Folia Geobotanica 38: 367­379, 2003
CHANGES IN SPECIES DENSITY ALONG THE SOIL pH
GRADIENT ­ EVIDENCE FROM GERMAN PLANT
COMMUNITIES
Brigitte Schuster* & Martin Diekmann
Vegetation Ecology and Conservation Biology, FB 2, University of Bremen, D-28359 Bremen, Germany; * fax
+49 421 2187052, e-mail schuster@uni-bremen.de
Abstract: The relationship between small-scale species richness and soil pH in plant communities was reviewed
using 40 (mainly phytosociological) studies primarily from the northern parts of Germany. Four formations were
distinguished (forest, grassland, mire and sand). To examine the above relationship, two approaches were
applied: (a) regression analysis using linear and quadratic models, and (b) calculation of the mean pH of the five
plots with the highest species density. Despite methodological problems (lack of single plot data in some studies,
varying plot sizes, different ways of measuring pH) some general patterns could be identified. In forests, the
regressions in most cases indicated a linear or curved increase in species density with increasing pH, whereas
maximum species density was, except in one case, found at an intermediate pH of around 5. In contrast, in
grasslands and mires, most studies showed hump-backed relationships between species density and soil acidity,
and, accordingly, maximum species density was found at moderately high pH between 5 and 6. For sand
vegetation, the few studies available revealed an increase in species density with increasing pH. The findings
imply that the predominance of calciphilous species vs. acidophilous species in the Central European flora does
not always translate into positive relationships between plant species density and pH on a local plot scale. The
hump-backed species density-pH pattern may be explained by confounding effects of other environmental
factors and of productivity, but also by the large overlap of species tolerances in the middle of the acidity
gradient.
Keywords: Calciphilous species, Hump-backed model, pH amplitude, Physiological tolerance, Regional
species pool, Soil acidity, Species tolerance
INTRODUCTION
Species richness varies across environments. On a regional scale, the species pool ­ the set
of all species physiologically capable of living in the environmental conditions of a certain
community (ZOBEL et al. 1998) ­ differs considerably between vegetation types. These differ
in species number also on a local scale, where factors such as competition are usually regarded
as the main determinants of species diversity (GRIME 1979). In recent years, however, the
species pool has also been recognized as playing an important role in regulating the plant
species richness of sites (ZOBEL 1997, GRACE 1999, PÄRTEL 2002).
EWALD (2003) shows that the regional pool of vascular plants in Central Europe on
base-rich and calcareous soils is much larger than the pool on more acidic soils. This appears
to be true both for the flora as a whole and for broadly defined formations such as forests.
PÄRTEL (2002) has shown that, in general, positive relationships between richness and pH
prevail in temperate areas with evolutionary centres on high-pH soils. Supposing that the size
of the regional species pool is positively correlated with the small-scale richness of plots, local
communities on base-rich soils should be more diverse than those on more acidic soils.
According to EWALD (2003), this prediction is supported by various studies.
It is unquestioned that there is an increase in species richness from very acidic to somewhat
less acidic sites. However, across the potential range of soil acidity, the relationship between
species number and pH must inevitably be unimodal as no species can survive pH values
higher than about 9. The basic question is at what pH is the optimum in species richness
reached, or, more specifically, is there a decrease in species number already on increasingly
base-rich sites. Whereas the relationship between richness and productivity has received
much attention in ecology (GRIME 1979), the effects of single factors such as pH have not
been studied in much detail (but see, for example, DUPRÉ et al. 2002).
The main objective of our study was to examine the general relationship between species
richness in sample plots (species density) and (soil) pH for different formations by means of
linear and quadratic regression, and to compare the resulting patterns between and within
these formations. For this, we gathered data across a wide array of plant communities, based
on published and unpublished data mainly from Germany. We deliberately avoided using
Ellenberg's indicator values for pH (reaction) to calculate site averages as surrogates for
actual measurements, because there is much evidence that the relation between indicated and
measured values is not linear at high-pH sites (SCHAFFERS & SÝKORA 2000).
MATERIAL AND METHODS
Compilation of data
For the compilation of data, phytosociological and other plant ecological literature was
screened for information on plot species density (mostly vascular plants, but often also
including cryptogams) linked with measurements of pH. A total of 40 studies are represented
in the data set (Appendix 1). Study regions included various parts of Germany with an
emphasis on the north of the country, as well as two surrounding countries (France and
Austria). We preferably used data where information was given for single plots. Many of the
studies, however, contained floristic and environmental information only for clusters
including several plots, i.e., mean values of species density and pH for certain community
types. Cluster data are less reliable than single plot data, because species density values
usually were the mean of all plots in a cluster, whereas pH was measured only in a few
selected plots that are not necessarily representative of the whole cluster.
