Facilitate tidal exchange to restore/create brackish/salt marshes from other land uses
Overall effectiveness category Beneficial
Number of studies: 14
Background information and definitions
This action includes actions that facilitate tidal exchange, in order to restore or create marshes from other land uses. The action may be a single permanent one (e.g. breaching sea walls or embankments, installing or widening culverts, excavating tidal creeks) or a reversible one (e.g. opening sluice gates once per day). However, the intervention must affect an area that does not retain substantial characteristics of the target habitat. This could be an upland area (e.g. the thousands of square kilometres of farmland that has been reclaimed from salt marsh in the Netherlands since the Middle Ages; Wolff 1992), an unvegetated wetland (e.g. mudflats), or a wetland other than the target type (e.g. swamp, where the habitat used to be a marsh). “Planned retreat” and “managed realignment” fall within the scope of this action.
Tidal wetlands may be brackish/saline (e.g. mangroves, coastal marshes) or freshwater (e.g. at the upstream end of estuaries, as in the Mississippi, Yangtze, and Elbe rivers; Baldwin et al. 2009). Studies of accidental realignment, such as when coastal defences are breached by a storm, have not been summarized as evidence (e.g. Onaindia et al. 2001; some sites in Williams & Orr 2002).
Related actions: Facilitate tidal exchange to restore degraded marshes; Reprofile/relandscape or Remove surface soil/sediment, both of which can alter patterns of tidal exchange; Facilitate tidal exchange to complement planting.
Baldwin A.H., Barendregt A. & Whigham D. (2009) Tidal Freshwater Wetlands. Backhuys Publishers, Lieden.
Onaindia M., Albizu I. & Amezaga I. (2001) Effect of time on the natural regeneration of salt marsh. Applied Vegetation Science, 4, 247–256.
Williams P.B. & Orr M.K. (2002) Physical evolution of restored breached levee salt marshes in the San Francisco Bay estuary. Restoration Ecology, 10, 527–542.
Wolff W.J. (1992) The end of a tradition: 1000 years of embankment and reclamation of wetlands in the Netherlands. Ambio, 21, 287–291.
Supporting evidence from individual studies
A before-and-after study in 1991–1994 of a coastal site where tidal exchange was restored in England, UK (Dagley 1995) reported that salt marsh vegetation colonized the site within one year, and that salt marsh plant communities developed within three years. Before intervention, the site was a coastal grassland containing 14 plant species. One year after intervention, the site contained 17 plant and algae species (1.6 species/m2) – mostly the alga Enteromorpha sp. (present in 88% of quadrats). After 2–3 years, the site contained 22–25 plant and algae species (3.0–3.5 species/m2). It had developed a recognizable salt marsh vegetation community. The most common species were glassworts Salicornia spp. (in 88–94% of quadrats) and seablite Suaeda maritima (in 64–74% of quadrats), along with Enteromorpha sp. (in 88–96% of quadrats). Methods: In July 1991, a sea wall was lowered to allow tidal exchange on 2 acres of Northey Island. The studied area was inundated by 100 tides/year. Vegetation was surveyed (50 randomly placed quadrats/year; ≤1 m2) before intervention (June 1991) and for three years after (summer 1992–1994). The restoration site was included in studies (7) and (11).Study and other actions tested
A before-and-after, site comparison study in 1993–1995 in an estuarine marsh in New Hampshire, USA (Burdick et al. 1996) reported that an area in which tidal exchange was improved (by modifying culverts under a road) developed cover of salt marsh vegetation within two years, although total vegetation cover declined. Statistical significance was not assessed. Before intervention, the tidally restricted area was a freshwater wet meadow dominated by grasses (39% cover) and asters Aster spp. (32% cover). Total vegetation cover was 94%. Two years after restoring tidal exchange, the area was dominated by the alga Vaucheria spp. (21% cover) and saltmeadow cordgrass Spartina patens (21% cover), with other salt-tolerant species present at lower abundance (e.g. smooth cordgrass Spartina alterniflora: <1% cover; see original paper for full data). Total vegetation cover was 63%. For comparison, a reference area of marsh downstream contained communities dominated by saltmeadow cordgrass (56% cover) or smooth cordgrass Spartina alterniflora (60% cover) with no algae, and 71–86% total vegetation cover. Methods: A road built across Mill Brook Marsh, in 1970, restricted tidal exchange to part of the marsh through a narrow gated culvert. In 1993, the gate on the old culvert was removed and a new, wider culvert was installed. This restored regular tidal exchange, raised the water table and increased soil salinity in the degraded area. Summer vegetation surveys were carried out, using 1-m2 quadrats, before (1993) and after (1995) intervention in the degraded/restored area. The undisturbed marsh below the road was also surveyed.Study and other actions tested
