Accepted Manuscript

Title: Role of substrate and landscape context in earlysuccession: an experimental approach

Author: Karel Prach Petr Pysek Klara Rehounkova

PII: S1433-8319(14)00046-8DOI: http://dx.doi.org/doi:10.1016/j.ppees.2014.05.002Reference: PPEES 25230

To appear in:

Received date: 2-10-2013Revised date: 28-4-2014Accepted date: 5-5-2014

Please cite this article as: Prach, K., Pysek, P., Rehounkova, K.,Roleof substrate and landscape context in early succession: an experimentalapproach, Perspectives in Plant Ecology, Evolution and Systematics (2014),http://dx.doi.org/10.1016/j.ppees.2014.05.002

This is a PDF file of an unedited manuscript that has been accepted for publication.As a service to our customers we are providing this early version of the manuscript.The manuscript will undergo copyediting, typesetting, and review of the resulting proofbefore it is published in its final form. Please note that during the production processerrors may be discovered which could affect the content, and all legal disclaimers thatapply to the journal pertain.

Page 1 of 25

Accep

ted

Man

uscr

ipt

1

Role of substrate and landscape context in early succession: an experimental

approach

Karel Prach1,2*, Petr Pyšek3,4, Klára Řehounková1,2

5

1Department of Botany, Faculty of Science, University of South Bohemia, Branišovská 31, CZ-370 05 České

Budějovice, Czech Republic

2Institute of Botany, Academy of Sciences of the Czech Republic, Dukelská 143, CZ-37982 Třeboň, Czech Republic

3Institute of Botany, Academy of Sciences of the Czech Republic, CZ-252 43 Průhonice, Czech Republic

4Department of Ecology, Faculty of Science, Charles University in Prague, Viničná 7, CZ- 128 44 Prague, Czech 10

Republic

* Corresponding author. E-mail address: prach@prf.jcu.cz

Abstract 15

Both local site conditions and landscape context influence the course of succession, but there is a

lack of experimental studies on the relative importance of these two factors. It is hypothesized that

convergence vs. divergence in succession is determined by the interplay of site factors, such as type

of substrate and the nature of the surrounding landscape. In order to evaluate the role of substrate

and surrounding landscape in the initial development of vegetation, experimental plots with tertiary 20

clay, sand, peat, sterilized local soil and undisturbed local soil as a control were established in two

contrasting regions, and the cover of all the species present was recorded annually for 10 years. In

early succession, vegetation was affected by both the substrate and surrounding landscape, but their

effects resulted in different trends. The importance of the substrate gradually decreased, while that

of the landscape context increased. In the course of succession the vegetation between the two 25

regions diverged and converged within each region. We concluded with regards to the divergence

vs. convergence dichotomy in succession: If contrasting habitats occur in the same or similar

Page 2 of 25

Accep

ted

Man

uscr

ipt

2

landscapes, convergence is expected, whereas if similar or the same habitats are located in

contrasting landscapes, divergence is expected. For the remaining combinations, i.e. contrasting

habitats in contrasting landscapes or the same habitats in the same or a similar landscape, 30

successions may exhibit no or only slight divergence or convergence.

Key-words: Central Europe; Convergence vs divergence; Landscape; Ordination; Substrate

manipulation; Vegetation succession

35

Introduction

The successional development of vegetation is determined by the available pool of species,

substrate quality, biotic interactions, disturbance regime and climatic conditions (Walker and del 40

Moral, 2003). Species that are available and establish at a given site (community species pool) are

determined by the local species pool, which largely depends on regional climate and the history of

landscape management in the region (Setelle et al., 1996). These external factors constitute the

‘landscape context’ in which succession proceeds at a particular locality. Biotic interactions in the

initial stages of succession are usually of much lesser importance than in the later stages, especially 45

in primary successions starting on bare ground (Callaway and Walker, 1997). In this study, there

were no additional disturbances at the plots under concern. Thus, only two basic groups of

environmental factors, i.e. substratum quality and landscape context, were considered.

The important influence of substrate quality on the course of succession was appreciated even in

the first studies on succession (Clements, 1916; see Walker and del Moral, 2003 for other 50

references). Many studies have investigated the influence of various soil factors especially nutrient

content (Tilman, 1988; van der Putten et al. 2013), soil moisture (Morecroft et al., 2004), pH (Prach

Page 3 of 25

Accep

ted

Man

uscr

ipt

3

et al., 2007a) and soil texture (Ejrnaes et al., 2003) on the course of succession. Some studies

experimentally manipulated these soil factors (Mitchley et al., 1996).

The role of landscape, especially the surrounding vegetation being a source of propagules, is also 55

well studied, and the importance of adjacent vegetation and land cover in the wider surroundings on

the course of succession documented (Rydin and Borgegård, 1991; Roche et al., 1998; del Moral et

al., 2005; Dovčiak et al., 2005; Benjamin et al., 2005; Novák and Konvička, 2006; Řehounková and

Prach, 2008). In some cases the surrounding landscape has a more important role than local site

conditions in the course of succession (Salonen and Setälä, 1992) or even than successional age 60

(Řehounková and Prach, 2006). The surrounding vegetation determines ecological succession via

the local species pool (Zobel et al., 1998) and especially the early stages of primary succession are

often “donor controlled”, with species composition closely depending on the pool of species

available in the close surroundings (Wood and del Moral, 1987).

