SEARCH

SEARCH BY CITATION

Keywords:

  • biodiversity;
  • Calamagrostis epigejos;
  • ecological restoration;
  • erosion control;
  • grasslands;
  • hay transfer;
  • mulch seeding;
  • species introduction;
  • succession

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

1. Highly disturbed areas such as surface-mined land provide a great challenge for ecological restoration. The goal is to identify appropriate restoration approaches in a continuum between technical reclamation and spontaneous succession. In particular, on slopes endangered by erosion, appropriate methods are needed that quickly establish vegetation cover but also take into account the natural potentials of the site.

2. In the mined area Roßbach (Saxony-Anhalt, Germany), we evaluated the effects of spontaneous succession and assisted site recovery (species introduction through hay transfer and sowing) during a 9-year experiment. We asked how rates and pathways of vegetation development differ between treatments and whether species composition converges over time owing to species exchange.

3. The application of green hay as well as the sowing of regional seed mixtures clearly accelerated vegetation development and led to the rapid establishment of species-rich grasslands. Hay transfer was most successful owing to the high amount of transferable target species. Moreover, both treatments facilitated the establishment of cryptogams and provided effective erosion control. Also, hay transfer and sowing clearly affected the pathway of succession. Calamagrostis epigejos migrated from nearby source populations and became increasingly dominant at sites with spontaneous succession. In contrast, the species-rich grasslands established after hay transfer and sowing were highly resistant to invasion of Calamagrostis and other ruderals.

4. Species exchange between treatments led to increasing similarity in vegetation composition over time. Nine years after implementation of the experiment, we did not find any significant differences between treatments in terms of total vegetation cover, species richness and the number of target species. However, the dominance ratio between target and nontarget species differed significantly. Species introduction through hay transfer and sowing led to a permanently higher abundance of grassland species and a lower coverage of ruderals compared with spontaneously developed sites. Hence, our results highlight the importance of initial floristic composition and the order of species arrivals for long-term vegetation development.

5.Synthesis and applications. Hay transfer and sowing of regional seed mixtures are appropriate restoration tools to achieve rapid revegetation when no potential seed sources of target species are available nearby or there are undesirable species that need to be suppressed. Our results show that introduced grassland species are able to grow under postmining site conditions and can migrate into adjacent spontaneously developing sites. A combination of spontaneous and assisted site recovery can promote the development of species-rich grasslands in postmining landscapes.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

Ecological restoration of highly disturbed areas often aims to accelerate and/or manipulate vegetation development, thus is intrinsically linked to succession (e.g. Walker, Walker & del Moral 2007; Prach & Hobbs 2008; Walker & del Moral 2009). During the last few decades, spontaneous and assisted site recovery has been studied in postmining landscapes in Eastern Germany (e.g. Felinks, Pilarski & Wiegleb 1998; Kirmer & Mahn 2001; Wiegleb & Felinks 2001a,b; Kirmer et al. 2008; Baasch, Tischew & Bruelheide 2009). Several studies have shown that nutrient-deficient sites in mined areas can enhance biodiversity and provide habitats for rare species and plant communities. Therefore, one method of restoring surface-mined land could be through spontaneous succession (e.g. Prach & Pyšek 2001; Prach 2003; Tischew & Kirmer 2007). However, assisted site recovery is more appropriate on sites endangered by erosion, where managers seek to reintroduce species with low dispersal abilities or when ‘problematic’ species are nearby. Appropriate methods are those that quickly establish vegetation cover to provide basic ecosystem services in a short time (e.g. erosion control) (Norris et al. 2008). For this purpose, technical reclamation is predominantly used, i.e. ameliorating nutrient-poor substrates and sowing low-diversity mixtures of grass cultivars (Tischew & Kirmer 2007). There is much evidence that these measures often destroy the natural potential of the sites, thereby counteracting the development of biological diversity on marginal land (e.g. Mudrák, Frouz & Velichová 2010; Tropek et al. 2010). Therefore, conservationists and ecologists call for restoration strategies that include the natural potential of the sites and result in the development of species-rich vegetation of high nature conservation value (e.g. Tischew & Kirmer 2007; Řehounková & Prach 2008; Tropek et al. 2010). This might be achieved through the creation of species-rich grasslands using alternative restoration methods, e.g. hay transfer to accelerate and direct vegetation development. During the last few decades, land use changes in Europe have led to a continuous decline in the area of semi-natural grassland rich in biodiversity (e.g. EU 2010 Biodiversity Baseline Report; WallisDeVries, Poschlod & Willems 2002; Cremene et al. 2005; Pärtel et al. 2005; Krauss et al. 2010). In the face of increasing pressures on productive sites, the restoration of species-rich grasslands should be given higher priority in restoration schemes for marginal land such as nutrient-deficient mined sites.

In the mined area Roßbach, we investigated the success of three approaches to restoration: (i) application of fresh, diaspore-rich green hay, (ii) sowing of autochthonous species with an additional mulch layer and (iii) spontaneous succession to initiate vegetation development on unvegetated slopes. The experiment was established in the vicinity of small patches of dry grassland and stands dominated by Calamagrostis epigejos. This strong, competitive, clonal grass is widely seen as a ‘problem’ species in the conservation of grasslands (e.g. Rebele & Lehmann 2001; Fiala et al. 2004; Somodi, Virágh & Podani 2008) and for vegetation restoration in postmining landscapes in Central Europe (e.g. Prach & Pyšek 2001; Wiegleb & Felinks 2001a; Tischew & Kirmer 2007).

