Natural History of Ecological Restoration

Tag: Prairie

  • Reconstructing and reconnecting native habitats post-pipeline construction at the Litzsinger Road Ecology Center

    Reconstructing and reconnecting native habitats post-pipeline construction at the Litzsinger Road Ecology Center

    By James Faupel & Caity Sims 

    James Faupel is the urban ecology restoration supervisor at the Litzsinger Road Ecology Center (LREC) and Caity Sims is the Project Clear restoration coordinator. The LREC is a private, urban outdoor education site in the heart of Metropolitan St. Louis, Missouri that is managed by the Missouri Botanical Garden for educational programming with K-12 teachers and students, and supported by many local university-based researchers.

    Part 1: by James Faupel

    Anticipating Loss

    In September of 2018, when I took on the lead restoration ecologist role at the  Litzsinger Road Ecology Center (LREC), I was aware of the impending utility pipeline project (Project Clear) that would impact around a quarter of this education center’s restored habitats and the adjoining urban Deer Creek. The 15.8-hectare (39-acre) ecology center combines a mix of reconstructed bottomland prairie and restored riparian woodlands, with a stretch of urban creek in surprisingly good ecological condition. I knew this was a unique opportunity to set up a long-term study to examine the processes of reconstructing tallgrass prairie following major anthropogenic soil and landscape disturbance, a topic with little previous research, especially in urban areas.

    The 6.3 km (3.9 mile) long Deer Creek branch of Project Clear was slated to dissect a 0.8 km (0.5 mile) long swath of LREC’s restored habitats – a pretty significant amount of collateral damage. To be clear, the purpose behind Project Clear’s sanitary pipeline path is well intentioned. The Metropolitan St. Louis Sewer District’s (MSD) Project Clear is a six billion dollar, 28-year long initiative to improve water quality and alleviate many wastewater concerns related to the frequent sewage overflows of the legacy sewer system of the St. Louis region. Cleaner water in Deer Creek would mean healthier breeding and feeding conditions for a wide variety of wildlife that live in or rely on the water of the creek, as well as healthier adjoining ecosystems, and much healthier domestic water supplies for human residents of St Louis County. The collateral damage of the MSD Deer Creek branch was 6.3 km (3.9 mile) of riparian habitat was damaged, degraded or destroyed.

    In 2019, the Litzsinger Road Ecology Center (LREC)’s oldest prairie reconstruction and woodlands were dissected by the Project Clear pipeline path. The LREC is an outdoor educational facility of the Missouri Botanical Garden, and relies on functional ecosystem services to introduce children to the intricacies of nature. The pipeline project construction equipment would directly impact the LREC’s grounds and classes for three years, but its impact to the habitats here will be felt for generations. Photo by James Faupel.

    As discussed in a previous NHER blog post on the topic, urban prairie reconstructions are important tools in the toolbox of prairie education to the general public and in saving North America’s prairie remnants. Remnant prairies are invaluable fragments of a once vast grassland ecosystem of the Midwest and are now one of the most endangered ecosystems in the world. Getting the chance to expand the area covered by reconstructed prairie at an educational site like LREC was an opportunity that I did not want to miss. The downside was that the LREC restoration team would have only one year to gather baseline data before the sanitary pipeline construction began, and I also had to fit in learning all of the other ins and outs of my new role.

    Throughout 2019, with the assistance of LREC volunteers, interns, and partner researchers from local universities and organizations, we were able to gather a wide array of baseline data that could be used as a comparison in the years ahead, once the future restored prairies are older. In 2019, the previously restored habitats at LREC were 30 years old, among some of the oldest restored habitats in the St. Louis region. Restoration projects of this age are incredibly valuable for ecological research projects looking to better understand anthropogenic impacts on the environment around urban centers and to learn best practices for effective, long-lasting restoration efforts. They are priceless.

     During the growing season of 2019, we were able to survey all the vascular and nonvascular plant species growing in the proposed pipeline path and adjoining areas, as well as survey for birds, bees, spiders, ants, and algae thanks to the varied expertise of all involved. Baseline samples were collected of the upper soil horizons along the entire pipeline path, in addition to a complete inventory of all of the trees and shrubs that would be removed during construction so we could assess their value. The most significant baseline dataset is that of Elizabeth Hasenmueller and her Lab at Saint Louis University, who had been studying the water quality of Deer Creek at LREC for five years prior to the sanitary pipeline installation and are excited to continue this research, hoping to capture more data and test hypotheses concerning the hydrology and geochemistry of this urban stream as water quality improves with the removal of the sewage overflows from the creek.

    LREC volunteers assisted staff in a tree inventory of 746 native trees and shrubs ahead of the pipeline installation in 2019. The canopy trees were measured and their value assessed using the US Forest Service’s iTree database. The USFS’s estimated replacement value of the 289 mature canopy trees was half a million US dollars. This dollar value is in addition to the countless ways the ecological and environmental services provided by these trees could be valued. Photo by James Faupel.

    Deconstruction of Habitat

    On September 30th, 2019, construction vehicles arrived on the grounds of LREC, and began the month-long habitat removal on site. In what felt like the blink of an eye, a large section of one of the oldest prairie reconstructions in the St. Louis region, along with 746 individual native trees and shrubs, were gone and no longer providing any ecological or educational services. It was quick, like a bandage being ripped from the skin.

    Before and after the 0.8 km (0.5 mile) MSD pipeline construction path on the property of the LREC. 
    Left: October 2018. Right: August 2020 (Google Earth 2019, 2021).

    The pipeline installation process, from start to finish on LREC property, lasted nearly three years. Following the removal of all the preexisting habitat, work began on trenching the ground open, dynamiting the bedrock, installing the sewer pipe, reworking the now homogenized limestone bedrock/C horizon soil mix back into the trenches, restabilizing the streambanks from their stream crossings, and finally grading the soil out to pre-construction elevations. We knew from the beginning of their trenching process that we would ultimately be trying to grow native prairie plants in a “novel ecosystem” with soil conditions that would never co-occur with them in nature. 