A site comparison study in 1987–1998 of two estuarine marshes in Washington, USA (Thom et al. 2002) reported that after breaching a dyke to restore tidal influx to one marsh, its plant community became more similar to an adjacent natural marsh. Statistical significance was not assessed. The restored marsh progressed through four key phases. Before breaching, it was dominated by freshwater wetland plant species (not quantified). In the first two years after breaching, the most abundant plant species was reed canary grass Phalaris arundinacea (in 43–57% of quadrats). After 4–5 years, the dominant species was pickleweed Salicornia virginica (52–53% cover). After 8–11 years, the dominant species were saltgrass Distichlis spicata (36–48% cover), arrowgrass Triglochin maritima (12–23% cover) and pickleweed (7–16% cover). Meanwhile, an adjacent natural marsh was dominated by saltgrass (24–47% cover), tufted hairgrass Dechampsia cespitosa (18–42% cover) and pickleweed (6–14% cover). Overall, the plant community composition in the restored marsh became more similar to the natural marsh over time (32–42% similarity after 4 years; 58–68% after 8 years; 48–80% after 11 years). After 4–11 years, total plant species richness was similar in both marshes (restored: 8–12 species/36 m2; natural: 9–14 species/30 m2; 14 species recorded in each marsh over all surveys). Methods: In early 1987, tidal exchange was restored to 23 ha of coastal land by breaching a dyke that had been built in the early 1900s. This area had subsided whilst dyked, and had a higher salinity than adjacent estuarine water after restoration. Vegetation was surveyed most summers in 1987–1998: initially species presence in seven 1-m2 quadrats in the restored marsh, but from 1991 presence and cover in 30–37 quadrats in the restored marsh and an adjacent natural marsh.Study and other actions tested
A replicated study in 1979–2000 of six coastal sites where tidal exchange was restored in California, USA (Williams & Orr 2002) reported that three of the sites developed ≥50% vegetation coverage within 15 years. These sites were 0.3–0.9 m above mean sea level when tidal exchange was restored. The other three sites had <50% vegetation coverage after 6–20 years (their maximum age during the study). These sites were 0.5–4.6 m below mean sea level when tidal exchange was restored. Methods: Between 1979 and 1995, levees were deliberately breached to restore tidal exchange in six coastal sites (farmland, mudflats, salt ponds or borrow pits). The area of each site covered by vegetation stands was estimated from historical aerial photographs and field surveys.Study and other actions tested
A replicated study in 1972–2000 of four filled and tidally restored coastal sites in California, USA (Williams & Orr 2002) reported that they developed 50% vegetation coverage within approximately 5–14 years. Any coverage beyond 50% was not quantified. Methods: Between 1972 and 1976, four coastal sites (historical land use not clear) were filled with dredged materials to restore suitable elevations for salt marsh plants (0.5–1.5 m above mean sea level). Then, tidal influx was restored by breaching levees. Note that this study evaluates the combined effect of these interventions. The area of each site covered by vegetation stands was estimated from historical aerial photographs.Study and other actions tested
A before-and-after study in 2002–2005 aiming to restore a salt marsh on farmland in England, UK (Badley & Allcorn 2006) reported that approximately three years after clearing existing vegetation and restoring tidal exchange, 70% of the site was covered by salt marsh vegetation. The first colonizers included glasswort Salicornia sp. and seablite Suaeda maritima, but sea purslane Halimione portulacoides and sea aster Aster tripolium were present after 2–3 years (data not reported). The study reported that plant species diversity in the managed site was similar to adjacent natural salt marsh (but this was neither quantified nor statistically tested). Methods: The study used 66 ha of cropland that had been claimed from the sea in 1983. In August 2002, tidal exchange was restored to the site by blocking some drainage ditches, excavating tidal channels and breaching the seawall. Existing vegetation was cleared before hydrological restoration, so note that this study evaluates the combined effect of these interventions. Details of vegetation monitoring were not reported.Study and other actions tested
A replicated, paired, site comparison study in 2004 of eight salt marshes in England, UK (Garbutt & Wolters 2008) reported that restored marshes (deliberately exposed to tidal influx) contained different vegetation communities to natural marshes, typically with lower species richness and taller vegetation. Although all restored sites contained salt marsh vegetation after 2–13 years, the specific community type differed from natural marshes in four of four comparisons. Further, vegetation communities in restored marshes were ≤44% similar to those in natural marshes (8% for a 2-year-old marsh; 35–44% for 9–13-year-old marshes). Four of 17 recorded species had significantly different cover in restored and natural marshes, including sea purslane Atriplex portulacoides (restored: 2%; natural: 30%) and common cordgrass Spartina anglica (restored: 21%; natural: 3%). Species with statistically similar cover in restored and natural marshes included saltmarsh grass Puccinellia maritima (restored 47%; natural: 33%) and glasswort Salicornia europaea (restored: 13%; natural: 5%). In two of four comparisons, restored marshes had significantly lower species richness than restored marshes (restored: 2–3 species/2 m2; natural: 8–10 species/2 m2; other comparisons no significant difference) and significantly taller vegetation than natural marshes (restored: 20–44 cm; natural: 9–22 cm; other comparisons mixed results). Methods: In July 2004, vegetation was surveyed in four pairs of adjacent restored and natural salt marshes. The restored marshes were former farmland, where embankments had been breached 2–13 years previously to restore tidal exchange. Plant/algal species and cover were recorded at a fixed elevation in five 2-m2 quadrats/marsh. This study included the restoration sites studied in (1) and (8). All sites in this study were included in (11).Study and other actions tested
