The Evergreen State College Institutional Repository

Prairie Fire as a Selective Agent: Second-generation Responses and Plant Community Shifts

Item

Title (dcterms:title)

Eng Prairie Fire as a Selective Agent: Second-generation Responses and Plant Community Shifts

Date (dcterms:date)

2014

Creator (dcterms:creator)

Eng Mann, Jaal G

Subject (dcterms:subject)

Eng Environmental Studies

extracted text (extracttext:extracted_text)

!

!

!

PRAIRIE FIRE AS A SELECTIVE AGENT:
SECOND-GENERATION RESPONSES
AND PLANT COMMUNITY SHIFTS

by
Jaal G. Mann

A Thesis
Submitted in partial fulfillment
of the requirements for the degree
Master of Environmental Studies
The Evergreen State College
June 2014

!

!

!

!

!

© 2014 by Jaal Mann. All rights reserved.

!

!

!

!

!

This Thesis for the Master of Environmental Studies Degree
by
Jaal Mann

has been approved for
The Evergreen State College
by

__________________________
Carri J. LeRoy
Member of the Faculty

__________________________
Date

!

!

!

!

!

ABSTRACT
Prairie fire as a selective agent: Second-generation responses and plant community shifts
Jaal G. Mann
Prescribed fire has been used extensively as a prairie restoration tool, but its specific
impacts on prairie plant communities are not always interpreted within a spatially or
evolutionarily explicit framework. Prairie management is also influenced by a variety of
social factors, including neighboring landowners, Native American culture, and military
training. Here, we used experimental paired prescribed fire and fire-exclusion plots to
examine the effects of realistic large-scale prescribed fire on prairie communities in
western Washington State. As a case study of fire’s potential selective effects, we
collected seeds from a common prairie plant (Lomatium utriculatum) to examine secondgeneration effects of burning (in the previous year) on seed vigor. We also explore some
of the interactions between prairie restoration and the social issues surrounding it. In
addition to these social issues, this research also addresses education and involvement of
an underserved audience: prisoners. Inmates had active roles in the care, tracking, and
measurement of the Lomatium utriculatum seed germination and seedling growth. This
partnership was made possible through the Sustainability in Prisons Project, and provided
a valuable opportunity for science education and inmate contributions to ecological
restoration. Lomatium utriculatum is an essential nectar source for the federally
endangered Taylor’s checkerspot butterfly (Euphydryas editha taylori), and it may serve
as a surrogate for the effects of fire on endangered species recovery. Plant communities
varied between burn treatments, and richness was increased in burned treatments.
Lomatium utriculatum seeds collected from burned plots showed increased germination
following the exposure of parent plants to fire, exhibiting a second-generation effect of
burning the parent plant. Overall, this study demonstrates that prescribed fire may play an
important role in altering prairie composition and may be a potential selective force
driving plant micro-evolution in ecologically important species.

!

!

!

!

!

TABLE OF CONTENTS
LIST OF FIGURES ...................................................................................................................... V
ACKNOWLEDGEMENTS ......................................................................................................... VI
CHAPTER 1: LITERATURE REVIEW ..................................................................................... 1
PUGET LOWLAND PRAIRIES ................................................................................................... 1
PUGET LOWLAND PRAIRIE RESTORATION......................................................................... 4
USES OF PRESCRIBED FIRE FOR RESTORATION ................................................................ 7
FIRE AND PLANT COMMUNITIES......................................................................................... 10
TECHNIQUES FOR EVALUATING PRESCRIBED BURNING OUTCOMES ...................... 12
FIRE AND HETEROGENEITY ................................................................................................. 13
FIRE AND POLLINATION ........................................................................................................ 15
SEED MORPHOLOGY AND VIABILITY ................................................................................ 16
FIRE AND SEED DISPERSAL .................................................................................................. 18
FIRE AS A SELECTIVE AGENT ............................................................................................... 19
LOMATIUM UTRICULATUM IMPORTANCE ....................................................................... 22
CONCLUSION ............................................................................................................................ 24
CHAPTER 2: MANUSCRIPT FORMATTED FOR THE JOURNAL ECOLOGY ............ 26
ABSTRACT................................................................................................................................. 26
INTRODUCTION ....................................................................................................................... 26
METHODS .................................................................................................................................. 29
STUDY AREA .......................................................................................................................... 29
EXPERIMENTAL DESIGN ..................................................................................................... 30
STATISTICAL ANALYSIS ........................................................................................................ 33
RESULTS .................................................................................................................................... 35
DISCUSSION .............................................................................................................................. 41
CHAPTER 3: EXTENDED DISCUSSION AND FUTURE RESEARCH ............................. 45
EXPANSION OF CONCLUSIONS ............................................................................................ 45
FURTHER RESEARCH ............................................................................................................. 47
PRAIRIE RESPONSE TO A CHANGING CLIMATE ............................................................. 51
HOW THIS RESEARCH FITS INTO PRAIRIE MANAGEMENT ........................................... 52
LAND MANAGEMENT ........................................................................................................... 52
RESTORATION ECOLOGY RESEARCH ............................................................................... 53
CONCLUSION ............................................................................................................................ 55
CHAPTER 4: INTERDISCIPLINARY CONNECTIONS ...................................................... 56
AN ANTHROPOGENIC ECOSYSTEM .................................................................................... 56
SMOKE POLLUTION AND WILDFIRE RISK ......................................................................... 57
COSTS OF MAINTAINING NATIVE PRAIRIE ....................................................................... 58
WORK WITH THE SUSTAINABILITY IN PRISONS PROJECT ............................................ 59
CONCLUSION ............................................................................................................................ 61
REFERENCES ............................................................................................................................. 63

!

iv!
!

!

!

!

LIST OF FIGURES
FIGURE 1-1. THIS MAP SHOWS THE HISTORIC AND PRESENT EXTENT OF PUGET LOWLAND PRAIRIES. OUR
STUDY SITE LOCATION IS CIRCLED. ........................................................................................................... 2
FIGURE 1-2. FROM DUNWIDDIE AND BAKKER (2011). SHOWS EXAMPLES OF ECOLOGICAL AND CULTURAL
SITE MANAGEMENT GOALS AND THEIR RESPECTIVE MANAGEMENT INTENSITY AND ECOLOGICAL
IMPACTS. ................................................................................................................................................... 5
FIGURE 1-3. FROM NATIONAL INSTITUTE OF INVASIVE SPECIES SCIENCE (N.D.). A TYPICAL MODIFIEDWHITTAKER PLOT LAYOUT FOR MEASURING PLANT COMMUNITIES. BECAUSE OF THE THREE DIFFERENT
SIZES OF SUBPLOTS, MODIFIED-WHITTAKER PLOTS ARE WELL SUITED TO QUANTIFY ABUNDANCE AND
DIVERSITY OF BOTH RARE AND COMMON SPECIES. .................................................................................. 13
FIGURE 1-4. TAYLOR’S CHECKERSPOT BUTTERFLY (EUPHYDRYAS EDITHA TAYLORI) RESTING ON LOMATIUM
UTRICULATUM. PHOTO COURTESY OF ROD GILBERT, JOINT BASE LEWIS-MCCHORD ............................. 23
FIGURE 2-1. MAP SHOWING THE PAIRED PLOT LOCATIONS ON JOHNSON AND UPPER WEIR PRAIRIES. THE
THREE PLOTS USED FOR THE LOMATIUM UTRICULATUM SAMPLING ARE CIRCLED. THESE WERE THE ONLY
PLOTS THAT HAD BEEN BURNED IN BOTH 2011 AND 2012, AND THEREFORE WERE HYPOTHESIZED TO
SHOW MORE PRONOUNCED EFFECTS OF BURNING. ONLY A SMALL PATCH OF PRAIRIE WAS BURNED IN
2013, AND NONE OF THE PLOTS WERE WITHIN IT. .................................................................................... 31
FIGURE 2-2: TIME-TO-EVENT ANALYSIS OF SEED GERMINATION DATA BY DAYS SINCE SEED IMBIBING.
GERMINATION RATES VARIED BETWEEN BURN TREATMENTS, WITH SEEDS FROM BURNED PARENT
PLANTS GERMINATING EARLIER AND MORE SUCCESSFULLY (LOG-RANK P<0.0001). .............................. 35
FIGURE 2-3. SPECIES RICHNESS WAS NEARLY 19% HIGHER IN BURNED PLOTS (P=0.0004), INDICATING THAT
SPECIES HAD BEEN LOST FROM UNBURNED PLOTS AFTER ONLY 1-2 YEARS OF BURN EXCLUSION. DATA
2
ARE FROM THE 2012 AND 2013 1000 M PRESENCE/ABSENCE PLOTS COMBINED AND BARS REPRESENT
MEANS ± STANDARD ERRORS. ................................................................................................................. 36
TABLE 2-1. INDICATOR SPECIES ANALYSIS (DUFRENE AND LEGENDRE 1997) DETERMINED THAT SEVERAL
SPECIES OF PLANTS WERE INDICATIVE OF THE VARIOUS BURNED AND UNBURNED TREATMENTS. THESE
INCLUDED A NUMBER OF NATIVE AND NON-NATIVE SPECIES. ................................................................. 37
FIGURE 2-4. SPECIES-AREA CURVES FOR 2012 AND 2013, WITH 95% CONFIDENCE INTERVALS ILLUSTRATING
THE LACK OF SIGNIFICANT DIFFERENCES BETWEEN BURNED AND UNBURNED PLOTS IN SPECIES
ACCUMULATION. ..................................................................................................................................... 38
FIGURE 2-5. NMS ORDINATIONS FOR SPECIES COVER FROM OUR 1 M2 PLOTS SHOW WEAK BUT SIGNIFICANT
DIFFERENCES BETWEEN BURN TREATMENTS IN BOTH YEARS (2012: F=2.1672, P=0.0002; 2013:
F=3.383, P=0.0002). THE 2013 DATA APPEAR TO BE MORE CLUSTERED IN BOTH TREATMENTS,
POSSIBLY DUE TO VARYING CLIMATIC CONDITIONS BETWEEN YEARS. .................................................... 39
FIGURE 2-6. COMPARISON BY YEAR FOR EACH BURN TREATMENT, TO EXAMINE THE EFFECTS OF TEMPORAL
SUCCESSION WITHIN PLOTS IN TERMS OF PLANT COMMUNITY CHANGE. YEAR SIGNIFICANTLY AFFECTED
PLANT COMMUNITIES IN BOTH TREATMENTS (BURNED: A=0.0269, P<0.0001; UNBURNED: A=0.0218,
P<0.0001). YEAR EXPLAINED 28.3% OF THE COMMUNITY VARIATION IN BURNED PLOTS, AND 21.5% IN
UNBURNED. ............................................................................................................................................. 40

!

!v!

!

!

!

ACKNOWLEDGEMENTS
I would like to thank everyone from the Field Ecology Lab at The Evergreen State
College who helped with community data collection in the field and peer review of this
manuscript: Carri LeRoy, Dylan Fischer, William Buck, Levi Travis, Mariah Smith, Abir
Biswas, Justin Kirsch, and many others. This research would have been impossible
without the Fish and Wildlife team at Joint Base Lewis-McChord who oversaw the
burning of the plot areas for three years and coordinated access to the sites, including
John Richardson, Jim Lynch, Evan Hayduk, and Dennis Aubrey. I also thank Melinda
Wood for her help collecting and counting Lomatium utriculatum seed and reviewing
sections of this manuscript, and Carl Elliott for providing the knowledge I needed to be
able to germinate and grow L. utriculatum in a nursery setting. Shotwell’s Landing
nursery, where the L. utriculatum germination tests and seedling growth measurements
took place, is owned by the Center for Natural Lands Management, and I thank them for
helping to coordinate my use of the germination chamber, with a special thanks to Sierra
Smith. I would also like to thank the Sustainability in Prisons Project for providing the
resources I needed to be able to grow and monitor the L. utriculatum seedlings, and the
Cedar Creek Corrections Center inmate prairie restoration crew for their invaluable
assistance collecting germination data and monitoring seedlings. Dylan Fischer provided
essential help with deciphering the burn history of the plots and statistical analysis.
Finally, I thank my thesis reader Carri LeRoy for her indispensable support with study
design, statistical analysis, and countless improvements to this manuscript.

!

vi!
!

!

!

!

Chapter 1: Literature Review
Puget lowland prairies are greatly diminished from their past range, with native prairies
occupying only 3% of historical grassland area (Crawford and Hall 1997). I will describe
Puget lowland prairies, their associated restoration goals, common restoration practices,
and the effectiveness of these practices. Next I will examine prescribed fire as a
restoration tool and the ideal methods to implement it, with a focus on one of the largest
remaining prairie remnants at Joint Base Lewis-McChord (JBLM). Plant community
alterations and increased heterogeneity can be outcomes of some prescribed fires, and I
include a discussion of the importance of heterogeneity for prairie habitats and the use of
prescribed fire as a tool to increase native plant cover. Next I discuss fire as a selective
agent and various plant adaptations to fire-dependent landscapes. I describe how fire can
increase target species both directly through beneficial chemicals and heat, and indirectly
through increased nutrients, thatch removal, and enhanced pollination. Finally, I discuss
Lomatium utriculatum, a forb native to Puget lowland prairies, which serves as a case
study in this thesis for the effects of fire on native target species.
PUGET LOWLAND PRAIRIES
Glacial prairies in North America were formed by warming and drying after glacial
recession at the end of the last ice age (Pielou 1991). Gravelly deposition from melting
glaciers created a soil that did not hold water well, and a warm climatic period around
10,000 years ago dried out the landscape and turned the newly formed forests into
prairies free of woody plants (Pielou 1991). As the climate cooled slightly, trees began to
encroach on the prairies, particularly in the maritime Pacific Northwest. Native

!

!1!

!

!

!

Americans in these areas, who prized the prairies for their edible crops and valuable
hunting areas, began to implement prescribed burning to prevent tree encroachment and
improve yields of staple food crops (White 1975). As white settlers arrived, burning was
largely halted and most prairie landscapes were converted to agriculture and grazing.
Today only small remnants of the original Pacific Northwest glacial prairies remain, and
these small patches are managed intensively to keep them in as close to historical
conditions as possible. Puget lowland prairies are anthropogenic ecosystems, and as such
they require continual maintenance in order to preserve characteristics needed to support
endemic and endangered species.

Figure 1-1. This map shows the historic and present extent of Puget lowland prairies. Our study
site location is circled.

!

!2!

!

!

!

