Full text of DOI:10.1111/j.1654-109X.2002.tb00540.x

Applied Vegetation Science 5: 107-120, 2002
IAVS; Opulus Press Uppsala. Printed in Sweden
EFFECTS OF LONG-TERM CUTTING IN A GRASSLAND SYSTEM: PERSPECTIVES FOR RESTORATION OF PLANT COMMUNITIES- 107
©
-
Effects of long-term cutting in a grassland system: perspectives for
restoration of plant communities on nutrient-poor soils
1
*
2
1
Bakker, J.P. ; Elzinga, J.A. & de Vries, Y.
1Laboratory of Plant Ecology, University of Groningen, P.O.Box 14, 9750 AA Haren, The Netherlands;
2
Present address: Department of Plant Population Biology, Netherlands Institute of Ecology NIOO-CTO, P.O.Box 40,
*
6
666 ZG Heteren, The Netherlands; Corresponding author; Fax +31503632273; E-mail j.p.bakker@biol.rug.nl
Abstract. The upper courses of brook valley systems harbour
Nardo-Galion saxatilis communities characteristic of oligo-
trophic soils under low-intensity farming. Most of these com-
munities have disappeared under intensified farming i.e. appli-
cation of fertilizers. We studied the possibilities of restoration
i.e. re-establishment of the former plant community by adopt-
ing various cutting regimes after the cessation of fertilization
in 1972. The various cutting regimes revealed different effects
after 25 yr. Regimes with cutting every second year with or
without removal of the swath, and complete abandonment
deviated from the other regimes that included annual haymak-
ing with different frequency and timing. The latter group of
cutting regimes came closer to the community of an adjacent
field where fertilization stopped in 1967. This field in turn
harboured several Nardo-Galion species after 25 yr of annual
cutting, and showed more resemblance with a local reference
community (at a distance of 500 m) that had not been fertilized
since the 1940s. The local reference still does not match
poorly developed Nardo-Galion saxatilis communities found
in the region of ca. 50 km around the study area, and is far from
well developed Nardo-Galion communities in the same re-
gion.
Introduction
Low-intensity farming systems (Bignal & Mc-
Cracken 1996) are disappearing over large areas. These
systems are characterized by a low nutrient input and
low outputs by cutting for hay and extensive grazing.
Intensification by increased nutrient input is the main
reason for their disappearance. As a result communities
harbouring plant species characteristic of oligotrophic
(nutrient-poor) soil conditions such as the Nardo-Galion
saxatilis (Schaminée et al. 1996) have almost disap-
peared in The Netherlands as they have been turned into
high productive stands with a low species richness, and
species typical of nutrient-rich soils (Klapp 1965).
To restore the Nardo-Galion saxatilis on fertilized
sites implies two goals: (1) a reduction in soil nutrient
conditions responsible for the high plant productivity in
the present intensive farming system, and (2) a return of
the plant species typical of the target community.
Soil nutrient conditions can be measured directly by
soil chemical analysis or indirectly from peak standing
crop and indicator values of constituent grassland plant
species (Schaffers & Sýkora 2000). After the cessation
of fertilization, a decrease of peak standing crop may be
expected (Berendse et al. 1992). Moreover, species in-
dicating nutrient-rich soil conditions may decrease,
whereas species indicating nutrient-poor soil conditions
are likely to increase, provided these species can estab-
lish in the target area. We hypothesize that management
regimes including removal of nutrients by a haymaking
regime will result in a quicker decrease of peak standing
crop and species indicating nutrient-rich soil conditions
than regimes with irregular haymaking or mulching
with less nutrients removal.
The study site still harbours several species characteristic
of eutrophic soil and few species characteristic of oligotrophic
soil after 25 yr of annual cutting and removal of the swath. The
soil seed bank harbours only few target species. Although
species characteristic of oligotrophic soil are present in an
adjacent field and Nardo-Galion saxatilis species occur at 500
m, they have not (yet) established in the target area.
Keywords: Grassland management; Long-term monitoring;
Nardo-Galion saxatilis; Permanent plot; Seed bank; Standing
crop, Target species.
Nomenclature: van der Meijden (1990).
Many species characteristic of the target Nardo-
Galion saxatilis community are not known for a long-
term persistent seed bank (Thompson et al. 1997). There-
fore, it is unlikely that they are still part of the commu-
nity pool sensu Zobel et al. (1998). Thus these species
have to reach the target area through dispersal from

1
08
BAKKER, J.P. ET AL.
elsewhere. We hypothesize that the soil seed bank in
this area harbours only few target species, and that
dispersal and subsequent establishment will be difficult
in the present fragmented landscape, despite the pres-
ence, often at short distance, in the local or regional
species pool sensu Zobel et al. (1998).