Macroclimate is another important landscape factor driving succession (Otto et al., 2006; Prach et 65

al., 2007a) as it can directly affect species establishment and have an indirect effect as it determines

the regional species pool (Settele et al., 1996). Dispersal and establishment are the main factors that

restrict the colonization of recently exposed habitats (Jones and del Moral, 2009). Dispersal is

associated with the local species pool, while whether a species becomes established or not is related

to abiotic site conditions, such as the character of the substrate and microclimate, and 70

competition/facilitation.

Quantification of the role of particular factors driving succession has both theoretical and practical

implications. The former may improve the understanding of succession, the latter in helping restore

vegetation at disturbed sites and indicate the ways in which certain factors may be manipulated in

order to direct the succession in a desired direction (Walker et al., 2007). 75

How is the nature of the substrate and landscape related to convergence or divergence during

succession? Answering this question may substantially help predict the course of succession in

various environments (del Moral, 2007; Walker et al., 2010). Early studies simply expected

Page 4 of 25

Accep

ted

Man

uscr

ipt

4

convergence towards a single climax community (Clements, 1916), but this was soon contradicted

and more diverse successions and endpoints suggested (see Walker and del Moral, 2003 for 80

references). It seems that the resulting trends in succession, i.e. divergence or convergence, are

largely determined by the initial (dis)similarity in local site conditions and how they change over

time, and by the space-temporal scale of a study (Lepš and Rejmánek, 1991; del Moral, 2007).

Divergence or convergence in succession is usually quantified by means of similarity indices or

multivariate methods based on species composition (Philippi et al., 1998). 85

In contrast to the many experimental studies on the influence of substrate quality on the course of

succession, there are only a few sites experimentally created in order to determine the role of

landscape in driving succession. They include reciprocally transplanting peat between two adjacent

peatlands differing in substrate quality (Salonen, 1990; Salonen and Setälä, 1992) or exposing small

boxes of the same sort of soil at two adjacent sites, which differ in surrounding vegetation, and 90

observing the course of succession in relation to the composition of the nearby vegetation (Lanta

and Lepš, 2009). To obtain a broader perspective of the role of substrate quality and landscape

context, we conducted an experiment using five contrasting types of substrate exposed for 10 years

at two contrasting localities, one in a relatively dry and warm region and the other in a cold and wet

region. This made it possible to ask the following questions: (i) To what extent is the course of 95

succession influenced by substrate quality and landscape context; (ii) How does the importance of

these driving factors change in the course of succession; and (iii) Is succession divergent or

convergent on the different substrates and between the two localities?

100

Methods

Site description, experimental design and data recording

Page 5 of 25

Accep

ted

Man

uscr

ipt

5

The experiment was established in spring 2002 at two climatically different localities (hereafter

called Locality), in the Czech Republic, central Europe: 105

1. A just abandoned part of an arable field (total size ca 0.3 km2) near the village of Vroutek,

located in a rather warm and dry lowland area (hereafter referred as Lowland); altitude 355 m a.s.l.;

latitude 50°11'44"N; longitude 13°21'24"E; average annual temperature 8.6 ºC; average annual

precipitation 461 mm (long-term data from nearby meteorological stations at Blšany and Kryry;

www.chmi.cz). This site is surrounded mostly by ruderal and weedy vegetation on and along arable 110

fields, by strips of mesic grassland dominated by Arrhenatherum elatius, scrubland along paths, and

semi-natural oak-hornbeam woodland about 30 m distant from the study plots.

2. A part of an arable field (ca 0.15 km2) abandoned shortly before the start of the experiment,

located near the village of Benešov, located in a relatively cold and wet upland area (hereafter

referred as Upland); altitude 665 m a.s.l.; latitude 49°19'51"N; longitude 15°00'13"E; average 115

annual temperature 6.7ºC; average annual precipitation 759 mm (long-term data from a nearby

meteorological station at Černovice; www.chmi.cz). This site is surrounded by regularly mown

meadow dominated by Phleum pratense, Festuca pratensis and F. rubra, and by arable land with

common weeds in the distance up to 30 m; the distance to the nearest forest (a Norway spruce

plantation) is ~100 m. 120

The following substrates (hereafter called Substrate) were used to establish experimental plots at

each locality: (i) Tertiary clay from the overburden of brown-coal (hereafter referred as Clay); (ii)

sand from an active sand pit (Sand); and (iii) peat from peat diggings (Peat). In addition, (iv) local

soil was excavated, placed in an oven at 110oC to kill plant propagules and then returned to the site

(Topsoil), and (v) untouched local soil used as a control (Control). The sterilization treatment was 125

not needed in the case of allochtonous substrates (Clay, Sand, Peat) because they were excavated

from the depth below the surface (clay ~100m, sand several meters, peat ~2 m). The three substrates

represented different seres, which are described in detail elsewhere, i.e. spoil heaps resulting from

brown-coal mining (Prach, 1987; Hodačová and Prach, 2003), sand pits (Řehounková and Prach,

Page 6 of 25

Accep

ted

Man

uscr

ipt

6

2006, 2008, 2010) and peat diggings (Konvalinková and Prach, 2010). The plots with local soil 130

represent the abandoned fields described by Prach et al. (2007b) and Jírová at al. (2011).