We hypothesized that the input of appropriate seeds via green hay and sowing would accelerate vegetation development leading to species-rich grassland communities within a few years, while revegetation through spontaneous succession would take longer and might lead to grasslands dominated by stands of Calamagrostis (Hypothesis 1). As in many restoration studies, it was not practical to weed the experimental site, thus treatments in the same locality should influence one another in the long term owing to species exchange (e.g. Bullock, Pywell & Walker 2007; Jongepierová, Mitchley & Tzanopoulos 2007; Lepšet al. 2007; Rydgren et al. 2010). Even so, we assumed that different rates and pathways of vegetation development on sites with and without species introduction will have a long-term influence on the composition of communities (Hypothesis 2), as species introduction will influence seed availability and dispersal stochasticity (e.g. Zobel et al. 2000), initial floristic composition (Egler 1954) and the sequence of species arrivals (e.g. Drake 1991). Long-term studies comparing different ecological restoration methods are rare for mining areas (see Kiehl et al. 2010). Our experiment evaluates the long-term effects of different methods to produce management recommendations for ecological restoration.

Materials and methods

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

Study Site and Sampling Design

The mined site Roßbach is part of the Geiseltal lignite mining district in Saxony-Anhalt (Germany). Our study area was situated in the southern part of the mined site and covered an area of 2·3 km2 (Kirmer et al. 2008). We set up an experiment in the centre of the study area on an artificial unvegetated slope with an inclination of c. 8°, an eastern aspect, and an altitude of c. 123 m a.s.l. (11°54′5·46′′E, 51°14′27·98′′N). In August 2000, the slope was shaped within the scope of remediation measures. The substrate consisted of dumped loess with a pH value (CaCl2) of 7·5. The material came from deeper soil layers, thus representing the features of primary succession (no soil development, no soil seed bank). In early September 2000, the experiment was established in a complete block design with three blocks and three treatments. (i) Sowing – S: 15 herbs and six grasses were sown at a density of 2 g/m2 or 860 seeds/m2 (site-specific seed mixture of autochthonous species, see Appendix S1, Supporting Information). Species were selected based on their ability to grow on dry, sunny sites (dry and mesic grassland species), satisfy aesthetic demands, ensure a fast vegetation development and prevent erosion in the first years. After sowing, a mulch layer with seed-poor green hay was applied by hand to a thickness of c. 5 cm. (ii) Hay transfer – H: The donor site for green hay was the Natura 2000 site ‘Göttersitz und Schenkenholz nördlich Bad Kösen’c. 20 km from the mined area. The vegetation consists of a mosaic of semi-natural dry grassland (Brometalia erecti Br.Bl. 1936) and mesic grassland (Dauco carotaeArrhenatheretum elatioris (Br.Bl. 1919) Görs 1966) including 97 higher plant species. There were 71 target species (Festuco-Brometea and Arrhenatheretea spp.) (see Appendix S2 Supporting Information). On 6 September 2000, seed-rich green hay was mown with cutter bar mowers, immediately transported to the mined site Roßbach, and evenly distributed by hand at a density of c. 1 kg per m2 (5 cm thickness). (iii) Control – C: Spontaneous succession.

The area of each treatment within a block was c. 1260 m2 (18 m × 70 m, see Fig. 1) amounting to 1·13 h for the total experimental site. Three 25-m2 permanent subplots were established per treatment and block. Vegetation was surveyed annually for each subplot at the end of June from 2001 until 2009, except in 2003. The percentage cover of all species and layers was recorded. Data from the subplots within treatments and blocks were pooled for statistical analyses. Management took place in August 2002 and 2007 (mowing), and July 2005 (manual removal of woody plants).

image

Figure 1.  Experimental site and schematic outline of the design.

Download figure to PowerPoint

Statistical Analysis

Mean values for plant species richness (number of species per 75 m2), percentage cover of vascular plants and cryptogam species were calculated for each treatment and each year (n = 3, data collected on subplots within a block were pooled prior to analyses). For the last observation year, we tested for treatment effects on species richness and vegetation coverage using generalized linear models (GLM) followed by planned comparisons of least square means between treatments with and without species introduction (simple contrast analysis; S, H vs. C).

We used nonmetric multidimensional scaling (NMDS) to analyse the differences in vegetation composition between the three different treatments (S, H, and C). NMDS ordination was based on Bray–Curtis distances ranging from 0 (completely identical) to 1 (completely dissimilar). We estimated Bray–Curtis distances between treatments for each year. ‘Stress’ indicates the goodness-of-fit with values <0·2 corresponding to an acceptable indication of the similarities between samples (Clarke 1993). NMDS was calculated using the statistical software package PAST (PAlaeontological STatistics, ver. 2.0, Hammer, Harper & Ryan 2001). We applied repeated-measures anova to test whether species composition became increasingly similar with time. Mauchly’s test indicated that the assumption of sphericity had been violated for the main effect of ‘year’ (χ2 = 80·49, d.f. = 27, < 0·001). Therefore, the degrees of freedom were corrected using Greenhouse-Geisser estimates of sphericity (ε = 0·55).

