• disturbance;
  • grassland restoration;
  • lime;
  • nutrient availability;
  • pH;
  • threatened plant species;
  • vegetation


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments

Specialist plant species in calcareous sandy grasslands are threatened by acidification and high nutrient levels in the topsoil. We investigated whether topsoil removal and soil perturbation in degraded sandy grasslands could lead to establishment of specialist species belonging to the threatened xeric sand calcareous grassland habitat. Restoration actions performed in 2006 resulted in increased soil pH and reduced nitrogen availability. We found early colonisztion of the perennial key species Koeleria glauca after both deep perturbation and topsoil removal, and high seedling establishment in topsoil removal plots 5 and 6 years following the restoration treatment (2011–2012). After topsoil removal, overall vegetation composition in 2012 had developed toward the undegraded community, with target species accounting for 20% of the community after topsoil removal, compared to 30% in the undegraded vegetation, and less than 1% in untreated controls. Deep perturbation led to 7% target species, while there were almost no effects of shallow perturbation 6 years following treatment. These results demonstrate that topsoil removal can promote colonization of target species of calcareous sandy grassland and highlights the importance of considering the regeneration niche for target species when implementing restoration measures.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments

Semi-natural calcareous grasslands in Europe and their biodiversity are threatened by abandonment, eutrophication, and acidification (Roem & Berendse 2000; Poschlod et al. 2005; Smits et al. 2008; Mårtensson & Olsson 2010). These factors lead to habitat degradation, reducing the populations of threatened specialist species. The remaining areas of semi-natural calcareous grassland are usually highly fragmented, which poses a further threat to the rarest species (Fischer & Stöcklin 1997). The survival of many threatened plant species in calcareous sandy grasslands has been shown to depend on regular soil disturbance as well as low nutrient availability and a high pH (Olsson et al. 2009; Eichberg et al. 2010). Nutrient enrichment may result in common generalists outcompeting threatened specialist species associated with nutrient-poor sandy grasslands, while acidification may result in heath-like vegetation, with a lower species-richness (Süss et al. 2004; Olsson et al. 2009; Mårtensson & Olsson 2010).

A number of restoration methods, such as topsoil removal, sod cutting, and soil inversion have been implemented to restore abiotic conditions suitable for high diversity and threatened species in semi-natural calcareous grasslands (Dolman & Sutherland 1994; Kiehl & Pfadenhauer 2007; Schnoor & Olsson 2010; Ödman et al. 2011). The responses of specialist species to different kinds of disturbance must be better understood in order to preserve disturbance-driven biodiversity and to improve our understanding of the dynamics in vegetation processes (Chapin III et al. 1996). One such disturbance-driven and very threatened habitat is the xeric sand calcareous grassland (Natura 2000 code 6120, 2002/83/EC Habitat Directive), which occurs on sandy soil and is restricted to glaciofluvial deposits and inland dunes. This habitat consists of open grasslands on calcareous, more or less humus-free, nutrient-poor, and well-drained sand with a discontinuous vegetation cover. In Sweden, it occurs in a small and very fragmented area in the southernmost parts of the country, and the vegetation type is locally called sand steppe (Sjörs 1967). Many of the sand steppes in Sweden have become degraded due to soil acidification and nutrient accumulation (Olsson 1994; Tyler 2003; Mårtensson & Olsson 2010). In the 18th century, von Linné (1751) noted extensive sand drift and sand fields in our study area and found many xeric sand calcareous grassland species growing on them. A drastic decline in sand steppes is thought to have occurred during the last century (Tyler 2003; Mattiasson 2009), and this is reflected in the high density of nationally red-listed (and thereby declining) plant species (Olsson et al. 2009). This decline is commonly believed to be caused mainly by changes in land use, that is, the old agricultural practices in these grasslands were abandoned, resulting in a reduction of bare sand areas (Tyler 2003; Mattiasson 2009), posing a threat to the disturbance-favored flora (Hobbs & Huenneke 1992; Schläpfer et al. 1998; Olsson et al. 2009) and fauna (Riksen et al. 2008).

