SEARCH

SEARCH BY CITATION

Keywords:

  • Festuco-Brometea;
  • Functional connectivity;
  • Habitat fragmentation;
  • Koelerio-Corynephoretea;
  • Post-dispersal fate;
  • Seed dispersal;
  • Seed rain;
  • Sheep grazing;
  • Spatial Analysis by Distance Indices (SADIE);
  • Soil seed bank

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Implications for restoration practice
  8. Acknowledgements
  9. References
  10. Supporting Information

Questions

Does epizoochorous dispersal via sheep lead to the establishment of populations of sandy grassland species on newly created, managed restoration sites on sandy bare soil? Do epizoochorously-induced spatial patterns persist during vegetation development? Does sandy grassland develop, which is rich in epizoochorously-dispersed target species?

Location

Upper Rhine Valley, Germany.

Methods

A 6-yr experiment on epizoochorous dispersal by sheep was conducted on three newly created deep sand deposition sites mimicking restoration areas with reduced nutrient availability. Establishment success and persistence of ten epizoochorously-dispersed species were assessed and spatial patterns were analysed using SADIE (Spatial Analysis by Distance Indices). Vegetation development of the experimental areas was related to a nearby nature reserve (relevés, target species ratios). In addition, seed rain and early-successional soil seed bank were sampled.

Results

All but one species dispersed by sheep became established and persisted during the 6-yr study. After establishment, most perennials did not change or increased in abundance over time, whereas annuals showed various population dynamics. Spatial patterns were aggregated for most study species. Similarity of spatial patterns between consecutive years varied by species, site and year, and was stronger in perennial than in annual species. Patterns of seed dispersal and establishment were positively associated with each other (a subset of three species was tested). Within 6 yrs, the proportion, but not the cover, of target species in the experimental areas reached a level similar to that of a nearby nature reserve; however, many species characteristic of the nature reserve were absent. The species compositions of both seed bank and seed rain were dominated by non-target species.

Conclusions

Sheep flocks may assist in colonization of grassland species on newly created bare soil areas via epizoochory. The incorporation of livestock into restoration projects might facilitate the regeneration and preservation of threatened plant populations. Livestock might be most successful in promoting biodiversity if they are moved from communities with target species to restoration areas.


Nomenclature
Wisskirchen & Haeupler (1998)

for vascular plants

Oberdorfer (2001)

for plant communities

Abbreviations
TSRqual

qualitative target species ratio

TSRquant

quantitative target species ratio

Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Implications for restoration practice
  8. Acknowledgements
  9. References
  10. Supporting Information

Loss and fragmentation of semi-natural grasslands are major threats to populations of rare plant species in the Central European landscape (Fischer & Matthies 1998; Kéry et al. 2000). Due to land-use changes such as agricultural intensification and afforestation, the connectivity between grassland fragments decreased dramatically during the 20th century (Soons et al. 2005; Hooftman & Bullock 2012). Therefore, restoration approaches are required that enlarge existing semi-natural grasslands and develop degraded areas as ‘stepping stones’ to connect existing habitats. Successful restoration of grassland habitats includes three major aspects: abiotic restoration, biotic restoration and follow-up management. Nutrient adjustments to soil can be achieved by either removing or replacing topsoil, depending on the specific site conditions and cost considerations (Hölzel & Otte 2003; Eichberg et al. 2010).

The availability of viable seeds for the restoration of target plant species may be limited by distance to well-developed plant communities and impoverished soil seed banks (Bakker & Berendse 1999). Additionally, the possibility that the regional species pool will act as a seed source for local colonization is limited by the dramatic decrease in the number of formerly widespread roaming livestock herds (‘dispersal infrastructure’) in the European landscape (Ozinga et al. 2009). Effective tools to overcome seed limitations in degraded grasslands include the transfer of seed-containing plant material and sowing of seed mixtures onto bare or disturbed soil (Kiehl et al. 2010; Coiffait-Gombault et al. 2012; Hölzel et al. 2012). However, seed transfer is normally employed only at the beginning of a restoration project and, as yet, little is known regarding the long-term success of anthropogenic seed transfer. Moreover, commercial seed mixtures are often dominated by seeds of graminoids and contain high amounts of perennial generalists, which can establish and spread quickly (Conrad & Tischew 2011). Seed mixtures of local provenance provide better results with regard to target species establishment in restored communities (Mitchley et al. 2012), so that local hay or raked plant material can be used (Stroh et al. 2007; Klimkowska et al. 2010).

Livestock transport a wide range of plant species through epi- and endozoochorous seed dispersal. In this study, we focused on epizoochory, but endozoochory plays an important role as well (e.g. Eichberg et al. 2007). With respect to epizoochory, seeds are dispersed by various livestock species, e.g. equids (Couvreur et al. 2004, 2005), cattle (Couvreur et al. 2004), goats (Shmida & Ellner 1983) and sheep (Fischer et al. 1996; Mouissie et al. 2005; Manzano & Malo 2006; Wessels et al. 2008). Due to their curly and greasy hair, specifically sheep have a very high epizoochorous dispersal potential. In sandy grasslands, roaming sheep flocks can disperse large numbers of seeds of many species (Wessels et al. 2008). Even though seed morphology and seed mass influence the attachment and detachment rates (Tackenberg et al. 2006; Will et al. 2007), epizoochory is possible for most if not all grassland species (Couvreur et al. 2004; Mouissie et al. 2005).

Grazing animals can provide effective long-distance epizoochorous seed transport (Boulanger et al. 2011; Purschke et al. 2012). Boulanger et al. (2011) demonstrated that epizoochory by deer led to a widened spatial distribution of the forest herb, Cynoglossum germanicum, which was rare in a region of France. In Swedish grasslands with a long history of continuous grazing management by livestock, plant species with a propensity for long-distance dispersal via wind and grazing animals were well represented (Purschke et al. 2012).

Domestic livestock species can serve as very effective management tools for increasing species richness, if the livestock are managed carefully (reviewed in Rosenthal et al. 2012). Reintroduction of livestock is an important way to overcome seed limitation and provide connectivity between semi-natural grasslands (Poschlod et al. 1998; Beinlich & Plachter 2010; Auffret et al. 2012). There is much evidence that livestock moving through the landscape can provide an effective source of seeds to certain ecosystems, but the mechanisms and quantities of zoochorous colonization are not well understood (Rosenthal et al. 2012).

Experimental studies on the effectiveness of epizoochorous dispersal by livestock have dealt only with the short-term establishment success of dispersed species (Bugla 2009; Wessels-de Wit & Schwabe 2010). Eichberg et al. (2005) and Wessels-de Wit & Schwabe (2010) studied the spatial distribution of the dispersed seeds, resulting in non-random seed shadows. A better understanding of spatial patterns and their consequences for community development will help to direct habitat management (Nathan & Muller-Landau 2000). Here we present results of a long-term experiment, originally started by Wessels-de Wit & Schwabe (2010). The main objectives of our study were to assess the establishment success and spatial distribution of sheep epizoochorously-dispersed seeds over a period of 6 yrs. The following questions were addressed in our study:

  1. Do epizoochorously-dispersed plant species establish and persist in restored communities?
  2. Do the spatial patterns related to epizoochory persist in restored communities?
  3. What is the structure of the plant community developing on sheep-affected bare soil sites in terms of species composition, vegetation cover, target species ratios and turnover ratio? Does the structure of the restored community after epizoochory by sheep correspond to the target community in nature protection areas?

Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Implications for restoration practice
  8. Acknowledgements
  9. References
  10. Supporting Information

Study site

The experiment in our study was carried out in an ex-arable sand field in the Northern Upper Rhine Valley, Germany (5.5 ha; 8°34′E, 49°50′N; 101 m a.s.l.). This site was used as a field until summer 2004 and, since then, managed by restorative sheep grazing to develop a new habitat for sandy grassland species (Koelerio–Corynephoretea). In autumn 2004, the site was inoculated once with small amounts of mown plant material from a nearby nature reserve. In 2005, three experimental areas were established by depositing nutrient-poor deep sand on this restoration site. Each sand deposition area was 300 m² in extent and raised 0.7 m above the former soil surface. The areas were arranged along a line separated by about 40–50 m. The sand material was from a construction site. The soil conditions were suitable for sand grassland restoration (pH = 7.7, Ntotal = 0.01 mg g-1, phosphate-P = 9.6 mg kg−1; Wessels-de Wit & Schwabe 2010). On each sand deposition area, an individually fenced experimental area was installed, comprising 81 1-m² plots arranged in a 9 m × 9 m grid. Until 2010, fencing protected the experimental areas against uncontrolled grazing by large sheep flocks and rabbits. As follow-up management, grazing by means of a small sheep flock was applied to the experimental areas to control ruderal species. In midsummer 2007–2009, four sheep were present on each experimental area on 1 d yr-1 for ca. 4–12 hrs, depending on the amount of available forage. In 2010, the fences were dismantled, and since then, the experimental areas have been grazed as a unit with their surroundings by a flock of 500–800 sheep during short rotations (about 1 d per grazing section with each section comprising one experimental area).

Experimental design

In October 2005, two Rhoen sheep were present for a single 24-h period on each experimental area. One sheep was prepared with experimentally attached seeds of 14 species typical for inland sand vegetation; the second sheep was included to increase the trampling effect and as a companion. The epizoochorously-dispersed species were: Alyssum montanum subsp. gmelinii, Armeria maritima subsp. elongata, Centaurea stoebe s.l., Cynoglossum officinale, Jasione montana, Koeleria glauca, Medicago minima, Myosotis stricta, Phleum arenarium, Scabiosa canescens, Silene conica, S. otites, Stipa capillata and Tragus racemosus. For characterization of the study species, see Appendix S1. For most of these species, natural epizoochorous dispersal had been documented (Wessels et al. 2008) or is likely (Couvreur et al. 2004; Mouissie et al. 2005). Seeds were collected in sandy habitats in the vicinity of the experimental site (covering an area of about 100 ha); the seeds originated from different sites and, depending on seed production, at least ten individuals. One species, K. glauca, was eliminated from consideration in the present study due to overlapping of the experiment presented here and a trampling experiment conducted simultaneously in 2005, in which additional seeds of this species were introduced into the experimental areas (see Wessels-de Wit & Schwabe 2010). Per body part (shoulder, flank and back on both sides of the sheep) and per plant species, 100 seeds were attached (in total, 600 seeds per species per sheep). Association of seed shadows and emerging seedlings were studied concurrently for three study species (C. officinale, M. minima and S. capillata), which can occur in high densities in sheep wool due to effective seed appendages (Wessels et al. 2008). The seeds of these species are large enough to be detected on the soil surface, and these seeds also were marked with yellow dye. For further details on seed and sheep preparation, see Wessels-de Wit & Schwabe (2010). Post-dispersal seedling emergence and establishment of all introduced species were recorded until 2011.

We did not include plots where sheep were not a part of the experiment. The study site was very open so that a wind-driven seed exchange between experimental plots was likely. In the course of longer-term experiments, it is nearly impossible to exclude such effects and to avoid plot-to-plot contamination (see Lepš et al. 2007; Eichberg et al. 2010). The use of almost seed-free deep sand (see also Eichberg et al. 2010) as a substrate for plant establishment allowed the exclusion of soil as a significant plant source. Seed rain was also sampled, and only data of those epizoochorously-introduced species were analysed, which were not detected in seed traps or were absent in the surrounding vegetation.

Vegetation relevés

Individuals of the introduced species were counted on every 1-m² plot over a 5-yr period (Nov. 2005, May/June 2006–2009). In addition, the percentage cover of the emerging vegetation was estimated on a per-plot basis (2006–2009) and in later surveys, for the entire experimental area (2009–2011), using the following scale: 0.1, 1, 2, 3, 4, 5, 6, 8, 10, 15, 20, 25,… 95, 100%.

To set vegetation composition of the experimental areas in context, five grid-based 80-m² circular patches were sampled in a nearby well-developed Koelerio-Corynephoretea sand ecosystem (nature reserve ‘Griesheimer Düne und Eichwäldchen’; inter-plot distance: 50–120 m; aerial distance to the experimental areas ca. 300 m). Assessment of vegetation cover on these plots used the same cover scale and took place in the same time period as in the experimental areas.

Seed rain

Seed rain was analysed starting in the main fruiting period of the first year of community development (July 2006 to June 2007) using funnel traps (Kollmann & Götze 1998). Each fenced area had ten funnel traps arranged evenly around the experimental area (a 0.5 m buffer zone between the fence and the experimental area was used for trap installation) 0.9 m above ground level. Total sampling area per experimental area was 0.452 m². Traps were emptied fortnightly. Vegetation surrounding the traps was cut regularly at ground level within a radius of ca. 0.5 m to avoid direct seed input into the traps. Trapped seeds were identified and counted; determination was conducted by means of a reference seed collection and literature (see Eichberg et al. 2010).

Soil seed bank

In March 2007, one growing season after the sand was deposited, the soil seed bank of the experimental areas was sampled. Previous studies carried out in our study region have shown that freshly deposited deep sand (gathered from ≥ 1 m depth) was almost free of seeds (Eichberg et al. 2010). Per experimental area, 100 individual soil samples were taken from the outer area edges in regular intervals using an Eijkelkamp liner sampler (diameter 4.7 cm; Giesbeek, NL). The samples were subdivided into an upper (1–6 cm depth) and a lower (11–16 cm) layer. Per experimental area and soil layer, ten composite samples were analysed by mixing ten individual samples, respectively (according to the method used by Eichberg et al. 2006).