Several cutting regimes were imposed to study to
which extent they could render different plant commu-
nities from an initially similar stand within a single field
with the same ‘agricultural history’. We hypothesized
that (1) abandonment would result in establishment of
shrubs and trees; (2) mulching i.e. cutting without re-
moval of the swath in September every 2 yr, and haymak-
ing in September every 2 yr would result in tall forb
communities; (3) 5 annual haymaking regimes, i.e. with
removal of the swath, different in frequency and in
timing would affect the nutrient output and hence stand-
ing crop and the rate of transformation into stands with
species characteristic of nutrient-poor soil conditions;
by a sandy subsoil with little accumulation of organic
matter. The hydrological conditions feature infiltration
in the source area and some superficial seepage water
that is relatively acid in the upper course area (Grootjans
1980). Duration lines of the dynamics of the groundwater
level are given by Bakker & de Vries (1985a). The
source area and upper course harboured plant communi-
ties characteristic of nutrient-poor soil conditions such
as Nardo-Galion saxatilis (Schaminée et al. 1996) which
is similar to the Violion caninae (Oberdorfer 1978).
According to the National Vegetation Classification in
the United Kingdom, Nardo-Galion includes Festuca-
Agrostis-Galium grassland (U4) and Nardus-Galium
grassland (U5) (Rodwell 1992). These communities
occurred under low-intensity farming regimes with com-
mon grazing and no input of fertilizer. After subdivision
of common grazing land, individual fields could also be
cut for hay. Yields were very low and did not surpass
150 g m dry weight (Klapp 1965). Increased fertiliza-
tion since the 1950s (Bakker & Olff 1995) transformed
most of these communities into stands characteristic of
nutrient-rich soil conditions (Klapp 1965; Bakker 1989).
The State Forestry Commission gradually acquired
fields in this nature reserve. One field in the Loefvledder
was acquired in 1972 and fertilization stopped in that
year (referred to in the text as ‘new field’). The vegeta-
tion in 1973 was dominated by grasses with high abun-
dance such as Agrostis stolonifera, Alopecurus genicu-
latus, Poa trivialis, Festuca pratensis and less abundant
species such as Lolium perenne, Holcus lanatus, Poa
pratensis and the herbs Ranunculus repens, Cardamine
pratensis and Rumex acetosa (Table 1, col. 1) (see also
Fig. 2). Species characteristic of both nutrient-rich and
intermediate soil conditions were present.
–
2
(4) annual haymaking both in July and September would
show the largest output of nutrients.
Methods
Study area and cutting regimes
The study site Loefvledder (53∞ 05' N, 6∞ 40' E), is at
the tributary of the Anloër Diepje (Fig. 1) and is part of
a larger brook valley system in the nature reserve
‘Stroomdallandschap Drentsche A’ at the sandy Pleisto-
cene Plateau of Drenthe (The Netherlands). It is located
in the upper course part (Everts & de Vries 1991) of the
brook valley at 7 m above sea level and is characterized
Several cutting regimes were imposed from 1973
onwards to study the extent they could render different
plant communities from an initially similar stand within a
single field with the same ‘agricultural history’ (Fig. 2).
A regime with burning every second year in winter was
also established, but it was difficult to perform this
practice at a regular basis, and although it is included in
the DCA analysis, it is omitted from further presenta-
tion. Detailed descriptions of the vegetation during the
first 15 yr of the regimes were published earlier (Bakker
&
de Vries 1985a, b; Bakker 1989).
An adjacent field was acquired in 1967 and fertiliza-
tion stopped in the same year (referred to further in the
text as ‘old field’) (Fig. 1). The vegetation was analysed
for the first time in 1973. In that year the vegetation was
characterized by the abundance of the grasses Festuca
rubra and Anthoxanthum odoratum. Also species such
as Carex nigra and Luzula campestris, characteristic of
nutrient-poor soil conditions, were found (Table 1, col.
4). The field was annually cut for hay in July.
Fig. 1. The Loefvledder study area indicating the ‘new’ and
‘old’ fields and the ‘local reference’.

-
EFFECTS OF LONG-TERM CUTTING IN A GRASSLAND SYSTEM: PERSPECTIVES FOR RESTORATION OF PLANT COMMUNITIES- 109
Table 1. Species composition of the vegetation of different regimes in ‘new’ with (hay) and without (no hay) annual haymaking, and
old’ fields of Loefvledder, the local reference (Loc) and ‘poorly’ (Reg p) and ‘well developed regional’ (Reg w) reference of the
‘
Nardo-Galion saxatilis. Indicated are the frequency of occurrence, mean cover percentage in parentheses, longevity index (LI, see
text) and indicator with respect to nutrient-rich, intermediate and nutrient-poor soil conditions (R, I and P, respectively, see text).