We manipulated soil in two degraded sand steppe grassland areas in Sweden: the Lyngsjö site, where acidification has been identified as the driving factor behind sand steppe degradation (Mårtensson & Olsson 2010; Ödman et al. 2011) and the Everöd site where eutrophication was suspected to be the driving factor behind sand steppe degradation (Olsson et al. 2009). Topsoil removal was performed at three sites in the Everöd area in order to restore nutrient poor conditions, and soil perturbation was performed at the acidified Lyngsjö site in an attempt to restore a lime-rich habitat by bringing up lime-containing sand from deeper soil layers. We tested the hypothesis that the treatments would lead to spontaneous establishment of the perennial grass Koeleria glauca, which is a key indicator species of this type of vegetation (Sjörs 1967; Rodwell et al. 2002) that occurs at all sites where experiments were performed. We also tested the hypothesis that the implemented treatments would increase the overall proportion of sand steppe specialist plant species. We compared the vegetation in treated areas to untreated controls at each site, as well as to an undegraded plot that was judged to be the best example of sand steppe vegetation within the study area.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments

Study Area and Restoration Target

The Everöd area was situated on redeposited wave washed sand around Kristianstad airport and consisted of large areas of sandy grasslands with low vegetation, managed by grazing, mowing, and harrowing. At the time of study, the sand steppe habitat in the area was deemed to be in a mostly degraded state (Tyler 2003), but with small patches of typical sand steppe vegetation remaining.

The Lyngsjö area was situated on a glaciofluvial deposit approximately 1 km west of Kristianstad airport. The area used to be a ploughed field with long periods of fallow, but during the last 30–40 years it has been a grazed pasture. The plant community predominantly consisted of heath-like vegetation resembling continental dunes with open Corynephorus and Agrostis grasslands (N2330, EU habitat directive). Patches of sand steppe remain in the area, mainly along the roadsides due to previous digging activities.

Bare sand is one of the key characters of xeric sand calcareous grassland and is related to diversity of plants and insects (Olsson et al. 2009; Ödman et al. 2011). Increased amount of bare sand is a restoration goal that is automatically achieved when implementing soil disturbance. A second restoration goal is to achieve colonization of the specialist plant species characteristic of this habitat. To evaluate the former, we used lists of specialist species (Table 6) from other studies in the same area (Olsson 1994; Tyler 2003). We selected a plot with typical sand steppe vegetation in the Everöd area as an example of an undegraded site.

Topsoil Removal Experiment

Topsoil removal was performed at three sites in the Everöd area (denoted Everöd 1, 2, and 3). At each of the three sites, nontreated control plots (10 m × 10 m) were established adjacent to the treated plots (Table 1). The vegetation in the three treated and three control plots of the Everöd area was classified as Fennoscandian lowland species-rich dry to mesic grassland (N6270, EU habitat directive). Everöd 1 and 2 were located 100 m apart, and Everöd 3 was situated 3 km south southeast of Everöd 1 and 2. Everöd 1 was situated close to large areas of sandy grassland with low vegetation, managed by mowing and harrowing. Slightly degraded sand steppe was in close proximity to the site, which still had rather large populations of target species such as Koeleria glauca, Dianthus arenarius, and Anthericum ramosum. Everöd 2 was situated in a pasture on a former pine plantation. This site was grazed 1–2 weeks every year and patches of two target species, K. glauca and Pulsatilla pratensis, were common. Everöd 3 was situated close to an area with a species-rich flora containing many red-listed and threatened plant species, and sand steppe specialists dominated some areas. Well-developed patches of xeric sand calcareous grassland (sand steppe) could be observed at this site, which is why we choose a plot (10 m × 10 m) from this site as an example of undegraded habitat.

Table 1. An overview of the treatments at the sites of this study, their position according to RT90 (Swedish grid), the area treated, the date of treatment and the distance from the nearest population of the key species Koeleria glauca
SiteLocationTreatmentDate of Treatment (d/month/year)Area (m2)Distance to K. glauca (m)
  1. Shallow and deep perturbation at Lyngsjö was carried out following a blocked design at the same location (Ödman et al. 2011).

Everöd 1X: 6201766 Y: 1392796Topsoil removal25/4/200690010–20
Everöd 2X: 6201875 Y: 1392814Topsoil removal25/4/2006500<10
Everöd 3X: 6198833 Y: 1393856Topsoil removal7/10/2006250<10
LyngsjöX: 6201499 Y: 1391798Deep/Shallow perturbation13/5/20064 × 645–10
TargetX: 6198806 Y: 13938441000

We performed topsoil removal at all Everöd sites to remove the nutrient rich topsoil layer with high organic matter content. Lime-rich soil layers had been detected at depths of 0–0.5 m in this area, and therefore we expected that topsoil removal would increase lime content and pH of the surface soil.

At all sites topsoil was removed using an excavator. At Everöd 1, topsoil was removed to a depth of 0.5 m in a 30 m × 30 m area. At Everöd 2, topsoil was removed from an area of 500 m2, consisting of one rectangular part (10 m × 20 m) and another more long-stretched and narrow part (3 m × 100 m). At Everöd 3, a 10 m × 25 m area was subjected to topsoil removal to a depth of 0.3 m. The exposed subsoil at Everöd 3 was perturbed to a depth of a further 30 cm to ensure exposure of lime-rich sand.