    Throughout the construction process, there was plenty of time for planning how we would restore the land once it came back into our hands, as well as how we would monitor it long term. There is little research that has been done on the topic of prairie reconstruction following major anthropogenic soil disturbance. Initially we thought it would be beneficial to test many different methods of soil remediation ahead of the seeding of native prairie plants in different study plots. However, during the three year construction period, there were several major flooding events from the adjacent Deer Creek that changed our minds. The regularity and increasing frequency of these flash flood events of this urban creek made us take pause and the aforementioned research idea was scrapped. These events aren’t likely to lessen anytime soon, as the area covered by impervious surfaces in the surrounding watershed continues to grow. Future flooding events would introduce too many confounding variables to any kind of research that compared the outcomes of various soil remediation methods to make it a worthwhile investment of our time and resources.

    Left: Throughout the construction of the pipeline path (2019-2022), the path was impacted by Deer Creek flooding beyond its banks 11 times. It was completely covered by 4 major flash flood events as seen here. Right: The pipeline path a few months after construction had finished in late summer of 2022, covered in a growing season’s worth of early successional species. Here a contractor puts down the first annual cover crop to help break up the compacted soils and help hold the soil in place during rain events. Photos by James Faupel.

    In late summer of 2022, the property was handed back to the staff at LREC to begin the habitat reconstruction work. By this point in time, I had been promoted to a supervisor position at LREC and had many more responsibilities that took me away from being able to dedicate significant time to the Project Clear pipeline restoration project. Our two summer interns that year assisted with the second round of soil sampling for comparison to the baseline soil data and an inventory of all of the plants that naturally colonized the disturbed areas along the entirety of the pipeline path that year. For information from that botanical inventory, you can read our article that was published in the 2022 volume of Missouriensis. Our immediate reconstruction plans were to bring in a contractor that could quickly get annual cover crops down and hopefully germinate and root into the compacted soil before any fall or winter flooding events occurred. We also planned for a major seed collecting effort in 2023, so that we could sow a wide variety of early successional native species onto the pipeline path in late fall of 2023, or before the first snow of winter.

    With my role having been redirected away from the pipeline path reconstruction, I lobbied for a new staff member that could replace me and focus all their efforts on the reconstruction, research, and monitoring of this important project. Caity Sims was hired for the role of coordinator of the project in January of 2023 and hit the ground running. It has been an incredible year mentoring her throughout this process and I look forward to exploring the reconstructed prairie habitat she will be establishing and chaperoning with our amazing team of volunteers, staff, and interns over the next few years.

    Part 2: by Caity Sims 

    Before Reconstruction Began

    I was fortunate enough to witness the LREC before Project Clear as a summer intern in 2019. That summer, I assisted with collecting pre-deconstruction baseline data on trees and the algae found in Deer Creek. After my internship, I left St. Louis for Jonesboro, Arkansas to earn a master’s degree focused on botany in the Mississippi Alluvial Plain. Ready to apply my botanical and research skills in the working world, I returned to the LREC in January 2023 as the Project Clear restoration coordinator. I was shocked to finally see first-hand the impact the pipeline construction project had had on the riparian, woodland, and prairie ecosystems of LREC. What seemed in my memory like a maze of bottomland woodland was reduced to a scanty wooded edge surrounded by a sparsely vegetated brown strip of fallow land. 

    Before I joined the LREC team, a cover crop of common wheat (Triticum aestivum) had been seeded onto a majority of the pipeline path in winter 2022. Around the same time, a six-species cover crop mix, consisting of crop grasses, common pea, and field mustard (Brassica rapa) was seeded onto the southern section of the path. Cover crops mend the soil by adding nutrients, perforating the soil for better water infiltration, outcompeting weedy species, and preventing erosion of bare soil. 

    Unfortunately, the first round of cover crop mixes did not do well, and by mid-April the common wheat was barely a foot tall and had already gone to seed. The mixed species cover crop did not grow on most of the seeded area of the southern section due to flooding washing away a majority of the seeds. The path desperately needed a second cover crop at that point, so we hired contractors to seed the warm season crop, Sudan grass (Sorghum bicolor), onto the majority of the path again and a nine species cover crop mix onto the southern section. The Sudan grass did much better, growing to heights of over six feet in some areas. Although the mixed species cover crop seemed to do better, the extreme summer heat drying out the compacted soils resulted in depauperate mature plants. There were many areas directly over the pipeline that seemed unlikely to hold any vegetation at all. 

    Cover crop comparison between common wheat, taken on May 23, 2023 (A) and Sudan grass taken on August 3, 2023 (B); both photos were taken from about the same position on the path, but facing in opposite directions. Pictured in panel C are size comparisons of common wheat individuals on May 26, 2023, the two on the left were collected from the middle of the pipeline and the one on the right from the outer edge of the pipeline path. Note the numerous bare patches of soil in panels A and C. Photos: Caity Sims.

    Despite environmental conditions being heavily anthropogenically impacted, a good number of native early successional and bottomland species grew alongside the cover crops on the path. I was surrounded by many familiar disturbance-prone and riparian species, as my master’s thesis had focused on plants growing on Mississippi River islands. I recorded 310 species growing on the entire path after completing vegetation surveys in 2023. As I expected, some of the most common species were yellow foxtail grass (Setaria pumila), southern crabgrass (Digitaria ciliaris), barnyard grass (Echinochloa crusgalli and E. muricata), hairy seed crowngrass (Paspalum pubiflorum), late boneset (Eupatorium serotinum), and tall goldenrod (Solidago altissima), all of which thrive in disturbed environments. 

    As I work toward reconstructing the land to improve structure and function of the ecosystem for many native organisms, I need to quantify the results of this effort. To measure the changes in plant communities along the pipeline path over time, I established dozens of permanent vegetation monitoring plots along the path that will be monitored every summer. For each plot, I collect the presence and percent cover of each species in the plot. I am hopeful that by 2027, native prairie plants will constitute 75% of the flora and vegetation in every plot. 