Puget lowland prairies often share a common soil type, and they are highly
suitable for various forms of land use. These prairies are largely located in the southern
Puget lowland area, with the majority of remaining prairies on Joint Base LewisMcChord (JBLM; Figure 1-1). JBLM’s prairie soils are in the Spanaway series (Tveten
and Fonda 1999), and are classified in Zulauf et al. (1979) as “somewhat excessively well
drained”—often shallow soils with a coarse texture and very little or deep free water
(Schoeneberger et al. 2002)—consisting of glacial outwash and volcanic ash. Mean
annual precipitation is 89-114 cm, mean air temperature is 10.6˚C, slope is 0-6% (Zulauf
and Washington Agricultural Experiment Station 1979), and urban development, gravel
mining, and grazing are suggested by Zulauf et al. (1979) as the primary human uses for
these soils.
Land use has been a major force in the destruction of native prairies, but
ironically, military use of prairies on JBLM has led to the preservation of the largest
remaining fragments. European settlers commonly used the flat, open landscapes as
agricultural lands and later as sites for development. While some prairie species can
coexist with grazing animals, tilling and fertilizing for agriculture or building over prairie
removed large tracts of native species. On JBLM, home to the majority of remnant Puget
lowland prairies, 40% of grassland areas remain “native prairie sites” (Crawford and Hall
1997). This large proportion of native prairie combined with frequent fires from military
training exercises make JBLM a local hotspot for prairie restoration management and
research. Federally endangered Euphydryas editha taylori W. H. Edwards (Taylor’s
checkerspot butterfly) are present in some areas along with other endangered or
threatened prairie-obligate birds and mammals, including Thomomys mazama (Mazama

!

!3!

!

!

!

pocket gopher), Eremophila alpestris strigata (streaked horned lark), and Sciurus griseus
(western gray squirrel). While JBLM’s prairies are some of the most intact, ongoing
management and restoration are necessary to keep them open and with a large proportion
of native cover.
PUGET LOWLAND PRAIRIE RESTORATION
Prairie restoration is a field fraught with moral dilemmas, and sometimes it can seem
impossible to keep up with the challenges of invasive species and land use change. While
in some cases restoration ecologists are able to facilitate functional assemblages
resembling pre-European-settlement prairies, it is impossible to truly return prairies to
their historical forms and completely eradicate the multitudes of established invasive
species (Seastedt 1995, Agee 1996). In addition, recent research in restoration ecology
argues that communities should not be restored to historic conditions but instead to
communities that can withstand future climatic conditions (Millar et al. 2007, Thorpe
2012). Immense quantities of resources are expended to restore native prairies, but how
do practitioners decide how much is enough in a field where complete success may be
unattainable?
A variety of ecological and cultural goals must be considered when determining
management plans. Dunwiddie and Bakker (2011) ranked several ecological and cultural
prairie restoration goals, respectively, by most to least conservative and low to high
ecological impact (Figure 1-2). This ranking serves to illustrate the broad and sometimes
divergent interests of stakeholders in prairie management, as a way of finding areas of
compromise that can satisfy as many parties as possible. Invasive species can be
suppressed in many cases, but are rarely eradicated, and often warrant continued removal

!

!4!

!

!

!

efforts (Dennehy et al. 2011). In this section, I will discuss some of the major challenges
to Puget lowland prairie restoration, and how they are addressed. This will be followed
by an examination of management techniques and their effectiveness at achieving various
ecological goals.

Figure 1-2. From Dunwiddie and Bakker (2011). Shows examples of ecological and cultural site
management goals and their respective management intensity and ecological impacts.

Restoration of Puget lowland prairie faces the continuing challenges of habitat
destruction, social reluctance to preserve prairies, and invasive species. Determining
achievable goals, functional assemblages, and necessary maintenance are essential to the
preservation of native prairie communities (Agee 1996, Dunwiddie and Bakker 2011,

!

!5!

!

!

!

Dennehy et al. 2011). Tools that are commonly used by land managers include
herbicides, mowing, carbon additions, and prescribed fire (Dennehy et al. 2011, Hamman
et al. 2011, Mitchell and Bakker 2011). Prescribed fire has numerous benefits for prairie
communities but can sometimes be met with social resistance due to health concerns such
as smoke inhalation (Bowman and Johnston 2005; see Chapter 4 for a complete
discussion of this issue). While social issues can usually be overcome with careful
planning and timing, in some cases they can conflict with ideal prescribed burn timing for
ecological restoration, which is still largely unknown. Invasive species are the threat that
most land managers are directly concerned with, and many tactics are employed with the
goal of reducing their populations.
Invasive species can alter overall prairie community types, converting them to
shrubland or forest in the absence of fire (Collins and Gibson 1990). Invasive species
include non-native Cytisus scoparius L. Link (Scotch broom) and native Pseudotsuga
menziesii (Mirb.) Franco (Douglas-fir). Cytisus scoparius is a nitrogen-fixing shrub that
may also inhibit growth of some nearby species through soil alterations (Haubensak and
Parker 2004). In the absence of fire, P. menziesii eventually colonizes prairies and
converts them to forest (Hamman et al. 2011). There are some researchers, however, who
argue that once a non-native species is well-established, impact assessments should be
done to see if it is still a threat or if it can be a non-disruptive or functionally desirable
species (Shackelford et al. 2013). Nevertheless, where invasive species do constitute a
major threat to the overall composition of the ecosystem, a variety of restoration methods
have been employed.

!

!6!

!

!

!

Herbicides, mowing, and manual removal are some of the main prairie restoration
tools used in Puget lowland prairies, and carbon addition has potential for more
widespread use. Herbicides are commonly used on prairies in the Pacific Northwest to
remove target invasive species. While they can be highly effective, these herbicides can
be toxic and caution must be exercised in choosing an herbicide to make sure to avoid
excessive toxicity to people, non-target species, and wildlife (Dennehy et al. 2011).
Mowing, pulling, and cutting of invasive vegetation can be successful if used at the right
time of year and repetitively (Dennehy et al. 2011). Manual removal methods can be
labor-intensive and repeated control is usually necessary. Carbon can be added to the soil
in the form of activated carbon or sugar, in order to adsorb excess available nitrogen or
increase microbial activity to use up soil nitrogen, respectively (Mitchell and Bakker
2011). Excess soil nitrogen can increase productivity, but favors invasive species, driving
some rare native species towards extinction (Suding et al. 2005, Clark et al. 2007).
Mitchell and Bakker (2011) found that by adding sugar to the soil they were able to
reduce the growth of several invasive forbs, potentially giving native species a boost. To
reach the various goals of native species enhancement, invasive biomass removal, and the
creation of beneficial disturbances, prescribed fires can also be used in prairie restoration.
USES OF PRESCRIBED FIRE FOR RESTORATION
Joint Base Lewis-McChord (JBLM) is a 36,684 ha military reservation (JBLM 2012) that
is home to the majority of remaining native Puget lowland prairie. Along with other
management practices, they began implementing regular prescribed fires on 3,000 ha of
native prairie and oak woodlands in 1978, using 3-5 year burn intervals (Tveten and
Fonda 1999). Military operations have also been essential to maintaining native cover,

!

!7!

!

!

!

and the largest native prairie remnant on the base is located in the artillery impact area,
which has burned nearly every year for 50 years because of artillery fire (Tveten and
Fonda 1999).
The optimal season and frequency of prescribed burns have been objects of much
study (Pollak and Kan 1998, Tveten and Fonda 1999), but methods to address the
questions have been varied. Experimental methods employed for this purpose include
intensive burning of small plots (Tveten and Fonda 1999) and placement of plots within
larger prescribed burns (Pollak and Kan 1998). I argue that placement of paired plots
within larger prescribed burns is more effective at evaluating real-world prescribed fire
effects because of the heterogeneity inherent in low-intensity grassland fires. Larger
burns may also increase flowering plant growth over a large enough area to attract more
insect pollinators (Van Nuland et al. 2013), an effect that would likely not be seen in
small-area burn treatments. Questions persist about how to use prescribed fire most
efficiently, but in most cases it is still superior to alternative restoration methods.
Fire has major benefits over herbicides and manual weed removal in both
effectiveness and cost. It has a long history of use to increase the prevalence of desirable
species and decrease invasive species abundance (Aldous 1934), and it can remove larger
woody vegetation to help maintain the open structure required by many native prairie
species (Hamman et al. 2011). Fire has historically been a common occurrence on Puget
lowland prairies and therefore the growth of many native species has been enhanced by it
(one of the major reasons for Native American burning; Boyd 1999). Fire has a relatively
low cost per acre that decreases with increasing treatment scale (Bidwell et al. 2002,
Hesseln and Berry 2008). Prescribed fires also provide a myriad of indirect benefits

!

!8!

!

!

!

including elimination of excess nitrogen (Seastedt 1995) and improving germination in
some species (Boyd 1999). While prescribed fire is an affordable tool with a number of
benefits, the various effects of fires that occur at different return intervals or over
different seasons are not well understood.
The ideal season for burning can vary depending on specific restoration goals, and
Native American burning practices may be an effective guide for some prescribed fires
today. However, changes in past management, habitat fragmentation, and invasive
species may have altered the effects of fire, and understanding the influences of fire on
modern native/non-native mixed plant communities is vital to restoration (Hamman et al.
2011, Sprenger and Dunwiddie 2011). Pollak and Kan (1998), studying a prairie in
California, mention that fall prescribed burns can kill seedlings of both native and nonnative plants indiscriminately, but late-spring burning does much less damage to plants
that have already dispersed their seeds—as most of the native plants have. In contrast, in
Washington State, Tveten and Fonda (1999) found that fall fires were more effective than
spring fires at promoting native communities. In tallgrass prairies of the Midwest, Howe
(1995) found that summer burns increased annual and biennial species, while many other
species were affected in different ways on an individual basis by fire season. Tveten and
Fonda (1999) also found that both annual burns and burning with intervals greater than 5
years increased invasive species populations, such as Hypochaeris radicata (hairy catsear). Sprenger and Dunwiddie (2011) used dendrochronology on P. menziesii trees on
Waldron Island, WA, to construct a fire history for the area, concluding that the majority
of fires were started by Native Americans who burned preferentially in the fall (Sprenger
and Dunwiddie 2011). Some of these results can be conflicting, and it is clear that more

!

!9!

!

!

!

research is needed locally on Puget lowland prairies with modern native/non-native
mixed plant assemblages to determine ideal burn regimes.
Fire can have important effects on prairie soils by altering nitrogen content and
changing plant behavior. When fire eliminates excess soil nitrogen, it simultaneously
increases evaporation and plant growth (Seastedt 1995). Faced with increased growth
concurrent with decreased water and nitrogen availability, plants are forced to allocate
more resources to further developing their root systems, an adaptation which is important
for building up soil organic matter (Seastedt 1995). The invasive species that often
dominate in nitrogen-enriched soils tend to have smaller root systems than native species,
thereby depositing less organic matter in the soil than the native plant communities and
altering the composition of prairie soils (Seastedt 1995). Over time this alteration of soil
organic matter content could lead to a soil environment for which native plant species are
poorly adapted.
FIRE AND PLANT COMMUNITIES
Fire may play a formational role in prairie plant communities (Collins and Gibson 1990),
but its role can be altered by soil fertility (Knops 2006). Knops (2006) found that in
infertile prairies with low productivity fire had a minimal effect on plant community
composition. Fire in many grassland ecosystems, however, has been found to alter
species richness, diversity, and community interactions (Collins and Gibson 1990).
Ewing and Engle (1988) argue that fire can cause persistent alterations to native
tallgrass prairie communities, and that fire intensity is important for predicting
community change. A large part of the observed community change was due to the
decrease in bunchgrass cover in high-intensity burned areas (Ewing and Engle 1988).

!

10!
!

!

!

!

Regardless, Puget lowland prairies may interact differently with fire intensity than
tallgrass prairies. Additionally, Howe (1995) argued that fire intensity may not be highly
relevant to herbaceous plant cover because most intensities of fire manage to kill the
majority of herbaceous aboveground vegetation. Because of the different ways that
communities in specific regions can respond to fire intensity, local studies may be needed
for reliable information. Other research topics such as community stability, however, may
be more easily generalized.
Prairie communities can be inherently unstable, with directional change over time
(gradual but continuous community change) recorded for tallgrass prairie communities
(Collins 2000). Over 10 years of data from the Konza Prairie Long-Term Ecological
Research site in Kansas show that native tallgrass prairie communities are not stable and
are progressing directionally instead. Prairies that were burned annually were progressing
more rapidly than occasionally burned prairies, and their plant communities varied
significantly. A 4-year fire frequency was hypothesized to be the most stable as it most
closely mimicked historical burn regimes, but instead the 20-year fire return interval
produced a more stable plant community. Consumer communities (birds, grasshoppers,
and small mammals) were also more stable under the 20-year fire return interval because
although the species composition was changing, the three-dimensional vegetation
structure remained fairly constant (Collins 2000).
Invasive species can change how fire interacts with plant communities,
complicating restoration efforts. McGranahan et al. (2012) found that a high-moisture
invasive grass required much greater wind speed to burn—often beyond safe levels—and
that this fire resistance could lead to large unburned areas that increase the risk of tree

!

11!
!

!

!

!

encroachment. Prescribed fires with the goal of reducing invasive species populations
ideally should occur after native species have released their seeds to the soil—where
there is less heat from the fire—and while invasive species still have theirs in the
understory canopy where they are more susceptible to heat (Ditomaso et al. 2006). This
technique means that invasive late-season annual forbs are much easier to control with
fire than winter annuals (Ditomaso et al. 2006). Unless burn timing is targeted for a
certain species, fire can increase populations of invasive perennial forbs, often requiring
other management techniques for their removal (Ditomaso et al. 2006). Ditomaso et al.
(2006) advocate integrated approaches to managing recalcitrant invasive species with
fire, using other control methods in conjunction with burning in order to avoid having to
create unnaturally short fire intervals—which could in turn promote other invasive
species.
TECHNIQUES FOR EVALUATING PRESCRIBED BURNING OUTCOMES
Previous studies of fire’s effects on herbaceous species at JBLM have used small plots (1
m2) as their experimental units (Tveten and Fonda 1999). This could put them at a
disadvantage because the small plots can be completely burned, and any larger patterns of
more natural and heterogeneous burning are lost. Designing studies with large burn
exclusion plot pairs within a matrix of heterogeneous burning is a preferable method to
capture the variability of large-scale prairie fires (Risser 1990). In addition, larger plot
sizes, like Modified-Whittaker plots (Figure 1-3) may be more suitable for identifying
rare species, analyzing diversity patterns, and identifying trends over time in long-term
plots (Stohlgren et al. 1995). Pollak and Kan (1998) put their plot pairs within a larger
burned area using transects within their plots and ¼ m2 frames to estimate cover classes

!

12!
!

!

!

!

of species guilds. This design places plots within large and heterogeneous burns, but their
method of collecting cover data within the plots may not provide the higher resolution
results that would be possible with a Modified-Whittaker sampling method and
identification of all species within the plots.

Figure 1-3. From National Institute of Invasive Species Science (n.d.). A typical ModifiedWhittaker plot layout for measuring plant communities. Because of the three different sizes of
subplots, Modified-Whittaker plots are well suited to quantify abundance and diversity of both
rare and common species.