There were two further methodological problems that we had to cope with. The first arose
from the fact that almost all plots were established as relevés according to the Braun-Blanquet
approach, meaning that plot sizes usually differed considerably. Non-standardized plot sizes
may distort the analysis if there is a systematic relationship between pH and plot size chosen,
as species density increases with increasing plot size. The relations between plot size, pH and
species density will be dealt with in more detail below. The second problem concerns the pH
measurements that were made with different methods. Most studies relied on pH (H2O) data,
while in several cases pH (KCl or CaCl2) was measured, resulting in considerably lower
values (cf. TYLER 1997). When both pH values were given in a study, however, the two ways
of measuring soil acidity resulted in very similar correlations with species density, and only
368 B. Schuster & M. Diekmann
pH (H2O) was used for further analysis. To be able to include data of both types in
comparative analyses, we calculated the mean difference between pH (KCl or CaCl2) and pH
(H2O) across all studies (0.62 units) and added this mean to all values of pH (KCl or CaCl2) to
arrive at corrected values.
To examine if the relation between species density and pH differed between formations, we
distinguished four broadly defined vegetation types (see Appendix 1):
Forest: deciduous and coniferous woodland on fresh and wet soils (classes Alnetea
glutinosae, Querco-Fagetea, Vaccinio-Piceetea and Vaccinietea uliginosae);
Grassland: intensively and extensively managed, open communities dominated by
graminoids (Molinio-Arrhenatheretea, Nardo-Callunetea [Nardetalia] and Festuco-
-Brometea);
Mire: bog, poor mire and fen communities (Oxycocco-Sphagnetea and Scheuchzerio-
-Caricetea);
Sand: vegetation of open sandy heaths and dunes (Sedo-Scleranthetea [Corynephoretalia
canescentis and Thero-Airetalia] and Nardo-Callunetea [Calluno-Ulicetalia]).
Statistical data analysis
The relationship between species density and pH was first analyzed separately for each
study by means of regression. We only used linear and quadratic models, because other model
types, though possible or even realistic, did not serve the purpose of our study aiming at
discriminating between linear or hump-shaped relationships. The model with the higher
adjusted R2
(R2
adj) was regarded as the one fitting the data best. The regression lines for
studies belonging to the same formation were displayed in one figure by overlaying the
graphs, using the pH (H2O) values and corrected pH (KCl or CaCl2) values. Differences in
species density between studies of the same formation were not analyzed and are mainly an
effect of different plot sizes.
Differences in the relationship between species density and pH between formations were
analyzed by calculating the slope of the fitted regression lines (either linear or quadratic) at
pH 5 for all studies. The slopes were then compared between formations by means of unpaired
t-tests.
The regression analysis describes the general relationship between species density and pH,
but does not tell us where along the acidity gradient species density is at its maximum. Also
a flat hump-backed curve may still rise at high pH. In addition, differences in the variation in
species density along the acidity gradient may imply that the peak of the optimum curve does
not fully correspond to the pH of maximum species density. Therefore, we calculated the
mean pH of the five most species-rich plots, cut the pH range into 3 equal sections of low
(2.00­3.99) , middle (4.00­5.99) and high (6.00­8.00) pH and then checked in which section
the above mean value was situated. The results were basically the same when we used the
mean pH of a larger or smaller number of the most species-rich plots. This analysis could only
be conducted on studies with single plot data.
Species density along the pH gradient in Germany 369
RESULTS
Plot size varies in all but a few
studies, but detailed information for
single plots was available only in 14
studies. In two cases, plot size was
significantly correlated with pH
(Table 1), and in only one of the two
(no. 28) species density was also
correlated with plot size. It therefore,
in general, appears unlikely that
varying plot size had a distorting
effect on the analysis between
species density and pH. In more than
half of the studies there was no
significant positive relationship
between species density and plot
size.
In forests, there was a significant
relationship between species density
and pH in 9 out of 11 studies (Table 2). All regression lines showed a rather steep rise at low
pH, while the curves differed in their slopes at higher pH (Fig. 1). Those with maximum
values around or lower than 6 were largely linear, whereas those exceeding pH 6 were
quadratic, exhibiting a reduced slope or even tapering off (no. 18).
The majority of studies in grasslands (9 out of 14) also gave significant fits for the
relationship between species density and pH (Table 2). Again, there was a steep increase in
species number from low to moderately high pH (Fig. 1). The hump-backed pattern was more
pronounced than in forests, with five studies having clear optimum curves (no. 4, 8, 16, 22,
24). One of the studies including pH values > 7 (no. 14) was found to have a positive linear
regression line.
Five of the ten studies of mires resulted in significant relationships between the two
variables (Table 2). The two studies with pH maximum values around 6 resulted in linear,
positively sloping regression lines, and those with higher top values had hump-backed curves,
two of which exhibited a clear optimum (Fig. 1).
For the sand formation the number of studies was lower (Table 2). Three regression lines
were found to be significant, two of which exhibited a linear, and the third a quadratic, but
concave shape (Fig. 1). There was, however, a clear increase in species density with pH on
increasingly neutral soils.