    Summer interns, Karen (left) and Cicely (right), learning to identify plants from a vegetation monitoring plot on the pipeline path. Photo: Adam Rembert.

    Also found post-pipeline construction were two species new to Missouri:  Centaurium pulchellum, and Myosotis arvensis, and three new to St. Louis County records: Sesamum indicum, Cyclospermum leptophyllum, and Krigia cespitosa. All record species are not native to Missouri except K. cespitosa, which has been recorded from neighboring counties. It’s not uncommon to find new state or county records in disturbed areas. For more detailed information about those species, their identification, and potential explanations for their introduction to the LREC, check out my article in the 2023 volume of Missouriensis. 

    We have classified several different habitat types on the pipeline path including wet depressions and ephemeral ponds, rocky upland (i.e., “glade”), creek bank, woodland edge, and fallow ground. Due to a permanent pipeline access easement built and maintained by the Metropolitan St. Louis Sewer District (MSD), we cannot reconstruct the bottomland woodland that was destroyed. However, we can reconstruct a bottomland prairie on a majority of the path where the fallow ground with rocky and compacted soil are the most prominent. In the wider southern section, we will reconstruct a savanna that will include tree plantings outside of the MSD’s easement. 

    Going Forward

    Staff and volunteers collected seed throughout the 2023 growing season. We focused on collecting species with Coefficient of Conservatism (C-Value) ranks of six or below, which is consistent with the range of C-values observed in the 2022 vegetation survey of the path. The coefficient of conservatism (or C-scores) is an index ranging from 0-10, representing a species’ fidelity to intact remnant or long-restored habitats, as determined by local botanical experts. Species with higher C-scores are less tolerant of environmental degradation and increasingly restricted to high quality sites. Species with C-Values from 0-6 are likely to be found in disturbed and early successional environments, while species with C-Values from 7-10 are more likely to be found in near-pristine and remnant environments (Ladd and Thomas 2015). High C-Value species are harder to establish in restored and reconstructed habitats. 

    Summer and fall interns focused mainly on seed collection, cleaning, and organization to prepare for the 2023/2024 winter seeding. It was a monumental effort with amazing results. We collected seeds from 96 species in 2023 for a total weight of 49 kg (108 lbs) of mostly Pure Live Seed (PLS), with some chaff. In addition to collecting seeds, contractors have drilled 39 prairie species onto the pipeline path seeding a total of 16 kg (36 lbs) of PLS. We also added seeds collected in previous years that were housed in seed storage at the LREC. Including the 2023 collected seed, the PLS from contractors, and the previously collected seed, a total of 189 species with a total weight of 74 kg (164 lbs) was seeded on the pipeline path. To view the results of our seed collecting effort in 2023, please see our interns’ compiled end of season unpublished report.

    Boxes of Culver’s root (Veronicastrum virginicum) seed ready to be cleaned (A), fall intern, Cailyn, posing with seed to be added to the broadcasting mix (B), staff and volunteers mixing seed in preparation for broadcasting the pipeline path on January 5, 2024 (C), and staff and volunteers broadcasting the “glade” mix onto the most well-drained and rocky portion of the pipeline path on January 11, 2024 (D). Top photos by Adam Rembert, bottom photos by Caity Sims.

    I am hopeful that the seedings will be successful and signs of a prairie will begin to emerge this year. However, flooding and poor soil conditions are major issues on the pipeline path. I have already dealt with the aftermath of floods washing away cover crops and soil, and remaining soils lacking suitable nutrients and structure for plant growth. These are out of my control, so all I can do is accept them, keep trying, have patience, and not let frustration get the best of me. 

    In three years, I anticipate less introduced and early successional species as they fail to compete with the maturing perennial natives. I envision more pollinator observations on the former pipeline path as the prairie species begin to bloom. After the third growing year, we may be able to burn the young prairie and savanna. In 30 years, I hope to return to see restored prairie, savanna, woodland, and creek habitats functioning as home to a diversity of plants, insects, birds, reptiles, amphibians, mammals, and microbiota like the undisturbed LREC habitats do today. 

    Staff and volunteers broadcasting a spring and summer flowering seed mix onto the pipeline path on October 13, 2023. Photo: Tom Kibby.

  • Land Abandonment, Succession, and Restoration: The Wolf Run Grassland Restoration Project at the Missouri Botanical Garden’s Shaw Nature Reserve

    Land Abandonment, Succession, and Restoration: The Wolf Run Grassland Restoration Project at the Missouri Botanical Garden’s Shaw Nature Reserve

    By: Mike Saxton and Calvin Maginel

    Mike Saxton (mike.saxton@mobot.org) is the Manager of Restoration and Land Stewardship, and Calvin Maginel (cmaginel@mobot.org) is the Ecological Resource Scientist at Shaw Nature Reserve. 

    Since 1950, over 1-billon acres of agricultural land have been abandoned worldwide. In certain landscape contexts, unassisted spontaneous recovery of high levels of native biodiversity in abandoned fields is possible while in others, like the Midwest USA, fallow fields rarely develop into biologically rich habitats. To achieve the ambitious goals of the UN Decade of Ecosystem Restoration, land managers and ecologists need to better understand how to assist the regeneration and ecological restoration of these highly altered landscapes. 

    In 1925, the Missouri Botanical Garden purchased 1,300 acres of battered farm ground in Gray Summit, MO – approximately 35 miles (56 km) west of St. Louis – to escape the deleterious impact urban air pollution was having on horticultural collections in St. Louis City. The intended uses of this acreage were to: (1) propagate and grow plants, trees, and shrubs for the main Garden’s displays and (2) establish an arboretum focusing on woody plant collections. The site was officially named the Shaw Arboretum in 1933. After decades of development, many non-essential operations at the Arboretum were eliminated in 1958 and many fields were abandoned to allow spontaneous successional change. In 2000, the Shaw Arboretum was renamed Shaw Nature Reserve to reflect its contemporary mission to demonstrate, test, and inspire responsible stewardship practices through education, restoration, and protection of natural habitats and public enjoyment of the natural world. Today the Nature Reserve consists of 2,400 acres of varied habitat in various stages of restoration and revised management, including the use of prescribed fire.  

