Canada: Restoration of a Fen Plant Community after Peat Mining in Southern Québec


Even several decades after the cessation of mining activities, natural revegetation of mined peat fields in North America is often constrained by a lack of suitable propagules, as the residual peat is devoid of plants and a viable seed bank (Salonen 1987), and natural areas surrounding the mined fields are typically bogs with few or no fen species present (Poulin et al. 1999; Campbell et al. 2003). This study was undertaken in the RiviÁ¨re-du-Loup peatland in southern Québec, Canada to test different restoration measures for their efficacy in contributing to the establishment of a fen plant community on minerotrophic surfaces after peat mining. Donor diaspore material, containing seeds, rhizomes, moss fragments and plant propagules, was collected from two nearby natural fens and tested as a possible means of facilitating revegetation. Also tested were the application of straw mulch as a means of increasing fen species cover and biodiversity compared to control plots, and terracing as a means of altering residual peat depths and thus influencing the process of revegetation. The results after two growing seasons showed the application of donor diaspore material to be an effective technique for establishing vascular fen plants, but they also suggested that rewetting measures are necessary in order to move beyond simply restoring fen species and begin creating a fen ecosystem.

Quick Facts

Project Location:
RiviÁ¨re-du-Loup, 47.835957, -69.53598540000002

Geographic Region:
North America

Country or Territory:


Freshwater Ponds & Lakes

Area being restored:
0.5 acres

Organization Type:
University / Academic Institution


Project Stage:

Start Date:

End Date:

Primary Causes of Degradation

Mining & Resource Extraction

Degradation Description

Restoration work was carried out on two mined peat fields abandoned in September 2000. Following peat mining, the environmental conditions of an abandoned field are extremely harsh for plant re-establishment (Salonen 1987, 1992; Campbell et al. 2003). The physical properties of peat deteriorate due to the effects of long-term drainage and compression from peat mining operations (Okruszko 1995; Price et al. 2003). In addition, the microclimatic conditions are harsh due to an absence of vegetation cover, and the surface peat may form impenetrable crusts prone to frost heaving (Salonen 1987; Groeneveld & Rochefort 2002).

Project Goals

The aim of this project was to restore a moderate-rich fen plant community on a minerotrophic peat surface degraded by mining activities.


The project does not have a monitoring plan.

