Spirit Mound
Explore, enjoy and protect the planet
|
|
Spirit Mound |
|
Explore, enjoy and protect the planet |
|
|
Part of the support for the Spirit Mound study described below was from the South Dakota
Chapter of the Sierra Club as well as the Living River Group of the Sierra Club.
Final Report Submitted by Karen L. Olmstead and Rustan Vote, Department of
Biology, University of South Dakota (605-677-6287, kolmstea@usd.edu) This report describes studies conducted at Spirit Mound Historic
Prairie in the summers of 2003 and 2004. Our goals
were to provide the State of South Dakota Game, Fish, and
Parks (SD GF&P) with an inventory of plant and selected insect species as
well as to establish baseline data on soil quality,
plant production, and variation in these around the
320-acre site. Funding for the study was provided by a SD GF&P Small
Wildlife Diversity Grant, The Sierra Club, the Spirit
Mound Trust, and the National Science Foundation-supported
Lewis and Clark REU site at the University of South Dakota. A prairie is defined as "an open community, dominated by
grass, and having less than one tree per acre."1 Before 1850, the great mid-continental grasslands, or
prairies, covered much of the interior United States. They stretched from southern Wisconsin to western
Montana, and from central Texas to Canada.2 The
tallgrass prairie covered 250 million acres of the
Midwest for thousands of years, including southeastern South Dakota. Today less than one percent of that original prairie remains, and it
is found mostly along roadsides, railroads, rivers, and on steep hills3.
The other 99% has been lost to agriculture and
development. These losses "exceed those reported for any other major
ecological community in North America."4 Prairies are valuable in terms of aesthetics,
as habitat, and as a means of erosion and flood control and other
ecosystem services. The challenge now is to determine mechanisms to
return the land to the way it was before the
introduction of the plow. Attempts at prairie restoration began in the 1930’s at the
University of Wisconsin-Madison Arboretum.1 Although
it has been practiced for several decades, prairie restoration
"involves much of the ongoing experimentation and trial of ideas characteristic
of a young, rapidly developing discipline"1
and the "success of many prairie
restorations is not well documented."5 Most
prairie restorations follow the same basic guidelines
(e.g see Kilde 6). Although initial results can be observed in just a few years,
it is anticipated that prairie restorations will take many years to restore
aspects of a native prairie, such as the plants and
invertebrate communities, soil structure, and vertebrate
populations.4 It is
possible and very likely that some aspects of the native prairie
system will never be completely restored because of changes in species complexes,
fragmentation, climate, fire frequency, or other important disturbance factors.4,5 Spirit Mound is one of only a few of the areas that we know with
certainty that Lewis and Clark visited. Many of the other areas along their route are now covered by
water because of the natural wandering of the Missouri River over the course of this
200-year period and because of the channelization of the river in most areas.
The
restoration at historic Spirit Mound is an effort to recreate the native tallgrass prairie that
once existed at this site. On April 14, 1986, the Lewis and Clark-Spirit Mound Trust formed
in an attempt to convert the site from agricultural fields, building sites, and a
feedlot, back to what it once was, a functional tallgrass prairie. The mission of the Spirit
Mound Trust (SM Trust) was to "acquire the 320-acre Spirit Mound site, restoring it as
closely as possible to the tallgrass prairie of 1804, and making it freely accessible to
the public." With the help of United States Senator Tim Johnson (D-SD), the National Park
Service (NPS), and the SD GF&P, the group petitioned for federal funds necessary to
purchase the site and reunite it under a single owner. Since the restoration began in
2001, the SM Trust, NPS, and SD GF&P, along with the University of South Dakota, have
remained involved with various aspects of the restoration. Spirit Mound is now managed by the SD GF&P in partnership
with the NPS and the SM Trust. The site was planted with a prairie seed mix purchased
from Ion Exchange in Harpers Ferry, Iowa in the fall of 2001. Four different seed
mixes were selected and planted in different areas at Spirit Mound (Appendix I).
The
seed mix that was used for the majority of the acres (227-acre Mesic Prairie Mix) was the
least diverse of the three main mixes, and is suited to the mesic habitat that is found
surrounding the mound. Species in this mix, which contained nine native grass and 19
native forbs, were chosen to emphasize tall and showy wildflowers. The second and most
diverse mix (10-acre Mesic/Xeric Mix) was planted in 50-foot strips on either side of
the walking trail, for a total area of 10 acres.8 This
mix included nine native grasses and 27 native forb species. The third seed mix (10-acre Wetter Mix) was for a small
area in the Northeast portion of the site, about 10 acres in size, which had been
identified as a wetland. This mix included species that are better adapted to seasonally wet
prairies, and was comprised of three grasses and 13 forbs. The final mix (40-acre
Xeric Mix, eight grasses and 26 forbs) was for the mound itself. This mix, with
plants better adapted to the steeper slopes and drier soils, was used for approximately
40 acres.8 Additionally, five grass species and 16 forb species were planted as plugs on
the east side of the mound in an area of approximately 10 acres. Seeds were planted at various seeding rates (Appendix I),
depending on the mix, in the fall of 2001 and spring of 2002. The fluffy seed was planted
using a Truax drill, while the heavy seed was not drilled, but was dropped by disconnecting
the drop tubes prior to the seed reaching the drilling mechanism (Mark Steck,
personal communication). The east slope of the mound was hand-planted with prairie plugs
in the spring of 2003 because of the steep slope in that area, which could not be
accessed by the drill. Interseeding with the original seed mixes was done as needed
throughout the prairie. Various sections of the site were mowed after plants became
established to prevent the weed species from setting seed. Additionally, herbicide was spot
sprayed throughout the growing season on problem weeds that could not be
effectively controlled with mowing. The western portion of the mound was burned in the
spring of 2003, and the eastern portion was burned in the spring of 2004 to stimulate
the growth of native plants and help control weedy species. Monitoring programs are a necessary part of any restoration
project to determine the success of the restoration, and to determine whether the current management
strategies are effective, or if they need to be modified. 1,9,10,11 Spirit Mound Historic Prairie was planted and is managed by the SD GF&P.
However,
SD GF&P does not have the necessary staffing to conduct a monitoring program of
the area and accepted a proposal for an inventory and monitoring program. The purpose
of this project was to provide the SD GF&P with information regarding the
establishment of planted species, a biological inventory, and to evaluate the likelihood of the
restoration’s success. This report is based in large part upon a graduate thesis12
resulting from the project. Several different aspects of prairie composition and production
were studied to determine the overall success of the restoration effort. We
addressed the following: What prairie and non-prairie species are present? Which of the
originally planted species became established? Do species richness and cover
increase over the season or between years? Do they vary among the different sampling
plots? Do they increase interannually? Does the ratio of forbs to grasses shift? Is
there spatial variation in productivity throughout the prairie, and is this tied to prior
land use histories? We expected plant species richness and total plant cover to
increase both throughout the season and interannually during the first few years of the
restoration, but to then decrease as the restoration progresses.5,13
With productivity, it is assumed that this will also increase over the season and
interannually, as the plants grow and become better established in the area.