New ’73
n = 16
New no hay ’95 New hay ’95
Old ’73
n = 2
Old ’95
n = 2
Loc
n = 5
Reg p
n = 5
Reg w
n = 6
LI
Species
n = 6
n = 10
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
.6
Alopecurus geniculatus
Lolium perenne
R
R
R
R
R
R
R
I
94 (2)
0
0
0
0
0
0
.2
62 (6)
0
0
0
0
0
0
0
.33
.69
.54
.3
.65
.44
.12
.24
.28
.42
.09
.48
.56
.28
Cirsium arvense
12 (1)
0
0
0
0
0
0
0
0
0
Rumex crispus
12 (1)
33 (1)
0
0
0
0
Ranunculus repens
94 (5)
100 (12)
100 (24)
100 (2)
40 (1)
100 (11)
60 (2)
80 (2)
100 (4)
100 (1)
50 (1)
0
100 (20)
100 (3)
0
0
0
0
Poa pratensis
94 (10)
67 (1)
67 (1)
0
20 (1)
0
0
Poa trivialis
87 (11)
100 (16)
0
0
0
0
Holcus lanatus
100 (37)
100 (1)
60 (1)
60 (5)
0
Festuca rubra
I
87 (3)
62 (1)
6 (1)
83 (14)
100 (6)
100 (4)
20 (10)
0
Rumex acetosa
I
I
R
R
I
I
R
100 (2)
100 (1)
40 (1)
20 (4)
40 (3)
20 (1)
17 (4)
Anthoxanthum odoratum
Agrostis stolonifera
Festuca pratensis
Cardamine pratensis
Cerastium fontanum
Trifolium repens
Taraxacum spec.
Ranunculus acris
Equisetum arvense
Potentilla anserina
Scrophularia nodosa
Galeopsis tetrahit
Elymus repens
0
100 (13)
100 (10)
100 (1)
100 (7)
100 (18)
33 (4)
100 (20)
50 (1)
0
0
0
0
100 (15)
0
0
40 (1)
0
0
0
0
0
0
0
0
94 (1)
60 (1)
0
0
0
56 (1)
33 (1)
20 (1)
0
0
20 (1)
0
44 (1)
0
0
100 (1)
0
0
0
0
37 (1)
0
40 (1)
100 (1)
0
0
20 (1)
0
0
.25
I
12 (1)
12 (1)
6 (1)
0
0
80 (1)
0
0
0
0
0
R
R
R
R
R
I
0
0
0
0
0
0
17 (2)
0
0
0
0
0
.17
.88
.41
.16
.23
83 (4)
40 (1)
0
0
0
0
0
17 (2)
0
0
0
0
0
0
0
0
17 (1)
0
0
0
0
0
0
17 (1)
0
0
0
0
0
0
Galium palustre
0
17 (1)
0
0
0
0
0
0
Polygonum hydropiper
I
0
17 (1)
0
0
0
0
0
0
0
0
Holcus mollis
P
R
P
I
0
17 (1)
0
0
0
0
0
0
0
.07
Heracleum sphondylium
Rhinanthus angustifolius
Lotus uliginosus
0
33 (1)
0
0
0
0
0
0
0
17 (1)
90 (11)
0
0
0
0
0
50 (2)
0
0
0
100 (1)
0
0
0
0
0
0
0
0
0
.5
.3
.69
.93
Epilobium obscurum
Cirsium palustre
Agrostis capillaris
Juncus effusus
0
50 (1)
0
0
0
0
P
P
0
17 (1)
100 (1)
0
0
0
0
0
83 (16)
100 (11)
50 (3)
30 (1)
0
100 (4)
100 (17)
80 (7)
17 (4)
33 (2)
33 (2)
0
0
33 (41)
100 (1)
100 (9)
60 (2)
0
Quercus robur
0
17 (1)
0
0
0
50 (1)
100 (1)
20 (1)
0
0
0
.38
.12
.39
Plantago lanceolata
Leontodon autumnalis
Hypochaeris radicata
Equisetum palustre
Rumex acetosella
Juncus conglomeratus
Carex nigra
I
I
0
90 (17)
0
0
0
0
0
40 (1)
50 (1)
0
0
0
0
100 (1)
100 (1)
50 (1)
50 (1)
0
0
40 (3)
50 (2)
33 (2)
17 (2)
P
P
0
P
P
I
0
0
20 (1)
0
0
80 (2)
0
0
20 (1)
0
0
0
0
0
0
0
0
0
0
0
0
0
0
.71
.4
.17
.38
.18
.53
0
10 (4)
0
0
60 (2)
0
0
10 (1)
0
0
0
0
17 (2)
0
0
0
100 (3)
50 (1)
50 (1)
80 (4)
Carex ovalis
0
0
20 (1)
Luzula campestris
Lychnis flos-cuculi
Prunus spec.
