristata and B. deltoidea showed a preference for
different microsite types across all three sites. B. deltoidea germinated at
higher rates in the mounds at Glacial Heritage and in the highland sites
(mounds + uplands), whereas G. aristata germinated at higher rates in the
inter-mounds of Glacial Heritage.



Density of aboveground growth differed strongly between mounds and
inter-mounds at Glacial Heritage. Although denser plots had slightly more
non-native species, it was not a higher presence of alien species per se that
influenced germination.



B. deltoidea and G. aristata should be kept in different seed mixes to be
sown in their preferred microsites/ soil type preference.



For plants that are more easily stressed out by late season heat, seeding
into areas that have allowed some growth after a burn would be preferred
compared to seeding immediately after a burn when there are few plants to
provide a thermal refugia.
54



Testing for microsite preference is a viable approach for hard-to-establish
native forbs. Defining too many microsites may dilute the effect a
microsite has on any given desirable outcome.

55

Thesis Works Cited
Applestein, C., Bakker, J. D., Delvin, E. G., & Hamman, S. T. (2018). Evaluating
Seeding Methods and Rates for Prairie Restoration. Natural Areas Journal, 38(5),
347–355.
Barnosky, C. W. (1985). Late Quaternary vegetation near Battle Ground Lake,
southern Puget Trough, Washington. Geological Society of America Bulletin
96:263–271.
Bond, W. J., & Parr, C. L. (2010). Beyond the forest edge: Ecology, diversity and
conservation of the grassy biomes. Biological Conservation, 143(10), 2395–2404.
Boyd, R. (1999). Indians, Fire, and the Land in the Pacific Northwest. Oregon State
University Press, Corvallis.
Bretz, J. H. (1913). Glaciation of the Puget Sound region. University of Chicago.
Cox, G. W. (1984). The Distribution and Origin of Mima Mound Grasslands in San
Diego County, California. Ecology, 65(5), 1397–1405.
Cox, G. W., & Roig, V. G. (1986). Argentinian Mima Mounds Occupied by
Ctenomyid Rodents. Journal of Mammalogy, 67(2), 428–432.
Dıá z, S., & Cabido, M. (2001). Vive la difference: plant functional diversity matters to
ecosystem processes. Trends in Ecology & Evolution, 16(11), 646–655.
Donath, T. W., Bissels, S., Hölzel, N., & Otte, A. (2007). Large scale application of
diaspore transfer with plant material in restoration practice – Impact of seed and
microsite limitation. Biological Conservation, 138(1), 224–234.

56

Dunwiddie, P. W., & Bakker, J. D. (2011). The Future of Restoration and Management
of Prairie-Oak Ecosystems in the Pacific Northwest. Northwest Science, 85(2),
83–92.
Dunwiddie, P. W., & Martin, R. A. (2016). Microsites Matter: Improving the Success
of Rare Species Reintroductions. PLoS ONE, 11(3), e0150417.
Endangered and Threatened Wildlife and Plants; Designation of Critical Habitat for
Taylor’s Checkerspot Butterfly and Streaked Horned Lark. (2013). Federal
Registrar.
Eriksson, O., & Ehrlén, J. (1992). Seed and microsite limitation of recruitment in plant
populations. Oecologia, 91(3), 360–364.
Fabbro, C., Güsewell, S., & Prati, D. (2014). Allelopathic effects of three plant
invaders on germination of native species: A field study. Biological Invasions,
16(5), 1035–1042.
Falk, D., Millar, C., & Olwell, M. (1996). Restoring Diversity: Strategies for
Reintroduction of Endangered Plants. Island Press, Washington DC.
Galen, C. and Stanton, M. L. (1991), Consequences of Emergence Phenology for
Reproductive Success in Ranunculus adoneus (Ranunculaceae). American
Journal of Botany, 78: 978-988.
Ghasemi, A., & Zahediasl, S. (2012). Normality Tests for Statistical Analysis: A Guide
for Non-Statisticians. International Journal of Endocrinology and Metabolism,
10(2), 486–489.

