Directly plant non-woody plants: brackish/saline wetlands
Overall effectiveness category Beneficial
Number of studies: 30
Background information and definitions
This action involves planting whole emergent plants, directly into soil or sediment, to restore/create marshes or swamps. These plants might be individual seedlings, rooted cuttings or mature plants. Plants may be raised in greenhouses/laboratories, or collected from natural sites (with potential damage to donor site; Laegdsgaard 2002).
Introduction of target vegetation might be useful in severely degraded or bare sites – which may lack remnant plants or seed banks to kick start revegetation with desirable species, and may be at risk of being taken over by undesirable species (Brown & Bedford 1997). It might also be useful in isolated wetlands, far from sources of marsh or swamp plant propagules. However, note that up-front costs can be high.
The effects of planting may be highly dependent on the environmental conditions in each study. Questions you might ask when interpreting the evidence include: Is the study site degraded? Where and when was vegetation planted? Was there any intervention to improve conditions before planting? What were the environmental conditions over the duration of the study?
The scope of this action does not include planting nurse plants; planting submerged or floating plants; planting to restore bogs, fens, fen meadows or peat swamp forests (see Taylor et al. 2018); planting facultative wetland plants in upland sites; or planting for commercial purposes (e.g. mangrove plantations; Kaly & Jones 1998). In contrast, the scope does include planting non-native species to conserve marshes or swamps – whilst acknowledging that this is often considered ethically unacceptable due to the risk of invasion (e.g. Ren et al. 2009).
Related actions: Introduce vegetation fragments; Introduce seeds or propagules; Transplant or replace blocks of vegetation; Transplant or replace wetland soil; Introduce organisms to control problematic plants; Introduce nurse plants; Restore/create marshes or swamps using multiple interventions, often including planting.
Brown S.C. & Bedford B.L. (1997) Restoration of wetland vegetation with transplanted wetland soil: an experimental study. Wetlands, 17, 424–437.
Kaly U.L. & Jones G.P. (1998) Mangrove restoration: a potential tool for coastal management in tropical developing countries. Ambio, 27, 656–661.
Laegdsgaard P. (2002) Recovery of small denuded patches of the dominant NSW coastal saltmarsh species (Sporobolus virginicus and Sarcocornia quinqueflora) and implications for restoration using donor sites. Ecological Management & Restoration, 3, 202–206.
Ren H., Lu H., Shen W., Huang C., Guo Q., Li Z. & Jian S. (2009) Sonneratia apetala Buch.Ham in the mangrove ecosystems of China: an invasive species or restoration species? Ecological Engineering, 35, 1243–1248.
Taylor N.G., Grillas P. & Sutherland W.J. (2018) Peatland Conservation: Global Evidence for the Effects of Interventions to Conserve Peatland Vegetation. Synopses of Conservation Evidence Series. University of Cambridge, Cambridge.
Supporting evidence from individual studies
A controlled study in 1973–1981 on a brackish, tidal sandflat in North Carolina, USA (Benner et al. 1982) reported that an area planted with wetland herbs contained 24 plant species after eight years, including four of nine planted species, whilst an adjacent unplanted area remained “unvegetated”. The four persisting planted species were black rush Juncus roemerianus, common reed Phragmites australis, broadleaf cattail Typha latifolia and smooth cordgrass Spartina alterniflora. Twenty other herb and shrub species colonized the planted area naturally. Methods: In spring and summer 1973, nine herbaceous species were transplanted from existing marshes to a 30 m stretch of brackish, tidal, sandy sediment. Plants were 60–90 cm apart. An adjacent area of sediment was not planted. Plant species were recorded in 1981.Study and other actions tested
A site comparison study in 1980–1981 involving reprofiled borrow pits in North Carolina, USA (Broome et al. 1982) reported 37–98% survival of four planted herb species after one growing season and that the biomass of survivors increased, but that vegetation in planted and natural marshes differed after two growing seasons. Statistical significance was not assessed. After one growing season, survival rates were 37–55% for smooth cordgrass Spartina alterniflora, 66–97% for big cordgrass Spartina cynosuroides, and 82–98% for saltmeadow cordgrass Spartina patens (data not clearly reported for black rush Juncus roemerianus). The above-ground biomass of surviving plants increased from 8–562 g/m2/species after one growing season to 297–3,105 g/m2/species after two growing seasons. Finally, planted and natural stands of big cordgrass and black rush were compared. After two growing seasons, planted stands contained only 297–1,525 g/m2 above-ground biomass (vs natural: 997–1,891 g/m2) and vegetation only 91–161 cm tall (vs natural: 155–293 cm). Planted stands contained more big cordgrass stems than natural marshes (planted: 223; natural: 44 stems/m2), but fewer black rush stems (planted: 509; natural: 884 stems/m2). Methods: In spring/early summer 1979–1980, four herb species were planted (60–90 cm apart) into reprofiled coastal land (dry during planting but rewetted after; salinity <20 ppt). Experimental design, including number of plants and plots, was not clearly reported. In October 1979–1981, planted vegetation and vegetation in a nearby natural marsh were surveyed. This included cutting, drying and weighing live vegetation from 0.25-m2 quadrats.Study and other actions tested