FIRE AND HETEROGENEITY
Using grazing by cattle and a variety of burn treatments Fuhlendorf et al. (2006)
attempted to determine the importance of prairie heterogeneity for conservation. In order
to accomplish this they used paired patches, burning one in spring and the other in late
summer or fall. They also allowed cattle to graze freely in order to assess their preference
of fire treatment. To compare this patchy burn structure to more traditionally managed

!

13!
!

!

!

!

pastures they also included samples in pastures that were all burned at once. Point counts
were used in all of the pastures in order to determine bird abundance by species, and
plant cover was recorded by functional group in each patch (Fuhlendorf et al. 2006).
Tallgrass and bare ground heterogeneity were increased up to 40 times in the patchy
treatments compared to traditional management (Fuhlendorf et al. 2006). They also found
that annual fires in traditional treatments removed nearly all of the litter, while in the
patchy experimental sections only portions were removed (Fuhlendorf et al. 2006). This
uneven litter buildup could lead to even more heterogeneity and diversity in the future.
Once they had determined optimal management techniques to foster heterogeneity,
Fuhlendorf et al. (2006) sought to examine how heterogeneity can help meet conservation
objectives.
In order to determine the value of increased heterogeneity Fuhlendorf et al. (2006)
recorded bird abundance in their different treatments and found that bird species varied
more in the patchy treatment. The buildup of litter in sections of the patchy treatment had
a strong effect on bird communities, with some species avoiding areas with low litter
cover (Fuhlendorf et al. 2006). Grazing animals were more likely to choose burned areas
to graze, and grazed areas were less likely to burn. This feedback loop would make
unburned areas increasingly likely to burn due to lower grazing, and burned areas less
likely to burn due to higher grazing, leading to less uniform disturbance overall
(Fuhlendorf et al. 2006).
Collins (1989) observed the effect of fire on plant community heterogeneity
within micro-disturbance areas and found that in areas with disturbances fire had no
effect on heterogeneity; however, in undisturbed areas fire decreased heterogeneity.

!

14!
!

!

!

!

Collins and Gibson (1990) also found that annual burning decreased heterogeneity. In
general, however, with longer fire return intervals the patchiness of burns due to variable
fuels, moisture levels, and wind can cause an increase in the heterogeneity of plant
communities (Risser 1990). Heterogeneity is especially important for prairie-associated
organisms—including pollinators—which depend on diverse plant assemblages for nectar
and larval hosts and could survive during fires in small unburned patches. Heterogeneity
can also extend the flowering season and attract more diverse pollinator species, and that
diversity could be vital if certain pollinators decline (Wratten et al. 2012).
FIRE AND POLLINATION
Insect pollination is an essential ecosystem service and most vegetables, nuts, fruits, and
edible oil crops depend on it (Gallai et al. 2009). In Puget lowland prairies, habitat
fragmentation can lead to pollinator-limitation in some species (Fazzino et al. 2011), but
this effect may be ameliorated by fire’s enhancement of pollination (Van Nuland et al.
2013).
A majority of the world’s crops depend on insect pollination, providing a service
valued at around $US2005 200 billion (Gallai et al. 2009). With recent declines of Apis
mellifera (honey bee) populations, native pollinators may become more important for
food pollination (Campbell et al. 2007). Monitoring pollinator decline is difficult due to
large annual variation in population sizes, and large-scale sampling efforts are required
(Lebuhn et al. 2013). While determining precise estimates of pollinator decline may be
cost prohibitive on Puget lowland prairies, we can hypothesize that as the prairies become
increasingly fragmented, pollination may be decreased (Fazzino et al. 2011). Fazzino et
al. (2011) found that Balsamorhiza deltoidea Nutt. (Puget balsamroot) was pollinator-

!

15!
!

!

!

!

limited on three prairie sites at JBLM (two of which were used in this study), but a
follow-up study by Husby (2012) did not find pollinator limitation for B. deltoidea or
Lupinus albicaulis (in both sites at JBLM used in this study as well as in a native plant
nursery setting). This indicates that while pollinator limitation may be a concern in some
situations, in different sites or years it may be less of a problem. Fire can lead to better
pollination, largely because many plants increase their density after fire, creating larger
nectar resources for pollinating insects (Van Nuland et al. 2013). On pollinator-limited
prairies, using prescribed fire to create higher-density flower clumps that are capable of
attracting more pollinators could be a viable management technique. Ground nesting bee
communities can be increased more directly by fire, since they depend on bare ground
(Potts et al. 2005, Campbell et al. 2007)—which tends to increase post-fire. Changes in
pollen availability due to fire may even affect seed size as plants can allocate their
resources differently based on pollination (Sakai 1995, Sakai and Sakai 1995, 1996).
SEED MORPHOLOGY AND VIABILITY
Changes in seed morphology following fire can be important for a variety of reasons.
Seed size can have an effect on germination success, and seed size can vary in response
to individual seed genetics, parent plant genetics, and environmental conditions
(Tavşanoğlu and Serter Çatav 2012). Tavşanoğlu and Serter Çatav (2012) examined a
Mediterranean species and found that larger seeds germinated at higher rates when
exposed to heat stratification but saw no pattern in those not exposed to heat. This means
that in some species seed size interacts with fire to increase germination. Larger seeds
can emerge from deeper under the soil surface, allowing them more protection from
excessive heat during high intensity fires (Bond et al. 1999), and this could potentially

!

16!
!

!

!

!

select for plants that produce larger seeds in areas with frequent, high-intensity fires. This
relationship is somewhat complicated, however, by Hanley et al.'s (2003) finding that
small seeds of some species may withstand higher temperatures than larger seeds.
Therefore, while smaller seeds can only emerge from closer to the surface—where heat is
more intense during a fire—they can tolerate higher heat than the deeper and larger seeds,
meaning that fire intensity could play a crucial role in determining optimal seed sizes for
survival.
Seed mass can vary greatly within species and even within individual plants
(Thompson 1984). Moles et al. (2005) examined seed mass data from nearly thirteen
thousand plant species and determined that differences in growth form—between
herbs/grasses, shrubs, and trees/vines—explained more than three times as much
variation in seed mass as any other variable. This supports the idea that while seed size
can vary due to other factors, the most important drivers of seed size are based on the
plant’s growth form (Moles et al. 2005). Thompson and Pellmyr (1989) studied
Lomatium salmoniflorum (salmonflower biscuitroot) and found that the majority of
variation in seed mass (46%) was among umbels within plants in contrast to only 19%
among plants. Thompson (1984) grew Lomatium grayi (Gray’s biscuitroot) seedlings
under controlled conditions and found that the heaviest seed on each plant weighed 4.7
times as much as the lightest seed. From his results, Thompson (1984) was able to assert
that weighing large numbers of seeds and averaging to obtain seed mass does not
adequately display the variance among seeds of the same individual plant (where 59.5%
of seed mass variation occurs). Intra-plant seed mass variation in Lomatium may be too
high for fire to have a discernible effect on it.

!

17!
!

!

!

!

Sakai and Sakai (1995) observed that when plants of the same species growing in
the same environment experienced higher resource availability they sometimes produced
seeds of heavier weights rather than increased numbers, terming this the “fertilization
efficiency hypothesis” (Sakai and Sakai 1996). Sakai (1995) refined this hypothesis to
note that as resource availability increases plants allocate more resources to producing
larger numbers of flowers (and hence seeds), as long as pollination increases in a linear
relationship with the number of flowers. If pollen availability is low, plants will produce
a smaller number of larger seeds because there is not enough pollen for additional seeds.
While pollen availability did appear to influence seed size in Lilium auratum (gold band
lily), Sakai (1995) noted that competition between flowers for resources also plays a role
in determining seed size. In a wind-dispersed species like Lomatium utriculatum, seed
size could be important for dispersal—in particular the weight-to-area ratio, or wingloading (Augspurger and Franson 1993)—where lighter seeds and/or those with larger
wing-to-mass ratios can be transported farther.
FIRE AND SEED DISPERSAL
Seed dispersal can play a role in how well a plant is able to colonize newly burned areas,
therefore the distance that a wind-dispersed seed can travel may be an adaptation to fire.
Lomatium utriculatum diaspores have winged achenes (or samara) that allow wind to
transport the seeds some distance from the parent plant. The diaspore wing-loading
determines in part the dispersal characteristics of a seed, where increased wing-loading
decreases how far the seeds can travel (Augspurger and Franson 1993). Wing-loading
may decrease (increasing dispersal distance) in plants that are exposed to more fire, just
as Pinus contorta (lodgepole pine) wing-loading has decreased as it migrates north

!

18!
!

!

!

!

(Pielou 1991). This occurs because the trees with seeds that can travel farthest disperse
more quickly and tend to also have seeds that can travel far, eventually leading to seed
dispersal distances increasing towards the extremes of the plants’ range (Pielou 1991).
Fire may also be able to aid in the dispersal of L. utriculatum seeds as large-scale
fires can cause powerful updrafts. This could lead to widespread long-distance seed
dispersal increases during years with fire (Murray 1987). In Taraxacum officinale
(common dandelion), Tackenberg et al. (2003) found that updraft is far more important
than horizontal wind-speed at determining dispersal distance. The potential increase in
the dispersal distance of L. utriculatum seeds when fire is present could be an important
driver of the distribution and evolution of this species.
FIRE AS A SELECTIVE AGENT
Fire may be a driver of evolution in some systems, playing a fundamental role in the
formation of grassland plant communities (Vogl 1974) and even causing plants to evolve
flammability in order to eliminate their neighbors (Bond and Midgley 1995). Hoffmann
and Franco (2003) found that savanna plants tend to allocate more resources to growing
large roots, which may be an adaptation to fire to be able to resprout quickly (Bond and
Keeley 2005). Even where woody plants do resprout they can lose years of growth and
their competitive advantage over herbaceous species, whose aboveground biomass often
dies back during winter and quickly resprouts in spring (Anderson 2006). Due to this
seasonal adaptation to resprout quickly, after a disturbance herbaceous species have an
advantage over woody species that take years to re-grow their more durable tissue and
may not be able to re-establish under the canopy of their fast-growing herbaceous
competitors.

!

19!
!

!

!

!

Fire can act indirectly as a selective agent through alteration of the local
environment following burning. “Postfire environments” can have altered soil conditions,
be more vulnerable to drought, and have reduced surface moisture availability—all of
which can select for species that are tolerant of these conditions (Keeley and
Fotheringham 2000). Andrusyszyn (2013) measured soil surface and sub-surface
temperatures in winter on two prairies at JBLM and found that areas burned four years
prior were significantly cooler than those burned within the subsequent three years. In
contrast, burns did not seem to affect soil moisture, possibly due to topographical
variation (Andrusyszyn 2013). Since many native species (including L. utriculatum)
germinate in the cold temperatures of early spring, higher soil temperatures may be
important for seedling survival. While some plants are fire-adapted simply due to their
tolerance of postfire conditions, others directly rely on fire for their survival.
Serotiny, heat scarification, and chemicals in smoke can be essential for many
plants that rely on the direct effects of fire. Trees with serotinous seeds retain the seeds
until a fire spreads through the area, when the heat releases the seeds and they can
establish on bare ground with less competition. Scarification weakens the seed coat,
speeding germination in seeds with longer dormancy periods. In some fire-adapted
species, their seeds will lay dormant in the soil until a fire passes over, cracking open the
seed coat and allowing the plant to germinate in a landscape free of larger competing
plants. Chemicals in smoke can enhance or inhibit germination, and they can be distilled
into “smoke water” used for the propagation of prairie plants (Yi-Fang Chou et al. 2012).
Smoke water, derived of compounds from plant-based smoke infused in water, can
enhance the germination of some species of prairie plants and inhibit others (Jefferson et

!

20!
!

!

!

!

al. 2008, Yi-Fang Chou et al. 2012). One possible reason for this enhancement effect
could be plant growth regulators (or karrikins; Chiwocha et al. 2009) from smoke being
stored in the parent plant’s tissue. Lomatium utriculatum does not exhibit serotiny or heat
scarification and has not been tested for smoke water response (likely because it has not
proven refractory to standard cultivation without the addition of smoke water), but may
show as yet undefined adaptations to fire.
Both parent plant genetics and “environmentally induced parental effects” can
alter germination and seedling vigor (Schmid and Dolt 1994, Lacey 1998). Lacey (1998)
notes that stimulus to the parent plant leading to an effect on offspring (secondgeneration effect) can be brief and occur long before the affected seeds are produced. In
the case of fire this means that the parent plant could be burned years before and still pass
the effects of this burning to its offspring. Lacey (1998) defines parental effects in order
to clarify the difference between these and genetic or direct environmental effects:
A parental effect is any parental influence on offspring phenotype that cannot be
attributed solely to offspring genotype, to the direct action of the nonparental
components of the offspring's environment, or to their combination.
In an area where a parent plant was burned, effects on its seeds, if germinated ex situ,
could certainly fall under this definition of a parental effect.
Donohue (2009) takes maternal effects beyond just germination and discusses
how they can also affect “overall phenology, projected population performance and even
the genetic basis of life stage transitions.” Flowering phenology, a maternal effect, can
vastly alter the life cycle of some species, where seeds of the same plant germinating in
fall exhibit a winter annual life history while those germinating in spring follow a
biennial life history (Donohue 2009). Donohue (2009) specifies what is considered a

!

21!
!

!

!

!

maternal effect, including effects based on direct genetic inheritance, endosperm, seed
coat, maternal nutrients and other provisions, and release of seeds into a particular
environment (i.e. in a particular season or following an event such as fire). It can be a
challenge to determine whether observed parental effects are maternal or based on
endosperm (two-thirds from the mother plant) or embryo (one-half from the mother plant;
Donohue 2009). Building a stronger understanding of parental effects on germination and
phenology is important to understand plant responses to both natural and artificial
selection (Donohue 2009).
LOMATIUM UTRICULATUM IMPORTANCE
Lomatium utriculatum (Nutt. ex Torr. & A. Gray) J.M. Coult. & Rose is found west of
the Cascade Mountains from British Columbia to California. It is a perennial forb on
Puget lowland prairies and a primary nectar source for two endangered butterflies:
Euphydryas editha taylori W. H. Edwards (Taylor’s checkerspot butterfly) in the Puget
lowland area (Shepard 2000), and Callophrys mossii bayensis R. M. Brown (San Bruno
elfin butterfly) in California (Arnold 1983). Puget lowland prairies are home to several
reintroduced populations of E. e. taylori, and L. utriculatum has been planted in
reintroduction sites for E. e. taylori at JBLM (Figure 1-4).

!

22!
!

!

!

!