To analyse community composition, we used species lists from the whole study area (2·3 km2) as well as from the donor site and the sown seed mixture and divided all plant species recorded in plots at the experimental site (1·13 ha) into the following groups: (i) target species that were introduced by sowing and/or hay transfer and which were not present in the study area; (ii) target species that were introduced by sowing and/or hay transfer but that may also have colonized the experimental site through seed rain from adjacent populations within the study area; (iii) target species that were not introduced having colonized the experimental site through seed rain from adjacent populations within the study area; and (iv) nontarget species. The target species included all grassland species typically occurring in and around Arrhenatherion as well as Festuco-Brometea-communities according to the regional phytosociological literature (Schubert 2001). Calamagrostis epigejos and all other species (mostly ruderals) were considered as nontarget species. Univariate GLMs were used to test for treatment effects on the number and coverage of target and nontarget species in the last observation year.

We applied the Markov approach to analyse and compare rates and pathways of vegetation development (e.g. Usher 1992). For each treatment, we constructed a transition matrix P having elements pij that represent transition probabilities from state i to j over time. We assessed the probability of colonization by target species and nontarget species (C. epigejos and ruderals) as well as the probability that one group is replaced by another. The assessment was based on the area covered by each group as well as by bare soil (standardized to sum to 100). We converted our multivariate ecological time series into transition probability matrices according to the approach presented in Baasch, Tischew & Bruelheide (2010). For each survey plot per treatment and each transition period, we first created an interim contingency table (s × s) based on the observed values and the following predefined model assumptions.

The entries in the diagonal vii define for a certain state (Si), the area that persisted from time t to time + 1 estimated by using the following equation:

  • image(eqn 1)

The i, j entries vij define the area initially covered by one state (Si), which was then replaced by another state (Sj) estimated by using the equation:

  • image(eqn 2)

As a result, we obtained 27 tables (three blocks × nine time periods) per treatment. By adding the corresponding elements across all 27 tables, we obtained for each treatment a final contingency table referring to the changes observed during the whole study period. Finally, we calculated the state transition probability matrix P for all three treatments by estimating the transition probabilities pij:

  • image(eqn 3)

The entries of P describe the processes that determine succession: colonization, disturbance, persistence and replacement. These can be quantified for a certain state or the entire system (for details, see Baasch, Tischew & Bruelheide 2010).

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

Overview – Development of Vegetation

In treatments with species introduction (S, H), plant species already covered more than 40% of the surface in the second year of observation (Fig. 2). From 2004 onwards, vegetation cover was continuously high ranging between 70 and 90%. In contrast, there was much less cover on control plots particularly in the first 5 years. Moreover, we observed a higher cover of cryptogams for treatments S and H compared to control plots in all years. In the final year of observation, we could still detect significant treatment effects on the cover of cryptogams but not on the total cover of vascular plant species (Table 1).

image

Figure 2.  Coverage of higher plant species and cryptogams and plant species richness on all three experimental treatments between 2001 and 2009. The field data collected on three 25-m2 subplots within treatments and blocks were pooled prior to the analyses. Values are means ± 1 SE, n = 3.

Download figure to PowerPoint

Table 1.   Results of generalized linear models (GLM) with treatment as explanatory variable (d.f. = 2,6). GLM was followed by contrast analyses (C, control, spontaneous succession; H, hay transfer; S, sowing)
YearDependent variableUnivariate GLMContrasts (simple, reference category: control)
FPP(S vs. C)P(H vs. C)
2009Coverage of cryptogams11·270·0090·007(S > C)0·006(H > C)
Coverage of plant species5·010·053  
Plant species richness1·620·274  
Number of nontarget species43·80<0·001<0·001(S < C)<0·001(H < C)
Coverage of nontarget species6·620·0300·022(S < C)0·018(H < C)
Number of target species3·370·104  
Coverage of target species9·360·0140·029(S > C)0·005(H > C)

In the first year of observation, treatments with species introduction exhibited higher species richness than control plots. However, the number of plant species decreased considerably in the following 3 years (Fig. 2). This was caused by a decrease in ruderal and annual species, but it was also owing to a rapid expansion of introduced legumes to the detriment of other target species. However, in the final observation year, we could not detect a significant treatment effect on species richness, indicating that all three treatments achieved the same level of diversity after 9 years (Table 1).

Erosion processes were controlled by the mulch layer and the fast vegetation development. Thus, we did not find erosion channels deeper than 5 cm in treatments with mulch seeding and hay transfer. Erosion was severe on control plots in the first two years where we found many erosion channels up to 1·5 m depth (particularly on the lower slope, see Fig. 3).

image

Figure 3.  Erosion on the experimental site in the mined area Roßbach (Saxony-Anhalt) two years after slope restoration by different methods (20 August 2002). To the left: Hay transfer treatment vs. control. Middle: Control. To the right: Control vs. sowing treatment with additional mulch layer.