    Prior to European settlement, the natural plant communities and ecosystems of the area were fire adapted, open oak-hickory woodlands and xeric glades with gallery forests along riparian corridors. Post-settlement, woodlands were clear-cut with some woodlots left to passively regenerate while others were converted to row crop agriculture. 

    Wolf Run Grassland Restoration 

    In 2016, Nature Reserve staff set an ambitious goal to bring all 2,400 acres of the site into active management to promote native biodiversity by the year 2030. This effort will include restoring open pastures and former row crop fields, a relatively simple process. A much more challenging effort will be reclaiming 120 acres of old fields with 60+ year successional development, which is our current Wolf Run Grassland Restoration project. The 120 acre project area was initially “wasted farm ground” that had erosion gullies “where a freight train could pass without you seeing it”, according to August Beilmann, former Arboretum Director from 1941 to 1956. The entire project area was re-sculpted and smoothed by a bulldozer in 1953 and then converted to bluegrass (Poa pratensis). “Every piece of this land that looks so likely to be just right was laboriously rebuilt,” said Beilmann in a 1974 interview.

    Wolf Run Grassland Restoration project area in ca. 1945 showing open fields (light green) maintained through cattle grazing and mowing with trees occupying wet-weather streams and ditches (dark green). Photo: MBG Archives.

    Since 1958 when areas including the Wolf Run Grassland Restoration were removed from mowing and grazing, the site was encroached upon and became dominated by eastern red cedar (Juniperus virginiana), the non-native invasive shrubs Amur honeysuckle (Lonicera maackii), and border privet (Ligustrum obtusifolium), as well as slippery elm (Ulmus rubra), shingle oak (Quercus imbricaria) and ash (Fraxinus) species, the latter of which which are in severe decline due to the emerald ash borer. 

    These degraded woodlands had little native ground flora and were highly infested with non-native shrubs. Tree abundance and species composition had no historic analog. When setting ecological restoration goals for the area, staff determined that much of the site could not be managed as an open oak-hickory woodland, which would have existed at the site pre-settlement.  A new vision was needed. 

    Goals for the Wolf Run Grassland Restoration project

    • Establish a mosaic of 80 acres (~32 ha) of prairie, 15 acres (~6 ha) of savanna and 25 acres (~10 ha) of oak-hickory woodlands
    • Maximize native flora diversity and aggressively control invasive shrub species 
    • Manage with periodic, dormant-season prescribed fire 

    In 2021, Nature Reserve staff marked hundreds of native trees to retain including white, red, bur and black oaks (Quercus alba, Q. rubra, Q. macrocarpa, Q. velutina) and shagbark hickory (Carya ovata) and bitternut hickory (C. cordiformis). Drainages and wet weather streams were left with a 50 ft. (~15 m)  untreated buffer zone while a perennial creek flowing through the unit retained a 150 ft. (~46 m) untouched buffer. A logger removed unmarked trees from the area, with the commercial value of the timber offsetting the cost of the removal. Following US Fish and Wildlife Service recovery management guidelines for the Indiana bat (Myotis sodalis) — a federally endangered species — trees were only removed from November 1st to April 1st.

    Wolf Run Grassland Restoration project area pre-thinning (2021), approximately 60 years after land abandonment. Note the dark green areas are dominated by eastern red cedar (Juniperus virginiana), a native tree that rapidly colonizes abandoned or disturbed lands in the Midwest. Historically, this fire-sensitive tree species primarily occurred on rocky outcrops and bluffs that served as refugia from periodic fires that were common in the pre-European settlement landscape. Photo: ESRI
    Wolf Run Grassland Restoration project area post-thinning (2023). Photo: ESRI
    Forestry contractor equipment was used to remove most woody biomass greater than 4.5 in. (~11 cm) in diameter and left behind mostly small-diameter slash. A bulldozer was used to collect the debris into 600 piles that were subsequently burned. Photo: Mike Saxton

    The Restoration team has spent the last 9 months focusing on the removal of stumps for the project area. Stumps can be a substantial hazard for vehicles, equipment and staff safety. To seed native species and effectively manage the area for invasive species in perpetuity, the stumps must be ground down or cut flush to the ground.

    Skid loader mounted stump grinder removing stumps. Photo: M. Saxton

    Concurrent to this effort has been the site preparation step of chemically treating all of the invasive species and the disturbance driven annual vegetation that emerged post-land clearing [primarily fireweed (Erechtites hieraciifolius), mare’s tail (Erigeron canadensis), ragweed (Ambrosia artemisiifolia) and fox tail (Setaria pumila)]. This step is necessary because diverse, healthy native plant communities have not existed in these areas in more than 100 years. Consequently, there is no native seedbank to support the unassisted spontaneous recovery of native perennial herbaceous species in these highly degraded acres. 

    In areas where stumps have not been cleared, traditional equipment (tractor boom sprayers and UTV mounted spray rigs) for applying herbicide are ineffective. The Nature Reserve hired a contractor that specializes in drone-assisted aerial herbicide applications. The drone flies 12 ft. (~37 m) above vegetation and can self-navigate around trees and other hazards. The unit carries a total of 8 gallons (~30 L) and sprays approximately 3 gallons of herbicide per acre. The effective width of each pass is 25 ft. (~8 m). A single battery powers the drone, with a flight time of 7.5 minutes and a re-charge time of 6 minutes. When the herbicide tank runs out, the drone re-deploys to the fill up location, is refilled by the contractor, and then returns to where it left off. 