Figure 1-4. Taylor’s checkerspot butterfly (Euphydryas editha taylori) resting on Lomatium
utriculatum. Photo courtesy of Rod Gilbert, Joint Base Lewis-McChord

Lomatium utriculatum serves here as a case study to examine changes in seed
size, viability, and germination following prescribed fire. Its seeds have not been shown
to be adapted to direct exposure to fire, but the parent plants are perennial and can
survive regular burns, and tolerance or adaptations to the post-fire environment may be
passed along to their offspring. In this way, Lomatium utriculatum is representative of
other fire-adapted species with seeds that do not depend on direct exposure to fire.
Lomatium utriculatum has a thickened taproot, compound umbels with many yellow
flowers (Hitchcock and Cronquist 1973) and seeds that are wind-dispersed with varying
samara size. Seed size may be influenced by pollinator availability, following the
“fertilization efficiency hypothesis” of diminishing returns from increasing numbers of
seeds (Sakai 1995, Sakai and Sakai 1996).

!

23!
!

!

!

!

While Lomatium plants can support endangered species, there are also several
federally endangered Lomatium species, including Lomatium cookii J.S. Kagan in Oregon
(Cook's lomatium; USDA NRCS National Plant Data Team n.d., Pendergrass 2010) and
Lomatium bradshawii (Rose) Math. & Const. in Oregon and Washington (Bradshaw's
lomatium; Pendergrass et al. 1999, Kaye et al. 2001). Information about L. utriculatum’s
response to fire may be applicable to these other species, and its relative abundance
makes it a good surrogate for studying these more protected species.
CONCLUSION
While prairie restoration is an active topic of continued discussion, there is still much left
to learn about the effects of various restoration practices. Prescribed fire is a commonly
used restoration tactic, but further study is needed to identify ideal seasonality and return
intervals in many ecological systems. Even if an ideal season and interval for burning is
determined, deviations from this schedule frequently occur for logistical reasons,
including local landowner opposition and weather. Depending on the seasonality,
intensity, and return interval, prescribed fire may result in a more diverse and
heterogeneous plant community.
Burning alters pollination and soil nitrogen, with results that are not well
understood. If fire increases both nutrient availability and pollination (Van Nuland et al.
2013), then based on Sakai’s (1995) “fertilization efficiency hypothesis” the increase in
resource availability and pollen abundance may lead plants to create more numerous
seeds rather than larger seeds. This is due to the higher abundance of pollen and thus
decreased costs of pollen capture, leading plants to invest resources into more numerous
smaller seeds since there is sufficient pollen to make this an effective strategy (Sakai

!

24!
!

!

!

!

1995). Increased soil nutrient availability may also increase pollination if nectar plants
are more abundant or have more numerous flowers.
Since some fire-adapted species have increased germination and/or sprouting
capabilities after fires (Agee 1996), and L. utriculatum is native to prairies with frequent
burn histories, L. utriculatum seeds may have higher germination and vigor if the parent
plant was burned, through environmentally induced parent effects (Lacey 1998). This
could be due to improved pollination, nutrient availability, or chemical compounds
released by the fire. Knowing the specific mechanisms by which the germination and
vigor of essential prairie species can be increased would allow for targeted management
to more easily achieve conservation objectives.
Prescribed fire may be a highly effective tool to alter plant communities, reduce
non-native plants, and increase fitness of key species—which are all important restoration
goals. Its continued use and study as a restoration tool is essential to the survival of many
endangered prairie species that are adapted to and depend on fire. Fire has maintained
Puget lowland prairies for thousands of years, and today it remains a cost-effective
restoration tool with a myriad of indirect benefits that we are only beginning to
understand.

!

25!
!

!

!

!

Chapter 2: Manuscript formatted for the journal Ecology
ABSTRACT
Prescribed fire has been used extensively as a prairie restoration tool, but its specific
impacts on prairie plant communities are not always interpreted within a spatially or
evolutionarily explicit framework. For example, plant community responses to fire can be
related to both treatment scale and the selective impact of fire on plant phenotypes, but
neither factor is accounted for in most plant community studies. Here, we used
experimental paired prescribed fire and fire-exclusion plots to examine the effects of
realistic large-scale prescribed fire on prairie communities in western Washington State.
As a case study of fire’s potential selective effects, we collected seeds from a common
prairie plant (Lomatium utriculatum) to examine second-generation effects of burning (in
the previous year) on seed vigor. Lomatium utriculatum seeds were germinated and
grown in cooperation with the Sustainability in Prisons Project, where inmates had active
roles in the care, tracking, and measurement of the plants, providing a valuable
opportunity for science education and inmate contributions to ecological restoration.
Lomatium utriculatum is an essential nectar source for the federally endangered Taylor’s
checkerspot butterfly (Euphydryas editha taylori) and may serve as a surrogate for the
effects of fire on endangered species recovery. Prescribed fire altered plant communities,
increased species richness, and enhanced germination of L. utriculatum. Lomatium
utriculatum seeds collected from burned plots had increased germination following the
exposure of parent plants to fire, exhibiting a second-generation effect of burning the
parent plant. Overall, this study demonstrates that prescribed fire could play an important
role in altering prairie composition and may be a potential selective force driving plant
micro-evolution in ecologically important species.
INTRODUCTION
Many grassland ecosystems around the world are subjected to regular burning and the
plants that inhabit these areas have become adapted to and thrive under the frequent
disturbance of fire. As plants in these environments have been burned regularly for
millennia, fire has played a fundamental role in shaping the communities that exist today
(Vogl 1974). In response to fire, plants can evolve flammability to eliminate their
neighbors (Bond and Midgley 1995), grow proportionately large roots in order to
resprout quickly (Bond and Keeley 2005), and become more well-suited to “postfire
environments” that often have reduced moisture availability and altered edaphic factors
(Keeley and Fotheringham 2000). Some species thrive under regular burning, but are

!

26!
!

!

!

!

more responsive to the indirect effects of fire. These plants may be able to utilize
resources in a postfire environment effectively, resprout quickly to outcompete
neighbors, or produce offspring that are more viable when the parent plant is exposed to
fire. One such species is Lomatium utriculatum (Nutt. ex Torr. & A. Gray) J.M. Coult. &
Rose (spring gold).
Lomatium utriculatum is a native perennial forb common to Puget lowland
prairies and an important nectar source for the federally endangered Euphydryas editha
taylori W. H. Edwards (Taylor’s checkerspot butterfly) as well as other important
pollinators (Shepard 2000). It is used here as a case study to identify patterns in seed size
and viability following fire, through the indirect second-generation fitness effects that
burning a parent plant can have on offspring.
Prescribed fire was commonly used by Native Americans on Puget lowland
prairies (Boyd 1999) to enhance native plant species and create open hunting ranges, and
it is still widely used in prairie restoration today. It is a particularly effective tool at large
scales, where it becomes far more affordable per acre than other treatment methods
(herbicide or manual weed removal; Bidwell et al. 2002, Hesseln and Berry 2008). Along
with maintaining the open structure that many native species require (Hamman et al.
2011), burning can increase the germination rates of some native species (Boyd 1999)
and remove excess nitrogen that may otherwise give invasive species a competitive edge
(Seastedt 1995).
Fire can alter plant competition and success in many ways. It can be a driver of
plant evolution (Bond and Midgley 1995), affect soil characteristics (Bond and Keeley
2005), and even alter soil surface temperatures for years afterwards (Andrusyszyn 2013).

!

27!
!

!

!

!

Fire can help disperse seeds with powerful updrafts (Murray 1987, Tackenberg et al.
2003) and enhance seed germination directly through heat and/or chemical cues (Keeley
and Fotheringham 2000, Yi-Fang Chou et al. 2012). In some species, germination may be
increased indirectly due to a higher abundance of pollinators (Van Nuland et al. 2013)
and/or healthier parent plants following fire.
We hypothesized that L. utriculatum seeds from burned plots would be smaller
than those from unburned plots, based on the “fertilization efficiency hypothesis” (Sakai
1995, Sakai and Sakai 1995, 1996) due to increased nutrient and pollination availability
(Van Nuland et al. 2013). Where pollination availability increases along with nutrient
availability, plants produce a larger number of smaller seeds, because pollination does not
limit the potential viability of the seeds (see Sakai 1995). These seeds were also expected
to have higher viability (because of increased parent plant fitness in postfire
environments) and therefore higher seed density due to a higher proportion of fully
developed seeds. Higher viability seeds from burned plots were expected to have more
complete and rapid germination.
We expected that prescribed burns would also alter overall plant communities
(Ewing and Engle 1988, Collins and Gibson 1990), enhancing the ability of prairies to
meet wildlife habitat and other restoration goals (Hamman et al. 2011). We predicted that
plant diversity and richness would increase after fire, since removing species that tend to
dominate an area would allow for several new species to take hold, and that evenness
would increase due to the potential removal of dominant non-native species.
Heterogeneity may be increased after fire because of the patchy burn pattern that occurs
in prairie fires with low fuel loads (Risser 1990, Fuhlendorf et al. 2006). We also

!

28!
!

!

!

!

expected to find a higher native/non-native ratio in burned plots, which is a major goal of
prescribed fire used for restoration. We anticipated differences in species
presence/absence between treatments, as some species may be present in unburned plots
that would not survive in burned plots or vice versa. We also hypothesized that overall
community composition would vary by burn treatment and by year: by burn treatment
due to increased heterogeneity and possible decreases in less fire-adapted species, and by
year because of different weather conditions between the two years.
METHODS
STUDY AREA
Puget lowland prairies are a rare and greatly diminished ecosystem, with “intact prairies”
extant on only 3% of historical grassland areas (Crawford and Hall 1997). As an
anthropogenic system, Puget lowland prairies require continual maintenance by humans,
usually in the form of prescribed burning. Prescribed fire has been increasingly used as a
restoration tool for Puget lowland prairies (Hamman et al. 2011), but monitoring and
evaluation of restoration efforts in these ecosystems have been lacking. This study
attempts to quantify some of the community and individual-species changes instigated by
prescribed fire.
Our study took place in remnant Puget lowland prairies southeast of Olympia,
Washington, on Joint Base Lewis-McChord (JBLM), home to some of the largest and
most native remaining prairie habitats in Washington State. This 36,684 ha military
reservation (JBLM 2012) has been subject to regular burns resulting from military
activities and widespread prescribed fires on 3,000 ha of prairie and oak woodlands since
1978 (Tveten and Fonda 1999). These prairies have soils in the Spanaway series (Tveten

!

29!
!

!

!

!

and Fonda 1999) that are well-drained and made up of glacial outwash and volcanic ash
(Zulauf and Washington Agricultural Experiment Station 1979). They receive mean
precipitation of 89-114 cm, air temperatures of 10.6˚C, and have average slopes of 0-6%
(Zulauf and Washington Agricultural Experiment Station 1979). The prairies of interest
in this study are located at approximately 46.92˚N, 122.73˚W.
These prairie sites host a wide variety of native plants and pollinator insects, and
are burned for ecological preservation. Invasive species are common, including some
such as Cytisus scoparius (Scotch broom) that alter the habitat type from grassland to
shrubland. Another species of concern on the prairies is Pseudotsuga menziesii (Douglasfir); in the absence of regular fires it will gradually turn prairies in this region into forests.
EXPERIMENTAL DESIGN
In order to observe patterns in plant community structure between burned and unburned
prairie sites at a relatively large and heterogeneous scale, 11 plot pairs (burned and
unburned; Figure 2-1) were measured for two consecutive years. The plots in this study
are located within a mosaic of burn histories at Johnson and Upper Weir prairies. A
Modified-Whittaker (20 m x 50 m) nested subplot design (Stohlgren et al. 1995) was
used in order to account for rare-but-present species and analyze species-area curves. Plot
pairs were established randomly in 2010 using Arcmap 9.2 (ESRI 2009) and designed to
have similar species composition and communities at the outset of the study. Large-scale
prescribed burns were completed in 2011 and 2012, with the unburned plots being
excluded from fire by the use of mowed fireguards in buffers around the plots. The
corners of plots were permanently marked, and GPS coordinates were taken at each
corner with a Trimble GeoXT and TerraSync software, using point-averaging for 200

!

30!
!

!

!

!

readings per point. All plots were burned in 2011, but only three sets of plots were burned
in 2012 (Figure 2-1). The three plots that were burned in both years were used to examine
the second-generation effects that burning a parent plant can have on its offspring.

Joint Base Lewis-McChord Paired Burn and Exclusion Plot Locations
Johnson Prairie

Upper Weir Prairie

Legend
Burn Plots
Control Plots
2013 Burns
2012 Burns
!

!

!

!

!

!

!

!

2011 Burns
2010 Burns
2009 Burns

O

!

0

0.25

0.5

1
Kilometers

Jaal Mann
Plots with Source:
more intensive
burn history
were
usedUSGS, AEX, Getmapping,
Esri, DigitalGlobe,
GeoEye,that
i-cubed,
USDA,
Aerogrid, State
IGN, IGP,
The Evergreen
College
swisstopo,
and the GIS User Community
for sampling
second-generation
effects are circled
2014

Figure 2-1. Map showing the paired plot locations on Johnson and Upper Weir prairies. The three
plots used for the Lomatium utriculatum sampling are circled. These were the only plots that had
been burned in both 2011 and 2012, and therefore were hypothesized to show more pronounced
effects of burning. Only a small patch of prairie was burned in 2013, and none of the plots were
within it.

31!
!

!

!

!

PLANT COMMUNITY RESPONSE TO PRESCRIBED FIRE
Within each Modified-Whittaker plot, 1 m2 subplots were measured using a 0.5 m x 2 m
sampling frame, recording species presence and estimating percent cover. Due to time
constraints in 2012, only four 1 m2 subplots were measured for each of the 22 plots and
the 10 m2 subplots were not measured; in 2013 all ten 1 m2 subplots (the typical number
for Modified-Whittaker plots) were measured for each plot along with the two 10 m2
subplots. In both years species presence/absence was determined for the 10 m x 20 m
subplots and the entire 20 m x 50 m plots at all 22 locations.
SECOND-GENERATION EFFECTS OF PRESCRIBED FIRE
Lomatium utriculatum seeds were collected in three burned and three unburned plots to
analyze how a fire-adapted species may display second-generation effects of burn
treatments. These collections focused on the three pairs of plots with the most recent burn
history (burned in 2011 and 2012). Plants were identified for collection by walking in a
random direction (determined by the second hand of a watch) and stopping at the first L.
utriculatum encountered. All of the seeds possible were collected when multiple stalks
were coming from the same plant. The seeds were then dried at room temperature for a
month and stored dry for another 2.5 months. After the dry storage period, the seeds were
weighed and measured to calculate mass per seed and approximate seed density
(mass/length). Seed mass and length were calculated for a subset of five seeds from each
plant. Up to 20 seeds from each plant were then placed into cold and moist dark
stratification for 64 days at 3˚C in a germination chamber (Hoffman Manufacturing, Inc.)
followed by one month of 15˚/7˚C 12-hour germination cycles with light during the 15˚C
“day” cycle. Time-to-germination—functionally defined as completion of germination, as

!

32!
!

!

!