All substrates were put in beds, 1.5 × 1.5 m in area and 0.3 m deep, dug into the local soil. Five

replicates were arranged in a Latin-square design, resulting in 25 plots at each site. The beds

containing the various substrates, except the controls, were isolated from the surrounding soil by

plastic foil perforated at the bottom to prevent vegetative expansion of clonal species in 135

underground. The controls were left without plastic foil because they were identical with the

surroundings. Strips 0.5 m in width around each plot, except controls, were sprayed annually in

May with Glyphosate to preclude vegetative colonization of the experimental plots by species from

the surrounding vegetation especially by surface stolones. Substrate chemistry, summarized in

Table 1, was assessed at the start and the end of the experiment using standard methods (Sparks et 140

al., 1996). A mixed sample was taken from each substrate just before transportation to the localities.

In the established experimental plots, a mixed sample consisting of five replicates was taken from

each of the plots from the 5 cm layer below a thin surface layer that was removed before the

sampling.

The central 1 m2 of each plot was sampled annually in July or August 2002–2011, at the time of 145

maximum development of the vegetation. All vascular plants were identified and their percentage

cover visually estimated (Kent and Coker, 1992). Nomenclature follows Flora Europaea (http://rbg-

web2.rbge.org.uk/FE/fe.html ).

Data analyses 150

The species cover data were processed using CANOCO version 4.5 with the ordination methods

Detrended Correspondence Analysis (DCA) and Canonical Correspondence Analysis (CCA) (ter

Braak and Šmilauer, 2002). The length of the gradient in DCA was 6.6 SD, thus the use of

unimodal methods was justified (Lepš and Šmilauer, 2003). In the DCA analysis, detrending by

segments was used and species with a weight of at least 3% are displayed in the ordination diagram 155

Page 7 of 25

Accep

ted

Man

uscr

ipt

7

(Fig. 1b). In the CCA analyses, the inter-sample distance and Hill scaling were applied. The use of

the Monte-Carlo permutation test (999 permutations) reflected the sequence of sampling the plots:

Data from 10 subsequent years in each plot were considered to form a whole “plot” and then a split-

plot design was applied. Within the CCA analyses, combining the factors and covariables following

the Monte Carlo test, allowed for testing partial effects of both Locality and Substrate in each year 160

separately. Marginal effects in the CCA were also calculated and tested for significance using the

Monte Carlo test. The marginal effects of environmental factors denoted the variability explained

by given environmental variables without considering other environmental factors, whereas partial

effects denoted the variability explained by given environmental variable with the other

environmental factors as covariables (ter Braak and Šmilauer, 2002). 165

Results

The greatest changes in substrate chemistry were recorded for allochtonous substrates where C 170

and N mostly increased, other trends were less clear. Variation coefficients of average values

among substrates decreased within each of the two localities from the beginning to the end of the

experiment (Lowland: from 232.1 to 213.9; Upland: from 194.9 to 184.9), which may indicate a

trend of increasing uniformity among substrates, but the differences were not statistically significant

(t-test). This suggests that the temporal changes in chemical soil characteristics in the course of 175

succession did not principally affect the differences in vegetation development among individual

substrates.

In the first year of succession, vegetation growing on the same substrate was similar at both

localities, exception for that on Topsoil, which differed between the localities. Later on, vegetation

on the different substrates became more similar at each locality, but between localities it became 180

increasingly dissimilar, indicating a convergence within and divergence between localities (Fig. 1a).

Page 8 of 25

Accep

ted

Man

uscr

ipt

8

The DCA ordination of samples is complemented by the ordination of species that best fit the model

(Fig. 1b). Increasing cover of species typical of meadows, such as Phleum pratense, Festuca

pratensis and F. rubra in all Upland plots was responsible for the convergence at this locality and

for their divergence from the Lowland plots. In Lowland plots, annual weedy species were 185

succeeded by perennial weeds and ruderal species, such as Elytrigia repens and Agrostis gigantea,

and later on by Arrhenatherum elatius, which were responsible for the convergence in succession in

the plots with the various substrates at this locality and for their divergence from the Upland plots

(Fig. 1b).

Partial and marginal effects of Locality steadily increased during the 10 years of the succession, 190

while those of Substrate decreased (Fig. 2). All these effects were significant (Table 1). The CCA

ordination of all the plots revealed significant summarized effects of Time, i.e. age of succession

(accounted for 15.7% of the variability), Locality (17.2%) and Substrate (5.3%).