A seedling emergence method (Eichberg et al. 2006) was used to assess seed content. Prior to outdoor exposure in the botanical garden of the ‘Technische Universität Darmstadt’, the sand samples were sieved (mesh width: 5 mm), filled into trays and dried at room temperature for 6 wks to eliminate vegetative propagules. Samples were placed on a transparently-roofed platform (0.9 m height) and covered by gauze as a protection against anemochorous seed input. As a control for aerial seed contamination, trays with autoclaved sand were positioned between the samples. The sand was kept moist and turned every second month. From May 2007 to November 2008, emerging seedlings were determined, counted and removed.

Data analysis

Spatial patterns of the introduced species on the experimental areas were studied for the time period 2006–2009 using SADIE (Spatial Analysis by Distance Indices; Perry 1998). For the SADIE approach, we excluded three species (C. stoebe s.l., M. stricta, S. conica) because we detected these species in the seed rain traps; therefore, seeds of these species might have reached the experimental areas from the surrounding grasslands. For all other species, seed dispersal did not come from external sources. These species were either not present in the surrounding vegetation or the seeds were detected on the ground directly after epizoochorous input.

The following variables were assessed to identify spatial patterns: (1) index of aggregation Ia, (2) clustering indices (patch cluster index vi, gap cluster index vj) and (3) spatial association X.

  1. SADIE measures the minimum total distance samples must extend to produce a completely regular arrangement (‘distance to regularity’ D). The index of aggregation is defined as Ia = D/Ea (with Ea: ‘mean distance to regularity’). Aggregated distributions have values of Ia > 1, random distributions of Ia = 1 and regular distributions of Ia < 1. A formal test of spatial randomness of the observed counts among the given sample units is provided as the value Pa. A test for significance at the 5% level indicates a regular arrangement when Pa > 0.975; Pa < 0.025 indicates aggregated distribution (Perry 1998).
  2. To identify areas of clustering, SADIE measures to what degree the subunits contribute towards clustering. Subunits with higher counts than the sample mean are ascribed a patch cluster index vi (positive value); vi > 1.5 indicates patches. The gap cluster index vj is defined similarly, except that subunits with smaller counts than the sample mean are assigned (negative values); values of vj < −1.5 belong to gaps (Perry et al. 1999). The level of significance for patches and gaps is provided by the values Pi, Pj < 0.025, respectively.
  3. To quantify the similarity between the spatial patterns of two consecutive years and the changes of spatial structure over time, the spatial association χ is measured by comparing the clustering indices (Conrad et al. 2006). The mean of these local values χ is the overall spatial association X (Perry & Dixon 2002). Positive association is indicated as overlapping of patches or of gaps in 2 yrs; negative dissociation is indicated by coincidence of opposite forms of spatial pattern. Significance of X is tested through randomizations, with values of the cluster indices reassigned amongst the sample units, after allowance for small-scale spatial autocorrelation (Winder et al. 2001). Spatial autocorrelation in a data set can be detected in SADIE (Dutilleul 1993). A two-tailed randomization test at the 5% level indicates significant association if Pa < 0.025 and significant dissociation if Pa > 0.975.

For analysis of spatial patterns SADIEShell v. 1.22 (IACR, Rothamsted, UK) was used and red-blue plots (given here in black-grey) were visualized using Surfer 8.02 (Golden Software Inc., Golden, CO, US).

Target species ratios (TSR) were calculated for the evaluation of the restoration success according to Eichberg et al. (2010) as:

  • display math

Denominated as target species were those from the classes Festuco-Brometea and Koelerio-Corynephoretea.

The turnover ratio was assessed as the percentage of exchanged plant species between two consecutive years.

Mixed linear models (SAS 9.2, PROC GLIMMIX; SAS Institute Inc., Cary, NC, US; Littell et al. 2006) were applied to calculate (1) the effect of ′year′ on individual numbers of the study species (2005–2009) and (2) the effects of variables ′year′ and ′area′ (experimental areas and nature reserve) as well as their interaction effect on both target species ratios (2006–2011). Individual numbers were log(x + 1)-transformed beforehand, and a hypothetic zero value (Sept. 2005) prior to epizoochorous seed input was included. Also in this case, we excluded the above-mentioned three species (see SADIE).

Mixed linear models are suitable for analysis of repeated-measures data (Littell et al. 1998). According to the corrected Akaike information criterion (AICC), 14 covariance structures were compared (Fernández 2007). If equal AICC values were assessed by two covariance structures, the simpler structure was chosen. Degrees of freedom were calculated using the Kenward–Roger approximation (Schaalje et al. 2002). For estimation of the model parameters, the restricted maximum likelihood method was used. Studentized residuals and conditional studentized residuals were examined for normality using a graphical display (histograms and quantile-residuum plots); a nearly Gaussian distribution could be ascertained. Post-hoc tests with adjustment by simulation tests were carried out to determine differences between years (significance level α = 0.05). In cases with significant interactions, the SLICE option in the LSMEANS statement was used to assess differences between the areas for each year (Schabenberger et al. 2000).

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Implications for restoration practice
  8. Acknowledgements
  9. References
  10. Supporting Information

Establishment and persistence

All species that had been manually attached to sheep established seedlings after epizoochorous dispersal (Table 1). Six years after the experiment had started, nine of ten species persisted on the experimental areas. These species varied in their establishment and persistence rates. Year had a significant effect on the density of all study species except J. montana and T. racemosus. Three hemicryptophyte perennials significantly increased in density until 2009, even though many individuals were immature. The other three hemicryptophytes decreased: S. capillata, S. canescens and J. montana (Table 1). Annuals showed various population dynamics. All study species had low mean cover (≤ 1 %) in 2011.

Table 1. Mean density of individuals per species and per m² (± SE; n = 3) between Sept. 2005 (hypothetic zero-value before the experiment started) and May/June 2009 and cover data for 2009–2011 (per entire experimental area of 81 m²; mean ± SE; n = 3)
     Number of individuals m−2Cover (%)
Speciesdf numdf de F PSept. 2005Nov. 20052006200720082009200920102011
  1. Mixed linear models were calculated for 2005–2009; different letters indicate significant differences between years at α = 0.05 (simulation tests). df num, degrees of freedom numerator, and df de, degrees of freedom denominator. Experimental areas were grazed since 2007 (2007–2009: four sheep, 2010–2011: flock of 500–800 sheep).

