IMLS Woodland Restoration Project
Missouri Botanical Garden's Shaw Nature Reserve
In 2016, with funding from the Institute for Museum and Library Services (IMLS), we launched a science-driven woodland restoration project at the Missouri Botanical Garden’s Shaw Nature Reserve (SNR). This project was developed and implemented by a collaborative team of restoration ecologists at SNR and in MBG’s Center for Conservation and Sustainable Development (CCSD). Our project fulfills two important functions: 1) advancing scientific knowledge of how Ozark Border woodland plant communities respond to restoration treatments including invasive species control, thinning, fire, and seed addition; and 2) serving as a demonstration site for other landowners to visit, learn, and appreciate the process of restoring damaged woodlands.
|
Restoration Ecology of Ozark Border Woodlands
Like many woodlands in the Ozark Border Region, our 27 ha study site was exceedingly degraded with an impoverished native plant community and a long-history of human-activities. A mosaic of different land use-histories were represented, including areas that were once partially or entirely cleared and a small abandoned tree nursery comprised mostly of species native to eastern North America, but not to the Eastern Ozark Border ecoregion. The latter is a legacy of SNR’s former days as an arboretum and focus on living collections of non-native woody plants, a few of which have become serious invasive pests. Long-term fire-exclusion has also resulted in mesophication, a positive feedback cycle whereby shade-tolerant plants create microenvironmental conditions (lower light levels, cooler temperatures, higher humidity, and less flammable litter) that are less favorable to fire-tolerant and more light demanding native species. In turn, this led to a dense infestation of non-native shrubs and vines (most notably bush honeysuckle, privet, and winter creeper), especially along the flood zone of a small, wet–weather stream that bisects the site.
When restoring ecosystems, restoration ecologists rely on historical ecological data and the biophysical setting to set benchmarks for the intended ecological function and community assemblage. A less disturbed portion (central subunit) of our project site was wooded in historical (1941) aerial photos and a few scattered large, open-grown oaks (wolf trees) remain in this area, providing clues this site once supported open oak-dominated woodland (see left figure). Yet, the impoverished herb-layer community provided little indication of the diversity that likely once existed at this site.
A majority of the site consists of gentle sloping terrain (<15%), moderately drained and slightly acidic to neutral silt-loam soils (pH: 6.3–7.1), an east– or south–facing aspect, and maps to a Loamy Upland Woodland based on the USDA–NRCS Ecological Site Type classification. Accordingly, the reference ecosystem state is a fire-maintained and moderately open oak woodland (65-85% closure) dominated by an overstory of white oak (Q. alba), along with black oak (Q. velutina) and hickory (Carya spp.), and a diverse herb layer of primarily perennial forbs and grasses such as elm-leaved goldenrod (Solidago ulmifolia), wild-ryes (Elymus spp.), and ticktrefoils (Desmodium spp). The riparian zone, which represents a smaller portion of the unit, maps as a Loamy Floodplain Forest, where periodic flooding creates frequent canopy gaps and open areas in the understory. The reference state for a Loamy Floodplain Forest includes greater canopy coverage, vertical structure, and representation of flood-tolerant and moisture-demanding species, such sedges (Carex spp.) and vines, than Loamy Upland Woodlands. |
In the Eastern Ozark border ecoregion, few remnants of even moderate quality remain that can serve as references for understanding the range of natural variability expressed in these community types. Although the prevailing paradigm is that herb-layer diversity can be restored through ecological thinning and reestablishment of low-severity fire regimes, plant community responses can vary widely among sites depending on the ecological context and land-use history. In damaged Ozark woodlands, herb-layer recovery can be limited by the absence of a soil seed-bank and loss of remnant populations of native species capable of expanding once conditions become more favorable, especially in sites heavily invaded by invasive shrubs over long time periods. In highly fragmented urbanized landscapes like SNR, dispersal limitation is another important barrier to the restoration of herb-layer diversity. Although seed additions can overcome dispersal barriers, they are seldom used and an understudied possibility for accelerating herb-layer abundance and richness when restoring Ozark woodlands.
Our project goals over the three-year period were to restore ecological structure and function in this damaged woodland by: 1) removing exotic shrubs and selectively thinning trees to create an open woodland structure, 2) reestablishing a low-intensity, high-frequency fire regime, and 3) testing whether seed additions could accelerate herb-layer community recovery.
IMLS Final Report Project Abstract Here
Our project goals over the three-year period were to restore ecological structure and function in this damaged woodland by: 1) removing exotic shrubs and selectively thinning trees to create an open woodland structure, 2) reestablishing a low-intensity, high-frequency fire regime, and 3) testing whether seed additions could accelerate herb-layer community recovery.
IMLS Final Report Project Abstract Here
Ecological Restoration Treatments
(hover over photo for treatment descriptions)
Photo Credits: Mike Saxton, Quinn Long, and John Behrer.
Photopoint Monitoring
To document ecosystem recovery and monitor project outcomes, we established eleven permanent photopoint monitoring locations throughout the site (see map below). Photopoints used tagged trees as permanent reference points and images were taken in the same cardinal direction every season over the three year period. The slideshows below illustrate dramatic ecosystem recovery over the three-year funding period.
Photopoint 05-East
Photopoint 02-South
|
Photopoint 10-North
Photopoint 08-East
|
Herb Layer Dynamics With and Without Seed Addition
To monitor the response of vegetation to restoration and test whether seed additions could accelerate the recovery of native species, we established ten transects each with five 1 × 1 squared meter quadrats throughout the project unit (see map below). Herb-layer abundance and composition was sampled in early– and late–summer beginning in 2017. Half the transects (n=5) were seeded in winter 2018 whereas the other half (n=5) served as controls. Vegetation data collection, description, and analysis was led by CCSD’s Noah Dell.
Local native species richness (per quadrat) increased over time in response to restoration treatments. However, local native richness increased at greater rates in seed addition relative to control (non-seeded) quadrats. Seed additions were conducted in winter 2018. See species list here.
|
Mean C-values (per quadrat) were significantly greater with seed additions compared to the controls (non-seeded quadrats). C-values are based on the coefficient of conservatism ranking assigned to each native species in the Missouri Flora (Ladd and Thomas 2015). Species with higher conservatism rankings are more dependent on remnant woodlands, or those that are less altered compared to pre-European settlement, than species with lower conservatism rankings.
|
Nonmetric multidimensional scaling (NMDS) ordination illustrating compositional differences in the herb layer community among seeded and control quadrats in 2019 (two-years after seed-addition treatments were applied). Letter codes represent those species that are significantly associated with the NMDS. Each symbol represents a quadrat.
|
Restoration Web Application (RWA)
To track restoration activities and enhance stewardship of natural communities at SNR, project staff developed a geospatial restoration web application (RWA) using ESRI’s ArcGIS Web AppBuilder for ArcGIS. The RWA combines publicly available environmental layers (e.g., satellite imagery, historical land cover, General Land Office survey data, topographic maps, digital elevation models, geology, and soils) with those specific to SNR (management units, structures, roads, research plots, and trails). With an iPad Mini in the field, project staff can use the RWA to track and monitor invasive species treatments, detail prescribed burns, relocate monitoring plots or transects, and document restoration activities such as seed additions. All data is permanently stored in a geodatabase and archived in MBG's Geospatial data server. Below are examples of the RWA's utility in our woodland restoration project.