Description of Project Activities:
The restoration site included two adjacent fields (30 m Á— 60 m) separated by a central drainage ditch. Residual peat at the centre of the fields averaged 65 cm, and decreased towards the drainage ditches, where the residual peat averaged 20 cm. No vegetation was present on the fields, which were abandoned in September 2000. The peat was composed of matted sedges interspersed with coniferous wood. Preliminary chemical analyses indicated that the peat was characteristic of a fen. The underlying mineral soil was primarily clay with deposits of sand, gravel, and occasional boulders. Field reconnaissance to locate donor sites revealed that there were few natural fens near the restoration site. However, two natural fens were found in the foothills of the Appalachian Mountains, ca. 25 km southwest of the restoration site. These fens were chosen as donor sites based on their proximity to the restoration site, accessibility, and contrasting plant communities and environmental site conditions. The first donor site is a basin fen (National Wetlands Working Group, Anon. 1997) dominated by Sphagnum species (hereafter referred to as Sphagnum fen). It is a small fen (0.5 ha, 47°77' N, 52°83' W, ca. 274 m a.s.l.) receiving minerotrophic water from a small stream to the north and surface runoff from a slope on its western side. The donor area (25 m Á— 25 m) was positioned in the centre of the peatland where the peat depth averaged 86 cm. The site is characteristic of a poor fen based on the vegetation composition and water and peat chemistry (Zoltai & Vitt 1995). The main species (in order of dominance at the site, cover > 2%) are Sphagnum centrale, Sphagnum flexuosum, Utricularia minor, Polytrichum strictum, Calamagrostis canadensis, Salix pyrifolia, Picea mariana, Glyceria canadensis, Sphagnum capillifolium, Carex canescens, and Sphagnum magellanicum. The second donor site is a riparian stream fen (Anon. 1997), dominated by Calamagrostis canadensis (hereafter referred to as Calamagrostis fen). It is a small fen (1.5 ha, 48°19' N, 52°81' W, ca. 320 m a.s.l.) receiving minerotrophic water from a stream entering the peatland on the north side, coursing through the main body of the fen and emptying into a small pond on the southern end. A beaver dam was found upstream of the fen in June 2001, which caused the water table to rise from below the surface to create flooded conditions for the remainder of the study period. The donor area (25 m Á— 25 m) was positioned in the centre of the fen where the peat depth averaged 85 cm. The fen is a moderate to rich fen (Zoltai & Vitt 1995). Other species that dominate (cover > 2%) the site are Warnstorfia exannulata, Carex utriculata, Scirpus cyperinus, Utricularia minor and Caltha palustris. The experiment was a split-plot factorial design. In total, 54 plots (3 terrace levels Á— 3 blocks (replicates) Á— 3 vegetation treatments Á— 2 straw mulch treatments) were established. Terrace levels were treated as main plots and were divided into three blocks to minimize the effect of variance within the site. The vegetation and straw treatments were treated as subplots and were randomly assigned within the blocks. The installation of the experiment commenced in April 2001, just after snow melt. The convex shape of the abandoned fields was modified to create three terraces of decreasing elevation, with different peat depths, on either side of and parallel to the main drainage ditch. The terraces were leveled with a machine grader that scraped excess peat off the site. The terrace levels are referred to as high, middle, and low, with an average peat depth of 56, 40 and 15 cm, respectively. The terrace levels could not be randomly positioned due to topographic constraints of the site. The central drainage canal was blocked, while a secondary ditch upslope of the site was unblocked. Peat mining operations continued on fields' upslope of the restoration site throughout the study period, and blocking of these drainage ditches was not permitted. Berms were created on the down slope side of the terrace levels to hold water on the site, and prevent erosion. Each berm was ca. 0.5m wide and 0.3 m in height. Prior to the application of the vegetation and straw treatments, plots were raked to break up the surface crust, minimize inconsistencies of compaction, and reduce microtopography resulting from the machinery. Phosphorus fertilizer (2 g.m - 2) was subsequently applied, as recommended for bog restoration to favour vascular plant establishment (Rochefort et al. 2003). Experimental plots (5 m Á— 5 m) were established on the terraces and were separated by a 1-m buffer. The vegetation treatments were (1) donor diaspore material from the Sphagnum fen, (2) donor diaspore material from the Calamagrostis fen, and (3) a control, without donor diaspore material applied. The donor diaspore material was collected from 18 (1.25 m Á— 1.25 m) random quadrats located within the donor area (25 m Á— 2 m). The ratio of donor diaspore collection area to restored area (1:16) was similar to that suggested for bog restoration (Campeau & Rochefort 1996). The top 10 cm of substrate and vegetation from each donor quadrat was collected by hand and transported to the restoration site, where it was broken into small pieces and spread by hand. The mulch treatments were (1) straw, and (2) a control without straw. The straw was applied with a density of 1500 kg.ha-1 and was spread to exceed the plot boundary to minimize edge effects. Vegetation and mulch treatments were applied to the restoration site during the week of 7-11 May 2001. Percent cover of the vegetation at the restoration site was surveyed from 10-14 October 2001 and from 8-13 August 2002. Ten quadrats (30 cm Á— 30 cm) in each experimental plot were sampled. The percent cover (visually estimated) for each plant species within each quadrat was recorded. Sampling omitted the border area (0.5 m on each side) of the plots to minimize the observation of edge effects.