Alyssum * gmelinii 59.1740.36<.00010a ± 01.3b ± <.051.6bd ± 0.112.5 bcd ± 6.610.0cd ± 3.912.8c ± 4.40.7 ± 0.30.7 ± 0.60.1 ± <.05
Armeria * elongata 59.49115.340.00030± 00.5b ± 0.20.4b ± 0.10.2b ± 0.12.1b ± 1.71.6b ± 1.20.1 ± <.050.4 ± 0.30.4 ± 0.3
Cynoglossum officinale 59.0692.87<.00010a ± 0<.05a ± <.052.3b ± 0.51.5b ± 0.415.3c ± 5.916.3c ± 6.70.1 ± <.050.1 ± <.050.1 ± <.05
Jasione montana 5120.840.54640 ± 0<.05 ± <.05<.05 ± <.050 ± 00 ± 00 ± 00 ± 00 ± 00 ± 0
Medicago minima 59.0692.87<.00010a ± 00.3b ± 0.10.6b ± 0.10.8b ± 0.26.8bc ± 5.814.1c ± 8.81 ± 14 ± 11 ± 0
Phleum arenarium 51291.96<.00010a ± 00.1b ± 0.11.4c ± 0.136.3d ± 10.129.8d ± 6.318.7d ± 7.70.1 ± <.050.1 ± 0.30.1 ± <.05
Scabiosa canescens 5109.120.00170a ± 00.1b ± <.05<.05a ± <.050a ± 00a ± 0<.05a ± <.05<.05 ± <.05<.05 ± <.05<.05 ± <.05
Silene otites 51212.000.00020a ± 00.3bc ± <.05<.05ab ± <.050.1ab ± <.051.1bc ± 1.03.1c ± 0.20.1 ± <.051.3 ± 0.31.0 ± 0.6
Stipa capillata 59.59126.75<.00010a ± 00 a ± 00.3b ± 0.10.3b ± 0.10.2b ± 0.10.2b ± 0.10.4 ± 0.30.7 ± 0.30.4 ± 0.3
Tragus racemosus 59.6593.310.05240 ± 00 ± 00.1 ± <.050.3 ± 0.22.1 ± 1.50.1 ± <.050.1 ± <.05<.05 ± <.05<.05 ± <.05

One species, A. montanum subsp. gmelinii, spread successfully into the surrounding area after dispersal by sheep. Individuals of this species originally did not occur in the adjacent vegetation.

Seed rain

In total, 41 taxa were detected in the funnel trap samples. Non-target species had a higher species richness and density than seeds of target species (Appendix S2). Accounting for 93% of the total amount of seeds, the neophyte species, Conyza canadensis, was most abundant in the aerial diaspore pool. Target species accounted for 24% of the taxa (ten species; among them four study species; Appendix S2), whereas their proportion of seeds was ca. 0.5%. Vulpia myuros and Corynephorus canescens had the highest abundances among target species. According to life form, the aerial seed rain was dominated by seeds of annuals (54% of taxa, 99% of seeds). A high proportion (78%) of species detected in the seed rain also established as seedlings or adults on the experimental areas and/or in the surrounding vegetation. Allochthonous species were mostly woody species (e.g. Betula pendula).

Soil seed bank

One growing season after deep sand deposition, the seed banks of the three areas were composed of 31 taxa, seven of which were target species of the Koelerio-Corynephoretea (Appendix S3). Mean total densities of seeds reached 1170 ± 217 seeds m−2 (mean ± SE) in the upper soil layer and 409 ± 39 seeds m−2 in the lower layer. Seeds of target species had a total abundance of only 7% (both layers pooled; e.g. Arenaria serpyllifolia and Saxifraga tridactylites). Among the species tested in the epizoochory experiment, only M. minima was detected at low density in the seed bank (two seedlings). The ruderal species, Conyza canadensis and Sisymbrium altissimum, were the most abundant species, together comprising more than 80% of the total seedlings. Annuals comprised 70% of the taxa and 96% of the seedlings. Most species present in the seed banks were also present in the standing vegetation of the experimental areas and/or the surrounding grassland.

Spatial patterns

In nearly all years of the study, most sheep-dispersed species showed overall aggregated spatial patterns on the three experimental areas (Appendix S4). A random spatial arrangement was recorded for six species, depending on the specific area and/or year. Three of these species had low emergence rates (J. montana, S. canescens, T. racemosus).

Clustering indices varied according to species, year and experimental area. Only A. montanum subsp. gmelinii had distinguishable patches and gaps in all experimental areas during the entire study (Fig. 1). All other species changed their aggregation pattern with respect to patches and gaps depending on year and experimental area (e.g. Cofficinale, S. capillata). Silene otites did not form significant patches or gaps on the experimental areas until 2009.

image

Figure 1. Development of spatial patterns of Alyssum montanum subsp. gmelinii recorded in an experiment on epizoochorous seed dispersal by sheep in the period 2006–2009 (‘red-blue plots’ sensu Perry et al. 1999). Patches (vi > 1.5) are indicated by black zones, gaps (vj < −1.5) are coloured in gray. Only significant values are mapped (see ‘Data analysis’). Each quadrat represents an area of 9 m × 9 m.

Download figure to PowerPoint

Spatial patterns of seed shadows and seedling establishment for Cofficinale, M. minima and S. capillata were associated with each other.

Overall spatial association X differed largely between study species (Table 2). Perennial species mostly showed association, except for S. otites (exp. area 3, all years). Positive association was also found for the annual P. arenarium. The only species showing dissociation in spatial patterns was T. racemosus on experimental area 3 between 2006 and 2007.

Table 2. Mean spatial association X between seed patterns (2005) and plant patterns in consecutive years (2006–2009) on the experimental areas (± SE; n = 3; T. racemosus n = 2)Thumbnail image of
  • For Jasione montana and Scabiosa canescens across the whole study period association was not calculable. –, no data for calculation collected. Highlighted in gray are those values showing significant association on all experimental areas. Spatial association was calculated using SADIE (see ′Data analysis′).

  • Community development

    The initial vegetation of the experimental areas was dominated by ruderal species. After the first year of the study, the mean number of ruderal species declined continuously (especially early-successional Stellarietea mediae species); the total number of plant species on the experimental areas also declined after a maximum peak in 2009 (Appendix S5). On the experimental areas, the mean number of target species was approximately stable; their cover increased after 2006. As a consequence, both qualitative and quantitative target species ratios of the experimental areas increased during the study period (Table 3); those of the nature reserve were stable (TSRquant) or declining (TSRqual). Until 2010, the qualitative TSR differed between the two areas; in 2011, mean TSRqual of the experimental areas reached the same value as the nature reserve. The TSRquant was not significantly affected by year and area, even though TSRquant on the experimental areas was still half the TSRquant of the reserve in 2011. Total cover of target species was much higher in the target community (e.g. S. capillata, Ononis repens and Phleum phleoides).

    Table 3. Effects of area and year on qualitative and quantitative target species ratios as tested by mixed linear models
     df numdf de F P 200620072008200920102011
    1. df num, degrees of freedom numerator; df de, degrees of freedom denominator. In the right part of the table mean values of the experimental areas and the nature reserve in each year (± SE) are given. Different letters indicate significant differences between areas sliced by year at α = 0.05 (simulation tests), only if area*year is significant.