Long-term changes in forb and grass
biomass are expected to vary as a function of burning, mowing, and other factors.1,
5,13,14 Plant Species Inventory, Richness, and Cover:
A plant species inventory was conducted to determine the presence/absence of native, invasive,
and planted species at Spirit Mound Historical Prairie as suggested by Masters.15
Plant sampling began in the summer of 2003. The prairie was divided into different areas
based on previous land ownership, and transects were established in each of these
areas. Transect locations were recorded using handheld GPS (Compass v.5.1b),
which was later used to relocate each transect. Standard plant ecology methods16
were used and plants were inventoried along thirteen permanent 100-meter line
transects, in seven different sites throughout the prairie (Figs. 1a and 1b).
Each transect
was subdivided into 10-m sections, and within each 10-m section a 1-m2
sampling frame was placed in front of a marker flag. Every plant species inside the 1-m2
quadrat was recorded. The percent of the area within the frame that was covered by living plants was
estimated by the observer and also recorded. Additionally, we used a modified pin-frame method to determine
the relative abundance of different plant species within each quadrat.17
The pin frame (or point frame) is a wooden frame one-meter in length
with ten evenly spaced holes across the top. A pin is inserted into each hole
and every species of plant (or bare ground) that contacts a pin is recorded,
as well as the number of pins that touch each species. In our study, we did not
differentiate between basal or foliar cover. Due to the
large number of quadrats to be sampled (> 130), we evaluated cover for 10 pins per plot
in contrast to the 100 pins/plot that is often recommended. Plants were identified using various field guides for the Great
Plains region (see Appendix II). Any plants that could not be identified were given
a unique designation as an unknown and digitally photographed. These images were then
entered into a database and were identified later in the growing season after
the plant had flowered. In August 2003, Dr. Ted Van Bruggen, Emeritus Professor of
Biology (USD) and expert on the flora of the Northern Great Plains toured Spirit Mound
Historic Prairie with us and confirmed our plant identifications. The plant species inventory was repeated once each month in
June, July, and August, with the exception of Transect 5. This transect was not recorded
after June because the area was sprayed with Roundup herbicide and then mowed to
control the weed problem in that area. In 2004 we added two more transects for a total of 14 (transect
5 still excluded), one on the west side of the site, which was previously unsampled, and
one in the feedlot area on the east side of the site (Fig. 1b). Transects were again
sampled once a month in June, July, and August. Data from transects within each year and
within each area were combined and referred to as ‘sites’. For example, the
two transects in the southwest plot were referred to as Site 1 in both 2003 and 2004.
Species richness was evaluated using a paired t-test comparing richness between years
2003 and 2004 within each site. For several analyses, plants were classified as prairie forbs,
prairie grasses, non-prairie forbs, or non-prairie grasses as determined by The Flora of the
Great Plains14 which was also used for all scientific nomenclature. We compared the percent cover, as determined using the pin-frame
method, of prairie forbs, prairie grasses, non-prairie forbs, and non-prairie
grasses using a three-way analysis of variance (ANOVA) with the effects of year, month,
and site, and their interactions as independent variables. When ANOVA’s were
significant, we used a Bonferonni t-test to compare percent cover among different
sites, months, and/or years. SAS Version 8 was used for all statistical analyses.
To evaluate the variability in relative cover among different
plant species, we evaluated the percent cover by site with a two-way ANOVA (month and year)
for species that were either desired prairie species or problem non-prairie species
that appeared to be common around the site. Native prairie-grass species analyzed
included Big Bluestem, Little Bluestem, Canada Wild Rye, Indian Grass, and Side-Oats
Grama. Non-prairie grasses analyzed were Green Foxtail and Yellow Foxtail.
Prairie
forbs included in the analysis were Wild Bergamot, Black-Eyed Susan, Stiff Goldenrod,
Hoary Vervain, Blue Vervain, and Gray-headed Coneflower. Non-prairie forbs included
Prickly Lettuce and Horseweed. We used a Spearman-Rank correlation to determine if our rough
estimates of cover (visually estimated percent area within the 1-m2
frame covered by plants) corresponded to the percent cover as determined by the
pin-frame method. Biomass Sampling: We sampled vegetative biomass on a monthly basis
to determine the relative production of different sites within Spirit Mound
Historic Prairie and the percentage of both grass and forb biomass present throughout the
season. For biomass sampling, the prairie was divided into five sampling
regions (Southwest, Northwest, Northeast, Creekside, and East Mound) that roughly
approximated the sites used in the transect sampling (Fig. 1b). Within each region,
four 1-m2 plots were haphazardly chosen, and all the above ground biomass was clipped
at the base using electric or manual clippers. The location of these four plots
within each site varied by month. The plant material was separated into grasses and forbs
(dead plant material was excluded), placed into paper bags, and returned to the lab
for drying. Biomass samples were dried to a stable dry weight (length of drying
varied with bulk of biomass and moisture present when collected). Samples were then weighed
and recorded. We conducted biomass sampling in June, July, August, and September.
In 2004, we used the same regions for biomass sampling, but instead sampled in
May, June, July, and August because the prairie production seemed to begin earlier
than in the previous year. We evaluated differences in grass, forb, and total biomass
among regions, months (June, July, August), and years using a three-way ANOVA. Soil Sampling and Land-Use Histories: Soil
samples were collected in October 2003 to determine the effect of differences in the soil
composition and previous land-use history on the types of plant species, both prairie and
non-prairie, growing in each sample location, and the overall restoration success of that
location. The following soil sampling protocol follows the recommendations of the Natural
Resources Conservation Service (Deron Ruesch, District Conservationist, NRCS, personal
communication). For soil sampling, the prairie was divided into nine areas that
roughly corresponded to those of the plant species inventory (Fig. 1c), with the
addition of one area that was sprayed with herbicide in 2003 (southeast corner, 5-1 in Fig.
1c). Two samples were taken in each of these areas. For each sample, 20 soil cores
were taken using a soilsampling probe that takes cores approximately 2.5 cm (1") in
diameter at two depths, 0-15.24 cm and 15.25-60.96 cm (0-6" and 6-24").