The slopes of the regression curves at pH 5 in forests were significantly higher than in
mires (t = 2.32, P = 0.033, df = 19), and they tended to be higher than in grasslands (t = 1.02,
P = 0.319, df = 22) and in sand vegetation (t = 1.02, P = 0.332, df = 13). The grassland slopes
were not different from those of mires (t = 0.89, P = 0.384, df = 21) nor did they differ from
those of sandy vegetation (t = -0.16, P = 0.872, df = 15). Finally, also mires and sandy
vegetation did not significantly differ from each other (t = -0.132, P = 0.227, df = 12).
370 B. Schuster & M. Diekmann
Table 1. Relationship between plot size and pH and relationship
between species density and plot size, given for studies with
varying plot size. R2
adj values are given; significant positive
correlations are indicated by plus signs, significant negative ones
by minus signs (+/- ­ P < 0.05; ++/-- ­ P < 0.01; +++/--- P
< 0.001).
Relationship between
No. plot size species density
of study and pH and plot size
Forest 6 0.0 0.0
Grassland 16 13.8 - 0.0
22 0.1 2.7 ++
28 2.8 ++ 2.1 ++
Mire 1 0.0 1.0
7 0.0 24.0 +++
17 0.0 0.0
38 0.0 0.0
Sand 13 1.4 0.2
20 1.8 11.4 +++
Species density along the pH gradient in Germany 371
Table 2. Basic data analysis of the 40 studies: pH (minimum, maximum and mean values), species number
(minimum, maximum and mean values), regression statistics of the relationship between species density and pH
(R2
adj are given for both linear and quadratic models; significant fits are marked with an asterisk: * ­ P < 0.05;
** ­ P < 0.01; *** ­ P < 0.001. (u) ­ EXPLAIN !!!), and pH of maximum species density (mean pH of the five
sites with the highest species number).
No. of pH No. of species Regression pH of highest
study - R2
adj species
Min. Max. Mean Min. Max. Mean Linear Quadratic density
1 3.7 6.4 4.7 7 39 19.1 5.3 * 8.8 * 4.8
2 4.0 7.3 5.5 6 36 21.4 60.3 *** 60.7 ** 6.3
3 4.9 6.3 5.6 18 41 32.6 22.5 27.4 5.3
4 3.9 7.2 5.9 27 61 49.2 0.0 16.8 * 5.5
5 2.6 4.1 3.3 5 9 4.8 25.3 15.9 2.8
6 2.9 6.3 4.8 12 65 37.3 58.5 *** 59.7 *** 5.4
7 4.8 7.7 6.8 5 45 18.1 0.0 0.0 6.0
8 3.6 7.6 5.5 7 71 33.8 15.3 *** 18.1 *** 5.9
9 4.4 7.0 5.6 14 59 37.0 4.9 14.7 * 5.8
10 3.8 4.9 4.4 13 29 20.0 4.7 12.8 * 4.4
11 3.3 5.7 4.6 14 59 22.0 37.6 *** 36.2 *** 5.1
12 3.5 6.1 5.0 15 36 24.4 62.7 * 55.7 5.4
13 3.4 6.5 5.0 1 44 22.8 36.6 *** 35.8 *** 5.4
14 4.0 8.0 6.0 18 61 35.6 10.9 *** 10.9 *** 6.5
15 3.1 5.7 3.8 3 39 15.1 35.2 *** 39.0 *** 4.7
16 3.5 5.4 4.3 18 50 34.7 23.8 ** 28.2 ** 4.5
17 4.0 7.0 5.0 13 42 27.9 0.0 2.2 5.2
18 2.7 7.1 4.3 22 55 33.9 35.0 ** 50.9 ** 5.6
19 4.0 7.2 5.3 4 29 16.3 37.0 ** 33.9 ** 6.0
20 6.2 7.6 7.1 11 66 35.9 2.9 6.6 (u) 7.0
21 4.0 7.7 5.8 5 25 13.8 0.0 0.8 6.0
22 3.4 8.0 7.3 10 71 37.7 3.4 *** 7.7 *** 7.0
23 4.9 8.0 6.3 20 61 40.1 6.7 ** 5.9 * 6.5
24 3.7 8.1 5.9 5 49 22.5 1.7 13.4 *** 5.2
25 3.3 5.2 4.6 12 28 18.1 3.2 2.9 4.7
26 4.3 7.1 5.0 31 55 44.0 0.9 0.0 5.0
27 6.0 8.1 7.0 1 29 11.5 0.0 0.0 7.2
28 3.0 5.7 4.5 7 72 31.4 41.3 *** 41.8 *** 4.8
29 3.4 7.2 4.6 2 59 25.1 13.1 *** 13.8 *** 5.8
30 3.9 4.5 4.1 5 16 10.7 16.5 0.0 4.1
31 4.3 5.6 4.8 14 44 24.7 0.0 12.2 4.9
32 4.0 6.1 5.0 16 56 30.8 33.4 *** 32.3 *** 5.4
33 2.5 3.4 2.9 11 47 21.6 40.9 *** 40.2 (u) *** 3.2
34 3.8 5.7 4.7 3 29 14.1 35.4 *** 61.6 *** 5.3
35 4.6 7.3 6.0 12 21 16.9 9.1 7.5 5.9
36 3.0 5.4 4.0 13 46 23.0 75.2 *** 74.1 *** 5.0
37 2.9 6.7 4.8 8 44 20.8 2.7 0.0 5.3
38 2.1 5.6 3.8 5 34 13.8 29.9 *** 29.4 *** 4.6
39 2.8 6.1 4.0 3 57 32.4 72.8 *** 75.9 *** 5.5
40 4.8 5.8 5.3 25 42 34.5 60.1 40.7 5.0
When calculating the mean soil pH of the five plots with the highest species number, the
emerging pattern was consistent (Table 2; Fig. 2): For forests, grasslands, mires and sand, the
maximum species density was found in the middle section (pH 4.00­5.99), i.e., on sites with
moderately high pH, followed by the high pH section (pH 6.00­8.00). In only a few cases was
species number at its maximum on low-pH sites. Across all 40 studies, maximum species
density was on average reached at pH 5.4, and of 26 studies with maximum pH values
exceeding 6, only 9 had values of maximum species richness at pH  6.
