Providing evidence to improve practice

Action: Cut/remove/thin forest plantations and rewet peat Peatland Conservation

Key messages

  • Eleven studies examined the effect of cutting/removing trees and rewetting peat (in combination): six in fens, two in bogs, and three in both fens and bogs. In four of the studies, the peatlands naturally contained some trees. Three studies were based on one experimental set-up, and two studies were based on another.
  • Plant community composition (5 studies): Of three replicated studies in fens, two in Finland found that removing trees/rewetting had no effect on the overall plant community composition whilst one in Sweden reported only a small effect. Two site comparison studies in bogs and fens in Finland found that removing trees/rewetting changed the overall plant community composition. It became less like sites that remained drained and forested.
  • Characteristic plants (2 studies): Two before-and-after studies (one site comparison, one controlled) in bogs and fens in Finland and Sweden reported that removing trees/rewetting increased the abundance of wetland-characteristic plants.
  • Moss cover (6 studies): Of five studies that examined the effect of removing trees/rewetting on Sphagnum moss, two replicated, paired studies in bogs and fens in Sweden and Finland found that the intervention increased Sphagnum cover. One replicated, before-and-after, site comparison study in forested fens in Finland found no effect. Two before-and-after studies in a bog in Finland and a fen in Sweden found mixed effects depending on site or species. Additionally, three studies (two replicated and paired) in peatlands in the UK and Finland found that removing trees/rewetting reduced cover of non-Sphagnum or forest-characteristic mosses. However, one replicated, before-and-after, site comparison study in forested fens in Finland found no effect of thinning trees/rewetting on forest mosses.
  • Herb cover (7 studies): Seven studies (including two replicated, paired, controlled) in bogs and fens in the UK, Finland and Sweden reported that removing trees/rewetting increased cover of at least one group of herbs, including cottongrasses and sedges. However, one of these studies reported loss of cottongrass from a fen where it was rare before intervention, along with reduced purple moor grass cover.
  • Vegetation structure (4 studies): One replicated site comparison study in a bog in the UK found that removing trees/rewetting increased ground vegetation height. One replicated, paired, controlled study in a fen in Sweden reported that removing trees/rewetting had no effect on canopy height after eight years. Two replicated, paired, site comparison studies in bogs and fens in Finland found that thinning trees/rewetting reduced the number of tall trees present for 1–3 years (although not to the level of natural peatlands).
  • Overall plant richness/diversity (4 studies): Two replicated, paired, controlled studies in rich fens in Sweden reported that removing trees/rewetting increased plant species richness. However, two replicated studies in fens in Finland found that removing trees/rewetting had no effect on total plant species richness or diversity.

Supporting evidence from individual studies

1 

A controlled, before-and-after study in 1994–1996 in a fen in Finland (Komulainen et al. 1999) reported that clear-cut and rewetted plots had greater cover of cottongrass Eriophorum vaginatum than plots that remained forested and drained. This was true after one year (clear-cut/rewetted: 5–80%; forested/drained: 1–20% cover) and after two years (clear-cut/rewetted: 20–90%; forested/drained: 1–40% cover). These results were not tested for statistical significance. Before intervention, cottongrass cover was 1% in eight of nine monitored plots (20% in the other). In February 1995, one area of a drained, tree-colonized fen was restored: trees were felled and removed, drainage ditches were filled or blocked, and an additional input ditch was excavated above the fen. In the restored area, the water table was 5–45 cm below the peat surface (during summer). The rest of the fen was left forested and drained (water table 20–65 cm below surface). Vegetation cover was visually estimated in 1994, 1995 and 1996, in six 60 x 60 cm plots in the clear-cut/rewetted area and three plots outside.

2 

A replicated, paired, controlled study in two afforested blanket bogs in Scotland, UK (Anderson 2010) reported that plots restored by tree felling and rewetting had greater cover of sheathed cottongrass Eriophorum vaginatum, and less cover of forest mosses, than plots that remained forested and drained. These results were not tested for statistical significance. After five years, restored plots had sheathed cottongrass cover of 44–45% (vs 11% in plots that remained forested and drained), plait moss Hypnum cupressiforme cover of 18–32% (vs 57%) and waved silk moss Plagiothecum undulatum cover of <1% (vs 6%). The combined effect of felling and rewetting was larger than the effect of felling or rewetting alone in 10 of 12 comparisons. Between 1996 and 1998, six blocks of six 40 x 100 m plots were established in drained, conifer-forested bogs. Each treatment was replicated once/block: rewetting with tree felling (debris left in place), rewetting with tree felling (debris removed), rewetting only, tree felling only, tree removal only, no intervention. Rewetting was achieved by damming plough furrows every 20 m. In rewetted plots, the water table was 8–32 cm below the peat surface during the growing season (vs drained plots: 11–38 cm below). Vegetation cover was recorded five years after intervention (details not reported).

