Impact of control structures on hydrologic restoration within the Great Dismal Swamp

Highlights

• Properly managed control structures positively impacted the rewetting efforts.

• Outcome will aid in the restoration of desired forest communities and reduction of fire susceptibility.

• Control structures prove to be a viable, low-cost tool for hydrologic restoration.

Abstract

The Great Dismal Swamp (GDS) is a 45,000-ha state and federally protected Coastal Plain peatland located on the border of North Carolina and Virginia that contains stands of Bald cypress and the globally threatened Atlantic white cedar. Centuries of drainage and logging have substantially altered the hydrology of the GDS, negatively affecting its ecosystem structure and function. To restore a seasonally flooded, saturated hydrologic regime to portions of the swamp, adjustable water control structures (WCS) were installed at strategic locations within existing drainage ditches. The objective of this study was to determine if the installation of the WCSs significantly altered the hydropatterns of two target restoration areas, resulting in hydrologic conditions comparable to nearby reference sites with desired forest communities. The water table (WT) was monitored for three years prior to WCS installation (pre-WCS) and three years after WCS installation (post-WCS). Comparison of WT data from the pre and post-WCS periods, using jurisdictional wetland criteria and empirical cumulative distribution functions (ECDFs), indicated increased saturated conditions within the target restoration areas following installation of the WCS. Paired Before-After Control-Impact (BACIP) statistical analysis revealed the WCS installation had a significant positive impact on WT levels in the target restoration areas relative to the reference sites. Hydrologic restoration will aid the effort to restore target forest communities within the swamp, reduce fire susceptibility, prevent peat oxidation, maintain carbon storage, and reduce non-target vegetation competition.

Introduction

Pocosins and associated peatlands once covered roughly 1.2 million ha of the southeastern Coastal Plain and spanned approximately 1000-km from Virginia to northern Florida (Richardson, 1983). These ecosystems have saturated, semi-permanently, intermittently, or seasonally flooded hydrologic regimes with deep peat accumulations (Cowardin et al., 1979). In this environment, Taxodium distichum (Bald cypress) and the globally rare Chamaecyparis thyoides (Atlantic white cedar) stands once formed dominant forest communities. However, extensive logging, conversion to agriculture, and commercial development in the Coastal Plain has greatly reduced the extent of these forests over the last 300 years (Drexler et al., 2017; Frost, 1987; Kuser and Zimmerman, 1995; Richardson, 2012). The remaining peatlands, although still intact, have been subjected to anthropogenic alteration via logging or drainage (Ferrell et al., 2007; Frost, 1987; Levy and Walker, 1979). One of these remaining anthropogenically altered peatlands is the Great Dismal Swamp (GDS) of North Carolina and Virginia.

Alteration of the hydrologic regime in GDS began in the late 18th century. By the end of the 19th century, five major ditches had been constructed, including the hydrologically imposing Dismal Swamp Canal (Levy, 1991; US, 2006). Since European settlement, approximately 320-km of ditches and canals have been excavated to allow for the harvest of the swamp's timber, the reclamation of land for development, and the transport of goods between the Chesapeake Bay and the Albemarle Sound (US, 2006). Extensive logging paired with the residual effects of the remaining drainage systems have hampered the regeneration of historic flora, and resulted in the transition of the GDS to a mesic forest community dominated by facultative red maple/black gum forest (Laderman, 1989; US, 2006). The once abundant stands of obligate wetland species, including Bald cypress (BC) and Atlantic white cedar (AWC), covered less than 20% of the total GDS area in 2006 (Frost, 1987; Laing et al., 2011; US, 2006). The drier conditions have also led to increased susceptibility to major peat fires, several of which occurred in the last century (1923 to 1926, 1975, 2008, and 2011). These peat fires have not only contributed to the altered ecological state of the GDS seen today (Richardson, 1982; Laderman, 1989), but also released tons of stored carbon to the atmosphere. The 2011 fire was estimated to have released between 0.06 and 1.10 Tg of stored carbon (Reddy et al., 2015).