DISCUSSION
Effects of varying plot size
Plot size was positively correlated with species density in less than half of the studies
(Table 1). This supports the observation made by CHYTRÝ (2001) who assumed that
researchers, especially those with a strong interest in phytosociology, tend to use larger plots
in species-poor vegetation to include more taxa of interest or potential value as diagnostic
species. However, there appeared to be no systematic bias as to the relationship between the
372 B. Schuster & M. Diekmann
8765432
60
50
40
30
20
10
0
pH
Speciesdensity
39*
40
18*
37
29*
12*
36*
32*
15*
33*
6*
Forest
2 3 4 5 6 7 8
0
10
20
30
40
50
60
pH
Speciesdensity
4*
26
24*
27
14*
28*
3125
3
23*
22*
8*
10*
Grassland
16*
8765432
60
50
40
30
20
10
0
pH
Speciesdensity
17
9*
2*
11*
5 30
735
38*
1*
Mire
2 3 4 5 6 7 8
0
10
20
30
40
50
60
pH
Speciesdensity
20
13*
19*
21
34*
Sand
Fig. 1. Relationship between species density and soil pH as assessed by regression analysis, shown separately
for the four formations: forest, grassland, mire and sand. Each regression line denotes one study, referred to by
its number (see Appendix 1). For regression statistics, see Table 1. Significant fits are marked with an asterisk
added to the study no. In most cases, pH (H2O) was measured; studies based on pH (KCl or CaCl2) that were
moved 0.62 pH units to the right to be comparable with pH (H2O) measurements are indicated by dashed lines
(see Material and Methods).
plot size chosen and pH, which justifies the use of phytosociological relevés for our purpose,
in absence of more standardized data.
Relationship between species density and pH
In all formations we found a distinct and sometimes steep increase in species number from
low to moderately high pH. This is in accordance with the literature (e.g. DUPRÉ et al. 2002)
and reflects the poor diversity of the most acidic and, at the same time, infertile sites. On soils
of higher pH, however, the formations differed from each other in their relationship between
species density and pH.
For forests, the two ways of examining this relationship gave slightly different results.
Whereas the regression analysis showed that species density generally continues to increase
even on sites with weakly acid or neutral pH of 6 to 7, the analysis of maximum species
density indicated that the top values are found on moderately acidic soils of around pH 5­6.
We believe that the latter pattern would have been even stronger if data had been available
from the most base-rich forest sites that are absent from or clearly under-represented in our
survey, namely beech and mixed deciduous forests on dry calcareous slopes (Cephalanthero-Fagion)
and mixed deciduous forests with Acer, Fraxinus, Tilia and Ulmus on highly fertile
slopes and glens (Tilio-Acerion). The first type has especially been associated with high
floristic diversity (ELLENBERG 1996, and DIEKMANN, unpubl. data). This result of our survey
Species density along the pH gradient in Germany 373
Forest
0
20
40
60
80
100
low medium high
pH
Speciesdensity
Grassland
0
20
40
60
80
low medium high
pH
Speciesdensity
Mire
0
20
40
60
80
low medium high
pH
Speciesdensity
Sand
0
20
40
60
80
low medium high
pH
Speciesdensity
Fig. 2. pH of maximum species density in relation to the pH gradient length. For each study, the mean pH of the
five sites with the highest species number was calculated. After dividing the whole pH range into equally long
sections of low (2.00­3.99), middle (4.00­5.99) and high (6.00­8.00), the section comprising the mean pH was
determined. The proportion of studies assigned to the three sections in each formation is given on the y axis.
is also in contrast to the findings of DUPRÉ et al. (2002), who, based on forest data mainly
from Scandinavia, reported a predominance of unimodal relationships between species
density and pH. More reliable data from forests are needed to understand if the relationship
between species richness and pH varies consistently between different forest subtypes, or if
there simply is no consistent pattern. The slopes of the regression lines found in forests were
or tended to be higher than in the other three formations, indicating that the highest species
richness in forests is reached at comparatively high pH.