3 

A before-and-after, site comparison study in 1994–2005 in two peatlands (one bog, one fen) in Finland (Haapalehto et al. 2010) found that restoration by tree removal and rewetting increased the abundance of wetland-characteristic plant species and some key peatland species. The overall plant community composition changed over ten years. Restored (clear-cut/rewetted) areas accumulated wetland-characteristic species, whilst unrestored (forested/drained) areas accumulated dryland- and forest-characteristic species (data reported as a graphical analysis). Specifically, cover of sheathed cottongrass Eriophorum vaginatum increased more in restored areas (from 4–11% before restoration to 20–21% ten years after) than unrestored areas (from 1–3% to 1–7%). In the fen, forest moss cover decreased in the restored area (from 17 to 4%) but increased in the unrestored area (from 26 to 42%). In the bog, Sphagnum moss cover increased more in the restored area (data not reported). In 1995, 10 ha of bog and 1 ha of fen were restored by clearing all trees/shrubs and blocking drainage ditches. In the restored areas, the water table was 5–17 cm below the peat surface (during summer). Comparisons were made with unrestored areas in each peatland (forested and drained; water table 23–45 cm below surface). In July 1994–1997 and 2005, vegetation cover was estimated in 9–12 permanent quadrats (0.5 m2) in each area.

4 

A replicated before-and-after study in 2002–2005 in two drained, tree-colonized, rich fens in Sweden (Mälson et al. 2010) reported that following rewetting and tree removal, there were small changes in plant community composition and cover. These results are not based on tests of statistical significance. The overall composition of the plant community changed following tree removal and rewetting (data reported as a graphical analysis). In one fen, cover of purple moor grass Molinia caerulea was 50% before intervention but 30% three years after. Common cottongrass Eriophorum angustifolium disappeared (having had 0.1% cover before intervention). Across both fens, cover of sedges Carex spp. was 0–1% before but 1–8% after. Cover of common reed Phragmites australis and Sphagnum mosses showed mixed responses by site or species respectively. In plots that had trees removed without rewetting, moor grass and sedge cover changed much less than under the combined treatment, whilst cottongrass cover increased (from <1 to 5%). Around winter 2002/2003, two 50 x 150 m plots (one plot/fen) were cleared of trees and rewetted by blocking drainage ditches (water table raised approximately 10 cm). Two adjacent plots (one plot/fen) were cleared of trees but remained drained. Vegetation cover was estimated before (2002) and after (2005) intervention, in 4–16 quadrats (each 0.25 m2) in the centre of each plot. This study was based on the same experimental set-up as (6) and (8).

5 

A replicated, before-and-after, site comparison study in 2006–2009 in 19 forested fens in Finland (Laine et al. 2011) found that restoration by tree thinning and rewetting did not affect the overall plant community, the number of plant species or cover of individual plant groups – except cottongrasses/sedges. The overall composition of the plant community was similar before restoration and 18 months after, and was similar in restored and natural fens (data reported as a graphical analysis). After 18 months, restored fens contained 44 plant species in total (vs 45 before restoration and 49 in natural fens). These results were not tested for statistical significance. Although cover of some plant groups changed significantly in restored fens (shrub cover from 33 to 42%; forb cover from 8 to 11%; cottongrass/sedge cover from 8 to 10%; Sphagnum moss cover from 64 to 55%; forest moss cover from 21 to 17%), the changes were mirrored in natural fens (with the exception of cottongrass/sedge cover, which decreased from 13 to 8%). In late 2007, eleven drained, densely forested fens were restored by filling drainage ditches (water table 8–16 cm below peat surface during summer) and thinning trees (from 940 to 317 stems/ha). Comparisons were made with eight nearby natural fens (water table 8–17 cm below surface; 373 trees/ha). In July 2006 and 2009, cover of every plant species was estimated in approximately twelve 1 m2 quadrats/fen.

6 

A replicated, paired, controlled, before-and-after study in 2002–2010 in three drained, tree-colonized, rich fens in Sweden (Hedberg et al. 2012) reported that following tree removal and rewetting, there were increases in plant species richness, bryophyte cover and sedge cover. These results are not based on tests of statistical significance. There were 9 plant species/0.25 m2 before intervention but 13 species/0.25 m2 eight years after. Cover of wetland-characteristic bryophytes was 33% before and 46% after, Sphagnum mosses 23% before and 33% after, sedges 1% before and 5% after. Similar changes in cover occurred in plots that were rewetted (without tree removal) or had trees removed (without rewetting). In control plots that remained both drained and forested, there was no change in the number of plant species or vegetation cover. In winter 2002/2003, four restoration treatments were applied in each drained and tree-colonized fen, in adjacent 50 x 150 m plots: cutting and removal of all trees, rewetting (by ditch blocking; water table raised by 12–25 cm), tree removal and rewetting, or none. Between 2002 (before intervention) and 2010, cover of every plant species was estimated at 20 points/plot, in 0.25 m2 quadrats. This study was based on the same experimental set-up as (4) and (8).