Download figure to PowerPoint

Similarity

Except in the first 2 years, dissimilarity was slightly higher between the two different treatments with species introduction (S-H) than between these treatments and control plots (C-S, C-H; Fig. 4). Using repeated-measures GLMs, we found a significant main effect of ‘year’ on the distance between treatments (F = 95·72, < 0·001). Contrasts revealed that the distance between treatments in the last observation year was significantly lower than the distances measured in all the previous years (except 2008), indicating that vegetation composition converged on different treatments. The between-subjects effect (comparison of treatment: S-H, C-S, C-H) was not significant (d.f. = 2, 24, F = 0·87, = 0·430). Thus, the results suggest that the similarity in vegetation composition increased with time across all three treatments. Despite this converging trend, the overall similarity between treatments remained quite low until the final observation year (dissimilarity S-H: 0·64 ± 0·07 SD, C-S: 0·62 ± 0·06 SD, C-H: 0·62 ± 0·07 SD).

image

Figure 4.  Similarity in species composition between treatments over time (C: control, spontaneous succession, H: hay transfer, S: sowing). (a) Nonmetric multidimensional scaling of plots using Bray–Curtis dissimilarity (stress = 0·13, R2 Axis 1 = 0·45, R2 Axis 2 = 0·28). We used the mean species composition for each treatment in each year. Lines connect mean values of treatments in different years. In all cases, the first year and the last year of the study are highlighted by filled symbols (grey = first year, black = last year). (b) Mean values of Bray–Curtis dissimilarity. Treatment plots are compared with each other.

Download figure to PowerPoint

Species Groups

We found a high number of nontarget species in all treatments in the first 2 years (Fig. 5a). In particular, short-lived ruderal and segetal species (e.g. Bromus sterilis, Chenopodium album, Descurainia sophia, Lepidium campestre, Papaver rhoeas, Polygonum aviculare, Reseda luteola, Thlaspi arvense, Tripleurospermum maritimum) frequently reached the experimental site through seed rain. Additionally, some less-competitive ruderals were purposefully introduced by the sowing treatment to ensure faster revegetation and prevent erosion (e.g. Poa annua, P. compressa).

image

Figure 5.  Mean values of species number and cumulative coverage of species groups for all treatments over time. (a) Number of nontarget species (mostly ruderals), including Calamagrostis epigejos. (b) Coverage of nontarget species (mostly ruderals) and coverage of C. epigejos. (c) Number of target species (dry and mesic grasslands). (d) Coverage of target species. Target species are additionally grouped according to their origin.

Download figure to PowerPoint

The number of nontarget species decreased with time for all three treatments; this decrease was more pronounced in treatments with species introduction. However, the coverage of nontarget species increased with time (Fig. 5b). Control plots showed a considerably higher increase in nontarget species’ cover, mainly caused by the clonal spread of C. epigejos between 2004 and 2009. For the final observation year, we found significant treatment effects on the number and coverage of nontarget species, which were both significantly higher in control plots than in plots with sowing or hay transfer treatments (Table 1).

Typical grassland community species developed best in the hay transfer treatment. We found between 30 and 40 target species with a cumulative vegetation cover of more than 70% from 2004 onwards (Fig. 5c,d). Transfer rates (i.e. the number of species found on treated plots as a percentage of the number of species that occurred in the donor site ‘Göttersitz’) differed between years, ranging from 55 to 69% with 64% in the final observation year (see also Appendix S2 Supporting Information). In addition to the common grassland species that already occurred in the mined site Roßbach (e.g. Arrhenaterum elatius, Galium album, Lotus corniculatus, Trifolium campestre), from the second year onwards, we found between 15 and 20 grassland species that had originated from the transfer of plant material (e.g. Bromus erectus, Centaurea scabiosa, Ononis spinosa, Salvia pratense).

Establishment rates in the sowing treatment (i.e. the number of species found on treated plots as a percentage of the number of sown species) were continuously high ranging from 81 to 100% (2009: 95%; see also Appendix S1, Supporting Information). However, the number of target species was not as stable as in the hay transfer treatment. This was caused by a rapid expansion of introduced legumes to the detriment of other target species. Between 2004 and 2007, the predominance of legumes was unexpectedly high (e.g. Securigera varia), which led to a significantly lower species richness in 2005 compared to the hay transfer treatment and the control plots (see also Fig. 2).

Over time, the number of target species also increased on the control plots (Fig. 5c). Some of these species clearly migrated into the control plots from adjacent treatment strips (e.g. B. erectus, C. scabiosa, O. spinosa, S. pratense, S. varia). Thus, we did not find any significant difference in the number of target species between treatments with and without species introduction at the end of our study period (Table 1). However, despite the high number of target species on the control plots, the percentage of target species within the total plant cover remained low during the whole study period. In 2009, total coverage of grassland species was around 30%, which was significantly lower than in treatments with hay transfer or sowing (Table 1).

Figure 6 represents a schematic outline of the mean transient dynamics between species groups over the whole study period. The probability that bare soil was colonized by any species group was much higher in treatments with hay transfer (0·43) and sowing (0·38) than in control plots with spontaneous succession (0·15). These treatments were mainly colonized by target grassland species, which persisted to a high degree. The colonization rate by ruderal species was low, and the probability that established ruderals would be replaced by target species over time was relatively high. Moreover, the rates of colonization and replacement by C. epigejos were notably low; thus, both treatments with species introduction prevented a rapid and massive invasion by this competitive clonal grass. On control plots with spontaneous succession, bare soil was colonized by target species as well as ruderal species and C. epigejos. However, both target and ruderal species were to some degree replaced by C. epigejos, which also had the highest probability of persistence.

image

Figure 6.  Rates of colonization, persistence and replacement by target and nontarget species for all treatments during the whole study period. Diagrammatic representation of the Markov transition matrix P. Transition probabilities for persistence are shown in italic fonts inside of curved arrows. Rates of colonization and replacement are shown in the middle of straight-lined arrows. All the probabilities ≥ 0·01 have been plotted. Probabilities ≥ 0·05 are indicated by continuous lines, while lower probabilities are indicated by dashed lines. Note: Probabilities <0·01 are only shown for the colonization and replacement by Calamagrostis epigejos indicated by short, dashed arrows.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