57

Guerrant Jr, E. O., & Kaye, T. N. (2007). Reintroduction of rare and endangered
plants: common factors, questions and approaches. Australian Journal of Botany,
55(3), 362.
Hamilton, J. G., Holzapfel, C., & Mahall, B. E. (1999). Coexistence and interference
between a native perennial grass and non-native annual grasses in California.
Oecologia, 121(4), 518–526.
Hegland, S. J., Nielsen, A., Lázaro, A., Bjerknes, A.-L., & Totland, Ø. (2009). How
does climate warming affect plant-pollinator interactions? Ecology Letters, 12(2),
184–195.
Horwath, J. L., & Johnson, D. L. (2006). Mima-type mounds in southwest Missouri:
Expressions of point-centered and locally thickened biomantles. Geomorphology,
77(3–4), 308–319.
Ingersoll, C. A., & Wilson, M. V. (1993). Buried propagule bank of a high subalpine
site: Microsite variation and comparisons with aboveground vegetation. Canadian
Journal of Botany, 71(5), 712–717.
Krock, S. (2016). Effects of sowing time and relative prairie quality on first year
establishment of 23 native prairie species. MES Thesis, The Evergreen State
College, Olympia, WA.
Lee-Mäder, E., Fowler, J., Vento, J., & Hopwood J. (2016). 100 Plants to Feed the
Bees. The Xerces Society, Storey Publishing, LLC, Portland, OR.

58

Linders, M. J., Hynson, L., Lewis, K., and Hamilton, L. (2015). Captive rearing and
translocation of Taylor’s checkerspot in south Puget Sound: 2014-2015. Annual
Progress Report to U.S. Fish and Wildlife Service and Joint Base Lewis-McChord
Army Compatible Use Buffer Program. Washington Department of Fish and
Wildlife. Olympia, Washington.
Macarthur, R., & Wilson, E. (1967). The Theory of Island Biogeography. Princeton
University Press.
Mangla, S., & Callaway, R. M. (2008). Exotic invasive plant accumulates native soil
pathogens which inhibit native plants. Journal of Ecology, 96(1), 58–67.
Memmott, J., Craze, P. G., Waser, N. M., & Price, M. V. (2007). Global warming and
the disruption of plant–pollinator interactions. Ecology Letters, 10(8), 710–717.
Morgan, J.W., (1995) Ecological Studies of the Endangered Rutidosis
leptorrhynchoides. II. Patterns of Seedling Emergence and Survival in a Native
Grassland. Australian Journal of Botany. 43, 13-24.
Porter, S. C., & Swanson, T. W. (1998). Radiocarbon age constraints on rates of
advance and retreat of the Puget Lobe of the Cordilleran ice sheet during the last
glaciation. Quaternary Research, 50(3), 205–213.
Rice, K. J. (1985). Responses of Erodium to Varying Microsites: The Role of
Germination Cueing. Ecology, 66(5), 1651–1657.
Rose, A.B., Frampton, C.M. (1999). Effects of microsite characteristics
on Hieracium seedling establishment in tall‐ and short‐tussock grasslands,
Marlborough, New Zealand, New Zealand Journal of Botany, 37:1, 107-118.
59

Ross, B. A., Tester, J. R., & Breckenridge, W. J. (1968). Ecology of Mima-Type
Mounds in Northwestern Minnesota. Ecology, 49(1), 172–177.
Ryser, P. (1993), Influences of neighboring plants on seedling establishment in
limestone grassland. Journal of Vegetation Science, 4: 195-202.
Schultz, C. B., & Dlugosch, K. M. (1999). Nectar and hostplant scarcity limit
populations of an endangered Oregon butterfly. Oecologia, 119(2), 231–238.
Schweiger, O. , Biesmeijer, J. C., Bommarco, R. , Hickler, T. , Hulme, P. E., Klotz, S. ,
Kühn, I. , Moora, M. , Nielsen, A. , Ohlemüller, R. , Petanidou, T. , Potts, S. G.,
Pyšek, P. , Stout, J. C., Sykes, M. T., Tscheulin, T. , Vilà, M. , Walther, G. ,
Westphal, C. , Winter, M. , Zobel, M. & Settele, J. (2010). Multiple stressors on
biotic interactions: how climate change and alien species interact to affect
pollination. Biological Reviews, 85: 777-795.
Severns, P. M. (2008). Seedling population size and microhabitat association in
Lupinus oreganus A. Heller var. Kincaidii C.P. Sm. (Fabaceae) a threatened plant
of western Oregon grasslands. Native Plants Journal, 9(3), 358–365.
Soons, M. B., & Heil, G. W. (2002). Reduced colonization capacity in fragmented
populations of wind-dispersed grassland forbs. Journal of Ecology, 90(6), 1033–
1043.
Stanley, A. G., Dunwiddie, P. W., & Kaye, T. N. (2011). Restoring Invaded Pacific
Northwest Prairies: Management Recommendations from a Region-Wide
Experiment. Northwest Science, 85(2), 233–247.

