Perspectives in Plant Ecology, Evolution and Systematics 16 (2014) 174–179
Contents lists available at ScienceDirect
Perspectives in Plant Ecology, Evolution and Systematics
journal homepage: www.elsevier.com/locate/ppees
Research article
Role of substrate and landscape context in early succession:
An experimental approach
Karel Prach a,b,∗ , Petr Pyšek c,d , Klára Řehounková a,b
a
Department of Botany, Faculty of Science, University of South Bohemia, Branišovská 31, CZ-370 05 České Budějovice, Czech Republic
Institute of Botany, Academy of Sciences of the Czech Republic, Dukelská 143, CZ-37982 Třeboň, Czech Republic
Institute of Botany, Academy of Sciences of the Czech Republic, CZ-252 43 Průhonice, Czech Republic
d
Department of Ecology, Faculty of Science, Charles University in Prague, Viničná 7, CZ-128 44 Prague, Czech Republic
b
c
a r t i c l e
i n f o
Article history:
Received 2 October 2013
Received in revised form 28 April 2014
Accepted 5 May 2014
Available online 15 May 2014
Keywords:
Central Europe
Convergence vs. divergence
Landscape
Ordination
Substrate manipulation
Vegetation succession
a b s t r a c t
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 hypothesised 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 clay, sand, peat,
sterilised 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 regions diverged and converged within each region.
We concluded with regard to the divergence vs. convergence dichotomy in succession: if contrasting
habitats occur in the same or similar 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,
successions may exhibit no or only slight divergence or convergence.
© 2014 Geobotanisches Institut ETH, Stiftung Ruebel. Published by Elsevier GmbH. All rights reserved.
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 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 (Settele 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 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
∗ Corresponding author at: Department of Botany, Faculty of Science, University
of South Bohemia, Branišovská 31, CZ-370 05 České Budějovice, Czech Republic.
E-mail address: [email protected] (K. Prach).
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 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 et al.,
2007a) and soil texture (Ejrnæs 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 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
http://dx.doi.org/10.1016/j.ppees.2014.05.002
1433-8319/© 2014 Geobotanisches Institut ETH, Stiftung Ruebel. Published by Elsevier GmbH. All rights reserved.
K. Prach et al. / Perspectives in Plant Ecology, Evolution and Systematics 16 (2014) 174–179
successional age (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 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 colonisation 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 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).
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 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 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).
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 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 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?
Methods
Site description, experimental design and data recording
The experiment was established in spring 2002 at two climatically different localities (hereafter called Locality), in the Czech
Republic, central Europe:
175
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 ◦ C; average annual precipitation 461 mm (longterm 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 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 annual temperature 6.7 ◦ C; 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 Festuca 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.
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 110 ◦ C to kill plant propagules and then returned to the site (Topsoil), and (v) untouched local soil used as a control (Control). The
sterilisation treatment was not needed in the case of allochtonous
substrates (clay, sand, peat) because they were excavated from
the depth below the surface (clay ∼100 m, sand several metres,
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, 2006, 2008, 2010) and peat diggings
(Konvalinková and Prach, 2010). The plots with local soil represent
the abandoned fields described by Prach et al. (2007b) and Jírová
et al. (2012).
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 Latinsquare 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 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 colonisation of the experimental plots by species from
the surrounding vegetation especially by surface stolones. Substrate chemistry, summarised in Table 1, was assessed at the start
and the end of the experiment using standard methods (Sparks
et 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 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).
176
K. Prach et al. / Perspectives in Plant Ecology, Evolution and Systematics 16 (2014) 174–179
Table 1
Substrate chemistry at the start and the end of the experiment. Only one mixed sample was taken for clay, sand, and peat at the beginning of the experiment because identical
allochthonous substrates were used at both localities. Average values with standard deviation are thus shown only for the end of experiment. L – lowland locality, U – upland
locality.