Galium saxatile
Agrostis canina
Nardus stricta
Betula pubescens
Hydrocotyle vulgaris
Viola palustris
Luzula multiflora
Potentilla erecta
Molinia caerulea
Deschampsia caespitosa
Achillea millefolium
Agrostis vinealis
Aira praecox
P
I
0
0
0
100 (3)
40 (7)
33 (2)
0
0
0
0
0
0
0
0
0
0
0
0
0
0
50 (1)
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
.32
.5
.18
.44
P
P
P
P
P
P
P
P
P
P
I
P
P
P
P
P
P
P
P
P
P
P
P
P
P
P
P
I
0
0
0
100 (1)
100 (3)
80 (31)
60 (5)
60 (15)
0
50 (5)
33 (4)
50 (3)
33 (6)
0
0
0
0
100 (11)
40 (3)
0
0
0
50 (1)
0
0
0
0
50 (1)
0
0
0
0
50 (1)
0
0
0
.17
.7
.34
.3
.28
.09
.38
0
0
0
50 (1)
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
60 (1)
20 (4)
60 (5)
20 (10)
0
17 (2)
100 (5)
83 (14)
0
0
0
0
20 (50)
0
0
0
20 (1)
0
0
0
20 (1)
0
0
0
0
20 (2)
20 (4)
20 (2)
20 (2)
20 (1)
20 (4)
100 (25)
100 (15)
80 (5)
80 (2)
80 (2)
80 (3)
60 (3)
40 (12)
20 (10)
20 (2)
20 (2)
20 (2)
20 (2)
20 (4)
20 (4)
20 (2)
0
0
0
0
0
0
0
0
0
0
0
0
Platanthera bifolia
Spergula morisonii
Viola canina
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
.75
.04
.27
.74
.32
0
0
0
0
0
Festuca ovina
0
0
0
0
83 (26)
67 (3)
83 (10)
83 (2)
33 (3)
17 (2)
0
Danthonia decumbens
Calluna vulgaris
Carex pilulifera
Juncus squarrosus
Hieracium pilosella
Hieracium laevigatum
Deschampsia flexuosa
Erica tetralix
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
.1
.42
0
0
0
0
17 (2)
100 (16)
83 (9)
33 (7)
33 (2)
33 (3)
17 (2)
17 (4)
17 (4)
83 (3)
83 (3)
50 (3)
33 (4)
33 (2)
17 (2)
17 (2)
14 (4)
17 (2)
17 (2)
17 (2)
17 (4)
17 (2)
17 (2)
17 (2)
17 (4)
0
0
0
0
Pedicularis sylvatica
Arnica montana
0
0
0
0
0
0
0
0
Rhamnus frangula
Trichophorum cespitosum
Hieracium umbellatum
Hieracium vulgatum
Succisa pratensis
Gentiana pneumonanthe
Polygala serpyllifolia
Carex panicea
0
0
0
0
0
0
P
P
P
P
P
P
P
P
P
I
0
0
0
0
.56
.22
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
.36
0
0
0
0
0
Dactylhoriza maculata
Drosera rotundifolia
Salix cinerea
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Carex arenaria
P
P
P
P
P
P
P
P
I
0
0
0
0
0
Cirsium dissectum
Cuscuta epithymum
Eriophorum angustifolium
Genista anglica
0
0
0
0
0
0
0
0
0
0
0
.15
0
0
0
0
0
0
0
0
0
0
Genista pilosa
0
0
0
0
0
Lycopodium clavatum
Narthecium ossifragum
Pinus sylvestris
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Salix repens
P
0
0
0
0
0

1
10
BAKKER, J.P. ET AL.
A third field in the study area near the Anloër Diepje
was acquired in the 1940s and had not been fertilized
since then (referred to in the text as ‘local reference’)
(Fig. 1). In 1999 the vegetation was characterized by
Agrostis capillaris and Anthoxanthum odoratum and
species characteristic of nutrient-poor soil conditions
such as Carex nigra, Luzula campestris, Galium saxatile,
Potentilla erecta and Agrostis canina (Table 1, col. 6).
The field was annually cut for hay in August.
Vegetation measurements
In the ‘old’ and ‘new’ fields two replicate permanent
plots of 2m ¥ 2m were established under each manage-
ment regime of subfields of ca. 10 m ¥ 50 m. These
subfields showed a slight elevation gradient towards the
brook. The permanent plots were located at the highest
part of each subfield where the initial vegetation was
most similar among the subfields (Fig. 2). The vegeta-
tion in the permanent plots was recorded in June every
year from 1973-1996 according to a decimal scale (Londo
1
976). The mean number of species per treatment was
derived from the two replicates. We counted the number
of flowering spikes of Dactylorhiza majalis in each
management subfield. The vegetation of the ‘local ref-
erence’ was described in 5 relevés of 2m ¥ 2m in 1999
(
‘
Table 1, col. 6). A description of the vegetation of
regional references’ was derived from the national
vegetation database used for the description of the veg-
etation of The Netherlands (Schaminée et al. 1996; see
also Smits et al. 2002, this issue). We received relevés
from the region of the Drenthe Plateau to be sure that we
did not include species which were only found in other
parts of The Netherlands. These relevés were divided
into a group of 6 relevés representing poorly developed
Nardo-Galion saxatilis, i.e. containing non-characteris-
tic species or missing characteristic species, and one of
5
relevés representing well developed Nardo-Galion
saxatilis. The relevés are referred to as ‘poorly’ and
well developed regional references’ (Table 1, col. 7 and
’
8
, respectively).