60

Stinson, D. W. (2005). Washington State Status Report for the Mazama Pocket
Gopher, Streaked Horned Lark, and Taylor’s Checkerspot. Washington
Department of Fish and Wildlife, Olympia. 129+ xii pp.
Storm, L., & Shebitz, D. (2006). Evaluating the Purpose, Extent, and Ecological
Restoration Applications of Indigenous Burning Practices in Southwestern
Washington. Ecological Restoration, 14(4), 256-268.
Tang, Y., Washitani, I., & Iwaki, H. (1992). Effects of microsite light availability on
the survival and growth of oak seedlings within a grassland. The Botanical
Magazine, 105(2), 281–288.
Tilman, D., & Downing, J. A. (1994). Biodiversity and stability in grasslands. Nature,
367(6461), 363–365.
Thill, D., Schirman, R., & Appleby, A. (1979). Influence of Soil Moisture,
Temperature, and Compaction on the Germination and Emergence of Downy
Brome (Bromus tectorum). Weed Science, 27(6), 625-630.
Thurston Regional Planning Council, WA. (2017). Population, Housing &
Employment Data Tables
Ugolini, F.C and K. Schlichte. (1973). The effect of Holocene environmental changes
on selected western Washington soils. Soil Science 116:218-227.
Wilson, M. V., Ingersoll, C.A., Wilson, M.G., and Clark, D. L. (2004). Why pest plant
control and native plant establishment failed: a restoration autopsy. Natural Areas
Journal. 24:23-31.

61

Wilson, E. O. (1992). The Diversity of Life. Bulletin of Science, Technology & Society,
14(1), 51–51.

62

Appendix 1: Figures

Figure 1 – boxplot of soil bulk densities (g/cm3). IM = intermound, M =
mound, SL = slope, SW = swale, UP = upland.

Figure 2 – boxplot of avg. % soil moisture at Glacial Heritage
broken down by date and microsite type.

63

Figure 3 - boxplot of avg. % soil moisture at JBLM broken down by date and microsite type.

Figure 4 – boxplot of plant density, broken down by microsite type and site. Plant
density represents the number of hits in a point-intercept grid, as described in methods.

64

Figure 5 – boxplot of microsite exotic cover broken down by microsite type and site.
Cover values of 1 indicate 100% cover while values of 0 indicate 0% cover.

Figure 6 – boxplot of microsite species richness broken down by microsite type and site.
Cover values of 1 indicate 100% cover while values of 0 indicate 0% cover.

65

Figure 7 - boxplot of native cover broken down by microsite type and site. Cover
values of 1 indicate 100% cover while values of 0 indicate 0% cover.

Figure 8 - boxplot of forb cover broken down by microsite type and site. Cover
values of 1 indicate 100% cover while values of 0 indicate 0% cover.

66

Figure 9 - boxplot of grass cover broken down by microsite type and site. Cover
values of 1 indicate 100% cover while values of 0 indicate 0% cover.

Figure 10 – avg. gaillardia germination % by microsite with standard deviation error bars.

67

Figure 11 – avg. GH balsamroot germination % by microsite with standard deviation error bars.

Figure 12 – avg. JBLM balsamroot germination % by microsite with standard deviation error bars.

68

Appendix 2: Tables
Table 1 – Descriptive statistics for soil moisture and bulk
density at different microsites at GH over time.