!

in McNair et al. (2012)—was recorded for each seed and after germination each seed was
planted into an individual cone-tainer (Bergeron and Leach 1975). Germination began at
approximately 45 days, therefore the majority of the seeds had already been sown by the
time germination temperature cycles began and germination rates did not seem to be
affected by the changes in temperature and lighting.
Germination testing took place at a Center for Natural Lands Management
nursery through the native prairie plant production program run by the Sustainability in
Prisons Project (www.sustainabilityinprisons.org). Several inmates had the opportunity to
be involved in germination counting, monitoring of the seedlings, and prairie plant
education programs (LeRoy et al. 2012). Seedling leaf area was measured after
approximately three months of growth by photographing a randomly selected subset of
68 plants and calculating leaf area using ImageJ 1.48v (Wayne Rasband, National
Institutes of Health).
STATISTICAL ANALYSIS
Lomatium utriculatum seed mass and density were compared by burn treatment to
observe any changes to the parent plant’s seed characteristics due to past fires. Statistical
analyses were carried out using JMP 10.0 (SAS Institute 2012) with a t-test for seed
mass, and resampling t-tests using the resampling stats 4.0 add-in (Resampling Stats,
Inc.) for Microsoft Excel 2010, with 10,000 iterations for intra-plant seed mass ratios,
seed density, and seed number, which were not normally distributed. Survival analysis, or
time-to-event analysis, was run in JMP to analyze time-to-germination with a log-rank
test to compare curves between treatments. This analysis was used because the
germination data were “right-censored” since ungerminated seeds, which could

!

33!
!

!

!

!

theoretically have germinated at some later date, remained at the end of the experiment
(McNair et al. 2012). To compare seedling leaf area between treatments, a resampling ttest with 10,000 iterations was used in the resampling stats add-in.
We hypothesized that prescribed burns would increase species diversity and
richness. Species richness was determined by burn treatment in PC-ORD 6.0 (McCune
and Mefford, 2005) and analyzed using the resampling stats add-in with 10,000
iterations. This analysis was carried out with combined data from both the 2012 and 2013
1,000 m2 species presence/absence (P/A) plots. Richness, evenness, and Shannon’s and
Simpson’s diversity indices were calculated for both 2012 and 2013 separately for the 1
m2 plots using PC-ORD. These diversity metrics were analyzed using resampling due to
non-normality for Simpson’s diversity index and evenness in both years, and for
Shannon’s diversity index in 2013. T-tests were used in JMP for richness in 2012 and
2013, and Shannon’s diversity index in 2012. Species-area curves were compared for
each year to examine the differences in species accumulation between treatments (Clay
and Holah 1999, McCune and Grace 2002). Native/non-native ratios between burn
treatments were analyzed for 2012 and 2013 combined using the resampling stats add-in
with 10,000 iterations due to non-normality.
To examine community variation due to burn treatment, we collected species
cover data in 1 m2 plots and P/A data in 1,000 m2 plots. Data from these plots were
analyzed separately for each year (2012 and 2013). For our 1,000 m2 species P/A plots,
Multiresponse Permutation Procedures (MRPP) and Nonmetric Multidimensional Scaling
(NMS) ordinations were run in PC-ORD. For analysis of our 1 m2 species cover plots,
NMS ordinations were run in PC-ORD. All NMS ordinations used 3 dimensions, 1,000

!

34!
!

!

!

!

runs, and 500 maximum iterations. Overall community change (plant cover by species)
was compared for each plot by burn treatment, with separate analyses for each year.
These analyses used MRPPs in PC-ORD to identify any changes in the structure of plant
cover in the community due to burn history. Data from the two collection years (2012 and
2013) were compared separately by burn treatment using MRPPs in PC-ORD to examine
succession following a burn and ensure that year of data collection was not a confounding
factor due to variable weather between the years. Indicator Species Analysis (Dufrene
and Legendre 1997) was run in PC-ORD to identify species that were more often present
in certain burn treatments.
RESULTS
SECOND-GENERATION EFFECTS OF PRESCRIBED FIRE
Lomatium utriculatum time-to-germination varied significantly between burn treatments,
with seeds from burned parent plants germinating at higher rates (Figure 2-2; P<0.0001).

Burned

Unburned

P<0.0001

Figure 2-2: Time-to-event analysis of seed germination data by days since seed imbibing.
Germination rates varied between burn treatments, with seeds from burned parent plants
germinating earlier and more successfully (log-rank P<0.0001).

!

35!
!

!

!

!

Germination of seeds from burned parent plants also occurred earlier and to an overall
higher percentage of the total seeds by the end of the experimental period.
Intra-plant seed mass ratios, seed density (g/mm), and seed number showed no
significant effect of burn treatment (intra-plant seed mass ratio P=0.5487; density
P=0.3701; seed number P=0.4309). Mass per seed also did not vary significantly by burn
treatment (DF=119, T=0.8820, P=0.3795). Seeds varied greatly in mass within plants,
with the largest seed nearly 16 times the mass of the smallest for one individual.
Seedling leaf area did not vary between treatments (P=0.2275), between plots
(P=0.8429), or by germination date (P=0.4333). While some seedlings did not survive,
those that did and were large enough to measure photographically varied in size nearly
five hundredfold (2.6 mm2 to 1,261.2 mm2).
PLANT COMMUNITY RESPONSE TO PRESCRIBED FIRE
1000 m2 Presence/Absence Scale
Richness was significantly higher
(increased nearly 19%) in burned
treatments for the 1000 m2 P/A

P=0.0004

plots (Figure 2-3; P=0.0004).

Figure 2-3. Species richness was nearly 19% higher in burned plots (P=0.0004), indicating that
species had been lost from unburned plots after only 1-2 years of burn exclusion. Data are from
the 2012 and 2013 1000 m2 presence/absence plots combined and bars represent means ± standard
errors.

!

36!
!

!

!

!

Several indicator species were identified for each treatment in both years, mostly
indicating for burned treatments. In 2012, four significant indicator species were
identified, all indicative of burned plots. Of these, only one was native (Zigadenus
venenosus var. venenosus). In 2013, eight burned indicator species and two unburned
indicator species were present. All five non-native indicators were for burned treatments,
while three natives indicated for burned and two for unburned treatments (Table 2-1).

Table 2-1. Indicator Species Analysis (Dufrene and Legendre 1997) determined that several
species of plants were indicative of the various burned and unburned treatments. These included a
number of native and non-native species.

!

37!
!

!

!

!

Multiple Spatial Scales
In order to examine our hypothesis that burned plots would be more diverse at all scales
than unburned plots, we created species-area curves for both treatments (Figure 2-4).

Figure 2-4. Species-area curves for 2012 and 2013, with 95% confidence intervals illustrating
the lack of significant differences between burned and unburned plots in species accumulation.

The confidence intervals for these curves overlapped heavily, indicating that a pattern in
species diversity at various scales was too subtle to determine using this method
(although richness did vary significantly at the 1000 m2 plot size). Species-area curves for
both years appear to approach asymptotes, demonstrating that our largest plot size (1000
m2) is representative of the overall number of species in the community.

!

38!
!

!

!

!

1 m2 Species Cover Scale
Native/non-native ratio was analyzed
for both 2012 and 2013 combined as a way to
determine if prescribed fire is increasing the
proportion of native cover. No significant
difference was found in native/non-native

A=0.0088
P=0.0188

ratio between the treatments (P=0.1172). In
both 2012 and 2013, our hypothesis that
burned and unburned communities would vary
based on cover was affirmed (Figure 2-5;
2012: A=0.0088, P=0.0188; 2013: A=0.0090,
P=0.0002). For the 2012 plots, burn
treatments explained 22.589% of the variation

A=0.0089
P=0.0002

between burned and unburned communities;
in 2013, burning explained 19.244% of the
variation in plant community structure.
Figure 2-5. NMS ordinations for species cover from our 1 m2 plots show weak but significant
differences between burn treatments in both years (2012: F=2.1672, P=0.0002; 2013: F=3.383,
P=0.0002). The 2013 data appear to be more clustered in both treatments, possibly due to varying
climatic conditions between years.

!

39!
!

!

!

!

Diversity metrics were compared by burn treatment separately for 2012 and 2013,
with no significant trends (2012: richness t=–1.0353, DF=94.5, P=0.3035; evenness
P=0.2687; Simpson’s diversity index P=0.1143; Shannon’s diversity index t=–1.1996,
DF=80.9, P=0.2338; 2013: richness t=–0.9164, DF=217.8, P=0.3603; evenness
P=0.9724; Simpson’s diversity
index P=0.4818; Shannon’s
diversity index P=0.3808).
We compared communities
between 2012 and 2013 in order to
determine how much weather
altered species cover. Species
cover changed significantly

A=0.0269
P<0.0001

between years for both
communities (Figure 2-6; Burned:
A=0.0269, P<0.0001; Unburned:
A=0.0218, P<0.0001), and
sampling year explained a higher
percentage of the variation for the
burned plots (28.3% for burned,

A=0.0218
P<0.0001

21.5% for unburned).
Figure 2-6. Comparison by year for each burn treatment, to examine the effects of temporal
succession within plots in terms of plant community change. Year significantly affected plant
communities in both treatments (Burned: A=0.0269, P<0.0001; Unburned: A=0.0218, P<0.0001).
Year explained 28.3% of the community variation in burned plots, and 21.5% in unburned.

!

40!
!

!

!

!

DISCUSSION
SECOND-GENERATION EFFECTS OF PRESCRIBED FIRE
Germination was increased in seeds from burned plots, which could be explained by
increased pollination, chemicals from smoke absorbed through the parent, or more
vigorous parent plants due to increased soil nitrogen following fire. Parent plants may
also be more fit in burned areas because healthier parent plants are more likely to survive
fires. Seed mass did not vary significantly between treatments, likely because of the huge
intra-plant variation in seed mass observed for many Lomatium species, both in this study
and elsewhere (Thompson 1984, Thompson and Pellmyr 1989). We were not able to
obtain accurate counts for the number of seeds on each plant because many plants had
already dispersed some of their seeds. For this reason, we were unable to test the
“fertilization efficiency hypothesis” that smaller numbers of heavier seeds are produced
with increased resource availability (Sakai and Sakai 1996).
While fire is known to directly increase germination in some species through heat
scarification and chemical compounds in smoke—sometimes distilled to “smoke water”
for use in plant propagation from seed (Yi-Fang Chou et al. 2012)—and indirectly
through “postfire environments” with altered growing conditions (Keeley and
Fotheringham 2000), the second-generation indirect effects seen here are novel because
they did not rely on fire’s direct effects on seeds. The effects of fire were seen here on
seeds that were germinated under controlled lab conditions and had never been directly
exposed to fire, fire surrogates, or the post-fire environment.
Many chemicals produced during the burning of vegetation may be able to
stimulate germination, including nitrogen oxides, cyanohydrins, and karrikins (Nelson et
al. 2012). It is possible that the second-generation germination increases seen in this

!

41!
!

!

!

!

study were the result of one or more of these germination-stimulating compounds being
taken up by the parent plant and imparted into the seeds before they were released, as
these compounds may have been more prevalent in the soil of burned plots. Another
potential explanation is that L. utriculatum can sense postfire conditions and alter its
development in response, as some fire-adapted plants can (Nelson et al. 2012). Finally,
the influences of compounds in smoke may be expressed in a plant’s genetic makeup
(Nelson et al. 2012), potentially with effects that could carry over to its seeds. DNA was
found to be changed by exposure to chemicals from smoke (Jain et al. 2008), and smokeresponse genes can increase stress during germination in order to better prepare seedlings
for environmental stressors (Soós et al. 2009).
PLANT COMMUNITY RESPONSE TO PRESCRIBED FIRE
Overall plant cover and richness varied between burn treatments, which is consistent with
most previous studies that found fire to have a formational role in prairie plant
communities (Collins and Gibson 1990). We expected to find increased native to nonnative ratios in the burned plots, but this may depend to some extent on the season in
which the plot was burned, which could obscure patterns (Howe 1995, Pollak and Kan
1998, Tveten and Fonda 1999). In this study, no significant differences in native/nonnative cover between treatments were found, but there was a slight trend towards
increased native cover in burned plots. This is consistent with research demonstrating that
fire can control some non-native species, but unless it is carefully targeted at a certain
species, prescribed fire can either be ineffective for invasive species control or lead to an
increase in some non-native species (Ditomaso et al. 2006).

!

42!
!

!

!

!

Depending on the restoration goals and local climate, the ideal season for
prescribed burning may vary. In some cases burning in late spring may be better suited
than fall for eliminating invasive species, as native plants tend to disperse their seeds
earlier in the year (Pollak and Kan 1998), while Tveten and Fonda (1999) found fall fires
more effective for promoting native species. Howe (1995) found variable effects of
season on many individual species along with an increase in annual and biennial species
following summer burning.
Prairies, which are dominated by herbaceous species, develop at different rates
each year due to variation in weather and other variables. While we did find patterns in
community change between years, hotter and drier weather in 2013 (compared to 2012)
combined with sampling later in the season may have driven some of these changes. This
created a challenge when comparing data between years and also led to the senescence of
some species, making them difficult or impossible to identify. While our timeframe for
this study was too short to be able to analyze long-term instability in plant communities,
as found by Collins (2000) in tallgrass prairie, future sampling in these plots could
identify any instability found in the communities at our study site.
This study supports recent findings that prescribed fire in Puget lowland prairies
can change plant community composition, but also demonstrates that further research is
necessary to determine optimal techniques for improving the ratio of native plants in
restoration sites. At least on the time scales involved in this research (two seasons of
burning), fire did not lead to a significantly increased proportion of native species. With
more time the trend towards native cover may strengthen, but for shorter-term restoration

!

43!
!

!

!

!

projects other methods such as herbicide and manual removal may need to be employed
in order to see rapid and consistent results.

!

44!
!

!

!

!

Chapter 3: Extended discussion and future research
In this chapter, I will expand on conclusions and analysis, make recommendations for
future research and actions, and explain how my results fit into the larger picture of
prairie research and management. There were some results and conclusions that were left
out of the manuscript because they were either non-significant or less relevant, and others
that can benefit from further in-depth explanation. Throughout the research and writing
of this thesis, I have encountered many areas where more research is needed, and have
developed ideas for how to implement modified land management practices using current
knowledge. Finally, I will share some of these thoughts on future research and
management techniques as a way for this study to become a building block in our
understanding of how prairie plant communities function.
EXPANSION OF CONCLUSIONS
Despite this study’s many fascinating findings, there were several hypotheses that did not
have their expected outcomes. Plant cover at the 1 m2 plot scale was found to be different
overall between treatments, but diversity metrics did not differ at this scale. At the 1000
m2 plot scale diversity and evenness still did not vary, but richness was increased in
burned treatments. This signifies that plant communities are changed by fire and that
unburned plots had fewer species present after only one to two years of burn exclusion.
Indicator Species Analysis (Dufrene and Legendre 1997) demonstrates that some species
tended to be present in larger numbers in burned areas, but 60% (6/10) of these were nonnative. The cover of only two species was increased in unburned areas and both were
native.

!

45!
!

!

!

!