The 20 cores from each sample were mixed together to make one representative sample for each
depth. The samples were sent to the SDSU Soil Testing Laboratory for determination of
the levels of nitratenitrogen, phosphorus, potassium, pH, soluble salts, texture, and organic
matter. The resulting levels were averaged for each sampling plot. We were
unable to statistically compare these due to low replication (only two samples from
several cores that were mixed). Butterfly Sampling: Butterflies are considered good indicator
species for prairies and are of keen interest to the public.19,20
Therefore, butterfly sampling was conducted to determine which butterfly species are
present at Spirit Mound, with special attention given to the Regal Fritillary (Speyeria idalia Drury (Lepidoptera: Nymphalidae)), which
was observed and digitally recorded, but not collected. Butterfly species were inventoried at several sites throughout the prairie by direct
observation, photographs taken while doing other sampling, and by direct capture with
butterfly nets. Sampling was rather informal, with a few hours during the morning spent
collecting different species seen once each month. Standard collection procedures
were followed to collect butterflies. Representative individuals were placed into glassine
envelopes and mounted in a reference collection. Butterflies were only collected in the
summer of 2003. We reviewed species observed to determine which butterfly species were present and
whether there were any prairie specialists at the site. We also sampled terrestrial
invertebrates but these samples have not yet been sorted or identified to lowest
taxonomic level. Birds: As part of this project, we solicited support for an avian survey at Spirit Mound in
summer 2003. Methods used in avian surveys can be found in the recent publication in
South Dakota Bird Notes.21
The avian survey was not directly supported by this SD
GF&P contract.
Results and Discussion
Plant Species Inventory, Richness, and Cover: In 2003, over 100 different species of
plants were observed and identified at Spirit Mound, as well as a few unknowns that
could not be identified to species. Of those that were identified, the majority were native
tallgrass prairie species representative of the mixes that were planted (Table 1).
Also,
nearly all of the planted species in all mixes were observed, with only a few species
from the wetland mix not observed. The ‘227-acre mesic-prairie mix’, which was
planted on most of the site (Appendix I), contained only three grass species and three
forb species that were not observed: rough dropseed, prairie dropseed, porcupine
grass, showy goldenrod, smooth blue aster, and round-headed bushclover.
We observed a few new plant species in 2004, including large-flowered beard-tongue,
poison ivy, ten-petal blazing star, silky aster, sky blue aster, and smooth blue aster as
well as a few additional unidentified species. There were also a few species that were
not observed in 2004 that had been seen the previous year, including Rocky Mountain
bee plant, little knotweed, dwarf indigo, toothed spurge, false boneset, blue grama
grass, sand bur, and pepper grass.
In 2003 and 2004 combined, 132 total species were observed and identified, along with
some other species that could not be positively identified to species (Table 1).
In both
years combined, there were 97 unknowns, most of which were later identified to species
and included in the analysis. Out of all of the identified species, 61% (n=81) were
tallgrass prairie species native to this area as determined by Flora of the Great
Plains.14 Of these native tallgrass prairie species observed,
35% (n=28) were not a part of any of 7
the seed mixes used in the restoration. A total of 53 different species were intentionally
planted from all four mixes combined. Of these, 74% of the forbs (n=43) and 70% of the
grasses (n=10) were observed in at least one of the two years. The 14 species (total
from all mixes) that were not observed in either year included three grasses (rough
dropseed, prairie dropseed, porcupine grass) and eleven forbs, most of which were
from the wetland mix or were planted as plugs. From the three main seed mixes used
to plant the prairie (Xeric, Mesic/Xeric, and Mesic), 86% of the planted species were
observed.
These data indicate that many different prairie species can germinate in just a few
years, although some prairie species may take several years to germinate.1
The low
establishment rates of the planted wetland species suggest that the area that was
formerly a wetland may no longer be as well suited for growing plants adapted to these
wetter conditions as it once was.
No significant differences in plant species richness existed between years (paired on
site; t = 0.412, P > 0.05).
The percent cover of prairie forbs, prairie grasses, non-prairie forbs, and non-prairie
grasses were also evaluated. A three-way ANOVA indicated that the main effects of
year (F1,704 = 11.03, P < 0.001) and site (F5,704 = 15.5, P < 0.001) were the only
significant predictors of prairie-forb cover. Prairie forb cover was lower in 2003 (6.11 ±
0.579%) than in 2004 (9.48 ± 0.752%). In 2003, Sites 1 and 2 (Southwest and East
Mound, respectively) had the highest percentages of prairie-forb cover, which was less
than 15% in all months. All of the remaining sites had on average than 6% prairie forb
cover. Similar patterns were observed in 2004, with the highest percentages of prairie
forbs again in Sites 1 and 2, and Site 7 (new sampling area in 2004) also having a
higher percentage of prairie forbs than the remaining areas (Fig. 2).
A significant interaction of year by site existed for non-prairie forbs (F5,704 = 12.06, P
< 0.001), but the greatest percentage of explainable variation was due to differences
among sites (78.8%; F5,704 = 52.13, P < 0.001). In 2003, non-prairie forbs were most
abundant in the East Mound area (Site 2), and represented a higher percentage of the
cover than in all other sites in 2003 (Fig 2, Table 2). At Site 2, non-prairie forb cover
increased over the season to nearly 50% of the total cover in August. In contrast, nonprairie
forb cover remained the same or decreased over the season in the other sites.
Dominant non-prairie forb species (weeds) at Site 2 were pigweed and lambsquarters.
In 2004, non-prairie forbs were again most abundant in the East Mound area (Site 2),
followed by the Northeast, Northwest, and West mound sites (Sites 4, 5, and 7
respectively). Again, pigweed and lambsquarters were abundant as was kochia.
A significant three-way interaction (year*site*month; F10,704 = 2.53, P < 0.005) existed
for prairie-grass cover, however, site again contributed the greatest percentage of
explainable variation (78.24%; F5,704 = 110.92, P < 0.001). In all months in both years,
the mound itself (Site 6) had very little prairie grass cover (dominated by the non-native smooth brome.
The East Mound (Site 2), dominated by non-prairie forbs in both years,
also had significantly less prairie grass cover than all other sites except Site 6 (Fig. 2).
Site 1 (Southwest) had the greatest percentage of prairie grass cover, followed by Site 4
(Northeast) and then Sites 5 and 3 (Northwest and Streamside, respectively) which did
not statistically differ from each other. In Sites 1, 2, and 4, Canada wild rye was the
dominant prairie grass in both years. Prairie grasses accounted for the largest
percentage of cover in all areas except Site 2 and Site 6 in 2004.
A three-way interaction also existed for non-prairie grasses (F10,704 = 7.61, P
< 0.001)
but the greatest percent of explainable variation (84.15%) was again due to differences
among sites (F5,704 = 425.3, P < 0.001). On the mound (Site 6), there was a constant
dominance by the non-prairie grass smooth brome. Non-prairie grasses were absent at
Sites 1 (Southeast) and nearly absent at Sites 2 and 4 (East Mound and Northeast).
At
Site 5, green and yellow foxtail became quite abundant later in the summer in 2003
(51% season average), but were much less abundant in 2004 (5%).
We also examined the percentage of bare ground to evaluate increases in total cover.
A year*site interaction (F5,681 = 12.95, P < 0.001) resulted from the large decrease in
percent bare ground at Site 4 (Northeast) between 2003 and 2004 (Table 2).