Substrate
Time
Locality
pH (H2 O)
C tot. [%]
N tot. [%]
C:N
Ca tot. [%]
Mg tot. [%]
Clay
Start
End
Sand
Start
End
Peat
Start
End
Topsoil
Start
Both
L
U
Both
L
U
Both
L
U
L
U
L
U
L
U
L
U
8.3
8.42 ± 0.20
8.18 ± 0.17
6.4
6.57 ± 0.30
5.85 ± 0.14
4.4
4.73 ± 0.18
5.62 ± 0.18
6.1
7.31 ± 0.24
7.4
6.15 ± 0.21
5.9
7.20 ± 0.20
7.0
6.14 ± 0.42
2.43
11.48 ± 1.02
20.14 ± 0.80
0.02
0.93 ± 0.23
0.85 ± 0.25
33.41
49.16 ± 2.26
20.13 ± 0.61
3.18
8.72 ± 0.95
1.82
19.46 ± 0.68
3.77
8.80 ± 0.82
1.54
21.86 ± 2.39
0.19
1.89 ± 0.27
1.86 ± 0.12
0.02
0.38 ± 0.15
0.30 ± 0.08
1.62
12.82 ± 1.78
6.76 ± 1.64
0.35
1.83 ± 0.31
0.20
2.47 ± 0.31
0.4
1.51 ± 0.87
0.18
2.60 ± 0.24
13:1
6:1
11:1
1:1
2:1
3:1
21:1
4:1
3:1
9:1
5:1
9:1
8:1
9:1
6:1
9:1
8:1
0.79
3.52 ± 0.35
2.73 ± 0.27
0.12
0.25 ± 0.09
0.16 ± 0.06
2.21
4.89 ± 0.80
3.21 ± 0.22
1.78
5.83 ± 0.23
3.6
2.55 ± 0.19
1.87
5.54 ± 0.41
3.94
2.42 ± 0.29
2.71
1.13 ± 0.07
1.13 ± 0.05
0.34
0.04 ± 0.01
0.02 ± 0.01
0.02
0.60 ± 0.09
0.41 ± 0.04
0.13
0.66 ± 0.07
2.20
0.21 ± 0.01
0.18
0.66 ± 0.04
2.45
0.23 ± 0.03
End
Start
Control
End
Data analyses
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 (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 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).
Results
The greatest changes in substrate chemistry were recorded for
allochtonous substrates where C 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 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 increasingly dissimilar, indicating a
convergence within and divergence between localities (Fig. 1a). 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 P. pratense, F. 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 succeeded by perennial weeds
and ruderal species, such as Elytrigia repens and Agrostis gigantea,
and later on by A. 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, while those of Substrate decreased
(Fig. 2). All these effects were significant (Table 2). The CCA ordination of all the plots revealed significant summarised effects of
Table 2
The results of the CCA of the partial and marginal effects for particular years. Covariables: Substrate/Locality. F-values for the F-statistics with probability levels *** P < 0.001,
**
P < 0.01, * P < 0.05 in the Monte-Carlo test. Percentages: marginal-variation attributed to environmental variables not considering the effects of other environmental variables,
partial-variance attributed to variables with the other environmental variables as covariables.
Year
Locality
partial %
Locality partial
F
Substrate
partial %
Substrate
partial F
Locality
marginal %
Locality
marginal F
Substrate
marginal %
Substrate
marginal F
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
1.8
3.7
7.9
7.7
8.5
12.0
13.3
14.4
19.3
20.7
1.97**
2.00**
4.38***
4.15***
5.93***
7. 03***
7.53***
8.05***
11.81**
11.37**
8.5
7.6
7.1
4.5
4.4
2.7
3.0
2.6
2.4
2.3
4.86***
4.10***
3.95***
3.68***
3.41***
2.19***
1.76***
1.47***
1.15**
1.65**
2.1
3. 9
8.0
7.8
8.6
12.1
13.4
14.5
19.4
20.9
2.05***
2.46***
4.15***
4.09***
5.69***
6.56***
7.40***
7.98***
11.50**
11.32*
8.8
7.8
7.2
4.7
4.5
2.8
3.1
2.7
2.5
2.4
6.97***
4.61***
3.71***
2.32***
2.35***
1.33***
1.54***
1.29***
1.52**
1.70**
K. Prach et al. / Perspectives in Plant Ecology, Evolution and Systematics 16 (2014) 174–179
177
Fig. 2. Percentage of the variability in the composition of the vegetation accounted
for by locality (black dots) and substrate (open circles) in the first 10 years of succession. Partial effects were calculated using CCA analyses.
time, i.e. age of succession (accounted for 15.7% of the variability),
Locality (17.2%) and Substrate (5.3%).
Discussion
Role of substrate
Fig. 1. DCA ordination of samples (a). The direction of succession on five at both
localities indicated by arrows (black lines – upland, grey lines – lowland). The arrows
connect centroids of samples for each substrate in particular years, i.e. from 2002
to 2011. DCA ordination of species (b). Species abbreviations: AgroCapi – Agrostis
capillaris, AgroGiga – Agrostis gigantea, AperSpic – Apera spica-venti, ArrhElat –
Arrhenatherum elatius, ArteVulg – Artemisia vulgaris, BromHord – Bromus hordeaceus
subs. hordeaceus, BromSter – Bromus sterilis, CalaEpig – Calamagrostis epigejos,
CampPatu – Campanula patula, CapsBuPa – Capsella bursa-pastoris, ChenAlbu –
Chenopodium album, CirsArve – Cirsium arvense, ConyCana – Conyza canadensis, DactGlom – Dactylis glomerata, DigiSang – Digitaria sanguinalis, EchiCrGa –
Echinochloa crus-galli, ElytRepe – Elytrigia repens, EpilCili – Epilobium ciliatum, FallConv – Fallopia convolvulus, FestArun – Festuca arundinacea, FestPrat – Festuca
pratensis, FestRubr – Festuca rubra, GaleTetr – Galeopsis tetrahit, GaliApar – Galium
aparine, LoliPere – Lolium perenne, LotuCorn – Lotus corniculatus, MatriMari – Matricaria maritima, MediLupu – Medicago lupulina, PhlePrat – Phleum pratense, PinuSylv
– Pinus sylvestris, PoaAngu – Poa angustifolia, PoaPalu – Poa palustris, PoaPrat – Poa
pratensis, PoaTriv – Poa trivialis, PolyHydr – Polygonum hydropiper, PolyLapa – Polygonum lapathifolium, QuerRobu – Quercus robur, RanuRepe – Ranunculus repens,
RosaCani – Rosa canina, RumeAcet – Rumex acetosella, SoncArve – Sonchus arvensis,
TanaVulg – Tanacetum vulgare, TaraOffi – Taraxacum officinale, TrisFlav – Trisetum
flavescens, ToriJapo – Torilis japonica, TrifRepe – Trifolium repens, TussFarf – Tussilago
farfara, UrtiDioi – Urtica dioica, ViolArve – Viola arvensis.