Because hydrology is the primary control of wetland structure and function (Hunt et al., 1999; National Research Council, 1992, National Research Council, 1995; Zedler, 2000), the reestablishment of a higher water table (WT) was recognized as a prerequisite for successful wetland restoration (Chimner et al., 2017). However, the necessary magnitude of the WT increase required in the GDS was unknown. In their extensive report on the GDS, Lichtler and Walker (1974) observed, “Restoring the (Great Dismal) swamp to its original condition is impossible because that condition is unknown.” Nevertheless, a strategy to raise the WT in portions of the GDS was viewed as the most prudent and cost-effective method to reduce fire susceptibility, prevent peat oxidation, maintain carbon storage, and begin the restoration of the historically dominant forest communities within the swamp.

A plan was developed to create a hydropattern that would increase the frequency and duration of saturation in the soil profile for a portion of the park without impact to existing roads and trails. This included the installation of adjustable water control structures (WCS) at strategic locations in the existing canal system. A WCS consists of a flashboard riser attached to a culvert. Adjustable riser boards, or stop-logs, in the riser structure are used to set the canal water level and slow drainage from target restoration areas (Fig. 1). When installed and managed properly, WCS decrease the hydraulic gradient from the surrounding swamp to the ditch, slowing the drainage of groundwater and producing wetter conditions within surrounding target wetland areas.

Similar ditch blocking approaches to hydrologic restoration have been applied to other peatlands with success. Landry and Rochefort (2012) provide a detailed overview of peatland drainage impacts and several different rewetting techniques, including recommendations for the design of WCSs. Jaenicke et al. (2010) investigated the potential use of similar ditch blocking structures to dam drainage canals in two catchments that drain peat domes in Indonesia. Hydrological modelling of the two systems predicted a 50 to 70 cm rise of groundwater levels during very dry conditions and an average groundwater level rise of 20 cm over a three-year period after the installation of ditch dams. A comprehensive study on the hydrologic restoration of drained peatlands was conducted by Menberu et al. (2016). The study evaluated the impact of ditch blocking techniques on peatland hydropattern in previously drained boreal peatlands of Finland. Hydropatterns were monitored at 24 previously drained sites for one to two years before restoration and one to six years after restoration; 19 pristine peatlands were also monitored and used as control sites. The WTs in 22 of the 24 drained sites were significantly deeper than the corresponding control sites before restoration. After restoration, 12 restoration sites had WTs significantly higher than their corresponding control site WT, 10 restoration sites still had WTs significantly lower than their corresponding control site WT, and 2 restoration sites had WTs that were not significantly different than their corresponding control site WT. The average increases in WT because of restoration were 35.0 cm, 17.4 cm, and 13.3 cm, for spruce mires, pine mires, and fens, respectively.

Locally, WCS have been used in North Carolina since the 1970s to improve downstream water quality by restricting the volume of drainage water released from artificially drained agricultural fields. Gilliam et al. (1979) observed that WCS installed on tile mains or in the outlet ditches increased water table elevation, provided effective control of the water table, and reduced annual drainage volume by approximately 50%. The ability of WCSs to increase upstream WT elevation led to the application of WCSs in hydrologic restoration projects for wetlands that had been ditched, drained, and converted to agriculture (Tweedy and Evans, 2001; Jarzemsky et al., 2013). Both studies indicated that the technique of plugging drainage ditches with WCSs had the potential to restore jurisdictional wetland hydrology to previously converted wetlands.

Although success has been demonstrated, Menberu et al. (2016) noted that “restoration of peatland hydrology is still poorly documented”, WCS have been implemented in North Carolina on drainage ditches in converted agricultural lands but there is limited data on the use of WCS to restore wetland hydrology in forested peatlands that have been ditched for silviculture. In one of the only papers on hydrologic restoration in North Carolina forested peatlands, Wurster et al. (2016) detail the use of WCS for hydrologic restoration, but there was no WT data associated with this report.

This study was conducted to monitor hydrologic conditions within two target restoration sites and two reference wetland areas with desirable forest communities (AWC and BC) before and after WCS installation. The objectives were to 1) assess the hydrology of the target restoration areas, 2) compare the hydrologic conditions in the target restoration areas to those in nearby reference wetland communities and 3) evaluate the impact of WCS installation on the hydropattern of the target restoration areas. To the best of our knowledge, this is the one of the few studies to quantify, over an extended period, the impact of WCS on the WT in previously drained forested peatlands in North Carolina.