This
statistical difference did not exist at the other sites, although the trend was always for a
reduction in percent bare ground from 2003 to 2004 (Fig. 2, Table 2).
As noted above, there were strong and consistent patterns of significant differences
among sites. In order to gain a greater understanding of what was going on within each
site, seven grass and eight forb species were evaluated individually within each site to
compare the variability in percent cover of these important plant species.
The species
that were chosen were either desired prairie species or problem non-prairie species that
appeared to be common in the restoration (see species listed on page 5 of this
document).
In Site 1 (Southwest), there were fewer non-prairie forbs overall in 2004 than in 2003
(Table 2), with the biggest differences in June. This is because there were many
different non-prairie species present in Site 1 in 2003, but by the following year, the
main non-prairie forb was horseweed, which seemed to replace many of the non-prairie
forb species seen the previous summer, which included pigweed and prickly lettuce, as
well as others.
There were two prairie-grass species that showed significant differences among months
or years at Site 1. There was a significant effect of month for Indian grass (F2,114 = 4.98,
P < 0.01) because the percentage of Indian grass decreased from June to July, but then
doubled by August. Side-Oats Grama showed a significant reduction between 2003
(8%) and 2004 (3%; F1,114 = 7.36, P < 0.01). Site 2, formerly a feedlot, was dominated by non-prairie forbs in both years, which
included pigweed, kochia, and lambsquarters. (It should be noted that after June 2004,
transect 14 was no longer used because it was sprayed and mowed to control
weeds, which were a problem in that area. In 2003, transect 13 was moved slightly after June
because of mowing as well.) The non-prairie grasses at Site 2 increased in July of
2003, and then decreased again in August. In 2004, there were low densities of nonprairie
grasses present in June, but they disappeared as the growing season
progressed (Fig. 2). The most common non-prairie grasses present both years were
green and yellow foxtail.
There was a significant difference between years in the amount of yellow foxtail present
at Site 2 (F1,134 = 3.85, P < 0.05). In 2003, yellow foxtail made up a very small
proportion of the total cover (2%), but was not present at all the following year.
There
was also a significant difference between years in the percent of prickly lettuce cover
observed in the former feedlot site (F1,114 = 33.25, P < 0.0001). The first summer,
prickly lettuce made up less than 3% of the total cover, but in 2004 it comprised nearly
18% of the total cover.
At Site 3 (Creekside), the prairie grass Canada Wild Rye had a significant month*year
interaction (F2,114 = 5.92, P < 0.01). No other species showed significant differences
among months or years at this site. In 2003, there was no Canada Wild Rye observed
in this location, but in 2004, this species was low in June, and then increased each
month, to over 21% by August. Overall, the percentage of prairie grasses decreased
slightly over the season in 2003, but increased by the end of the season in 2004.
The
increase in prairie grasses reflects the increase in prairie cordgrass and Canada wild
rye observed in August than in June or July.
There was a striking reduction in non-prairie grasses at Site 3 in August 2004.
This
reduction may have resulted from the flooding that occurs in this area. In late June of
2004, large amounts of rain caused the creek to flood and spill out of its banks.
This
caused many of the grasses to be bent down instead of standing straight up.
In July
when the sampling was done, many of the grasses were still prostrate, which may have
made it more difficult for us to sample species that were underneath the top layer of
grass. Further matting of the grass could have shaded out any species that were below
the topmost layer of grass. In August, the percent grass cover increased, but this time
there were more prairie grasses and fewer non-prairie species. This could possibly be
due to prairie grasses out-competing the other species after the flood, because there
was an increase in two of the main prairie grasses in that area after that time.
At Site 4, prairie-grass cover remained about the same throughout the summer in 2003.
In 2004, however, prairie-grass densities were low relative to non-prairie species in
June and then increased in July and August. This difference between years resulted in
a significant month by year interaction for the prairie grass species Canada Wild Rye
(F2,114 = 5.94, P < 0.01). In 2003, the percentage of Canada Wild Rye stayed about the
same each month (64-67%). In 2004, however, the amount of Canada Wild Rye
increased each month, starting at only 28% in June, but increasing to over 65% by the
end of the summer in August. In 2004 we began to observe more little bluestem and
Indian grass in July and August as well, although these increases were not statistically
significant.
The percentage of non-prairie grasses at Site 4 increased each month in 2003, mainly
due to the growth of yellow and green foxtail. In 2004, the amount of non-prairie
grasses started high, again largely due to green and yellow foxtail, but then disappeared
in July and August as the foxtail disappeared. Green foxtail showed significant
differences between years at Site 4 (F1,114 = 4.95, P < 0.05). The amount of green
foxtail at this site was higher in 2003 than in 2004 when it was not observed at all,
although it was relatively low in 2003 as well, making up less than 3% of the total cover.
The dominant non-prairie grass species in Site 4 was yellow foxtail. Yellow foxtail
showed a significant interaction of month*year in the Northeast site (F2,114= 41.45, P <
0.0001). The percentage of yellow foxtail was zero in June, and very low in July and
August (less than 2% in both months) in 2003. In June of 2004, there was a very high
percentage of yellow foxtail (27.25%), which then disappeared in July and August.
At Site 4, there were more non-prairie forbs in all months in 2004 than there were in
2003 (Fig. 2), and the amounts decreased each month as the summer went on in both
years. The increase in non-prairie forbs in 2004 is most likely due to two species,
prickly lettuce and horseweed, which were seen in much higher numbers the second
year. There was a significant month*year interaction for prickly lettuce at Site 4
(Northeast) (F2,114 = 6.36, P < 0.01). In 2003, the percent cover of prickly lettuce was
highest in July, and zero in the other two months. In 2004, the amount of prickly lettuce
was highest in June, and then decreased each month. Horseweed showed a significant
difference between years (F1,114 = 56.01, P < 0.0001). The total percent cover of
horseweed was higher in 2004 (5.6%) than in 2003 (0.4%). As noted earlier, the
percent bare ground at this site also decreased dramatically between 2003 and 2004.
At Site 5, prairie grasses started high and then decreased steadily in 2003, while nonprairie
grasses (mainly green and yellow foxtail) increased each month. In 2004, prairie
grasses started low, then increased slightly in July and August, while non-prairie
grasses (again green and yellow foxtail) started high and then disappeared (Fig. 2).
There was a significant year effect at Site 5 for Canada Wild Rye (F1,114 = 35.5, P <
0.001). The mean percentage was nearly double in 2004 (62%) when compared to
2003 (32%). After June of 2003, some of the little bluestem and Indian grass that had
been observed at this site was no longer present. Similar to Canada Wild Rye, in 2004
Little Bluestem and Side-Oats Grama gained in cover as the summer went on, although
these differences were not statistically significant.
