Citation
From extirpation to recolonization

Material Information

Title:
From extirpation to recolonization
Alternate title:
Bubonic plague mitigates the effects of prairie dogs (cynomys ludovicianus) on plant communities
Creator:
Visel, Mark David ( author )
Language:
English
Physical Description:
1 electronic file (66 pages) : ;

Subjects

Subjects / Keywords:
Plague ( lcsh )
Prairie dogs ( lcsh )
Prairie conservation ( lcsh )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Review:
Plague, caused by Yersinia pestis, was introduced to the western United States in the 1940s. Black-tailed prairie dogs (Cynomys ludovicianus) are highly susceptible to the disease and plague has become a significant disturbance that results in periodic epizootics on prairie dog colonies. While there have been many studies of the effects of prairie dogs on vegetation, most have been conducted in rural rather than urban areas and most have not been related to plague. Examining vegetation on prairie dog colonies before and after plague in urban and suburban areas can have critical management implications because, in urban areas, other global change factors such as urbanization, climate warming, nitrogen deposition, and propagule pressure from non-native plant species interact with prairie dogs to result in novel plant communities. Urban prairie dog colonies differ significantly from rural colonies by having higher density of prairie dogs, higher richness and cover of exotic vegetation, and higher percentages of bare ground. This study evaluated whether periodic plague epizootics mitigate the effects prairie dog have on vegetation in an urban environment, Boulder, Colorado. Using point-intercept methods we examined the effects of plague epizootics on colony vegetation. Main findings are that 1) prairie dogs reduce native species richness and cover and increase bare ground, but recovery after seven years of abandonment is seen and 2) historical tillage has a large effect on the landscape as well, with tilled sites showing lower native species richness and cover. Our work has significant implications for management of lands at the urban wild-land interface and contributes to broader theories about how multiple change drivers can interact in ecosystems.
Thesis:
Thesis (M.S.) - University of Colorado Denver
Bibliography:
Includes bibliographic references
System Details:
System requirements: Adobe Reader.
General Note:
Department of Integrative Biology
Statement of Responsibility:
by Mark David Visel.

Record Information

Source Institution:
|University of Colorado Denver
Holding Location:
|Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
945932182 ( OCLC )
ocn945932182
Classification:
LD1193.L45 2015m V57 ( lcc )

Downloads

This item has the following downloads:


Full Text
FROM EXTIRPATION TO RECOLONIZATION: BUBONIC PLAGUE MITIGATES THE EFFECTS
OF PRAIRIE DOGS (CYNOMYS LUDOVICIANUS) ON PLANT COMMUNITIES
by
MARK DAVID VISEL
B.A., Bethel University, 2011
B.S., Bethel University, 2011
A thesis submitted to the
Faculty of the Graduate School of the
University of Colorado in partial fulfillment
of the requirements for the degree of
Master of Science
Integrative Biology
2015


This thesis for the Master of Science degree by
Mark David Visel
has been approved for the
Department of Integrative Biology
by
Laurel Hartley, Chair
Timothy Seastedt
Michael Wunder
November 20th 2015


Visel, Mark David (M.S., Biology)
From Extirpation to Recolonization: Bubonic Plague Mitigates the Effects of Prairie Dogs
(Cynomys ludovicianus) on Plant Communities.
Thesis directed by Associate Professor Laurel M. Hartley
ABSTRACT
Plague, caused by Yersinia pestis, was introduced to the western United States in the
1940s. Black-tailed prairie dogs (Cynomys ludovicianus) are highly susceptible to the
disease and plague has become a significant disturbance that results in periodic
epizootics on prairie dog colonies. While there have been many studies of the effects of
prairie dogs on vegetation, most have been conducted in rural rather than urban areas
and most have not been related to plague. Examining vegetation on prairie dog colonies
before and after plague in urban and suburban areas can have critical management
implications because, in urban areas, other global change factors such as urbanization,
climate warming, nitrogen deposition, and propagule pressure from non-native plant
species interact with prairie dogs to result in novel plant communities. Urban prairie dog
colonies differ significantly from rural colonies by having higher density of prairie dogs,
higher richness and cover of exotic vegetation, and higher percentages of bare ground.
This study evaluated whether periodic plague epizootics mitigate the effects prairie dog
have on vegetation in an urban environment, Boulder, Colorado. Using point-intercept
methods we examined the effects of plague epizootics on colony vegetation. Main
findings are that 1) prairie dogs reduce native species richness and cover and increase
bare ground, but recovery after seven years of abandonment is seen and 2) historical


tillage has a large effect on the landscape as well, with tilled sites showing lower native
species richness and cover. Our work has significant implications for management of
lands at the urban wild-land interface and contributes to broader theories about how
multiple change drivers can interact in ecosystems.
The form and content of this abstract are approved. I recommend its publication.
Approved: Laurel Hartley
IV


ACKNOWLEDGEMENTS
I thank Lynn Riedel and Heather Swanson from Boulder Open Space and
Mountain Parks for assistance with permits, historical data, and plant identification. I
would also like to thank David Buckner and ESCO Associates Inc. for loaning us the
optical-point projection device and teaching us to use it and the method. I also thank
Stower Beals for his assistance with non-metric multidimensional scaling analyses and
interpretation. I also thank my committee members, Timothy Seastedt and Michael
Wunder for their assistance in getting me started and seeing me through. Lastly, I would
like to thank my advisor, Laurel Hartley, for her guidance, insight, and much help that
she poured into this project. This project was funded by a grant from the National
Science Foundation (Ecosystem Science #1120390). This study complies with current
laws of the U.S.
v


CONTENTS
CHAPTER
I. INTRODUCTION................................................................1
General Effects of Prairie Dogs on Plant Communities.........................2
Effects of Prairie Dogs on Plant Communities in Urban Environments...........3
The Role of Plague in Mitigating the Effects of Prairie Dogs on Plant Communities.4
II. METHODS....................................................................7
Study Area...................................................................7
Field Methods................................................................8
Data Analysis...............................................................10
III. RESULTS..................................................................13
Species Diversity and Evenness..............................................13
Species Richness............................................................14
Plant Cover.................................................................21
Litter and Bare Ground Cover................................................30
Notable Species.............................................................30
Plant Community Similarity Based on Nonmetric Multidimensional Scaling......36
IV. DISCUSSION................................................................37
Influence of Historical Tillage.............................................38
Relationship between Tillage and Prairie Dog Occupation History.............39
Relationship of Our Findings to Previous Work...............................41
Effects of Prairie Dogs and Tillage and Plague on Specific Plant Species....41
Management Implications.....................................................43
REFERENCES....................................................................47
APPENDIX......................................................................53
VI


LIST OF FIGURES
Figure
1: Map of Prairie Dog Colony Extent...........................................10
2: Plant Richness, Diversity, Evenness 2013-2014............................. 15
3: Graminoid Richness 2013-2014...............................................17
4: Forb Richness 2013-2014................................................... 18
5:Forb Richness 2013..........................................................19
6: Forb Richness 2014.........................................................20
7: Graminoid Cover 2013-2014................................................. 23
8: Graminoid Cover 2013.......................................................24
9: Graminoid Cover 2014.......................................................25
10: Forb Cover 2013-2014......................................................26
11: Forb Cover 2013...........................................................27
12: Forb Cover 2014...........................................................28
13: Other Cover 2013-2014.....................................................29
14: NMDS Community Composition Plots..........................................37
15: Bromus after Flood........................................................45
vii


LIST OF TABLES
Table
1: Mean Notable Graminoid Species Cover...................................32
2: Mean Notable Forb Species Cover........................................33
3: Percentage Replicates of Notable Forb Species..........................34
4: Percentage Replicates of Notable Graminoid Species.....................35
Al: MANOVA output for Evenness & Diversity................................53
A2: MANOVA output for Richness............................................54
A3: MANOVA output for Cover...............................................55
A4: ANOVA for Graminoid Cover.............................................56
A5: ANOVA for Forb Cover..................................................57
A6: ANOVA for Other Cover.................................................58
viii


CHAPTER I
INTRODUCTION
Increasing human impact on ecosystems has resulted in some ecosystems being
subject to the effects of multiple global change drivers in concert with often
unpredictable results (Berg et al. 2015, Beals et al. 2014, Hobbs et al. 2014, Prevey and
Seastedt 2014, Spasojevic et al. 2013, Vitousek et al. 1997). Our research focuses on
how the effects of a native keystone herbivore on plant communities are impacted by
both by introduced disease and urbanization. Black-tailed prairie dogs (Cynomys
ludovicianus) are semi-fossorial rodents found throughout the Great Plains of North
America and have been considered a keystone species and an ecosystem engineer
(Forrest 2005, Vermeire et al. 2004, Bangert & Slobodchikoff 2000, Miller et. al. 2000).
Prairie dogs have also been regarded as competitors with livestock, as they forage on
similar species (Mellado et. al. 2005, Cid et al. 1991). While prairie dogs are found
throughout much of their historic range and the total acreage occupied by prairie dogs
has been severely decreased (Miller & Cully 2001), they are a species than can tolerate
urbanization. Prairie dog colonies can be found occupying grassland remnants within
suburban and urban landscapes throughout the Front Range of Colorado. It is well
known that prairie dogs alter the vegetation structure and plant community
composition in their colonies, with the majority of research being conducted in rural
areas (Fahnestock et al. 2003, Whicker & Detling 1988, Coppock et al. 1983). The effects
of prairie dogs on vegetation vary among grassland types. The general effects of prairie
1


dogs on vegetation in mixed grass prairie are less dramatic than in tall grass prairie and
even less dramatic on short grass prairie (Hartley et al. 2009, Cid et al. 1991, Coppock et
al. 1983, Osborn & Allan 1949). We know that the effects of prairie dogs on plant
communities can be influenced by introduced plague (Hartley et al. 2009, Augustine et
al. 2007) and by urbanization (Beals et al. 2014, Magle et al. 2010, Magle & Crooks
2008). Our study examines the effects of plague epizootics on prairie dog colony plant
communities in an urban setting, Boulder County, Colorado.
General Effects of Prairie Dogs on Plant Communities
In the mixed grass prairie of Badlands National Park, South Dakota, vegetation
on black-tailed prairie dog colonies is dominated by forbs, whereas off of the colonies
vegetation is dominated by graminoids (Fahnestock et al. 2003). Rural short grass
steppe follows the same pattern of vegetation change after prairie dog colonization
(Hartley et al. 2009, Bonham & Lerwick 1976). On colonies that have been grazed
heavily by prairie dogs for several years, species richness may be limited to a handful of
forbs or small shrubs throughout the colony (Whicker & Detling 1988). Richness of
species is the greatest on young colonies (2-3 years) (Whicker & Detling 1988; Coppock
et al. 1983). Prairie dogs typically do not occupy tall grass prairies, although Osborn and
Allan (1949) observed seven different stages of plant succession in and around an active
prairie dog colony in Oklahoma.
2


Effects of Prairie Dogs on Plant Communities in Urban Environments
As in rural environments, urban environments in Colorado show decreased
graminoids and increased forbs with prairie dog occupation (Beals et al. 2014, Magle &
Crooks 2008). Plant height is also shorter on than off active colonies in an urban
environment (Magle & Crooks 2008). In Boulder County, Colorado, prairie dog
occupation results in a significant reduction of perennial plants on colonies and an
increase in introduced plant species (Beals et al. 2014). Urban and rural prairie dog
colonies also show an increase in bare ground cover compared to sites unoccupied by
prairie dogs (Magle & Crooks 2008, Hartley et al. unpublished). This and research in
rural areas (e.g. Hartley et al. 2009) suggests a release from grazing pressure should
result in a decrease in bare ground cover.
Invasive plant encroachment and soil erosion are significant land management
issues in urban and suburban Colorado. Prairie dogs profoundly change vegetation in
their colonies as well as increase bare ground cover and this may significantly contribute
to soil erosion (Munson et al. 2011). Prairie dog colonies in urban areas also have more
than double the amount of disturbed soil per 1000 m2, indicating a higher density of
prairie dogs and increased erosion. (Hartley, unpublished). Prairie dogs may also
contribute to invasive plant establishment by selective consumption, resulting in an
alteration of competitive dynamics. Urban colonies have almost three and a halftimes
the cover of non-native plant species (Beals et al. 2014).
3


The Role of Plague in Mitigating the Effects of Prairie Dogs on Plant Communities
Within the last 70 years, prairie dogs have been faced with a new challenge:
plague. Plague is caused by a flea-borne bacterium (Yersinia pestis). Plague is extremely
detrimental to prairie dog populations, as it kills all or nearly all prairie dogs within an
infected colony (Pauli et al. 2006, Cully and Williams 2001). Plague can also spread from
direct animal to animal contact, whether that is prairie dog to prairie dog or a predator
or small rodent to prairie dog (Cully et al. 2010). Prairie dog colony complexes that have
been affected by plague have been severely decreased in area, which results in greater
distances between colonies. These greater distances between colonies result in greater
isolation, which lowers the probability of being recolonized after a plague event while
also lowering the probability of a plague epizootic occurring again (Cully et al. 2010).
Urban environments include many more barriers, such as roads and buildings, when
compared rural environments, which restrict prairie dog movements. These urban
barriers are also negatively associated with plague occurrence (Collinge et al. 2005).
Despite these barriers to plague, Friggens and Beier (2010) reported that anthropogenic
disturbances were associated with an increase in flea-borne diseases.
Plague also creates opportunities for vegetation changes on prairie dog
colonies. In studying the effect of plague on rural prairie dog colonies, it was observed
that the biomass of both graminoids and forbs increased following a plague epizootic,
although recovery of biomass may be related to colony age at the time of the epizootic
because the degree of impact of prairie dogs on colony vegetation increased with colony
age (Hartley et al. 2009).
4


In the Pawnee National Grassland of Colorado, USA, less than 2% of colonies go
more than 20 years without experiencing a plague epizootic, and colonies, on average,
experience plague epizootics every six and a half years (Hartley et al. 2009). In addition,
the probability that a colony will be recolonized within five years of an epizootic is only
50% (Hartley et al. 2009).
Most of the research dealing with plague, vegetation, prairie dogs, or any
combination of them, has been conducted in rural sites. Urban and suburban habitats
possess very different characteristics and dynamics, such as increased density of prairie
dogs, increased habitat fragmentation, increased migration obstacles, and increased
anthropogenic disturbances, which are important for land managers to consider
(Seastedt et al. 2013, Seastedt et al. 2008). While prairie dogs may be subject to local
extinctions due to development, with proper management they can persist in an
increasingly urbanized landscape (Magle et al. 2010). Climate change is also an
important factor in both vegetation change and plague (Seastedt et al. 2013, Snail et al.
2008). It is yet unclear how climate change will interact with grazing and management
pressures (Coffin et al. 1996). Understanding how these variables affect plant
communities in urban areas is critical to making proper land and prairie dog
management decisions.
Understanding how periodic plague outbreaks mitigate vegetation change
caused by prairie dogs colonization has important management considerations (e.g.
should prairie dogs be removed from some areas?, should plague be controlled?, what
should management be between plague cycles?). Currently we only have pre-post
5


plague vegetation data from rural colonies that don't have the high densityies of prairie
dogs or high non-native plant propagule pressure that we see in urban and suburban
colonies.
Prairie dogs in Boulder County are accelerating the transformation of grasslands
into a 'novel ecosystem' in their urban environments with increased bare soil well
beyond that seen in rural colonies, increased encroachment by invasive plants, and a
detrimental decrease in native grasses (Beals et al. 2014). We don't currently
understand the role that introduced plague plays in shaping plant communities in urban
areas. It could potentially mitigate prairie-dog induced changes to vegetation as is seen
in rural areas (Hartley et al. 2009). In an effort to understand how plague influences the
role of prairie dogs in urban areas, we studied plant communities on prairie dog colonies
impacted by a plague epizootic in 2006 that extirpated more than 50 prairie dog
colonies in the urban Front Range of Colorado, USA. The goals of our work were
investigate plant community responses to prairie dog occupation and subsequent prairie
dog extirpation due to plague. Further, we were interested in whether a history of
previous tillage would influence those plant community responses.
6