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; Bardgett
and Wardle, 2010). 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
summarised 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 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., 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 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 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 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
colonised by species typical of disturbed peaty soils (Konvalinková
and Prach, 2010) (Fig. 1a and 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 towards plots on other substrates. Vegetation on the
178
K. Prach et al. / Perspectives in Plant Ecology, Evolution and Systematics 16 (2014) 174–179
Topsoil plots at the Upland locality also differed from that on other
substrates because Taraxacum officinale colonised 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 monopolise the space and at least temporarily deflect succession (Samuels and Drake, 1997). However, the vegetation on
Topsoil plots also gradually 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
experiment. The experimental substrates we used differed in their
chemistry, but none of them was really extreme (Table 1; Prach
et al., 2007a).
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).
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 vegetation on the course of succession are observational and
usually based on the space-for-time substitution approach, which
may limit some generalisations (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).
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 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., A.
elatius in Lowland plots, and P. pratense and F. pratensis in Upland
plots. The experimental plots were rather small because of technical limitations. Thus both, amelioration of the substrates and
colonisation from the surroundings are 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
an additional factor in determining colonisation. 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 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 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 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), 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 specialised (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
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 experimental plots as the substrate specificity decreased during succession.
The trends in divergence vs. convergence during succession are
summarised 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 localities in this study. For the remaining combinations cited in Table 3,
Table 3
A general scheme of the trends in divergence vs. convergence in succession on
different habitats and in different landscapes.
Landscape
Habitats
Contrasting
Similar or the same
Similar or the
same
Clear divergence
No
or
Slight convergence
or
Slight divergence
Contrasting
No
or
Slight convergence
or
Slight divergence
Clear convergence
K. Prach et al. / Perspectives in Plant Ecology, Evolution and Systematics 16 (2014) 174–179
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).
Acknowledgements
The work was supported by grant nos. GACR-P505/11/0256,
MSM6007665801 (K.P. and K.Ř.), long-term research plans
RVO67985939, and MSM0021620828, and project no. LC06073
(P.P.). P.P. acknowledges the Praemium Academiae Award from the
Academy of Sciences of the Czech Republic. We thank Tony Dixon
for English revision and anonymous reviewers for their valuable
comments.
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 abandoned farmland: effects of ecological, historical and spatial factors.
Landsc. 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.
Callaway, R.M., Walker, L.R., 1997. Competition and facilitation: a synthetic
approach to interactions in plant communities. Ecol. 78, 1958–1965.
Clements, F.E., 1916. Plant Succession: An Analysis of the Development of Vegetation. Carnegie 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.
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.
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.
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 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. spontaneous re-vegetation. Restor. Ecol. 11, 385–391.
Jírová, A., Klaudisová, A., Prach, K., 2012. Spontaneous restoration of target vegetation in old fields in a central European landscape: a repeated analysis after three
decades. Appl. Veget. Sci. 15, 245–252.
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. Lett. 11, 419–431.
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 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.
179
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.
Lepš, J., Rejmánek, M., 1991. Convergence or divergence: what should we expect
from vegetation succession? Oikos 62, 261–264.
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: 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 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
Environ. 65, 156–178.
Philippi, T.E., Dixon, P.M., Taylor, B.E., 1998. Detecting trends in species composition.
Ecol. Appl. 8, 300–308.
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 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. In: 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 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.
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.
Řehounková, K., Prach, K., 2008. Spontaneous vegetation succession in gravel-sand
pits: a potential for restoration. Restor. 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.
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. Evol. 12, 427–432.
Settele, J., Margules, C., Poschlod, P., Henle, K. (Eds.), 1996. Species Survival in Fragmented Landscapes. Kluwer Academic Publishers, 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. Madison, Soil Science Society of America.
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 University 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, 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 University Press, Cambridge.
Walker, L.R., Walker, J., Hobbs, R.J. (Eds.), 2007. Linking Restoration and Ecological
Succession. 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.
Wood, D.M., del Moral, 1987. Mechanisms of early primary succession in subalpine
habitats on Mount St. Helens. Ecol. 68, 780–790.
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.