    Aerial drone sprayer used to eliminate undesirable vegetation. Photo: M. Saxton

    The last step after the undesirable trees have been removed, biomass/debris has been burned, the stumps have been ground and invasive species have been controlled, is the final ground preparation. Currently, in 2023, we are again smoothing out erosion gullies and clearing away the last remnants of woody debris with a bulldozer. This effort will ensure effective seed-to-soil contact when we sow native seed in January 2024 and will enhance our ability to successfully search for invasive species in the coming years by eliminating deep ruts and rills.  

    Above: Bulldozer in 1953 eliminating erosion rills in Wolf Run Grassland Restoration project area. Below: Bulldozer completing site preparation for native seed addition (2023).  Top Photo: St. Louis Globe Democrat, Bottom Photo: M. Saxton

    Native Seeding and Experimentation to promote Biodiversity Recovery 

    During the growing season of 2023, the Restoration team at the Nature Reserve has been feverishly collecting seed for this 2024 seeding effort. More than 1,100 lbs. (~500 kg) of bulk, milled seed from ~200 locally collected native tallgrass prairie and open oak-hickory woodland species will be used in the restoration planting. Additionally, these acres are enrolled in the Environmental Quality Incentives Program, EQIP – part of the Natural Resources Conservation Service, and we must purchase viability-tested seed to meet the minimum required specifications of the contract. The hand-collected seed together with the pure live seed (PLS) – percentage of viable seed in a seed lot – purchased from commercial vendors will provide us with ample species and volume of seed to effectively cover the 40 acres (~16 ha) to be planted this winter. 

    Concurrent with the preparation effort, we initiated a research study, which will help inform our work and the broader research community on plant recruitment amongst scraped soils that have been inoculated with mycorrhizal fungi and those that have not. This study includes paired species from the same genus that have different CC-values, short for coefficients of conservatism, which represents a species’ tolerance of environmental degradation, or its fidelity to intact remnant or long-restored habitats, as determined by local botanical experts. Ecologists generally expect species that are dependent on stable intact communities (higher CC-values) to be more reliant on mycorrhizae connections to establish and flourish. Species with high CC-values tend to establish poorly in restoration sites, which is one of the reasons to pursue this study. Some examples from the ten herbaceous pairs of native species include sedges (Carex bushii [CC = 4] and Carex bicknellii [CC = 10]), grasses (Sporobolus compositus [CC = 3] and Bouteloua curtipendula [CC = 7]), and forbs (Oligoneuron rigidum [CC = 5] and Oligoneuron album [CC = 9]).

    We added all 20 species at the same rate of pure live seed to provide each species an equal opportunity to establish. Initial analyses after one growing season indicate that low CC-value species germinated more successfully, producing more seedlings and greater percent cover than the high CC-value species, regardless of inoculation. We expect the addition of mycorrhizal fungi to have the greatest effects on species during the first couple of years after germination. If the mycorrhizae associate with the roots of the high-CC species more than the low-CC species, this may help them grow faster or be more resistant to future stress. Future monitoring will show us if there are long-term effects of inoculation.

    To check for updates on restoration activities and results from experimental studies, please visit our webpage.

    Barrels of hand-collected seed connected to a seed dryer, which pumps air through tubes into the barrels to eliminate mold & moisture. Photo: M. Saxton
  • Identifying regional and restoration species pools for the Ozark Highlands

    Identifying regional and restoration species pools for the Ozark Highlands

    Andrew Kaul is a Restoration Ecology Post-doc in the Center for Conservation and Sustainable Development working with Matthew Albrecht at the Missouri Botanical Garden, and Michael Barash is a junior Biology major at Washington University in St. Louis. Here they describe Michael’s undergraduate research on commercial native seed availability for woodland restoration.

    One of the largest barriers to restoration of degraded terrestrial habitats is availability of seed for use in reintroduction of desirable native plant species. Over the past few decades, the industry of native plant seed production has grown rapidly, but most native species in the US (and globally) are still not commercially available, and there can be strong biases in which types of species tend to be selected by seed producers.

    In ecological parlance, a “species pool” represents all of the species which can colonize and occupy a certain region. Not all of the species in a regional species pool are available commercially, which is how many restoration practitioners acquire seeds, so the subset of species pool that contains only the species that are commercially available in a given region is sometimes called the “restoration species pool”.

    For most ecosystems around the world, it is not well documented what proportion of the species pool is commercially available, and why these species have been selected for commercial trade. The few studies that have been conducted on commercial seed availability for restoration have found consistently that herbaceous (rather than woody) and rare species (rather than common ones) are less likely to be available, and there are strong taxonomic biases in which plant families are more represented. In the US, these studies have focused on open-canopy habitats with few trees, such as grasslands, rather than on more closed-canopy systems like woodlands and forests.

    An open rocky glade (left), and a glade to woodland transition (middle), a woodland understory (right; Shaw Nature Reserve, Gray summit MO).

    To address this information gap, we assessed the capacity of the native seed industry to support ecological restoration across terrestrial habitats in the Ozark region of the midcontinent USA. The use of seed additions to accelerate recovery of plant diversity in Ozark woodlands and forests is not well studied, and little information is available on how to best select species for reintroduction from seed. The specific goals of this project were to:

    1) Identify the species pool of native herbaceous (non-woody) vascular plants appropriate for restoration of glades, woodlands, and forests in the Ozark Highlands;

    2) Define the restoration species pool by identifying which of these species are commercially available;

    3) Quantify biases in this restoration species pool with respect to growth form, rarity, habitat affinity, and a few important functional traits;

    4) Identify candidate species which are not available from seed vendors, but should be a priority for seed production due to their importance for Ozark habitats.

    The spatial scope of this study is the Ozark Highlands, level III ecoregion 39, which covers all of Southern Missouri, as well as parts of Northern Arkansas, and NE Oklahoma.

    We began this project by developing a targeted species list of 1,178 herbaceous species native to upland habitats in the Ozark region, based on existing datasets from the Ecological Checklist of the Missouri Flora, the Flora of Missouri, and the Biota of North America Program (BONAP).