Ecological Outcomes Achieved

Eliminate existing threats to the ecosystem:
VEGETATION Two growing seasons after the re-introductions the majority of established plants were fen species, in terms of percent cover and richness. The total species cover at the restoration site doubled during the second year from 12 to 35%. The abundance of fen plants within the experimental plots increased from 5% in the first year to 20 % in the second year. The plant community was composed primarily of forbs and graminoids. There was a small component of woody plants, while bryophytes were largely absent. Trace amounts of bryophyte species were observed in several plots in the first year that were no longer present in the second year. Control plots without straw or diaspore material were primarily devoid of vegetation, while many treatment plots had abundant growth. Several species established at the restoration site from the donor fen diaspore material, including Glyceria canadensis, Carex canescens, Galium trifidum, Calamagrostis canadensis, Viola macloskeyi, Ranunculus pensylvanicus, Fragaria virginiana, Epilobium ciliatum, and Juncus brevicaudatus. Several other fen species were either introduced via the donor diaspore material or present in the local seed rain, such as Agrostis hyemalis, Scirpus cyperinus, Juncus effusus, and Lycopus uniflorus. Equisetum arvense and Tussilago farfara established an extensive cover on the low terraces. These species are commonly observed growing along the ditches of the cut-over peat fields; their rhizomes were likely left in the ground during preparation of the terraces. Straw mulch introduced a few agricultural species including Secale cereale, Rorippa palustris, and Avena sativa. By the end of the second year only Rorippa palustris was still present. DONOR DIASPORE TREATMENTS Donor diaspore treatments (from Sphagnum fen and Calamagrostis fen) increased the abundance of fen species after the first and second growing seasons. During the first year the combination of Sphagnum donor diaspore and straw mulch treatments favoured the abundance of fen species cover and produced the highest total fen species cover of all experimental treatments (9 +/- 1%). Several herbaceous species proliferated with the combined treatments of Sphagnum fen diaspore material and straw mulch including: Viola macloskeyi, Lycopus uniflorus, and Galium trifidum. After the second growing season plots treated with Calamagrostis fen diaspore material (31 +/- 5%) tended to have higher fen cover than Sphagnum fen diaspore material (22 +/- 3%), although these differences were not significant. Both diaspore treatments were significantly higher than control plots (8 +/- 1%). After the first growing season, the fen plant richness was significantly highest where Sphagnum fen diaspore material (18 +/- 1 taxa) had been applied, intermediate with the application of Calamagrostis fen diaspore material (13 +/- 1 taxa) and lowest without donor diaspore material (6 +/- 1 taxa). The richness of fen plants decreased from the first to second year, and there was no longer a significant difference between the types of donor diaspore material applied (Sphagnum fen = 13 +/- 1; Calamagrostis fen = 12 +/- 1 taxa). Nevertheless, the application of donor diaspore material increased the fen plant richness compared to the control (7 +/- 1 taxa). STRAW MULCH The application of straw mulch did not improve the abundance of fen plants after two growing seasons. Only during the first year did straw mulch improve the cover of fen plants from 5 to 9 % in combination with Sphagnum fen diaspore material. These initial increases in fen plant cover did not extend to the second year. More notably, straw mulch clearly increased the richness of fen species from both diaspore materials after two years. Fen plant richness was higher for plots treated with straw mulch (12 +/- 1 taxa) compared to plots without straw mulch (9 +/- 1 taxa). Rorippa palustris was the only species introduced with the straw mulch that persisted over two growing seasons at the restoration site. TERRACE LEVEL After two growing seasons there was significantly more fen species cover on the middle terrace level (27 +/- 5%) than the high terrace level (14 +/- 2%). The mean fen species cover on the low terrace level (20 +/- 4%) was between the values observed at the middle and high terrace levels, and was not significantly different from either. Equisetum arvense and Tussilago farfara, the second and third most dominant species after two growing seasons, were dominant on the low terrace level (26 +/- 8 %), whereas they formed only a minor component of the plant cover on the middle (1 +/- 1%), and high terrace (2 +/- 1%) levels. No other experimental treatments had an effect on their establishment and their dominance on the low terrace is presumed to be due to their establishment along the ditches prior to the experiment, rather than a treatment effect. HYDROLOGY From May to August 2001 and 2002 the total rainfall was 286 and 253 mm, respectively, compared to the mean 30-yr seasonal total of 353 mm (Environment Canada 1993). The lower terraces were more proximal to the water table and to the underlying clay substrate. The mean depth to water table was - 29, - 34 and - 45 cm for low to high terraces, respectively. This resulted in a water table that sloped toward the central drainage ditch, with a gradient of approximately 0.032 and 0.048 during the wettest (June 4, 2001) and driest (August 16, 2001) period, respectively. Except for brief periods immediately following rain events, the water table in the lowest terrace was always within the clay substrate. In the middle and upper terrace the water table was generally within the peat layer except for during the driest periods. The water table depth at the restoration site was far lower than the donor sites throughout the 2001 growing season. The water table at the donor sites was around - 4 cm at the Sphagnum fen and + 8 cm at the Calamagrostis fen. Mean soil-water pressure decreased below - 100 mb from mid-July to mid-August. Soil-water pressure was less than - 100 mb at the low, middle, and high terraces 16%, 24%, and 24% of the time, respectively. Soil-water pressure in the upper and middle terrace was similar (averaging - 66 and - 62 mb, respectively), in spite of a notable difference in water table. WATER CHEMISTRY In general, the concentrations of minerals and nutrients at the restoration site were higher than at the natural fen donor sites. The mean pH levels of the terraces at the restoration site did not vary greatly (ca. 5.9), and were similar to the Calamagrostis fen donor site (5.8), while the Sphagnum fen donor site had a lower mean pH (5.5). The electrical conductivity of the restoration site was higher than that of the natural fens and the low terrace values were higher than the other two terraces. The major cations followed a pattern similar to the electrical conductivity. The nitrogen and phosphorus levels at the restoration site were moderate and are considered mesotrophic (Bridgham et al. 1996). The mean concentration of available P at the restoration site was higher than the natural fens, but within the observed range of variability. The concentrations of ammonium and nitrate on the lowest terrace level of the restoration site were similar to the natural fens, whereas the middle and high terrace levels tended to have higher concentrations. PEAT CHEMISTRY Total nutrients (N, P, and K) of the restoration site tended to be lower or equal to the donor fens. The mean total nitrogen concentration at the restoration site (17.96 mg.g-1) was between the mean concentration of the Sphagnum fen (10.11 mg.g-1) and Calamagrostis fen (19.92 mg.g-1). Total phosphorus concentrations of the peat were lower at the restoration site (0.34 mg.g-1) than the Sphagnum (0.58 mg.g-1) and Calamagrostis (1.18 mg.g-1) fens. Similarly, potassium concentrations were generally lower at the restoration site than the natural fens. In contrast, the concentrations of the other mineral elements present in the peat including calcium, magnesium, sodium, and iron, were higher at the restoration site than at the reference sites.

