Recipient Organization
VIRGINIA POLYTECHNIC INSTITUTE
(N/A)
BLACKSBURG,VA 24061
Performing Department
School of Plant and Environmental Sciences
Non Technical Summary
a. Problem statement(s) Environmental restoration is now a global priority to slow or reverse the combined effects of climate change, habitat loss, and associated biodiversity loss. The U.S. and dozens of other countries have pledged to restore more than 172 million hectares of degraded habitat by 2030 - an area larger than the size of Alaska (IUCN 2020). The United Nations has declared 2021-2030 the "Decade on Ecosystem Restoration" (UN Environment 2019). Depending on how restoration is enacted, this international effort could significantly and cost-effectively contribute to stabilizing atmospheric carbon concentration and could directly prevent thousands of species extinctions (Bastin et al. 2019, Brancalion et al. 2019, Strassburg et al. 2019, Fagan et al. 2020). Global restoration rhetoric often highlights trees and forests, but nowhere is the need for restoration greater than in the world's grasslands (Temperton et al. 2019). Temperate grasslands, like the North American prairie, represent the most threatened biome on Earth (Hoekstra et al. 2005). These ancient ecosystems are characterized by dominant, diverse, and often endemic grasses and forbs - a unique vegetation type that is maintained by fires, herbivores, and soil conditions that limit tree growth (Veldman et al. 2015a). Grasslands are at risk almost everywhere they occur from interrelated threats including agriculture, atmospheric carbon enrichment, defaunation, fire suppression, invasive species, and human habitat expansion (Bond and Parr 2010). In many cases, even well-meaning restoration efforts further degrade grasslands. For example, many misguided projects have planted and continue to plant trees in historically diverse grasslands - a process that converts unique, high quality ecosystems into low quality secondary woodlands (Veldman et al. 2015b, Veldman et al. 2015c). Unsurprisingly, restoring grasslands is fraught with complexity. Some of the biggest challenges are that the characteristic disturbances that maintain grasslands are hard to reproduce in human-dominated landscapes. For example, landscape-scale fires and migratory herds of bison once maintained grasslands in the southeastern U.S. (Noss 2013), but are hard to recreate in the modern context. Challenges to grassland restoration are compounded by over-fertilization; excess nutrients favor invasive species over native grasses and forbs (Daehler 2003). This project addresses the problem of how to restore biodiversity to northern Virginia's native grasslands. Although the Southeast is often thought of as dominantly a forested region and has sufficient precipitation to allow forests to develop in most places, early explorers in Virginia and the Carolinas described landscapes with head-high grasses and few trees (Noss 2013). Grasslands, savannas, and woodlands were maintained in some areas by natural processes such as thin soil, wetland hydrology, and grazing by large animals. For example, one of the first sightings of bison in eastern North America was made near the Potomac River close to Washington D.C. in 1612 (Hornaday 1889). Across the broader landscape, Native Americans maintained some open woodland and grassland by clearing and burning forests (Byrd 1737). Contemporarily, the Virginia Department of Conservation and Recreation describes several grassy plant communities occurring in the state including pine-oak woodlands, ultramafic woodlands and barrens, oak-hickory woodlands and savannas, and Piedmont prairies. Native grassland restoration has multiple co-benefits for agriculture and human societies, including (but not limited to) preventing species extinctions, reducing greenhouse gas emissions, supporting pollinator populations, and increasing crop pollination on farms (Paterson et al. 2019, Scott et al. 2019). In northern Virginia, constituencies for native grassland restoration include livestock owners, private land owners, and managers of public and private parks and conservation areas. Some of these stakeholders are supported in their projects by funding from the USDA, such as through the Conservation Reserve Program. This program is focused on northern Virginia because of the large constituency of interested land managers in that region.b. Relevance to advancing Virginia and the U.S.This project will advance Virginia and the U.S. in several ways:Technical knowledge gained will increase technical capacity for grassland restoration in Virginia, including on solar farms. Four new large solar farms have been approved for Virginia in the past year and will soon generate 192 megawatts of electricity on facilities spanning thousands of acres, and more solar farms are likely; last year Governor Northam signed an executive order aiming for Virginia's power to be carbon-free by 2050. The Virginia Department of Conservation and others are encouraging solar farms to vegetate their developments with native prairie plants. This project will inform what plant species to prioritize for restoration as well as technical strategies to establish native grassland on degraded sites.This project is designed based on input from Virginia stakeholders, including land managers focused on conservation and production. Land managers in Virginia seek information about how to restore native grasslands and meadows. Stakeholders include private landowners interested in conserving