    We predicted there would be selection, implicit or explicitly, by seed producers for species based on their growth form, conservatism score wetness rating, rarity, and functional traits. Each species’ physiognomy (growth form), conservatism score (that is, sensitivity to disturbance), and wetness ratings (a type of habitat affinity) were included in the Missouri Checklist. For each species in our pool, we compiled data on two measures of rarity including a qualitative measure –  the official Missouri State conservation ranking, and a quantitative measure – the range size of the species in the US, as measured by the number counties in the nation where there have been recorded occurrences in the BONAP database. We compiled trait data on height of adult plants, and bloom timing and duration from species descriptions in the Flora of Missouri. For each grass species, we compiled data on photosynthetic pathway from published literature.

    We made predictions that species with certain growth strategies, traits, range sizes, and habitat preferences would be under- or over-represented in the pool of species produced by seed vendors. We predicted that compared to other growth forms, perennial forbs would be over-represented in the restoration species pool because the aesthetic value of restoration projects is often a high priority, and perennial forbs with their big flowers, are “showier” and will return year after year. Similarly, taller species, and those with a longer bloom period may be selected preferentially because their blooms are more noticeable. We expected that species which have a smaller range or are not abundant in sites where they do occur are less likely to be in demand by restoration practitioners, so are less likely to be commercially available. Based on this pattern we expected species with a lower conservatism score, larger range size, and higher conservation rank (less concern for conservation) to be more commonly produced by seed vendors.

    We predicted that species with a larger range such as pale purple coneflower (Echinacea pallida; map on the left), would be more likely to be available from at least one producer than species with a smaller range, such as the related yellow coneflower (Echinacea paradoxa; right). Maps are from BONAP, and light green areas denote counties where the species has been reported.

    The cottage seed industry for prairie plants has grown especially rapidly in recent years, so we expected that species which are generally found in more open habitats like glades, prairies, and savannas, would more likely to have been selected by at least one producer than the species which occur mostly in shady habitats like woodlands and forests. Similarly, since open habitats tend to have drier soils than shaded ones, we predicted there may be a bias toward species with a higher (drier) wetness rating. Many grass species that grow in open habitats have evolved a more efficient way of conducting photosynthesis under hot sunny conditions. There are fewer of these “warm-season” grasses than “cool-season” ones, but we predict that proportionally more warm-season grasses will be commercially available, because they are common in the prairie seed market.

    Inflorescences of big bluestem (Andropogon gerardii), and Indian grass (Sorghastrum nutans) can be seen at this glade to woodland transition at Victoria Glades Conservation Area in Hillsboro, MO. These warm-season grasses are dominant in many prairies and common in glades, but generally do not occur under the canopy of wooded areas.

    In order to test these predictions, we needed to compile information on which species in our pool are available from seed vendors. We identified ten seed vendors that are likely potential sources of seed materials for species native to the Ozark Highlands. These include five seed vendors within Missouri, four large regional seed vendors located in Iowa, Minnesota, and Kentucky, and one very large seed vendor that produces seed for regions all across all the US. We were able to get information on which species each vendor produces from their website, or if they did not have a website, then through personal communication. Many vendors sell a combination of seeds and potted plants, with most species only being available in one form or the other. For this study, we were only interested in seed products because restoration of herbaceous communities through seed additions is the most common and affordable approach.

    Based on preliminary analyses, we found that 501 (43%) species were commercially available from at least one vendor. We found the strongest trends supporting the prediction that species differ in their likelihood of commercial availability based on physiognomy or “growth form”. Perennial species were twice as likely to be available as shorter-lived annual or biennial species, and as predicted, forbs were better represented in seed vendor catalogues than grasses or sedges.

    We predicted that more common species would be better represented in the restoration species pool and our results somewhat support this prediction. Conservatism scores are assigned to species by expert botanists in each region, so they reflect how rare and how disturbance tolerant species are within local areas. In the US, these scores are often assigned at the state level. In order to avoid over-interpreting these designations, we binned scores into three groups including ruderal (0-3), matrix (4-6), and conservative (7-10) for use in our analysis. We found that “matrix” species with middling conservatism scores were more likely to be available than conservative or ruderal species. This may be because ruderal species can be somewhat weedy and may be expected to recruit into restored areas as volunteers. And on the other hand, highly conservative species may be difficult to grow for seed production, or have a small range, and thus limited restoration potential or demand. The state of Missouri has designations for the conservation concern of all native species. We found that species classified as “vulnerable” (S3), “imperiled” (S2), or “critically imperiled” (S1) were less likely to be available from seed vendors, as species classified as “secure” (S5) or “apparently secure” (S4). And finally, as predicted, we found that species with larger ranges are more likely to be commercially available.

    We expected species which mostly occur in open habitats with little tree cover to be more likely to be commercially available. We classified each species as belonging to one of three habitat affinity groups, being an open habitat specialist, closed habit specialist, or a generalist. We found no bias in species availability based on habitat affinity or based on the wetness rating for Missouri. Based on the prediction that the prairie-focused seed market would promote availability of warm-season grasses, we thought they would have greater proportional representation in the seed market, but we also did not find evidence for that prediction. Warm and cool season grasses were equally likely to be available, with about a third of all species belonging to each group being available.

    While we did not find that species with affinity to open habitats were more likely available from at least one producer than species from closed habitats, we did notice that the species which were sold by the most producers tended to be “prairie species” like butterfly milkweed (Asclepias tuberosa; left), which was available from 9 of the 10 vendors we surveyed, or stiff goldenerod (Solidago rigida; right), which was available from 8 vendors.

    Traits of species may also contribute to seed vendors’ interest in propagating them. We found evidence that within perennial wildflowers (forbs), species with a taller maximum height are more likely to be available. We also predicted that species with a longer potential bloom period would be better represented in the seed market, but surprisingly our data shows a negative relationship, where species that can bloom for many months are less represented in the restoration species pool. This pattern may be driven by differences between functional groups or plant families and deserves further investigation.