CHAPTER II
METHODS
Study Area
This study was conducted during the months of May-August 2013 and 2014 on
Boulder Open Space and Mountain Parks (BOSMP) public land in Boulder, Colorado, USA
(40.0176 N, 105.2797 W, mean elevation of 1640m). Mean annual precipitation is 486
mm. Precipitation in 2012 was 398 mm. Precipitation in 2013 was a record 867 mm,
with 461 mm falling in a matter of days in September (Boulder Climate Data, 2015).
May-August precipitation was 145 mm in 2013 and 290 mm in 2014 (Boulder Climate
Data, 2015). At all study sites, soils consist of mostly clay loam to sandy clay loam, at 0
to 12 percent slopes (BOSMP, unpublished data). Vegetation is primarily a mixed-grass
prairie mosaic, with some properties used for grazing cattle. During the study period,
there was no cattle grazing on the prairie dog colonies. Despite the agricultural usage,
the sites were located in a variable landscape with a range of urban factors (Beals et al.
2014).
We chose study sites from among the 151 properties containing prairie dogs on
BOSMP land. BOSMP has a total of 410 colonies ranging in size from less than .01 ha to
215 ha with a mean area of 4.5 ha. BOSMP maps active prairie dog colonies each year
and has done so since 1991. Because we were interested in vegetation change after
plague epizootics, only colonies that were extirpated during a wide-spread plague event
7


in 2006 were considered. Furthermore, we avoided sites that were closed because of
nesting threatened or endangered birds and extremely small colonies (less than 4
square ha). Of the 30 sites selected based on these criteria, 13 sites had never been
tilled, 17 had been tilled historically ranging from 20 to 30+ years ago (BOSMP,
unpublished data). From the BOSMP digital maps of prairie dog colonies, we could
determine prairie dog occupation history. Using Google Earth (2014), we divided the
colony into polygons based on prairie dog occupation history and established a 50 x 2 m
sampling transect within each polygon. One transect was located in an area of the
colony that was recolonized by prairie dogs in 2007, less than one year after the plague
epizootic. A second transect was in an area recolonized in 2010, about three years after
the epizootic. A third transect was in an area that was still uncolonized by prairie dogs,
but had been active before the plague event in 2006, allowing 6-7 years of vegetation
recovery. The fourth transect was established in an area that had never been colonized
by prairie dogs. The never colonized by prairie dog transects showed similar soil type,
slope, and degree of urbanization when compared with the sites that had been
previously occupied by prairie dogs. All of the prairie dog colonies in our study had been
active for approximately 4-7 years prior to the plague epizootic.
Field Methods
To estimate plant cover by species, we used a Cover-Point Optical Point
Projection Device (ESCO Associates, Boulder, Colorado, USA) which allows the standing
operator to visualize the species located in the crosshairs of a low-powered optical
8


scope located above and perpendicular to the ground (Norland et al. 1993). This method
was chosen because is relatively easy to set-up, affording us time to sample all of the
prairie dog colonies that met our criteria for inclusion in the study. In addition, this
method can be used for long term monitoring because it provides accurate and
repeatable data (Buckner 1985, Kinsinger 1960, Johnston 1957) and is currently the
method that BOSMP uses for the monitoring of vegetation transects. We used an
orange spray-painted rebar stake to mark the beginning of each transect and recorded
the GPS (Global Positioning System) location of the stake so that we could locate the
transect for the duration of the study. The coordinate direction of the transect was
randomly assigned. Data were collected at points 0.5 m to the left and 0.5 m to the right
of the transect tape (50 x 2 m) at 1 m intervals for a total of 100 points per transect. We
recorded whether the crosshair was above bare soil, litter, rock, standing dead
vegetation, or a live plant species. These crosshair "hits" were used to estimate cover of
plants by species, litter cover, and bare ground. We calculated richness, evenness, and
Shannon-Wiener diversity values using all data from the transect sampling, which
included species noted during the point sampling and species present in the 50 X 2 m
area but not captured during the point sampling. Vegetation was conventionally
classified by names, origins, and functional groups following the nomenclature of the
USDA Plants Database (USDA, NRCS).
Sites were sampled twice per year during the growing season, which allowed us
to capture both early and late season plant species. The first sampling period began in
the end of May and ended in the beginning of June. The second sampling period started
9


in early August and ended in late August. The order in which sites were sampled was
random.
Figure 1: Map of Prairie Dog Colony Extent
Map of prairie dog colony Extent before plague (orange), extent in 2007 (purple)
and extent in 2010 (pink). One transect was located in each of these areas. Our
fourth transect was established outside the orange colony extent.
Data Analysis
Our analysis choices were guided by our overarching goals of contributing to
management of prairie dogs and plague and to relating our work to previous findings in
urban and rural settings, thus we focused primarily on effects of prairie dogs, plague,
and previous tillage on species diversity and evenness, native and introduced graminoid
and forb richness and cover, and on bare ground and litter cover. All analyses were
performed using the statistical program R (R Development Core Team 2011).
We calculated Shannon-Wiener diversity and species evenness for each of our
transects. Species that were present in a transect, but not "hit" during the point
10


sampling were given cover values of 0.001 so that they would be counted, but less than
actual "hits". We used a two way factorial analysis of variance (ANOVA) to determine
differences between tillage and transect type for evenness and diversity.
We conducted comparisons between tillage (untilled and tilled) and site type
(recolonized in 2007, recolonized in 2010, never recolonized, and never colonized) to
examine differences in species diversity, evenness, richness and cover of plants by
functional group (graminoids and forbs) and origin (native and introduced). Because
richness data should encompass the total number of species in an area across the
growing season, we pooled our early and late season data; therefore season was a
factor in our analysis of cover data, but not species richness data. We used a two way
factorial multivariate analysis of variance (MANOVA) since we had four independent
variables for cover (year, season, tillage history, prairie dog occupation history), three
factors for richness (year, tillage history, prairie dog occupation history), and multiple
response variables for cover and richness. Since a MANOVA also takes into account
possible intercorrelations among the response variables, we ran one for all cover related
variables (total plant cover, bare ground cover, litter cover, graminoid cover, native
graminoid cover, introduced graminoid cover, forb cover, introduced forb cover, native
forb cover) and one for all species richness related variables (graminoid richness, native
graminoid richness, introduced graminoid richness, forb richness, introduced forb
richness, native forb richness). These variables have ecological significance for land
managers.
We also chose to examine differences among treatments with respect to the
11


presence and cover of notable species. We considered species to be notable if they
were very common and/or a species of concern to BOSMP land managers (pers.
Communication). These species included introduced forbs (Convolvulus arvensis,
Erodium cicutarium, Carduus nutans, Dipsacus follonum), native forbs (Heterotheca
villosa, Sphaeralcea coccinea, Psoralidium tenuiflorum, Gaura coccinea), introduced
graminoids (Bromus inermis, B. tectorum, B. arvensis, Poa pratensis) and native
graminoids (Pascopyrum smithii, Bouteloua dactyloides, B. gracilis, Aristida purpurea).
To assess notable species we counted the number of times a species was found in a
replicate within each treatment group and calculated the percentage of replicates in
which that species was found. In addition we used ANOVA to investigate whether there
were significant differences among on treatments with respect to cover of Convolvulus
arvensis, Bromus inermis, and Pascopyrum smithii.. The cover of other species was too
small to warrant statistical analysis.
We used non-metric multidimensional scaling (NMDS) based on Bray-Curtis
dissimilarity to examine plant community differences among our treatments
(recolonized in 2007, recolonized in 2010, not recolonized, and never colonized), years
(2013, 2014), and season (early, late). The plots are a 95% confidence interval of the
different treatment communities. We used the Permutational Multivariate Analysis of
Variance (PERMANOVA) function in R to quantify differences among treatments.
12


CHAPTER III
RESULTS
The goals of our work were investigate plant community responses to prairie dog
occupation and subsequent prairie dog extirpation due to plague. Further, we were
interested in whether a history of previous tillage would influence those plant
community responses.
Species Diversity and Evenness
Our MANOVA showed no biologically relevant significant interactions between
year and our other factors (Table Al), therefore we report data averaged across years.
Tillage reduced species evenness across treatments, however only significantly on the
sites that were not occupied by prairie dogs (not recolonized, never colonized) (Fig. 2).
Prairie dogs suppressed the effect of tillage by reducing evenness on active sites (Fig. 2)
On the sites where prairie dogs were active there was not a significant difference
between the 2007 and 2010 recolonized treatments (Fig. 2). Species diversity,
measured by the Shannon-Wiener diversity index, showed similar trends. Prairie dogs
reduced species diversity generally, however tillage also had an effect (Fig. 2). In areas
without active prairie dogs tillage decreased diversity significantly (Fig. 2).
13


Species Richness
Our MANOVA showed no biologically relevant significant interactions between
year and our other factors (Table A2), therefore we report richness data averaged across
years. Both prairie dog occupation and tillage had significant influences on species
richness. Sites that were currently occupied by prairie dogs had lower total plant species
richness than sites that had not been recolonized post plague or never colonized when
tillage was accounted for (Fig. 2). Sites with a history of previous tillage had lower total
plant species richness than sites that had not been tilled (Fig. 2).
14


2013-2014
l/>
0}
u
cu
Q.
LO
M
O
_Q
E
3
>
to
>
Cl
to
to
CU
c
c
cu
>
LU
o
o
fO
o
CM
O
O
LO
o
to
o
to
d
o
o
oo
o
to
o
o
fM
o
o
o
Recolonized
In 2007
Recolonized
Recolonized ln 2010 Recolonized
Not Yet
Recolonized
Not Yet
Never
Colonized Never
In 2007 In 2010 Recolonized Colonized
Figure 2: Plant Richness, Diversity, Evenness 2013-2014
Box plots demonstrating differences in species richness, diversity and
evenness across site types (recolonized in 2007 after one year of extirpation,
recolonized in 2010 after three years of extirpation, not yet recolonized, and
never colonized) and tillage history (tilled and untilled). Data presented are
averages from the 2013 and 2014 early and late growing season sampling
events. Different letters indicate significant differences between treatments
at the P < 0.01 level.
15


Graminoid richness was most influenced by whether or not a colony was
currently occupied by prairie dogs. Sites that had been plague extirpated, but were
reoccupied by prairie dogs in 2007 and 2010 had lower total, native, and introduced
graminoid richness compared to sites that were not currently occupied (not yet
recolonized or never colonized) (Fig. 3). If there was an effect of plague extirpation on
graminoid richness, that effect could no longer be seen at the time of our sampling,
which commenced 7 and 4 years after the sites had been recolonized. Tilled sites without
prairie dogs generally had lower native and higher non-native graminoid richness (Fig. 3).
All treatment types that were disturbed by either prairie dogs or tillage had low richness
(0.981.0) of native graminoids. Introduced graminoid richness was highest on sites not
currently occupied by prairie dogs regardless of tillage history (Fig 3).
Both prairie dog occupation and tillage influenced forb richness. Tillage had the
largest effect on native forb richness, with tilled sites having significantly fewer native
forbs (7.13.5) than untilled sites (13.35.7) (Fig 4). Tilled and untilled transects were not
significantly different with respect to introduced forb richness (Fig 4). Prairie dog
occupation had a small effect on introduced forb richness, with currently occupied sites
having lower introduced forb richness (4.02.4) than unoccupied sites (5.32. 8). We saw
higher numbers of native and introduced forb species across all transects regardless of
tillage history in 2014 than in 2013, which we attribute to the anomalous large rainfall
event in September 2013 (Figs. 5, 6).
16


2013-2014
i/i
a;
'a
0)
a.
to
cu
-Q
Tilled
Untilled
Total Graminoid Richness
to
oj
'<_>
aj
CL
to
aj
_Q
Recolonized Recolonized
ln 2007 n , . . In 2010 _ . . ,
Recolonized Recolonized
Not Yet
Recolonized
In 2007
In 2010
Never
Not Yet Colonized Never
Recolonized Colonized
Box plots demonstrating differences in total, native and introduced graminoid
richness across site types (recolonized in 2007 after one year of extirpation,
recolonized in 2010 after three years of extirpation, not yet recolonized, and
never colonized) and tillage history (tilled and untilled). Data presented are
averages from the 2013 and 2014 early and late growing season sampling
events. Different letters indicate significant differences between treatments
at the P < 0.01 level.
17


Number of Species Number of Species Number of Species
2013-2014
O
m
o
CM
o
o
o
CM
O
O

o
LO
o
Recolonized
In 2007
Recolonized
Recolonized
In 2010
Recolonized
Not Yet
Recolonized
Not Yet
Never
Colonized Never
In 2007 In 2010 Recolonized Colonized
Box plots demonstrating differences in total, native and introduced
forb richness across site types (recolonized in 2007 after one year of
extirpation, recolonized in 2010 after three years of extirpation, not
yet recolonized, and never colonized) and tillage history (tilled and
untilled). Data presented are averages from the 2013 and 2014 early
and late growing season sampling events. Different letters indicate
significant differences between treatments at the P < 0.01 level.
18


Number of Species Number of Species Number of Species
O
rsi
o
O
o
CM
o
O
LH
o
LO
o
Recolonized
In 2007
Recolonized
Recolonized
In 2010
Recolonized
Not Yet
Recolonized
Not Yet
Never
Colonized .
Never
In 2007 In 2010 Recolonized Colonized
Box plots demonstrating differences in total, native and introduced forb
richness across site types (recolonized in 2007 after one year of extirpation,
recolonized in 2010 after three years of extirpation, not yet recolonized,
and never colonized) and tillage history (tilled and untilled). Data presented
are averages from the 2013 early and late growing season sampling events.
Different letters indicate significant differences between treatments at the
P < 0.01 level.
19


Number of Species Number of Species Number of Species
2014
Tilled
Total Forb Richness
O
CO
Recolonized Recolonized
ln2007 q . . . In 2010 _ , . ,
Recolonized Recolonized
Not Yet
Recolonized
In 2007
In 2010
Never
Not Yet Colonized Never
Recolonized Colonized
Box plots demonstrating differences in total, native and introduced forb
richness across site types (recolonized in 2007 after one year of
extirpation, recolonized in 2010 after three years of extirpation, not yet
recolonized, and never colonized) and tillage history (tilled and
untilled). Data presented are averages from the 2014 early and late
growing season sampling events. Different letters indicate significant
differences between treatments at the P < 0.01 level.
20


Plant Cover
Our MANOVA results (Table A3) indicate biologically relevant and statistically
significant interactions between year and our other factors. Therefore, we report plant
cover data for 2013 and 2014 combined and separately. Both prairie dog occupation
and tillage had significant effects on total, native, and introduced graminoid cover. Sites
that were currently occupied had significantly lower total, native and introduced
graminoid cover than unoccupied sites (not yet recolonized and never colonized (Figs. 7-
9). There were no significant differences between the 2007 recolonized and 2010
recolonized sites (Figs. 7-9) indicating that there were not detectable differences
between a one and three year period of release from grazing pressure post plague. Sites
(both tilled and untilled) that had not yet been recolonized post-plague had
approximately eight times the total graminoid cover of the currently occupied sites.
(Figs. 7-9). However tilled sites without prairie dogs showed a higher introduced
graminoid cover (42.618.6%) than untilled sites (19.820.8%) (Fig. 7). Uncolonized and
untilled sites had higher native graminoid cover than any other site type (Figs. 7-9). In
2014, introduced graminoid cover increased slightly when compared to the previous
year, but native graminoid cover was not effected (Table A4) (Figs. 8, 9). In summary,
prairie dog occupation and tilling together greatly reduce native graminoid cover. When
prairie dogs are extirpated, graminoid cover is higher but because of higher introduced
rather than native graminoid cover.
21


Forb cover was significantly affected by whether sites had been tilled or not and
whether sites were currently occupied or not. Introduced forb cover was highest
(38.516.5%) on sites actively grazed by prairie dogs (Fig. 10). Sites that were not
currently colonized showed a significantly lower amount of introduced forb cover
regardless of tillage history (13.712.0%) (Fig 10). Native forb cover, however, was not
affected by current or past prairie dog grazing. Native forb cover was most affected by
tillage with tilled sites having very low native forb cover (1.83.7%) (Fig. 10). In 2014
forb cover showed more pronounced differences between occupied sites and
unoccupied sites than in 2013 (Figs. 11,12). Year had an effect on introduced forb cover
(Table A5), driving up total forb cover, however native forbs showed only a slight
increase.
22


% Cover % Cover % Cover
2013-2014
O
00
O
O
o
rsi
O
O
LD
O
o
ro
O
rsl
O
O
O
00
O

o
,=a*
o
CN
o
Recolonized ln 2010 Recolonized Reco'on'zet* Not Yet Colonized
In 2007 In 2010 Recolonized Colonized
Box plots demonstrating differences in total, native and introduced
graminoid cover across site types (recolonized in 2007 after one year of
extirpation, recolonized in 2010 after three years of extirpation, not yet
recolonized, and never colonized) and tillage history (tilled and untilled).
Data presented are averages from the 2013 and 2014 early and late
growing season sampling events. Different letters indicate significant
differences between treatments at the P < 0.01 level.
23


% Cover % Cover % Cover
2013
O
00
O
KD
O
*3- *
o
CM -
Graminoid Cover
Tilled
Untilled

Native Graminoid Cover
O
o
m
o
CM
o
00
o
o
Recolonized Recolonized
In 2007 . In 2010 _ . . .
Recolonized Recolonized
In 2007
In 2010
Not Yet Never
Recolonized NotYet Colonized
Recolonized
Never
Colonized
Box plots demonstrating differences in total, native and introduced
graminoid cover across site types (recolonized in 2007 after one year
of extirpation, recolonized in 2010 after three years of extirpation,
not yet recolonized, and never colonized) and tillage history (tilled
and untilled). Data presented are averages from the 2013 early and
late growing season sampling events. Different letters indicate
significant differences between treatments at the P < 0.01 level.