Prairie plant community studies often use small-scale fires, but this study
attempted to accurately represent the variable intensity of prairie fires by using largescale burns that were more similar to natural fires. Within a burned plot there are many
different fire intensities evident due to variations in topography and fuel load. It was
hypothesized that this varying fire intensity may create novel microhabitats and lead to an
increase in plant diversity, but this pattern was not seen. This may be because these plots
have only experienced different burn treatments for two years and it would take more
time for plant communities to shift.
A primary goal of prescribed fire is increasing the native/non-native ratio of the
plant community. While native cover did appear to increase slightly, these differences
were not statistically significant (P=0.1172). This could indicate that native cover is being
increased, but only slightly, and larger-scale surveying would be needed to determine if
there is an actual statistically significant trend. Community analysis determined that plant
communities did vary due to prescribed fire although the differences were subtle. Our
weak trend towards increased native cover in burned plots could imply that current
prescribed fire practices are effective at increasing the proportion of native cover.
When collecting and analyzing L. utriculatum seeds, my methods limited some of
the analyses I could perform. I collected data on the number of seeds per plant, but these
were not truly representative of all of the seeds on the plant in many cases. When I
arrived to collect seeds some of the plants that were riper had already lost many seeds in
the wind, and the seeds I collected were only a subset of the original seed number. While
seed number was significantly different between plots, this may not be indicative of a real
pattern because one of the plants was a large outlier (several times the second-largest’s

!

46!
!

!

!

!

number of seeds) and caused this result. Future studies could include a mesh bag to
collect any seeds that fall off before sampling or count the numbers of unripe seeds on the
plants prior to collection in order to have a more accurate count of the total number of
seeds on the plant.
Insect damage can influence seed size but high variability in L. utriculatum seed
size may obscure any effects. Ellison and Thompson (1987) found that weevils feeding
on Lomatium grayi seeds could have either a positive or negative effect on surviving seed
size depending on the weevil species. While the intact seeds from the weevil-damaged
groups had lower germination than control groups, those that germinated grew to an
equal size at 45 days compared to the controls (Ellison and Thompson 1987). They also
found that 45-day plant mass was correlated with seed mass except in greenhouse
conditions with no competition and large pots. Seed size and mass for L. utriculatum in
this study were highly variable among plants (but fairly consistent within plants) and no
trend due to burning was identified. Seed size in L. utriculatum may be driven by other
factors such as plant genetics, obscuring any influence that changes in pollination or other
effects of fire could have on seed size or mass. Further research into L. utriculatum seed
size could seek to identify the primary causes of variation, by controlling for genetic
lineage and/or pollination.
FURTHER RESEARCH
While I found increases in germination due to the second-generation effects of fire, the
direct cause of this increase is difficult to identify. Increased germination could occur
because of direct effects of chemicals from burning that could be sequestered in the
parent plant or simply due to increased parent plant vigor after burning for a variety of

!

47!
!

!

!

!

reasons including altered competition, pollinator availability, and/or soil nutrient
composition.
Maternal effects from a seed’s parent plant can be passed through hereditary
pathways and also through the mother plant’s environment (Luzuriaga et al. 2006).
Schmid and Dolt (1994) used Solidago altissima L. as a case study of environmental vs.
genetic effects on seedling vigor and found both maternal and paternal genetic effects
along with effects based on the parent plant’s environmental conditions (i.e. secondgeneration effects). They grew S. altissima in garden and greenhouse experiments and
found that seedlings from plants growing in different soil media had varying germination
and seedling size (Schmid and Dolt 1994). In this study I found similar effects based on
parental environment, but using a field-based approach and the environmental variable of
exposure to fire instead of varied growth media.
It is possible that parent plant age can affect seed germinability, with older plants
producing fewer viable seeds (Fenner 1991). If burning killed some older L. utriculatum
plants and increased the abundance of young plants, subsequent seeds could have higher
germination due to the younger average parent plant age. Unfortunately, it would be
impossible to test for parent plant age in this experimental system because L. utriculatum
has no age indicators.
Increased temperature can promote germination in many species of plants, and at
Joint Base Lewis-McChord’s Johnson and Upper Weir prairies Andrusyszyn (2013)
found that soil surface temperature was higher in more recently burned areas. While my
L. utriculatum seeds in this research were germinated ex situ so germination temperature

!

48!
!

!

!

!

did not alter results, warmer temperatures in burned plots may still correlate with more
vigorous parent plants or increased in situ germination.
The order of umbel that a seed is borne on—which was not analyzed in this
study—can also influence germination rates (Fenner 1991). Working with Apium
graveolens L. (celery), in the same family as L. utriculatum (Umbelliferae), Thomas et al.
(1979) found that primary umbels produced heavier seeds but that viable seeds from
quaternary umbels germinated more readily, while the heaviest seeds produced the largest
seedlings. Future research with Lomatium could separate each plant’s seeds by umbel
during collection and analysis to determine if these effects occur.
While the second-generation germination increases seen here on L. utriculatum
did not rely on direct exposure to fire, testing its seeds for response to smoke water would
help build a more complete understanding of the species’ germination. Thomas and
Staden (1995) found that smoke water increased the germination of some varieties of A.
graveolens, and if similar effects were identified for L. utriculatum it would further
complicate our understanding of how this species responds to fire.
Tetrazolium chloride testing (Cottrell 1947, Smith 1951, Witty 2012) is a method
that verifies what percentage of seeds were viable and had the potential to germinate.
This technique has been successfully used on Umbelliferae species (Rita et al. 2012).
Tetrazolium chloride testing prior to germination tests would allow a better
understanding of why some seeds do not germinate and what percentage of viable seeds
do successfully germinate. The test kills the seed being examined, so tests would have to
be carried out on a separate subsample of seeds from the same parent plant as those being
germinated.

!

49!
!

!

!

!

Increased pollination could lead to a larger abundance of viable seeds, and higher
pollination can occur in burned sites (Van Nuland et al. 2013). While this higher
pollination rate may simply be due to the fact that more vigorous plants after burning
produce more flowers (Van Nuland et al. 2013), it could still be an important
environmental effect that cannot be controlled for by germinating seeds ex situ. Research
observing pollinator abundance on sites where seeds are collected or excluding
pollinators and pollinating by hand would help to determine how pollination might alter
the second-generation effects of fire.
The plants that were produced in this study could be planted back onto the prairie
that their seeds were collected from into reciprocal burn treatments and monitored to
determine if observed differences between treatments are due to their environment or
their genetics. If the plants originally from the burned plots continue to produce more
viable seed even when planted in unburned areas then it would imply that increased seed
viability in burned areas is due to genetic adaptations and not environmental conditions.
While some research has discussed the optimal season for burning to increase
native cover (Pollak and Kan 1998, Tveten and Fonda 1999), there is a lack of studies
designed to directly answer this question. Further research experimentally comparing
burns during different seasons at the same location could help land managers plan their
burns in accordance with maximizing ecological benefit. One of the reasons that such
studies do not exist currently is the difficulty of burning during most of the year. If the
weather is not optimal it can either be too wet (natural restriction) or too hazardous
(human restriction) to burn.

!

50!
!

!

!

!

While it appeared that the proportion of native cover might be increased slightly
by prescribed fire, future research to solidify this trend would be helpful to encourage
prescribed fire as a technique to increase native plant dominance. Such research would be
especially relevant to conservation groups whose main goals with prescribed fire often
include increasing the proportion of native species.
PRAIRIE RESPONSE TO A CHANGING CLIMATE
In the 21st century, the Pacific Northwest is expected to have fairly constant annual
precipitation but with wetter winters and drier summers (Mote and Salathé 2009). As
climate change progresses, droughts will become more common, reducing prairie
productivity and favoring a plant community shift towards species that grow quickly,
early in the year (Thorpe 2011). Native prairie species tend to grow quickly and early in
spring, dispersing their seeds before many non-native species (Pollak and Kan 1998),
which may be a benefit to native species as climate change progresses. Species with
longer dispersal distances will be more able to migrate and keep up with the changes in
climate (Thorpe 2011). Species’ ranges are broadly shifting northward as the climate
warms and prairie ecosystems are no exception. Two responses that a prairie land
manager can take to facilitate this northward shift are increasing landscape connectivity
and/or assisted migration (Thorpe 2012).
The three categories of possible action in response to a changing climate are
resistance, resilience, and response (Millar et al. 2007). Resistance means attempting to
maintain the current state of the ecosystem, at least temporarily. This likely will not be a
permanent solution, so resilience may often be a more powerful tool that can allow
ecosystems to more easily return to a previous state after a disturbance. In the case of

!

51!
!

!

!

!

climate change, however, the disturbance may be ongoing, leaving response as the
preferred adaptation method. Response from a management perspective means taking
actions that can accommodate long-term climate change, including assisted migration and
planting species outside their core historical ranges (Millar et al. 2007) in areas that may
become preferred under the projected climate changes.
HOW THIS RESEARCH FITS INTO PRAIRIE MANAGEMENT
LAND MANAGEMENT
An understanding of the second-generation effects of fire is important to prairie
restoration to plan and meet goals. Land managers seeking to improve habitat for
butterflies and other pollinators can use knowledge of the second-generation effects of
fire to select seeds from plants that were burned previously, even in cases where the seeds
are not exposed directly to fire. Planning with second-generation effects in mind may also
help more effectively predict outcomes of prescribed fires.
Soil nutrients can alter both plant production and community composition on
prairies, and they can be affected by burning. Excess soil nitrogen can increase
productivity but favor invasive species (Suding et al. 2005, Clark et al. 2007). Since
prescribed fire alone is often not enough to increase native/non-native ratios, excess soil
nitrogen reduction—sometimes together with herbicide and manual removal—can be a
powerful tool to improve the proportion of native species on prairies. This can be
accomplished by adding activated carbon or sugar to the soil to adsorb nitrogen or
increase microbial activity to consume excess nitrogen, respectively (Mitchell and
Bakker 2011).

!

52!
!

!

!

!

If further research into second-generation effects finds them in other target
enhancement species, burn timing could be altered to promote these species. If the target
species in an area exhibit second-generation increases in germination and/or growth after
fire then an ideal burn interval would be to burn before each generation begins to set
seeds. If a target plant takes six years to begin producing seeds then burning every five
years may be most effective to generate strong second-generation effects. With burn
intervals based on phenological timing, each larger generation occurring following a fire
year would have time to develop and be burned before it entered seed production.
RESTORATION ECOLOGY RESEARCH
Understanding the second-generation effects of fire on native prairie plants is an
important new area of research because second-generation effects can change how fire
affects prairie communities in unexpected ways. An important piece of prairie ecosystem
research is knowing how historic processes such as fire continue to influence plant
communities today. While fire’s beneficial direct effects on plants and seeds are fairly
well understood, the second-generation indirect effects of fire explored in this study open
up a new area of research into prairie plants’ adaptations to fire.
Germinability can be increased by environmental factors such as high
temperatures and nitrogen levels (Fenner 1991), both of which could be outcomes of
prescribed fire. Therefore, fire may be selecting for phenotypes which germinate more
readily following burning of the parent plant and driving rapid contemporary evolution
(Carroll et al. 2007). More research is needed into second-generation effects on other
species of prairie plants to see if this effect is widespread or an isolated occurrence.

!

53!
!

!

!

!

The fractal geometry of trees can vary based on genetics, with important
implications for organisms that use these habitats (Bailey et al. 2004). Species such as L.
utriculatum that can have highly varied umbel geometry could also inherit these traits
from their parent plants. Valuable research could be done to confirm if genetic variation
drives changes in umbel geometry and if these changes could be affected by fire. The
outcomes of this research would be important for studying arthropods, microbes, and
mycorrhizal fungi, which are all affected by plant genotypic diversity (Bailey et al. 2009).
If there is a broad pattern of burned plants producing more viable seeds (as L.
utriculatum did) it might explain part of fire’s major role in prairie plant evolution, by
promoting the continuation of genes from fire-adapted plants. This novel mechanism of
natural selection could help explain how adaptations to fire and other disturbances have
evolved throughout many ecosystems.
Most available research on grassland habitat is in the tallgrass prairie of the
Midwest. In many cases, these habitats are similar and these studies can be applied to
research on Puget lowland prairies. Midwest prairie ecosystems are much more expansive
and widely studied and so applying many of these studies to Puget lowland prairies is an
effective way to understand topics that have not been studied regionally, such as the
effects of heterogeneity on birds (Fuhlendorf et al. 2006). Studies in the Midwest often do
use prescribed fire but it can be for different reasons, including livestock grazing.
Grazing is often one of the main variables considered in research on prairie disturbance.
While the plant species on Midwest prairies are different, they usually fall under similar
functional groups to Pacific Northwest prairie plants.

!

54!
!

!

!

!

CONCLUSION
While prescribed fire has been used as a prairie restoration tool for years, there is a lack
of solid research specific to the Puget lowland area supporting some of its ecological
benefits. This study identified second-generation effects of fire on plants and reinforced
fire’s alteration of prairie plant communities, the former with the potential to alter how
fire’s community interactions are analyzed and the latter supporting prescribed fire’s use
as a land management tool. We also identified a possible trend towards an increased
proportion of native plant cover with prescribed fire, which is often a goal of land
management plans.
While further research is essential to determine the specific mechanisms of
second-generation effects of fire on plants, their existence opens up a new realm of study
beyond the direct effects that have been the focus of plant-fire interactions research in the
past. If the seeds of a parent plant are affected by fire indirectly it could have important
implications for plant evolution as another mechanism by which plants could evolve to
adapt to regular disturbance by fire. Future studies examining a wider range of species
under more varied conditions will help to uncover the specific mechanisms and extent of
fire’s second-generation effects on prairie plants and bring a stronger understanding of
how our landscapes can be altered by this powerful management tool.

!

55!
!

!

!

!

Chapter 4: Interdisciplinary connections
In this chapter I will discuss the human history of prairie habitats and why we should
continue to protect them. Next, I will examine some of the conflicts that local landowners
and developers have with prairie restoration in the areas of smoke pollution and wildfire
risk, followed by whether our conservation efforts can be (or need to be) sustained
indefinitely. Finally, I will discuss how prairie research can be intimately connected to a
variety of fields surrounding restoration ecology including social justice and
environmental education.
AN ANTHROPOGENIC ECOSYSTEM
Puget lowland prairies are an anthropogenic ecosystem, formed after glaciation around
8,000-12,000 years ago (Pielou 1991) and maintained by human activities since then in
order to maximize hunting and gathering opportunities (Boyd 1999). When considering
prairie restoration, people have to put aside the notion of “leaving nature be” and realize
that the natural process on prairies is one of human intervention. Native American
cultures shaped the prairies for thousands of years, and the species that are abundant
today may even be an outcome of the cultivation techniques that they used to promote
certain desirable food species:
“Rather than being major Indian food sources because they dominated the
prairies, bracken [Pteridium aquilinum (L.) Kuhn] and camas [Camassia quamash
(Pursh) Greene] more likely dominated the prairies because they were major
Indian food sources [emphasis added]” (White 1975).
The formational role that Native Americans played in Puget lowland prairie history
should not be discounted, and the cultural significance of these prairie ecosystems is one
of their many values.

