Runnels as a Climate Adaptation Tool
Runnels used for climate adaptation to SLR are a new application of an existing mosquito control technique used in the USA and Australia (Hulsman et al. 1989; Wolfe 1996). Historic mosquito ditches were excavated > 60 cm deep, causing peat oxidation and subsidence of the inter-ditch marsh platform (Burdick et al. 2020). More recently, mosquito control programs began constructing runnels that resembled natural channels to drain standing water (mosquito larvae habitat) and allow fish passage (mosquito larvae predators), with minimal impact to marshes (Hulsman et al. 1989).
A runnel is a small channel (generally ≤ 30 cm wide and deep) that drains standing water on the marsh surface. Runnels are constructed using hand-digging and low-ground pressure excavators or ditchers (Supplemental File 1) to follow topographical low areas, and only drain water within the rooting zone (Hulsman et al. 1989; Wigand et al. 2017). Runnels are similar in principle to tidal creek extension projects that connect an area of inundation to the tidal creek network, though tidal creek extensions are larger in scale than runnels (Raposa et al. 2019; Taylor et al. 2020; Wetland restoration at Farm Creek Marsh 2021). After observing rapid expansion of shallow water within northeastern US marshes, restoration ecologists began working with mosquito control agencies to use runnels for the dual purpose of mosquito abatement and marsh adaptation to SLR. Practitioners used the technique to target shallow water features that were expanding, had formed within the last few decades, and where an anthropogenic topographic feature was impairing water flow (ditch spoils, plugged ditches, embankments) (Wigand et al. 2017; Adamowicz et al. 2020; Perry et al. 2021; Wolfe et al. 2021). True ponds that remained flooded throughout the tide cycle, with unconsolidated sediments in the basin, and that appeared stable in dimension on decadal timescales, were not targeted with this technique (workshop communications). Save the Bay (STB), an environmental non-profit, launched a series of projects using runnels in Rhode Island (RI), USA beginning in 2010. In our 2020 workshop, resource managers reported projects on dozens of marshes across six northeastern US states, and another half dozen northeastern and mid-Atlantic marshes were under consideration for runnel-adaptation by land trusts, NGOs, and government agencies (Supplemental File 1). The majority of projects from the workshop, and one recently published study on runnels (Perry et al. 2021), have reported some vegetation recovery within 1–5 years (Table 2 and Supplemental File 1).
Runnel Case Study: Winnapaug Marsh, RI
An STB restoration project provides a case study on patterns of vegetation recovery. We selected this project because it has the longest monitoring record (8 years) of the STB projects, including pre-treatment data. Habitat restoration using runnels can be summarized by three phases. Phase I: “Drainage” is characterized by a loss of standing surface water. Phase II: “Early colonizers” is characterized by bare sediment which is colonized by Salicornia spp. and Spartina alterniflora. Phase III: “High-marsh species” is characterized by Distichlis spicata, Spartina patens, and Juncus gerardii succeeding early colonizers.
Winnapaug back barrier salt marsh in RI (41.3306°N, − 71.7684°W) is a grid-ditched marsh with significant surface water cover and platform degradation (Fig. 3). Tidal range at the nearest tide station in Newport, RI is 1.05 m; however, tidal amplitudes are restricted in back-barrier environments such as Winnapaug. Ditches were created during the 1930s, and peat spoils were placed along ditch edges, creating linear impoundments. Altered topography in combination with RSLR in RI (5.26 mm yr−1 between 1999 and 2015) (Raposa et al. 2017) led to the “waffle-maple-syrup” pattern (Adamowicz et al. 2020) seen in aerial imagery (Fig. 3). As of 2011, large mats of filamentous algae were growing in shallow water areas (Fig. 1e), and mosquito larvae were observed. Initial depths of shallow water areas ranged from a few centimeters up to about 25 cm and were generally less than 15 cm deep.
STB and Town of Westerly, RI, secured funds and permits to create runnels targeting shallow water areas. Environmental and vegetation monitoring was conducted prior to runnel creation in 2011, and post-implementation monitoring was repeated in 2013–2015, 2017, and 2019. Initial hand excavation of a few small runnels began in summer 2012. In May 2013, STB and RI Department of Environmental Management’s Mosquito Abatement Program used a low-ground pressure excavator to expand the runnel network, and volunteers hand dug smaller runnels. Clogged mosquito ditches were cleared, and the material was used to fill selected ditches and degraded areas. Hand digging continued in 2013–2014 to facilitate additional drainage. In total, around 33 runnels were created ranging from 2 to 8 m in length. Runnel widths ranged from 10 to 24 cm, and depths ranged from 10 to 18 cm.
Surveys of vegetation and surface water were conducted using quadrat sampling along transects (Roman et al. 2001). Vegetation and ground cover was estimated as percent cover of each transect (Fig. 3). In the text below, transect data was aggregated to present coverages by species or cover type for the entire marsh. Initially (2011), algal mats covered 44%, open water 14%, and bare peat 4.5% of the marsh platform. The marsh was dominated (57% cover) by Spartina alterniflora, a species which tolerates frequent inundation (Fig. 3). Less-flood tolerant, “high-marsh” species included Distichlis spicata (26%), Spartina patens (18%), and Juncus gerardii (2.7%).
Ecosystem responses to runnels proceeded as across the marsh (Fig. 4). During Phase I open water decreased to 5% by 2013, and was absent in 2014 across the entire marsh. Algal mats disappeared by 2013. During Phase II, bare peat initially increased as water drained from the site (maximum of 26% by 2013), but then declined (3.2% by 2019) as areas were recolonized. Salicornia depressa, a flood-tolerant, early-colonizing species, increased rapidly from 3.3% prior to runnels to 73% in 2 years (2014). Salicornia then declined to 4.3% by 2019 as less flood-tolerant, high-marsh species increased. After 3 years (2015), Phase III-high marsh species began to increase. After 7 years (2019), Distichlis had increased to 42% cover, Spartina patens to 24% cover, and Juncus to 3.8%. Spartina alterniflora remained the dominant cover, increasing to 65% cover after 3 years, and 68% after 7 years (2019). The increase in vegetation, especially high-marsh species, suggests that runnels have potential for short-term restoration of marsh plants.
While vegetation recovered across the marsh on the whole, responses differed across the marsh. Platform elevations along transect T1 were conducive to high-marsh species growth prior to runnel creation; as a result, draining the shallow water areas allowed bare peat to recolonize with high-marsh species quickly (Fig. 3). In contrast, the shallow water areas at the northern ends of transects T2 and T3 (Fig. 3) showed minimal response to runnels. Water levels decreased, but the features never fully drained and vegetation did not recover. As a result, T2 and T3 vegetation responses differed from T1 (Fig. 3). Based on water table monitoring (Supplemental File 2), STB believes that basin elevations in some of the northern shallow water areas were too low in elevation for vegetation to recover. Long-term monitoring at this and other runnel project sites is important for assessing which marshes are good candidates for runnels, and how much time we can “buy” using this technique (a few years, decades, or more).