195

Discussion

Role of substrate

Since the studies of Clements many others have shown that the chemical and physical properties

of the substrate determine the rate and direction of succession (Glenn-Lewin et al., 1992). 200

Vegetation-soil feedback loops are expected to operate especially during primary successions in

extreme habitats (Walker and del Moral, 2003; Laliberté et al., 2013; van der Putten et al., 2013). In

this study, the particular chemical characteristics among the substrates largely differed (see Table 1)

and their summarized effects on seral vegetation, expressed as the substrate types, was evident

especially at the beginning of the experiment (Fig. 2). Unfortunately, we cannot measure substratum 205

moisture which could contribute to differences in vegetation (Morecroft et al., 2004). The

differences in chemistry between substrates are expected to gradually decrease (Laliberté et al.,

Page 9 of 25

Accep

ted

Man

uscr

ipt

9

2013), although the differences between the start and the end of our experiment were not

statistically significant. The given time frame was not probably long enough to better demonstrate

the increasing similarity among substrates. The increasing uniformity among substrates within each 210

locality could be generally explained by the influence of three main factors: climate (leaching by

rainfall), mixing to some extent with the surrounding autochtonous soil (in the case of small

experimental plots it is unavoidable due to the effect of wind and rainfall), and amelioration effects

of plants and other successional organisms. All these factors probably contributed to the decrease of

the role of substrate in the course of succession. In our previous study (Prach et al. 2007a) we 215

demonstrated that only pH significantly influenced the course of succession across various

substrates. This, together with rather inconsistent trends found in the present study and especially

the lack of data between the start and the end of our experiment, was reason for not analysing

substrate chemical data as explanatory variables of vegetation changes.

The trajectories of vegetation changes on two substrates (Peat, Topsoil) differed slightly from the 220

overall pattern. Vegetation on the peat substrate at both localities differed from that on other

substrates in the first years of the experiment, as the peat was colonized by species typical of

disturbed peaty soils (Konvalinková and Prach, 2010) (Fig. 1a,b). Because some of these species

were not present in the surrounding vegetation it is likely that the peat was slightly contaminated

with their propagules during extraction at the original locality. But later on, the peat plots converged 225

towards plots on other substrates. Vegetation on the Topsoil plots at the Upland locality also

differed from that on other substrates because Taraxacum officinale colonized and dominated

Topsoil plots immediately after the experiment started. This would appear to be a priority effect in

which the first, often random arrival may monopolize the space and at least temporarily deflect

succession (Samuels and Darke, 1997). However, the vegetation on Topsoil plots also gradually 230

converged to that on the other plots at this locality. That changes in substratum chemistry could

have played some role in vegetation development cannot be excluded but the nature of these effects

cannot be clarified without information available for the period between the start and the end of the

Page 10 of 25

Accep

ted

Man

uscr

ipt

10

experiment. The experimental substrates we used differed in their chemistry, but none of them was

really extreme (Table 1; Prach et al. 2007a). 235

Our results suggest that substrates, despite the great differences among them, were generally less

important in determining the course of succession than landscape context and age of the succession.

The lower importance of substrate compared to landscape context is reported for systems largely

differing in the landscape context (del Moral et al., 2005).

240

Role of landscape context

Although landscape context has long been thought to be important in determining succession it

has only relatively recently been quantitatively evaluated (Rydin and Borgegård, 1991; Roche et al.,

1998; del Moral and Ellis, 2004; Benjamin et al., 2005; Řehounková and Prach, 2006; Kirmer et al.,

2008; Lanta and Lepš, 2009; del Moral et al. 2010). Most of the studies on the role of surrounding 245

vegetation on the course of succession are observational and usually based on the space-for-time

substitution approach, which may limit some generalizations (Johnson and Miyanishi, 2008; Walker

et al., 2010). A review of studies on succession indicates that the surrounding vegetation had a

significant effect on the course of succession in each study that addressed its effect (Prach and

Řehounková, 2006). 250

Some quantitative studies indicate that landscape factors are more important in determining

successions than substrate characteristics (Salonen and Setälä, 1992; Řehounková and Prach, 2006,

2008). Based on their study on Mt St Helens, del Moral et al. (2005) conclude that “plant succession

is determined as much by chance factors and landscape context as by characteristics of the site

itself” and that “interactions between site amelioration and proximity to colonists affect the arrival 255

sequence”; our results seem to be in accordance with this. The proximity of colonists certainly

affected succession also in our case: all dominants of later stages of the experiment dominated also

in the close proximity, i.e., Arrhenatherum elatius in Lowland plots, and Phleum pratense and

Festuca pratensis in Upland plots. The experimental plots were rather small because of technical

Page 11 of 25

Accep

ted

Man

uscr

ipt

11

limitations. Thus both, amelioration of the substrates and colonization from the surroundings are 260

expected to be easier and faster than at the extensive original sites from which the substrates came,

i.e. large spoil heaps, sand pits and abandoned peat diggings (Prach, 1987; Řehounková and Prach,

2006; Konvalinková and Prach, 2010).

Only rarely is the role of the surroundings determined based on repeatedly analysed experimental

plots. Salonen and Setälä (1992) conclude that seed supply is the major factor and soil quality only 265

an additional factor in determining colonization. Their study is probably the only one similar in

principle to that reported here. Lanta and Lepš (2009) conclude that differential seed inputs lead to

different successions even when all other environmental conditions are equal. However, the above-

mentioned studies did not assess temporal changes in the role of local site conditions and landscape

factors. Thus, our study experimentally demonstrates probably for the first time the continuously 270

decreasing role of substrate quality and increasing role of landscape context during succession.