Restoration through direct addition of appropriate seeds via green hay and sowing clearly accelerated successional progress at our experimental sites. Moreover, both treatments not only favoured the rapid establishment of plant species but also facilitated the establishment of cryptogams (see also Poschlod & Biewer 2005; Jeschke & Kiehl 2006) and provided effective erosion control (see also Kirmer & Mahn 2001). The hay mulch in both treatments may also have facilitated seedling establishment and growth on the nutrient-deficient slope (see also Kirmer, Baasch & Tischew 2011; Tongway & Ludwig 2011) owing to improvements in soil structure, moisture capture and retention, fertility and biological activity (e.g. Bradshaw 1983; Grant, Campbell & Charnock 2002; Bakker et al. 2003; Cummings et al. 2005).

The spread of target species that were not present in the mined site prior to the experiment indicates that these species generally are able to grow under postmining conditions but were seed-limited before hay transfer or direct sowing. Limited dispersal capability has often been shown to be an obstacle for successful grassland restoration (e.g. Bakker et al. 1996; Bullock et al. 2002; Donath, Hölzel & Otte 2003; Bossuyt & Honnay 2008; Bischoff, Warthemann & Klotz 2009), and propagule pressure has been shown to be a major driver of community assembly (e.g. Von Holle & Simberloff 2005). Large-scale mined sites act as huge seed traps in the landscape (Kirmer et al. 2008), providing a large number of niches for (re)colonization of species well adapted to open habitats (e.g. Bradshaw 1983; Ash, Gemmell & Bradshaw 1994; Schulz & Wiegleb 2000; Kirmer et al. 2008) and thus can generate highly diverse habitats (e.g. Prach & Pyšek 2001; Prach 2003; Tischew & Kirmer 2007; Řehounková & Prach 2008). However, these processes take time. The transfer of green hay and sowing of site-specific seed mixtures of local provenance can accelerate this process considerably leading to the rapid establishment of species-rich grasslands. Thus, our results are in line with several studies from other restoration sites (e.g. ex-arable fields), showing that the use of near-natural restoration methods leads to high establishment rates of target grassland species (e.g. Pywell et al. 2002; Hölzel & Otte 2003; Kiehl, Thormann & Pfadenhauer 2006; Jongepierová, Mitchley & Tzanopoulos 2007; Lepšet al. 2007; Török et al. 2010; for review see Kiehl et al. 2010).

Moreover, application of green hay as well as direct sowing not only affected the speed of succession, but also the pathway. In spontaneously developing sites, C. epigejos migrated from nearby source populations, expanded by vegetative means, showed a high persistence rate and thus became increasingly dominant over time. In Central European man-made habitats, Calamagrostis is very common, often forming dense swards and can even arrest succession altogether for a long time period (e.g. Prach & Pyšek 2001; Rebele & Lehmann 2001; Wiegleb & Felinks 2001a; Prach 2003; Tischew & Kirmer 2007; Mudrák, Frouz & Velichová 2010). Therefore, if potentially ‘problematic’ species (e.g. ruderal or invasive species) occur in the proximity of restoration sites, spontaneous succession might not successfully recreate species-rich grasslands (see also Prach & Hobbs 2008). These results confirm our first hypothesis.

Species introduction via hay transfer and sowing had a long-term effect on species composition and the number of target species. Once established, the species-rich grasslands were highly resistant to invasion by Calamagrostis and other ruderals even though management was rather sporadic. Stable grassland communities developed best in the hay transfer treatment, while in the sowing treatment, the high productivity of legumes caused a temporary decrease in species diversity in 2005 and 2006. However, as in other experiments, the predominance of legumes did not persist over the long term (e.g. Pfisterer et al. 2004; Petermann et al. 2010). Compared to the artificial seed mixture, the seed composition of the harvested green hay may have been more balanced (species number and species identity), resulting in stable and continuous vegetation development towards the intended target community. Moreover, the number of potentially transferable species in hay was five times higher than the number of sown species. Other studies (e.g. Pakeman, Pywell & Wells 2002; Pywell et al. 2002; Walker et al. 2004; Lepšet al. 2007) have also found that highly diverse seed mixtures are more successful in creating species-rich grasslands and that species turnover is low compared to less diverse seed mixtures.

As has been found in other grassland re-creation experiments (e.g. Pakeman, Pywell & Wells 2002), species introduced elsewhere in the experiment migrated into the other treatments. In our experimental site, the number of target species continues to increase on the control plots owing to both migration of target species that were already present in the mining area and migration of target species that have been introduced in adjacent treatments. Thus, successful propagation of local species and species exchange led to a converging trend in species composition between treatments. Interestingly, 9 years after the start of the experiment, we did not find any significant differences between treatments with and without species introduction in terms of total vegetation cover, species richness and the number (but not cover) of target species. However, our experimental site offered ideal conditions with nearby seed sources from adjacent treatments and thus our results are not representative of field-scale restoration projects. In extensively disturbed areas, appropriate seed sources are often long distances away. If connectivity is low, the possibility of target species colonizing the sites may decrease drastically (e.g. Walker et al. 2004). Successful colonization via long-distance dispersal through natural processes takes longer than a few years (Ash, Gemmell & Bradshaw 1994; Brändle et al. 2003; Kirmer et al. 2008). Finally, despite the possibility of species exchange, the overall similarity in community composition between treatments remained quite low until the final observation year. The dominance ratio between target and nontarget species still differed significantly, as treatments where species were introduced still showed a higher abundance of target species and a lower coverage of ruderals. Hence, our results confirm our second hypothesis highlighting the importance of initial floristic composition and the order of species arrivals (e.g. priority effects of competitive species) in vegetation development and convergence to the same species composition.