A study in 1982 in an estuary in Maryland, USA (Earhart & Garbisch 1983) reported approximately 70% survival of saltmeadow cordgrass Spartina patens planted into deposited dredge sediment. Survival was measured after 5–6 months. Methods: In June–July 1982, nursery-reared saltmeadow cordgrass was planted on an area of fine-grained dredge sediment deposited in Tar Bay. Approximately 65,520 plants were planted, 60 cm apart, 40–70 cm above mean low water. Each plant was fertilized with 30 g of slow-release Osmocote® fertilizer. Survival was recorded in December 1982.Study and other actions tested
A replicated study in 1976–1977 on two intertidal mudflats in Texas, USA (Tanner & Dodd 1985) reported 3–89% survival of planted smooth cordgrass Spartina alterniflora after one growing season, and increases in stem height, density and biomass over two growing seasons. Unless specified, statistical significance was not assessed. After two growing seasons, cordgrass stems were 84–140 cm tall (vs 20–100 cm when planted). There were 18–252 cordgrass stems/m2 (vs <1–35 stems/m2 after one growing season and 4 stems/m2 when planted). Above-ground cordgrass biomass was 466–1,840 g/m2 (vs 20–104 g/m2 after one growing season). Amongst planted plots, results depended on the mudflat and the age/form of the cordgrass (see original paper) and whether plants were treated with root dip before planting (see Action: Apply root dip to plants before planting). Fertilizer typically had no significant effect on the results (see Action: Add inorganic fertilizer before/after planting). Methods: In July 1976, smooth cordgrass was transplanted into seventy-two 12.5-m2 plots (50 plants/plot, 50 cm apart) across two intertidal mudflats. Transplants were dug from existing salt marshes (young, mature short-form or mature tall-form). Some plants were treated with root dip before planting, and some plots were fertilized after planting. Cordgrass was monitored over the growing season in 1976 and 1977. Monitoring included counting stems, measuring representative flowering stems, and cutting, drying and weighing three cordgrass plants/plot.Study and other actions tested
A replicated, site comparison study of salt marshes in North Carolina, USA (Seneca et al. 1985) reported that transplanted smooth cordgrass Spartina alterniflora did not clearly change in height over three growing seasons, but increased in density and biomass (with biomass reaching similar levels to natural marshes). Statistical significance was not assessed. After three growing seasons, cordgrass shoots were 125–158 cm tall on average (vs 109–173 cm after one growing season). Plots contained 203–275 cordgrass stems/m2 (vs 52–96 stems/m2 after one growing season) and 676–1,241 g/m2 above-ground cordgrass biomass (vs 121–272 g/m2 after one growing season). After three growing seasons, plots planted with tall-form cordgrass supported an above-ground cordgrass biomass of 1,068 g/m2, compared to an average of 1,168 g/m2 in five nearby natural, tall-form marshes. Methods: In April, smooth cordgrass was planted (90 cm apart) into fifteen plots on an area of recently deposited and graded intertidal sediment (year and number of plants not reported). Cordgrass plants were dug from four natural marshes and had different initial growth forms (short, intermediate or tall). In September after 1–3 growing seasons, transplanted cordgrass growing 50–60 cm above sea level was monitored: height of five shoots/plot; density and dry above-ground biomass in one 0.25–1 m2 quadrat/plot.Study and other actions tested
A replicated study in 1977 on intertidal sediment in Texas, USA (Webb & Dodd 1989) reported that 20–91% of planted smooth cordgrass Spartina alterniflora survived for two months, with increases in cordgrass density and height over one growing season. Statistical significance was not assessed. After one growing season, planted plots contained 21–230 cordgrass stems/m2 (vs 4 stems/m2 when planted). Cordgrass plants were 54–157 cm tall (vs 48–59 cm when planted). Cordgrass cover was <10–50% (initial cover not reported). Amongst the plots, results depended on planting date, elevation and the combination of the two (see original paper for full details). For example, after one growing season, plots planted in February supported higher cordgrass densities than plots planted in May at the lowest elevation (February: 153; May: 2 stems/m2), but the opposite was true at highest elevation (February: 21; May: 56 stems/m2). Fertilizer typically had no significant effect on the results (see Action: Add inorganic fertilizer before/after planting). Methods: In 1977, four hundred and fifty 15-m2 plots were established, in three sets of 150, at varying elevations on created intertidal land (sediment deposited and graded, protected by a breakwater and fenced). All plots were planted with field-collected cordgrass (60 plants/plot): half in February and half in May. Most plots were also fertilized. After two months (April or July) and one growing season (November), the central 30 cordgrass plants in each plot were surveyed.Study and other actions tested