A before-and-after, site comparison study in 1995–2003 of three salt marshes in England, UK (Wolters et al. 2008) found that a marsh restored by breaching an embankment around farmland was colonized by salt marsh vegetation, and developed a similar species richness to nearby natural marshes within five years. Plant species colonized gradually: glassworts Salicornia spp. were the first species to establish (within two years), then seablite Suaeda maritima, then long-lived salt marsh species (see original paper for frequency data). From five years after breaching, plant species richness on the restored marsh was within the range of two nearby natural marshes (data reported as a saturation index). After eight years, 11 salt marsh plant species had established. The plant communities on the restored marsh matched recognized salt marsh community types, characterized on higher ground by saltmarsh grass Puccinellia maritima (found in 100% of quadrats) and on lower ground by glassworts (in 55–100% of quadrats, depending on elevation). Methods: In August 1995, a 50-m-wide opening was made in an embankment around agricultural land, allowing the tide to enter twice a day. Plants were allowed to colonize naturally. Annually from 1997–2003, vegetation cover was recorded in 7,500 quadrats (each 1 m2). Quadrats were arranged in three 20 x 125 m transects, perpendicular to the shoreline. Long-term data on plant species in two nearby natural marshes were used for comparison. The restoration site was included in studies (7) and (11).Study and other actions tested
A before-and-after study in 1993–2004 in an estuary in New South Wales, Australia (Howe et al. 2010) reported that after removing culverts to improve tidal exchange to an island, the area of salt marsh vegetation increased. Salt marsh vegetation covered 44 ha of the study area two years before culvert removal, 52 ha three years after culvert removal, and 53 ha nine years after culvert removal. Other habitats present in the study site included mangrove forests (before: 1 ha; after nine years: 12 ha), tidal pools/mudflats (before: 33 ha; after nine years: 32 ha) and upland pasture (before: 42 ha; after nine years: 22 ha). Methods: The study focused on an island in the Hunter River Estuary, which had been partially drained for agriculture. In 1995, two 0.5-m diameter culverts in a tidal inlet were removed, restoring full tidal exchange to approximately one fifth of the island. Tidal exchange was slightly improved across the rest of the marsh, where culverts remained in place. Habitats were mapped from aerial photographs taken in 1993, 1998 and 2004.Study and other actions tested
A site comparison study in 2007 of two salt marshes in the UK (Kadiri et al. 2011) reported that a restored salt marsh (where the sea wall was breached after depositing sediment) contained fewer plant species and less vegetation cover than a natural salt marsh. Statistical significance was not assessed. After 15 months, the restored marsh contained only one plant species: glasswort Salicornia europaea. Its cover was 11%. A nearby natural marsh contained eight plant species: mostly common saltmarsh grass Puccinellia maritima (50% cover), sea lavender Limonium vulgare (23% cover) and common cordgrass Spartina anglica (10% cover). Glasswort cover was 2%. The study also noted differences in sediment properties, including salinity and organic matter content, between the restored and natural marsh. Methods: In October 2007, plant species and their cover were recorded in ten 0.5-m2 quadrats, in each of two salt marshes. One marsh had been restored by depositing dredged sediment onto farmland, to raise the ground to an appropriate level for marsh vegetation (May 2005), then breaching the sea wall to restore tidal exchange (July 2006). The other, natural marsh had never been tidally restricted. Note that this study evaluates the combined effect of depositing sediment and restoring tidal exchange.Study and other actions tested