Practical Recommendations

The application of green hay and direct sowing regional seed mixtures considerably affected both the speed and the pathway of vegetation succession in the restoration sites. Both methods are appropriate restoration tools when rapid vegetation development is desired for aesthetic reasons or erosion control, when no potential seed sources of target species are available nearby, when species with low dispersal capacity are preferred and finally when there is a need to suppress unwanted species. An additional mulch layer is recommended on sites susceptible to erosion. Our results demonstrate that introduced grassland species are able to colonize adjacent nutrient-poor but hospitable sites. Therefore, in large-scale restoration areas with suitable site conditions, we recommend the application of seed-rich green hay or sowing of regional seed mixtures in strips to allow for natural colonization in the intervening gaps resulting in mosaic biotopes of different grassland types.

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

The study was funded by the German Foundation for the Environment (DBU), the European Union (Interreg IIIB CADSES, Interreg IVB Central Europe), the German Federal Ministry for Education, Science, Research and Technology (BMBF) and the Lausitz and Central-German Mining Administration Company (LMBV). We are grateful to all people who helped to collect the data and to Keith Edwards who kindly checked the English. Finally, we thank the editor and two anonymous reviewers for providing constructive comments.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information
  • Ash, H.J., Gemmell, R.P. & Bradshaw, A.D. (1994) The introduction of native plant species on industrial waste heaps: a test of immigration and other factors affecting primary succession. Journal of Applied Ecology, 31, 7484.
  • Baasch, A., Tischew, S. & Bruelheide, H. (2009) Insights into succession processes using temporally repeated habitat models: results from a long-term study in a post-mining landscape. Journal of Vegetation Science, 20, 62938.
  • Baasch, A., Tischew, S. & Bruelheide, H. (2010) Twelve years of succession on sandy substrates in a post-mining landscape: a Markov chain analysis. Ecological Applications, 20, 113647.
  • Bakker, J.P., Poschlod, P., Strykstra, R.J., Bekker, R.M. & Thompson, K. (1996) Seed banks and seed dispersal: important topics in restoration ecology. Acta Botanica Neerlandica, 45, 46190.
  • Bakker, J.D., Wilson, S.D., Christian, J.M., Li, X., Ambrose, L.G. & Waddington, J. (2003) Contingency of grassland restoration on year, site, and competition from introduced grasses. Ecological Applications, 13, 13753.
  • Bischoff, A., Warthemann, G. & Klotz, S. (2009) Succession of floodplain grasslands following reduction in land use intensity: the importance of environmental conditions, management and dispersal. Journal of Applied Ecology, 46, 2419.
  • Bossuyt, B. & Honnay, O. (2008) Can the seed bank be used for ecological restoration? An overview of seed bank characteristics in European communities. Journal of Vegetation Science, 19, 87584.
  • Bradshaw, A.D. (1983) The reconstruction of ecosystems. Journal of Applied Ecology, 20, 117.
  • Brändle, M., Durka, W., Krug, H. & Brandl, R. (2003) The assembly of local communities: plants and birds in non reclaimed mining sites. Ecography, 26, 65260.
  • Bullock, J.M., Pywell, R.F. & Walker, K.J. (2007) Long-term enhancement of agricultural production by restoration of biodiversity. Journal of Applied Ecology, 44, 612.
  • Bullock, J.M., Moy, I.L., Pywell, R.F., Coulson, S.J., Nolan, A.M. & Caswell, H. (2002) Plant dispersal and colonization processes at local and landscape scales. Dispersal Ecology (eds J.M. Bullock, R.E. Kenward & R.S. Hails), pp. 297302. Blackwell, Oxford, UK.
  • Clarke, K.R. (1993) Non-parametric multivariate analyses of changes in community structure. Australian Journal of Ecology, 18, 11743.
  • Cremene, C., Groza, G., Rakosy, L., Schileyko, A.A., Baur, A., Erhardt, A. & Baur, B. (2005) Alterations of steppe-like grasslands in eastern Europe; a threat to regional biodiversity hotspots. Conservation Biology, 19, 160618.
  • Cummings, J., Reid, N., Davies, I. & Grant, C. (2005) Adaptive restoration of sand-mined areas for biological conservation. Journal of Applied Ecology, 42, 16070.
  • Donath, T.W., Hölzel, N. & Otte, A. (2003) The impact of site conditions and seed dispersal on restoration success in alluvial meadows. Applied Vegetation Science, 6, 1322.
  • Drake, J.A. (1991) Community-assembly mechanics and the structure of an experimental species ensemble. American Naturalist, 137, 126.
  • Egler, F.E. (1954) Vegetation science concepts. I. Initial floristic composition-a factor in oldfield vegetation development. Vegetatio, 4, 4127.