Soil Perturbation Experiment

Two types of restoration treatments at the Lyngsjö site, deep perturbation (to a depth of 1 m) and shallow perturbation (to a depth of 0.30 m), were compared to control plots with no disturbance (Ödman et al. [2011] for details). Lime-rich sand began at 0.5 m depth at this site. The soil was unearthed and mixed using an excavator. Deep perturbation was designed to incorporate lime-rich soil into the surface soil, whereas the shallow perturbation was designed to disturb the soil, but not alter the lime content. All treatments were repeated four times, once in each of four blocks in treatment plots that were 8 m × 8 m. The blocks were situated 2–20 m from each other along a transect. The short-term effects of this soil perturbation experiment are described in Ödman et al. (2011).

Soil Sampling and Analyses

The topsoil (0–10 cm) from each plot at all sites was sampled (two to four subsamples per plot) in September 2011 using a 3 cm in diameter soil corer. Soil samples were stored 1–2 days in a refrigerator before N extraction, after which they were stored in a freezer until further analysis.

Extractable phosphate-P was determined using Bray 1 extraction, followed by flow injection analysis (Fransson et al. 2003). Exchangeable nitrate-N and ammonium-N were determined using flow injection analysis (ISO 13395 and ISO 11732:2005, respectively) following 2 hour extraction of 20 g soil at field moisture with 100 mL 0.2M BaCl2. For pH measurements, 10 g of soil were dissolved in 50 mL distilled water for 2 hours on a rotator, and the pH was measured electrometrically in the supernatants after 24 hours sedimentation.

Total soil organic and inorganic carbon was analyzed using a PrimacsSLC TOC Analyzer (Skalar Analytical B. V., Breda, The Netherlands). The method is based on the measurement of the total C from the CO2 emitted after combustion at 1,050°C, and the measurement of the total inorganic C from the CO2 produced after treatment with 20% phosphoric acid at 100°C. Total organic C was calculated by subtracting the inorganic C from the total C. We assumed that all inorganic C originated from CaCO3, and recalculated the values to determine % CaCO3 (lime).

Vegetation Sampling

The total number of tussocks of the specialist species K. glauca was surveyed at all sites in June each year from 2007 to 2012. The frequency of each plant species was estimated between 12th and 14th of June 2012 by recording the occurrence of each species of vascular plants, bryophytes, and lichens in 1 m2 quadrats. We used 10 quadrats (each species could thus have a score between 0 and 10 in each of the plots) for each plot at the Lyngsjö site, as well as the controls and undegraded plots at the Everöd sites. The three topsoil removal plots at the Everöd sites were larger and had a more sparse vegetation cover (Table 1). Thus, we used 30 quadrats for Everöd 1, and 20 quadrats for Everöd 2 and 3. Quadrats were placed in a systematically in each plot along two (Lyngsjö controls and undegraded) or three transects. We also estimated the total vegetation cover, including bryophytes, in each quadrat.

Statistical Analyses

Soil chemical properties and vegetation characteristics were analyzed using a paired t-test for the Everöd study and a one-way analysis of variance (ANOVA) followed by Fisher's least significant difference (LSD) means comparison procedure test for the Lyngsjö study. Plant communities were analyzed using nonmetric multidimensional scaling (NMDS) based on plant records at all plots and Bray–Curtis distances using PRIMER-E (Clarke & Gorley, 2006).


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments

Soil Chemistry

Topsoil removal at the Everöd sites exposed soil that had pH levels of around 8 and carbonate-C indicating around 8% lime (Table 2). Organic C was reduced to below 0.5% by the restoration action, although not significantly different from the controls. The pH after topsoil removal was similar to that of the undegraded site, but the content of both organic C and lime were higher at the undegraded site. Available nitrate concentrations were much lower than in the control soils, and similar to those of the undegraded site (Table 2). Bray phosphate, however, was not significantly reduced by the restoration action, and Bray phosphate of the undegraded site was almost 10 times lower (Table 2).

Table 2. Soil chemical characteristics (pH, lime content, and soil organic matter [OM] content) and nutrient availability (0–10 cm depth) in control and topsoil removal treatments
ParameterControlTopsoil RemovalPaired t-testUndegraded
  1. Values represent means ± SE (n = 3, df = 2), differences indicated by results of paired t-tests (p-values <0.05 were considered significant and displayed in bold). Note that SE for the undegraded site describes only the variation among four samples within one plot and were not included in the statistical test.

pH (H2O)6.9 ± 0.198.0 ± 0.140.0266.068.14 ± 0.10
Lime (%)2.5 ± 1.37.9 ± ± 3.7
OM (%)3.2 ± 1.20.13 ± ± 0.20
NH4-N (µg/g)2.2 ± 0.650.17 ± 0.080.10−2.900.26 ± 0.03
NO3-N (µg/g)0.96 ± 0.210.18 ± 0.090.029−5.770.41 ± 0.10
PO4-P (µg/g)29 ± 1517 ± 4.50.38−1.251.82 ± 0.63

Deep perturbation increased the soil pH and decreased organic matter content, NO3-N and extractable P at the Lyngsjö site (Table 3). However, pH was only elevated to slightly over 7, and no lime was detected in the topsoil after deep perturbation. Furthermore, NH4-N was not reduced significantly. Shallow perturbation neither increased pH nor decreased the organic matter content. Perturbation plots differed from the undegraded site, with lower pH and lime content, and higher P availability.