    TSR qualexp. area
    area16.8928.020.00120.30a ± 0.020.32a ± 0.040.36a ± 0.020.36a ± 0.010.41a ± 0.020.44a ± 0.01
    year528.770.680.6392nature reserve
    area*year528.778.30<.00010.60b ± 0.040.57b ± 0.020.53b ± 0.040.54b ± 0.030.51b ± 0.020.43a ± 0.02
    TSR quantexp. area
    area17.87175.39<.00010.16 ± 0.030.32 ± 0.030.36 ± 0.030.39 ± 0.090.45 ± 0.090.40 ± 0.11
    year528.22.240.0775nature reserve
    area*year528.21.450.23820.77 ± 0.050.81 ± 0.050.84 ± 0.030.78 ± 0.070.79 ± 0.030.79 ± 0.03

    The similarity of species spectra of experimental areas and nature reserve plots was not very high (Appendix S5): 42% of the 148 recorded species occurred on both sites, and 26% solely in the nature reserve. With regard to target species, 23 species (46%) occurred only in the reserve (in high presence, e.g. Alyssum alyssoides, Euphorbia seguieriana, O. repens, P. phleoides); four species (8%) occurred only on the experimental areas (three of these were study species). Turnover ratios of species on the experimental areas varied between sets of years. Nearly 23% of all species were replaced between 2006 and 2007, whereas the ratio decreased to 10% in 2008/09 and increased again in the last year to 15 %. The nature reserve had a consistently high yearly species turnover of about 22%, with a slight increase in turnover from 2010 to 2011.

    Discussion

    1. Top of page
    2. Abstract
    3. Introduction
    4. Methods
    5. Results
    6. Discussion
    7. Implications for restoration practice
    8. Acknowledgements
    9. References
    10. Supporting Information

    Establishment and persistence of epizoochorously-introduced species

    Traditional pastoralism practices have the potential to disperse grassland species and aid in the colonization of restoration sites via epizoochory. In this study, most species dispersed by sheep established on the experimentally-created bare sand areas and persisted during the 6-yr study. In response to question 1, seeds clearly dispersed and established in relation to sheep grazing. Successful establishment of plant species on bare soil after sheep and goat dispersal was found also by Bugla (2009). This means that sheep can assist grassland restoration by increasing the availability of seeds and encouraging post-dispersal establishment. Information on post-dispersal processes has been revealed to be as important as information on landscape structure to adequately model landscape connectivity (Rico et al. 2012).

    The observed differences in the study species′ establishment rates, recorded as individual numbers, can be explained mainly by differences in life history requirements. Specific environmental requirements determine which species establish successfully in restoration sites. Therefore, future ecological studies should elaborate conditions for the establishment of threatened species. With respect to the first year, differences in detachment rates from sheep wool or in trampling-induced burial effects (Rotundo & Aguiar 2004; Eichberg et al. 2005) were likely important reasons for different establishment rates of the study species. Additionally, soil chemistry was important; species that successfully established were characteristic of base-rich, nutrient-poor sand, whereas species that are typical of acidic (Jmontana) or more fertile soils (T. racemosus) (see Ellenberg et al. 2001) did not perform well.

    Another factor influencing the establishment rate and persistence of introduced species is the successional stage of the restoration site. The first years were characterized by low vegetation cover, favouring most annual study species. Nevertheless, annual species typically show fluctuating abundances across years (e.g. Špačková & Lepš 2004; Winkler et al. 2011); probably this is caused by variability in environmental conditions. Increasing soil consolidation and total vegetation cover may explain why the density of annual species decreased after 2007/2008 (except for M. minima). Annuals are often displaced by perennials during early succession (Alday et al. 2011). Although sheep may facilitate the appearance of annual species through soil disturbance, the experimental grazing of the individual experimental areas with four sheep from 2007–2009 probably did not reach optimal disturbance intensity. Prolonging the grazing period of a small flock cannot fully replace short-term impacts of a typical sheep flock comprising some hundreds of animals.

    In contrast, some perennial study species presumably benefited from the initial consolidation of the experimental areas (e.g. S. canescens and S. capillata). Re-emergence and establishment of S. canescens, a threatened mid-successional species, may be explained by seeds incorporated into the soil during the experiment in 2005, which reached favourable conditions for establishment, e.g. after disturbance. Soil disturbance usually leads to an activation of the soil seed bank (Milton et al. 1997) and has the potential to enhance seedling establishment (Jentsch et al. 2002). Stipa capillata, a typical later-successional species (Allio-Stipetum, Festuco-Brometea), needs longer time spans to establish on restoration sites (Eichberg et al. 2010) than the other study species.

    Spatial patterns of epizoochorously-introduced species

    Several studies have assessed seed dispersal distances in livestock species (e.g. Fischer et al. 1996; Mouissie et al. 2005) but very few have looked at spatial patterns of establishment following livestock epizoochory. As an answer to question 2, we showed that the epizoochorously-induced spatial patterns persisted throughout our study for most study species. One explanation for this phenomenon could be that dispersal distances of the established plant individuals were short (Cousens et al. 2008).

    For woody species, previous studies suggest that seed fall and seedling emergence vary from year to year (Hampe et al. 2008). The spatial similarity between life-history stages is often weak for these species (Schupp & Fuentes 1995). This does not correspond to the findings in our study; the initial seed distribution and subsequent plant establishment in the following year was significantly related for the three study species. Secondary dispersal (by animals or abiotic processes; reviewed in Vander Wall et al. 2005) or seed predation may not have been important in our study.

    Spatial patterns of most introduced species were aggregated the year after epizoochorous dispersal. Initial patterns of aggregation were usually preserved during the following years. Aggregation and association were weaker in annuals than in perennials, reflecting the ephemeral life-history strategy of therophytes. Similarly, the spatial association for adult desert shrubs was stable for at least 20 yrs or more (Miriti 2007). The clumped distribution we observed in our experiment may have been related to the uneven patterns of sheep movement across the areas (Wessels-de Wit & Schwabe 2010); uneven patterns of sheep movement have also been observed in sheep grazing larger areas (Rosenthal et al. 2012).

    The observed clustering could be beneficial for the colonization on bare sandy soil patches, since in harsh environments establishment of some species profits from sheltering effects of vegetation, such as shading (Ryser 1993; Maestre et al. 2003; Padilla & Pugnaire 2006). A better understanding of spatial patterns and their consequences for community development will help to direct habitat management and restoration (Nathan & Muller-Landau 2000; Brooker et al. 2008).

    Development of the entire plant community on the experimental areas

    In response to question 3, the initial vegetation of the newly created experimental areas was dominated by ruderal species, which is also true for many other restoration projects (Jongepierová et al. 2007; Rydgren et al. 2010). As stated above, the deep sand used can presumably be excluded as source of seeds for most of these species. Most of the 31 taxa found in our soil samples probably entered the seed bank during the 18-mo period after sand deposition. One indication that seeds in the seed bank increased during the study was the high degree of similarity between soil seed bank and standing vegetation of the experimental areas and their surroundings. A further indication is the fact that we found a higher density of seeds in the upper than the lower soil layer. However, at least one species growing on the experimental areas, Melilotus albus, likely originated from an extant seed bank, because this species emerged almost exclusively in the sample trays of the lower sand layer and was not found in the vicinity of the study site.