  • European Environment Agency. (2010) EU 2010 biodiversity baseline. EEA Technical report, 12. EEA, Copenhagen.
  • Felinks, B., Pilarski, M. & Wiegleb, G. (1998) Vegetation survey in the former brown coal mining area of eastern Germany by integrating remote sensing and ground-based methods. Applied Vegetation Science, 1, 23340.
  • Fiala, K., Záhora, J., Tůma, I. & Holub, P. (2004) Importance of plant matter accumulation, nitrogen uptake and utilisation in expansion of tall grasses (Calamagrostis epigejos and Arrhenatherum elatius) into an acidophilous dry grassland. Ekológia, 23, 22540.
  • Grant, C.D., Campbell, C.J. & Charnock, N.R. (2002) Selection of species suitable for derelict mine site rehabilitation in New South Wales, Australia. Water, Air, & Soil Pollution, 139, 21535.
  • Hammer, Ø., Harper, D.A.T. & Ryan, P.D. (2001) PAST: paleontological statistics software package for education and data analysis. Palaeontologia Electronica, 4, 19.
  • Hölzel, N. & Otte, A. (2003) Restoration of a species-rich flood meadow by topsoil removal and diaspore transfer with plant material. Applied Vegetation Science, 6, 13140.
  • Jeschke, M. & Kiehl, K. (2006) Auswirkung von Renaturierungs- und Pflegemaßnahmen auf die Artendiversität von Gefäßpflanzen und Kryptogamen in neu angelegten Kalkmagerrasen. Tuexenia, 26, 22342.
  • Jongepierová, I., Mitchley, J. & Tzanopoulos, J. (2007) A field experiment to recreate species-rich hay meadows using regional seed mixtures. Biological Conservation, 139, 297305.
  • Kiehl, K., Thormann, A. & Pfadenhauer, J. (2006) Evaluation of initial restoration measures during the restoration of calcareous grasslands on former arable fields. Restoration Ecology, 14, 14856.
  • Kiehl, K., Kirmer, A., Donath, T.W., Rasran, L. & Hölzel, N. (2010) Species introduction in restoration projects – evaluation of different techniques for the establishment of semi-natural grasslands in Central and Northwestern Europe. Basic and Applied Ecology, 11, 28599.
  • Kirmer, A., Baasch, A. & Tischew, S. (2011) Sowing of low and high diversity seed mixtures in ecological restoration of surface mined-land. Applied Vegetation Science, 14, in press, Doi: 10.1111/j.1654-109X.2011.01156.x.
  • Kirmer, A. & Mahn, E.-G. (2001) Spontaneous and initiated succession on unvegetated slopes in the abandoned lignite-mining area of Goitsche, Germany. Applied Vegetation Science, 4, 1927.
  • Kirmer, A., Tischew, S., Ozinga, W.A., von Lampe, M., Baasch, A. & van Groenendael, J.M. (2008) Importance of regional species pools and functional traits in colonisation processes: predicting re-colonisation after large-scale destruction of ecosystems. Journal of Applied Ecology, 45, 152330.
  • Krauss, J., Bommarco, R., Guardiola, M., Heikkinen, R.K., Helm, A., Kuussaari, M., Lindborg, R., Öckinger, E., Pärtel, M., Pino, J., Pöyry, J., Raatikainen, K.M., Sang, A., Stefanescu, C., Teder, T., Zobel, M. & Steffan-Dewenter, I. (2010) Habitat fragmentation causes immediate and time-delayed biodiversity loss at different trophic levels. Ecology Letters, 13, 597605.
  • Lepš, J., Doležal, J., Bezemer, T.M., Brown, V.K., Hedlund, K., Igual Arroyo, M., Jörgensen, H.B., Lawson, C.S., Mortimer, S.R., Peix Geldart, A., Rodríguez Barrueco, C., Santa Regina, I., Šmilauer, P. & van der Putten, W.H. (2007) Long-term effectiveness of sowing high and low diversity seed mixtures to enhance plant community development on ex-arable fields. Applied Vegetation Science, 10, 97110.
  • Mudrák, O., Frouz, J. & Velichová, V. (2010) Understory vegetation in reclaimed and unreclaimed post-mining forest stands. Ecological Engineering, 36, 78390.
  • Norris, J.E., Stokes, A., Mickovski, S.B., Cammeraat, E., van Beek, R., Nicoll, B.C. & Achim, A. (2008) Slope Stability and Erosion Control: Ecotechnological Solutions. Springer, Dordrecht, The Netherlands.
  • Pakeman, R.J., Pywell, R.F. & Wells, T.C.E. (2002) Species spread and persistence: implications for experimental design and habitat re-creation. Applied Vegetation Science, 5, 7686.
  • Pärtel, M., Kalamees, R., Reier, Ü., Tuvi, E.-L., Roosaluste, E., Vellak, A. & Zobel, M. (2005) Grouping and prioritization of vascular plant species for conservation: combining natural rarity and management need. Biological Conservation, 123, 2718.
  • Petermann, J.S., Fergus, A.J.F., Roscher, C., Turnbull, L.A., Weigelt, A. & Schmid, B. (2010) Biology, chance, or history? The predictable reassembly of temperate grassland communities. Ecology, 91, 40821.
  • Pfisterer, A.B., Joshi, J., Schmid, B. & Fischer, M. (2004) Rapid decay of diversity-productivity relationships after invasion of experimental plant communities. Basic and Applied Ecology, 5, 514.
  • Poschlod, P. & Biewer, H. (2005) Diaspore and gap availability are limiting species richness in wet meadows. Folia Geobotanica, 40, 1334.