Table 3. Soil chemical characteristics (pH, soil organic matter [OM] content) and nutrient availability (0–10 cm depth) in control and shallow and deep perturbation treatment
ParameterSoil PerturbationOne-way ANOVA
  1. Values represent means ± SE (n = 4, df = 2), differences indicated by results of one-way analysis of variance (ANOVA) and significant differences between types of treatments are indicated by different letters (Fisher's least significant difference [LSD]).

pH (H2O)5.6 ± 0.02A5.7 ± 0.04A7.1 ± 0.28B<0.00125.3
OM (%)0.94 ± 0.07A0.84 ± 0.08A0.46 ± 0.02B<0.00165.4
NH4-N (µg/g)0.50 ± 0.18A0.29 ± 0.07A0.31 ± 0.05A0.420.95
NO3-N (µg/g)0.87 ± 0.09A0.88 ± 0.12A0.45 ± 0.11B0.0186.5
PO4-P (µg/g)74 ± 5.3AB95 ± 5.5A47 ± 15B0.0206.3

Specialist Species and Vegetation Composition

All three topsoil removal plots at the Everöd sites were colonized by Koeleria glauca, while this species did not occur in the adjacent controls. Initially, only a few tussocks appeared; however, between 2010 and 2012 number of tussocks increased along with many new seedlings (Fig. 1a). The density of K. glauca tussocks in the deep perturbation plots at the Lyngsjö site increased during the 6-year period, although a slight increase also occurred in the surrounding shallow perturbation plots and in the control plots (Fig. 1b).


Figure 1. Population development of Koeleria glauca following (a) topsoil removal in each individual plot at Everöd and (b) in response to the two soil perturbation treatments and controls at Lyngsjö (mean ± SE). As no K. glauca occurred in controls at the Everöd sites, these were not included.

Download figure to PowerPoint

Species richness (per m2) tended to decrease in response to topsoil removal (Table 4), but not following soil perturbation (Table 5). The highest species richness was recorded in the undegraded plot (Table 3). The proportion of specialist species increased in response to both topsoil removal and deep perturbation, and was highest after topsoil removal (20%) and in the undegraded plot (30%). Target species did not exceed 1% in any of the control plots. Among the specialist species, annuals were only found after topsoil removal and deep perturbation, and not in the undegraded plot. Among annual species, there was no significant difference due to treatments.

Table 4. Vegetation characteristics in topsoil removal plots and adjacent nontreated control plots
ParameterControlTopsoil RemovalPaired t-testUndegraded
  1. Species richness (SR) represents the mean number of species m2. The percentage target species was calculated from the number of all species occurrences in a plot that were target species. The percentage annual target species was calculated from the number of target species occurrences in a plot that were annual species. Values represent means ± SE (n = 3, df = 2), differences indicated by results of paired t-tests (p-values <0.05 were considered significant and displayed in bold).

SR (no m−1)8.1 ± 1.05.9 ± 0.410.06−3.714.2
Target species (%)0.34 ± 0.3419.6 ± 1.50.00811.029.6
Annual target species (%)020.1 ± 17.40.361.20
Annual species (%)10.0 ± 5.218.8 ±
Vegetation cover (%)100 ± 025.5 ± 2.5<0.00150.971
Table 5. Vegetation characteristics in perturbated (shallow and deep) plots and in nontreated control plots
ParameterSoil PerturbationOne-way ANOVA
  1. Species richness (SR) represents the mean number of species in each analyzed m2. The percentage target species was calculated from the number of all species occurrences in a plot that were target species. The percentage annual target species was calculated from the number of target species occurrences in a plot that were annual species. Values represent means ± SE (n = 4, df = 2), differences indicated by results of one-way analysis of variance (ANOVA, p-values <0.05 were considered significant and displayed in bold) and significant differences between types of treatments are indicated by different letters (Fisher's least significant difference [LSD]).