    According to our seed trap study, the dominant source for ruderal species was aerial seed input. In line with other restorations carried out in our study region (Stroh et al. 2002; Eichberg et al. 2010), the seed rain comprised to a major extent non-target species. Moreover, all target and most other species detected in the seed rain could be related to the standing vegetation on the experimental areas and the nearby surroundings. Similar results were obtained by Auffret & Cousins (2011) on former arable fields and Faust et al. (2012) in a Koelerio-Corynephoretea community.

    Many species of grassland ecosystems (including sand species like Corynephorus canescens) have short dispersal distances and lack the ability to colonize quickly (see e.g. Jentsch & Beyschlag 2003 for sandy grassland; Erfanzadeh et al. 2010 for salt marshes). Particularly when there is no source population in the direct vicinity of a restoration site, natural colonization processes via seed rain may be slow (e.g. Verhagen et al. 2001; Buisson et al. 2006). Therefore, differences in species composition between restoration and target sites persist for several years (e.g. Pywell et al. 2002; Conrad & Tischew 2011). We found in our experimental areas that species composition differed from that of nature reserves. Some species that are characteristic of nature reserves were missing on the experimental areas, including A. alyssoides, E. seguieriana, O. repens and P. phleoides.

    We conclude that even though sheep grazing in many cases might not be sufficient as an exclusive restoration measure, especially within short time periods, grazing has the potential to establish new populations of threatened plant species in sandy grasslands, e.g. M. minima and S. capillata.

    Implications for restoration practice

    1. Top of page
    2. Abstract
    3. Introduction
    4. Methods
    5. Results
    6. Discussion
    7. Implications for restoration practice
    8. Acknowledgements
    9. References
    10. Supporting Information

    Techniques that create bare soil patches and implement livestock grazing in restoration areas may be useful to increase richness of target species. Additional seeds of plant species might be introduced through man-made seed transfer to support a specific community structure. This study suggests that restoration goals to re-establish target species can be aided by livestock grazing, if properly managed.

    Acknowledgements

    1. Top of page
    2. Abstract
    3. Introduction
    4. Methods
    5. Results
    6. Discussion
    7. Implications for restoration practice
    8. Acknowledgements
    9. References
    10. Supporting Information

    We thank Saskia Wessels-de Wit (Houten, NL) for supplying data on the first year of the project. Species that potentially escaped from our experiment into the adjacent surroundings were recorded by Nadine Schweda (Langen). Klaus Birkhofer (Lund, SE) provided an introduction to the SADIE programme; assistance in statistical analysis and valuable comments were provided by Christian Storm (Darmstadt). We thank Beth Middleton and three anonymous referees whose comments greatly improved earlier versions of the manuscript, and Ann Thorson (Oxford) for editing the English text. Many thanks to Reiner Stürz (Ober-Ramstadt) for good cooperation concerning the grazing regime. ‘Regierungspräsidium Darmstadt’ gave permission to work in the target area ‘Griesheimer Düne und Eichwäldchen’.