The species were characterized as indicating nutri-
ent-rich, intermediate and nutrient-poor soils (Bakker
989; see also Bakker et al. 1987). This classification
Fig. 2. Simplified vegetation maps of the ‘new’ field of the
Loefvledder site in 1973, 1978 and 1984, respectively. The
management regimes imposed in 1973 after the cessation of
fertilizer application are (from left to right): 1= abandon-
ment; 2 = burning every second year; 3 = mulching in
September every second year; 4 = haymaking in September
every second year; 5 = annual haymaking in September; 6 =
annual haymaking alternating in July or September; 7 =
annual haymaking in July; 8 = annual haymaking in July by
scythe; 9 = annual haymaking in both July and September
1
was derived from Klapp (1965), Kruijne et al. (1967)
and Ellenberg (1979). A few species were not classified
by these authors. Ellenberg’s indicator values were de-
veloped for central Europe, but it is assumed that they
have broader applicability (e.g. Thompson et al. 1993).
Recently it is argued that the term N-value should be
replaced by ‘productivity value’ (Hill & Carey 1997;
Schaffers & Sýkora 2000). Mean frequencies of species
indicating the different soil conditions were derived
from the two replicates per treatment.
(
after Bakker & de Vries 1985). Burning turned out to be a
difficult practice. Therefore, its effect is not discussed in the
present paper.

-
EFFECTS OF LONG-TERM CUTTING IN A GRASSLAND SYSTEM: PERSPECTIVES FOR RESTORATION OF PLANT COMMUNITIES- 111
Expected changes in above-ground biomass were
measured as dry weight after clipping plots of 20cm ¥
0 cm (n = 10) in each treatment just before cutting.
Hence, treatments cut every 2 yr lack data in the years
when no cutting took place. The same holds for the
abandonment regime.
The relevés were ordinated with Detrended Corre-
spondence Analysis (program CANOCO; ter Braak
1988). As higher eigenvalues of ordination axes indi-
cate greater importance of these axes, we used the two
resulting axes with the highest eigenvalues for analysis
(Jongman et al. 1987). All relevés in all treatments were
entered after square-root transformation as a mean relevé
of the two replicates in each management regime in the
2
Nitrogen removal
’old’ and ‘new’ fields. Individual relevés of the ‘local
The nitrogen concentration in above-ground biomass
was measured by clipping samples of 20 cm ¥ 20 cm
n = 1 in 1975, n = 3 in 1983, n = 5 in 1999) just before
reference’, the ‘poorly’ and ‘well developed regional
references’ were also included.
(
cutting in the regime both cut in July and September.
Total nitrogen was determined with a modified Kjehl-
dahl technique (Wieniger 1936). Nitrogen removal was
estimated from the concentration and the biomass as
measured above.
Results
Vegetation changes compared among regimes
The main conclusions regarding the hypotheses put
forward were: (1) abandonment resulted in the estab-
lishment of shrubs, although it takes more than 20 yr to
completely cover the site; (2) mulching resulted in tall
forb communities; (3) different periods of annual
haymaking regimes did not reveal different plant com-
munities; (4) haymaking twice a year resulted in the
largest removal of nutrients and a decrease of species
number after an initial strong increase.
In 1995 the vegetation in the permanent plots of the
‘new’ field with less frequent cutting and mulching was
characterized byHolcus lanatus, Festuca rubra, Agrostis
capillaris, Potentilla anserina, Ranunculus repens and
sporadically tall plants such as Scrophularia nodosa,
Elymus repens, Heracleum sphondyleum, Cirsium
palustre (Table 1, col. 2). The vegetation in the plots cut
annually for hay showed the establishment of species
indicating intermediate soil conditions such as Antho-
xanthum odoratum, Plantago lanceolata, and Ranuncu-
lus acris (Table 1, col. 3). The vegetation of the ‘old’
field had also changed by 1995 and showed the estab-
lishment of Agrostis capillaris, Juncus effusus, and
species indicating nutrient-poor soil conditions such as
Carex nigra, Agrostis canina, Galium saxatile and
Nardus stricta (Table 1, col. 5).
The above-mentioned changes are reflected in the
change in mean frequency of species indicating nutrient-
rich, intermediate and nutrient-poor soil conditions. The
regimes with less frequent cutting, mulching and aban-
donment remained similar in species frequencies indi-
cating nutrient-rich conditions (Fig. 3). The regimes
with annual haymaking showed a decrease of species
indicating nutrient-rich soil conditions and an increase
of intermediate species (Fig. 3). The vegetation in the
‘old’ field showed a disappearance of species indicating
nutrient-rich soil conditions and an increase of species
indicating nutrient-poor soil conditions (Fig. 4).