SR (no m−1)9.3 ± 0.5710.3 ± 0.789.8 ± 0.320.560.52
Target species (%)0.69 ± 0.69 A2.1 ± 0.90 A6.9 ± 1.5 B0.0069.3
Annual target species (%)005.6 ±
Annual species (%)19.7 ± 3.014.6 ± 3.219.8 ± 1.60.321.3
Vegetation cover (%)79.4 ± 6.0 A74.0 ± 2.9 A39.2 ± 3.4 B<0.00126

The NMDS showed that topsoil removal induced changes that clearly separated treated plots from controls, but that the former had a vegetation composition rather different from undegraded vegetation (Fig. 2a). Furthermore, the NMDS showed that the perturbation treatments at the Lyngsjö site altered vegetation composition in the case of deep perturbation, but not in the case of shallow perturbation (Fig. 2b). The deep perturbation plots were more similar to the undegraded plot along axis 1 than to the controls.


Figure 2. Nonmetric multidimensional scaling (NMDS) ordination of plant communities based on plant records (frequency) and Bray-Curtis similarities. (a) The deep perturbation, shallow perturbation, and nondisturbed controls at the Lyngsjö site (final 2D stress = 0.055). An undegraded plot in the Everöd area was included as a reference in the analysis. (b) The three topsoil removal sites, their controls, and an undegraded site in the Everöd area (final 2D stress = 0.067).

Download figure to PowerPoint

High abundance of generalist species such as Festuca rubra, Festuca brevipila, and Poa pratense was observed in topsoil removal plots, while these species were not present at all in the undegraded vegetation (data not shown). Furthermore, the sand heath specialist Corynephorus canescens was found after topsoil removal and deep perturbation, but it was not present in the undegraded vegetation.

Nine specialist species were found in topsoil removal plots (Table 6), some of which were not recorded in the undegraded plot. Only three specialist species were recorded after deep perturbation, and two of these also in the control (Table 6). Overall, only few occurrences of specialist species were recorded in control plots.

Table 6. The frequency (proportion occurrences in 1 m2 squares) of specialist species listed in order of abundance in the undegraded plots. National red-listing according to Gärdenfors (2010) in brackets (EN = Endangered and VU = Vulnerable)
SpeciesaEveröd, Topsoil RemovalLyngsjö, Soil Perturbation
  1. Dianthus arenarius ssp. arenarius is protected by the Natura 2000 network (EU habitat directive, Annex II).

Koeleria glauca (EN)1.00.60 ± 0.140.38 ± 0.100.08 ± 0.000.03 ± 0.00
Festuca polesica1.00.09 ± 0.09
Dianthus arenarius ssp arenarius (EN)0.800.01 ± 0.01
Helianthemum nummularium0.70
Euphrasia stricta var. stricta0.30
Potentilla tabernaemontani0.20
Pulsatilla pratensis0.100.15 ± 0.130.03 ± 0.03
Scabiosa canescens (VU)0.100.01 ± 0.01
Helichrysum arenarium0.06 ± 0.030.25 ± 0.060.15 ± 0.100.05 ± 0.00
Andosaceae septentrionalis0.03 ± 0.030.05 ± 0.03
Silene conica0.17 ± 0.14
Satureja acinos0.02 ± 0.02


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments

The occurrence of bare soil is a key factor for many threatened species, and topsoil removal is particularly efficient in creating bare soil that remains for several years after treatment (Jentsch et al. 2009). Topsoil removal is a recognized method of restoring grasslands and increasing biodiversity (Kiehl & Pfadenhauer 2007). This measure has been used to restore grey dunes and to enhance sand mobility in order to preserve species that depend on such habitat conditions (Arens et al. 2005; Van Til & Kooijman 2007). The main purpose of such measures is to create a nutrient-poor space for the establishment of an early successional stage, preferred by many threatened species. Topsoil removal and soil disturbances may also reach deeper soil layers with a high pH and limestone content (Dolman & Sutherland 1994; Jones et al. 2010).

We found establishment of specialist species after severe disturbance treatments, and annual target species occurred only after disturbance treatments. The fact that disturbance did not increase the total proportion of annual plant species (considered as disturbance adapted) may reflect that these disturbed habitats, with extreme abiotic conditions (drought and nutrient stress), select for stress tolerance strategies (Grime 1977; Bahr et al. 2012). In another area where soil plots were subjected to deep perturbation (unpublished data), we observed that continuous disturbance caused by ungulate grazing at early successional stages after soil perturbation, may be too harsh a pressure for the development of target vegetation, in particular if restored areas are small compared to the grazed grassland in which they are situated.

The fact that annual target species were present in the treated plots, but not in the undegraded plot, highlights the importance of disturbance events and a dynamic approach to management practices in sandy grasslands. Continuous disturbance may be necessary to maintain environments suitable for the species of different successional stages, which is why some military training areas have a large number of threatened species and become protected areas (Gazenbeek 2005; Leis et al. 2005; Jentsch et al. 2009). The success of topsoil removal described in this study in promoting natural establishment of several specialist species may explain the fact that many of the remaining sand steppes occur on former sand pits and on military training ground.