Site

Topography

Date

n

Moisture mean

Moisture BD
BD
sd
mean sd

GH

Intermound

March

12

35.58%

6.88

0.36

0.08

GH

Intermound

Early Apr

12

32.67%

5.82

0.36

0.08

GH

Intermound

Late Apr

12

30.16%

6.31

0.36

0.08

GH

Intermound

May

12

1.43%

1.65

0.36

0.08

GH

Mound

March

12

27.96%

5.62

0.25

0.07

GH

Mound

Early Apr

12

22.46%

4.84

0.25

0.07

GH

Mound

Late Apr

12

26.59%

5.71

0.25

0.07

GH

Mound

May

12

0.93%

1.03

0.25

0.07

Table 2 – Estimated marginal means for GLM post-hoc test for
differences in bulk density between microsites at JP.

EMM post-hoc for JP bulk density
Contrast

Estimate

SE

Z ratio

P-value

Mound - Slope

-0.09959

0.0332

-2.997

0.0145

Mound - Swale

-0.08626

0.0344

-2.507

0.0588

Mound - Upland

-0.08031

0.0344

-2.334

0.0903

Slope - Swale

0.01333

0.0266

0.5

0.959

Slope - Upland

0.01929

0.0266

0.724

0.8877

Swale - Upland

0.00595

0.0281

0.212

0.9966

69

Table 3 – Estimated marginal means for GLM post-hoc test for
differences in bulk density between microsites at TA15.

EMM post-hoc for TA15 bulk density
Contrast

Estimate

SE

Z ratio

P value

Mound - Slope

-0.0652

0.0337

-1.936

0.2129

Mound - Swale

-0.0395

0.0312

-1.263

0.5864

Mound - Upland

-0.0812

0.0345

-2.355

0.086

Slope - Swale

0.0257

0.0229

1.124

0.6747

Slope - Upland

-0.016

0.0271

-0.591

0.9347

Swale - Upland

-0.0418

0.0241

-1.734

0.3058

Table 4 – Estimated marginal means for GLM post-hoc test for
differences in moisture between microsites at GH.

EMM post-hoc for GH moisture
Contrast

Estimate SE

Z ratio p-value

Intermound, March - Mound, March

7.625

1.45

5.241

p < 0.0001

Intermound, Early April - Mound, Early April

10.208

1.45

7.017

p < 0.05

Intermound, Late April - Mound, Late April

3.567

1.45

2.452

p = 0.2166

Intermound, Early May - Mound, May

0.508

1.45

0.349

p = 1.000

70

TA15

TA15

JP

JP

JP

JP

GH

GH

Site

Upland

Swale

Slope

Upland

Swale

Slope

Mound

Mound

Intermound

4

6

4

4

4

4

2

6

6

17.75

17.67

16.00

15.50

22.25

19.50

12.50

14.00

12.50

3.86

3.08

0.82

3.87

2.50

3.70

0.71

4.20

2.66

27.25

36.17

24.50

35.75

24.00

32.75

35.50

35.17

22.83

4.43

10.23

4.36

8.46

2.16

1.26

9.19

7.41

4.62

0.45

0.60

0.66

0.56

0.23

0.49

0.00

0.47

0.55

0.05

0.20

0.35

0.36

0.09

0.11

0.00

0.29

0.17

0.55

0.40

0.34

0.44

0.77

0.51

1.00

0.53

0.45

0.05

0.20

0.35

0.36

0.09

0.11

0.00

0.29

0.17

M
SD
M
SD
M
SD
M
SD
Topography n richness richness density density native native exotic exotic

0.40

0.26

0.35

0.31

0.53

0.26

0.14

0.47

0.49

M
forb

0.22

0.14

0.21

0.18

0.28

0.13

0.10

0.20

0.28

SD
forb

0.54

0.70

0.63

0.69

0.47

0.74

0.86

0.39

0.43

M
grass

0.17

0.15

0.21

0.18

0.28

0.13

0.10

0.26

0.20

SD
grass

Table 5 – mean species richness, plant density, native cover,
exotic cover, forb cover, and grass cover with standard
deviation for microsites within each site.