Soil seed bank
Soil seed bank samples were collected in January
1
988. Natural stratification has already taken place in
the field at that time of the year. In each management
regime 5 replicate soil samples were taken with an auger
of 7 cm diameter of the 0-6 cm top layer. The replicate
samples were pooled and spread in a layer of 1 -2 cm
over a sterile layer of potting soil in trays (30cm ¥ 30cm
¥
10 cm) in a glasshouse. The samples were watered
daily. The temperature was 25 ∞C during day time and
5 ∞C at night. Seedlings were identified, counted and
1
removed as soon as possible. This procedure was con-
tinued during 6 months. Unidentified seedlings were
grown individually in order to identify them when they
were bigger or started to flower.
The records in the seed bank database of Thompson
et al. (1997) were used to calculate a longevity index of
the species in the present study. This is the proportion of
‘persistent records’, i.e. (short-term + long-term persist-
ent records) / (transient + short-term + long-term per-
sistent records). Hence, the longevity index can vary
between 0 (no persistent records) and 1 (all persistent
records) (Bekker et al. 1998). Only species with 5 or
more records in the database were taken into account.
Species with 3 or 4 records were only considered when
their classification was unanimous.
Data analyses
Significant trends in mean species numbers of two
replicate permanent plots during the study period were
analysed with SPSS (de Vocht 1999). We fitted a linear,
quadratic or cubic regression. The simplest model was cho-
sen, unless a higher model showed a significant (P < 0.05)
higher correlation according to the Z-test (Zar 1996).

1
12
BAKKER, J.P. ET AL.
Fig. 3. Changes since 1973 in
the frequency of species indicat-
ing nutrient-poor, nutrient-rich
and intermediate soil conditions
in the ‘new’ field with different
regimes.
The mean number of species in the permanent plots
did not change with abandonment and haymaking every
second year (Fig. 5). It decreased with haymaking by
scythe and increased in all other regimes (Fig. 5). The
increase was fastest in the regime cut for hay in both
July and September and reached the highest species
richness of over 20 species m after ca. 12 yr. However,
it started to decrease after 15 yr. Also annually haymak-
ing in July in the ‘new’ field showed a decrease of the
number of species, but the decrease started after 20
years. The number of species declined somewhat in the
was about 50% of that during the first decade after the
start of the experiment. The concentration of nitrogen did
not change, but the yield strongly decreased (Table 3).
-2
Table 2. Changes in flowering spikes of Dactylorhiza majalis
at different cutting regimes. Sometimes two spikes were ob-
served at September mulching every two years (- = no
obserservation).
July +
September
July
scythe
July
July or September
September
‘old’ field annually cut for hay (Fig. 4).
The first flowering spikes of Dactylorhiza majalis
1
1
1
983 first spikes observed since start in 1973
were found in the ’new’ field in 1983, 10 yr after the
start of the experiment. The numbers remained low
during the 1980s, and increased strongly from 1991
onwards, but only in the regime cut both in July and
September (Table 2).
The standing crop at cutting date did not change in
the regimes with less frequent cutting and mulching
Fig. 6). It decreased significantly from about 600-750 g
984
985
-
0
2
10
6
-
-
-
-
1
0
1
0
0
-
0
0
1
0
0
-
1
2
3
0
0
-
1
0
1
0
0
-
1986
1987
1988
1989
5
-
1990
1991
1992
1993
13
30
38
91
0
2
3
3
6
0
0
0
1
0
1
0
1
4
0
5
4
1
1
5
4
3
4
3
5
2
0
0
0
0
1
5
1
0
3
3
(
-
2
-2
dw m to about 300 g dw m in the regimes annually
cut for hay (Fig. 6). Total removal was greatest in the
1994
1995
1996
1997
142
147
192
ca. 300
ca. 600
-
2
twice-annually cut regime, namely, from 900 g dw m
in 1973 to about 400 g dw m in 1998 (Fig. 6). Total
nitrogen removal 24 yr after the start of the experiment
-2
1998
999
1

-
EFFECTS OF LONG-TERM CUTTING IN A GRASSLAND SYSTEM: PERSPECTIVES FOR RESTORATION OF PLANT COMMUNITIES- 113
Table 4. Composition of the established vegetation and the
soil seed bank in January 1988 in the ‘new’ and ‘old’ field of
the plots with annual haymaking in July. For the established
vegetation presence/absence in permanent plots in July 1987
is indicated. For the soil seed bank the number of germinated
2
seeds from 38.5 cm (0-6 cm) in January 1988 is indicated.
