Conservation Evidence strives to be as useful to conservationists as possible. Please take our survey to help the team improve our resource.

Providing evidence to improve practice

Action: Remove upper layer of peat/soil (without planting) Peatland Conservation

Key messages

Read our guidance on Key messages before continuing

  • Ten studies evaluated the effects of removing the upper layer of peat or soil (without planting afterwards) on peatland vegetation. Nine studies were in fens or fen meadows and one was in an unspecified peatland.
  • Plant community composition (6 studies): Five studies (including one replicated, randomized, paired, controlled) in a peatland in the USA and fens or fen meadows in the Netherlands and Poland reported that plots stripped of topsoil developed plant communities with a different composition to those in unstripped peatlands. In one study, the effect of stripping was not separated from the effect of rewetting. Two studies in fen meadows in Germany and Poland reported that the depth of soil stripping affected plant community development.
  • Characteristic plants (5 studies): Four studies in fen meadows in Germany and the Netherlands, and a peatland in the USA, reported that stripping soil increased cover of wetland-characteristic or peatland-characteristic plants plants after 4–13 years. In the Netherlands, the effect of stripping was not separated from the effect of rewetting. One replicated site comparison study in fens in Belgium and the Netherlands found that stripping soil increased fen-characteristic plant richness.
  • Herb cover (4 studies): Three studies (including one replicated, paired, controlled) in fens or fen meadows in Germany, the UK and Poland found that stripping soil increased cover of rushes, reeds or sedges after 2–6 years. However, one controlled study in a fen meadow in the Netherlands reported that stripping soil had no effect on sedge or bentgrass cover after five years. Two controlled studies in a fen meadow in the Netherlands and a fen in the UK found that stripping soil reduced purple moor grass cover for 2–5 years.
  • Vegetation structure (3 studies): Two studies in fens or fen meadows in the Netherlands and Belgium found that stripping soil reduced vegetation biomass (total or herbs) for up to 18 years. One replicated, randomized, paired, controlled study in a peatland in the USA found that stripping soil had no effect on vegetation biomass after four years.
  • Overall plant richness/diversity (6 studies): Three studies (including one replicated, paired, controlled) in fens or fen meadows in the UK, Belgium and the Netherlands reported that stripping soil increased total plant species richness over 2–18 years. In one study, the effect of stripping was not separated from the effect of rewetting. One replicated, controlled study in a fen in Poland found that stripping soil had no effect on plant species richness after three years. One replicated, randomized, paired, controlled study in a peatland in the USA found that stripping soil increased plant species richness and diversity, after four years, in one field but decreased it in another. One replicated study in a fen meadow in Poland reported that plant species richness increased over time, after stripping soil.

Supporting evidence from individual studies

1 

A controlled study in 1991–1996 in a degraded fen meadow in the Netherlands (van Duren et al. 1998) found that stripping topsoil reduced vegetation biomass after three months, but typically had no effect on vegetation cover after five years. After three months, above-ground biomass was significantly lower in a stripped area (20–240 g/m2) than in an area that had not been stripped (200–490 g/m2). After five years, both areas were dominated by velvety bentgrass Agrostis canina (36–37% cover) and contained the same three Carex sedge species (1–4% cover) but no Sphagnum moss. However, cover of purple moor grass Molinia caerulea was only 1% in the stripped area, compared to 21% in the unstripped area. Cover results were not tested for statistical significance. In 1991, 15–20 cm of topsoil was removed from 0.5 ha of degraded fen meadow. An adjacent area was not stripped. The meadow was historically drained but had been rewetted five years before stripping. Both areas were partially fertilized, partially limed and mown every August. In August 1994, above-ground vegetation was harvested in 60 x 60 cm quadrats (number not reported), then dried and weighed. Vegetation cover was estimated in 1996 (details not reported).

2 

A study in 1991–1997 in a degraded fen meadow in Germany (Patzelt et al. 2001) reported that a plant community developed following topsoil removal, but its composition depended on the depth of soil removed. In plots with 40–60 cm of soil removed, the community contained wetland-characteristic herbs and tall rush species after six years. In plots with 20 cm of soil removed, species from drier grasslands were more abundant. All data were reported as a graphical analysis. The results were not tested for statistical significance. In February 1991, topsoil was removed from three 4,500 m2 plots in a fen meadow historically used for agriculture. A different depth of soil was removed from each plot: 20, 40 or 60 cm. None of these plots were sown with hay. From 1992 to 1997, vegetation cover was estimated annually in five 4 m2 quadrats/plot.