A site comparison study in 1989 of two estuarine salt marshes in California, USA (Langis et al. 1991) found that a marsh created by reprofiling, planting California cordgrass Spartina foliosa and fertilizing contained less cordgrass biomass, after 4–5 years, than an adjacent natural marsh. The created marsh contained 192 g/m2 above-ground California cordgrass biomass: significantly lower than the 454 g/m2 in the natural marsh. Methods: In July 1989, California cordgrass was cut from 9–12 quadrats at a similar elevation in the two marshes, then dried and weighed. One marsh (same marsh as in Study 8) had been created by reprofiling into islands and creeks (autumn 1984), planting California cordgrass along creek banks (March 1985) and fertilizing with urea (25 g/m2; four times 1985–1986). This study evaluates the combined effect of these interventions on any non-planted cordgrass. A nearby natural marsh, exposed to similar tides, was chosen for comparison.Study and other actions tested
A site comparison study in 1989 of four estuarine salt marshes in California, USA (Zedler 1993) found that a marsh created by reprofiling, planting California cordgrass Spartina foliosa and fertilizing supported a similar cordgrass density to adjacent natural marshes, but with shorter plants. Statistical significance was not assessed. Four years after planting, four of four transects in the created marsh supported a cordgrass density (133–173 stems/m2) within the range of nearby natural marshes (73–193 stems/m2). However, cordgrass was shorter in the created than natural marshes, with a greater proportion of stems in shorter height classes (see original paper for data). Methods: In September 1989, California cordgrass was surveyed in 0.1-m2 quadrats. Twelve quadrats (four transects) were surveyed in a created marsh (reprofiled into islands and creeks in 1984, planted with California cordgrass in 1985, fertilized with urea in 1985–1986; same marsh as in Study 7). This study evaluates the combined effect of these interventions on any non-planted cordgrass. Fifty-four quadrats (seven transects) were surveyed in three nearby natural marshes.Study and other actions tested
A replicated study in 1989–1991 in an estuary in California, USA (Gibson et al. 1994) reported that in an excavated salt marsh planted with California cordgrass Spartina foliosa, there were increases in California cordgrass density and biomass. Statistical significance was not assessed. After one growing season, there were 25 cordgrass stems/m2 and 60 g/m2 dry above-ground biomass. After two growing seasons, there were 50 cordgrass stems/m2 and 220 g/m2 dry above-ground biomass. Methods: In March 1990, California cordgrass plants were planted into four 5-m2 plots in a salt marsh that had been excavated in 1989 (ten 4-L pots of cordgrass/plot). None of these four plots received any additional treatment. California cordgrass stems were counted and measured until October 1991. Note that this study does not distinguish between the effects of planting and excavation on any non-planted cordgrass.Study and other actions tested
A replicated, paired, before-and-after, site comparison study in 1993–1997 of four salt marshes in New York, USA (Bergen et al. 2000) reported that most planted smooth cordgrass Spartina alterniflora survived the first month, and that the average height and biomass of cordgrass in planted areas typically became similar to natural cordgrass stands within 2–4 growing seasons. After one month, cordgrass survival was 50%, 60% and 99% in the three planted marshes. Cordgrass stems were 56–136 cm tall after one growing season, then 114–182 cm tall after 2–4 growing seasons. At the same time, the planted marshes had developed 15–80% cordgrass cover, 68–236 cordgrass stems/m2 and 641–2,144 g/m2 cordgrass biomass (dry, above-ground). In the majority of pairwise comparisons (see original paper), these metrics were statistically similar to existing mature cordgrass stands (where height: 137–158 cm; cover: 66–80%; density: 136–196 stems/m2; biomass: 1,477–2,138 g/m2) and greater than in degraded areas that had not been planted (where height: 34–46 cm; cover: 2–4%; density: 6–9 stems/m2; biomass: not reported). Methods: Between 1993 and 1995, smooth cordgrass was planted into bare intertidal sediment in three salt marshes (denuded by an oil spill in 1990). Plants were mostly nursery-reared seedlings (planted 30 cm apart), but some mature individuals were also planted (1–10 m apart). All seedlings were fertilized, and the sites were fenced to exclude geese. Vegetation was surveyed in September, for up to four growing seasons after planting: in planted areas (three marshes), unplanted degraded areas (four marshes) and natural cordgrass stands (four marshes).Study and other actions tested
A replicated study in 1996–1997 in two degraded, coastal, brackish marshes in Manitoba, Canada (Handa & Jeffries 2000) reported 24–100% survival of two transplanted herb species after two growing seasons, and that their cover had increased. Statistical significance was not assessed. Two growing seasons after planting, creeping alkaligrass Puccinellia phryganodes had a survival rate of 47–100% and black estuary sedge Carex subspathacea had a survival rate of 24–50%. Cover of surviving alkaligrass was 1,600–5,400 mm2/m2 (vs 200 mm2/m2 when planted). Cover of surviving estuary sedge was 700–2,800 mm2/m2 (vs 200 mm2/m2 when planted). Adding mulch or fertilizer significantly increased the cover of alkaligrass but not estuary sedge, and did not significantly affect survival rates (see Actions: Add surface mulch before/after planting and Add inorganic fertilizer before/after planting). Methods: In June 1996, plugs of alkaligrass and estuary sedge were transplanted from natural stands to 1-m2 plots within brackish marsh vegetation damaged by geese (one species/marsh; 12 plots/species; 42 plugs/plot). All plots were fenced to exclude geese. Mulch (5 mm peat layer) and/or fertilizer (N and P) were added to three quarters of each planted plot. Survival and cover (basal area) of planted vegetation in the centre of each plot were monitored until mid-August 1997.Study and other actions tested