Green foxtail showed a significant interaction of month*year at Site 5 (F2,114 = 15.23, P <
0.0001). In 2003, the amount of green foxtail increased each month, to almost 71% by
August. In 2004, there was no green foxtail recorded in any month at this site.
Yellow
foxtail showed a significant interaction of month*year as well (F2,114 = 13.92, P <
0.0001). In 2003, there was no yellow foxtail observed until August, and in 2004, it
started at its highest level in June (15.3%), and then disappeared by July.
At Site 5, overall, there were more non-prairie forbs in all months in 2004 than there
were in 2003 (Table 2), again due to the large amounts of horseweed and prickly lettuce seen the second summer.
There was a significant difference between years for the
total amount of prickly lettuce observed at the Northwest site (F1,114 = 12.45, P < 0.001).
Again the total amount of prickly lettuce increased in the second summer, from 0.1% in
2003 up to 5.4% in 2004. Horseweed showed a significant difference between years at
Site 5 as well (F1,114 = 32.10, P < 0.0001). The total amount of horseweed was again
higher in 2004 (13%) than in 2003 (0%).
At Site 5, there was a significant month*year interaction (F2,114 = 5.55, P < 0.01) for Wild
Bergamot which was the only prairie forb with significant differences among months or
years for any of the sites. This was explained by the fact that in 2003, there was no
Wild Bergamot present along our transects in any month at this location, but in 2004, it
was present in June, and then decreased in July and again in August. Wild Bergamot is
an early flowering species, so it is more common earlier in the summer, and then begins
to die or fade as the summer goes on, which explains this observation.
On the Mound (Site 6), non-prairie grasses dominated in both years, with the next
highest percentage of cover from non-prairie forbs. There were no significant
differences among months or years for any of the analyzed species at this site, probably
because most of the species analyzed did not occur at this location. Non-prairie forbs
present on the Mound included ragweed, creeping jenny, and black medick. Some of
the planted prairie plugs that were observed on the east slope of the mound included
hoary vervain, plains (whorled) milkweed, stiff goldenrod, and side-oats grama.
Most of
these were observed in 2004. The plugs may take a few years to become established,
so hopefully more of them will emerge and become successful in the coming years.
At Site 7 (West Mound), the most common type of plants observed were prairie
grasses, mainly Canada Wild Rye. There were also quite a few prairie forbs, which is
encouraging. Non-prairie forbs found at this site were mainly prickly lettuce and
horseweed. Amounts of grasses, prairie forbs, and non-prairie forbs remained relatively
constant throughout the growing season. This site is most comparable to the
Southwest (Site 1).
Visual estimates of cover and the cover estimates as determined using the pin-frame
were highly correlated (R = 0.39, P < 0.0001, N=685) indicating that the modified pinframe
method was consistent with our visual estimates of cover.
Results from the cover estimates corresponded well with what we observed within the
different sites. The Southwest area (Site 1) seemed to be the most successful in terms
of the establishment of prairie species and the lack of weedy forbs and grasses.
The
Northeast and Northwest sites were not as successful in the first year, but the
germination of planted seed began to catch up to the Southwest site by the middle of
the second summer. In 2003, these sites were dominated by non-prairie grasses which
were replaced by prairie grasses and some non-prairie forbs during 2004. Additionally,
there was more bare ground in these areas the first year than the second, indicating
that as time went on, more plants became established here. It was noted that the levels
of non-prairie forbs in the Southwest site decreased in 2004, which may possibly indicate that levels of non-prairie forbs in the northern sites will decrease next summer,
as these areas seem to be about a year behind in their establishment. Because these
weedy species most likely make up a large portion of the seed bed, it may take a few
years before they are completely out-competed. Non-prairie forbs are most prominent
in the East Mound area, and are probably preventing most of the planted prairie species
from becoming established. The same is true with the smooth brome on the mound.
It
may be difficult for the planted prairie species to become established when the area is
already dominated by another competing species. Further management may be
needed in these two areas for them to become successful.
Biomass Sampling: There was a significant year*month*plot interaction for total
biomass (F8,89 = 4.76, P < 0.0001; only data from June, July, and August analyzed).
In
2003, in the Northwest (Site 5), East Central (Site 2), and Southwest (Site 1) sites, the
total amount of biomass increased each month. In the Northeast (Site 4) and Creekside
(Site 3) sites, total biomass increased from June to July, but then decreased in August.
In 2004, the Northeast (Site 4), Northwest (Site 5), and Southwest (Site 1) sites showed
an increase in total biomass each month, whereas the East Central (Site 2) and
Creekside (Site 3) sites showed an increase from June to July, and then a decrease
from July to August. By August of the first summer, the highest total biomass was
observed in the East Central (Site 2) area, followed by Northwest (Site 5), Southwest
(Site 1), Creekside (Site 3), and Northeast (Site 4). In 2004, however, total biomass in
August was highest in the Northeast site (Site 4), followed by Southwest (Site 1),
Northwest (Site 5), East Central (Site 2), and the Creekside area (Site 3).
Overall, the
amount of total biomass increased each month from June-August (Fig. 3), as would be
expected in a newly restored prairie.
There was also a significant year*month*plot interaction for the biomass of grasses
sampled (F 8,90 = 2.81, P < 0.01). Most of this could be explained by the difference
between months, as the total amount of grass increased each month as the grasses got
taller.
There was no significant difference among years, months, plots, or their interactions in
forb biomass, although it was greater in all months in 2004 than in the same month in
2003. The ratio of percent grasses to percent forbs did not significantly vary between
years (F29,118= 1.28, P > 0.05) as is expected in a tallgrass prairie
restoration.3 The
reason for this is that it is probably still too early in the restoration (only third growing
season) for this shift to occur. In the first few years following the restoration, Canada
wild rye, gray-headed coneflower, ox-eye, and wild bergamot will become well
established and bloom, and may be persist as the dominant species for the first three to
four years. Three to five years into the restoration, big bluestem, little bluestem,
switchgrass, Indian grass, side-oats grama, purple coneflower, compass plant, and
white and purple prairie clover will become more common and begin to flower.
By this
point, most weed species will also have been out-competed and replaced by these longlived
perennials. Long-lived perennials, such as compass plant, prairie cinquefoil, and
sky-blue aster will become established six or more years after the initial restoration.
Some species, such as prairie dropseed and lead plant, may take 13 to 40 years
to establish themselves and often represent the final stage of succession in the restoration
process.3
Soil Sampling and Land-Use Histories: Results of the soil sampling showed differences
among the sampling sites in soil nutrients. Most striking was the high nutrient loading in
the East Mound site (Site 2) which was the former feedlot site. This site had higher
levels of nitrogen, phosphorus, potassium, and organics than any of the other sites.
The levels shown in Figure 4 are averages of the two samples from each site.