A replicated, site comparison study in 2004–2010 of 52 salt marshes in England, UK (Mossman et al. 2012) reported that restored marshes (deliberately exposed to tidal influx) were colonized by salt-tolerant plants within one year, but found that they had a different plant community with lower diversity and cover than natural salt marshes. After 1–14 years, restored marshes contained 21–80% of all salt-tolerant plant species recorded in the study (vs 27–77% in natural marshes; statistical significance not assessed). However, the overall composition of the plant community significantly differed between restored and natural marshes (data reported as a graphical analysis). Plant diversity was also lower in quadrats from restored marshes (data reported as a diversity index). In three of three comparisons, restored marshes had lower overall vegetation cover (53–83%) than natural marshes (84–98%). Restored marshes had significantly lower cover of Atriplex portulacoides in three of three comparisons (restored: 7–9%; natural: 17–21%) and significantly greater cover of glasswort Salicornia europaea in two of three comparisons (for which restored: 12–21%; natural: 5–12%), but similar cover of saltmarsh grass Puccinellia maritima in two of three comparisons (for which restored: 23–32%; natural: 29–31%; see original paper for full cover data). Methods: In summer–autumn 2004–2010, vegetation was surveyed in 52 salt marshes: 18 marshes restored from agricultural land 1–14 years previously by deliberately breaching sea walls, and 34 nearby natural marshes. Cover of all vascular plant species, and bare ground, were estimated in at least fifty 0.25-m2 quadrats/marsh (along transects perpendicular to shoreline). Species within 20 m of transects were also noted. This study included the sites studied in (1), (7) and (8).Study and other actions tested
A before-and-after study in 2001–2012 aiming to restore a salt marsh on pasture land in Scotland, UK (Elliott 2015) reported that salt marsh vegetation colonized the site within one year of breaching the sea wall, and dominated the site within three years. Before breaching, all sixty surveyed quadrats contained wet grassland/rush pasture plant communities. After one summer, 18% of quadrats contained salt marsh plant communities, with 57% of quadrats bare mud. Within three years, 65% of quadrats contained salt marsh plant communities, with only 2% bare mud. Wet grassland/rush pasture persisted in 21% of quadrats, at higher elevations. After nine years, 93% of quadrats contained salt marsh plant communities, 6% wet grassland and 1% bare mud. After breaching, there were only 25–32 plant species on the marsh each year, compared to 37 before. Methods: In February 2003, two 20-m-long breaches were dug in a sea wall. This restored tidal exchange to a 25-ha pasture created in the 1950s. Plant species and community types were recorded in sixty permanent quadrats across the site, before breaching (August 2001) and for up to nine years after (summer 2003–2011).Study and other actions tested
A before-and-after study in 1987–2011 aiming to restore a brackish/salt marsh on grassland in the Netherlands (Chang et al. 2016) reported that within ten years of restoring regular tidal exchange, marsh plant communities had developed. Before intervention, the site was a grassland containing 35–70% (depending on the elevation) of the target marsh species (typical of brackish or saline marshes in the region). After 10 years, the site was completely covered by a range of brackish and saline marsh plant communities, each containing 78–96% of the target marsh species. Of the 23 target species, only common reed Phragmites australis was not found along any surveyed transects – but it was noted elsewhere in the site. Methods: Between 1997 and 2001, regular tidal exchange (i.e. more than just high spring tides and storm surges) was restored to grassland behind an embankment: first (1997) by opening two culverts, then (2000) by excavating creeks and filling drainage ditches, and finally (2001) by creating three breaches, each 20–40 m wide, in the embankment. Plant communities were assessed from maps made 14 years before and up to 10 years after breaching. Plant species were recorded in three permanent transects (each containing 250–310 contiguous 10 x 10 m cells) one year before and up to 10 years after breaching.Study and other actions tested
A site comparison study of four brackish marshes in an estuary in Oregon, USA (Flitcroft et al. 2016) reported that after removing levees to restore tidal exchange, the plant community became more similar to that of a nearby natural marsh – but remained significantly different after >30 years. In all three restored marshes, freshwater pasture grasses were gradually replaced by native salt-tolerant species such as pickleweed Salicornia virginica and saltgrass Distichlis spicata (data not reported). However, in a marsh where tidal exchange had been restored for the longest time (>30 years), the overall plant community composition remained significantly different from the natural marsh (data not reported). This restored marsh lacked some “diagnostic” brackish marsh species, such as Baltic rush Juncus balticus and black bent Agrostis alba. Methods: Vegetation was surveyed in four brackish marshes within the Salmon River estuary (years and survey methods not reported; salinity obtained from Gray et al. 2002). In three marshes, tidal influx had been restored. Levees that kept these sites as freshwater pasture were removed in 1978, 1987 or 1996. The other site was a natural marsh, where tidal influx had never been modified.
Additional Reference: Gray A., Simenstad C.A., Bottom D.L & Cornwell T.J. (2002) Contrasting functional performance of juvenile salmon in recovering wetlands of the Salmon River estuary, Oregon USA. Restoration Ecology, 10, 514–526.Study and other actions tested
Referenced paperFlitcroft R.L., Bottom D.L., Haberman K.L., Bierly K.F., Jones K.K., Simenstad C.A., Gray A., Ellingson K.S., Baumgartner E., Cornwell T.J. & Campbell L.A. (2016) Expect the unexpected: place-based protections can lead to unforeseen benefits. Aquatic Conservation: Marine and Freshwater Ecosystems, 26, 39-59.