Cut/mow herbaceous plants to maintain or restore disturbance: brackish/salt marshes

Supporting evidence from individual studies

1. A controlled study in 1977–1979 in a brackish marsh in Mississippi, USA (Hackney & de la Cruz 1981) reported that cutting reduced black rush Juncus roemerianus height and dominance, and reduced big cordgrass Spartina cynosuroides dominance. Statistical significance was not assessed. In an initially rush-dominated area, black rush reached a maximum height of 130–134 cm in cut plots, in the year following the final cut (vs 203 cm in uncut plots). Rushes comprised only 39–80% of all plant biomass in cut plots (vs 62–94% in uncut plots). In an initially cordgrass-dominated area, cordgrass comprised only 33–97% of all plant biomass in cut plots (vs 61–99% in uncut plots). Methods: Plots in a tidal brackish marsh, dominated by black rush or big cordgrass, were cut once (1979), twice (1978 and 1979) or three times (1977, 1978 and 1979). Some additional plots were left uncut. Cutting was done in winter and cuttings were removed. The marsh was historically burned, but not since 1973. Vegetation was surveyed from April to November 1979. The study does not report further details of the methods.

   Study and other actions tested

   Referenced paper

   Hackney C.T. & de la Cruz A.A. (1981) Effects of fire on brackish marsh communities: management implications. Wetlands, 1, 75-86.

2. A site comparison study in 1994 of two reedbeds in a fresh/brackish wetland in Denmark (Kristiansen 1998) found that a recently cut reedbed supported a lower density of tall common reed Phragmites australis with a smaller basal area than a more mature reedbed, although reed diameter was similar in both areas. In a reedbed cut two years before measurement, tall common reed stems were less dense (217 stems/m²) than in a more mature reedbed, cut seven years before measurement (422 stems/m²). However, the diameter of these reed stems did not significantly differ between the reedbeds (recently cut: 2.9; more mature: 3.1 mm). Combining these metrics, tall reed stems occupied a smaller proportion of the more recently cut reedbed (1,497 mm²/m²) than the more mature reedbed (3,014 mm²/m²). Methods: In 1994, vegetation was surveyed around 14 points in each of two reedbeds: one last cut in 1992 and one last cut in 1987. The reedbeds had been commercially harvested for “many years” previously. Most points were around (but approximately 2 m from) greylag goose Anser anser nests. At each point, all reed stems >75 cm tall were counted in four 0.25-m² quadrats. Twenty stems/quadrat were measured.

   Study and other actions tested

   Referenced paper

   Kristiansen J.N. (1998) Nest site preference by greylag geese Anser anser in reedbeds of different harvest age. Bird Study, 45, 337-343.

3. A replicated, site comparison study in 1998–1999 of 13 brackish reedbeds in southern France (Poulin & Lefebvre 2002) found that cut reedbeds contained a similar number of plant species and a similar live reed structure to uncut reedbeds, but a lower density of dead reeds. Plant species richness did not significantly differ between cut and uncut reedbeds. This was analyzed separately for emergent species (cut: 0.3; uncut: 0.6 species/quadrat) and terrestrial species (cut: 0.2; uncut: 0.5 species/quadrat). The structure of live (green) common reed Phragmites australis also did not significantly differ between cut and uncut reedbeds. This was true for density (cut: 203; uncut: 164 stems/m²), stem diameter (cut: 4.6; uncut: 4.1 mm), height (cut: 146; uncut: 137 cm) and above-ground biomass (cut: 1,853; uncut: 1,291 g/m²). However, cut reedbeds contained a lower density of dead reeds (11 stems/m²) than uncut reedbeds (311 stems/m²). Other structural metrics were not reported for dead reeds. Methods: In May–June 1998 or 1999, vegetation was surveyed in five cut reedbeds (commercially harvested each winter) and eight uncut reedbeds (never harvested, or not harvested for at least eight years). The average salinity was 2–3 ppt. In each reedbed, vegetation was surveyed in 25 quadrats (25 x 25 cm) along each of two transects (250 m long). All standing reed stems were counted. One random living reed stem was measured in each quadrat. Biomass was calculated from density, diameter and height measurements.

   Study and other actions tested

   Referenced paper

   Poulin B. & Lefebvre G. (2002) Effect of winter cutting on the passerine breeding assemblage in French Mediterranean reedbeds. Biodiversity and Conservation, 11, 1567-1581.