3 

A replicated, randomized, paired, controlled study in 2000–2004 in a degraded peatland in Ohio, USA (Hausman et al. 2007) found that plots stripped of topsoil contained significantly different plant communities to unstripped plots after four years, whilst plant species richness and diversity showed mixed results and biomass did not differ. Overall, the plant community in stripped plots contained more wetland-characteristic species and fewer upland-characteristic species than the community in unstripped plots (data reported as a graphical analysis). In one of two fields, stripped 25 m2 plots contained more plant species than unstripped plots (24 vs 15) and were more diverse (data reported as a diversity index). In the other field, stripped plots contained fewer species than unstripped plots (13 vs 20) and were less diverse. Above-ground plant biomass did not differ between treatments in either field (stripped: 168–405; unstripped: 171–415 g/0.5 m2). In 2000, twelve pairs of plots were established across two historically farmed peat fields. Topsoil (40–50 cm depth) was stripped from one plot in each pair but not from the other. Across the whole study area, the water table was raised and some seeds were sown (although the study states that most plants colonized naturally). In 2004, cover of every plant species was estimated in one 25 m2 quadrat/plot. Above-ground biomass was collected from three 0.5 m2 quadrats/plot, then dried and weighed.

4 

A replicated, paired, controlled study in 2001–2005 in a degraded fen meadow in Germany (Rasran et al. 2007) reported that topsoil removal increased the abundance of bog/fen characteristic plants. The results are not based on tests of statistical significance. Five years following topsoil stripping, bog- and fen-characteristic plants occurred in up to 3% of quadrats with up to 3% cover/plot. Peatland-characteristic plants were not present in unstripped plots. In 2001, sixteen 6 x 6 m plots (in four blocks of four) were established in a drained, abandoned, nutrient-enriched fen meadow. Topsoil (30 cm depth) was stripped from eight plots but left on the others. None of these plots were sown with hay, but four stripped and four unstripped plots were fenced to exclude cattle. Between 2002 and 2005, vegetation cover was estimated in 16 permanent 1 m2 quadrats/plot.

5 

A replicated, paired, controlled study in 1991–2002 in a degraded fen meadow in the Netherlands (van der Hoek et al. 2007) reported that plots stripped of topsoil contained different plant communities to unstripped plots. In particular, stripped plots were characterised by the absence of common haircap moss Polytrichum commune and star sedge Carex echinata. Plant communities in stripped plots also changed over time, whilst they remained stable in unstripped plots. All data were reported as a graphical analysis. The results were not tested for statistical significance. In 1991, surface vegetation and 10–15 cm of organic soil were stripped from two plots in an acidified fen meadow. Two adjacent plots were not stripped. Excess rainwater was drained by ditches. In 1993, 1997, 1999 and 2002, vegetation cover was estimated in representative areas of all four plots.

6 

A replicated, paired, controlled study in 2006–2008 in a degraded, grassy fen in Northern Ireland, UK (Reid & McEvoy 2009) found that plots stripped of surface peat had greater plant species richness than unstripped plots after two years, greater cover of rushes Juncus spp. and sedges Carex spp., but less cover of total vegetation and purple moor grass Molinia caerulea. Stripped plots contained more species/0.5 m2 than unstripped plots (5.1 vs 3.9) and had greater cover of rushes (55 vs 2%) and sedges (13 vs <1%). However, stripped plots had less cover than unstripped plots of vegetation in total (65 vs 100%) and purple moor grass (11 vs 78%). Results were similar after one year, with the exception of species richness which did not differ significantly between stripped and unstripped plots (3.6 vs 3.9 species/0.5 m2). In autumn 2006, four pairs of 5 x 5 m plots were established in a fen dominated by moor grass. Surface peat (15 cm depth) and vegetation were stripped from one plot in each pair, but not from the other. In July 2007 and October 2008, cover of every plant species was estimated in eight 70 x 70 cm quadrats/plot.