biodiversity, beef producers interested in incorporating native plants into pastures, and government agencies tasked with conserving natural areas. Several private and government stakeholders are contributing to this project by donating land, time, and other resources to the research program. This work will also inform ongoing efforts to conserve habitat diversity on farmlands, such as the USDA Conservation Reserve Program, which supports rural communities in Virginia.Students trained in this project will contribute to the large and growing U.S. restoration economy. As of 2015, the U.S. restoration economy directly employed ~126,000 workers and generated ~$9.5 billion in annual economic output (BenDor et al. 2015). This project will form part of the training for undergraduate and graduate students that will move into the restoration workforce.c. Approach To advance understanding of native grassland restoration in Virginia, I propose a two-tiered approach. First, I will document the plant species and soil attributes that characterize high-quality grasslands in northern Virginia (Culpeper, Fairfax, Fauquier, and Warren Counties), and I will compare these to plant communities and soil attributes in fescue fields located in the same region, which represent the main land use that stakeholders seek to restore to native grassland. Second, I will monitor vegetation recovery in a grassland restoration experiment replicated at multiple private and government-owned sites in northern Virginia. The experiment will test different methods of establishing and maintaining native grasslands.d. Anticipated outcomes and impacts. This project will produce new technical knowledge to support grassland restoration efforts in Virginia and elsewhere in the temperate grassland biome. I will disseminate the study to relevant stakeholders including private landowners, students, Virginia citizens, and government agencies through co-production of knowledge, formal and informal publications, and presentation to stakeholder groups. Knowledge gained through this project will inform grassland restoration undertaken in programs such as the Conservation Reserve Program. Ultimately, the knowledge gained from this project will contribute to biodiversity conservation and carbon sequestration, which are two major global challenges.
Animal Health Component
65%
Research Effort Categories
Basic
25%
Applied
65%
Developmental
10%
Goals / Objectives
This project has three major goals. Specific objectives are listed below each one.To characterize vascular plant diversity in native grasslands.Survey and analyze vascular plant composition in three remnant grasslands in northern Virginia.To determine if soil physical and biological conditions underly variation in plant composition.Collect soil and analyze its chemical and biological attributes for three high-quality grasslands.Analyze multivariate relationships between plant composition and soil characteristics.To elucidate the early outcomes of different restoration strategies.Establish and maintain a field experiment testing restoration and maintenance strategies for Virginia grasslands on three sites in northern Virginia.Monitor vegetation communities annually to evaluate changes under different management regimes.Analyze plant community changes in relation to remnant grassland plant communities as a target.
Project Methods
Study area. This study focuses on the northern extent of southeastern grasslands, in northern Virginia between Front Royal and Manassas (Fig. 2). The climate is humid subtropical (Köppen climate zone cfb) and within USDA plant hardiness zone 7a. The climate is characterized by warm, humid summers (mean max. temp. <90° F and cold winters with mean min. temp. >32° F). Annual precipitation is variable but averages 90-110 cm/year across the study area (1981-2010 averages; www.prism.oregonstate.edu). Underlying rock formations vary from Paleozoic metamorphic metabasalt and gneiss to Mesozoic sedimentary and igneous rocks. Soils are predominantly loams on gentle to steep slopes. Three high-quality grasslands were selected as reference sites based on consultation with experts, searches in the Virginia Working Landscapes (VWL) grassland plant survey database, and ad hoc observations. Sites were filtered by searching for occurrences of 35 species or genera of heliophytic grasses and forbs characteristic of northern piedmont prairies (Fleming et al. 2017). The species include grasses (Aristida purpurescens, Eragrostis spectabilis, Andropogon gyrans, Elymus virginicus, Schizachyrium scoparium, Sorghastrum nutans) and forbs (Liatris pilosa, L. squarrosa, Desmodium marilandicum, Lespedeza capitata, L. virginica, Chamaecrista sp., Senna sp., Sabatia angularis, Agalinis sp., Silphium trifoliatum, Chrysopsis mariana, Eupatorium hysoppifolium, E. rotundifolium, Packera paupercula, Coreopsis verticillata, Solidago nemoralis, S. juncea, Cirsium pumilum, C. discolor, Penstemon laevigatus, Platanthera lacera, Spiranthes sp., Asclepias tuberosa, A. viridiflora, Pycnanthemum incanum, P. tenuifolium, Monarda fistulosa, Euphorbia corollata, Polygala sanguinea). If a site had any of these species, we analyzed the total species list for the site, and if it seemed to be dominated by natives, and had a diverse herbaceous flora, we included it in the list of potential research sites. In addition, three sites were selected for a replicated grassland restoration experiment. All three sites are former pastures dominated by tall fescue. Sites range in size from 25-100 