In the
East Mound site (Site 2), the maximum levels were even higher than those shown in the
figure. Many of the nutrient levels observed in the former feedlot site were several
times greater than the ideal levels for the successful establishment of prairie plants
which may explain why this area supports mainly weed species that are better able to
tolerate a broader nutrient range.22 Soil levels of phosphorus that are > 16 ppm are
considered very high for new seedings of grass.23 Above that level, farmers should not
add any more phosphorus fertilizer, because it this inhibits rather than promotes healthy
establishment of the grass species. Soil phosphorus levels measured in the east
mound area (former feedlot site, Site 2) were 640 ppm on average.
Potassium showed the same pattern. Soil test levels of 161+ ppm are considered very
high for new seeding of grass,23 and the feedlot area (Site 2) had potassium
measurements of 3240 ppm. The Creekside site (Site 3) had the highest salt levels of
any site, which may be attributed to erosion and salt deposition that occurs in this area
during flooding events. The high levels of potassium in the southeast site (not a plantsampling
site) could be the result of the different type of soil that is found there or due to
machine storage at that site (Ron Thaden, current land manager at Spirit Mound and
former Clay County extension agent, personal communication).
High soil nutrient levels have been found to be one of the most important abiotic factors
affecting the success of a prairie restoration.22 High nutrient levels are likely to promote
the growth and establishment of weedy perennials and competitive grass species which
often become the dominant plant types in the earliest stages of prairie restoration
because they are able to out-compete the desired prairie species and can tolerate these
higher nutrient levels.22 Rates of prairie seed establishment on former agricultural soils
where nutrient levels are high have been found to be very slow.22
Butterfly Sampling: Eleven different species of butterflies were observed and/or
collected during the summer of 2003 at the site (Table 3). The three most common
species recorded were clouded sulphur, red admiral, and cabbage white. All observed
butterfly species have been previously recorded in Clay County, South Dakota.24
Additionally, all species observed are generalist species. It should also be noted that
the regal fritillary was observed in the Northern portion of the prairie. The regal fritillary
is a more specialist species, preferring areas of tallgrass prairie.25
Birds: Over 50 species of breeding birds were observed at Spirit Mound in the summer
of 2003. For more information, please consult Carlisle et al. 2004.
Conclusions
To determine whether the initial stages of the restoration at Spirit Mound were
successful, we compared the species observed at the restoration site to the intentionally
planted species. With a majority of the planted species (74%) observed after only three
complete growing seasons, we conclude that the initiation of the Spirit Mound Historic
Prairie restoration has been very successful. The area has been converted from
buildings, a feedlot, and several independently owned agricultural fields which
supported one or two species of crops, along with a few weed species, to a tallgrass
prairie comprised of more than 80 different tallgrass prairie species. As the site
continues to develop, it is expected that many non-prairie species of grasses and forbs
will be out-competed and replaced by native prairie species, which are better adapted to
the tallgrass prairie ecosystem.3,6 Additionally, as the site remains out of cultivation,
weed species in the seedbed should become depleted and should present less of a
problem in the future. Rotational burning and interseeding with the original seed mix will
favor prairie species and keep non-prairie species at a disadvantage. Additional
species may also be added to the seed mix in the future to increase the overall diversity
of this prairie.
We expect plant productivity to become more homogeneous from year to year as the
prairie matures and the planted species become better established throughout the site.
The forb cover observed should decrease and be replaced by prairie grasses.3
This has
not yet occurred, since the forb production in 2004 was higher each month than in 2003.
Soil nutrient levels are not expected to change much throughout the prairie, with the
exception of the East Mound area, which should be mined (reverse fertilization, see
below) to attempt to reduce the levels of some nutrients and promote the growth of
prairie plants.
As the restored prairie site matures, we expect to see more species of animals taking
advantage of the area as habitat as well. In the few years since the restoration began,
several species of birds, insects, and mammals have already been observed.
A 2003
study indicated that more than 50 species of breeding birds have already become
established on the site.21 Additionally, from our own observations, it is evident that the
pheasant population has also increased since the restoration began, which is good
evidence that the vertebrate community is already increasing at Spirit Mound.
According the SER Primer on Ecological Restoration,11 a restoration can be considered
successful when “it will sustain itself structurally and functionally and demonstrate
resilience to normal ranges of environmental stress and disturbance.” Even though the
restoration at Spirit Mound appears to be going well, there are several factors to
consider with respect to the long term success of the project. First, nutrient levels at the
East Central site (Site 2, former feedlot) are extremely high. A recommended approach
for removing excessive soil nutrients is ‘soil mining’ or ‘reverse fertilization’ by planting
and harvesting crops at the contaminated site. By cropping a contaminated area, it is
expected that more nutrients, especially phosphorus, will be removed from the soil than are added as inputs by the
crop.22 However, reverse fertilization is not always
successful. “Increased off-take of nutrients within the crop has not always been
followed by significant reductions in soil concentrations” and “cropping has been
ineffective where large reductions in soil nutrients have been required for restoration.”22
Even if the East Central site (Site 2) is planted to alfalfa which is harvested and
removed from the site for a number of years, this does not guarantee that soil nutrient
levels will be reduced. Thus, we can not be confident that the former feedlot area will
ever have the conditions favored for the establishment of native prairie plants.
A second issue of concern is the frequency of fire at the Spirit Mound Historic Prairie
site. Because this site is fairly small (only 320 acres), only a small portion can be
burned each year to ensure that there is a large enough unburned portion to provide a
refuge for the vertebrate and insect species that inhabit the site. Therefore, the
recommended three to four year burn cycle for the entire prairie will probably not be
able to be followed. This could lead to a loss in overall species diversity, and especially
in shorter plants such as the forb understory and short to medium height grasses26 that
require the bare soil conditions created by fires for re-colonization and growth.
With nearly 75% of the planted species observed just a few years after initial planting,
this site seems to be on its way to successful establishment of prairie species.
With
continued management of the restoration site, especially in terms of reducing nutrient
levels in the former feedlot site, controlling and eliminating problem weed species with
mowing or herbicide applications, and rotational burning of different portions of the site,
we would expect this restoration to continue to be successful. Without this
management, at least in the initial stages of the restoration, however, the overall
success of the restoration may decrease, and the site may become dominated by only
one or a few species. As the restoration progresses, the extent of management
necessary should become less, and the prairie should become more self-sustaining,
perhaps with the exception of the necessary burning. Compared to the feedlot and
fields that comprised Spirit Mound just three years ago, hundreds of visitors are now
enjoying “a most butifull landscape” (Capt. William Clark, August 25, 1804).
Over the past two years, we have made several presentations
about the restoration efforts at Spirit Mound: Lewis &Clark NSF-Research Experiences for Undergraduates
Cohorts (both summers).
W.H. Over Museum in April and August, 2003.
Wyandotte County (Kansas City) Chamber of Commerce-August 2003.