In accordance, the regression lines in grasslands in most cases had their optima at a
moderately high pH of 5­6, a result that was supported also by the analysis of pH of maximum
species number. This was somewhat unexpected, as the survey included dry calcareous
grasslands of the class Festuco-Brometea occurring on high-pH soils that are well-known for
their high species number. Mires also tended to have their highest species density at
intermediate pH, although the patterns were less clear and the changes in diversity along the
acidity gradient generally less pronounced than in grasslands. Because species density
included both vascular plants and cryptogams, mainly bryophytes, and because the latter are
of particularly great importance in mires with respect to cover and species number, we
assumed that the numbers of bryophytes and vascular plants were reciprocal to each other.
However, this was not the case: in those studies that allowed us to test this assumption, species
densities of the two groups were significantly positively correlated.
The species density-pH relationship in the sand formation was difficult to assess and
interpret, partly due to the small number of studies included. In the following, the discussion
will therefore focus on forests, grasslands and mires.
Species pools and local species richness
As there are physiological pH limits to the survival of plants, species densities beyond
these limits must be zero (PÄRTEL 2002). This implies that, close to the upper pH limit of
about 9, species density-pH relationships are expected to be negative. This may be relevant in
grasslands, where pH exceeded 8 in several studies and where, except in one case, richness
decreased between pH 7 and 8. On such alkaline soils, species suffer from the low availability
of iron, manganese and phosphate and their incapacity to render these soluble for uptake
(TYLER 1999). However, species density in grasslands in most cases was at maximum at pH 5
to 6.5, i.e., at somewhat or considerably higher acidity, and this also applies to mires where pH
values often exceeded 7. Thus, it is unlikely that the predominance of hump-backed curves
found in grasslands and mires is solely explained by the physiological tolerances of species. In
contrast, the often sharp decrease in species density on increasingly acidic soils found in most
formations may be primarily explained by the small pool of species physiologically capable
of surviving the high hydrogen ion concentration and overcoming the Al3+
toxicity of these
soils (cf. TYLER 1999).
Our results contradict the general statement by EWALD (2003) that there is a "positive
relationship between pH and plant species density throughout the temperate zone". Even
though this survey shows that the relationship is more positive than negative (when only
linear regression is applied, the large majority of studies had positive slopes), at high pH it is
in fact no longer positive. The results of the analysis of pH of maximum species density also
374 B. Schuster & M. Diekmann
give evidence for the hump-backed pattern. Thus, the large pool of calciphilous species
present in the flora of Central Europe may explain the higher floristic diversity of the most
base-rich sites compared to the most acidic sites, but it cannot explain the high species density
found in sites with intermediate pH. Here, we attempt to offer some possible explanations that
are not mutually exclusive:
(a) The term species pool was actually coined for communities affected by many
environmental variables at the same time, not by a single one such as pH. There may be
confounding effects of other factors that are correlated with pH across different environments.
If high pH were associated with a certain value of another variable, for example low light
availability in forests for which the species pool is low, then species density on such sites
would be low despite the large pool of calciphilous species. The importance of analyzing and
interpreting such habitat-specific species pools was emphasised by DUPRÉ et al. (2002).
However, the problem remains of how to identify or calculate the species pools of
communities in a statistically sound way (ERIKSSON 1993, PÄRTEL et al. 1996, DUPRÉ 2000,
GRACE 2001).
(b) This explanation is somewhat related to the one presented under (a). Productivity
increases with decreasing soil acidity from low to intermediate pH, whereas the pattern
becomes more complicated at higher pH. For example, base-rich forest sites can both be
highly productive (Tilio-Acerion communities) and relatively unproductive (Cephalanthero-Fagion,
cf. ELLENBERG 1996). In some formations such as grasslands, high pH sites are
probably often more productive than soils with intermediate pH, although we do not know of
any systematic study of the relationship between the two variables across different
formations. If this holds true, the relationship between species density and pH may interfere
with the hump-backed relationship between species density and productivity (GRIME 1979).
(c) The overlap of pH amplitudes of species with different indicator values, hypothetically
shown by EWALD (2003) in Fig. 2, implies another explanation: as the indicator values for
reaction denote the supposed species optima relative to pH, the high number of species with
scores between 7 and 9 does not necessarily result in high species densities on high-pH sites,
because species density is first of all dependent on the number of species tolerant of these
conditions. At pH of about 5 to 6, both acidophilous and calciphilous species may be able to
(co-)exist, while at lower or higher pH the numbers of tolerant species decline. In other words,
intermediate pH conditions exclude only few species from growing there, and therefore we
may argue that the species pool in such conditions is large. The explanation relates to the
mid-domain effect described by COLWELL & LEES (2000).
It may be concluded that the large pool of calciphilous species in Central Europe certainly
affects the small-scale richness of many plant communities on high-pH soils, but that it cannot
fully account for the patterns of hump-backed species density-pH relationships found in this
study, especially in grasslands and mires. We believe that the decline in species richness on
increasingly base-rich soils partly results from the correlation of pH with other environmental
factors across the landscape. To get a clearer view of the issues raised by EWALD (2003) we
need to collect more measurement data of pH and other variables accompanied by species
richness data. The collection of these data should be made with more or less standardized
sampling methods to avoid the drawbacks of the phytosociological approach in this context.