Factors limiting recovery of the ecosystem:
Water availability is an important factor affecting fen restoration success (Roth et al. 1999), fen species distribution (e.g. Bridgham & Richardson 1993), and number of plant niches (Silvertown et al. 1999). Reintroduction of fen species by plantings was the most effective with water table levels slightly below the surface or with shallow standing water (Cooper & MacDonald 2002). The extremely dry conditions of the restoration site are likely limiting the establishment of a fen plant community. The water tables at the donor sites were consistently just below or above the surface, whereas the surface peat of the restoration site was dry throughout the growing seasons, except immediately following rain events. Further rewetting measures are considered necessary to create fen-like hydrological setting at the restoration site, and ultimately to create a fen system rather than simply restoring fen species (Grootjans & van Diggelen 1995). However, the freedom to manipulate the sites hydrology was constrained by drainage requirements of adjacent peat extraction activities. Water quality is considered as important as water availability for fen restoration (Charman 2002; Lamers et al. 2002). In this experiment, the higher concentration of solutes in deeper peat results from the diffusion of salts from the marine clay (Van Seters & Price, unpubl. data for Cacouna peatland, Québec, 1999). Some of these solutes may migrate more quickly above the water table by capillary rise, and become more concentrated by evaporation. During the summer the net direction of solute transport is upwards, similar to the net water flow direction (i.e. driven by evaporative water losses). The concentrations of solutes within the peat substrate were considerably higher than the donor sites and may be one factor that lowered the success of fen plant recolonization.