acres. The experiment is a randomized block design with three grassland establishment treatments, each of which is divided into two maintenance treatments.Establishment treatments include:Control - no establishment intervention.Spray - a single growing season treatment with glyphosate to weaken tall fescue and allow the seed bank to express itself.Spray + Seed - two growing season glyphosate treatments plus seeding a diverse mix of native plants. Seeds will be chosen based on availability from the list recommended for northern piedmont grasslands by the Virginia Department of Conservation and Recreation (https://cliftoninstitute.org/wp-content/uploads/2019/12/Suggested-Species-for-Meadow-Plantings-November-2019.pdf).Maintenance treatments will include:Mowing - Annual or biannual mowing to prevent woody species encroachment.Prescribed fire - Annual or biannual dormant season-controlled fire to prevent woody species encroachment. Vegetation surveys. To characterize vegetation at each site, we will use a modified Whittaker plot that has been adapted for sampling small, fragmented grasslands (Miller et al. 2015). At each site, we will establish one 2 x 50 m (100 m2) study plot. We will measure species richness and composition within this plot. To measure environmental characteristics, we will establish five 1-m2 quadrats evenly spaced throughout the larger plot. Within these subplots we will visually estimate the cover of each plant species. We will also estimate the cover of rock, mineral soil, organic matter, woody debris, and bryophytes, and we will estimate canopy cover using a spherical densiometer. We expect canopy cover to be non-negligible given the small size and adjacent forest edges in many sites. Vegetation sampling will be timed to visit each site during both the early (May-early July) and late (late July-August) blooming season to maximize species detections. Plants will be identified using the Flora of Virginia (Weakley et al. 2013). A reference collection will be compiled and deposited at the Massey Herbarium (VPI; Curator Jordan Metzgar has indicated his support for this research and specimen deposition). Soil sampling. We will collect and analyze soils from each site to characterize the biology and chemistry of grassland soils. We will collect five 21-inch (53-cm) soil cores evenly distributed across each study plot, and we will aggregate the samples for analysis. Biological analysis will be performed by Ward Laboratory using phospholipid fatty acid analysis (Frostegård et al. 2011). This technique yields 27 biological soil variables including total microbial biomass, the ratio of bacteria to fungi, mycorrhizal biomass, total microbial diversity, and predator: prey ratio (https://www.wardlab.com/plfa/). Chemical analysis will be performed by the Cornell University Soil Laboratory using their standard soil health test plus total carbon and nitrogen (https://soilhealth.cals.cornell.edu/testing-services/comprehensive-soil-health-assessment/). The ratio of carbon and nitrogen provides particularly important information about grassland soil fertility (Blumenthal et al. 2003). We will sample soils in late October and early November to capture late growing season and early dormant season microbial communities. Microbial soil communities are known to change seasonally (Bardgett et al. 1999). Data analysis. Data analyses will focus on describing the plant communities and soil attributes of reference grasslands and comparing these to experimental restoration plots over time. We will describe the composition of northern piedmont prairies by analyzing the frequency at which each plant species is recorded. We will use this analysis to define the dominant and indicative species of the community. We will compare the native prairie data that we collect to compositional information on fescue-dominated pastures from (a) VWL's database of plant surveys that have been done in fescue-dominated pastures in the northern Virginia Piedmont from 2010-2019 and (b) 280 1 x 1 m plots from our 2019 surveys of grassland restoration experimental plots in Fauquier and Warren Counties. Compositional analyses will be done using non-metric multidimensional scaling with Bray-Curtis dissimilarities, which account for abundance-based changes. Species not present in at least 5% of samples will be removed as per McCune and Grace (2002). NMDS solutions will only be interpreted when stress is <0.2 (Clarke 1993). Mean compositional differences among habitats and years will be assessed using permutational multivariate analysis of variance, and difference in variability will be assessed using analysis of multivariate homogeneity of group dispersions (Anderson 2006). Analyses will be performed in R using the vegan community ecology package (Oksanen et al. 2019). We will use generalized linear mixed-effects regression and ordination techniques (NMDS or PCA) to compare the biological soil attributes of native prairies to pastures that are dominated by tall fescue. We will also analyze the relationships between biological and chemical soil variables and site quality, as well as the effects of soil type (https://websoilsurvey.sc.egov.usda.gov/App/HomePage.htm) on site quality. To evaluate the influence of experimental restoration treatments on vegetation and soil over time compared to reference sites, we will use generalized linear mixed effects regression and multivariate trajectory analysis (De Cáceres et al. 2019). Regressions will be performed in R using the lme4 package (Bates et al. 2019).