% Cover % Cover % Cover
2014
Recolonized
In 2007
Recolonized Not Yet
Recolonized ln 2010 Recolonized Recolonized Not Yet
In 2007 In 2010 Recolonized
Never
Colonized
Never
Colonized
Box plots demonstrating differences in total, native and introduced
graminoid cover across site types (recolonized in 2007 after one year of
extirpation, recolonized in 2010 after three years of extirpation, not yet
recolonized, and never colonized) and tillage history (tilled and untilled).
Data presented are averages from the 2014 early and late growing season
sampling events. Different letters indicate significant differences between
treatments at the P < 0.01 level.
25


% Cover % Cover % Cover
2013-2014
Recolonized
In 2007
Recolonized Not Yet
In 2010 Recolonized ......
Recolonized Recolonized Not Yet
In 2007 in 2010 Recolonized
Never
Colonized Mrnmr
Never
Colonized
Box plots demonstrating differences in total, native and introduced
forb cover across site types (recolonized in 2007 after one year of
extirpation, recolonized in 2010 after three years of extirpation, not
yet recolonized, and never colonized) and tillage history (tilled and
untilled). Data presented are averages from the 2013 and 2014 early
and late growing season sampling events. Different letters indicate
significant differences between treatments at the P < 0.01 level.
26


% Cover % Cover % Cover
O
00
2013
Forb Cover
O
ID
O
O
CM
O
O
m
o
CM
o
o
o
CO
o
cd
o
o
o
t-----------------------------1--------------1---------------1--------------1--------------r
Native Forb Cover
i i
;

o B I BC
j AB A A A
AC | i : i

i i i----------------------------------------------------1---------------------------------------------------1 i i
Introduced Forb Cover
T---------------1--------------?--------------1---------------1--------------1--------------1--------------T
Recolonized
In 2007
Recolonized
Recolonized
In 2010
Recolonized
Not Yet
Recolonized
Not Yet
Never
Colonized Never
In 2007 In 2010 Recolonized Colonized
Box plots demonstrating differences in total, native and introduced
forb cover across site types (recolonized in 2007 after one year of
extirpation, recolonized in 2010 after three years of extirpation, not
yet recolonized, and never colonized) and tillage history (tilled and
untilled). Data presented are averages from the 2013 early and late
growing season sampling events. Different letters indicate significant
differences between treatments at the P < 0.01 level.
27


% Cover % Cover % Cover
2014
O
CO
o
o
Cvi
o
o
m
o .
rM
Native Forb Cover

c

o o
C 7 0 BC
A
h n | L A 1 i : i 1 AB l
O
00
o
D
O
<3-
Recolonized Recolonized
,n 2007 , . In 2010 n . . .
Recolonized Recolonized
Not Yet
Recolonized
In 2007
In 2010
Never
Not Yet Colonized Never
Recolonized Colonized
Box plots demonstrating differences in total, native and introduced
forb cover across site types (recolonized in 2007 after one year of
extirpation, recolonized in 2010 after three years of extirpation, not
yet recolonized, and never colonized) and tillage history (tilled and
untilled). Data presented are averages from the 2014 early and late
growing season sampling events. Different letters indicate
significant differences between treatments at the P < 0.01 level.
28


% Cover % Cover % Cover
2013-2014
O
00
O

O
o
CM
O
O
CO
O
o
o
(N
O
O
VD
O
o
CM
o
T---------------1 I I I I---------------1 I
Bare Ground Cover
Litter Cover
Recolo'nized Recolonized ' Not Yet ' Never
In 2007
. . In 2010 . . Recolonized w tv . Colonized
Recolomzed Recolonized Not Yet coioruzeu |sjever
In 2007 In 2010 Recolonized Colonized
Box plots differences in total plant, bare ground, and litter cover
across site types (recolonized in 2007 after one year of extirpation,
recolonized in 2010 after three years of extirpation, not yet
recolonized, and never colonized) and tillage history (tilled and
untilled). Data presented are averages from the 2013 and 2014 early
and late growing season sampling events. Different letters indicate
significant differences between treatments at the P < 0.01 level.
29


Litter and Bare Ground Cover
The statistical patterns across years showed the same trends when comparing
treatment types, so we report findings averaged across years (Fig. 13). Litter and bare
ground cover showed opposing trends. Bare ground cover was significantly higher on
active prairie dog colonies regardless of tillage or time since recolonization post-plague
(Table A6, Figs 13). Conversely, litter cover was significantly higher on uncolonized sites,
again regardless of tillage and time since recolonization post-plague (Table A5, Fig 13).
Year also was a factor in bare ground cover, with bare ground cover in 2014 decreasing
on all treatments, anywhere from 9.6% to 51.7%, when compared to bare ground cover
in 2013. We attribute this year-to- year difference (Table A6) to the record rainfall in
September of 2013. Litter cover also decreased across years for all but one treatment
(Tilled, never colonized) with the lowest percentage decrease being 0.4% and the
highest at 60.3%.
Notable Species
Relatively few species made up a large proportion of cover on our sites. The four
most abundant species from each life history group (native graminoid, introduced
graminoid, native forb, introduced forb) have percent cover shown in Tables 1 and 2 and
the percentage of replicates per treatment type in Tables 3 and 4. Not surprisingly, the
most abundant native graminoid (Pascopyrum smithii) had higher cover on sites that
had not been tilled and that had been released from grazing pressure or never grazed.
The most abundant introduced graminoid (Bromus inermis) was highest on tilled sites
30


not occupied by prairie dogs. Native forbs did not have much of a presence individually,
with Heterotheca villosa having the highest cover percentage on transects that hadn't
been tilled. Lastly, Convolvulus arvensis cover was positively associated with prairie dog
occupation. Active prairie dog colonies showed significantly higher cover of bindweed
than the inactive sites (P<0.001), with an average of 30% cover regardless of tillage
(Table 1). The former prairie dog colonies that had not yet been recolonized showed a
greater (P<0.01) cover of Convolvulus arvensis than the never colonized sites.
31


Table 1: Mean Notable Graminoid Species Cover
Not tilled, never colonized Tilled, never colonized Not tilled, not yet recolonized Tilled, not yet recolonized Not tilled, recolonized 2010 Tilled, recolonized in 2010 Not tilled, recolonized in 2007 Tilled, recolonized in 2007 Mean Percent Cover of Notable Graminoid Species in Transect
7.5 i 15.3 32.6 1 20.0 4.0110.0 20.0118.3 0.812.6 6.3110.1 0,311.3 2,014.9 Bromus inermis Introduced
4.317.6 2.116.5 12.9116.7 3.9110.3 0.612.1 CD cn 14- cn 2.2110.6 0.211.6 Bromus tectorum
cd CT) 14- 0.110.4 14- -P=* CD cn 14- CD 0.0210.1 CD ro 14- CD CD 14- CD 0.0110,1 Bromus arvensis
1.615.3 5.1110.7 CD cn 14- bo 4.018.7 CD H- CD 0.412.2 0.0210.1 CD 14- CD Poo pratensis
9.018.2 1.012.9 5.716.4 3.115.1 1.112.9 0.511.4 0.811.8 0.612.8 S Co Co a: ^ c: 3 Native
0.310.9 010 0.110.4 0.612.6 1 010 0.110.5 CD 14- CD 0.0210.1 Bouteloua dactyloides
2.113.3 0.0210.1 CD 14- l 0.210.6 010 CD 14- CD 0.110.6 1.015.4 Bouteloua gracilis
CD NJ CD CD CD CD
O') k) CD CD CD n_> CD
H- i-*- 14- 14- 14- 14- 14- 14-
CD CD ^ §-
OD ro
We noted the most commonly occurring introduced and native graminoids within our study. For
each of those species, we found the average cover of that species for each treatment group.
Tilled treatments had 13 replicate transects and untilled treatments had 17 replicate transects.
Data shown here are averages of four sampling events (early and late season 2013 and 2014).


Table 2: Mean Notable Forb Species Cover
We noted the most commonly occurring introduced and native forbs within our study. For each
of those species, we found the average cover of that species for each treatment group. Tilled
treatments had 13 replicate transects and untilled treatments had 17 replicate transects. Data
shown here are averages of four sampling events (early and late season 2013 and 2014).
33


Table 3: Percentage Replicates of Notable Forb Species
We noted the most commonly occurring introduced and native forbs within our study. For each
of those species, we calculated the percentage of replicates in each treatment group that
contained a particular species. Tilled treatments had 13 replicate transects and untilled
treatments had 17 replicate transects. Data shown are averages of four sampling events (early
and late season 2013 and 2014).
34


Table 4: Percentage Replicates of Notable Graminoid Species
We noted the most commonly occurring introduced and native forbs within our study. For each
of those species, we calculated the percentage of replicates in each treatment group that
contained a particular species. Tilled treatments had 13 replicate transects and untilled
treatments had 17 replicate transects. Data shown here are averages of four sampling events
(early and late season 2013 and 2014).
35


Plant Community Similarity Based on Nonmetric Multidimensional Scaling
NMDS plots were separated by season and year. The plots are 95% confidence intervals
of the different treatment communities. For the early and late season in 2013 all
treatments showed significant differences from each other (p<0.001). For the early
season in 2014, all sites except the sites that were recolonized in 2007 and 2010 were
significantly different (P<0.001). The sites recolonized in 2007 and 2010 were similar
(P=0.44). The late season in 2014 showed the same trend as in the early season, with all
sites being different except for the sites recolonized in 2007 and 2010 (P=0.06).


Early Season 2013
Early Season 2014
Late Season 2013 Late Season 2014
nmds1 mmos1
NMDS plots showing community composition grouped by occupation history
(recolonized in 2007, recolonized in 2010, not yet recolonized, and never colonized)
at 95% confidence intervals
CHAPTER IV
DISCUSSION
37


We studied plant communities after a major plague epizootic in Boulder County,
Colorado, USA, to see if episodes of colony inactivity as a result of plague mitigate the
vegetation change caused by prairie dogs. We found that whether a not a colony was
currently occupied by prairie dogs and whether a not a site had been previously tilled or
not had the greatest effects on the plant community metrics we examined. We found no
differences between the sites that had been released from prairie dog grazing for one
versus three years because of the plague epizootic. This suggests that 1) history of
tillage, even 30+ years ago, still impacts the vegetation composition of habitats, 2) seven
years of release from grazing pressure is sufficient to return a prairie dog colony to a
state somewhat comparable to a never colonized site, and 3) a three year release from
grazing pressure did not result in lasting effects on vegetation once the site had been
recolonized. On the urban Front Range of Colorado, plague seems to be a potential
mitigating factor with respect to the plant community effects of prairie dogs. However,
there are interesting interactions between the disturbances of tilling and prairie dog
occupation in this urban habitat.
Influence of Historical Tillage
Historical tillage is an important factor in determining vegetation community
composition. Tillage is known to alter soil attributes and plant communities, even 60
years after cessation (McGranahan et al. 2014, Brudvig et al. 2013, Jangid et al. 2011,
38


Stylinski & Allen 1999). In this study, tillage severely decreased native species richness
and this was most apparent on the sites that had never been colonized by prairie dogs.
Native graminoid richness on the tilled, never colonized site was less than half that of
the untilled never colonized site. The differences were even more pronounced with
native graminoid cover, with the untilled sites having over three times more native grass
cover than the tilled sites. Native forb cover and richness was significantly higher on all
untilled transects regardless of prairie dog occupation. Tillage effects on unoccupied
sites show higher introduced graminoid richness and cover but not higher introduced
forb cover. These findings suggest that tillage plays a vital role in the resilience of native
species to other disturbances, such as prairie dogs. Some distinct agricultural legacies
remain on some of the tilled sites, such as the planting of hay grasses like Bromus
inermis. Tilled sites also appear to have been influenced by plant propagule pressure
from neighboring properties. Our finding that disturbance from tillage impacts plant
communities is not surprising given that, in disturbed landscapes, invasive plants often
outcompete native plant species (Fenesi et al. 2015, Hagen et al. 2014, Wang et al.
2013). However, our findings related to how the disturbance of prairie dogs interacts
with the disturbance of tillage is of importance both to our understanding of how
multiple change drivers interact and to our ideas about management of prairie dogs in
urban areas.
Relationship between Tillage and Prairie Dog Occupation History
39


Seven years after being released from prairie dog grazing, unfilled, not yet
recolonized sites are approaching the never colonized sites, showing similarities in
richness, diversity, evenness, and cover. Native graminoid cover and richness is on the
rise, as is native forb richness. These recolonized sites have already surpassed their tilled
equivalent in native richness and cover. On unfilled sites, this indicates that prairie dog
effects are not irreversible, at least not when periodic plague outbreaks continue to
occur. Sites experiencing prairie dog extirpation will likely eventually revert to a state
closer to the never colonized site. However, on tilled sites, native forb richness and
cover does not increase after prairie dog exptirpation. Instead there is an increase in
cover and richness of introduced graminoids. For tilled sites that have not yet been
recolonized, there is slightly higher species diversity than never colonized sites. It's
possible these sites experience multiple propagule pressures after the prairie dogs are
gone when bare soil cover is high, thus increasing the number of species for a short time
before the more dominant species colonize the site. It is also possible that the increase
in species diversity resulting from prairie dog occupation is long lasting.
Three years of recovery before being occupied by prairie dogs did not seem to
affect the vegetation community metrics in a different way than one year of recovery.
Sites that were colonized the year after the plague epizootic, in 2007, were similar to
the sites colonized three years later, in 2010. With the exception of native forb richness
and cover, the effects of previous tillage could not be distinguished from the effects of
prairie dog occupation. Introduced and native graminoid richness and cover, species
diversity, and evenness were the same on the tilled and not tilled occupied sites.
40


Whatever recovery the vegetation gains after three years of no occupation is quickly
reversed once prairie dogs begin actively occupying the site.
Relationship of Our Findings to Previous Work
This study corroborates the findings from both rural and urban landscapes that
prairie dogs decrease graminoid cover and prairie dogs increase forb cover (Agnew et al.
1986, Whicker and Detling 1988, Magle & Crooks 2008). However, in our urban sites,
this increase in forb cover is not due to an increase in richness; instead the number of
forb species declines, particularly the native forbs, and the cover of introduced forbs
increases. Though historical tillage severely reduces native forb richness, the effect of
prairie dogs on sites that are not tilled also drives richness lower.
Our work also augments the findings of Beals et al. (2014) which demonstrated
that this region (urban Boulder County, Colorado, USA) has been experiencing
directional declines in native graminoid cover, increases in bare ground and increases in
native and non-native forbs and that prairie dog occupation amplifies those directional
changes by four to 10 times. Like Beals et al (2014), we also found that prairie dog
occupied sites had lower native graminoid cover, more bare ground, and higher forb
cover. We can add that plague may be a mitigating factor that reduces the landscape
level impact of prairie dogs and that vegetation changes are highly influenced by tillage
history.
Effects of Prairie Dogs and Tillage and Plague on Specific Plant Species
41


We examined specific notable species and found interesting trends. The
presence of the toxic morning glory species Convolvulus arvensis (field bindweed) seems
to be closely associated with the prairie dog specific disturbance. Seven years after
prairie dogs have been extirpated the cover of C. arvensis decreases by more than half.
C. arvensis seems to have a harder time competing without the grazing pressure on
other plant species put forth by prairie dogs. Other introduced forb species such as
Erodium cicutarium (Redstem stork's bill) show some preference for active prairie dog
sites, but are able to persist longer than C. arvensis after the prairie dogs are gone. We
suggest that the relationship between C. arvensis and prairie dogs deserves further
attention. Does the high primary production of C. arvensis allow prairie dogs to persist
in high densities on sites that would otherwise be abandoned because of lack of forage?
What are the characteristics of C. arvensis that promote its success on prairie dog
colonies? How does herbivory of C. arvensis influence its toxicity?
Graminoid species increase in cover and richness after the extirpation of prairie
dogs, but the recovery of community composition is affected by tillage history. Bromus
inermis (Smooth brome), the most dominant introduced graminoid, shows a greater
percentage of cover on tilled, unoccupied sites, than any other graminoid. On sites that
haven't been tilled Pascopyrum smithii (Western wheatgrass) makes up the most
abundant graminoid, but has much less cover than the most abundant graminoid on the
tilled sites. This could be due to the increase in species diversity and evenness on the
untilled, never colonized site, with many different species being represented in the
cover data, instead of the site being dominated by one species.
42


Management Implications
In the urban to suburban Front Range of Colorado, plant communities are
experiencing novel conditions that have a large impact on composition. Native prairie
dogs are pushing the urban Boulder vegetation community towards a novel state faster
than the ecosystems would be with the absence of prairie dogs (Beals et al. 2014). In
this study and in Beals et al. (2014) it is shown that prairie dogs amplify trends not seen
in rural settings: lower richness, evenness, and diversity. When tillage interacts with
prairie dog grazing these two factors can drive richness, evenness, and diversity even
lower. Prairie dogs face many challenges in the urban environment particularly
movement restrictions due to anthropogenic features. Increased road density was
positively correlated with an increase in the density of prairie dog burrows (Johnson and
Collinge 2004, Magle et al. 2010). This increased burrow density and inability to migrate
may result in overpopulation in colonies, which would increase grazing. This may be an
explanation for the increase in bare ground cover on urban prairie dog colonies when
compared to rural ones (Hartley et al. 2009). Hopson et al. (2015) also reports that
urban prairie dog density is higher than on exurban sites. On the windy Front Range, this
may have devastating effects on the native seed bank, where seeds are blown away and
the propagule pressure from introduced species takes over. Thus prairie dog overgrazing
may have severe consequences for the recovery of native vegetation. Plant community
composition is determined by availability of seed and vegetative sources and conditions
(i.e. soil nutrients, moisture, temperature) for plant establishment, and stochastic
43


community assembly processes. It is possible that directional changes in urbanization
and climate could alter how prairie dog colony vegetation will respond to prairie dog
extirpation in the future. We suggest further exploration into the effects of prairie dogs
on soil erosion, soil nutrients, and soil seed banks. We also suggest further exploration
into how the surrounding landscape matrix influences vegetation recovery after
extirpation of prairie dogs.
Further, extreme weather events (e.g. rainfall or drought) may provide
temporary relief to the colonies greater bare ground cover and reduced vegetative
cover or exacerbate the effects of prairie dogs on vegetation. During September 2013
shortly after the fall vegetation sampling, the Front Range experienced one of the most
intense precipitation events in its recorded history. This record amount of precipitation
is a probable explanation for the decrease of bare ground and litter in the spring of 2014
and the increase in plant cover. This also accelerated the spread of introduced grasses,
of which many are spring annuals (Bromus tectorum and B. japonicus), into certain
colonies, which may suppress native species even further (Fig. 15). This event also may
have caused the observed increase in forb richness across treatments regardless of
tillage, increasing native forb richness by 41.6% (mean across transects) and increasing
introduced forb richness by 56.4% (mean across transects). Without this increase, the
effect of prairie dogs on forb richness may have been more pronounced.
44