Socio-Economic & Community Outcomes Achieved

Key Lessons Learned

The application of donor diaspore material from natural fens clearly increased the cover and richness of fen species compared to control plots. Diaspore material led to the establishment of several species at the restoration site that were absent from neighbouring abandoned fields and the nearby natural bog. This provides evidence that the establishment of fen vegetation is hampered by dispersal, rather than conditions at the restoration site. After two years, several graminoid and herbaceous species from the Calamagrostis fen formed a dominant vegetation cover at the restoration site, despite unfavourable water table conditions.

The advantages of applying diaspore material as a plant introduction technique are numerous. Firstly, the variety of propagule species and types contained within the donor diaspore material increases the chances that some of the species biological requirements will match the environmental conditions of a site and the particular climatic conditions of a given year. Secondly, the inclusion of the substrate with the diaspore material means that soil mycorrhizal fungi associated with the plant community are also brought to the site. Mycorrhizal fungi may be of great importance in wetland plant communities (Turner & Friese 1998; Thormann et al. 1999). Thirdly, insect larvae and other disseminules may also be brought to the site within the substrate, further aiding plant community establishment by acting as dispersal agents (Middleton 1999). Finally, if the donor diaspore material is collected in the spring, it allows propagules to fulfil their natural dormancy cycle under their native conditions. This may be of great importance for establishing Carex species, an important component of fen plant communities, which have been shown to have complex dormancy cycles and species-specific germination traits (Baskin et al. 1996; Schultz 1998; Patzelt et al. 2001).

A disadvantage of applying donor diaspore material as a reintroduction method is the disturbance caused to the donor wetland community during diaspore collection. However, informal observations of the disturbed quadrats (> 4% of the reference unit) of the current study revealed that there was 25 to 40 % recovery by the end of the second growing season. Monitoring the donor sites is warranted if this method is adopted for large-scale restoration projects in the future.

After two years, straw mulch significantly increased the diversity of fen plants. Only during the first year was there a synergistic effect on fen species abundance with straw mulch and Sphagnum fen diaspore material treatments. The straw mulch may facilitate the germination and early stages of establishment of some species due to improved water conditions (Price et al. 1998), whereas mulch may inhibit other species and later stages of plant development by shading, obstructing plants, attracting predators or promoting pathogens (Xiong & Nilsson 1997). In previous studies, cover treatments improved the germination of vascular plants, including Eriophorum angustifolium, Molinia and Calluna; however, other species such as E. vaginatum failed to respond to the same treatments (Sliva & Pfadenhauer 1999). Straw mulch did not improve the establishment of vascular plants in bog restoration attempts, whereas the establishment of mosses, particularly Sphagnum species, was substantially improved (Rochefort et al. 2003). For fen restoration sites where better rewetting can be achieved, mulching could prove to have a greater effect, particularly for helping mosses to establish.

It was hypothesized at the project’s outset that the terrace level with environmental conditions closest to the natural fen donor sites would support the highest fen plant establishment. The intermediate terrace level had the highest fen species cover after two years. The environmental conditions of the middle terrace level may represent a compromise between the extremely dry conditions of the high terrace level, and the more minerotrophic conditions of the lowest terrace level.

Long-Term Management

Several non-target species became established at the restoration site, most notably, Equisetum arvense and Tussilago farfara, but these species appear to have spread from the adjacent ditch to the lower terrace. The establishment of invasive, weedy species has been observed elsewhere in Europe (Rowlands 2001) and North America (Cooper & MacDonald 2002) on cut-over sites with minerotrophic peat. In Finland, the abundance of E. arvense and T. farfara was lower on older minerotrophic peat fields, suggesting a decrease in dominance over time (Salonen 1990), so these species may not be a concern in the long-term. Further monitoring of the vegetation community is required to determine whether there is convergence with the donor fen plant communities over time, or increased dominance of non-target species.

Sources and Amounts of Funding

Financial support for this project was provided by the Natural Sciences and Engineering Research Council of Canada, the Canadian Peat Moss Association, and the Berger Peat Moss Company.

Other Resources

Cobbaert, D., L. Rochefort & J.S. Price. 2004. Experimental restoration of a fen plant community after peat mining. Applied Vegetation Science 7: 209-220.

Primary Contact

Organizational Contact