‘New’ field
‘Old’ field
Established
vegetation
Seed Established
bank vegetation
Seed
bank
Agrostis canina
-
-
+
-
-
-
+
-
+
-
-
-
-
+
-
-
+
+
-
-
-
26
4
2
-
-
-
1
-
-
-
-
3
-
1
-
+
+
-
-
-
+
-
-
115
2
8
-
-
15
-
26
-
-
8
93
-
-
-
-
-
1
5
-
Agrostis capillaris
Agrostis stolonifera
Alopecurus geniculatus
Alopecurus pratensis
Anthoxanthum odoratum
Anthriscus sylvestris
Callitriche spec.
Cardamine pratensis
Carex curta
Carex nigra
Carex ovalis
Carex panicea
Cerastium fontanum
Cirsium palustre
Deschampsia cespitosa
Festuca pratensis
Festuca rubra
-
+
+
+
+
-
+
-
-
+
-
-
3
111
Glyceria fluitans
Holcus lanatus
+
-
Hydrocotyle vulgaris
Juncus bufonius
Juncus bulbosus
Juncus effusus
Leontodon autumnalis
Lolium perenne
-
-
-
+
+
-
-
13
1
38
-
+
-
-
+
-
4
8
3
1018
-
3
-
-
Luzula campestris
Luzula multiflora
Lychnis flos-cuculi
Lythrum portula
Plantago major
-
-
-
-
-
-
3
3
+
+
-
-
-
1
-
-
2
-
-
1
Poa pratensis
Poa trivialis
+
+
-
-
-
+
-
-
-
10
63
1
-
1
130
1
19
2
-
-
-
-
+
-
+
+
+
-
-
-
3
2
-
Fig. 4. Changes since 1973 in the frequency of species indicat-
Polygonum persicaria
Potentilla erecta
Ranunculus flammula
Ranunculus repens
Rhinanthus angustifolius
Rumex acetosa
Rumex obtusifolius
Stellaria media
Taraxacum spec.
Viola palustris
ing nutrient-poor, nutrient-rich and intermediate soil condi-
-
-
2
7
21
1
1
-
1
2
2
tions, the mean number of species 4m and the yield in the
old’ field with haymaking in July.
‘
-
-
-
4
-
+
Table 3. Average nitrogen content at peak standing crop
yield ± S.D.) and nitrogen removal in the first half of July in
different cutting regimes and different years (nitrogen content:
975: n = 1; 1983: n = 3; 1999: n = 5; yield all years: n = 10).
(
Vegetation changes compared to reference situations
1
The ordination demonstrated a shift of the regimes
cut for hay annually of the ‘new’ field towards the ‘old’
field. That moved in turn towards the ‘local reference’.
The ‘local reference’ was still not similar to the ‘poorly
developed regional reference’, and even further from
the ‘well developed regional reference’ (Fig. 7). The
position of species indicating nutrient-rich and nutrient-
poor soil conditions made clear that the first axis of the
ordination represents an axis from eutrophic towards
oligotrophic soil conditions, where the Nardo-Galion
N-concentration
%)
Yield
(g dw m-2)
N-removal
-1
(
(kg ha
)
1
1
1
975 July
1.31
1.39
625 ± 92
810 ± 97
82
113
July + Sept
983 July
July + Sept
1.49
1.77
723 ± 64
641 ± 69
108
113
999 July
July + Sept
1.89
1.41
269 ± 23
291 ± 43
51
41

1
14
BAKKER, J.P. ET AL.
Fig. 5. Changes since 1973 in
the mean number of species
2
per 4 m in the ‘new’ field
with different regimes.
Fig. 6. Changes since 1973 in
the yield at cutting periods in
July or September in the ‘new’
field with different regimes.

-
EFFECTS OF LONG-TERM CUTTING IN A GRASSLAND SYSTEM: PERSPECTIVES FOR RESTORATION OF PLANT COMMUNITIES- 115
Fig. 7. Ordination by Detrended Correspondence Analysis of changes in the ‘new’ and ‘old’ field covering more than 20 yr of
succession since 1973 with different regimes, and of the relevés representing the local reference, poorly and well developed
regional references of Nardo-Galion saxatilis communities. Changes between 5 time intervals (1973, 1978, 1983, 1987, 1995) are
shown. Eigenvalues in parentheses on the axis labels.
Fig. 8. Ordination by Detrended Correspondence Analysis of all species in the ‘new’ and ‘old’ field, the local reference and the
poorly and well developed regional references of Nardo-Galion saxatilis communities. Eigenvalues in parentheses in the axis labels.