A replicated, randomized, paired, controlled study in 1997–2000 in an estuary in California, USA (Keer & Zedler 2002) found that plots planted with salt marsh vegetation typically contained more canopy layers and taller vegetation than unplanted plots, after four growing seasons. In three of three comparisons, planted plots had more canopy layers (2.1–2.8) than unplanted plots (1.7). Plots planted with three or six species had a greater maximum vegetation height (53–56 cm) than unplanted plots (41 cm). Plots planted with only one species had a similar vegetation height (46 cm) to the unplanted plots. The study also reported data from planted plots after two growing seasons. Plots planted with multiple species had greater overall vegetation cover, more canopy layers and a greater maximum vegetation height than plots planted with single species – but a similar average vegetation height (see original paper for data). Methods: In spring 1997, eight salt marsh herbs/succulents were planted into recently reprofiled intertidal sediment. In each of five areas, 14 random 4-m2 plots were planted with 90 greenhouse-reared seedlings (eight single-species plots, three three-species plots, three six-species plots) and three random plots were left unplanted. The planting areas had recently been excavated, amended with fine sediment, tilled and levelled. Non-planted vegetation was cleared from all plots during the first two growing seasons (1997–1998), but was left to grow from the third (1999–1998). Vegetation was surveyed using transects and point quadrats, in autumn 1997–2000. This study was based on the same experimental set-up as (13), (15) and (16).Study and other actions tested
A replicated, randomized, controlled study in 1997–1998 in an estuary in California, USA (Lindig-Cisneros & Zedler 2002) found that planting salt marsh succulents reduced seedling recruitment for one species, but increased recruitment for two others. Over the second growing season after planting, there were fewer unplanted pickleweed Salicornia virginica seedlings in plots where pickleweed had been planted (71–99 seedlings/4 m2) than where it had not been planted (167–380 seedlings/4 m2). In contrast, there were more unplanted seedlings of dwarf saltwort Salicornia bigelovii and estuary seablite Suaeda esteroa in plots where each species had been planted (saltwort: 395–920; seablite: 21–137 seedlings/4 m2) than where they had not been planted (saltwort: 14–102; seablite: 3–10 seedlings/4 m2). Methods: In April 1997, eighty-five 4-m2 plots were established (in five sets of 17) on an area of recently reprofiled intertidal sediment. All plots were amended with fine sediment, tilled and levelled. Seventy plots were then planted with 90 greenhouse-reared seedlings (random mix of one, three or six plant species: sometimes including the focal species and sometimes not). The other 15 plots were left unplanted. Seedlings were counted in all plots throughout the 1998 growing season. This study was based on the same experimental set-up as (12), (15) and (16).Study and other actions tested
A replicated study in 1998–1999 in cleared and reprofiled former farmland in Florida, USA (Anastasiou & Brooks 2003) reported that 60–100% of planted saltmeadow cordgrass Spartina patens plants survived for 20 days. Survival rates varied with soil pH (acidic: 74–86%; weakly acidic: 100%; alkaline: 60–69%) but not elevation (low: 63–100%; moderate: 60–100%; high: 69–100%). Statistical significance was not assessed. Methods: In October 1998, saltmeadow cordgrass plants (nursery-reared from locally-collected seed) were planted into three 4 x 9 m plots (100 plants/plot). The plots were in an area farmed for approximately 100 years, then cleared of invasive plants and lowered to the elevation of surrounding wetlands. All plots had brackish soils (2–7 ppt). Soil pH varied between plots (acidic: 5.2; weakly acidic: 6.4; alkaline: 8.5). Elevation varied within plots (low: <30 cm; moderate: 30–60 cm; high: >60 cm above mean tide level; approximate values). Cordgrass plants that were “severely stressed” (<25% green stems, no new growth, wilted) 20 days after planting, and that did not recover over the following 300 days, were considered dead.Study and other actions tested
A replicated, randomized, paired, controlled study in 1997–2000 in an estuary in California, USA (Callaway et al. 2003) found that plots planted with salt marsh vegetation contained more above-ground plant biomass, after three growing seasons, than unplanted plots. Results summarized for this study exclude litter and are not based on assessments of statistical significance. On average, plots planted with 3–6 species contained 372–431 g/m2 biomass whilst plots planted with a single species contained 277 g/m2 biomass. Unplanted plots contained only 94 g/m2 biomass. In single-species plots, the biomass of the planted species ranged from <1 g/m2 (arrowgrass Triglochin concinna) to 490 g/m2 (pickleweed Salicornia virginica). The biomass of unplanted species in these plots was 1–102 g/m2. Methods: In spring 1997, eight salt marsh herbs/succulents were planted into recently reprofiled intertidal sediment. In each of five areas, 14 random 4-m2 plots were planted with 90 greenhouse-reared seedlings (eight single-species plots, three three-species plots, three six-species plots) and three random plots were left unplanted. The planting areas had recently been excavated, amended with fine sediment, tilled and levelled. Non-planted vegetation was cleared from all plots during the first two growing seasons (1997–1998), but was left to grow in the third (1999). In January 2000, standing vegetation was cut from a 20 x 120 cm quadrat in each plot, then dried and weighed. This study was based on the same experimental set-up as (12), (13) and (16).Study and other actions tested