Transect in May 2014 showing an explosion of Bromus tectorum where flood waters
passed. In September 2013, 46.1 cm of precipitation fell in Boulder, CO over a three
day period. Above the flood path persists native graminoids including Pascopyrum
smithii and Hesperostipa comata.
Urban Boulder is a novel ecosystem that is rapidly changing. Human factors,
climate change, agricultural changes, and new diseases may accelerate the directional
changes of plant communities. The extirpation of prairie dogs by periodic plague
epizootics may slow down their effect on the vegetation communities and allow native
species a chance to rebound.
The results and implications from this study should be used in conjunction with
others to maintain or improve the stability and function of urban grassland ecosystems.
We contend that untilled sites provide the best place to maintain grassland ecosystems
45


in urban Colorado without major re-vegetation projects. This study also shows that a
reduction in prairie dog numbers could benefit these sites. Prairie dogs in their current
urban state may be better suited to historically tilled sites where the vegetation
composition has already been altered and the impact that prairie dogs have on native
vegetation is lessened. Plague may offer a viable means of control without the expense
to land managers. Plague outbreaks occur at intervals around every 6-12 years in rural
settings (Augustine et al. 2008, Hartley et al. 2009, Cully et al. 2010) but it remains to be
seen if these cycles are similar in urban areas. Should prairie dogs evolve resistance to
the plague bacteria, the consequences to the vegetative communities will need
additional studies.
46


REFERENCES
Agnew W, Uresk DW, and Hansen RM. 1986. Flora and fauna associated with prairie dog
colonies and adjacent ungrazed mixed-grass prairie in western South Dakota.
Journal of Range Management 39:135-139.
Augustine DJ, Derner JD, and Detling JK. 2014. Testing for thresholds in a semiarid
grassland: the influence of prairie dogs and plague. Rangeland Ecology &
Management 67:701-709.
Augustine DJ, Matehett MR, Toombs TP, Cully JF, Johnson TL, and Sidle JG. 2008.
Spatiotemporal dynamics of black-tailed prairie dog colonies affected by plague.
Landscape Ecology 23:255-267.
Bangert RK and Slobodchikoff CN. 2000. The Gunnison's prairie dog structures a high
desert grassland landscape as a keystone engineer. Journal of Arid Environments
46:357-69.
Beals SC, Hartley LM, Prevey JS, and Seastedt TR. 2014. The effects of black-tailed prairie
dogs on plant communities within a complex urban landscape: An ecological
surprise? Ecology 95:1349-1359.
Berg MD, Sorice MG, Wilcox BP, Angerer JP, Rhodes EC, and Fox WE. 2015. Demographic
changes drive woody plant cover trends- and example from the Great Plains.
Rangeland Ecology & Management 68:315-321.
Bonham CD and Lerwick A. 1976. Vegetation changes induced by prairie dogs on
shortgrass range. Journal of Range Management 29:221-5.
Boulder Monthly Climate Data: Precipitation (2015) Earth System Research Laboratory
http://www.esrl.noaa.gov/psd/boulder/Boulder.mm.precip.html
47


Brudvig LA, Grman E, Habeck CW, Orrock JL, and Ledvina JA. 2013. Strong legacy of
agricultural land use on soils and understory plant communities in longleaf pin
woodlands. Forest Ecology and Management 310: 944-955
Buckner DL. 1985. Point-intercept sampling in revegetation studies: maximizing
objectivity and repeatability. Paper presented at American Society for Surface
Mining and Reclamation Meeting, Denver, CO. 1985.
Cid MS, Detling JK, Whicker AD, and Brizuela MA. 1991. Vegetational responses of a
mixed-grass prairie site following exclusion of prairie dogs and bison. Journal of
Range Management 44:100-5.
Coffin DP, Lauenroth WK, and Burke 1C. 1996. Recovery of vegetation in a semiarid
grassland 53 years after disturbance. Ecological Applications 62:538-55.
Collinge SK, Johnson WC, Ray C, Matchett R, Grensten J, Cully JF Jr., Gage KL, Kosoy MY,
Loye JE, and Martin AP. 2005. Landscape structure and plague occurrence in
black-tailed prairie dogs on grasslands of the western USA. Landscape Ecology
2941-55.
Coppock DL, Detling JK, Ellis JE, and Dyer Ml. 1983. Plant-herbivore interactions in a
North American mixed-grass prairie. I. effects of black-tailed prairie dogs on
intraseasonal aboveground plant biomass and nutrient dynamics and plants
species diversity. Oecologia 56:1-9.
Cully JF Jr., Johnson TL, Collinge SK, and Ray C. 2010. Disease limits populations: plague
and black-tailed prairie dogs. Vector-borne and Zoonotic Diseases 10:7-15.
Cully JF and ES Williams. 2001. Interspecific comparisons of sylvatic plague in prairie
dogs. Journal of Mammalogy 82:894-905.
48


Dearborn DC and Kark S. 2010. Motivations for conserving urban biodiversity.
Conservation Biology 24:432-40.
Fahnestock JT, Larson DL, Plumb GE, and Detling JK. 2003. Effects of ungulates and
prairie dogs on seed banks and vegetation in a North American mixed-grass
prairie. Plant Ecology 167:255-68.
Fenesi A, Gered J, Meiners SJ, Tothmeresz B, Torok P, and Ruprecht E. 2015. Does
disturbance enhance the competitive effect of the invasive Solidao Canadensis
on the performance of two native grasses? Biological Invasions 17: 3303-3315.
Forrest S. 2005. Getting the story right: A response to Vermeire and colleagues.
Bioscience 55:526-30.
Friggens MM, Beier P. 2010. Anthropogenic disturbance and the risk of flea-borne
disease transmission. Oecologia 164: 809-820.
Hagen D, Hansen Tl, Graae BJ, and Rydgren K. 2014. To seed or not to seed in alpine
restoration: introduced grass species outcompete rather than facilitate native
species. Ecological Engineering 64: 255-261.
Hartley, LM, Detling JK, and Savage LT. 2009. Introduced plague lessens the effects of an
herbivorous rodent on grassland plant communities. Journal of Applied Ecology
46:861-869.
Hobbs RJ, Higgs E, Hall CM, Bridgewater P, Chapin FS, Ellis EC, Ewell JJ, Hallett LM, Harris
J, Hulvey KB, Jackson ST, Kennedy PL, Kueffer C, Lach L, Lantz TC, Lugo AE,
Mascaro J, Murphy SD, Nelson CR, Perring MP, Richardson DM, Seastedt TR,
Standish RJ, Starzomski BM, Suding KN, Tognetti PM, Yakob L, Yung L. 2014.
Managing the whole landscape: historical, hybrid, and novel ecosystems.
Frontiers in Ecology and the Environment 12(10):557-564.
49


Hopson R, Meiman P, and Shannon G. 2015. Rangeland dynamics: investigating
vegetation composition and structure of urban and exurban prairie dog habitat.
PeerJ 3:e736.
Jangid K, Williams MA, Franzluebbers AJ, Schmidt TM, Coleman DC, and Whitman WB.
2011. Land-use history has a stronger impact on soil microbial community
composition than aboveground vegetation and soil properties. Soil Biology &
Biochemistry 43:2184-2193
Johnston A. 1957. A comparison of the line interception, vertical point quadrat, and loop
methods as used in measuring basal area of grassland vegetation. Canadian
Journal of Plant Science 37:34-42.
Kinsinger FE, Eckert RE, and Currie PO. A comparison of the line-interception, variable-
plot, and loop methods as used to measure shrub-crown cover. Journal of Range
Management 13: 17-21.
Magle SB and Crooks KR. 2008. Interactions between black-tailed prairie dogs (Cynomys
ludovicianus) and vegetation in habitat fragmented by urbanization. Journal of
Arid Environments 72:238-46.
Magle SB, Reyes P, Zhu J, and Crooks KR. 2010. Extirpation, colonization, and habitat
dynamics of a keystone species along an urban gradient. Biological Conservation
143:2146-55.
McGranahan DA, Daigh AL, Veenstra JJ, Engle DM, Miller JR, and Debinski DM. 2014.
Connecting soil organic carbon and root biomass with land-use and vegetation in
temperate grassland. The Scientific World Journal 2014:487563
Mellado M, Olvera A, Quero A, Mendoza G. 2005. Dietary overlap between prairie dog
(Cynomys mexicanus) and beef cattle in a desert rangeland of northern Mexico.
Journal of Arid Environments 62: 449-458.
50


Miller B, Reading R, Hoogland J, Clark T, Ceballos G, List R, Forrest S, Hanebury L,
Manzano P, and Pacheco J. 2000. The role of prairie dogs as a keystone species:
Response to Stapp. Conservation Biology 14:318-21.
Miller SD, and Cully JF Jr. 2001. Conservation of black-tailed prairie dogs (Cynomys
ludovicianus). Journal of Mammalogy 82:889-93.
Munson SM, Belnap J, and Okin GS. 2011. Responses of wind erosion to climate-induced
vegetation changes on the Colorado plateau. Proceedings of the National
Academy of Sciences 108:3854-3859.
Norland MR, Veith DL, and Dewar SW. 1993. Standing crop biomass and cover on
amended coarse taconite iron ore tailing. Pages 385-415 in Proceedings of the
10th Meeting, Spokane, Washington, May 16-19,1993. American Society for
Surface Mining and Reclamation, Champaign, Illinois, USA.
Osborn B and Allan PF. 1949. Vegetation of an abandoned prairie-dog town in tall grass
prairie. Ecology 30:322-32.
Pauli JN, Buskirk SW, Williams ES, Edwards WH. 2006. A plague epizootic in the black-
tailed prairie dog (Cynomys ludovicianus). Journal of Wildlife Diseases 42:74-80.
Prevey JS and Seastedt TR. 2014. Seasonality of precipitation interacts with exotic
species to alter composition and phenology of a semi-arid grassland. Journal of
Ecology 102:1549-1561.
Seastedt TR, Hartley LM, and Nippert J. 2013. Case Study: Ecosystem transformations
along the Colorado Front Range: prairie dog interactions with multiple
components of global environmental change. Pp. 142-149 In Novel Ecosystems:
Intervening in the New Ecological World Order, First Edition, (e.d. R. J. Hobbs, E.
S. Higgs, and C. M. Hall). John Wiley & Sons.
Seastedt TR, Hobbs RJ, and Suding KN, 2008. Management of novel ecosystems: are
novel approaches required? Frontiers in Ecology and the Environment 6: 547-
553.
51


Snail T, O'Hara RB, Ray C, and Collinge SK. 2008. Climate-driven spatial dynamics of
plague among prairie dog colonies. American Naturalist 171:238-48.
Spasojevic MJ, Bowman WD, Humphries HC, Seastedt TR, and Suding KN. 2013. Changes
in alpine vegetation over 21 years: are patterns across a heterogeneous
landscape consistent with predictions? Ecosphere 4:1-18.
Stylinski CD and Allen EB. 1999. Lack of native species recovery following severe exotic
disturbance in southern California shrublands. Journal of Applied Ecology 36:544-
554.
USDA, NRCS. 2015. The PLANTS Database (http://plants.usda.gov, 11 October 2015).
National Plant Data Team, Greensboro, NC 27401-4901 USA.
Vermeire LT, Heitschmidt RK, Johnson PS, and Sowell BF. 2004. The prairie dog story: Do
we have it right? Bioscience 54(7): 689-695.
Vitousek PM, D'Antonio CM, Loppe LL, Rejmanek M, and Westbrooks R. 1997.
Introduced species: a significant component of human-caused global change.
New Zealand Journal of Ecology 21:1-16.
Wang XL, Wang YQ, and Wang YJ. Use of exotic species during ecological restoration can
produce effects that resemble vegetation invasions and other unintended
consequences. Ecological Engineering 52:247-25.1
Whicker AD, and Detling JK. 1988. Ecological consequences of prairie dog disturbances.
BioScience 38:778-85.
52


APPENDIX
Table Al: MANOVA output for Evenness & Diversity
Factor Df Wilk's Lambda Approx. F Num df Den df Pr(>F)
Tillage History (Tilled or untilled) 1 0.679 102.922 2 435 <0.001
Prairie Dog Occupation History (Never colonized, not yet recolonized, recolonized 2007, recolonized 2010) 3 0.649 35.048 6 870 <0.001
Year (2013, 2104) 1 0.929 16.543 2 435 <0.001
Season (early, late) 1 0.897 24.866 2 435 <0.001
Tillage History: Occupation History 3 0.952 3.577 6 870 0.002
Tillage History: Year 1 0.997 0.672 2 435 0.511
Occupation History: Year 3 0.974 1.958 6 870 0.069
Tillage History: Season 1 0.994 1.270 2 435 0.282
Occupation History: Season 3 0.992 1.291 6 870 0.276
Year: Season 1 0.994 1.218 2 435 0.297
Tillage History: Occupation History:Year 3 0.994 0.467 6 870 0.932
Tillage History: Occupation History:Season 3 0.993 0.532 6 870 0.784
Tillage History: Year: Season 1 0.998 0.403 2 435 0.669
Occupation History: Year: Season 3 0.994 0.385 6 870 0.889
Tillage History: Occupation History: Year: Season 3 0.990 0.694 6 870 0.654
Statistical output for MANOVA with dependent variables of evenness and diversity.
53


Table A2: MANOVA output for Richness
Factor Df Wilk's Lambda Approx. F Num df Den df Pr(>F)
Tillage History (Tilled or untilled) 1 0.498 90.237 5 448.0 <0.001
Prairie Dog Occupation History (Never colonized, not yet recolonized, recolonized 2007, recolonized 2010) 3 0.393 33.232 15 1237.1 <0.001
Year (2013, 2104) 1 0.771 26.684 5 448.0 <0.001
Tillage History: Occupation History 3 0.840 5.364 15 1237.1 <0.001
Tillage History: Year 1 0.980 1.814 5 448.0 0.109
Occupation History: Year 3 0.964 1.097 15 1237.1 0.354
Tillage History: Occupation History: Year 3 0.970 0.899 15 1237.1 0.565
Statistical output for MANOVA with dependent variables of total species richness, total
graminoid richness, native graminoid richness, introduced graminoid richness, total forb
richness, native forb richness, and introduced forb richness
54


Table A3: MANOVA output for Cover
Factor Df Wilk's Lambda Approx. F Num df Den df Pr(>F)
Tillage History (Tilled or untilled) 1 0.669 26.478 8 429.0 <0.001
Prairie Dog Occupation History (Never colonized, not yet recolonized, recolonized 2007, recolonized 2010) 3 0.209 37.084 24 1244.8 <0.001
Year (2013, 2104) 1 0.757 17.224 8 429.0 <0.001
Season (early, late) 1 0.769 16.137 8 429.0 <0.001
Tillage History: Occupation History 3 0.711 6.485 24 1244.8 <0.001
Tillage History: Year 1 0.962 2.130 8 429.0 0.031
Occupation History: Year 3 0.850 2.991 24 1244.8 <0.001
Tillage History: Season 1 0.987 0.682 8 429.0 0.707
Occupation History: Season 3 0.920 1.504 24 1244.8 0.056
Year: Season 1 0.971 1.624 8 429.0 0.116
Tillage History: Occupation History:Year 3 0.934 1.239 24 1244.8 0.197
Tillage History: Occupation History:Season 3 0.976 0.441 24 1244.8 0.991
Tillage History: Year: Season 1 0.974 1.413 8 429.0 0.189
Occupation History: Year: Season 3 0.978 0.396 24 1244.8 0.996
Tillage History: Occupation History: Year: Season 3 0.972 0.510 24 1244.8 0.976
Statistical output for MANOVA with c ependent variables of tota cover, bare ground
cover, litter cover, total graminoid cover, native graminoid cover, introduced graminoid
cover, total forb cover, native forb cover, and introduced forb cover.
55