    References

    1. Top of page
    2. Abstract
    3. Introduction
    4. Methods
    5. Results
    6. Discussion
    7. Implications for restoration practice
    8. Acknowledgements
    9. References
    10. Supporting Information
    • Alday, J.G., Pallavicini, Y., Marrs, R.H. & Martínez-Ruiz, C. 2011. Functional groups and dispersal strategies as guides for predicting vegetation dynamics on reclaimed mines. Plant Ecology 212: 17591775.
    • Auffret, A.G. & Cousins, S.A.O. 2011. Past and present management influences the seed bank and seed rain in a rural landscape mosaic. Journal of Applied Ecology 48: 12781285.
    • Auffret, A.G., Schmucki, R., Reimark, J. & Cousins, S.A.O. 2012. Grazing networks provide useful functional connectivity for plants in fragmented systems. Journal of Vegetation Science 23: 970977.
    • Bakker, J.P. & Berendse, F. 1999. Constraints in the restoration of ecological diversity in grassland and heathland communities. Trends in Ecology & Evolution 14: 6368.
    • Beinlich, B. & Plachter, H. 2010. Sheep: a functional corridor system. In: Plachter, H. & Hampicke, U. (eds.) Large-scale Livestock Grazing: A Management Tool for Nature Conservation, pp. 281288. Springer, Berlin, DE.
    • Boulanger, V., Baltzinger, C., Saïd, S., Ballon, P., Ningre, F., Picard, J.-F. & Dupouey, J.-L. 2011. Deer-mediated expansion of a rare plant species. Plant Ecology 212: 307314.
    • Brooker, R.W., Maestre, F.T., Callaway, R.M., Lortie, C.L., Cavieres, L.A., Kunstler, G., Liancourt, P., Tielborger, K., Travis, J.M.J., Anthelme, F., Armas, C., Coll, L., Corcket, E., Delzon, S., Forey, E., Kikvidze, Z., Olofsson, J., Pugnaire, F.I., Quiroz, C.L., Saccone, P., Schiffers, K., Seifan, M., Touzard, B. & Michalet, R. 2008. Facilitation in plant communities: the past, the present, and the future. Journal of Ecology 96: 1834.
    • Bugla, B. 2009. Untersuchung von dynamischen Ausbreitungsprozessen in fragmentierten Sandhabitaten. Dissertationes Botanicae 410. J. Cramer, Berlin, DE.
    • Buisson, E., Dutoit, T., Torre, F., Römermann, C. & Poschlod, P. 2006. The implications of seed rain and seed bank patterns for plant succession at the edges of abandoned fields in Mediterranean landscapes. Agriculture, Ecosystems & Environment 115: 614.
    • Coiffait-Gombault, C., Buisson, E. & Dutoit, T. 2012. Using a two-phase sowing approach in restoration: sowing foundation species to restore, and subordinate species to evaluate restoration success. Applied Vegetation Science 15: 277289.
    • Conrad, M.K. & Tischew, S. 2011. Grassland restoration in practice: do we achieve the targets? A case study from Saxony-Anhalt/Germany. Ecological Engineering 37: 11491157.
    • Conrad, K.F., Perry, J.N., Woiwod, I.P. & Alexander, C.J. 2006. Large-scale temporal changes in spatial pattern during declines of abundance and occupancy in a common moth. Journal of Insect Conservation 10: 5364.
    • Cousens, R., Dytham, C. & Law, R. 2008. Dispersal in Plants – A Population Perspective. Oxford University Press, Oxford, UK.
    • Couvreur, M., Christiaen, B., Verheyen, K. & Hermy, M. 2004. Large herbivores as mobile links between isolated nature reserves through adhesive seed dispersal. Applied Vegetation Science 7: 229236.
    • Couvreur, M., Cosyns, E., Hermy, M. & Hoffmann, M. 2005. Complementarity of epi- and endozoochory of plant seeds by free ranging donkeys. Ecography 28: 3748.
    • Dutilleul, P. 1993. Modifying the t-test for assessing the correlation between 2 spatial processes. Biometrics 49: 305314.
    • Eichberg, C., Storm, C. & Schwabe, A. 2005. Epizoochorous and post-dispersal processes in a rare plant species: Jurinea cyanoides (L.) Rchb. (Asteraceae). Flora 200: 477489.
    • Eichberg, C., Storm, C., Kratochwil, A. & Schwabe, A. 2006. A differentiating method for seed bank analysis: validation and application to successional stages of Koelerio-Corynephoretea inland sand vegetation. Phytocoenologia 36: 161189.
    • Eichberg, C., Storm, C. & Schwabe, A. 2007. Endozoochorous dispersal, seedling emergence and fruiting success in disturbed and undisturbed successional stages of sheep-grazed inland sand ecosystems. Flora 202: 326.
    • Eichberg, C., Storm, C., Stroh, M. & Schwabe, A. 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.
    • Ellenberg, H., Weber, H.E., Düll, R., Wirth, V., Werner, W. & Paulissen, D. 2001. Zeigerwerte von Pflanzen in Mitteleuropa. 3. Aufl., Scripta Geobotanica 18. Erich Goltze KG, Göttingen, DE.
    • Erfanzadeh, R., Garbutt, A., Petillon, J., Maelfait, J.P. & Hoffmann, M. 2010. Factors affecting the success of early salt-marsh colonizers: seed availability rather than site suitability and dispersal traits. Plant Ecology 206: 335347.
    • Faust, C., Storm, C. & Schwabe, A. 2012. Shifts in plant community structure of a threatened sandy grassland over a 9-yr period under experimentally induced nutrient regimes: is there a lag phase? Journal of Vegetation Science 23: 372386.
    • Fernández, G. 2007. Model selection in PROC MIXED – A user-friendly SAS® macro application. Proceedings of 2007 SAS Global Forum April 16-18 Orlando, FL, US, paper 191-2007.
    • Fischer, M. & Matthies, D. 1998. Effects of population size on performance in the rare plant Gentianella germanica. Journal of Ecology 86: 195204.
    • Fischer, S.F., Poschlod, P. & Beinlich, B. 1996. Experimental studies on the dispersal of plants and animals on sheep in calcareous grasslands. Journal of Applied Ecology 33: 12061222.
    • Hampe, A., García-Castaño, J.L., Schupp, E.W. & Jordano, P. 2008. Spatio-temporal dynamics and local hotspots of initial recruitment in vertebrate-dispersed trees. Journal of Ecology 96: 668678.
    • Hölzel, N. & Otte, N. 2003. Restoration of a species-rich flood meadow by topsoil removal and diaspore transfer with plant material. Applied Vegetation Science 6: 131140.
    • Hölzel, N., Buisson, E. & Dutoit, T. 2012. Species introduction – a major topic in vegetation restoration. Applied Vegetation Science 15: 161165.
    • Hooftman, D.A.P. & Bullock, J.M. 2012. Mapping to inform conservation: a case study of changes in semi-natural habitats and their connectivity over 70 years. Biological Conservation 145: 3038.
    • Jentsch, A. & Beyschlag, W. 2003. Vegetation ecology of dry acidic grasslands in the lowland area of Central Europe. Flora 198: 325.
    • Jentsch, A., Friedrich, S., Beyschlag, W. & Nezadal, W. 2002. Significance of ant and rabbit disturbances for seedling establishment in dry acidic grasslands dominated by Corynephorus canescens. Phytocoenologia 32: 553580.
    • Jongepierová, I., Mitchley, J. & Tzanopoulos, J. 2007. A field experiment to recreate species-rich hay meadows using regional seed mixtures. Biological Conservation 139: 297305.
    • Kéry, M., Matthies, D. & Spillmann, H.H. 2000. Reduced fecundity and offspring performance in small populations of the declining grassland plants Primula veris and Gentiana lutea. Journal of Ecology 88: 1730.
    • 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: 285299.
    • Klimkowska, A., Kotowski, W., Van Diggelen, R., Grootjans, A.P., Dzierża, P. & Brzezińska, K. 2010. Vegetation re-development after fen meadow restoration by topsoil removal and hay transfer. Restoration Ecology 18: 924933.
    • Kollmann, J. & Götze, D. 1998. Notes on seed traps in terrestrial plant communities. Flora 193: 3140.
    • Lepš, J., Doležal, J., Bezemer, T.M., Brown, V.K., Hedlund, K., Igual, A.M., Jorgensen, H.B., Lawson, C.S., Mortimer, S.R., Peix Geldart, A., Rodriguez 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.
    • Littell, R.C., Henry, P.R. & Ammermann, C.B. 1998. Statistical analyses of repeated measures data using SAS procedures. Journal of Animal Science 76: 12161231.
    • Littell, R.C., Miliken, G.A., Stroup, W.W., Wolfinger, R.D. & Schabenberger, O. 2006. SAS® for mixed models, 2nd edn. SAS Institute, Cary, NC, US.
    • Maestre, F.T., Bautista, S. & Cortina, J. 2003. Positive, negative, and net effects in grass–shrub interactions in mediterranean semiarid grasslands. Ecology 84: 31863197.
    • Manzano, P. & Malo, J.E. 2006. Extreme long-distance seed dispersal via sheep. Frontiers in Ecology and the Environment 4: 244248.
    • Milton, S.J., Dean, W.R.J. & Klotz, S. 1997. Effects of small-scale animal disturbances on plant assemblages of set-aside land in Central Germany. Journal of Vegetation Science 8: 4554.
    • Miriti, M.N. 2007. Twenty years of changes in spatial association and community structure among desert perennials. Ecology 88: 11771190.
    • Mitchley, J., Jongepierová, I. & Fajmon, K. 2012. Regional seed mixtures for the re-creation of species-rich meadows in the White Carpathian Mountains: results of a 10-yr experiment. Applied Vegetation Science 15: 253263.
    • Mouissie, A.M., Lengkeek, W. & Van Diggelen, R. 2005. Estimating adhesive seed-dispersal distances: field experiments and correlated random walks. Functional Ecology 19: 478486.
    • Nathan, R. & Muller-Landau, H.C. 2000. Spatial patterns of seed dispersal, their determinants and consequences for recruitment. Trends in Ecology & Evolution 15: 278285.
    • Oberdorfer, E. 2001. Pflanzensoziologische Exkursionsflora für Deutschland und angrenzende Gebiete. 8. Aufl., Ulmer, Stuttgart, DE.
    • Ozinga, W.A., Romermann, C., Bekker, R.M., Prinzing, A., Tamis, W.L., Schaminee, J.H., Hennekens, S.M., Thompson, K., Poschlod, P., Kleyer, M., Bakker, J.P. & van Groenendael, J.M. 2009. Dispersal failure contributes to plant losses in NW Europe. Ecology Letters 12: 6674.
    • Padilla, F.M. & Pugnaire, F.I. 2006. The role of nurse plants in the restoration of degraded environments. Frontiers in Ecology and the Environment 4: 196202.
    • Perry, J.N. 1998. Measures of spatial pattern for counts. Ecology 79: 10081017.
    • Perry, J.N. & Dixon, P.M. 2002. A new method to measure spatial association for ecological count data. Ecoscience 9: 133141.
    • Perry, J.N., Winder, L., Holland, J.M. & Alston, R.D. 1999. Red-blue plots for detecting clusters in count data. Ecology Letters 2: 106113.
    • Poschlod, P., Kiefer, S., Tränkle, U., Fischer, S. & Bonn, S. 1998. Plant species richness in calcareous grasslands as affected by dispersability in space and time. Applied Vegetation Science 1: 7590.
    • Purschke, O., Sykes, M.T., Reitalu, T., Poschlod, P. & Prentice, H.C. 2012. Linking landscape history and dispersal traits in grassland plant communities. Oecologia 168: 773783.
    • 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.
    • Rico, Y., Boehmer, H.J. & Wagner, H.H. 2012. Determinants of actual functional connectivity for calcareous grassland communities linked by rotational sheep grazing. Landscape Ecology 27: 199209.
    • Rosenthal, G., Schrautzer, J. & Eichberg, C. 2012. Low-intensity grazing with domestic herbivores: a tool for maintaining and restoring plant diversity in temperate Europe. Tuexenia 32: 167205.
    • Rotundo, J.L. & Aguiar, M.R. 2004. Vertical seed distribution in the soil constrains regeneration of Bromus pictus in a Patagonian steppe. Journal of Vegetation Science 15: 515522.
    • Rydgren, K., Jørn-Frode, N., Ingvild, A., Inger, A. & Einar, H. 2010. Recreating semi-natural grasslands: a comparison of four methods. Ecological Engineering 36: 16721679.
    • Ryser, P. 1993. Influences of neighboring plants on seedling establishment in limestone grassland. Journal of Vegetation Science 4: 195202.
    • Schaalje, G.B., McBride, J.B. & Fellingham, G.W. 2002. Adequacy of approximations to distributions of test statistics in complex mixed linear models. Journal of Agricultural Biological and Environmental Statistics 7: 512524.
    • Schabenberger, O., Gregoire, T.G. & Kong, F.Z. 2000. Collections of simple effects and their relationship to main effects and interactions in factorials. American Statistician 54: 210214.
    • Schupp, E.W. & Fuentes, M. 1995. Spatial patterns of seed dispersal and the unification of plant population ecology. Ecoscience 2: 267275.
    • Shmida, A. & Ellner, S. 1983. Seed Dispersal on Pastoral Grazers in Open Mediterranean Chaparral, Israel. Israel Journal of Botany 32: 147159.
    • Soons, M.B., Messelink, J.H., Jongejans, E. & Heil, G.W. 2005. Habitat fragmentation reduces grassland connectivity for both short-distance and long-distance wind-dispersed forbs. Journal of Ecology 93: 12141225.
    • Špačková, I. & Lepš, J. 2004. Variability of seedling recruitment under dominant, moss, and litter removal over four years. Folia Geobotanica 39: 4155.
    • Stroh, M., Storm, C., Zehm, A. & Schwabe, A. 2002. Restorative grazing as a tool for directed succession with diaspore inoculation: the model of sand ecosystems. Phytocoenologia 32: 595625.
    • Stroh, M., Storm, C. & Schwabe, A. 2007. Untersuchungen zur Restitution von Sandtrockenrasen: das Seeheim-Jugenheim-Experiment in Südhessen (1999–2005). Tuexenia 27: 287306.
    • Tackenberg, O., Römermann, C., Thompson, K. & Poschlod, P. 2006. What does diaspore morphology tell us about external animal dispersal? Evidence from standardized experiments measuring seed retention on animal-coats. Basic and Applied Ecology 7: 4558.
    • Vander Wall, S.B., Kuhn, K.M. & Beck, M.J. 2005. Seed removal, seed predation, and secondary dispersal. Ecology 86: 801806.
    • Verhagen, R., Klooker, J., Bakker, J.P. & van Diggelen, R. 2001. Restoration success of low-production plant communities on former agricultural soils after top-soil removal. Applied Vegetation Science 4: 7582.
    • Wessels, S., Eichberg, C., Storm, C. & Schwabe, A. 2008. Do plant-community-based grazing regimes lead to epizoochorous dispersal of high proportions of target species? Flora 203: 304326.
    • Wessels-de Wit, S. & Schwabe, A. 2010. The fate of sheep-dispersed seeds: plant species emergence and spatial patterns. Flora 205: 656665.
    • Will, H., Maussner, S. & Tackenberg, O. 2007. Experimental studies of diaspore attachment to animal coats: predicting epizoochorous dispersal potential. Oecologia 153: 331339.
    • Winder, L., Alexander, C.J., Holland, J.M., Woolley, C. & Perry, J.N. 2001. Modelling the dynamic spatio-temporal response of predators to transient prey patches in the field. Ecology Letters 4: 568576.
    • Winkler, E., Dienst, M. & Peintinger, M. 2011. Markov simulation model: flooding, competition, and the fate of the endemic plant Myosotis rehsteineri. Basic and Applied Ecology 12: 620628.
    • Wisskirchen, R. & Haeupler, H. 1998. Standardliste der Farn- und Blütenpflanzen Deutschlands. Ulmer, Stuttgart, DE.