A replicated study in 1997–1998 in an estuary in California, USA (Zedler 2003) reported ≥81% survival of eight planted salt marsh species through their first growing season, and ≥58% survival through their first winter. Over the first growing season, the survival rate ranged from 81% for saltwort Batis maritima to 99% for alkali heath Frankenia salina (overall: 93%). Over the first winter, the survival rate ranged from 58% for arrowgrass Triglochin concinna to >99% for pickleweed Salicornia virginica (overall: 82%). The study suggests that winter mortality was related to smothering by algae and sediment, and feeding/trampling by waterbirds. Colonization by non-planted seedlings was also reported (see Study 13). Methods: In April 1997, greenhouse-reared herbs/succulents were planted into seventy-two 4-m2 plots, on an area of recently reprofiled intertidal sediment (90 seedlings/plot; 1–6 species/plot; eight species total). Plots were amended with fine sediment, tilled and levelled before planting. Seedlings were watered regularly after planting. Dead planted seedlings were replaced. Non-planted seedlings were removed. Survival was assessed in July 1997 and March 1998. This study was based on the same experimental set-up as (12), (13) and (15).Study and other actions tested
A replicated study in 2000–2002 in an estuary in California, USA (Zedler 2003) reported <1–80% survival of six planted salt marsh species. Of 1,332 seedlings planted in April 2000, only 9% survived their first growing season (from <1% of salt marsh daisy Jaumea carnosa to 16% of alkali heath Frankenia salina). For 180 seedlings planted in December 2000, the survival rate over one year was 48% (from 30% for estuary seablite Suaeda esteroa to 80% for saltwort Batis maritima). For 504 seedlings planted in March 2001, the survival rate was 70% over the first growing season (68–90% per species) then 62% over the first winter (45–82% per species). The study identified high salinities, waterlogging, limited tidal flushing and sediment deposition as possible causes of mortality. Methods: Between April 2000 and March 2001, greenhouse-reared herbs/succulents were planted into an area of recently reprofiled intertidal sediment (36–48 plots/trial; 1–6 species/plot; 1–42 seedlings/plot, ≥5 cm apart). In two of three trials, dead planted seedlings were replaced. Survival was assessed in July 2000, December 2001, July 2001 and January 2002. One trial in this study used a subset of the plots in (20).Study and other actions tested
A replicated, site comparison study in 1998 of a salt marsh near Christchurch, New Zealand (Thomsen et al. 2005) reported 0–100% survival of planted herbs after nine months, and that surviving plants had a similar biomass to those in a nearby natural marsh. Statistical significance was not assessed. Three-square bulrush Schoenoplectus pungens did not recover following the expected winter die-back, two months after planting. In contrast, 100% of planted wire rush Leptocarpus similis and 73% of planted sea rush Juncus maritimus were alive, but appeared stressed, after nine months. Surviving wire rush and sea rush had a dry above-ground biomass of 0.08–0.19 g/plant, compared to 0.12–0.22 g/plant in a nearby natural marsh. Methods: In March 1998, the three herb species were planted into thirty-six 0.25-m2 plots (12 plots/species; four plants/plot) within an estuarine salt marsh. The plots started as bare sediment: half were re-filled with marsh mud and half were re-filled with treated sewage/industrial waste. Within each plot, two plants were sourced from a nearby marsh and two were nursery-reared. All plants were clipped to 20 cm height. Plots were regularly cleared of debris. Planted vegetation was monitored for up to nine months. Biomass was estimated from height measurements.Study and other actions tested
A replicated, site comparison study in 1998–2002 involving a reprofiled, planted and fenced salt marsh in California, USA (Armitage et al. 2006) reported 88–98% survival of planted salt marsh species after one growing season, but lower cover of vegetation overall and the dominant plant species compared to a natural marsh after three growing seasons. Statistical significance was not assessed. After one growing season, survival rates ranged from 88% for saltgrass Distichlis spicata to 98% for salt marsh daisy Jaumea carnosa. After three growing seasons, total vegetation cover in the created marsh was 62% (mostly pickleweed Salicornia virginica: 39% cover). In a nearby natural marsh total vegetation cover was 87% (mostly pickleweed: 62% cover). Each species had greater cover in plots where it was planted (pickleweed: 38–72%; other species: 10–57%) than where it colonized by itself (pickleweed: 23–27%; other species: <0.5%). For some species, the final cover and canopy height depended on planting density (see original paper). Methods: In autumn 