  • Prach, K. (2003) Spontaneous succession in Central-European man-made habitats: what information can be used in restoration practice? Applied Vegetation Science, 6, 1259.
  • Prach, K. & Hobbs, R.J. (2008) Spontaneous succession versus technical reclamation in the restoration of disturbed sites. Restoration Ecology, 16, 3636.
  • Prach, K. & Pyšek, P. (2001) Using spontaneous succession for restoration of human-disturbed habitats: experience from Central Europe. Ecological Engineering, 17, 5562.
  • Pywell, R.F., Bullock, J.M., Hopkins, A., Walker, K.J., Sparks, T.H., Burke, M.J.W. & Peel, S. (2002) Restoration of species-rich grassland on arable land: assessing the limiting processes using a multi-site experiment. Journal of Applied Ecology, 39, 294309.
  • Rebele, F. & Lehmann, C. (2001) Biological flora of Central Europe: Calamagrostis epigejos (L.) Roth. Flora, 196, 32544.
  • Řehounková, K. & Prach, K. (2008) Spontaneous vegetation succession in gravel-sand pits: a potential for restoration. Restoration Ecology, 16, 30512.
  • Rydgren, K., Nordbakken, J.-F., Austad, I., Auestad, I. & Heegaard, E. (2010) Recreating semi-natural grasslands: a comparison of four methods. Ecological Engineering, 36, 16729.
  • Schubert, R. (2001) Prodromus der Pflanzengesellschaften Sachsen-Anhalts. Mitteilungen zur floristischen Kartierung Sachsen-Anhalts, Sonderheft 2. Herausgegeben vom Botanischen Verein Sachsen-Anhalt e.V. Halle/Saale.
  • Schulz, F. & Wiegleb, G. (2000) Development options of natural habitats in a post-mining landscape. Land Degradation and Development, 11, 99110.
  • Somodi, I., Virágh, K. & Podani, J. (2008) The effect of the expansion of the clonal grass Calamagrostis epigejos on the species turnover of a semi-arid grassland. Applied Vegetation Science, 11, 18792.
  • Tischew, S. & Kirmer, A. (2007) Implementation of basic studies in the ecological restoration of surface-mined land. Restoration Ecology, 15, 3215.
  • Tongway, D.J. & Ludwig, J.A. (2011) Restoring Disturbed Landscapes – Putting Principles into Practice. Island Press, Washington, Covelo, London.
  • Török, P., Deák, B., Vida, E., Valkó, O., Lengyel, S. & Tóthmérész, B. (2010) Restoring grassland biodiversity: sowing low-diversity seed mixtures can lead to rapid favourable changes. Biological Conservation, 143, 80612.
  • Tropek, R., Kadlec, T., Karesova, P., Spitzer, L., Kocarek, P., Malenovsky, I., Banar, P., Tuf, I.H., Hejda, M. & Konvicka, M. (2010) Spontaneous succession in limestone quarries as an effective restoration tool for endangered arthropods and plants. Journal of Applied Ecology, 47, 13947.
  • Usher, M.B. (1992) Statistical models of succession. Plant Succession: Theory and Prediction (eds D.C. Glenn-Lewin, R.K. Peet & T.T. Veblen), pp. 215248. Chapman & Hall, London.
  • Von Holle, B. & Simberloff, D. (2005) Ecological resistance to biological invasion overwhelmed by propagule pressure. Ecology, 86, 32128.
  • Walker, L.R. & del Moral, R. (2009) Lessons from primary succession for restoration of severely damaged habitats. Applied Vegetation Science, 12, 55.
  • Walker, L.R., Walker, J. & del Moral, R. (2007) Forging a new alliance between succession and restoration. Linking Restoration and Ecological Succession (eds L.R. Walker, J. Walker & R.J. Hobbs), pp. 118. Springer, New York.
  • Walker, K.J., Stevens, P.A., Stevens, D.P., Mountford, J.O., Manchester, S.J. & Pywell, R.F. (2004) The restoration and re-creation of species-rich lowland grassland on land formerly managed for intensive agriculture in the UK. Biological Conservation, 119, 118.
  • WallisDeVries, M.F., Poschlod, P. & Willems, J.H. (2002) Challenges for the conservation of calcareous grasslands in northwestern Europe: integrating the requirements of flora and fauna. Biological Conservation, 104, 26573.
  • Wiegleb, G. & Felinks, B. (2001a) Predictability of early stages of primary succession in post-mining landscapes of Lower Lusatia, Germany. Applied Vegetation Science, 4, 518.
  • Wiegleb, G. & Felinks, B. (2001b) Primary succession in post-mining landscapes of Lower Lusatia – chance or necessity. Ecological Engineering, 17, 199217.
  • Zobel, M., Otsus, M., Liira, J., Moora, M. & Möls, T. (2000) Is small-scale species richness limited by seed availability or microsite availability? Ecology, 81, 327482.

Supporting Information

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

Appendix S1. Seed mixture used for slope restoration (sowing treatment) in the mined site Roßbach.

Appendix S2. List of 97 higher plant species from the donor site ‘Göttersitz’ used for hay transfer.

As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials may be re-organized for online delivery, but are not copy-edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.

FilenameFormatSizeDescription
JPE_2086_sm_AppendixS1-S2.doc145KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.