Spirit Mound Trust, Oct. 2003
Missouri River Institute Symposium, Vermillion, SD, February 2004
USD’s Graduate Student Research Forum in April 2004.
August 25, 2004 Lewis & Clark bicentennial celebration organized by SD GF&P.
North American Prairie Conference, Madison Wisconsin, August, 2004.
SD Biocomplexity Conference, Sioux Falls, SD, August 2004.
SD NSF-EPSCoR Conference, Rapid City, August 2004.
Informally talk with people at the site on a regular basis.
Press coverage included:
A grass roots approach: project is restoring prarie at Spirit Mound. June 27, 2003.
Native prairie, birds and bugs thriving as Spirit Mound enjoys transformation. Sioux
City Journal. August 3, 2003.
Prairie restoration of Spirit Mound continues. Yankton Daily Press and Dakotan.
August 5, 2003.
Prairie restoration of Spirit Mound continues. Rapid City Journal. August 20, 2003.
Students restore Spirit Mound: Site provides students with research opportunities.
USD Volante. October 15, 2003.
Project Website: www.usd.edu/prairie
includes images of many plants and a table of plant species observed in 2003. Packard, S. and C. F. Mutel. "The Tallgrass Restoration
Handbook: For Prairies, Savannas, and Woodlands," Island Press, Washington DC,
1997.
Native American Seed . Where have all the prairies gone?"
http://www.seedsource.com/index.htm. 2002.
Accessed June 15, 2003.
Shirley S. "Restoring the Tallgrass Prairie: An
Illustrated Manual for Iowa and the Upper Midwest," University of Iowa Press, Iowa City, IA,
1994.
Sampson, F. and F. Knopf. 1994. Prairie conservation in North
America. Bioscience. 44:418-421.
Sluis, W.J. 2002. Patterns of species richness and
composition in re-created grassland. Restoration Ecology 10: 667-684.
Kilde, R. 2000. Going Native: A Prairie Restoration Handbook
for Minnesota Landowners. Minnesota Department of Natural Resources, Section
of Ecological Services, Scientific and Natural Areas Program. St. Paul, MN.
Spirit Mound Trust. Brochure. 2002 Spirit Mound Historic
Prairie: Recapture the spirit of Lewis and Clark.
Breyfogle, D.D. 2002. Spirit Mound Restoration Plan Update:
The Prairie Stewardship Project.
White, P.S. and J.L. Walker. 1997. Approximating nature’s
variation: selecting and using reference information in restoration ecology.
Restoration Ecology 5: 338-349.
Ehrenfeld, J.G. 2000. Defining the limits of restoration:
the need for realistic goals.Restoration Ecology 8: 2-9.
Society for Ecological Restoration website.
http://www.ser.org/reading. Accessed
September 13, 2003.
Vote, R. 2004. Monitoring the progress of a tallgrass
prairie restoration: Spirit Mound Historic Prairie. Unpublished M.S. Thesis, University of
South Dakota.
Howe, H.F. 1994. Managing species diversity in tallgrass
prairie: assumptions and implications. Conservation Biology 8: 691-704.
Barkley, T.M. 1986. Flora of the Great Plains. University
Press of Kansas. 1402 pp.
Masters, L. A. 1997. Monitoring Vegetation In: Packard, S.
and C.F. Mutel, editors. The Tallgrass Restoration Handbook: For Prairies, Savannas, and
Woodlands. Island Press, Washington DC. pp. 279-301.
Brower, J.E., J.H. Zar, C.N. Von Ende. 1997. Field and
Laboratory Methods for General Ecology. 4th Edition. WCB McGraw Hill.
Biondini, M. Scientific method and vegetation.
http://www.ndsu.nodak.edu/instruct/biondini/ars452/intro_eq.pdf
Accessed March 11, 2004.
SAS/STAT Users Guide, Version 8e. 2001. Cary, NC: SAS
Institute Inc.
Taron, D.J. 1997. Insects. In: Packard, S. and C.F. Mutel,
editors. The Tallgrass Restoration Handbook: For Prairies, Savannas, and Woodlands.
Island Press, Washington DC. pp. 305-318.
Ries, L., D.M. Debinski, and M.L. Wieland. 2000.
Conservation value of roadside prairie restoration to butterfly communities.
Conservation Biology 15: 301-411.
Carlisle, J.D., H.S. Hoff, and P.M. Mabee. 2004. Breeding bird inventory of Spirit
Mound Historic Prairie in Clay County – an area being restored to tallgrass prairie.
South Dakota Bird Notes 56: 32-41
Walker, K. J., P. A. Stevens, D.P. Stevens, J.O. Mountford, S.J. Manchester, and
R.F. Pywell. 2004. The restoration and re-creation of species-rich lowland
grassland on land formerly managed for intensive agriculture in the UK. Biological
Conservation 119-1: 1-18.
South Dakota State University. 2002. Fertilizer Recommendations Guide. USDA in
cooperation with SDSU.
Marrone, G.M. 2002. Field checklist of South Dakota butterflies.
Marrone, G.M. 2002. Field Guide to Butterflies of South Dakota. South Dakota
Game, Fish, and Parks, Pierre, SD. .
Leach, M.K. and T.J. Givnish. 1996. Ecological determinants of species loss in
remnant prairies. Science 273: 1555-1558.
Breyfogle, D.D. 2001. Spirit Mound Restoration Plan Update: The Prairie
Stewardship Project.
Table 1. Plant
species observed at Spirit Mound Historic Prairie during 2003 and 2004.
P=Prairie species, NP=Non-prairie species, U=Unknown, T=Woody plants. Shading
indicates intentionally planted species. Table 2. A comparison of 2003 and
2004 percentages for five different plant cover types at Spirit Mound Historic
Prairie. Common letters indicate no significant
differences between years within sites for each of the five cover types. Table 3. Butterfly species
recorded in Clay County24 and
whether or not they were observed at Spirit Mound Historic Prairie in Summer 2003. Family Common Name Scientific Names Observed at Spirit Mound? Figure 1a. Aerial photograph
showing different land-use site divisions at Spirit Mound Historic Prairie. East Mound is referred to as East
Central in the rest of this document. Map is oriented with North at the top.
Figure 1b. Approximate plant
sampling transect locations at Spirit Mound Historic
Prairie. Transects 14 and 15 were added in 2004. Figure 1c. Approximate soil
sampling locations at Spirit Mound Historic Prairie in October 2003. Map is
oriented with north at the top. Figure 2. Percent plant cover in
different research sites at Spirit Mound Historic Prairie. Figure 3. Plant biomass production
in 2003 and 2004 at Spirit Mound Historic Prairie. Figure 4: Soil nutrients at Spirit Mound Historic Prairie.