!

56!
!

!

!

!

Puget lowland prairies were maintained by Native American burning for
thousands of years, but these burn practices may no longer produce the same desired
outcomes as they have historically. Since European settlement, development, invasive
species, and even climate change have altered how fire interacts with prairies. Some
invasive species can even be benefited by fire, necessitating other control methods for
removal. Finally, nearby landowners can reduce the opportunities for prescribed fire,
limiting when burns can be completed.
SMOKE POLLUTION AND WILDFIRE RISK
Particulate smoke pollution is linked to asthma and other health problems, but smoke can
be kept under safe thresholds (Bowman and Johnston 2005). Many cost/benefit analyses
of prescribed fire smoke pollution focus mainly on burning to reduce fuel loads and
prevent catastrophic wildfire (Yoder 2004, Black et al. 2008) but some argue that the
risks of not burning are smaller in prairie restoration. For this reason land managers
seeking to use prescribed fire for prairie restoration have a more difficult task of bringing
the public on board.
A major barrier to prescribed fire is landowner concern about smoke pollution. If
conditions are less than optimal, smoke can spread widely throughout areas and become a
cause for concern among nearby residents. Limiting burning to days with the right
weather conditions to achieve a strong updraft and disperse the smoke more widely can
largely alleviate this issue.
Prairies are level sites that are often ideally suited for easy building, and much
pristine prairie habitat has been lost to development. This encroachment of the built
environment also poses another problem by increasing the potential cost presented by

!

57!
!

!

!

!

fires escaping control. Nearby landowners are often worried about fire damage and can
oppose prescribed fire. Notifying landowners prior to burns, leaving large buffers, and
burning only under optimal conditions can help reduce these concerns.
In the Puget lowland region, burn bans due to wildfire and/or pollution risk play a
large role in determining when prescribed fires are permitted. In many years there is a
narrow window where burns are both possible and allowed, reducing the potential for
experimental changes in burn timing. These legal restrictions on prescribed fire have
made research into ideal burn timing challenging since timing can vary greatly between
years—and in some years burns may not be possible in all areas. Restoration of prairie
landscapes is not without risk, and this problem is magnified by the ongoing maintenance
that prairie ecosystems require.
COSTS OF MAINTAINING NATIVE PRAIRIE
Resource expenditure in prairie restoration can be high, with herbicides, hours of labor,
and fossil fuels all being used in the name of restoring prairie habitat. These each have
associated costs and care must be taken to minimize the collateral damage of restoration.
At the most basic level, maintaining our prairies is a necessity if we are to conserve the
numerous endangered plant and animal species that depend on them. The recent federal
endangered species listing of Euphydryas editha taylori (Taylor’s checkerspot butterfly)
has brought increased urgency to prairie restoration and provided a boost to programs
that have already been in place to conserve this butterfly for years, previously as a
candidate for federal endangered species listing.
Many restoration measures are costly and much of the funding comes through the
Army Compatible Use Buffer (ACUB) program at Joint Base Lewis-McChord (JBLM).

!

58!
!

!

!

!

Conservation of federally endangered species also provides money needed for restoration
projects, as there is funding available to remediate the organisms’ critical habitats. JBLM
was approved for the ACUB program in 2005, primarily for protection of Polites mardon
(Mardon skipper butterfly), Euphydryas editha taylori (Taylor’s checkerspot butterfly),
Eremophila alpestris (streaked horned lark), and Thomomys mazama (Mazama pocket
gopher; “JBLM ACUB Program fact sheet” n.d.). The Center for Natural Lands
Management (CNLM) is JBLM’s ACUB “Cooperative Agreement Partner” (“Joint Base
Lewis-McChord Army Compatible Use Buffer program fact sheet” n.d.), and the
Sustainability in Prisons Project’s prairie restoration activities largely take place through
partnership with CNLM.
WORK WITH THE SUSTAINABILITY IN PRISONS PROJECT
The Sustainability in Prisons Project (SPP) works in several areas of restoration ecology
as part of its mission to “bring science and nature into prisons” (LeRoy et al. 2012).
Along with raising endangered frogs, butterflies, and caring for endangered turtles, SPP
has three conservation nursery sites that grow native prairie plants. These plants are used
largely to establish butterfly reintroduction sites through intensive planting with native
species that are beneficial to Euphydryas editha taylori (Taylor’s checkerspot butterfly)
and to plant in seed-production beds to produce larger quantities of seed for directseeding restoration of prairie sites.
Inmates were involved in this research in the counting and sowing of germinants,
monitoring plant growth, and measuring plant vigor. They benefited from involvement
with this project by getting an opportunity to contribute to scientific knowledge and
increasing their interest in the science of plant restoration, and this study benefited by

!

59!
!

!

!

!

being able to more rapidly attain large numbers of measurements. I benefited from
working with inmates by gaining assistance with data collection and monitoring and
being able to clarify my results and their implications by discussing them with an
audience that does not have a background in restoration.
A large part of what SPP does is science education. This ranges from a science
and sustainability lecture series—monthly at two prisons—to regular workshops
conducted at the conservation nursery sites. Workshops at the nursery sites are helpful
not only to make sure that everyone knows how to complete the tasks at hand but also
that they understand how their work fits into the big picture of restoration ecology.
Inmates involved in this research were excited to have the opportunity to
contribute to the scientific process, and they worked hard to ensure the quality of the data
they collected. Working in a nursery facility can be monotonous, and it helps to have a
sense of contributing to a larger goal. Showing inmates how they can contribute to
scientific research can be an empowering part of work in a nursery and help them
understand the larger purpose of what they do.
Meaningful work while imprisoned can help develop a sense of self-worth, an
improved work ethic, and has the potential to decrease the likelihood of infractions
(LeRoy et al. 2012). Many inmates involved in SPP’s projects have become interested in
studying science when they are released or getting jobs in fields where they could work
with living things. Working with living plants and animals can be restorative and can
reduce aggression in prisons and possibly even bring down rates of recidivism (Rice and
Remy 1998, Richards and Kafami 1999, LeRoy et al. 2012).

!

60!
!

!

!

!

Watching plants grow can be therapeutic and it can help impart a sense of
regularity. Gardening programs in a San Francisco Sheriff’s Department jail were found
to increase inmate “desire for help” with drug-use related issues (Rice and Remy 1998).
Richards and Kafami (1999) found that horticultural therapy programs in prison decrease
vulnerability to addiction (defined as “psychological symptoms, tension, and distress”),
but do not increase resistance to addiction (defined as “behavioral controls and positive
self esteem”).
Growing native prairie plants requires good communications skills, attention to
detail, and the work ethic to sow, weed, and water tens of thousands of plants. Seeds are
scarce for some species and every seed has to count. Many of the inmates in SPP’s
conservation nurseries have become highly skilled at quickly and accurately dispensing
seeds into containers, even competing against themselves to beat their previous records.
The prairie restoration crew from Cedar Creek Corrections Center has learned many
applicable job skills and provided immense benefits to SPP, its many partners, and the
prairie ecosystem by making restoration that would otherwise be impossible a reality.
CONCLUSION
Despite the risks and challenges associated with prescribed fire, it is still a valuable
restoration tool. Puget lowland prairies must be preserved in order to maintain
biodiversity and their cultural heritage, although we must also accept that the prairies
may never return to the way they were before European settlement. Funding is likely to
continue for as long as prairie ecosystems are imperiled, through concerned non-profits
and federal and state endangered species programs. In areas where prescribed fire alone
cannot remediate a site, caution must be taken before investing large amounts of labor

!

61!
!

!

!

!

and/or herbicide into a project since they can quickly expend valuable resources and
potentially cause further harm to the environment. One of the major goals for prairie
restoration in the future should be determining what the target state for a modern “native
prairie” will be, because while they will never return to historical conditions, Puget
lowland prairies can be maintained in a state that preserves the functional assemblages of
prairie communities and the birds, animals, and insects that depend on them.

!

62!
!

!

!

!

REFERENCES
Agee, J. K. 1996. Achieving conservation biology objectives with fire in the Pacific
Northwest. Weed Technology 10:417–421.
Aldous, A. E. 1934. Effect of burning on Kansas bluestem pastures. Pages 1–65.
Technical Bulletin, Kansas State College of Agriculture and Applied Science,
Manhattan, Kansas.
Anderson, R. C. 2006. Evolution and origin of the Central Grassland of North America:
Climate, fire, and mammalian grazers. The Journal of the Torrey Botanical
Society 133:626–647.
Andrusyszyn, R. 2013. Using fire for butterflies: Soil characteristics across a burn
gradient in western Washington. Master’s thesis, The Evergreen State College.
Arnold, R. A. 1983. Ecological studies of six endangered butterflies (Lepidoptera,
Lycaenidae): Island biogeography, patch dynamics, and the design of habitat
preserves. University of California Press, Oakland, California.
Augspurger, C. K., and S. E. Franson. 1993. Consequences for seed distributions of intracrop variation in wing-loading of wind-dispersed species. Vegetatio 107/108:121–
132.
Bailey, J. K., R. K. Bangert, J. A. Schweitzer, R. T. T. III, S. M. Shuster, and T. G.
Whitham. 2004. Fractal geometry is heritable in trees. Evolution 58:2100–2102.
Bailey, J. K., J. A. Schweitzer, F. Ùbeda, J. Koricheva, C. J. LeRoy, M. D. Madritch, B.
J. Rehill, R. K. Bangert, D. G. Fischer, G. J. Allan, and T. G. Whitham. 2009.
From genes to ecosystems: A synthesis of the effects of plant genetic factors

!

63!
!

!

!

!

across levels of organization. Philosophical Transactions: Biological Sciences
364:1607–1616.
Bergeron, D. G., and R. E. Leach. 1975, June 17. Seedling tree growing apparatus. U.S.
Patent, Oregon.
Bidwell, T. G., J. R. Weir, and D. M. Engle. 2002. Economics of redcedar control in
forest and range management. Pages 33–41. Government document, Oklahoma
Department of Agriculture, Food and Forestry, Oklahoma City, Oklahoma.
Black, A., M. Williamson, and D. Doane. 2008. Wildland fire use barriers and
facilitators. Fire Management Today 68:10–14.
Bond, W. J., M. Honig, and K. E. Maze. 1999. Seed size and seedling emergence: An
allometric relationship and some ecological implications. Oecologia 120:132–
136.
Bond, W. J., and J. E. Keeley. 2005. Fire as a global “herbivore”: The ecology and
evolution of flammable ecosystems. Trends in Ecology & Evolution 20:387–394.
Bond, W. J., and J. J. Midgley. 1995. Kill thy neighbour: An individualistic argument for
the evolution of flammability. Oikos 73:79–85.
Bowman, D. M. J. S., and F. H. Johnston. 2005. Wildfire smoke, fire management, and
human health. EcoHealth 2:76–80.
Boyd, R. 1999. Indians, fire, and the land in the Pacific Northwest. Oregon State
University Press, Corvallis, Oregon.
Campbell, J. W., J. L. Hanula, and T. A. Waldrop. 2007. Effects of prescribed fire and
fire surrogates on floral visiting insects of the blue ridge province in North
Carolina. Biological Conservation 134:393–404.

!

64!
!

!

!

!

Carroll, S. P., A. P. Hendry, D. N. Reznick, and C. W. Fox. 2007. Editorial: Evolution on
ecological time-scales. Functional Ecology 21:387–393.
Chiwocha, S. D. S., K. W. Dixon, G. R. Flematti, E. L. Ghisalberti, D. J. Merritt, D. C.
Nelson, J.-A. M. Riseborough, S. M. Smith, and J. C. Stevens. 2009. Karrikins: A
new family of plant growth regulators in smoke. Plant Science 177:252–256.
Clark, C. M., E. E. Cleland, S. L. Collins, J. E. Fargione, L. Gough, K. L. Gross, S. C.
Pennings, K. N. Suding, and J. B. Grace. 2007. Environmental and plant
community determinants of species loss following nitrogen enrichment. Ecology
Letters 10:596–607.
Clay, K., and J. Holah. 1999. Fungal endophyte symbiosis and plant diversity in
successional fields. Science 285:1742–1744.
Collins, S. L. 1989. Experimental analysis of patch dynamics and community
heterogeneity in tallgrass prairie. Vegetatio 85:57–66.
Collins, S. L. 2000. Disturbance frequency and community stability in native tallgrass
prairie. The American Naturalist 155:311–325.
Collins, S. L., and D. J. Gibson. 1990. Effects of fire on community structure in tallgrass
and mixed-grass prairie. Pages 81–98 in S. L. Collins and L. L. Wallace, editors.
Fire in North American tallgrass prairies. University of Oklahoma Press, Norman,
Oklahoma.
Cottrell, H. J. 1947. Tetrazolium salt as a seed germination indicator. Nature 159:748.
Crawford, R. C., and H. Hall. 1997. Changes in the south Puget prairie landscape. Pages
11–15 in P. Dunn and K. Ewing, editors. Ecology and Conservation of the South
Puget Sound Prairie Landscape. The Nature Conservancy, Seattle, Washington.

!

65!
!

!

!

!

Dennehy, C., E. R. Alverson, H. E. Anderson, D. R. Clements, R. Gilbert, and T. N.
Kaye. 2011. Management strategies for invasive plants in Pacific Northwest
prairies, savannas, and oak woodlands. Northwest Science 85:329–351.
Ditomaso, J. M., M. L. Brooks, E. B. Allen, R. Minnich, P. M. Rice, and G. B. Kyser.
2006. Control of invasive weeds with prescribed burning. Weed Technology
20:535–548.
Donohue, K. 2009. Completing the cycle: Maternal effects as the missing link in plant
life histories. Philosophical Transactions of the Royal Society B: Biological
Sciences 364:1059–1074.
Dufrene, M., and P. Legendre. 1997. Species assemblages and indicator species: The
need for a flexible asymmetrical approach. Ecological Monographs 67:345–366.
Dunwiddie, P. W., and J. D. Bakker. 2011. The future of restoration and management of
prairie-oak ecosystems in the Pacific Northwest. Northwest Science 85:83–92.
Ellison, R. L., and J. N. Thompson. 1987. Variation in seed and seedling size: The effects
of seed herbivores on Lomatium grayi (Umbelliferae). Oikos 49:269–280.
Ewing, A. L., and D. M. Engle. 1988. Effects of late summer fire on tallgrass prairie
microclimate and community composition. American Midland Naturalist
120:212–223.
Fazzino, L., H. E. Kirkpatrick, and C. Fimbel. 2011. Comparison of hand-pollinated and
naturally-pollinated Puget balsamroot (Balsamorhiza deltoidea Nutt.) to
determine pollinator limitations on South Puget Sound lowland prairies.
Northwest Science 85:352–360.

!

66!
!

!

!

!