    The final goal of this project was to identify candidate species to recommend to seed producers as valuable for restoration potential. We identified such species based on the highly detailed descriptions provided in a keystone reference for this region, Paul Nelson’s The Terrestrial Natural Communities of Missouri (2005). This book describes the geologic, climatic, and natural features of natural community types in Missouri. We only considered habitats within the broad designations of forests, woodlands, savannas, prairies, and glades, and we narrowed our focus to only habitat types that occur within the Ozark Ecoregion. For each of these 37 Ozark habitats, this reference provides lists of plant species that are “dominant”, “characteristic”, or “restricted” to that habitat. We propose that a good starting place in assessing the capacity of the native seed industry to support ecological restoration across terrestrial habitats in the Ozark region is to examine whether all of the “dominant” plant species in habitats within the Ozarks are available from vendors. Of the 120 species identified by Nelson as “dominant” in Ozark habitats, 80 of them (66%) were commercially available. This is encouraging, since it is higher than the overall availability rate of 43%, however there are still 40 species which would be difficult for restoration practitioners to acquire without hand collecting from wild populations. This highlights how biases in the restoration species pool could potentially make assembling a high-quality seed mix more difficult, if the species for sale represent those which are easiest to cultivate, rather than being the ones which have the most biological significance to restoration.

    Birdfoot violet (Viola pedata) is classified as a dominant species for dry sandstone woodlands and is common on dolomite glades. Fortunately, we found it is commercially available from two vendors. Two other violets, wood violet (Viola palmata), and arrowleaf violet (Viola sagittata) are dominant in other Ozark habitats, but are not available from any of the vendors we surveyed.

    Here, we are only scratching the surface in terms of identifying ways in which the seed production industry may inadvertently be biasing the restoration species pool and consequently the diversity and composition of restored plant communities. In the future we recommend continued collaboration between seed producers, restoration practitioners, and conservation scientists, to identify the limitations of available seed stocks and better align supply and demand for native seeds. Most seed vendors do not label products at taxonomic designations below the species level. However, conservation goals are sometimes identified for subspecies or varieties. The extent to which these taxa are commercially available is difficult to assess. Additionally, many restoration projects call for seed from a local provenance, but obtaining information on ecotypes of native seed lots from vendors can be difficult. While nearly 40% of our species pool for restoration projects in the Ozark Highlands are commercially available, the proportion of those species that are available from an Ozark ecotype is likely much lower.

    We are currently preparing this project for publication. If you are interested in learning more, or have any questions, feel free to email Andrew (akaul@mobot.org).

  • A prairie resurgence?

    A prairie resurgence?

    James Faupel is the urban ecology restoration supervisor at the Litzsinger Road Ecology Center, a suburban outdoor education site managed by the Missouri Botanical Garden. The property is a mix of reconstructed bottomland prairie and restored riparian woodlands in St. Louis County, Missouri.  

    North American prairie remnants are invaluable pieces of a once vast grassland ecosystem, critical for the survival of so many plants and animals. Prairies are one of the most endangered ecosystems in the world, removed from existence by our agricultural development for crop production. According to the National Park Service, less than 1% of original prairies now remain in North America. The Missouri Prairie Foundation states that less than half of 1% of pre-settlement prairie is left here in my home state. These few remaining North American prairie remnants are vital seed banks for local ecotypes of thousands of native plant species, such as the federally endangered Mead’s milkweed (Asclepias meadii), and they are home to many species of animals that just cannot be found in any other type of habitat. Most prairie specialist species cannot survive once a fragmented prairie has been plowed or bulldozed under. Species such as the regal fritillary butterfly (Argynnis idalia) only occur on remnant prairies in Missouri and have not appeared in our human-made prairie reconstructions.

    A now rare, prairie horizon view. This one is visible at the National Tallgrass Prairie Preserve in Kansas. Photo Credit: James Faupel.

    Undiscovered remnant prairies are generally only spared thanks to practices such as consistent haying or grazing, and have sometimes been found protected in unused areas of historical sites, such as old cemeteries. Unfortunately, remnant prairies are now mostly found in rural settings far from the eyes of our growing urban populations. These sometimes small patches of prairie habitat do not have large dramatic features, such as mountains or canyons that draw vacationers’ attention from states away. Most remaining prairies are also no longer large enough to host their once charismatic herds of grazing megafauna, the American bison. The amazing views of these smaller, modern-day prairies must be experienced up close and personal. This is a problem if you want to educate the public on the importance of protecting these fragile habitats, that are now fragmented and spread far from each other across such a vast continent.

    Middle school children getting to know prairies up close and personal, at a prairie managed by the Missouri Botanical Garden. Photo credit: James Faupel.

    My home city of St. Louis was once 61% prairie pre-European settlement. The only remnant prairie still existing here is a small plot at Calvary Cemetery, which has had to have extensive restoration work done to remove trees, shrubs, and exotic invasive plants from smothering it out of existence.

    The prairie remnant at Calvary Cemetery in St. Louis City looking very open after some much needed restoration work, consisting of tree and shrub removal. Photo credit: James Faupel.

    Many organizations in St. Louis have begun to reconstruct prairies here over the years, to help regain this lost habitat for local wildlife and to be able to get these valuable grasslands back in view of the public. Some of the earliest prairie reconstructions in the Greater St. Louis Region started in the 1970s and 80s. Specifically within St. Louis City & County, this practice didn’t begin until the 80s. I have the pleasure of working on one of those prairies reconstructed in the 1980’s, at the Litzsinger Road Ecology Center, a prairie started and managed by Missouri Botanical Garden staff. The ecology center is a private education site dedicated to working with K-12 teachers, to improve upon their ability to engage their students in place-based education, using our local ecology as the framework.

    Recent work at the Litzsinger Road Ecology Center suggests that St. Louis prairies are making a comeback. Our 2021 spring intern, Lydia Soifer, began work on an independent research project looking at prairie habitat connectivity within St. Louis City & County. Through this project Lydia and I generated a count of 58 small-scale, urban prairie reconstructions managed by various entities within this highly populated area. There are also many more prairie reconstructions in the 7 surrounding counties within Missouri and Illinois.  