Divergent vs convergent succession: a context-dependent phenomenon

The changing role of substrate and locality during early succession revealed by this study is

related to the often discussed topic of divergence vs. convergence in succession. Both convergence 275

and divergence in early succession have been reported from various successions (Walker and del

Moral, 2003; Walker et al., 2010). For example, Odland (1997) reports divergence on an artificially

constructed island as the vegetation gradually differentiated along a steep moisture gradient. Del

Moral (2007) also concludes that divergence prevails in early succession on substrates of volcanic

origin and only weak convergence occurs in plots that are located close together. Similar 280

conclusions are presented by Tsuyuzaki (2009). On the other hand, Borgegård (1990) found

increasing influence of the surrounding vegetation on species composition of seral stages in

abandoned sand-gravel pits resulting in the late stages of succession being more uniform. The

expectation is that the composition of vegetation in the initial stages of succession is determined

mainly stochastically and in later stages more deterministically (Walker and del Moral, 2003), 285

Page 12 of 25

Accep

ted

Man

uscr

ipt

12

which supports the convergent character of succession (del Moral, 2009). On the other hand,

species in the early stages of succession are usually ruderals with broad ecological amplitudes and

in the late stages are usually more specialized (Grime, 2002). This supports divergence. Obviously,

the extent to which succession is divergent or convergent generally depends on the participating

species, space and temporal scales, differences in local conditions and in landscape if the succession 290

occurs in landscapes that differ in climatic and other features. Taking this into account, the

convergence recorded in the experimental plots with very different substrates at one locality in this

study can be attributed to amelioration of the substrates and divergence between the two localities

to differences in the species pools in climatically contrasting landscapes. The effect of the different

species pools between the two contrasting localities became more and more noticeable in the 295

experimental plots as the substrate specificity decreased during succession.

The trends in divergence vs. convergence during succession are summarized in Table 3. If

contrasting habitats are located in the same or a similar landscape, convergence is expected,

whereas if similar or the same habitats are located in contrasting landscapes, divergence is expected.

The first occurred on all the substrates at each locality and the latter on the same substrate at the two 300

localities in this study. For the remaining combinations cited in Table 3, i.e. contrasting habitats in

contrasting landscapes or the same habitats in the same or a similar landscape, no or slight

convergence or divergence of successions can be expected. This scheme provides a general

framework for interpreting the results and conclusions also of previous studies on divergence vs.

convergence during succession (Lepš and Rejmánek, 1991; del Moral, 2009; Walker et al. 2010). 305

Acknowledgements

The work was supported by grants no. GACR-P505/11/0256, MSM6007665801 (K.P. and K.Ř.),

long-term research plans RVO67985939, and MSM0021620828, and project no. LC06073 (P.P.). 310

P.P. acknowledges the Praemium Academiae Award from the Academy of Sciences of the Czech

Page 13 of 25

Accep

ted

Man

uscr

ipt

13

Republic. We thank Tony Dixon for English revision and anonymous reviewers for their valuable

comments.

315

References

Bardgett, R.D., Wardle, D.A., 2010. Aboveground-belowground linkages: Biotic interactions,

ecosystem processes and global change. Oxford University Press Oxford, UK.

Benjamin, K., Domon, G., Bouchard, A., 2005. Vegetation composition and succession of 320

abandoned farmland: effects of ecological, historical and spatial factors. Landscape Ecol. 20,

627-647.

Borgegård, S.O., 1990. Vegetation development in abandoned gravel pits: effects of surrounding

vegetation, substrate and regionality. J. Veget. Sci. 1, 675–682.

Clements, F.E., 1916. Plant Succession: an Analysis of the Development of Vegetation. Carnegie 325

Institution, Washington.

del Moral, R., 2007. Vegetation dynamics in space and time: an example from Mount St. Helens. J.

Veget. Sci. 18, 479–488.

del Moral, R., 2009. Increasing deterministic control of primary succession on Mount St. Helens,

Washington. J. Veget. Sci. 20, 1145–1154. 330

del Moral, R., Ellis, E.E. 2004. Gradients in heterogeneity and structure on lahars, Mount St.

Helens, Washington, USA. Plant Ecol. 175, 273-286.

del Moral, R., Wood, D.M., Titus, J.H., 2005. Proximity, microsites and biotic interactions during

early primary succession. In: Dale, V.H., Swanson, F.J., Crisafulli, C.M. (Eds.) Ecological

Recovery after the 1980 Eruptions of Mount St. Helens. Springer, New York, pp. 93-109. 335

del Moral, R., Saura, J.M., Emenegger, J.N., 2010. Primary succession trajectories on a barren

plain, Mount St. Helens, Washington. J. Veget. Sci. 22: 857-867.

Page 14 of 25

Accep

ted

Man

uscr

ipt

14

Dovčiak, M., Frelich, L.E., Reich, P.B., 2005. Pathways in old-field succession to white pine: seed

rain, shade, and climate effects. Ecol. Monogr. 75, 363–378.