This study demonstrates that drastic measures such as topsoil removal can lead to rapid colonization of the a long-lived target perennial grass species, Koeleria glauca if the abiotic conditions being restored approximate to those of near undegraded levels. Initially, the colonization with K. glauca was slow, but this species steadily increased throughout the investigated period. This may reflect a typical population growth curve, and was similar to the population growth of perennial species observed at landfill sites in Berlin (Rebele & Lehmann 2002). Although the soil chemistry of most sand steppes in this region has been investigated (Olsson et al. 2009), this may not reflect the conditions required for the regeneration of sand steppe species as the regeneration niche may differ from that of species growing in established vegetation (Grubb 1977). Facilitation may also be important during succession following restoration treatment (Gomez-Aparicio 2009), especially as the establishment of favorable abiotic and biotic conditions for a specific species, or a vegetation type, may require time to develop.

Soil perturbation may provide a means to decrease nutrient availability and increase pH in grasslands (Luscombe et al. 2008; Ödman et al. 2011). Although deep soil perturbation increased the pH at the Lyngsjö site, there were fewer occurrences of K. glauca and other specialists in these plots compared to topsoil removal plots in the Everöd sites. Generalist species dominated 2 years after restoration in the Lyngsjö site (Ödman et al. 2011), and continued to do so after 6 years. Koeleria glauca has a wide pH range, occurring in soil with pH as low as 5 (Olsson et al. 2009, 2010). The seedling establishment of K. glauca at pH near 7 following deep perturbation was still slow. This might indicate that the regeneration niche of this long-lived calcicole species, as well as others, differs from their distribution range, with a regeneration niche at higher pH levels. The aforementioned might, therefore, be an example of extinction dept, as areas with pH above 7.5 in the topsoil are rare (Olsson et al. 2009, 2010).

Positive responses of target species to topsoil removal have also been found by Buisson et al. (2006) and Eichberg et al. (2010) when combined with seeding and plant transplantation. We found that the colonization by sandy grassland target species may be particularly rapid when topsoil is removed in a site close to undegraded vegetation. Succession after topsoil removal was still at an early stage, but it is clear that low N availability and high lime content contributed to the establishment of target species and that these conditions prevailed 6 years after treatment. This partly supports the conceptual model proposed by Ödman et al. (2012), where topsoil removal and reduced nutrient levels are proposed to promote a succession of specialist species. However, the restoration measures did not affect available P levels. This might explain why some generalist species, such as Festuca rubra, were common in restored plots. Süss et al. (2004) found that unwanted tall grass invaded sandy grasslands when extractable P was over 20 mg/kg, as measured with the Ca-lactate method. It is difficult to measure plant-available P when measuring over a gradient that includes both calcareous and noncalcareous soils (Fransson et al. 2003; Mårtensson et al. 2012), and it is unclear how the Ca-lactate method compares to that used in this study. It is possible that the high Bray-P availability in restored sites may inhibit further development of species-rich sandy grassland.

Anthropogenic N deposition increases productivity of natural ecosystems (Bobbink et al. 1998) and poses a threat to specialist species and plant diversity (Roem & Berendse 2000; Carroll et al. 2003). Nitrogen availability may be the most important growth-limiting factor in calcareous grassland (Storm & Süss 2008; Mårtensson et al. 2012) and in the area of this study it has been found that N enrichment may lead to invasion of the competitive grass Arrhenatherum elatius at the expense of sandy grassland specialist species (Mårtensson & Olsson 2010). Topsoil removal resulted in N availability below that of the undegraded site, which promotes stress-tolerant specialist species. In the soil perturbation experiment, N availability was measured both in 2008 (Ödman et al. 2011) and 2011 (this study). Between these two occasions, there was no evidence for increase in N availability Succession may need to proceed for N supply to increase during soil formation and achieve conditions suitable for the establishment of desired specialist species (Ödman et al. 2012).

Implications for Practice

  • Topsoil removal can be used as an initial restoration measure on degraded calcareous sandy grasslands where nontarget species dominate, but it can only be successful in restoring calcareous conditions when the exposed underlying material is lime-rich.
  • In areas where topsoil removal is not possible, deep perturbation may be an alternative for restoring calcareous sandy grasslands. At least when the perturbation can reach lime rich soil and bring it to the surface.
  • Habitats for the key species Koeleria glauca and annual sand steppe specialists can be restored by removing both the organic and the leached soil layer of degraded sites, exposing calcareous and N poor sandy soil for primary succession. With nearby donor populations, successful colonization can be reached without seeding.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments

The Swedish Environmental Protection Agency, The County Administrative Board, The Swedish Research Council, and the Oskar and Lilli Lamm Memorial Foundation funded this work. We thank Klaus Birkhofer for statistical advice, Maritza Florian for technical help and Peter Poschlod for valuable comments, Carina Wettemark and Sam Skällberg at the municipality of Kristianstad, and Torgny Roswall helped in planning the experiments.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  • Arens, B., L. Geelen, R. Slings, and H. Wondergem. 2005. Restoration of dune mobility in the Netherlands. in J. -L. Herrier, J. Mees, A. Salman, J. Seys, H. Van Nieuwenhuyse, and I. Dobbelaere, editor. Proceedings of Dunes and Estuaries 2005—International Conference on Nature Restoration. Practices in European Coastal Habitats, Koksijde, Belgium, 19–23 September 2005:129–138.
  • Bahr, A., M. Ellström, T. K. Schnoor, L. Påhlsson, and P. A. Olsson. 2012. Long-term changes in vegetation and soil chemistry in a calcareous and sandy semi-natural grassland. Flora 207:379387.
  • Bobbink, R., M. Hornung, and J. G. M. Roelefs. 1998. The effects of air-borne nitrogen pollutants on species diversity in natural and semi-natural European vegetation. Journal of Ecology 86:717738.
  • Buisson, E., K. D. Holl, S. Anderson, E. Corcket, G. F. Hayes, F. Torre, A. Peteers, and T. Dutoit. 2006. Effect of seed source, topsoil removal, and plant neighbor removal on restoring California coastal prairies. Restoration Ecology 14:569577.
  • Carroll, J. A., S. J. M. Caporn, D. Johnson, M. D. Morecroft, and J. A. Lee. 2003. The interactions between plant growth, vegetation structure and soil processes in semi-natural acidic and calcareous grasslands receiving long-term inputs of simulated pollutant nitrogen deposition. Environmental Pollution 121:363376.
  • Chapin, F. S. III, M. S. Torn, and M. Tateno. 1996. Principles of ecosystem sustainability. American Naturalist 148:10161037.
  • Clarke, K. R., and R. N. Gorley. 2006. PRIMER v6: User Manual/Tutorial. PRIMER-E, Plymouth, UK.
  • Dolman, P. M., and W. J. Sutherland. 1994. The use of soil disturbance in the management of Breckland grass heaths for nature conservation. Journal of Environmental Management 41:123140.
  • Eichberg, C., C. Storm, M. Stroh, and A. Schwabe. 2010. Is the combination of topsoil replacement and inoculation with plant material an effective tool for the restoration of threatened sandy grassland? Applied Vegetation Science 13:425438.
  • Fischer, M., and J. Stöcklin. 1997. Local extinctions of plants in remnants of extensively used calcareous grasslands 1950–1985. Conservation Biology 11:727737.
  • Fransson, A.-M., I. M. vanAarle, P. A. Olsson, and G. Tyler. 2003. Plantago lanceolata L. and Rumex acetosella L. differ in their utilisation of soil phosphorus fractions. Plant and Soil 248:285295.
  • Gärdenfors, U., editor. 2010. The 2010 Red List of Swedish Species. ArtDatabanken, SLU, Uppsala, Sweden.
  • Gazenbeek, A. 2005. LIFE, Natura 2000 and the military. European Commission, Environmental Directorate General, Brussels, Belgium.
  • Gomez-Aparicio, L. 2009. The role of plant interactions in the restoration of degraded ecosystems: a meta-analysis across life-forms and ecosystems. Journal of Ecology 97:12021214.
  • Grime, J. P. 1977. Evidence for the existence of three primary strategies in plants and its relevance to ecological and evolutionary theory. American Naturalist 111:11691194.
  • Grubb, P. J. 1977. The maintenance of species richness in plant communities: the importance of the regeneration niche. Biological Reviews 52:107145.
  • Hobbs, R. J., and L. F. Huenneke. 1992. Disturbance diversity and invasion implications for conservation. Conservation Biology 6:324337.
  • Jentsch, A., S. Friedrich, T. Steinlein, W. Beyschlag, and W. Nezadal. 2009. Assessing conservation actions for substitution of missing dynamics on former military training areas in Central Europe. Restoration Ecology 17:107116.
  • Jones, M. L. M., K. Norman, and P. M. Rhind. 2010. Topsoil inversion as a restoration measure in sand dunes, early results from a UK field-trial. Journal of Coastal Conservation 14:139151.