Samples were taken in October, 2003. Appendix I: Seed Mixes (lbs./acre)
planted at Spirit Mound Historic Prairie.8,27 Common Name Scientific Name 40-Acre Xeric Mix* 227-Acre Mesic Mix 10-Acre Mesic/Xeric Mix (Trail Mix) 10-Acre Wetland Mix 10-Acre Xeric Plugs Prairie Plugs (plugs/acre) Leadplant Amorpha canescens 0.05 0.01 0.1 0.0 0.0 Ironweed Vernonia faciculata 0.0 0.2 0.0 0.3 0.0 Silver Sage Artemisia ludoviciana 0.05 0.01 0.01 0.0 32.7 Butterfly Milkweed Asclepias tuberosa 0.1 0.01 0.5 0.0 32.7 Swamp Milkweed Asclepias incarnata 0.0 0.0 0.0 0.25 0.0 Whorled Milkweed Asclepias verticillata 0.0 0.0 0.0 0.0 32.7 Sky Blue Aster Aster azureus 0.01 0.0 0.01 0.0 32.7 Smooth Blue Aster Aster laevis 0.1 0.01 0.1 0.0 0.0 Silky Aster Aster sericeus 0.02 0.0 0.1 0.0 32.7 New England Aster Aster novae-angliae 0.0 0.01 0.1 0.05 0.0 Milkvetch Astragalus canadensis 0.05 0.0 0.1 0.0 0.0 Partridge Pea Cassia chamaecrista 0.25 0.0 0.5 0.0 0.0 White Prairie Clover Dalea candida 0.25 0.0 0.5 0.0 0.0 Purple Prairie Clover Dalea purpurea 0.5 0.0 0.5 0.0 0.0 Roundheaded Bushclover Lespedeza capitata 0.05 0.01 0.05 0.0 0.0 Purple Coneflower Echinacea angustifolia 0.2 0.1 0.2 0.0 32.7 Grey-headed Coneflower Ratibida pinnata 0.0 0.6 0.5 0.0 0.0 Upright Yellow Coneflower Ratibida columnifera 0.5 0.0 0.0 0.0 32.7 Stiff Sunflower Helianthus rigidus 0.1 0.01 0.05 0.0 32.7 Yellow Ox-Eye Heliopsis helianthoides 0.2 0.5 0.2 0.0 0.0 Rough Blazingstar Liatris aspera 0.05 0.01 0.1 0.0 32.7 Beebalm Monarda fistulosa 0.2 0.08 0.2 0.1 0.0 Large-flowered Beard-tongue Penstemon grandiflorus 0.05 0.0 0.1 0.0 0.0 Prairie Phlox Phlox pilosa 0.05 0.0 0.1 0.0 0.0 Prairie Cinquefoil Potentilla arguta 0.1 0.0 0.2 0.0 0.0 Black-eyed Susan Rudbeckia hirta 0.2 0.1 0.3 0.0 0.0 Grey Goldenrod Solidago nemoralis 0.05 0.0 0.1 0.0 32.7 Stiff Goldenrod Solidago rigida 0.1 0.3 0.2 0.0 32.7 Showy Goldenrod Solidago speciosa 0.05 0.01 0.1 0.0 32.7 Spiderwort Tradescantia bracteata 0.05 0.0 0.05 0.0 32.7 Hoary Vervain Verbena stricta 0.2 0.0 0.4 0.0 32.7 Blue Vervain Verbena hastata 0.0 0.04 0.0 0.2 0.0 Joe-pye Weed Eupatorium maculatum 0.0 0.0 0.0 0.250 0.0 Boneset Eupatorium perfoliatum 0.0 0.0 0.0 0.250 0.0 Sneezeweed Helenium autumnale 0.0 0.0 0.0 0.250 0.0 Bur Marigold Bidens cernua 0.0 0.0 0.0 0.250 0.0 Flat-topped Aster Aster umbellatus 0.0 0.0 0.0 0.2 0.0 Wild Licorice Glycyrrhiza lepidota 0.0 0.0 0.0 0.01 0.0 Monkey Flower Mimulus ringens 0.0 0.0 0.0 0.05 0.0 Cup Plant Silphium perfoliatum 0.0 0.5 0.0 0.5 0.0 Compass Plant Silphium laciniatum 0.0 0.05 0.0 0.0 0.0 Prairie Smoke Geum triflorum 0.0 0.0 0.0 0.0 32.7 Pasque Flower Anemone patens 0.0 0.0 0.0 0.0 32.7 Big Bluestem Andropogon gerardii 0.1 1.0 0.1 1.0 0.0 Little Bluestem Andropogon scoparius 2.5 0.5 2.19 0.0 32.7 Side-oats Grama Bouteloua curtipendula 1.7 0.5 2 0.0 32.7 Canada Wild Rye Elymus canadensis 0.2 1.0 0.2 0.0 0.0 Indian Grass Sorghastrum nutans 0.2 1.69 0.2 0.0 0.0 Rough Dropseed Sporobolus aspera 0.1 0.1 0.1 0.0 32.7 Prairie Dropseed Sporobolus heterolepis 0.1 0.1 0.1 0.0 32.7 Porcupine Grass Stipa spartea 0.1 0.1 0.1 0.0 32.7 Switchgrass Panicum virgatum 0.0 0.01 0.01 0.1 0.0 Prairie Cordgrass Spartina pectinata 0.0 0.0 0.0 4 0.0 *40-acre xeric mix: 3.53 lbs. forb
seed/acre, 5 lbs. grass seed/acre; 227-acre mesic mix: 2.56 lbs. forb seed/acre,
5 lbs. grass seed/acre; Trail mix: 5.37 lbs. forb
seed/acre, 5 lbs. grass seed/acre; Wetland mix: 2.66 lbs. forb seed/acre, 5.1
lbs. grass seed/acre. Appendix II: Field guides used in this
study. Jackson, L. and L. Dittmer. 1997. Prairie Seedlings Illustrated:
An Identification Guide. Volume 1: Twenty selected grasses and forbs. University
of Northern Iowa, Cedar Falls, Iowa. Johnson, J.R., and G.E. Larson. 1999. Grassland Plants of South
Dakota and the Northern Great Plains. South Dakota State University, Brookings,
South Dakota. Ladd, D. 1995. Tallgrass Prairie Wildflowers. Globe Pequot
Press. Guilford, Connecticut. Marrone, G.M. 2002. Field Guide to Butterflies of South Dakota.
South Dakota Game, Fish, and Parks, Pierre, South Dakota. Peterson, R.T., and M. McKenny. 1996. A Field Guide to
Wildflowers of Northeastern/North Central North America. Houghton Mifflin
Company, New York, New York.. Van Bruggen, T. 1983. Wildflowers, Grasses, and Other Plants of
the Northern Plains and Black Hills. University of South Dakota, Vermillion,
South Dakota. Vance, F.R., R. Jowsey, J.S. McLean, and F.A. Switzer. 1999.
Wildflowers of the Northern Great Plains Third Edition. University of Minnesota
Press, Minneapolis, Minnesota. |