Ejrnæs, R., Hansen D.N., Aude, E., 2003. Changing course of secondary succession in abandoned 340

sandy fields. Biol. Conserv. 109, 343–350.

Glenn-Lewin, D.C., Peet, R.K., Veblen, T.T., (Eds.) 1992. Plant Succession: Theory and Prediction.

Chapman and Hall, London.

Grime, J.P. 2002. Plant Strategies and Vegetation Processes. 2nd Ed.. J. Wiley and Sons, Chichester.

Hodačová, D., Prach, K., 2003. Spoil heaps from brown coal mining: technical reclamation vs. 345

spontaneous re-vegetation. Rest. Ecol. 11, 385–391.

Jones, C.C., del Moral, R., 2009. Dispersal and establishment both limit colonization during

primary succession on a glacier foreland. Plant Ecol. 204, 217-230.

Johnson, E.A., Miyanishi, K., 2008. Testing the assumptions of chronosequences in succession.

Ecol. Letters 11, 419–431. 350

Kent, M., Coker, P., 1992. Vegetation Description and Analysis. Belhaven Press, London.

Kirmer, A., Tischew, S., Ozinga, W.A., von Lampe, M., Baasch, A., van Groenendael, J.M., 2008.

Importance of regional species pools and functional traits in colonization processes: predicting

re-colonization after large-scale destruction of ecosystems. J. Appl. Ecol. 45, 1523–1530.

Konvalinková, P., Prach, K., 2010. Spontaneous vegetation succession in extracted peatlands: a 355

multi-site study. Preslia 82, 423-435.

Laliberté, E., Grace, J.B., Huston, M.A., Lambers, H., Teste, F.P., Turner, B.L., Wardle, D.A.,

2013. How does pedogenesis drive plant diversity? Trends Ecol. Evol. 28, 331-340.

Lanta, V., Lepš, J., 2009. How does surrounding vegetation affect the course of succession: A five-

year container experiment. J. Veget. Sci. 20, 686–694. 360

Lepš, J., Rejmánek, M., 1991. Convergence or divergence: What should we expect from vegetation

succession? Oikos 62, 261-264.

Page 15 of 25

Accep

ted

Man

uscr

ipt

15

Lepš, J., Šmilauer, P. 2003. Multivariate Analysis of Ecological Data Using CANOCO. Cambridge

University Press, Cambridge.

Mitchley, J., Buckley, G.Y., Helliwell, D.R. (1996) Vegetation establishment on chalk marl spoil: 365

The role of nurse grass species and fertilizer application. J. Veget. Sci. 7, 543–548.

Morecroft, M.D., Masters, G.J., Brown, V.K., Clarke, I.P., Taylor, M.E., Whitehouse, A.T., 2004.

Changing precipitation patterns alter plant community dynamics and succession in ex-arable

grassland. Funct. Ecol.18, 648–655.

Novák, J., Konvička, M., 2006. Proximity of valuable habitats affects succession patterns in 370

abandoned quarries. Ecol. Eng. 26, 113–122.

Odland, A., 1997. Development of vegetation in created wetlands in western Norway. Aquat. Bot.

59, 45–62.

Otto, R., Krüsi, B.O., Burga, C.A., Fernandéz-Palacios, J.M., 2006. Old-field succession along a

precipitation gradient in the semi-arid coastal region of Tenerife. J. Arid Env. 65, 156-178. 375

Prach, K., 1987. Succession of vegetation on dumps from strip coal mining, N. W. Bohemia,

Czechoslovakia. Folia Geobot. Phytotax. 22, 339–354.

Prach, K., Řehounková, K., 2006. Vegetation succession over broad geographical scales: which

factors determine the patterns? Preslia 78, 469-480.

Prach, K., Pyšek, P., Jarošík, V., 2007a. Climate and pH as determinants of vegetation succession in 380

Central-European human-made habitats. J. Veget. Sci. 18, 701–710.

Prach, K., Lepš, J., Rejmánek, M., 2007b. Old field succession in central Europe: local and regional

patterns. Cramer V.A., Hobbs R. J. (Eds.) Old Fields: Dynamics and Restoration of Abandoned

Farmland. Island Press, Washington, pp. 180-201.

Roche, P., Tatoni, T., Medail, F., 1998. Relative importance of abiotic and land use factors in 385

explaining variation in woody vegetation in a French rural landscape. J. Veget. Sci. 9, 221–228.

Rydin, H., Borgegård, S.-O., 1991. Plant characteristics over a century of primary succession on

islands, Lake Hjalmaren. Ecol. 72, 1089–1101.

Page 16 of 25

Accep

ted

Man

uscr

ipt

16

Řehounková, K., Prach, K., 2006. Spontaneous vegetation succession in disused gravel-sand pits:

role of local site and landscape factors. J. Veget. Sci. 17, 493–500. 390

Řehounková, K., Prach, K., 2008. Spontaneous vegetation succession in gravel-sand pits: a

potential for restoration. Rest. Ecol. 16, 305–312.