4. A replicated, site comparison study in 1999 of eight brackish reedbeds in southern France (Schmidt et al. 2005) found that cut reedbeds contained fewer dead reeds than uncut reedbeds, but that cutting had no significant effect on live reed density, live reed height, plant species richness and non-reed cover. Cut reedbeds contained a significantly lower density of dead common reed Phragmites australis (5 stems/m²) than uncut reedbeds (224 stems/m²). However, there was no significant difference between treatments for live reed density (cut: 198; uncut: 107 stems/m²), live reed height (cut: 129; uncut: 165 cm), total plant species richness (cut: 5.0; uncut: 5.0 species/reedbed) and cover of plants other than common reed (cut: 12%; uncut: 10%). Methods: In late July 1999, vegetation was surveyed in five cut reedbeds (harvested each winter for ≥5 years) and eight uncut reedbeds (not harvested for >5 years). The average salinity was 3 ppt. Vegetation was surveyed in 24 quadrats (50 x 50 cm) in each reedbed. This included counting all standing reed stems and measuring one random living reed stem/quadrat.

   Study and other actions tested

   Referenced paper

   Schmidt M.H., Lefebvre G., Poulin B. & Tscharntke T. (2005) Reed cutting affects arthropod communities, potentially reducing food for passerine birds. Biological Conservation, 121, 157-166.

5. A replicated, randomized, controlled, before-and-after study in 2004–2006 in a brackish marsh in South Africa (Russell & Kraaij 2008) found that cutting common reed Phragmites australis had mixed effects on vegetation structure after 1–2 years, depending on water levels. For example in the driest zone, cut plots always contained more reeds than uncut plots (cut: 67–68; uncut: 17–28 shoots/1.5 m²), but there was no difference in overall reed volume (cut: 5,150–5,620; uncut: 3,280–6,880 cm³/1.5 m²). In cut plots, reeds were always shorter (cut: 21–23; uncut: 30–35 m/1.5 m²) and thinner (cut: 6; uncut: 8–9 mm diameter). Meanwhile in the wettest zone, cut plots always contained fewer reeds than uncut plots (cut: 3–6; uncut: 68–104 shoots/1.5 m²) with a smaller volume (cut: 180; uncut: 1,190–1,870 cm³/1.5 m²). The low density of reeds in wetter cut plots negates meaningful interpretation of length and diameter. Before intervention, vegetation structure did not significantly differ between plots (averaged over moisture zones; data not reported). Methods: In 2004, eight plots (each 200–400 m²) were established in a reed-invaded marsh on the edge of a brackish lake (5–8 ppt). Stabilized water levels, reduced disturbance from large herbivores and reduced fire frequency likely contributed to reed encroachment. In three random plots, reeds were clipped to ground level in late summer 2004 and 2005. Cuttings were removed. The other five plots were left uncut. All plots were perpendicular to the lake edge so were divided into three moisture zones: dry (flooded for roughly 30% of the study), moist (50%) and wet (100%). Reed structure was surveyed before (late summer 2004) and after 1–2 cuts (late summer 2005 and 2006), in six 0.25-m² quadrats/zone/plot.

   Study and other actions tested

   Referenced paper

   Russell I.A. & Kraaij T. (2008) Effects of cutting Phragmites australis along an inundation gradient, with implications for managing reed encroachment in a South African estuarine lake system. Wetlands Ecology and Management, 16, 383-393.

6. A replicated, randomized, controlled, before-and-after study in 2003–2007 in two brackish wet grasslands in Estonia (Berg et al. 2012) found that annual cutting altered the overall plant community composition, but typically had no significant effect on plant species richness or diversity. Over four years of cutting, the plant community composition in cut plots became less similar to that in uncut plots – especially in the wetter of the two grasslands (data reported as a graphical analysis). Cover of 5–6 individual plant species – including common reed Phragmites australis – significantly differed between cut and uncut plots in at least one grassland and at least some measured years (see original paper for data). In most comparisons, cut and uncut plots had statistically similar plant species richness (six of eight comparisons, for which cut: 9–19 species/4 m²; uncut: 8–17 species/4 m²; other two comparisons lower in cut than uncut plots) and diversity (16 of 16 comparisons; data reported as diversity indices). Before cutting and within each grassland, plots destined for each treatment had statistically similar plant communities, richness and diversity (see original paper for data). Methods: In August 2003, sixteen 2 x 2 m plots were established in two degraded wet grasslands. The vegetation was historically grazed, but had not been for the past 40 years. Each summer, eight random plots/grassland were cut (with shears, cuttings removed). The other plots were not cut. All plots were fenced to exclude wild boar. Plant species and their cover were recorded annually (before each cut), from 2003 to 2007.

   Study and other actions tested

   Referenced paper

   Berg M., Joyce C. & Burnside N. (2012) Differential responses of abandoned wet grassland plant communities to reinstated cutting management. Hydrobiologia, 692, 83-97.