1997, an upland area was reprofiled to form an intertidal mudflat. In March 1998, rooted cuttings of four salt marsh herb/succulent species were planted into fifty-five 4-m2 plots around the edge of the mudflat (25–81 plants/plot; combinations of 1 or 2 species/plot). After one growing season, the plots were protected with rabbit-proof fencing. Debris and colonizing vegetation were regularly removed during the first two growing seasons, but left in place thereafter. The study does not distinguish between the effects of reprofiling, planting and fencing on the non-planted vegetation. Survival of planted individuals was monitored after one growing season. Total vegetation cover was measured in 0.25-m2 quadrats: in the created marsh (1 quadrat/plot/year until October 2000) and a nearby natural marsh of similar elevation (10 quadrats in July 1999).Study and other actions tested
A replicated study in 2000–2002 in an estuary in California, USA (O’Brien & Zedler 2006) reported 31–83% survival of five planted salt marsh species over the second year after planting began, and that the average size of survivors increased. In December 2001, the study site contained 108 plants of each of five species (one plant/species in 108 plots). In August 2002, 33–90 plants/species were still alive, with an average of 2.7–3.5 surviving plants/plot. Initial planting occurred in December 2000, but dead plants were replaced until December 2001 to maintain the total of 108 plants/species (129–290 replacements/species). Between October 2001 and August 2002, the average size of surviving plants increased in 15 of 15 comparisons (statistical significant not assessed; data reported as an index combining height and lateral extent). Survival rates and plant size were typically increased by adding kelp compost to plots (see Action: Add below-ground organic matter before/after planting) but not significantly affected by the spacing of planting or excavation of tidal creeks (see original paper and Action: Facilitate tidal exchange before/after planting). Methods: In December 2000, one-year-old, greenhouse-reared herbs/succulents were planted into intertidal sediment excavated the previous winter. The species were saltwort Batis maritima, alkali heath Frankenia salina, salt marsh daisy Jaumea carnosa, California sea lavender Limonium californicum and estuary seablite Suaeda esteroa. Half of the plots were in the catchment of excavated tidal creeks. Kelp compost was tilled into some plots, some plots were tilled, and some were left undisturbed. Colonizing vegetation was removed until October 2001. This study included some of the plots used in (17).Study and other actions tested
A replicated, paired, controlled study in 2000–2004 in a degraded brackish marsh in New Jersey, USA (Wang et al. 2006) reported 67–100% survival of five planted herb species over two years, and that survivors grew in 8 of 10 cases. Statistical significance was not assessed. Survival rates were lowest for saltmarsh rush Juncus gerardii and saltmarsh bulrush Scirpus robustus (67% or 100% across two cases) and highest for black rush Juncus roemerianus (100% in two of two cases). In 8 of 10 cases, surviving plants grew in height (4–241% increase) and circumference (21–251% increase) over the second year after planting. In the other two cases, plant circumference decreased by 16–78% and height changed by ≤15%. The study also reported that areas planted with the herbs (and some shrubs) contained fewer common reed stems (7–25 stems/m2) than adjacent unplanted areas (66–149 stems/m2). Methods: In summer–autumn 2000–2002, five herb and three shrub species were planted in three areas on the edge of a marsh (4–7 species/area; 4–48 plants/species/area; individual plants 60–100 cm apart). Plants were collected from the wild or grown from tissue in a laboratory. Invasive common reed Phragmites australis had been cleared <1 year before planting, by applying herbicide and cutting. Plant survival and size were recorded 1–2 years after planting. Common reed stems were counted in the planted areas and three adjacent unplanted areas, 2–4 years after reed clearance.Study and other actions tested
A study in 1995–2003 of brackish wetland patches within a park in New York, USA (Galbraith-Kent & Handel 2007) reported that 11 of 20 wetland herbs planted in 1995 were still present two years later, and that eight of these increased in area over the second year after planting. Statistical significance was not assessed. Between one and two years after planting, ovate spikerush Eleocharis ovata was the species that increased most in area (from 10 m2 to 112 m2). Lizard’s tail Saururus cernuus was the species that declined most in area (from 102 m2 to 0 m2). The study also reported that most wetland patches, especially the smallest ones, were invaded by common reed Phragmites australis and purple loosestrife Lythrum salicaria 7–8 years after restoration (not quantified). Methods: In late spring 1995, twenty wetland herb species were planted in nine wetland patches next to a brackish lake. Approximately 10,000 nursery-reared plants were planted at appropriate elevations. The site had been disturbed by pipeline maintenance, but then graded to create nine wetland patches (125–536 m2) and cleared of common reed using herbicide and plastic sheeting. The study does not distinguish between the effects of these interventions and planting on any non-planted individuals. The area covered by each planted species was mapped in early summer 1996 and 1997. Vegetation was surveyed qualitatively in 2002 and 2003.Study and other actions tested