Species density along the pH gradient in Germany 375
Acknowledgements: We thank Marlies Rückmann and Marion Ahlbrecht for compiling much of the data on
species densities and soil pH. Toby Spribille and an unknown reviewer made many helpful comments to improve
the paper.
REFERENCES
BAUMANN K. (1996): Kleinseggenriede und ihre Kontaktgesellschaften im westlichen Unterharz
(Sachsen-Anhalt). Tuexenia 16: 151­177.
BAUMANN K. (2000): Vegetation und Ökologie der Kleinseggenriede des Harzes. Cuvillier Verlag, Göttingen.
BEER W. (1995): Methodische und standortsökologische Untersuchungen zum Nährstoffumsatz im Grünland.
Diss. Bot. 242: 1­216.
BOHNER A., SOBOTIK M. & ZECHNER L. (2001): Die Iris-Wiesen (Iridetum sibiricae PHILIPPI 1960) im
Mittleren Steirischen Ennstal (Steiermark, Österreich) ­ Ökologie, Soziologie und Naturschutz. Tuexenia 21:
133­151.
BORCHERT R. (1990): Artenkombination und Standorte des Rhynchosporetum in der westfälischen Bucht
(Nordrhein-Westfalen, BRD). Acta Biol. Bendrodis 2: 1­18.
BRAND J. (2000): Untersuchungen zur synsystematischen Umgrenzung und Untergliederung sowie zur
standörtlichen und landschaftsräumlichen Bindung von Feuchtwäldern im nordwestdeutschen Tiefland.
Diss. Bot. 323: 1­344.
BRAUN W. (1968): Die Kalkflachmoore und ihre wichtigsten Kontaktgesellschaften im bayerischen
Alpenvorland. Diss. Bot. 1: 1­134.
BRUELHEIDE H. (1995): Die Grünlandgesellschaften des Harzes und ihre Standortsbedingungen. Mit einem
Beitrag zum Gliederungsprinzip auf der Basis von statistisch ermittelten Artengruppen. Diss. Bot. 244:
1­338.
BRUELHEIDE H. & FLINTROP T. (1994): Ordnen von Vegetationstabellen nach Arten-Aufnahmen-Gruppen mit
dem Programm ESPRESSO. Tuexenia 14: 493­502.
CHYTRÝ M. (2001): Phytosociological data give biased estimates of species richness. J. Veg. Sci. 12: 439­444.
COLWELL R.K. & LEES D.C. (2000): The mid-domain effect: geometric constraints on the geography of species
richness. Trends Ecol. Evol. 15: 70­76.
DÖRING-MEDERAKE U. (1991): Pflanzensoziologische Gliederung und Standortsbedingungen von Bruch-,
Quell- und Uferwäldern im niedersächsischen Tiefland. Dissertation, Universität Göttingen, Göttingen.
DUPRÉ C. (2000): How to determine a regional species pool: a study in two Swedish regions. Oikos 89: 128­136.
DUPRÉ C., WESSBERG C. & DIEKMANN M. (2002): Species richness in deciduous forests: Effects of species
pools and environmental variables. J. Veg. Sci. 13: 505­516.
ELLENBERG H. (1996): Vegetation Mitteleuropas mit den Alpen. Ed. 5. Verlag Eugen Ulmer, Stuttgart.
ERIKSSON O. (1993): The species-pool hypothesis and plant community diversity. Oikos 68: 371­374.
EWALD J. (2003): The calcareous riddle: Why are there so many calciphilous species in the Central European
flora? Folia Geobot. 38: 357­366 (this issue).
FISCHER P. (1998): Sandtrockenrasen von Binnendünen in der Unteren Mittelelbe-Niederung zwischen Dömitz
und Boizenburg. Tuexenia 18: 119­151.
GOEBEL W. (1995): Die Vegetation der Wiesen, Magerrasen und Rieder im Rhein-Main-Gebiet. Diss. Bot. 237:
1­533.
GÖNNERT T. (1989): Ökologische Bedingungen verschiedener Laubwaldgesellschaften des Nordwestdeutschen
Tieflandes. Diss. Bot. 136: 1­224.
GRACE J.B. (1999): The factors controlling species density in herbaceous plant communities: an assessment.
Perspect. Pl. Ecol. Evol. Syst. 2: 1­28.
GRACE J.B. (2001): Difficulties with estimating and interpreting species pools and the implications for
understanding patterns of diversity. Folia Geobot. 36: 71­83.
GRIME J.P. (1979): Plant strategies and vegetation processes. John Wiley, Chichester.
HACHMÖLLER B. (2000): Vegetation, Schutz und Regeneration von Bergwiesen im Osterzgebirge. Diss. Bot.
338: 1­300.
HARM S. (1990): Kleinseggenriede (Scheuchzerio-Caricetea fuscae) im Südwest-Harz. Tuexenia 10: 173­183.