    Supporting Information

    1. Top of page
    2. Abstract
    3. Introduction
    4. Methods
    5. Results
    6. Discussion
    7. Implications for restoration practice
    8. Acknowledgements
    9. References
    10. Supporting Information
    FilenameFormatSizeDescription
    avsc12052-sup-0001-AppendixS1.pdfapplication/PDF198KAppendix S1. Plant species investigated by Wessels-de Wit & Schwabe (2010) and in this study.
    avsc12052-sup-0002-AppendixS1.txtplain text document2K 
    avsc12052-sup-0003-AppendixS2.pdfapplication/PDF197KAppendix S2. Seed rain data, collected via funnel traps from July 2006 to June 2007.
    avsc12052-sup-0004-AppendixS2.txtplain text document1K 
    avsc12052-sup-0005-AppendixS3.pdfapplication/PDF240KAppendix S3. Seed bank data of deep sand, sampled 1.5 yrs after deposition in March 2007.
    avsc12052-sup-0006-AppendixS3.txtplain text document2K 
    avsc12052-sup-0007-AppendixS4.pdfapplication/PDF337KAppendix S4. Spatial patterns of established plants after epizoochorous seed dispersal between 2006 and 2009.
    avsc12052-sup-0008-AppendixS4.txtplain text document3K 
    avsc12052-sup-0009-AppendixS5.pdfapplication/PDF448KAppendix S5. Presence table for all species growing on the experimental areas (n = 3) and the nearby nature reserve (n = 5) in the years 2006 to 2011.
    avsc12052-sup-0010-AppendixS5.txtplain text document9K 

    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.