7 

A replicated site comparison study in 2011 across 21 blanket bogs in Scotland, UK (Gilbert 2013) found that restoration by tree felling and rewetting increased vegetation height and cover of grass-like herbs. After 5–13 years, restored bogs had significantly taller ground vegetation (21 cm) than forested/drained bogs (3 cm) and naturally open bogs (17 cm). Amongst restored sites, vegetation height declined with time since restoration (see original paper for data and statistical model). Restored bogs had significantly greater cover of grass-like herbs than forested/drained sites (mostly bare ground covered in pine needles) and natural bogs (moss-dominated; data and species not reported). In summer 2011, twenty-one bogs were surveyed: eight restored (conifers cut and drainage ditches blocked 5–13 years before surveying, raising the water table “close to the ground surface”), six degraded (conifer-forested/drained) and seven natural (unforested/undrained). In each bog, ground vegetation height (i.e. excluding trees) was measured at 45 points, distributed along fifteen 10 m transects. Details of cover measurements were not reported.

8 

A replicated, paired, controlled, site comparison study in 2002–2010 involving three degraded rich fens in Sweden (Hedberg et al. 2013) reported that clear-cut and rewetted plots developed greater plant species richness than plots that remained forested and drained, but that vegetation grew to a similar height. Most of these results were not tested for statistical significance. After eight years, clear-cut and rewetted plots contained 14 species/0.25 m2, compared to 9 species/0.25 m2 in plots that remained forested and drained. Canopy height (of vegetation that grew following intervention) in clear-cut/rewetted plots was 5 m, compared to 6 m in drained/forested plots. Plots only rewetted or cleared had similar species richness (13–14 species/0.25 m2) and vegetation height (5–6 m) to the plots both rewetted and cleared. For comparison, a nearby natural (undrained and unforested) fen contained 9 plant species/0.25 m2 and had a canopy height of 1 m. These were significantly greater in the clear-cut/rewetted plots. Around winter 2002/2003, four restoration treatments were applied in each drained and tree-colonized fen, in adjacent 50 x 150 m plots: cutting and removal of all trees, rewetting (by ditch blocking), tree removal and rewetting, or none. In 2010, plant species and canopy height were recorded at 20 points/plot, in 0.25 m2 quadrats. The natural fen was sampled in 1978. This study was based on the same experimental set-up as (3) and (6).

9 

A replicated, paired, site comparison study in 2003–2007 in nine bogs and fens in Finland (Noreika et al. 2015) reported that areas restored by tree thinning and rewetting had moss cover and tree structure intermediate between degraded (forested and drained) and natural (sparse trees, never drained) areas. After 1–3 years, restored areas had greater Sphagnum moss cover but less cover of other mosses than degraded areas, but less Sphagnum moss cover and greater cover of other mosses than natural areas. Restored areas had fewer tall trees (>3 m) than degraded areas, but more tall trees than natural areas. All data were reported as graphical analyses and differences were not tested for statistical significance. Between 2003 and 2006, in each of nine degraded peatlands, one area was managed by removing excess trees (above the natural tree density) and blocking drainage ditches. In each peatland, one degraded and one pristine area were also monitored. In 2007, vegetation cover was visually estimated in twenty-four 1 m2 quadrats/area (72 quadrats/peatland). Trees were counted and measured in six 100 m2 plots/area (18 plots/peatland). This study was based on the same experimental set-up as (11).

10 

A replicated site comparison study in 2007 in 19 forested fens in Finland (Daza Secco et al. 2016) found that restoration by tree thinning and rewetting had no effect on plant taxon richness, diversity or community composition. After 3–12 years, there were no significant differences between treatments for plant taxon richness (restored: 11–12; degraded: 11; natural: 12 taxa/m2) or diversity (data reported as a diversity index). Overall plant community composition did not differ between restored and degraded sites, but was significantly different from natural sites in both (data reported as a graphical analysis). Of the 19 studied forested fens, 10 had been restored 3–12 years before sampling (trees thinned and drainage ditches filled; water table 16 cm below peat surface). Four fens were degraded (with excess tree growth and drained; water table 32 cm below surface). Five fens were natural (sparsely forested and undrained; water table 19 cm below surface). In summer 2007, cover of plant taxa was estimated in three 1 m2 plots at each site.

11 

A replicated, paired, before-and-after, site comparison study in 2003–2007 in nine bogs and fens in Finland (Noreika et al. 2016) reported that a combination of tree thinning and rewetting reduced the number of tall trees for 1–3 years. Areas that were rewetted and cleared of trees contained fewer tall (>3 m) trees 1–3 years after restoration than before. Thus, the number of tall trees in restored areas became more like natural areas and less like degraded areas. Data were reported as graphical analyses and differences were not tested for statistical significance. Between 2003 and 2006, in each of nine peatlands, one area previously drained for forestry was restored by removing excess trees (above the natural tree density) and blocking drainage ditches. This was compared to one area that remained degraded (drained and fully forested) and one pristine area (never drained, sparsely forested). In 2003 (before intervention) and 2007, trees were counted and measured in six 100 m2 plots/area (18 plots/peatland). This study was based on the same experimental set-up as (9).

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.