376 B. Schuster & M. Diekmann
HÄRDTLE W. (1995): Vegetation und Standort der Laubwaldgesellschaften (Querco-Fagetea) im nördlichen
Schleswig-Holstein. Mitt. Arbeitsgem. Geobot. Schleswig-Holstein & Hamburg 48: 1­441.
HEINKEN T. (1990): Pflanzensoziologische und ökologische Untersuchungen offener Sandstandorte im
östlichen Aller-Flachland (Ost-Niedersachsen). Tuexenia 10: 223­257.
HEMP A. (2001): Die Dolomitsand-Trockenrasen (Helichryso-Festucetum) der Frankenalb. Tuexenia 21:
91­112.
ISERMANN M. (1997): Vegetations- und standortkundliche Untersuchungen in Küstendünen Vorpommerns.
Dissertation, Universität Greifswald, Greifswald.
JANDT U. (1999): Kalkmagerrasen am Südharzrand und im Kyffhäuser. Diss. Bot. 322: 1­246.
KLESCZEWSKI M. (2000): Die Glatthaferwiesen im Bergmassiv des Mont Aigoual (Cevennen, Südfrankreich).
Tuexenia 20: 189­212.
KRATZERT G. & DENGLER J. (1999): Die Trockenrasen der Gabower Hänge am Oderbruch. Berliner
Botanischer Verein 132: 285­330.
LIENKAMP G. (1993): Grundwasseruntersuchungen ausgewählter Grünlandbestände des mittleren Ostetals.
Diploma thesis, Universität Bremen, Bremen.
LISBACH I. & PEPPLER-LISBACH C. (1996): Magere Glatthaferwiesen im Südöstlichen Pfälzerwald und im
Unteren Werraland. ­ Ein Beitrag zur Untergliederung des Arrhenatheretum elatioris BRAUN 1915.
Tuexenia 16: 311­336.
MEHRTENS L. (1995): Synmorphologische und bodenkundliche Untersuchungen an den salzbeeinflussten
Grünland- und Röhrichtgesellschaften des Spiekaer Sommergrodens. Unpubl. Diploma thesis, Universität
Bremen, Bremen.
PÄRTEL M. (2002): Local plant diversity patterns and evolutionary history at the regional scale. Ecology 83:
2361­2366.
PÄRTEL M., ZOBEL M., ZOBEL K. & VAN DER MAAREL E. (1996): The species pool and its relation to species
richness: evidence from Estonian plant communities. Oikos 75: 111­117.
PEPPLER C. (1992): Die Borstgrasrasen (Nardetalia) Westdeutschlands. Diss. Bot. 193: 1­402.
PFLUME S. (1999): Laubwaldgesellschaften im Harz: Gliederung, Ökologie und Verbreitung. Arch. Naturwiss.
Diss. 9: 1­238.
PIETSCH W. (1986): Vegetationsverhältnisse und ökologische Situation im NSG Sauerkopfmoor, Kreis Gotha.
Arch. Naturschutz Landschaftsf. 26: 19­47.
ROSENTHAL G. (1991): Erhaltung und Regeneration von Feuchtwiesen. Vegetationskundliche Untersuchungen
auf Dauerquadraten. Dissertation, Universität Bremen, Bremen.
SCHAFFERS A.P. & SÝKORA K.V. (2000): Reliability of Ellenberg indicator values for moisture, nitrogen and
soil reaction: a comparison with field measurements. J. Veg. Sci. 11: 225­244.
SCHÖNERT T. (1994): Die Bruchwälder des westlichen Rheinischen Schiefergebirges. Diss. Bot. 228: 1­143.
STORM C. (1996): Analyse der Beziehungen zwischen Standort und Vegetation mit der logistischen Regression
­ eine Studie in Nadelwäldern des Schwarzwaldes. Phytocoenologia 26: 273­412.
TÄUBER T. (1994): Vegetationsuntersuchungen auf einem Panzerübungsgelände im Naturschutzgebiet
Lüneburger Heide. Tuexenia 14: 197­228.
TÄUBER T. (2002): Zwergbinsen-Gesellschaften (Isoeto-Nanojuncetea) in Niedersachsen. Cuvillier Verlag,
Göttingen.
TAUX K. (1981): Wald- und Forstgesellschaften der Rasteder Geest. Oldenburger Jahrb. 81: 325­380.
TYLER G. (1997): Soil chemistry and plant distribution in rock habitats of southern Sweden. Nord. J. Bot. 16:
609­635.
TYLER G. (1999): Plant distribution and soil-plant interactions on shallow soils. In: RYDIN H., SNOEIJS P. &
DIEKMANN M. (eds.), Swedish plant geography, Acta Phytogeogr. Suec. 84: 21­32.
WIEBE C. (1989): Ökologische Charakterisierung von Erlenbruchwäldern und ihren Entwässerungsstadien:
Vegetation und Standortverhältnisse. Mitt. Arbeitsgem. Geobot. Schleswig-Holstein & Hamburg 56: 1­156.