Table A4: ANOVA for Graminoid Cover
Graminoid Cover
Introduced Native
Factor Df F Value Pr(>F) F Value Pr(>F)
Tillage History (Tilled or untilled) 1 91.90 <0.001 42.45 <0.001
Year (2013, 2104) 1 14.74 <0.001 3.63 0.057
Prairie Dog Occupation History (Never colonized, not yet recolonized, recolonized 2007, recolonized 2010) 3 154.60 <0.001 56.94 <0.001
Season (early, late) 1 5.11 0.024 0.61 0.435
Tillage History: Year 1 0.97 0.326 0.00 0.985
Tillage History: Occupation History 3 26.32 <0.001 23.05 <0.001
Occupation History: Year 3 4.90 0.002 0.43 0.731
Tillage History: Season 1 0.05 0.825 0.04 0.832
Year: Season 1 0.02 0.878 0.79 0.372
Occupation History: Season 3 1.07 0.360 0.29 0.831
Tillage History: Occupation History:Year 3 0.31 0.822 0.21 0.892
Tillage History: Year: Season 1 0.17 0.678 0.11 0.736
Tillage History: Occupation History:Season 3 0.23 0.875 0.61 0.611
Occupation History: Year: Season 3 0.18 0.908 0.06 0.983
Tillage History: Occupation History: Year: Season 3 0.13 0.940 0.15 0.927
Statistical output for summary ANOVA with
cover and introduced graminoid cover.
dependent variables of native graminoid
56


Table A5: ANOVA for Forb Cover
Forb Cover
Introduced Native
Factor Df F Value Pr(>F) F Value Pr(>F)
Tillage History (Tilled or untilled) 1 0.001 0.922 109.85 <0.001
Year (2013, 2104) 1 21.69 <0.001 8.47 0.004
Prairie Dog Occupation History (Never colonized, not yet recolonized, recolonized 2007, recolonized 2010) 3 135.19 <0.001 1.57 0.196
Season (early, late) 1 25.65 <0.001 3.84 0.051
Tillage History: Year 1 0.26 0.613 5.21 0.022
Tillage History: Occupation History 3 0.64 0.593 1.88 0.132
Occupation History: Year 3 11.07 <0.001 1.25 0.289
Tillage History: Season 1 0.52 0.469 0.37 0.544
Year: Season 1 0.49 0.486 0.47 0.492
Occupation History: Season 3 2.97 0.032 0.45 0.717
Tillage History: Occupation History:Year 3 0.43 0.73041 1.93 0.123
Tillage History: Year: Season 1 2.20 0.138 0.25 0.615
Tillage History: Occupation History:Season 3 0.53 0.663 0.63 0.597
Occupation History: Year: Season 3 0.34 0.795 0.37 0.774
Tillage History: Occupation History: Year: Season 3 0.25 0.864 0.51 0.685
Statistical output for summary ANOVA with dependent variables of forb graminoid
cover and introduced forb cover.
57


Table A6: ANOVA for Other Cover
Other Cover
Bare Ground Litter
Factor Df F Value Pr(>F) F Value Pr(>F)
Tillage History (Tilled or untilled) 1 0.40 0.529 2.80 0.095
Year (2013, 2104) 1 44.20 <0.001 10.28 0.001
Prairie Dog Occupation History (Never colonized, not yet recolonized, recolonized 2007, recolonized 2010) 3 291.79 <0.001 151.45 <0.001
Season (early, late) 1 47.90 <0.001 24.46 <0.001
Tillage History: Year 1 1.07 0.302 5.99 0.015
Tillage History: Occupation History 3 3.80 0.010 1.76 0.155
Occupation History: Year 3 2.82 0.039 3.27 0.021
Tillage History: Season 1 0.06 0.804 1.18 0.279
Year: Season 1 1.83 0.177 0.02 0.896
Occupation History: Season 3 5.86 <0.001 1.65 0.177
Tillage History: Occupation History:Year 3 0.46 0.709 0.15 0.933
Tillage History: Year: Season 1 0.06 0.805 0.15 0.698
Tillage History: Occupation History:Season 3 0.20 0.897 0.23 0.878
Occupation History: Year: Season 3 0.35 0.788 0.18 0.911
Tillage History: Occupation History: Year: Season 3 0.27 0.848 0.27 0.846
Statistical output for summary ANOVA with dependent variables of bare ground cover
58


Full Text

PAGE 1

FROM EXTIRPATION TO RECOLONIZATION: BUBO NIC PLAGUE MITIGATES THE EFFECTS OF PRAIRIE DOGS (CYN OMYS LUDOVICIANUS) O N PLANT COMMUNITIES by MARK DAVID VISEL B.A., Bethe l University, 2011 B.S., Bethel University, 2011 A thesis submitted to the Faculty of the Graduate School of the University of Colorado in partial fulfillment of the requirements for the degree of Master of Science Integrative Biology 2015

PAGE 2

ii This thesis for the Master of Science degree by Mark D avid Visel has been approved for the Department of Integrative Biology by Laurel Hartley, Chair Timothy Seastedt Michael Wunder November 20 th 2015

PAGE 3

iii Visel, Mark David (M.S., Biology) From E xtirpation to Recolonization: Bubonic Plague Mitigates the Effects of Prairie D ogs ( Cynomys ludovicianus ) on Plant C ommunities. Thesis directed by Ass ociate Professor Laurel M. Hartley ABSTRACT Plague, caused by Yersinia pestis was introduced to the western U nited States in the 1940s. Black tailed prairie dogs ( Cynomys ludovicianus ) are highly susceptible to the disease and plague has become a significant disturbance that results in periodic epizootics on prairie dog colonies. While there have been many studies of the effects of prairie dogs on vegetation, most have been conducted in rural rather than urban areas and most have not been related to plague. Examining vegetation on prairie dog colonies before and after plague in urban and suburban areas can have critica l management implications because, in urban areas, other global change factors such as urbanization, climate warming, nitrogen deposition, and propagule pressure from non native plant species interact with prairie dogs to result in novel plant communities Urban prairie dog colonies differ significantly from rural colonies by having higher density of prairie dogs, higher richness and cover of exotic vegetation, and higher percentages of bare ground. T his study evaluated whether periodic plague epizootics mi tigate the effects prairie dog have on vegetation in an urban environment, Boulder, Colorado Using point intercept methods we examine d the effects of plague epizootics on colony vegetation. Main findings are that 1) prairie dogs reduce native species rich ness and cover and increase bare ground, but recovery after seven years of abandonment is seen and 2) h istorical

PAGE 4

iv tillage ha s a large effect on the landscape as well, with tilled sites showing lower native species richness and cover. Our work has significan t implications for management of lands at the urban wild land interface and contributes to broader theories about how multiple change drivers can interact in ecosystems. The form and content of this abstract are approved. I recommend its publication. Approved: Laurel Hartley

PAGE 5

v ACKNOWLEDGEMENTS I thank Lynn Riedel and Heather Swanson from Boulder Open Space and Mountain Parks for assistance with permits, historical data, and plant identification. I would also like to thank David Buckner and ESCO A ssociates Inc. for loaning us the optical point projection device and teaching us to use it and the method. I also thank Stower Beals for his assistance with non metric multidimensional scaling analyses and interpretation I also thank m y committee members, Timothy Seastedt and Michael Wunder for their assistance in getting me started and seeing me through. Lastly, I would like to thank my advisor, Laurel Hartley, for her guidance, insight, and much help tha t she poured into this project. This project was funded by a grant from the National Science Foundation (Ecosystem Science #1120390) This study complies with current laws of the U.S.

PAGE 6

vi CONTENTS CHAPTER I. INTRODUCTION ................................ ................................ ................................ ................ 1 General Effects of Prairie Dogs on Plant Communities ................................ ................... 2 Effects of Prairie Dogs on Plant Communities in Urban Environments .......................... 3 The Role of Plague in Mitigating the Effects of Prairie Dogs on Plant Communities ...... 4 II. METHODS ................................ ................................ ................................ ........................ 7 Study Area ................................ ................................ ................................ ....................... 7 Field Methods ................................ ................................ ................................ .................. 8 Data Analysis ................................ ................................ ................................ ................. 10 III. RESULTS ................................ ................................ ................................ ........................ 13 Species Diversity and Evenness ................................ ................................ ..................... 13 Species Richness ................................ ................................ ................................ ............ 14 Plant Cover ................................ ................................ ................................ .................... 21 Litter and Bare Ground Cover ................................ ................................ ....................... 30 Notable Species ................................ ................................ ................................ ............. 30 Plant Community Similarity Based on Nonmetric Multidimensional Scaling ............... 36 IV. DISCUSSION ................................ ................................ ................................ ................. 37 Influence of Historical Tillage ................................ ................................ ........................ 38 Relationship between Tillage and Prairie Dog Occupation History .............................. 39 Relationship of Our Findings to Previous Work ................................ ............................ 41 Effects of Prairie Dogs and Tillage and Plague on Specific Plant Species ..................... 41 Management Implications ................................ ................................ ............................ 43 REFERENCES ................................ ................................ ................................ ...................... 47 APPENDIX ................................ ................................ ................................ .......................... 53

PAGE 7

vii LIST OF FIGURES Figure 1: Map of Prairie Dog Colony Extent ................................ ................................ ................. 10 2: Plant Richness, Diversity, Evenness 2013 2014 ................................ ............................ 15 3: Graminoid Richness 2013 2014 ................................ ................................ .................... 17 4: Forb Richness 2013 2014 ................................ ................................ .............................. 18 5:Forb Richness 2013 ................................ ................................ ................................ ........ 19 6: Forb Richness 2014 ................................ ................................ ................................ ....... 20 7: Graminoid Cover 2013 2014 ................................ ................................ ......................... 23 8: Graminoid Cover 2013 ................................ ................................ ................................ .. 24 9: Graminoid Cover 2014 ................................ ................................ ................................ .. 25 10: Forb Cover 2013 2014 ................................ ................................ ................................ 26 11: Forb Cover 2013 ................................ ................................ ................................ .......... 27 12: Forb Cover 2014 ................................ ................................ ................................ .......... 28 13: Other Cover 2013 2014 ................................ ................................ .............................. 29 14: NMDS Community Composition Plots ................................ ................................ ........ 37 15: Bromus after Flood ................................ ................................ ................................ ..... 45

PAGE 8

viii LIST OF TABLES Table 1: Mean Notable Graminoid Species Cover ................................ ................................ ...... 32 2: Mean Notable Forb Species Cover ................................ ................................ ................ 33 3: Percentage Replicates of Notable Forb Species ................................ ........................... 34 4: Percentage Replicates of Notable Graminoid Species ................................ ................. 35 A1 : MANOVA output for Evenness & Diversity ................................ ................................ 53 A2 : MANOVA output for Richness ................................ ................................ .................... 54 A3 : MANOVA output for Cover ................................ ................................ ......................... 55 A4 : ANOVA for Graminoid Cover ................................ ................................ ...................... 56 A5 : ANOVA for Forb Cover ................................ ................................ ................................ 57 A6 : ANOVA for Other Cover ................................ ................................ .............................. 58

PAGE 9

1 CHAPTER I INTRODUCTION Increasing human impact on ecosystems has resulted in some ecosystems being subject to the effects of multiple global change drivers in concert with oft en unpredictable results ( Berg et al. 2015, Beals et al. 2014, Hobbs et al. 2014, Prevey and Seastedt 2014, Spasojevic et al. 2013 Vitousek et al. 1997 ). Ou r research focuses on how the effects of a native keystone herbivore on plant communities are impacted by both by introduced disease and urbanization. Black tailed p rairie dogs ( Cynomys ludovicianus ) are semi fossorial rodents found throughout the Great Plains of North America and have been considered a keystone species and an ecosystem engineer (Forrest 2005, Vermeire et al. 2004 Bangert & Slobodchikoff 2000 Miller et. al. 2000). Prairie dogs have also been regarded as competitors with livestock, as th ey forage on similar species (Mellado et. al. 2005 Cid et al. 1991). While prairie dogs are found throughou t much of their historic range and the total acreage occupied by prairie dogs has been severely decreased (Miller & Cully 2001), they are a species than can tolerate urbanization. Prairie dog colonies can be found occupying grassland remnants within suburban and urban landscapes throughout the Front Range of Colorado It is well known that prairie dogs alter the vegetation structure and plant communi ty composition in their colonies, with the majority of research being conducted in rural areas (Fahnestock et al. 2003 Whicker & Detling 1988 Coppock et al. 1983). The effects of prairie dogs o n vegetation vary among grassland types The general effects of prairie

PAGE 10

2 dogs on vegetation in mixed grass prairie are less dramatic than in tall grass prairie and even less dramatic on short grass prairie ( Hartley et al. 2009 Cid et al. 1991 Coppock et al. 1983 Osborn & Allan 1949 ). We know that the effects of pr airie dogs on plant communities can be influenced by introduced plague ( Hartley et al. 2009 Augustine et al. 2007 ) and by urbanization (Beals et al. 2014 Magle et al. 2010 Magle & Crooks 2008 ). Our study examines the effects of plague epizootics on prai rie dog colony plant communities in an urban setting, Boulder County C olorado General Effects of Prairie Dogs on Plant Communities In the mixed grass prairie of Badlands National Park, South Dakota, vegetation on black tailed prairie dog colonies is dominated by forbs, whereas off of the colonies vegetation is dominated by gra minoids (Fahnestock et al. 2003). Rural short grass steppe follows the same pattern of ve getation change after prairie dog colonization (Hartley et al. 2009, Bonham & Lerwick 1976). On colonies that have been grazed heavily by prairie dogs for several years, species richness may be limited to a handful of forbs or small shrubs throughout the c olony (Whicker & Detling 1988). Richness of species is the g reatest on young colonies (2 3 years) (Whicker & Detling 1988; Coppock et al. 1983) Prairie dogs typically do not occupy tall grass prairies, although Osborn and Allan (1949) observed seven diffe rent stages of plant succession in and around an active prairie dog colony in Oklahoma.

PAGE 11

3 Effects of Prairie Dogs on Plant Communities in Urban Environments As in rural environments, urban environments in Colorado show decreased graminoids and increased f orbs with prairie dog occupation ( Beals et al. 2014, Magle & Crooks 2008). Plant height is also shorter on than off active colonies in an urban environment (Magle & Crooks 2008). In Boulder County Colorado prairie dog occupation results in a significant reduction of perennial plants on colonies and an increase in in troduced plant species (Beals et al. 2014). Urban and rural prairie dog colonies also show an increase in bare ground cover compared to sites unoccupied by prairie dogs (Magle & Cro oks 2008 Hartley et al. unpublished) T his and research in rural areas ( e.g. Hartley et al. 2009) suggests a release from grazing pressure should result in a decrease in bare ground cover. Invasive plant encroachment and soil erosion are significant land management issues in urban and suburban Colorado. Prairie dogs profoundly change vegetation in their colonies as well as increase bare ground cover and this may significantly contribute to soil erosion (Munson et al. 2011). Prairie dog colonies in urban a reas also have more than double the amount of disturbed soil per 1000 m, indicating a higher density of prairie dogs and increased erosion. (Hartley, unpublished). Prairie dogs may also contribute to invasive plant establishment by selective consumption, resulting in an alteration of competitive dynamics. Urban colonies have almost three and a half times the cover of non native plant species (Beals et al. 2014).

PAGE 12

4 The Role of Plague in Mitigating the Effects of Prairie Dogs on Plant Communities Within the last 70 years, prairie dogs have been faced with a new challenge: plague. Plague is caused by a flea borne bacterium ( Yersinia pestis ). Plague is extremely detrimen tal to prairie dog populations, as it kills all or nearly all prairie dogs within an infecte d colony (Pauli et al. 2006 Cully and Williams 2001 ). Plague can also spread from direct animal to animal contact, whether that is prairie dog to prairie dog or a predator or small rodent to prairie dog (Cully et al. 2010). Prairie dog colony complexes th at have been affected by plague have been severely decreased in area, which results in greater distances between colonies. These greater distances between colonies result in greater isolation, which lowers the probability of being recolonized after a plagu e event while also lowering the probability of a plague epizootic occurring again (Cu lly et al. 2010). U rban environments include many more barriers, such as roads and buildings, when compared rural environments, which restrict prairie dog movements. These urban barriers are also negatively associated with plague occurrence (Collinge et al. 2005). Despite these barriers to plague, Friggens and Beier (2010) reported that anthropogenic disturbances were associated with an increase in flea borne diseases. Pla gue also creates opportunities for vegetation changes on prairie dog colonies. In studying the effect of plague on rural prairie dog colonies, it was observed that the biomass of both graminoids and forbs increased following a plague epizootic, although re covery of biomass may be related to colony age at the time of the epizootic because the degree of impact of prairie dogs on colony vegetation increased with colony age (Hartley et al. 2009).