A replicated, paired, site comparison study in 2001–2004 of six salt marshes in North Carolina, USA (Currin et al. 2008) found that restored marshes – planted with cordgrasses Spartina spp. and protected with breakwaters – typically contained less, and shorter, smooth cordgrass than natural marshes. Averaged over the 22 or 31 months after planting, smooth cordgrass cover was lower in restored than natural marshes in three of three comparisons (restored: 10–26%; natural: 33–46%). Smooth cordgrass density was lower in restored than natural marshes in two of three comparisons (for which restored: 70–162 stems/m2; natural: 150–222 stems/m2; other comparison no significant difference). Smooth cordgrass plants were shorter in restored than natural marshes in three of three comparisons (restored: 50–62 cm; natural: 64–82 cm). Methods: Between autumn 2001 and summer 2002, three degraded salt marshes were restored. Rocky breakwaters were built offshore, then cordgrasses Spartina spp. (mainly smooth cordgrass Spartina alterniflora) were planted. The study does not distinguish between the effects of planting and the breakwaters on non-planted vegetation. For each planted marsh an adjacent, physically similar, natural marsh was selected for comparison. Smooth cordgrass was monitored along transects each spring and autumn for up to 31 months after intervention. Cover was estimated in 1-m2 plots, stems were counted in 0.25-m2 subplots, and the three tallest stems/plot were measured.Study and other actions tested
A replicated study in 2006 in an estuarine salt marsh in California, USA (Varty & Zedler 2008) reported that survival of transplanted dwarf saltwort Salicornia bigelovii seedlings depended on plot elevation and thinning of the dominant competitor. After one growing season, <40% of seedlings transplanted into 10-cm depressions were still alive. In contrast, 70% of seedlings transplanted into 5-cm depressions or level plots were still alive. The survival rate of transplants was 2.4 times greater in plots where dominant pickleweed Salicornia virginica had been thinned (to 50% cover) than where it had not been thinned (>75% cover). Methods: In March 2006, four dwarf saltwort seedlings were planted in each of seventy-two 0.25-m2 plots on a pickleweed-dominated salt marsh. Dwarf saltwort was also sown onto each plot (1.25 ml seed/plot). In some of the plots, the surface was lowered by 5–10 cm and/or pickleweed stems were cut and removed before planting. Survival was monitored in September 2006. This study was in the same area as (16b) and (19), but used different plots.Study and other actions tested
A replicated, before-and-after study in 2006–2008 on estuarine mudflats in southern Spain (Castillo & Figueroa 2009) reported that planted clumps of herbaceous vegetation survived and expanded, but that an invasive grass colonized some sites. After one year, 75–99% of planted small cordgrass Spartina maritima clumps had survived. Survival varied with location (flat plain < sloping banks). Surviving clumps had expanded horizontally by 1.1 cm/month, on average. Clumps of glasswort Sarcocornia perennis, introduced as fragments within the cordgrass clumps, had also expanded horizontally by 1.8 cm/month. Seedlings of invasive denseflower cordgrass Spartina densiflora appeared in three sites (abundance not quantified). Methods: Between November 2006 and January 2007, salt marsh vegetation was planted into polluted, unvegetated, tidal mudflats in the Odiel Estuary (number of sites not reported). All sites were planted with cordgrass-dominated clumps, collected from natural marshes (1 clump/m2; approximately 20 cordgrass shoots/clump). Sea purslane Atriplex portulacoides, was also planted around the edge of some sites. Expansion was monitored for 21–76 clumps/herb species (further details not reported).Study and other actions tested
A replicated, controlled, site comparison study in 2010–2012 in a salt marsh in Georgia, USA (Cain & Cohen 2014) found that the density and height of smooth cordgrass Spartina alterniflora increased in plots planted with cordgrass plants, and that after three growing seasons cordgrass density was similar in planted plots and natural marshes. The total number of live stems in plots planted with cordgrass increased from 30–35 stems/m2 at planting to 345–369 stems/m2 after three growing seasons. Maximum cordgrass height increased from 45–48 cm to 56–58 cm. Adding alginate did not significantly affect cordgrass density or height (see Action: Add below-ground organic matter before/after planting). After three growing seasons, planted plots contained taller cordgrass than mature natural marshes, but at a similar density (natural marshes: 29 cm tall; 427 stems/m2). Planted plots contained a greater density of cordgrass than unplanted plots, but of a similar height (unplanted plots: 89 stems/m2; 40 cm tall). Methods: In May 2010, thirty 1-m2 plots were established in an estuarine salt marsh. Twenty bare mud plots were planted with swards of cordgrass from nearby natural marsh, in nine holes 45 cm apart. Alginate (a carbon-rich seaweed extract) was added to half of these plots. Five bare mud plots were not planted. The final five plots were situated in patches of natural marsh. Cordgrass stems were counted, and the five tallest stems/plot measured, in each plot over three growing seasons.Study and other actions tested