WITTIG B. (1999): Vegetationskundliche und ökologische Untersuchungen über nordwestdeutsche
Kleinseggen-Sümpfe, insbesondere im Hinblick auf Gefährdung und Regenerationspotentiale. Diss. Bot.
310: 1­207.
Species density along the pH gradient 377
WULF M. (1985): Vegetationskundliche Untersuchungen ausgewählter Laubwaldgesellschaften auf der
Beverstedter Geest unter Berücksichtigung von Naturschutzaspekten. Unpubl. Diploma thesis, Universität
Bremen, Bremen.
WULF M. (1992): Vegetationskundliche und ökologische Untersuchungen zum Vorkommen gefährdeter
Pflanzenarten in Feuchtwäldern Nordwestdeutschlands. Diss. Bot. 185: 1­246.
ZOBEL M. (1997): The relative role of species pools in determining plant species richness: an alternative
explanation of species coexistence? Trends Ecol. Evol. 12: 266­269.
ZOBEL M., VAN DER MAAREL E. & DUPRÉ C. (1998): Species pool: the concept, its determination and its
significance for community restoration. Appl. Veg. Sci. 1: 55­66.
Received 4 January 2003, revision received and accepted 16 September 2003
Encl. Appendix p. 379
378 B. Schuster & M. Diekmann
APPENDIX 1
Survey of references from which data were compiled. The studies were ordered alphabetically and referred to in
the text by their numbers. The vegetation data were assigned to four broadly defined types (see text) and consist
either of plots or groups of plots (Clusters). Plot size differed between studies and partly also within studies
(then, range and mean are given; means are in parentheses); in some cases, no information on plot size was
available. pH was measured either in water (H2O) or in CaCl2/KCl solution.
No. Reference Vegetation Data type No. of plots/ Plot size Type of
of study clusters (m2
) pH data
1 BAUMANN (1996) Mire Plots 66 4­16 (15) H2O
2 BAUMANN (2000) Mire Clusters 14 4 H2O
3 BEER (1995) Grassland Plots 8 no data H2O
4 BOHNER et al. (2001) Grassland Plots 29 no data CaCl2
5 BORCHERT (1990) Mire Clusters 9 no data H2O
6 BRAND (2000) Forest Plots 167 200­400 (100) H2O
7 BRAUN (1968) Mire Plots 126 0.5­100 (13) H2O
8 BRUELHEIDE (1995) Grassland Plots 127 10 H2O
9 BRUELHEIDE (1995) Mire Plots 44 10 H2O
10 BRUELHEIDE & FLINTROP (1994) Grassland Plots 36 no data H2O
11 BRUELHEIDE & FLINTROP (1994) Mire Plots 42 no data H2O
12 DÖRING-MEDERAKE (1991) Forest Clusters 8 100­400 H2O
13 FISCHER (1998) Sand Plots 58 2­12 (5) H2O
14 GOEBEL (1995) Grassland Plots 168 15­70 KCl
15 GÖNNERT (1989) Forest Plots 36 100­500 H2O
16 HACHMÖLLER (2000) Grassland Plots 30 16­25 KCl
17 HARM (1990) Mire Plots 15 5­15 (10) H2O
18 HÄRDTLE (1995) Forest Clusters 17 100 H2O
19 HEINKEN (1990) Sand Plots 22 8­53 (18) H2O
20 HEMP (2001) Sand Plots 52 4­75 (17) CaCl2
21 ISERMANN (1997) Sand Clusters 17 1­25 KCl
22 JANDT (1999) Grassland Plots 274 3­16 H2O
23 KLESCZEWSKI (2000) Grassland Plots 101 no data H2O
24 KRATZERT & DENGLER (1999) Grassland Plots 84 10 H2O
25 LIENKAMP (1993) Grassland Clusters 13 no data CaCl2
26 LISBACH & PEPPLER-LISBACH
(1996) Grassland Plots 99 15­20 H2O
27 MEHRTENS (1995) Grassland Plots 14 no data CaCl2
28 PEPPLER (1992) Grassland Plots 231 4­40 (18) H2O
29 PFLUME (1999) Forest Plots 387 100­400 H2O
30 PIETSCH (1986) Mire Clusters 6 0.4­60 H2O
31 ROSENTHAL (1991) Grassland Clusters 14 25 CaCl2
32 SCHÖNERT (1994) Forest Plots 47 80­100 H2O
33 STORM (1996) Forest Plots 77 300 CaCl2
34 TÄUBER (1994) Sand Clusters 16 9­22 (16) H2O
35 TÄUBER (2002) Mire Clusters 20 1­2 H2O
36 TAUX (1981) Forest Clusters 13 no data H2O
37 WIEBE (1989) Forest Clusters 20 no data KCl
38 WITTIG (1999) Mire Plots 95 1­30 (18) CaCl2
39 WULF (1985) Forest Plots 20 100­400 CaCl2
40 WULF (1992) Forest Clusters 5 100­200 CaCl2
Species density along the pH gradient 379