PAGE 13

5 In the Pawnee National Grassland of Colorado, USA less than 2% of colonies go more than 20 years without experiencing a plague epizootic, and colonies, on average, experience plague epizootics every six and a half years (Hartley et al. 2009). In addition, the probability that a colony will be recolonized within five y ears of an epizootic is only 50% (Hartley et al. 2009). Most of the research dealing with plague, vegetation, prairie dogs, or any combination of them, has been conducted in ru ral sites Urban and suburban habitats possess very different characteristics a nd dynamics, such as increased density of prairie dogs, increased habitat fragmentation, increased migration obstacles, and increased anthropogenic disturbances, which are important for land managers to consider ( Seastedt et al. 2013, Seastedt et al. 2008) While prairie dogs may be subject to local extinctions due to development, with proper management they can persist in an increasingly urbanized landscape (Magle et al. 2010). Climate change is also an important factor in both vegetation change and plague ( Seastedt et al. 2013, Snll et al. 2008). It is yet unclear how climate change will interact with grazing and management pressures (Coffin et al. 1996) Understanding how these variables affect plant communities in urban areas is critical to making proper land and prairie dog management decisions Understanding how periodic plague outbreaks mitigate vegetation change caused by prairie dogs colonization has important management considerations (e.g. should prairie dogs be removed from some areas ? s hould plague be controlled ? what should management be between plague cycles ? ). Currently we only have pre post

PAGE 14

6 plague vegetation data from ru the hig h density ies of prairie dogs or high non native plant propagule pressure that we see in urban and suburban colonies. Prairie dogs in Boulder County are accelerating the transformation of grasslands into a beyond that seen in rural colonies, increased encroa chment by invasive plants, and a detrimental decrease in native grasses (Beals et al. 2014) understand the role that introduced plague plays i n shaping plant communities in urban areas. It could potentially mitigate prairie dog induced changes to vegetation as is seen in rural areas (Hartley et al. 2009). In an effort to understand how plague influences the role of prairie dogs in urban areas, we studied plant communities on prairie dog colonies impacted by a plague epizootic in 2006 that extirpated more than 50 prairie dog colonies in the urban Front Range of Colorado USA The goals of our work were investigate plant community responses to prairie dog occupation and subsequent prairie dog extirpation due to plague. Further, we were i nterested in whether a history of previous tillage would influence those plant community responses.

PAGE 15

7 CHAPTER II METHODS Study Area This study was conducted during the months of May August 2013 and 2014 on Boulder Open Space and Mountain Parks (BOSMP) public land in Boulder, Colorado, USA (40.0176 N, 105.2797 W mean elevation of 1640m ) Mean annual precipitation is 486 mm. Precipitation in 2012 was 398 mm Precipitation in 2013 was a record 867 mm with 461 mm falling in a matter of days in September (Boulder Climate Data 2015 ) May August precipitation was 145 mm in 2013 and 290 mm in 2014 (Boulder Climate Data, 2015). At all study s ites, s oils consist of mostly clay lo am to sandy clay loam, at 0 to 12 percent slopes (BOSMP, unpublished dat a ) Vegetation is primarily a mixed grass prairie mosaic, with so me properties used for grazing cattle During the study period, t here was no cattle grazing on the prairie dog colonies Despite the agric ultural usage, the sites were located in a variable landscape with a range of urban factors (Beals et al. 2014). We chose study sites from among the 151 properties containing prairie dogs on BOSMP land. BOSMP has a total of 410 colonies ranging in size f rom less than .01 ha to 215 ha with a mean area of 4.5 ha. BOSMP maps active prairie dog colonies each year and has done so since 1991 Because we were interested in vegetation change after plague epizootics, only colonies that were extirpated during a wid e spread plague event

PAGE 16

8 in 2006 were considered Further more we avoided sites that were closed because of nesting threatened or endangered birds and extremely small colonies (less than 4 square ha ). Of the 30 sites selected based on th ese criteria 13 sites ha d never been tilled, 17 ha d been tilled historically ranging from 20 to 30+ years ago (BOSMP, unpublished data) From the BOSMP digital maps of prairie dog colonies, we could determine prairie dog occupation history. Usin g Google Earth (2014), we divide d the colony into polygons based on prairie dog occupation history and established a 50 x 2 m sampling transect within each polygon. One transect was located in an area of the colony that was recolonized by prairie dogs in 2007, less than one year after th e plague epizootic A second transect was in an area recolonized in 2010, about three years after the epizootic A third transect was in an area that was still uncolonized by prairie dogs, but had been active before the plague event in 2006 allowing 6 7 years of vegetation recovery The fourth transect was established in an area that had never been colonized by prairie dogs. The never colonized by prairie dog transects showed similar soil type, slope, and degree of urbanization when compared with the site s that had been previously occupied by prairie dogs All of the prairie dog colonies in our study had been active for approximately 4 7 years prior to the plague epizootic Field Methods To estimate plant cover by species, w e used a Cover Point Optical P oint Projection D evice (ESCO Associates, Boulder, Colorado, USA ) which allows the standing operator to visualize the species located in the crosshairs of a low powered optical

PAGE 17

9 scope located above and perpendicular to the ground (Norland et al. 1993) This method was chosen because is relatively easy to set up, affording us time to sample all of the prairie dog colonies that met our criteria for inclusion in the study. In addition, this method can be used for long term monitoring because it provides accurate and repeatable data (Buckner 1985, Kinsinger 1960, Johnston 1957) and is currently the method that BOSMP uses for the monitoring of vegetation transects We used a n orange spray painted rebar stake to mark the beginning of each transect and recorded the GPS (Global Positioning System) location of the stake so that we could locate the transect for the duration of the study. The coordinate direction of the transect was randomly assigned Data w ere collected at points 0 .5 m to the left and 0 .5 m to the right o f the transect tape (50 x 2 m) at 1 m intervals for a total of 100 points per transect. We recorded whether the crosshair was above bare soil, litter, rock, standing dead vegetation or a live plant species These crosshair hits were used to estimate cover of plants by species, litter cover and bare ground. We calculate d richness, evenness, and Shannon Wiener diversity values using all data from the transect sampling, which included species noted during the point sampling and species present in the 50 X 2 m area but not captured during the point sampling Vegetation was conventionally classified by names, origins, and functional groups following the nomenclature of the USDA Plants Database (USDA, NRCS) Sites were sampled twice per ye ar during the growing season which allowed us to capture both early and late season plant species. The first sampling period began in the end of May and en ded in the beginning of June. The second sampling period started

PAGE 18

10 in early August and ended in late A ugust. The order in which sites were sampled was random. Data Analysis Our analysis choices were guided by our overarching goals of contributing to management of prairie dogs and plague and to relating our wor k to previous findings in urban and rural settings thus we focused primarily on effects of prairie dogs, plague, and previous tillage on species diversity and evenness, native and introduced graminoid and forb richness and co ver and on bare ground and litter cover All analyses were performed using the statistical program R (R Development Core Team 2011 ) We calculated Shannon Wiener diversity and species evenness for each of our transects. Species that were present in a transect, but not hit ing the point Map of prairie dog colony E xtent before plague (orange), extent in 2007 (purple) and extent in 2010 (pink). One transect was located in each of these areas. Our fourth transect was established outside the orange colony extent. Figure 1 : Map of Prairie Dog Colony Extent

PAGE 19

11 sampling were given cover values of 0.001 so that they would be counted, but less than actual hits We use d a two way factorial analysis of variance (ANOVA) to determine differences between tillage and transect type for evenness and diversi ty. We conducted comparisons between tillage (untilled and tilled) and site type (recolonized in 2007, recolonized in 2010, never recolonized, and never colonized) to examine differences in species diversity, evenness, richness and cover of plant s by func tional group (graminoids and forbs) and origin (native and introduced) Because richness data should encompass the total number of species in an area across the growing season we pooled our early and late season data ; therefore season was a factor in our analysis of cover data, but not species richness data. We used a two way factorial multivariate analysis of variance (MANOVA) since we had four independent variables for cover ( y ear s eason t illage h istory p rairie d og o ccupation h istory ), three factors f or richness ( y ear, t illage h istory, p rairie d og o ccupation h istory ), and multiple response variables for cover and richness. Since a MANOVA also takes into account possible intercorrelations among the response variabl es we ran one for all cover related va riables (total plant cover, bare ground cover, litter cover, graminoid cover, native graminoid cover, introduced graminoid cover, forb cover, introduced forb cover, native forb cover) and one for all species richness related variables (graminoid richness, native graminoid richness, introduced graminoid richness, forb richness, introduced forb richness, native forb richness) These variables have ecological significance for land managers. We also chose to examine differences among treatments with respect to the

PAGE 20

12 presence and cover of notable species. We considered species to be notable if they were very common and/or a species of concern to BOSMP land managers (pers. Communication) These species included introduced forbs ( Convolvulus arvensis, Erodium cicutarium, Carduus nutans, Dipsacus follonum), native forbs ( Heterotheca villosa, Sphaeralcea coccinea Psoralidium tenuiflorum Gaura coccinea ), introduce d graminoids ( Bromus inermis B. tectorum B. arvensis Poa pratensis ) and native graminoids ( Pascopyrum smithii, Bouteloua dactyloides B. gracilis, Aristida purpurea ). To assess notable species we counted the number of times a species was found in a replicate within each treatment group and calculated the percentage of rep licates in which that species was found. In addition we used ANOVA to investigate whether there were significant differences among on treatments with respect to cover of Convolvulus arvensis Bromus inermis and Pascopyrum smithii . The cover of other species was too small to warrant statistical analysis. We used non metric multidimensional scaling (NMDS) based on Bray Curtis dissimilarity to examine plant community differences among our treatments (recolonized in 2007, recolonized in 2010, not recolonized, and never colonized), years (2013, 2014), and season (early, late). The plots are a 95% confidence interval of the different treatment communities. We used the Permutational Multivariate Analysis of Variance (PERMANOVA) function i n R to quantify differences among treatments.

PAGE 21

13 CHAPTER III RESULTS The goals of our work were investigate plant community responses to prairie dog occupation and subsequent prairie dog extirpation due to plague. Further, we were interested in whether a history of previous tillage would influence those plant community resp onses. Species Diversity and Evenness Our MANOVA showed no biologically relevant significant interactions between year and our other factors (Table A1 ) therefore we report data averaged across years Tillage reduced species evenness across treatments however only significantly on the sites that were not occupied by prairie dogs (not recolonized, never colonized) (Fig. 2) Prairie dogs suppressed the effect of tillage by reducing evenness on active sites ( Fig. 2 ) O n the sites where prairie dogs were a ctive there was not a significant difference between the 2007 and 2010 recolonized treatments (Fig 2) Species diversity, measured by the Shannon Wiener diversity index, showed similar trends. Prairie d ogs reduced species diversity generally however til lage also had an effect (Fig 2) In areas without active prairie dogs tillage decr eased diversity significantly ( Fig. 2 ).

PAGE 22

14 Species Richness Our MANOVA showed no biologically relevant significant interactions between year and our other factors (Table A 2 ) therefore we report richness data averaged across years Both prairie dog occupation and tillage had significant influences on species richness. Sites that were currently occupied by prairie dogs had lower total plant species ri chness than sites that had not been recolonized post plague or never colonized when tillage was accounted for (Fig 2 ). Sites with a history of previous tillage had lower total plant species richness than sites that had not been tilled (Fig 2 ).

PAGE 23

15 Box plots demonstrating differences in species richness, diversity and evenness across site types (recolonized in 2007 after one year of extirpation, recolonized in 2010 after three years of extirpation, not yet recolonized, and never colonized) and tillag e history (tilled and untilled). Data presented are averages from the 2013 and 2014 early and late growing season sampling events. Different letters indicate significant differences between treatments at the P < 0.01 level. Figur e 2 : Plant Richness, Diversity, Evenness 2013 2014

PAGE 24

16 Graminoid richness was most influenced by whether or not a colony was currently occupied by prairie dogs Sites that had been plague extirpated, but were reoccupied by prairie dogs in 2007 and 2010 had lower total, native, and introduced graminoid richness compared to sites that were not currently occupied (not yet recolonized or never colonized) (Fig 3 ) If there was an effect of plague extirpation on graminoid richness, that effect could no longer be seen at the time of our sampling which commenced 7 and 4 years after the sites had been recolonized. Tilled sites without prairie dogs generally had lower native and higher non native graminoid richness (Fig 3 ) All tr eatment types that were disturbed by either prairie dogs or tillage had low richness ( 0 .981.0) of native graminoids. Introduced graminoid richness was highest on sites not currently occupied by prairie dogs regardless of tillage history (Fig 3 ) Both prairie dog occupation and tillage influenced forb richness. Tillage had the largest effect on native forb richness, with tilled sites having significantly fewer native forbs (7.13.5) than untilled sites (13.35.7) (Fig 4 ). Tilled an d untilled trans ects were not significantly different with respect to introduced forb richness ( Fig 4 ). Prairie dog occupation had a small effect on introduced forb richness, with currently occupied sites having lower introduced forb richness (4.02.4) than unoccupied si tes (5.32. 8). We saw higher numbers of native and introduced forb species across all transects regardless of tillage history in 2014 than in 2013, which we attribute to the anomalous large rainfall event in September 2013 (Fig s. 5, 6)

PAGE 25

17 Box plots demonstrating differences in total, native and introduced graminoid richness across site types (recolonized in 2007 after one year of extirpation, recolonized in 2010 after three years o f extirpation, not yet recolonized, and never colonized) and tillage history (tilled and untilled). Data presented are averages from the 2013 and 2014 early and late growing season sampling events. Different letters indicate significant differences between treatments at the P < 0.01 level.

PAGE 26

18 Box plots demonstrating differences in total, native and introduced forb richness across site types (recolonized in 2007 after one year of extirpation, r ecolonized in 2010 after three years of extirpation, not yet recolonized, and never colonized) and tillage history (tilled and untilled). Data presented are averages from the 2013 and 2014 early and late growing season sampling events. Different letters in dicate significant differences between treatments at the P < 0.01 level.

PAGE 27

19 Box plots demonstrating differences in total, native and introduced forb richness across site types (recolonized in 2007 after one year of extirpation, recolonized in 2010 after three years of extirpation, not yet recolonized, and never colonized) and till age history (tilled and untilled). Data presented are averages from the 2013 early and late growing season sampling events. Different letters indicate significant differences between treatments at the P < 0.01 level.

PAGE 28

20 Box plots demonstrating differences in total, native and introduced forb richness across site types (recolonized in 2007 after one year of extirpation, recolonized in 2010 after three years of extirpation, not yet recolonized, and n ever colonized) and tillage history (tilled and untilled). Data presented are averages from the 2014 early and late growing season sampling events. Different letters indicate significant differences between treatments at the P < 0.01 level.

PAGE 29

21 Plant Cover Our MANOVA results (Table A 3 ) indicate biologically relevant and statistically significant interactions between year and our other factors. Therefore, we report plant cover data for 2013 and 2014 combined and separately. Both prairie dog occupation and tillage had significant effects on total, native, and introduced graminoid cover. Sites that were currently occupied had significantly lower total, native and introduced graminoid cover than unoccupied sites (not yet recolonized and nev er colonized (Fig s. 7 9 ). There were no significant differences between the 2007 recolonized a nd 2010 recolonized sites (Fig s. 7 9 ) indicating that there were not detectable differences between a one and three year period of release from grazing pressure post plague Sites (b oth tilled and untilled) that had not yet been recolonized post plague had approximately eight times the total g raminoid cover of the currently occupied sites. (Figs. 7 9 ). However tilled sites without prairie dogs showed a higher introduced graminoid cov er (42.618.6%) than untilled sites (19.820.8 %) ( Fig. 7). Uncolonized and untilled sites had higher native gramino i d cover than any other site type (Figs. 7 9) In 2014, introduced graminoid cover increased slightly when compared to the previous year, but native graminoid cover was not effected (Table A4 ) (Fig s. 8, 9). In summary, prairie dog occupation a nd tilling together greatly reduce native graminoid cover. When prairie dogs are extirpated, graminoid cover is higher but because of higher introduc ed rather than native graminoid cover

PAGE 30

22 Forb cover was significantly affected by whether sites had been tilled or not and whether sites were currently occupied or not Introduced forb cover was highest (38.516.5 %) on sites actively grazed by prairie dogs (Fig 10 ) S ites that were not currently colonized showed a significantly lower amount of introduced forb cover regardless of tillage history ( 13.712.0 %) (Fig 10 ) Native forb cover, however, was not a ffected by current or past prairie dog grazing. Native forb cover was most affected by tillage with tilled sites having very low native forb cover (1.83.7%) (Fig 10 ). In 2014 forb cover showed more pronounced differ ences between occupied sites and unoccupied sites than in 2013 (Fig s. 11 12). Year had an effect on introduced forb cover (Table A5 ), driving up total forb cover, however native forbs showed only a slight increase.

PAGE 31

23 Box plots demonstrating differences in total, native and introduced graminoid cover across site types (recolonized in 2007 after one year of extirpation, recolonized in 2010 after three years of extirpation, not yet recolonized, and never colonized) and ti llage history (tilled and untilled). Data presented are averages from the 2013 and 2014 early and late growing season sampling events. Different letters indicate significant differences between treatments at the P < 0.01 level.

PAGE 32

24 Box plots demonstrating differences in total, native and introduced graminoid cover across site types (recolonized in 2007 after one year of extirpation, recolonized in 2010 after three years of extirpation, not ye t recolonized, and never colonized) and tillage history (tilled and untilled). Data presented are averages from the 2013 early and late growing season sampling events. Different letters indicate significant differences between treatments at the P < 0.01 le vel.

PAGE 33

25 Box plots demonstrating differences in total, native and introduced graminoid cover across site types (recolonized in 2007 after one year of extirpation, recolonized in 2010 after three years of extirpation, not yet recolonized, and never colonized) and ti llage history (tilled and untilled). Data presented are averages from the 2014 early and late growing season sampling events. Different letters indicate significant differences between treatments at the P < 0.01 level.