Řehounková, K., Prach, K., 2010. Life-history traits and habitat preferences of colonizing plant

species in long-term spontaneous succession in abandoned gravel-sand pits. Basic Appl. Ecol.

11, 45–53. 395

Salonen, V., 1990. Early plant succession in two abandoned cut-over peatland areas. Holarctic Ecol.

13, 217–223.

Salonen, V., Setälä, H., 1992. Plant colonization of bare peat surface - relative importance of seed

availability and soil. Ecography 15, 199–204.

Samuels, C.L., Drake, J.A., 1997. Divergent perspectives on community convergence. Trends Ecol. 400

Evol. 12, 427-432.

Settele, J., Margules, C., Poschlod, P., Henle, K. (Eds.) 1996. Species Survival in Fragmented

Landscapes. Kluwer Acad. Publ., Dordrecht.

Sparks, D.L., Page, A.L., Helmke, P.A., Loeppert, R.H., Soltanpour, P.N., Tabatabai, M.A., Johjnson, C.T.,

Sumner, M.E., 1996. Methods of soil analysis. Part 3 – Chemical methods. Soil Science Society of 405

America, Madison.

ter Braak, C.J., Šmilauer, P., 2002. CANOCO Reference Manual and CanoDraw for Windows

User´s Guide: Software for Canonical Community Ordination (version 4.5). Microcomputer

Power. Ithaca.

Tilman, D., 1988. Plant Strategies and the Dynamics and Structure of Plant Communities. Princeton 410

Univ. Press, Princeton.

Tsuyuzaki, S., 2009. Causes of plant community divergence in the early stages of volcanic

succession. J. Veget. Sci. 20, 959-969.

van der Putten, W.H., Bardgett, R.D., Bever, J.D., Bezemer, T.M., Casper, B.B., Fukami, T.,

Kardol, P., Klironomos, J.N., Kulmatiski, A., Schweitzer, J.A., Suding, K.N., Van der Voorde, 415

Page 17 of 25

Accep

ted

Man

uscr

ipt

17

T.F.J., Wardle, D.A., 2013. Plant-soil feedbacks: the past, the present and future challenges. J.

Ecol. 101, 265-276.

Walker, L.R., del Moral, R., 2003. Primary Succession and Ecosystem Rehabilitation. Cambridge

Univ. Press, Cambridge.

Walker, L.R., Walker, J., Hobbs R.J. (Eds.) 2007. Linking Restoration and Ecological Succession. 420

Springer, New York.

Walker, L.R., Wardle, D.A., Bardgett, R.D., Clarkson, B.D. 2010. The use of chronosequences in

studies of ecological succession and soil development. J. Ecol. 98, 725-736.

Zobel, M., van der Maarel, E., Dupré, C., 1998. Species pool: the concept, its determination and

significance for community restoration. Appl. Veget. Sci. 1, 55–66. 425

Page 18 of 25

Accep

ted

Man

uscr

ipt

18

Figure captions: 1

2

Fig. 1. DCA ordination of samples (a). The direction of succession on five at both 3

localities indicated by arrows (black lines – Upland, grey lines – Lowland). The 4

arrows connect centroids of samples for each substrate in particular years, i.e. from 5

2002 to 2011. DCA ordination of species (b). Species abbreviations: AgroCapi – 6

Agrostis capillaris, AgroGiga – Agrostis gigantea, AperSpic – Apera spica-venti, 7

ArrhElat – Arrhenatherum elatius, ArteVulg – Artemisia vulgaris, BromHord – 8

Bromus hordeaceus subs. hordeaceus, BromSter – Bromus sterilis, CalaEpig – 9

Calamagrostis epigejos, CampPatu – Campanula patula, CapsBuPa – Capsella 10

bursa-pastoris, ChenAlbu – Chenopodium album, CirsArve – Cirsium arvense, 11

ConyCana – Conyza canadensis, DactGlom – Dactylis glomerata, DigiSang – 12

Digitaria sanguinalis, EchiCrGa – Echinochloa crus-galli, ElytRepe – Elytrigia 13

repens, EpilCili – Epilobium ciliatum, FallConv – Fallopia convolvulus, FestArun – 14

Festuca arundinacea, FestPrat – Festuca pratensis, FestRubr – Festuca rubra, 15

GaleTetr – Galeopsis tetrahit, GaliApar – Galium aparine , LoliPere – Lolium 16

perenne, LotuCorn – Lotus corniculatus, MatriMari – Matricaria maritima, MediLupu 17

– Medicago lupulina, PhlePrat – Phleum pratense, PinuSylv – Pinus sylvestris, 18

PoaAngu – Poa angustifolia, PoaPalu – Poa palustris, PoaPrat – Poa pratensis, 19

PoaTriv – Poa trivialis, PolyHydr – Polygonum hydropiper, PolyLapa – Polygonum 20

lapathifolium, QuerRobu – Quercus robur, RanuRepe – Ranunculus repens, 21

RosaCani – Rosa canina, RumeAcet – Rumex acetosella, SoncArve – Sonchus 22

arvensis, TanaVulg – Tanacetum vulgare, TaraOffi - Taraxacum officinale, TrisFlav – 23