7 

A replicated site comparison study in 2004–2007 in a drained fen meadow in Poland (Klimkowska et al. 2010) reported that topsoil stripping changed the plant community composition, and that vascular plant cover and plant species richness increased over time after stripping. These results are not based on tests of statistical significance. Over three years following topsoil stripping, the overall plant community composition changed: it became less like degraded fen meadows, but also less like target fen meadow vegetation. The community also differed between plots stripped to different depths (data reported as a graphical analysis). Over the same time period, there were increases in vascular plant cover (from 2–3% to 58–75%) and plant species richness (from 5–8 species/4 m2 to 18–19 species/4 m2). In 2004, topsoil was stripped from eight 8 x 16 m plots in a drained fen meadow: 40 cm from four plots and 20 cm from the other four. All of these plots were left open to grazing by boar and deer, and were mown in 2006 and 2007. None of these plots were sown with hay. Vegetation cover and plant species were recorded annually between 2004 (after soil removal) and 2007, in each plot and in nearby degraded and target (reference) meadows.

8 

A replicated, controlled study in 2008–2011 in a degraded fen in Poland (Hedberg et al. 2014) reported that plots stripped of topsoil developed a different plant community to unstripped and natural plots, but found that all plots contained a similar number of vascular plant species. After three years, the overall composition of the plant community differed between stripped, unstripped and natural plots. In particular, stripped plots had greater cover of jointleaf rush Juncus articulatus, cattail Typha latifolia and common duckweed Lemna minor than both unstripped and natural plots (data reported as a graphical analysis; differences not tested for statistical significance). However, the number of vascular plant species did not significantly differ between treatments (data not reported). In December 2008, 60 cm depth of topsoil was stripped from two 0.5 ha plots in the drained, degraded fen. Soil was not stripped from five adjacent plots. None of these plots were sown with hay. Ten plots in two natural fens were also monitored. In summer 2011, cover of every vascular plant species was estimated in each plot.

9 

A replicated site comparison study in 2013 in six degraded rich fens in Belgium and the Netherlands (Emsens et al. 2015) found that plots stripped of surface peat contained significantly less herb biomass than unstripped plots after 3–18 years, but had significantly greater bryophyte cover and plant species richness (data not reported). The higher species richness in stripped plots applied to both the total number of plant species and the number of fen-characteristic plant species. The differences in herb biomass and species richness were consistent in all eight sites. In June 2003, vegetation cover was recorded in eight 2 x 2 m quadrats/fen: four quadrats in an area stripped of surface peat and four in an unstripped area. In the stripped areas, 15–30 cm of peat had been removed (leaving some peat below) 3–18 years previously. Historically, all eight fens were drained and used for agriculture. Two had since been rewetted.

10 

A controlled study in 1995–2008 in a degraded fen meadow in the Netherlands (Klimkowska et al. 2015) reported that plots stripped of topsoil (and rewetted) developed different plant communities to unstripped (and drier) plots, with more plant species and greater moss/liverwort cover. Most of these results were not tested for statistical significance. Over 13 years, restored plots developed a different plant community (with fen meadow-characteristic species) to unrestored plots (dominated by species characteristic of drier, nutrient-rich sites; data reported as a graphical analysis). After 13 years, there were 24 species/4 m2 in restored plots (vs 21 species/4 m2 in unrestored plots). Restored plots also had significantly greater moss/liverwort cover (78–83%) than unrestored plots (23%). Results were similar in areas with shallow and deep topsoil removal. In 1995, an area of drained fen meadow was restored by stripping topsoil (shallow: 20cm; deep: 40 cm) and rewetting (by blocking local drainage ditches). The study does not distinguish between the effects of these interventions. An adjacent area was not stripped of topsoil or locally rewetted. Two interventions affected the whole site: additional rewetting by blocking a large drainage ditch in 2000, and reinstated annual mowing from 2001. Between 1997 and 2008, cover of every plant species was estimated in permanent 4 m2 plots: 16 in the restored area and two in the unrestored area.

Referenced papers

Please cite as:

Taylor N.G., Grillas P. & Sutherland W.J. (2018) Peatland Conservation. Pages 329-392 in: W.J. Sutherland, L.V. Dicks, N. Ockendon, S.O. Petrovan & R.K. Smith (eds) What Works in Conservation 2018. Open Book Publishers, Cambridge, UK.