    With an increase of 58 small prairies slowly over 40 years, this may seem like a time to celebrate, but this prairie resurgence should not be taken lightly. Some of these new prairies are now at risk of failure. Prairie reconstructions cannot be left to their own devices in our modern, highly human influenced world. Investment in both ongoing habitat maintenance and the continued education of staff is a necessity, or these prairie reconstructions can quickly turn into fields of exotic invasive weeds or full of aggressive trees and shrubby growth. Even at 32 years old, the urban prairie I work at still needs continued maintenance to keep it a “native prairie”.

    Challenges facing urban prairie stewards range from intense seed pressure from surrounding invasive plants, severe runoff and volatile urban waterways, minimal funding and educational resources, fire & smoke restrictions that limit the chance of using prescribed fire, and heavy browsing from oversized whitetail deer populations. Many businesses and organizations outsource with private contractors for their prairie maintenance, which can have some beneficial and detrimental outcomes. There is not a constant visual presence overseeing the land they hold, but it can be much more affordable than permanent staff. Sometimes the only maintenance is periodic visits from dedicated volunteers. The decision to reconstruct a prairie should be well thought out and planned for optimal long-term care. Placement should be targeted for areas where a new prairie could help connect existing fragmented habitats to improve urban wildlife corridors.

    A prairie planting reclaims some space previously occupied by turf grass on a steep hillside at Bellerive Park, a St. Louis City Park. Photo credit: James Faupel.

    Are these human-made prairies working?

    So, it appears prairie reconstructions are gaining some ground within St. Louis and surrounding areas of the Midwest. How do we know if these reconstructions are being successful? What is success? Data collection of any kind is minimal to non-existent across these local sites, so assigning a value to these lands could be considered speculative at best.  

    When I transferred to the Litzsinger Road Ecology Center in 2018, I took notice of data previously collected there relating to pollinators (I have a passion for animal associations with native flora.). There were collection records from around the year 2000, of the now federally endangered rusty patched bumble bee (Bombus affinis), the endangered (IUCN Red List) Southern plains bumble bee (Bombus fraternus), and the vulnerable (IUCN Red List) American bumble bee (Bombus pensylvanicus). This is the only confirmed record of the rusty patched bumble bee in St. Louis, and its range has now shrunk considerably in recent times and can only be found much farther to our northeast. After surveying the reconstructed prairies at my work, I was able to find these two latter bumble bee species of concern. I was curious. Could more of the prairies around St. Louis be supporting the potentially declining populations of Southern plains and American bumble bees?  

    Left image – A Southern plains bumble bee queen nectaring on spider milkweed (Asclepias viridis). Right image – An American bumble bee worker visiting great blue lobelia (Lobelia siphilitica). Photo credits: James Faupel.

    Previously, not much was known specifically about the rare Southern plains bumble bee in the St. Louis region. According to the Checklist of the Bees of St. Louis, MO (Camilo et al. 2017) only two records within the city had been collected, in addition to the collection I mentioned earlier from the Litzsinger Road Ecology Center in St. Louis County. According to many local bee specialists, the American bumblebee used to be commonly seen all around St. Louis, but the Checklist notes only 3 sites that it was recorded at during their recent surveys. After spending a lot of my free time surveying St. Louis prairies, woodlands, and gardens over the last three years, I have found very promising results in the prairies.  

    Six of the larger and older prairie reconstructions in St. Louis City and County, with moderately rich species lists of native plants, were found to contain and support the Southern plains bumble bee, sometimes two to three years in a row. Many more of the prairies I visited supported the American bumblebee. Shutterbee, a local citizen science project I partner with, has recorded 3 Southern plains bumble bees and over a hundred American bumble bees from bi-weekly bee surveys in private home gardens in St. Louis City and County over the last two years. This shows there may be increased value in native plant gardens placed near prairies, for enlarging the foraging areas of bumble bees. 

    I am also beginning to see a trend with these two species’ floral choices. These two species of conservation concern seem much more reliant on native prairie plants than some of their more common bumble bee counterparts, that are flexible enough in their diets to visit many more exotic flowers. For the moment, this is just observational data, but at least it is showing that there is value in the hard work being done bringing these grasslands back to urban spaces. There are many other ways we could begin to assign value to man-made prairies, but more data collection needs to be done across the board on urban prairies. 

    All of these same prairie reconstructions containing milkweeds, blazing stars, sunflowers, asters, or goldenrods have also been recorded to attract in the majestic, migrating monarch butterfly (Danaus plexippus). Last December, the monarch was nearly put on the U.S. endangered species list. The US Fish and Wildlife Service put off this decision for a few years and will revisit it. If the well-known monarch butterfly does indeed get listed as endangered in the near future, will there be a vast new interest in prairie reconstruction? Will there be more investment in prairie protection and reconstruction from municipalities, utilities, corporations and other large land holders? If a quick surge of interest arises, education about these unique ecosystems and their management will be needed more than ever. 

    There are current opportunities to capitalize on the revitalized interest in the outdoors that the pandemic brought about, and with urban populations projected to outpace their rural counterparts in the future, native ecosystems will need to be brought to the people, to spark their curiosity and passion with nature. Without urban prairie reconstructions, we won’t be able to inspire the future volunteers, donors, conservation voters, and land stewards needed to care for and protect remnant lands. Urban prairie reconstructions are therefore integral in the process of preserving our rural remnant prairies, while also being ecologically biodiverse and important in their own right. We need more prairie reintroduced into North America and we need continued investment in their long-term care and monitoring. We aren’t just hoping to save endangered species, we are also hoping to save our continent’s most endangered ecosystem.

    A monarch butterfly visiting New England aster (Symphyotrichum nova-angliae) at the Donald Danforth Plant Science Center’s 5-year-old prairie reconstruction. Photo credit: James Faupel.