Trisetum flavescens, ToriJapo – Torilis japonica, TrifRepe – Trifolium repens, 24

TussFarf – Tussilago farfara, UrtiDioi – Urtica dioica, ViolArve – Viola arvensis. 25

Page 19 of 25

Accep

ted

Man

uscr

ipt

19

1

2

Fig. 2. Percentage of the variability in the composition of the vegetation accounted for by 3

Locality (black dots) and Substrate (open circles) in the first 10 years of succession. Partial 4

effects were calculated using CCA analyses. 5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

Page 20 of 25

Accep

ted

Man

uscr

ipt

20

1

Table 1. Substrate chemistry at the start and the end of the experiment. Only one mixed 2

sample was taken for Clay, Sand, and Peat at the beginning of the experiment because 3

identical allochthonous substrates were used at both localities. Average values with standard 4

deviation are thus shown only for the end of experiment. L – Lowland locality, U – Upland 5

locality. 6

Substrate Time Locality pH (H2O) C tot. [%] N tot. [%] C:N Ca tot. [%] Mg tot. [%]

Clay Start Both 8.3 2.43 0.19 13:1 0.79 2.71

End L 8.42±0.20 11.48±1.02 1.89±0.27 6:1 3.52±0.35 1.13±0.07

U 8.18±0.17 20.14±0.80 1.86±0.12 11:1 2.73±0.27 1.13±0.05

Sand Start Both 6.4 0.02 0.02 1:1 0.12 0.34

End L 6.57±0.30 0.93±0.23 0.38±0.15 2:1 0.25±0.09 0.04±0.01

U 5.85±0.14 0.85±0.25 0.30±0.08 3:1 0.16±0.06 0.02±0.01

Peat Start Both 4.4 33.41 1.62 21:1 2.21 0.02

End L 4.73±0.18 49.16±2.26 12.82±1.78 4:1 4.89±0.80 0.60±0.09

U 5.62±0.18 20.13±0.61 6.76±1.64 3:1 3.21±0.22 0.41±0.04

Topsoil Start L 6.1 3.18 0.35 9:1 1.78 0.13

U 7.31±0.24 8.72±0.95 1.83±0.31 5:1 5.83±0.23 0.66±0.07

End L 7.4 1.82 0.20 9:1 3.6 2.20

U 6.15±0.21 19.46±0.68 2.47±0.31 8:1 2.55±0.19 0.21±0.01

Control Start L 5.9 3.77 0.4 9:1 1.87 0.18

U 7.20±0.20 8.80±0.82 1.51±0.87 6:1 5.54±0.41 0.66±0.04

End L 7.0 1.54 0.18 9:1 3.94 2.45

U 6.14±0.42 21.86±2.39 2.60±0.24 8:1 2.42±0.29 0.23±0.03

7

Page 21 of 25

Accep

ted

Man

uscr

ipt

21

Table 2. The results of the CCA of the partial and marginal effects for particular years. 1

Covariables: Substrate/Locality. F-values for the F-statistics with probability levels 2

***P < 0.001, **P < 0.01, *P < 0.05 in the Monte Carlo test. Percentages: marginal - variation 3

attributed to environmental variables not considering the effects of other environmental 4

variables, partial – variance attributed to variables with the other environmental variables as 5

covariables. 6

7

Year Locality

Partial %

Locality

Partial F

Substrate

Partial %

Substrate

Partial F

Locality

Marginal %

Locality

Marginal F

Substrate

Marginal %

S

M

2002 1.8 1.97 ** 8.5 4.86*** 2.1 2.05*** 8.8 6

2003 3.7 2.00 ** 7.6 4.10*** 3. 9 2.46*** 7.8 4

2004 7.9 4.38 *** 7.1 3.95*** 8.0 4.15*** 7.2 3

2005 7.7 4.15 *** 4.5 3.68*** 7.8 4.09*** 4.7 2

2006 8.5 5.93 *** 4.4 3.41*** 8.6 5.69*** 4.5 2

2007 12.0 7. 03 *** 2.7 2.19*** 12.1 6.56*** 2.8 1

2008 13.3 7.53 *** 3.0 1.76*** 13.4 7.40*** 3.1 1

2009 14.4 8.05 *** 2.6 1.47*** 14.5 7.98*** 2.7 1

2010 19.3 11.81** 2.4 1.15** 19.4 11.50** 2.5 1

2011 20.7 11.37** 2.3 1.65** 20.9 11.32* 2.4 1

8

9

Page 22 of 25

Accep

ted

Man

uscr

ipt

22

1

Table 3. A general scheme of the trends in divergence vs. convergence in succession on 2

different habitats and in different landscapes. 3

4

Landscape

Contrasting Similar or the same

Similar or

the same

Clear divergence

No or

Slight convergence or

Slight divergence

Habitats

Contrasting

No or

Slight convergence or

Slight divergence

Clear convergence

5

6

7

Page 23 of 25

Accep

ted

Man

uscr

ipt

Figure 1a

Page 24 of 25

Accep

ted

Man

uscr

ipt

Figure 1b

Page 25 of 25

Accep

ted

Man

uscr

ipt

Figure 2