  • Kiehl, K., and J. Pfadenhauer. 2007. Establishment and persistence of target species in newly created calcareous grasslands on former arable fields. Plant Ecology 189:3148.
  • Leis, S. A., D. M. Engle, D. M. Leslie Jr.. , and J. S. Fehmi. 2005. Effects of short- and long-term disturbance resulting from military maneuvers on vegetation and soils in a mixed prairie area. Environmental Management 36:849861.
  • von Linné, C. 1751. Carl Linnaei skånska resa på höga öfwerhetens befallning förrättad år 1749 (Carl Linnaeus's Journey to Scania as ordered by the high authorities in 1749). Stockholm.
  • Luscombe, G., R. Scott, and D. Young. 2008. Pages 56. Soil inversion works. Landlife National Flower Centre, Court Hey Park, Liverpool, UK.
  • Mårtensson, L.-M., and P. A. Olsson. 2010. Soil chemistry of local vegetation gradients in sandy calcareous grasslands. Plant Ecology 206:127138.
  • Mårtensson, L.-M., T. K. Schnoor, and P. A. Olsson. 2012. Allocation of carbon to mycorrhiza in the grasses Koeleria glauca and Corynephorus canescens in sandy grasslands Applied Soil Ecology 54:5562.
  • Mattiasson, G. 2009. Åtgärdsprogram för sandnörel (Minuartia viscosa) 2008–2014. Report no. 5949. Swedish Environmental Protection Agency, Stockholm, Sweden.
  • Ödman, A. M., L.-M. Mårtensson, C. Sjöholm, and P. A. Olsson. 2011. Immediate responses in soil chemistry, vegetation and ground beetles to soil perturbation. Biodiversity and Conservation 20:30393058.
  • Ödman, A. M., T. K. Schnoor, J. Ripa, and P. A. Olsson. 2012. Soil disturbance as a restoration measure in dry sandy grasslands. Biodiversity and Conservation 21:19211935.
  • Olsson, K.-A. 1994. Sandstäpp i Skåne—ett upprop. Lunds Botaniska Förening Medlemsblad 1994:47.
  • Olsson, P. A., L.-M. Mårtensson, and H. H. Bruun. 2009. Acidification of sandy grasslands—consequences for plant diversity. Applied Vegetation Science 12:350361.
  • Olsson, P. A., T. K. Schnoor, and S.-Å. Hanson. 2010. pH preferences of red-listed fungi in calcareous sandy grasslands: implications for conservation and restoration. Fungal Ecology 3:357365.
  • Poschlod, P., J. P. Bakker, and S. Kahmen. 2005. Changing land use and its impact on biodiversity. Basic and Applied Ecology 6:9398.
  • Rebele, F., and C. Lehmann. 2002. Restoration of a landfill site in Berlin, Germany by spontaneous and directed succession. Restoration Ecology 10:340347.
  • Riksen, M., W. Spaan, and L. Stroosnijder. 2008. How to use wind erosion to restore and maintain the inland drift-sand ecotype in the Netherlands? Journal for Nature Conservation 16:2643.
  • Rodwell, J. S., J. H. J. Schaminée, L. Mucina, S. Pignatti, J. Dring, and D. Moss. 2002. The Diversity of European Vegetation. An overview of phytosociological alliances and their relationships to EUNIS habitats. EC-LNV, Wageningen, Netherlands.
  • Roem, W. J., and F. Berendse. 2000. Soil acidity and nutrient supply ratio as possible factors determining changes in plant species diversity in grassland and heathland communities. Biological Conservation 92:151161.
  • Schläpfer, M., H. Zoller, and C. Körner. 1998. Influences of mowing and grazing on plant species composition in calcareous grassland. Botanica Helvetica 108:5767.
  • Schnoor, T. K., and P. A. Olsson. 2010. Effects of soil disturbance on plant diversity of calcareous grasslands. Agriculture Ecosystems and Environment 139:714719.
  • Sjörs, H. 1967. Nordisk växtgeografi, 2:a upplagan. Svenska Bokförlaget, Bonniers, Stockholm, Sweden.
  • Smits, N. A. C., J. H. Willems, and R. Bobbink. 2008. Long-term effects of fertilisation on the restoration of calcareous grasslands. Applied Vegetation Science 11:279286.
  • Storm, C., and K. Süss. 2008. Are low-productive plant communities responsive to nutrient addition? Evidence from sand pioneer grassland. Journal of Vegetation Science 19:343354.
  • Süss, K., C. Storm, A. Zehm, and A. Schwabe. 2004. Succession in inland sand ecosystems: which factors determine the occurrence of the tall grass species Calamagrostis epigejos (L.) Roth and Stipa capillata L.? Plant Biology 6:465476.
  • Tyler, T. 2003. The state of the sand-steppe vegetation of eastern Scania (Skåne) in early spring 2003. Botaniska Notiser 136:122.
  • Van Til, M., and A. Kooijman. 2007. Rapid improvement of grey dunes after shallow sod cutting. Coastline Reports 7:5360.