PAGE 34

26 Box plots demonstrating differences in total, native and introduced forb cover across site types (recolonized in 2007 after one year of extirpation, recolonized in 2010 after three years of extirpation, not yet recolonized, and never colonized) and tillage history (tilled and untilled). Data presented are averages from the 2013 and 2014 early and late growing season sampling events. Different letters indicate significant differences between treatments at the P < 0.01 level.

PAGE 35

27 Box plots demonstrating differences in total, native and introduced forb cover across site types (recolonized in 2007 after one year of extirpation, recolonized in 2010 after three years of extirpati on, not yet recolonized, and never colonized) and tillage history (tilled and untilled). Data presented are averages from the 2013 early and late growing season sampling events. Different letters indicate significant differences between treatments at the P < 0.01 level.

PAGE 36

28 Box plots demonstrating differences in total, native and introduced forb cover across site types (recolonized in 2007 after one year of extirpation, recolonized in 2010 after three years of extirpation, not yet recolonized, and never colonized) and tillage history (tilled and untilled). Data presented are averages from the 2014 early and late growing season sampling events. Different letters indicate significant differences between treatments at the P < 0.01 level.

PAGE 37

29 Box plots differences in total plant, bare ground, and litter cover across site types (recolonized in 2007 after one year of extirpation, recolonized in 2010 after three years of extirpation, not yet recolonized, and never colonized) and tillage history (tilled and untilled). Data presented are averages from the 2013 and 2014 early and late growing season sampling events. Different letters indicate significant differences between treatments at the P < 0.01 level.

PAGE 38

30 Litter and Bare Ground Cover The statistical patte rns across years showed the same trends when comparing treatment types, so we report findings averaged across years (Fig. 13). Litter and bare ground cover showed opposing trends. Bare ground cover was significantly higher on active prairie dog colonies re gardless of tillage or time since recolonization post plague (Table A6, Fig s 13 ) Conversely, litter cover was significantly higher on uncolonized sites again regardless of tillage and time since recolonization post plague (Table A5, Fig 13 ) Year also was a factor in bare ground cover, with bare ground cover in 201 4 decreasing on all tr eatments anywhere from 9. 6% to 51.7% when compared to bare ground cover in 2013. We attribute this year to year difference (Table A6 ) to the record rainfall in September of 2013. Litter cover also decreased across years for all but one treatment (Tilled, never colonized) with the lowest percentage decrease being 0.4% and the highest at 60.3%. Notable Species Relatively few species made up a large pro portion of cover on our sites The four most abundant species from each life history group ( n ative graminoid, i ntroduced graminoid, n ative forb, i ntroduced forb ) have percent cover shown in Tables 1 and 2 and the percentage of replicates per treatment type in Tables 3 and 4. Not surprisingly, the most abundant native graminoid ( Pascopyrum smithii ) had higher cover on sites that had not been tilled and that had been released from grazing pressure or never grazed. The most abundant introduced graminoid ( Bromu s inermis ) was highest on tilled sites

PAGE 39

31 not occupied by prairie dogs. Native forbs did not have much of a presence individually, with Heterotheca villosa been tilled. Lastly, Convolvulus arvensis cover was positively associated with prairie dog occupation. Active prairie dog colonies showed significantly higher cover of bindweed than the inactive sites ( P <0.001), with an average of 30% cover regardless of tillage (Table 1 ) The former prairie dog c olonies that had not yet been recolonized showed a greater ( P <0.01) cover of Convolvulus arvensis than the never colonized sites

PAGE 40

Table 1 : Mean Notable Graminoid Species Cover We noted the most commonly occurring introduced and native graminoids within our study. For each of those species, we found the average cover of that species for each treatment group. Tilled treatments had 13 replicate transects and untilled treatments had 17 replicate transects. Da ta shown here are averages of four sampling events (early and late season 2013 and 2014).

PAGE 41

33 Table 2 : Mean Notable Forb Species Cover We noted the most commonly o ccurring introduced and native forb s within our study. For each of those species, we found the average cover of that species for each treatment group. Tilled treatments had 13 replicate transects and untilled treatments had 17 replicate transects. Data shown here are averages of four sampling events (early and late season 2013 and 2014).

PAGE 42

34 Table 3 : Percentage Replicates of Notable Forb Species We noted the most commonly occurring introduced and native forbs within our study. For each of those species, we calculated the percentage of replicates in each treatment group that contained a particular species. Tilled treatments had 13 replicate transects and untilled treatments had 17 replicate transects. Data shown are average s of four sampling events (early and late season 2013 and 2014).

PAGE 43

35 Table 4 : Percentage Replicates of Notable Graminoid Species We noted the most commonly occurring introduced and native forbs within our study. For each of those species, we calculated the percentage of replicates in each treatment group that c ontained a particular species. Tilled treatments had 13 replicate transects and untilled treatments had 17 replicate transects. Data shown here are average s of four sampling events (early and late season 2013 and 2014).

PAGE 44

Plant Community Similarity Based on N onmetric Multidimensional Scaling NMDS plots were separated by season and year. The plots are 95% confidence intervals of the different treatment communities. For the early and late season in 2013 all treatments showed significant differences from each other (p<0.001). For the early season in 2014, all sites except the sites that were recolonized in 2007 and 2010 were significantly different ( P <0.001). The sites recolonized in 2007 and 2010 were similar ( P =0.44). The late season in 2014 showed the same trend as in the early season, with all sites being different except for the sites recolonized in 2007 and 2010 ( P =0.06).

PAGE 45

37 CHAPTER IV DISCUSSION NMDS plots showing community composition grouped by occupation history (recolonized in 2007, recolonized in 2010, not yet recolonized, and never colonized) at 95% confid ence intervals

PAGE 46

38 We studied plant communities after a major plague epizootic in Boulder County, Colorado, USA, to see if episodes of colony inactivity as a result of plague mitigate the vegetation change caused by prairie dogs. We found that whether a not a colony was currently occupied by prairie dogs and whether a not a site had been previously tilled or not had the greatest effects on the plant community m etrics we examined. We found no differences between the sites that had been released from prairie dog grazing for one versus three years because of the plague epizootic. This suggests that 1) history of tillage, even 30+ years ago, still impacts the vegeta tion composition of habitats, 2) seven years of release from grazing pressure is sufficient to return a prairie dog colony to a state somewhat comparable to a never colonized site, and 3) a three year release from grazing pressure did not result in lasting effects on vegetation once the site had been recolonized. On the urban Front Range of Colorado, plague seems to be a potential mitigating factor with respect to the plant community effects of prairie dogs. However, there are interesting interactions betw een the disturbances of tilling and prairie dog occupation in this urban habitat. Influence of Historical Tillage Historical tillage is an important factor in determining vegetation community composition. Tillage is known to alter soil attributes and plant communities, even 60 years after cessation (McGranahan et al. 2014, Brudvig et al. 2013, Jangid et al. 2011,

PAGE 47

39 Stylinski & Allen 1999). In this study, tillage severely decreased native species richness and this was most apparent on the sites that had n ever been colonized by prairie dogs. Native graminoid richness on the tilled, never colonized site was less than half that of the untilled never colonized site. The differences were even more pronounced with native graminoid cover, with the untilled sites having over three times more native grass cover than the tilled sites. Native forb cover and richness was significantly higher on all untilled transects regardless of prairie dog occupation. Tillage effects on unoccupied sites show higher introduced gramin oid richness and cover but not higher introduced forb cover. These findings suggest that tillage plays a vital role in the resilience of native species to other disturbances, such as prairie dogs. Some distinct agricultural legacies remain on some of the t illed sites, such as the planting of hay grasses like Bromus inermis Tilled sites also appear to have been influenced by plant propagule pressure from neighboring properties. Our finding that disturbance from tillage impacts plant communities is not surpr ising given that, in disturbed landscapes, invasive plants often outcompete native plant species (Fenesi et al. 2015, Hagen et al. 2014, Wang et al. 2013). However, our findings related to how the disturbance of prairie dogs interacts with the disturbance of tillage is of importance both to our understanding of how multiple change drivers interact and to our ideas about management of prairie dogs in urban areas. Relationship between Tillage and Prairie Dog Occupation History

PAGE 48

40 Seven years after being released from prairie dog grazing, untilled, not yet recolonized sites are approaching the never colonized sites, showing similarities in richness, diversity, evenness, and cover. Native graminoid cover and richness is on the rise, as is native forb richne ss. These recolonized sites have already surpassed their tilled equivalent in native richness and cover. On untilled sites, this indicates that prairie dog effects are not irreversible, at least not when periodic plague outbreaks continue to occur. Sites e xperiencing prairie dog extirpation will likely eventually revert to a state closer to the never colonized site. However, on tilled sites, native forb richness and cover does not increase after prairie dog exptirpation. Instead there is an increase in cov er and richness of introduced graminoids. For tilled sites that have not yet been possible these sites experience multiple propagule pressures after the prairie dogs a re gone when bare soil cover is high, thus increasing the number of species for a short time before the more dominant species colonize the site. It is also possible that the increase in species diversity resulting from prairie dog occupation is long lastin g. Three years of recovery before being occupied by prairie dogs did not seem to affect the vegetation community metrics in a different way than one year of recovery. Sites that were colonized the year after the plague epizootic, in 2007, were similar to the sites colonized three years later, in 2010. With the exception of native forb richness and cover, the effects of previous tillage could not be distinguished from the effects of prairie dog occupation. Introduced and native graminoid richness and cover, species diversity, and evenness were the same on the tilled and not tilled occupied sites.

PAGE 49

41 Whatever recovery the vegetation gains after three years of no occupation is quickly reversed once prairie dogs begin actively occupying the site. Relationship of Our Findings to Previous Work This study corroborates the findings from both rural and urban landscapes that prairie dogs decrease graminoid cover and prairie dogs increase forb cover (Agnew et al. 1986, Whicker and Detling 1988, Magle & Crooks 2008). However, in our urban sites, this increase in forb cover is not due to an increase in richness; instead the number of forb species declines, particularly the native forbs, and the cover of introduced forbs increases. Though historical tillage severely redu ces native forb richness, the effect of prairie dogs on sites that are not tilled also drives richness lower. Our work also augments the findings of Beals et al. (2014) which demonstrated that this region (urban Boulder County, Colorado, USA) has been ex periencing directional declines in native graminoid cover, increases in bare ground and increases in native and non native forbs and that prairie dog occupation amplifies those directional changes by four to 10 times. Like Beals et al (2014), we also found that prairie dog occupied sites had lower native graminoid cover, more bare ground, and higher forb cover. We can add that plague may be a mitigating factor that reduces the landscape level impact of prairie dogs and that vegetation changes are highly inf luenced by tillage history. Effects of Prairie Dogs and Tillage and Plague on Specific Plant Species

PAGE 50

42 We examined specific notable species and found interesting trends. The presence of the toxic morning glory species Convolvulus arvensis (field bindwe ed) seems to be closely associated with the prairie dog specific disturbance. Seven years after prairie dogs have been extirpated the cover of C. arvensis decreases by more than half. C. arvensis seems to have a harder time competing without the grazing pr essure on other plant species put forth by prairie dogs. Other introduced forb species such as Erodium cicutarium show some preference for active prairie dog sites, but are able to persist longer than C. arvensis after the prairie do gs are gone. We suggest that the relationship between C. arvensis and prairie dogs deserves further attention. Does the high primary production of C. arvensis allow prairie dogs to persist in high densities on sites that would otherwise be abandoned becaus e of lack of forage? What are the characteristics of C. arvensis that promote its success on prairie dog colonies? How does herbivory of C. arvensis influence its toxicity? Graminoid species increase in cover and richness after the extirpation of prairie dogs, but the recovery of community composition is affected by tillage history. Bromus inermis ( Smooth brome ), the most dominant introduced graminoid, shows a greater percentage of cover on tilled, unoccupied sites, than any other graminoid. On sites that Pascopyrum smithii (Western wheatgrass) makes up the most abundant graminoid, but has much less cover than the most abundant graminoid on the tilled sites. This could be due to the increase in species diversity and evenness on the until led, never colonized site, with many different species being represented in the cover data, instead of the site being dominated by one species.

PAGE 51

43 Management Implications In the urban to suburban Front Range of Colorado, plant communities are experiencing novel conditions that have a large impact on composition. Native prairie dogs are pushing the urban Boulder vegetation community towards a novel state faster than the ecosystems would be with the absence of prairie dogs (Beals et al. 2014). In this study a nd in Beals et al. (2014) it is shown that prairie dogs amplify trends not seen in rural settings: lower richness, evenness, and diversity. When tillage interacts with prairie dog grazing these two factors can drive richness, evenness and diversity even l ower. Prairie dogs face many challenges in the urban environment particularly movement restrictions due to anthropogenic features. Increased road density was positively correlated with an increase in the density of prairie dog burrows (Johnson and Collinge 2004, Magle et al. 2010). This increased burrow density and inability to migrate may result in overpopulation in colonies, which would increase grazing. This may be an explanation for the increase in bare ground cover on urban prairie dog colonies when compared to rural ones (Hartley et al. 2009). Hopson et al. (2015) also reports that urban prairie dog density is higher than on ex urban sites. On the windy Front Range, this may have devastating effects on the native seed bank, where seeds are blown away and the propagule pressure from introduced species takes over. Thus prairie dog overgrazing may have severe consequences for the re covery of native vegetation. Plant community composition is determined by availability of seed and vegetative sources and conditions (i.e. soil nutrients, moisture, temperature) for plant establishment, and stochastic

PAGE 52

44 community assembly processes. It is po ssible that directional changes in urbanization and climate could alter how prairie dog colony vegetation will respond to prairie dog extirpation in the future. We suggest further exploration into the effects of prairie dogs on soil erosion, soil nutrients and soil seed banks We also suggest further exploration into how the surrounding landscape matrix influences vegetation recovery after extirpation of prairie dogs. Further, extreme weather events (e.g. rainfall or drought) may provide temporary relief to the colonies greater bare ground cover and reduced vegetative cover or exacerbate the effects of prairie dogs on vegetation. During September 2013 shortly after the fall vegetation sampling, the Front Range experienced one of the most intense precipitat ion events in its recorded history. This record amount of precipitation is a probable explanation for the decrease of bare ground and litter in the spring of 2014 and the increase in plant cover. This also accelerated the spread of introduced grasses, of which many are spring annuals ( Bromus tectorum and B. japonicus ), into certain colonies, which may suppress native species even further (Fig. 15 ). This event also may have caused the observed increase in forb richness across treatments regardless of tillag e, increasing native forb richness by 41.6% (mean across transects) and increasing introduced forb richness by 56.4% (mean across transects). Without this increase, the effect of prairie dogs on forb richness may have been more pronounced.

PAGE 53

45 Urban Boulder is a novel ecosystem that is rapidly changing. Human factors, climate change, agricultural changes, and new diseases may accelerate the directional changes of plant communities. The extirpation of prairie dogs by periodic plague epizootics may slow down their effect on the vegetation communities and allow native species a chance to rebound. The results and implications from this study should be used in conjunction with others to maintain or improve the stability and function of urban grassland ecosystems. We contend that untilled sites provide the best place to maintain grassland ecosystems Figure 15 : Bromus after Flood Transect in May 2014 showing an explosion of Bromus tectorum where f lood waters passed. In September 2013, 46.1 cm of precipitation fell in Boulder, CO over a three day period Above the flood path persists native graminoids including Pascopyrum smithii and Hesperostipa comata.

PAGE 54

46 in urban Colorado without major re vegetation projects. This study also shows that a reduction in prairie dog numbers could benefit these sites. Prairie dogs in their current urban state may be better suited to historically tilled sites where the vegetation composition has already been altered and th e impact that prairie dogs have on native vegetation is lessened. Plague may offer a viable means of control without the expense to land managers. Plague outbreaks occur at intervals around every 6 12 years in rural settings (Augustine et al. 2008, Hartley et al. 2009, Cully et al. 2010) but it remains to be seen if these cycles are similar in urban areas. Should prairie dogs evolve resistance to the plague bacteria, the consequences to the vegetative communities will need additional studies.