A 2016 systematic review of salt marsh restoration studies around the world (Bayraktarov et al. 2016) reported a 65% average survival rate of planted and sown vegetation. Survival ranged from 0% (2 of 64 cases) to ≥95% (7 of 64 cases). Methods: These results are based on 64 cases (e.g. different species, environments or intervention methods) from 16 publications and five countries, 63 of which involved planting or sowing salt marsh vegetation (mostly herbs and succulents, sometimes shrubs; see Appendix to original paper). Literature searches were carried out in 2014. Planting and sowing were sometimes into environments thought to be suitable (but sometimes into hostile environments) and sometimes preceded by site preparation (but sometimes not). Study duration ranged from 20 days to 13 years. Survival was sometimes estimated from other metrics, such as cover. The review does not separate results for planting vs sowing. The review includes studies (10), (14), (16), (17), (19), (20) and (25) summarized above.Study and other actions tested
A study in 2005–2013 of an excavated, planted and harvested water treatment marsh in Sardinia, Italy (De Martis et al. 2016) reported that it supported 275 plant taxa. This included 201 plant species in 161 genera. Approximately 63% of the taxa were Mediterranean (found predominantly or solely in this region) and approximately 16% were known non-natives in Italy. As expected in the study area, 56% of the taxa were annual plants that complete their life cycle rapidly in favourable conditions (“thereophytes”). Only 2% of taxa had underwater resting buds (“hydrophytes”). Methods: Between 2005 and 2013, plant taxa were recorded in the 37-ha EcoSistema Filtro marsh, which had been constructed with the dual aims of habitat creation and water treatment. There were monthly surveys (a) across the whole site, including banks and upland areas, and (b) in three 16-m2 plots, each April–July and September–December. The wetland had been constructed by excavating basins of varying salinity and levees (including removal of all existing vegetation; beginning 1990) and planting bundles of 2-m-tall common reed Phragmites australis (2004). Some “plant biomass” was mechanically removed between 2005 and 2007. Note that this study evaluates the combined effect of these interventions, and does not separate results from fresh, brackish and saline areas.Study and other actions tested
A replicated, site comparison study in 2013–2015 around two fresh/brackish lakes in South Australia (Jellinek et al. 2016) found that planted stands of river club-rush Schoenoplectus tabernaemontani became more similar to mature natural stands over time – in terms of structure and rush abundance – and supported similar near-shore vegetation to the natural stands within 8 years. Older planted rush stands were more similar to mature natural stands in terms of stand width (young planted: 1–3 cm; old planted: 5–12 cm; natural: 35 cm), maximum height (young: 60–142 cm; old: 131–152 cm; natural: 155 cm) and stem density (data not reported). All stands were a similar average height (data not reported). Near-shore vegetation (i.e. between the rush stands and the shoreline) behind older planted rush stands was similar to that behind mature natural stands, whereas young planted stands supported similar near-shore vegetation to areas without rush stands. This was true for overall community composition (data reported as graphical analyses; statistical significance of differences not assessed), plant species richness (no rushes: 30; young: 45; old: 150; natural: 330 species/site) and abundance (no rushes: 940; young: 1,370; old: 14,000; natural: 31,300 plants/site). Methods: In autumn 2013–2015, vegetation was surveyed at 21 sites on the margins of two connected fresh/brackish lakes. Ten sites had been planted with nursery-reared rushes (1 m apart): six sites ≤3 years ago (young plantings) and four sites 8–11 years ago (old plantings). Three sites had mature natural rush stands (≥20 years old) and eight had no rushes. All sites were fenced to exclude livestock. Rush stands were surveyed in five 1-m2 quadrats/site/year. Other near-shore vegetation was surveyed in approximately thirty-six 3-m2 quadrats/site/year.Study and other actions tested
A before-and-after study in 2011–2016 of an intertidal site in Florida, USA (Donnelly et al. 2017) reported 53% survival of transplanted smooth cordgrass Spartina alterniflora after two years, and increases in cordgrass abundance and height over five years. Before planting, the mid-intertidal zone was sparsely vegetated (<1 cordgrass shoots/m2; 2% cover). Five years after planting smooth cordgrass into this zone, its density had increased to 56 shoots/m2 and its cover had increased to 52%. The average height of planted cordgrass had increased, from 37 cm when planted to 67 cm after five years (statistical significance not assessed). No natural recruitment was observed within the first three years after planting (data not reported after this). Methods: The study took place along a 200 m stretch of shoreline, on the edge of an ancient shell waste dump. In April/May 2011, smooth cordgrass Spartina alterniflora was planted in the mid-intertidal zone (620 nursery-reared plants; 3 plants/m). Mangrove seedlings were planted in the upper intertidal zone and oyster-shell mats were placed in the lower intertidal zone. The study does not distinguish between the effects of these interventions on natural recruitment. Cordgrass in the mid-intertidal zone was surveyed before planting (presumably April 2011) and for five years after (2011–2016).Study and other actions tested