Fenner, M. 1991. The effects of the parent environment on seed germinability. Seed
Science Research 1:75–84.
Fuhlendorf, S. D., W. C. Harrell, D. M. Engle, R. G. Hamilton, C. A. Davis, and D. M. L.
Jr. 2006. Should heterogeneity be the basis for conservation? Grassland bird
response to fire and grazing. Ecological Applications 16:1706–1716.
Gallai, N., J.-M. Salles, J. Settele, and B. E. Vaissière. 2009. Economic valuation of the
vulnerability of world agriculture confronted with pollinator decline. Ecological
Economics 68:810–821.
Hamman, S. T., P. W. Dunwiddie, J. L. Nuckols, and M. McKinley. 2011. Fire as a
restoration tool in Pacific Northwest prairies and oak woodlands: Challenges,
successes, and future directions. Northwest Science 85:317–328.
Hanley, M. E., J. E. Unna, and B. Darvill. 2003. Seed size and germination response: A
relationship for fire-following plant species exposed to thermal shock. Oecologia
134:18–22.
Haubensak, K. A., and I. M. Parker. 2004. Soil changes accompanying invasion of the
exotic shrub Cytisus scoparius in glacial outwash prairies of western Washington
[USA]. Plant Ecology 175:71–79.
Hesseln, H., and A. H. Berry. 2008. The effect of the wildland-urban interface on
prescribed burning costs in the pacific northwestern United States. Proceedings of
the Second International Symposium on fire Economics, Planning, and Policy: A
Global View. PSW-GTR-208. USDA Forest Service, Pacific Southwest Research
Stn., Albany, CA.

!

67!
!

!

!

!

Hitchcock, L. C., and A. Cronquist. 1973. Flora of the Pacific Northwest. University of
Washington Press, Seattle, Washington.
Hoffmann, W. A., and A. C. Franco. 2003. Comparative growth analysis of tropical
forest and savanna woody plants using phylogenetically independent contrasts.
Journal of Ecology 91:475–484.
Howe, H. F. 1995. Succession and fire season in experimental prairie plantings. Ecology
76:1917–1925.
Husby, J. F. 2012. Pollinators may not limit native seed viability for Puget lowland
prairie restoration. Master’s thesis, The Evergreen State College.
Jain, N., V. Soós, E. Balazs, and J. Staden. 2008. Changes in cellular macromolecules
(DNA, RNA and protein) during seed germination in tomato, following the use of
a butenolide, isolated from plant-derived smoke. Plant Growth Regulation 54.
Jefferson, L. V., M. Pennacchio, K. Havens, B. Forsberg, D. Sollenberger, and J. Ault.
2008. Ex situ germination responses of midwestern USA prairie species to plantderived smoke. American Midland Naturalist 159:251–256.
Joint Base Lewis-McChord. 2012, November. Orientation briefing, Nov 2012.
Powerpoint presentation.
Joint Base Lewis-McChord Army Compatible Use Buffer program fact sheet. (n.d.). U.S.
Army Environmental Command.
Kaye, T. N., K. L. Pendergrass, K. Finley, and J. B. Kauffman. 2001. The effect of fire on
the population viability of an endangered prairie plant. Ecological Applications
11:1366–1380.

!

68!
!

!

!

!

Keeley, J. E., and C. J. Fotheringham. 2000. Role of fire in regeneration from seed. Page
410 in M. Fenner, editor. Seeds: The ecology of regeneration in plant
communities. CABI Publishing, Wallingford, UK.
Knops, J. M. H. 2006. Fire does not alter vegetation in infertile prairie. Oecologia
150:477–483.
Lacey, E. R. 1998. What is an adaptive environmentally induced parental effect? Pages
54–66 in T. A. Mousseau and C. W. Fox, editors. Maternal Effects as
Adaptations. Oxford University Press, New York, New York.
Lebuhn, G., S. Droege, E. F. Connor, B. Gemmill-Herren, S. G. Potts, R. L. Minckley, T.
Griswold, R. Jean, E. Kula, D. W. Roubik, J. Cane, K. W. Wright, G. Frankie, and
F. Parker. 2013. Detecting insect pollinator declines on regional and global scales.
Conservation Biology 27:113–120.
LeRoy, C. J., K. Bush, J. Trivett, and B. Gallagher. 2012. The Sustainability in Prisons
Project: An overview (2004-2012). Sustainability in Prisons Project, The
Evergreen State College, Olympia, Washington.
Luzuriaga, A. L., A. Escudero, and F. Pérez-García. 2006. Environmental maternal
effects on seed morphology and germination in Sinapis arvensis (Cruciferae).
Weed Research 46:163–174.
McCune, B., and J. B. Grace. 2002. Analysis of ecological communities. MjM Software
Design, Gleneden Beach, OR.
McGranahan, D. A., D. M. Engle, S. D. Fuhlendorf, J. R. Miller, and D. M. Debinski.
2012. An invasive cool-season grass complicates prescribed fire management in a
native warm-season grassland. Natural Areas Journal 32:208–214.

!

69!
!

!

!

!

McNair, J. N., A. Sunkara, and D. Frobish. 2012. How to analyse seed germination data
using statistical time-to-event analysis: Non-parametric and semi-parametric
methods. Seed Science Research 22:77–95.
Millar, C. I., N. L. Stephenson, and S. L. Stephens. 2007. Climate change and forests of
the future: Managing in the face of uncertainty. Ecological applications 17:2145–
2151.
Mitchell, R. M., and J. D. Bakker. 2011. Carbon addition as a technique for controlling
exotic species in Pacific Northwest prairies. Northwest Science 85:247–254.
Moles, A. T., D. D. Ackerly, C. O. Webb, J. C. Tweddle, J. B. Dickie, A. J. Pitman, M.
Westoby, and C. B. Field. 2005. Factors that shape seed mass evolution.
Proceedings of the National Academy of Sciences of the United States of America
102:10540–10544.
Mote, P. W., and E. P. J. Salathé. 2009. Future climate in the Pacific Northwest.
University of Washington College of the Environment Climate Impacts Group,
Seattle, Washington.
Murray, D. R. 1987. Seed dispersal. Academic Press, Waltham, Massachusetts.
National Institute of Invasive Species Science. (n.d.). Modified Whittaker plot.
Educational. http://ibistest.nrel.colostate.edu/cwis438/websites/niiss/FieldMethods/ModWhit.php?WebS
iteID=1.
Nelson, D. C., G. R. Flematti, E. L. Ghisalberti, K. W. Dixon, and S. M. Smith. 2012.
Regulation of seed germination and seedling growth by chemical signals from
burning vegetation. Annual Review of Plant Biology 63:107–130.

!

70!
!

!

!

!

Van Nuland, M. E., E. N. Haag, J. A. M. Bryant, Q. D. Read, R. N. Klein, M. J. Douglas,
C. E. Gorman, T. D. Greenwell, M. W. Busby, J. Collins, J. T. LeRoy, G.
Schuchmann, J. A. Schweitzer, and J. K. Bailey. 2013. Fire promotes pollinator
visitation: Implications for ameliorating declines of pollination services. PLoS
ONE 8:e79853.
Pendergrass, K. 2010. Introduction to Cook’s lomatium, a federally-listed endangered
species, and a key and photo guide to the Lomatium species that occur within its
range. Page 23. Technical Notes, U.S. Department of Agriculture, Portland,
Oregon.
Pendergrass, K. L., P. M. Miller, J. B. Kauffman, and T. N. Kaye. 1999. The role of
prescribed burning in maintenance of an endangered plant species, Lomatium
bradshawii. Ecological Applications 9:1420–1429.
Pielou, E. C. 1991. After the Ice Age: The return of life to glaciated North America.
University of Chicago Press, Chicago, Illinois.
Pollak, O., and T. Kan. 1998. The use of prescribed fire to control invasive exotic weeds
at Jepson Prairie Preserve. Pages 241–249 Ecology, conservation, and
management of vernal pool ecosystems: Proceedings from a 1996 conference.
California Native Plant Society, University of California, California.
Potts, S. G., B. Vulliamy, S. Roberts, C. O’Toole, A. Dafni, G. Ne’eman, and P. Willmer.
2005. Role of nesting resources in organising diverse bee communities in a
Mediterranean landscape. Ecological Entomology 30:78–85.

!

71!
!

!

!

!

Rice, J. S., and L. L. Remy. 1998. Impact of horticultural therapy on psychosocial
functioning among urban jail inmates. Journal of Offender Rehabilitation 26:169–
191.
Richards, H. J., and D. M. Kafami. 1999. Impact of horticultural therapy on vulnerability
and resistance to substance abuse among incarcerated offenders. Journal of
Offender Rehabilitation 29:183–193.
Risser, P. G. 1990. Landscape processes and the vegetation of the North American
grassland. Pages 133–146 in S. L. Collins and L. L. Wallace, editors. Fire in
North American tallgrass prairies. University of Oklahoma Press, Norman,
Oklahoma.
Rita, P., M. Aninda, and D. K. Animesh. 2012. Cumin (Cuminum cyminum L.;
Umbelliferae) cultivation in West Bengal Plains, Kalyani, Nadia. International
Research Journal of Pharmacy 3.
Sakai, S. 1995. A model for seed size variation among plants. Evolutionary Ecology
9:495–507.
Sakai, S., and A. Sakai. 1995. Flower size-dependent variation in seed size: Theory and a
test. The American Naturalist 145:918–934.
Sakai, S., and A. Sakai. 1996. Why is there variation in mean seed size among plants
within single populations? Test of the fertilization efficiency hypothesis.
American Journal of Botany 83:1454–1457.
Schmid, B., and C. Dolt. 1994. Effects of maternal and paternal environment and
genotype on offspring phenotype in Solidago altissima L. Evolution 48:1525–
1549.

!

72!
!

!

!

!

Schoeneberger, P. J., D. A. Wysocki, E. C. Benham, and W. D. Broderson (Eds.). 2002.
Field book for describing and sampling soils. 2.0 edition. Natural Resources
Conservation Service, National Soil Survey Center, Lincoln, Nebraska.
Seastedt, T. R. 1995. Soil systems and nutrient cycles of the North American prairie. in
A. Joern and K. H. Keeler, editors. The changing prairie: North American
grasslands. Oxford University Press, New York, New York.
Shackelford, N., R. J. Hobbs, N. E. Heller, L. M. Hallett, and T. R. Seastedt. 2013.
Finding a middle-ground: The native/non-native debate. Biological Conservation
158:55–62.
Shepard, J. H. 2000. Status of five butterflies and skippers in British Columbia. British
Columbia Ministry of Environment, Lands, and Parks, Wildlife Branch;
Resources Inventory Branch, Victoria, BC.
Smith, F. E. 1951. Tetrazolium salt. Science 113:751–754.
Soós, V., E. Sebestyén, A. Juhász, J. Pintér, M. Light, J. Staden, and E. Balázs. 2009.
Stress-related genes define essential steps in the response of maize seedlings to
smoke-water. Functional & Integrative Genomics 9.
Sprenger, C. B., and P. W. Dunwiddie. 2011. Fire history of a Douglas-fir-Oregon white
oak woodland, Waldron Island, Washington. Northwest Science 85:108–119.
Stohlgren, T. J., M. B. Falkner, and L. D. Schell. 1995. A Modified-Whittaker nested
vegetation sampling method. Vegetatio 117:113–121.
Suding, K. N., S. L. Collins, L. Gough, C. Clark, E. E. Cleland, K. L. Gross, D. G.
Milchunas, S. Pennings, and F. S. Chapin III. 2005. Functional- and abundance-

!

73!
!

!

!

!

based mechanisms explain diversity loss due to N fertilization. Proceedings of the
National Academy of Sciences of the United States of America 102:4387–4392.
Tackenberg, O., P. Poschlod, and S. Kahmen. 2003. Dandelion seed dispersal: The
horizontal wind speed does not matter for long-distance dispersal - it is updraft!
Plant Biology 5:451–454.
Tavşanoğlu, Ç., and Ş. Serter Çatav. 2012. Seed size explains within-population
variability in post-fire germination of Cistus salviifolius. Annales Botanici Fennici
49:331–340.
Thomas, T. H., N. L. Biddington, and D. F. O’Toole. 1979. Relationship between
position on the parent plant and dormancy characteristics of seeds of three
cultivars of celery (Apium graveolens). Physiologia Plantarum 45:492–496.
Thomas, T. H., and J. V. Staden. 1995. Dormancy break of celery (Apium graveolens L.)
seeds by plant derived smoke extract. Plant Growth Regulation 17:195–198.
Thompson, J. N. 1984. Variation among individual seed masses in Lomatium grayi
(Umbelliferae) under controlled conditions: Magnitude and partitioning of the
variance. Ecology 65:626–631.
Thompson, J. N., and O. Pellmyr. 1989. Origins of variance in seed number and mass:
Interaction of sex expression and herbivory in Lomatium salmoniflorum.
Oecologia 79:395–402.
Thorpe, J. 2011. Vulnerability of prairie grasslands to climate change. Saskatchewan
Research Council.
Thorpe, J. 2012. Adaptation to climate change in management of prairie grasslands.
Saskatchewan Research Council.

!

74!
!

!

!

!

Tveten, R. K., and R. W. Fonda. 1999. Fire effects on prairies and oak woodlands on Fort
Lewis, Washington. Northwest Science 73:145–158.
USDA NRCS National Plant Data Team. (n.d.). Plants profile for Lomatium cookii (agate
desert parsley). United States Department of Agriculture Natural Resources
Conservation Service.
Vogl, R. J. 1974. Effects of fire on grasslands. Pages 139–194 in T. T. Kozlowski, editor.
Fire and Ecosystems. Academic Press, Waltham, Massachusetts.
White, R. 1975. Indian land use and environmental change: Island County, Washington:
A case study. Arizona and the West 17:327–338.
Witty, M. 2012. Topographischer nachweis der keimfähigkeit der getreidefrüchte durch
tetrazoliumsalze (topographic detection of germination in cereal crops by
tetrazolium salts) — A translation of Lakon’s 1942 paper on tetrazolium seed
testing. Seed Technology 34:275–282.
Wratten, S. D., M. Gillespie, A. Decourtye, E. Mader, and N. Desneux. 2012. Pollinator
habitat enhancement: Benefits to other ecosystem services. Agriculture,
Ecosystems & Environment 159:112–122.
Yi-Fang Chou, R. D. Cox, and D. B. Wester. 2012. Smoke water and heat shock
influence germination of shortgrass prairie species. Rangeland Ecology &
Management 65:260–267.
Yoder, J. 2004. Playing with fire: Endogenous risk in resource management. American
Journal of Agricultural Economics 86:933–948.
Zulauf, A. S., and Washington Agricultural Experiment Station. 1979. Soil survey of
Pierce County area, Washington. The Service: [U.S. G.P.O], Washington.

!

75!
!

Item sets

The Evergreen State College Masters Theses: Environmental Studies

Media

Mann_JMESthesis2014.pdf