PAGE 55

47 REFERENCE S Agnew W, Uresk DW, and Hansen RM. 1986. Flora and fauna associated with prairie dog colonies and adjacent ungrazed mixed grass prairie in western South Dakota. Journal of Range Management 39:135 139. Augustine DJ, Derner JD, and Detling JK. 2014. Tes ting for thresholds in a semiarid grassland: the influence of prairie dogs and plague. Rangeland Ecology & Management 67:701 709. Augustine DJ, Matehett MR, Toombs TP, Cully JF, Johnson TL, and Sidle JG. 2008. Spatiotemporal dynamics of black tailed prair ie dog colonies affected by plague. Landscape Ecology 23:255 267. Bangert RK and Slobodchikoff CN. 2000. The prairie dog structures a high desert grassland landscape as a keystone engineer. J ournal of Arid Environ ments 46:357 69. Beals SC, Hartley LM, Prevy JS, and Seastedt TR. 2014. The effects of black tailed prairie dogs on plant communities within a complex urban landscape: An ecological surprise? Ecology 95:1349 1359 Berg MD, Sorice MG, Wilcox BP, Angerer JP, Rhodes EC, and Fox WE. 2 015. Demographic changes drive woody plant cover trends and example from the Great Plains Rangeland Ecology & Management 68:315 321. Bonham CD and Lerwick A 1976. Vegetation changes induced by prairie dogs on shortgrass range. J ournal of Range Manage me nt 29:221 5. Boulder Monthly Climate Data: Precipitation (2015) Earth System Research Laboratory http://www.esrl.noaa.gov/psd/boulder/Boulder.mm.precip.html

PAGE 56

48 Brudvig LA, Grman E, Habeck CW, Orrock JL, and Ledvina JA. 2013. Strong legacy of agricultural la nd use on soils and understory plant communities in longleaf pin woodlands. Forest Ecology and Management 310: 944 955 Buckner DL. 1985. Point intercept sampling in revegetation studies: maximizing objectivity and repeatability. Paper presented at America n Society for Surface Mining and Reclamation Meeting, Denver, CO. 1985. Cid MS, Detling JK, Whicker AD, and Brizuela MA. 1991. Vegetational responses of a mixed grass prairie site following exclusion of prairie dogs and bison. J ournal of Range Manage ment 44:100 5. Coffin DP, Lauenroth WK, and Burke IC. 1996. Recovery of vegetation in a semiarid grassland 53 years after disturbance. Ecol ogical Appl ications 62:538 55. Collinge SK, Johnson WC, Ray C, Matchett R, Grensten J, Cully JF Jr., Gage KL, Kosoy MY, Loye JE, and Martin AP. 2005. Landscape structure and plague occurrence in black tailed prairie dogs on grasslands of the western USA Landscape Ecol ogy 2941 55. Coppock DL, Detling JK, Ellis JE, and Dyer MI. 1983. Plant herbivore interactions in a North A merican mixed grass prairie. I. effects of black tailed prairie dogs on intraseasonal aboveground plant biomass and nutrient dynamics and plants species diversity. Oecologia 56:1 9. Cully JF Jr., Johnson TL, Collinge SK, and Ray C. 2010. Disease limits p opulations: plague and black tailed prairie dogs. Vector borne and Zoonotic Diseases 10:7 15. Cully JF and ES Williams. 2001. Interspecific comparisons of sylvatic plague in prairie dogs. Journal of Mammalogy 82:894 905.

PAGE 57

49 Dearborn DC and Kark S. 2010. Mot ivations for conserving urban biodiversity. Conserv ation Biol ogy 24:432 40. Fahnestock JT, Larson DL, Plumb GE, and Detling JK. 2003. Effects of ungulates and prairie dogs on seed banks and vegetation in a North A merican mixed grass prairie. Plant Ecol ogy 167:255 68. Fenesi A, Gered J, Meiners SJ, Tothmeresz B, Torok P, and Ruprecht E. 2015. Does disturbance enhance the competitive effect of the invasive Solidao Canadensis on the performance of two native grasses? Biological Invasions 17: 3303 3315. Forr est S. 2005. Getting the story right: A response to V ermeire and colleagues. Bioscience 55:526 30. Friggens MM, Beier P. 2010. Anthropogenic disturbance and the risk of flea borne disease transmission. Oecologia 164: 809 820. Hagen D, Hansen TI, Graae BJ and Rydgren K. 2014. To seed or not to seed in alpine restoration: introduced grass species outcompete rather than facilitate native species. Ecological Engineering 64: 255 261. Hartley, LM, Detling JK, and Savage LT. 2009. Introduced plague lessens the effects of an herbivorous rodent on grassland plant communities. Journal of Applied Ecology 46:861 869. Hobbs RJ, Higgs E, Hall CM, Bridgewater P, Chapin FS, Ellis EC, Ewell JJ, Hallett LM, Harris J, Hulvey KB, Jackson ST, Kennedy PL, Kueffer C, Lach L, Lantz TC, Lugo AE, Mascaro J, Murphy SD, Nelson CR, Perring MP, Richardson DM, Seastedt TR, Standish RJ, Starzomski BM, Suding KN, Tognetti PM, Yakob L, Yung L. 2014. Managing the whole landscape: historical, hybrid, and novel ecosystems. Frontiers in Ecol ogy and the Environment 12(10):557 564.

PAGE 58

50 Hopson R, Meiman P, and Shannon G. 2015. Rangeland dynamics: investigating vegetation composition and structure of urban and exurban prairie dog habitat. PeerJ 3:e736. Jangid K, Williams MA, Franzluebbers AJ, Schmi dt TM, Coleman DC, and Whitman WB. 2011. Land use history has a stronger impact on soil microbial community composition than aboveground vegetation and soil properties. Soil Biology & Biochem ist ry 43:2184 2193 Johnston A. 1957. A comparison of the line in terception, vertical point quadrat, and loop methods as used in measuring basal area of grassland vegetation. Canadian Journal of Plant Science 37:34 42. Kinsinger FE, Eckert RE, and Currie PO. A comparison of the line interception, variable plot, and loo p methods as used to measure shrub crown cover. Journal of Range Management 13: 17 21. Magle SB and Crooks KR. 2008. Interactions between black tailed prairie dogs ( Cynomys ludovicianus ) and vegetation in habitat fragmented by urbanization. Journal of Arid Environments 72:238 46. Magle SB, Reyes P, Zhu J, and Crooks KR. 2010. Extirpation, colonization, and habitat dynamics of a keystone species along an urban gradient. Biological Conservation 143:2146 55. McGranahan DA, Daigh AL, Veenstra JJ, Engle DM Miller JR, and Debinski DM. 2014. Connecting soil organic carbon and root biomass with land use and vegetation in temperate grassland. The Scientific World Journal 2014:487563 Mellado M, Olvera A, Quero A, Mendoza G. 2005. Dietary overlap between prair ie dog ( Cynomys mexicanus ) and beef cattle in a desert rangeland of northern Mexico. J ournal of Arid Envi ronments 62: 449 458.

PAGE 59

51 Miller B, Reading R, Hoogland J, Clark T, Ceballos G, List R, Forrest S, Ha nebury L, Manzano P, and Pacheco J 2000. The role of prairie dogs as a keystone species: Response to S tapp. Conserv ation Biol ogy 14:318 21. Miller SD, and Cully JF Jr. 2001. Conservation of black tailed prairie dogs (Cynomys ludovicianus). J ournal of Mamma logy 82 : 889 93. Munson SM, Belnap J, and Okin GS. 2011. Responses of wind erosion to climate induced vegetation changes on the Colorado plateau. P roceedings of the National Academy of Sciences 108:3854 385 9. Norland MR, Veith DL, and Dewar SW. 1993. Standing crop biomass and cover on amended coarse taco nite iron ore tailing. Pages 385 415 in Proceedings of the 10 th Meeting, Spokane, Washington, May 16 19, 1993. American Society for Surface Mining and Reclamation, Champaign, Illinois, USA. Osborn B and Allan PF. 1949. Vegetation of an abandoned prairie d og town in tall grass prairie. Ecology 30:322 32. Pauli JN, Buskirk SW, Williams ES, Edwards WH. 2006. A plague epizootic in the black tailed prairie dog ( Cynomys ludovicianus ). J ournal of Wildlife Diseases 42:74 80. Prevey JS and Seastedt TR. 2014. Seas onality of precipitation interacts with exotic species to alter composition and phenology of a semi arid grassland. Journal of Ecology 102:1549 1561. Seasted t TR, Hartley LM, and Nippert J 2013. Case Study: Ecosystem transformations along the Colorado Fro nt Range: prairie dog interactions with multiple components of global environmental change. Pp. 142 149 In Novel Ecosystems: Intervening in the New Ecological World Order, First Edition, (e.d. R. J. Hobbs, E. S. Higgs, and C. M. Hall). John Wiley & Sons. Seastedt TR, Hobbs RJ, and Suding KN, 2008. Management of novel ecosystems: are novel approaches required? Front iers in Ecol ogy and the Environ ment 6 : 547 553

PAGE 60

52 and Collinge SK. 2008. Climate driven spatial dynamics of plague among prairie dog colonies. Am erican Nat uralist 171:238 48. Spasojevic MJ, Bowman WD, Humphries HC, Seastedt TR, and Suding KN. 2013. Changes in alpine vegetation over 21 years: are patterns across a heterogeneous landscape consistent with predictions? Ec osphere 4 : 1 18. Stylinski CD and Allen EB. 1999. Lack of native species recovery following severe exotic disturbance in southern California shrublands. Journal of Applied Ecology 36:544 554. USDA, NRCS. 2015. The PLANTS Database (http://plants.usda.gov, 11 October 2015). National Plant Data Team, Greensboro, NC 27401 4901 USA. Vermeire LT, Heitschmidt RK, Johnson PS, and Sowell BF. 2004. The prairie dog story: Do we have it right? Bios cience 54(7): 689 695 M, and Westbrooks R. 1997. Introduced species: a significant component of human caused global change. New Zealand Journal of Ecology 21:1 16. Wang XL, Wang YQ, and Wang YJ. Use of exotic species during ecological restoration can produce effects that resem ble vegetation invasions and other unintended consequences. Ecological Engineering 52:247 25.1 Whicker AD, and Detling JK. 1988. Ecological consequences of prairie dog disturbances. BioScience 38:778 85.

PAGE 61

53 APPENDIX Table A1 : MANOVA output for Evenness & Diversity Factor Df Lambda Approx. F Num df Den df Pr(>F) Tillage History (Tilled or untilled) 1 0.679 102.922 2 435 <0.001 Prairie Dog Occupation History (Never colonized, not yet recolonized, recolonized 2007, recolonized 2010) 3 0.649 35.048 6 870 <0.001 Year (2013, 2104) 1 0.929 16.543 2 435 <0.001 Season (early, late) 1 0.897 24.866 2 435 <0.001 Tillage History: Occupation History 3 0.952 3.577 6 870 0.002 Tillage History: Year 1 0.997 0.672 2 435 0.511 Occupation History: Year 3 0.974 1.958 6 870 0.069 Tillage History: Season 1 0.994 1.270 2 435 0.282 Occupation History: Season 3 0.992 1.291 6 870 0.276 Year: Season 1 0.994 1.218 2 435 0.297 Tillage History: Occupation History:Year 3 0.994 0.467 6 870 0.932 Tillage History: Occupation History:Season 3 0.993 0.532 6 870 0.784 Tillage History: Year: Season 1 0.998 0.403 2 435 0.669 Occupation History: Year : Season 3 0.994 0.385 6 870 0.889 Tillage History: Occupation History: Year: Season 3 0.990 0.694 6 870 0.654 Statistical output for MANOVA with dependent variables of evenness and diversity.

PAGE 62

54 Table A2 : MANOVA output for Richness Tillage History (Tilled or untilled) 1 0.498 90.237 5 448.0 <0.001 Prairie Dog Occupation History (Never colonized, not yet recolonized, recolonized 2007, recolonized 2010) 3 0.393 33.232 15 1237.1 <0.001 Year (2013, 2104) 1 0.771 26.684 5 448.0 <0.001 Tillage History: Occupation History 3 0.840 5.364 15 1237.1 <0.001 Tillage History: Year 1 0.980 1.814 5 448.0 0.109 Occupation History: Year 3 0.964 1.097 15 1237.1 0.354 Tillage History: Occupation History: Year 3 0.970 0.899 15 1237.1 0.565 Statistical output for MANOVA with dependent variables of total species richness, total graminoid richness, native graminoid richness, introduced graminoid richness, total forb richness, native forb richness, and introduced forb richness

PAGE 63

55 Table A3 : MANOVA output for Cover Factor Df Lambda Approx. F Num df Den df Pr(>F) Tillage History (Tilled or untilled) 1 0.669 26.478 8 429.0 <0.001 Prairie Dog Occupation History (Never colonized, not yet recolonized, recolonized 2007, recolonized 2010) 3 0.209 37.084 24 1244.8 <0.001 Year (2013, 2104) 1 0.757 17.224 8 429.0 <0.001 Season (early, late) 1 0.769 16.137 8 429.0 <0.001 Tillage History: Occupation History 3 0.711 6.485 24 1244.8 <0.001 Tillage History: Year 1 0.962 2.130 8 429.0 0.031 Occupation History: Year 3 0.850 2.991 24 1244.8 <0.001 Tillage History: Season 1 0.987 0.682 8 429.0 0.707 Occupation History: Season 3 0.920 1.504 24 1244.8 0.056 Year: Season 1 0.971 1.624 8 429.0 0.116 Tillage History: Occupation History:Year 3 0.934 1.239 24 1244.8 0.197 Tillage History: Occupation History:Season 3 0.976 0.441 24 1244.8 0.991 Tillage History: Year: Season 1 0.974 1.413 8 429.0 0.189 Occupation History: Year : Season 3 0.978 0.396 24 1244.8 0.996 Tillage History: Occupation History: Year: Season 3 0.972 0.510 24 1244.8 0.976 Statistical output for MANOVA with dependent variables of total cover, bare ground cover, litter cover, total graminoid cover, native graminoid cover, introduced graminoid cover, total forb cover, native forb cover, and introduced forb cover.

PAGE 64

56 Table A4 : ANOVA for Graminoid Cover Graminoid Cover Introduced Native Factor Df F Value Pr(>F) F Value Pr(>F) Tillage History (Tilled or untilled) 1 91.90 <0.001 42.45 <0.001 Year (2013, 2104) 1 14.74 <0.001 3.63 0.057 Prairie Dog Occupation History (Never colonized, not yet recolonized, recolonized 2007, recolonized 2010) 3 154.60 <0.001 56.94 <0.001 Season (early, late) 1 5.11 0.024 0.61 0.435 Tillage History: Year 1 0.97 0.326 0.00 0.985 Tillage History: Occupation History 3 26.32 <0.001 23.05 <0.001 Occupation History: Year 3 4.90 0.002 0.43 0.731 Tillage History: Season 1 0.05 0.825 0.04 0.832 Year: Season 1 0.02 0.878 0.79 0.372 Occupation History: Season 3 1.07 0.360 0.29 0.831 Tillage History: Occupation History:Year 3 0.31 0.822 0.21 0.892 Tillage History: Year: Season 1 0.17 0.678 0.11 0.736 Tillage History: Occupation History:Season 3 0.23 0.875 0.61 0.611 Occupation History: Year : Season 3 0.18 0.908 0.06 0.983 Tillage History: Occupation History: Year: Season 3 0.13 0.940 0.15 0.927 Statistical output for summary ANOVA with dependent variables of native graminoid cover and introduced graminoid cover.

PAGE 65

57 Table A5 : ANOVA for Forb Cover Forb Cover Introduced Native Factor Df F Value Pr(>F) F Value Pr(>F) Tillage History (Tilled or untilled) 1 0.001 0.922 109.85 <0.001 Year (2013, 2104) 1 21.69 <0.001 8.47 0.004 Prairie Dog Occupation History (Never colonized, not yet recolonized, recolonized 2007, recolonized 2010) 3 135.19 <0.001 1.57 0.196 Season (early, late) 1 25.65 <0.001 3.84 0.051 Tillage History: Year 1 0.26 0.613 5.21 0.022 Tillage History: Occupation History 3 0.64 0.593 1.88 0.132 Occupation History: Year 3 11.07 <0.001 1.25 0.289 Tillage History: Season 1 0.52 0.469 0.37 0.544 Year: Season 1 0.49 0.486 0.47 0.492 Occupation History: Season 3 2.97 0.032 0.45 0.717 Tillage History: Occupation History:Year 3 0.43 0.73041 1.93 0.123 Tillage History: Year: Season 1 2.20 0.138 0.25 0.615 Tillage History: Occupation History:Season 3 0.53 0.663 0.63 0.597 Occupation History: Year : Season 3 0.34 0.795 0.37 0.774 Tillage History: Occupation History: Year: Season 3 0.25 0.864 0.51 0.685 Statistical output for summary ANOVA with dependent variables of forb graminoid cover and introduced forb cover.

PAGE 66

58 Table A6 : ANOVA for Other Cover Other Cover Bare Ground Litter Factor Df F Value Pr(>F) F Value Pr(>F) Tillage History (Tilled or untilled) 1 0.40 0.529 2.80 0.095 Year (2013, 2104) 1 44.20 <0.001 10.28 0.001 Prairie Dog Occupation History (Never colonized, not yet recolonized, recolonized 2007, recolonized 2010) 3 291.79 <0.001 151.45 <0.001 Season (early, late) 1 47.90 <0.001 24.46 <0.001 Tillage History: Year 1 1.07 0.302 5.99 0.015 Tillage History: Occupation History 3 3.80 0.010 1.76 0.155 Occupation History: Year 3 2.82 0.039 3.27 0.021 Tillage History: Season 1 0.06 0.804 1.18 0.279 Year: Season 1 1.83 0.177 0.02 0.896 Occupation History: Season 3 5.86 <0.001 1.65 0.177 Tillage History: Occupation History:Year 3 0.46 0.709 0.15 0.933 Tillage History: Year: Season 1 0.06 0.805 0.15 0.698 Tillage History: Occupation History:Season 3 0.20 0.897 0.23 0.878 Occupation History: Year : Season 3 0.35 0.788 0.18 0.911 Tillage History: Occupation History: Year: Season 3 0.27 0.848 0.27 0.846 Statistical output for summary ANOVA with dependent variables of bare ground cover