TA15

71

Table 6 – GH count model, goodness of fit. Models were considered a good fit if the model residual
deviance was less than the 5% critical chi-squared model deviance value.

Table 7 – Balsamroot germination GLM by microsite at GH.

GH balsamroot germination
estimated marginal means post-hoc

Contrast

Date

Estimate SE

Z - ratio P - value

Mound - Intermound

March

-0.734

0.589

-1.245

0.213

Mound - Intermound

Early April

-0.582

0.553

-1.052

0.2928

Mound - Intermound

Late April

-0.168

0.505

-0.334

0.7387

Mound - Intermound

May

-1.253

0.575

-2.177

0.0295

Table 8 – Gaillardia germination GLM by microsite at GH.

GH gaillardia germination
estimated marginal means post-hoc

Contrast

Date

Estimate

SE

Z - ratio P - value

Mound - Intermound

March

na

na

na

na

Mound - Intermound

Early April

0.588

0.614

0.958

0.3382

Mound - Intermound

Late April

0.802

0.421

1.908

0.0564

Mound - Intermound

May

0.492

0.46

1.07

0.2848

72

Table 9 – Johnson Prairie balsamroot germination GLM.

JP germination by microsite
estimated marginal means post-hoc
contrast

date

estimate

Highland - Lowland

Early April

Highland - Lowland

SE

z.ratio

p-value

0.356674944 0.285556909

1.249050302

0.211647

Late April

0.508119262 0.273334413

1.858965566

0.063032

Highland - Lowland

March

0.307025035 0.33872677

0.906409126

0.364719

Highland - Lowland

May

0.349375641 0.293965395

1.188492413

0.234639

Table 10 – Training Area 15 balsamroot germination GLM.

TA15 germination by microsite
estimated marginal means post-hoc
contrast

date

estimate

SE

z ratio

p-value

Highland - Lowland

Early April

0.731066

0.391641

1.866673

0.061947

Highland - Lowland

Late April

0.81871

0.385741

2.122435

0.033801

Highland - Lowland

March

0.230524

0.372971

0.618075

0.536526

Highland - Lowland

May

0.868712

0.378839

2.29309

0.021843

Table 11 – GH balsamroot GLM on bulk density and moisture predictors.

GH BALDEL
negative binomial:
count ~ abiotic

Std.
Estimate Error

(Intercept)

1.929808 0.546924 3.528474 0.000418

bulk density

-4.30705

moisture

0.014547 0.010747 1.353653 0.175847

z value

1.705991 -2.52466

Pr(>|z|)

0.011581

73

Table 12 – GH gaillardia GLM on bulk
density and moisture predictors.

GH GAIARI negative binomial:
count ~ abiotic

Estimate

Std.
Error

(Intercept)

-1.17501

0.608889 -1.92976

bulk density

4.084012

1.785779 2.286964 0.022198

moisture

-0.01161

0.011451 -1.0143

z value

Pr(>|z|)
0.053637

0.31044

Table 13 – Biotic parameters influence on balsamroot
germination using negative binomial GLM.

Biotic GH BALDEL negative
binomial: count ~ biotic

Estimate

Std.
Error

z value

Pr(>|z|)

(Intercept)

-7.98144

3.180735

-2.50931

0.012097

Richness

-0.01417

0.061841

-0.22912

0.818779

Density

0.070466

0.027169

2.593647

0.009496

Exotic

0.041494

0.99594

0.041663

0.966767

Forb

7.169098

2.927689

2.448722

0.014336

Grass

7.672913

2.929281

2.619384

0.008809

74

Table 14 – Biotic parameters influence on gaillardia
germination using negative binomial GLM.

GH GAIARI negative
binomial: count ~ biotic

Std.
Estimate Error

z value

Pr(>|z|)

(Intercept)

-2.73164

3.702057

-0.73787

0.460593

Richness

0.110107 0.088473

1.244525

0.213306

Density

-0.05626

0.035829

-1.57022

0.116364

Exotic

-1.70108

1.396615

-1.218

0.223224

Forb

5.071007 2.935217

1.727643

0.084052

Grass

3.369795 3.261795

1.033111

0.301552

75

fin.
76

Media

Thesis_MES_2019_MessinaJ.pdf