1
16
BAKKER, J.P. ET AL.
saxatilis communities thrive (Fig. 8). At the right part of
the ordination characteristic species of nutrient-poor
soil conditions are found such as Nardus stricta, Carex
nigra, Luzula multiflora, Juncus squarrosus, Succisa
pratensis, Arnica montana, Erica tetralix and Calluna
vulgaris.
already reached after 25 yr, but it should be further
reduced. The yield of the ‘old’ field decreased from
–2
580 to ca. 200 g dw m , and is still higher than that of
the data for Nardo-Galion saxatilis reported by Klapp
(1965). Only the local reference community showed
-2
yields of ca. 150 g dw m (unpubl.). The yields meas-
ured before cutting or mulching every 2 yr, did remain
The second axis might be explained by the occur-
rence of tall forbs such as Urtica dioica, Heracleum
sphondylium, Cirsium arvense, the tree Sorbus aucu-
paria, and annual species indicating accumulation of
litter such as Senecio sylvaticus and Galeopsis tetrahit
at the left lower part of the ordination (Fig. 8). These
species especially occur in the regimes that were not cut
annually or abandoned. At the left higher part of the
ordination species such asLolium perenne, Poa trivialis,
P. pratensis, Bromus hordeaceus, Trifolium repens and
Plantago major are found (Fig. 8). These species mainly
occur during the first years in the regimes annually cut
for hay. The eigenvalues of the axes indicate that the
first axis explains the variation much more than the
second one.
-2
ca. 650 g dw m . Therefore, we conclude that cutting
annually is a prerequisite for a decrease in the yield
after cessation of fertilization.
Similar conclusions were reached by Oomes &
Mooi (1981, 1985) for Arrhenatherion elatioris grass-
lands in The Netherlands, and by Poschlod & Kahmen
(2002) for calcareous grassland communities in Ger-
many. A stand of mixed elements of Nardo-Galion and
Festuco-Gentianelletum in Germany was compared
for grazing with low stocking density and mulching
twice a year (Schiefer 1981). The stand had received
some fertilizer in the 1950s and yielded 300-400 g dw
-
2
m . Mulching enhanced both the grassland species
Poa pratensis and species indicating nutrient-poor soil
conditions such as Luzula campestris, Agrostis
capillaris, Viola canina, Rumex acetosella and Poten-
tilla erecta. This suggests that the latter low-statured
species not only need nutrient-poor soil conditions, but
also little competition for light from a taller canopy.
Although similar decreases in yields were recorded
in two regimes (July haymaking by machinery and by
scythe) the number of species increased in the first and
decreased in the latter regime, even though the fields
are adjacent to each other and (vegetative) dispersal
should be possible. As removal of the swath also took
place by hand, no machinery entered the regime cut by
scythe since 1973. No soil compaction took place and
the sward lost direct contact with the soil. Hence the
sward is very dry, and this may explain the low species
number compared to the regimes including cutting by
machinery.
Seed bank
Species characteristic of the Nardo-Galion saxatilis
were absent from the seed bank (Table 4). They have a
low longevity index, such as Pedicularis sylvatica (0),
Arnica montana (0),Festuca ovina (0.04), Nardus stricta
(0.18), Succisa pratensis (0.22), Galium saxatile (0.22),
Danthonia decumbens (0.27), Molinia caerulea (0.3),
Potentilla erecta (0.34). Viable seeds of Hydrocotyle
vulgaris and Viola palustris were found in the seed bank
of the ‘old’ field (Table 4), although their longevity
index is 0. As they also occurred in the established
vegetation, their occurrence in the seed bank probably
reflects recent seed rain.
The removal of nitrogen from the field decreased
–
1
–1
Discussion
dramatically from ca. 100 to 40-50 kg N ha yr , not
because of the N concentration but as a result of
decreasing yield (Table 3). Nowadays the removal of
nitrogen approaches that of the input from the average
atmospheric deposition in The Netherlands: in 1980
Abiotic conditions
The initial standing crop of the fertilized grassland
in the ‘new’ field ranged from 600 to 750 g dw m and
averaged 680 g dw m (Fig. 6). Although strong
-2
–1 –1
–1
–1
this was 43 kg N ha yr against 39 kg N ha yr in
1993 (Erisman & Draaijers 1995). The relatively high
N input may coincide with the very high cover (up to
80%) of the moss Rhytidiadelphus squarrosus
(unpubl.). This moss only occurs in fields that have
been cut for decades without fertilization (Pegtel et al.
1996), and is spreading in a range of low-productive
plant communities in The Netherlands during the last
15 yr (Londo 2002). Extra N input in very low produc-
tive systems may be taken up by mosses and not by
-2
fluctuations could be noticed, an overall decrease to
00 - 300 g dw m- occurred when cuttings were re-
2
2
moved once or twice a year. As the standing crop was
measured in the first half of July, it resembled peak
standing crop and hence yield. The yields after 25 yr of
cutting are well above the data given by Klapp (1965)
for Nardo-Galion saxatilis communities: 100 - 150 g
-2
dw m . Apparently a strong decrease in yield was

-
EFFECTS OF LONG-TERM CUTTING IN A GRASSLAND SYSTEM: PERSPECTIVES FOR RESTORATION OF PLANT COMMUNITIES- 117
phanerogams as is shown in the high Arctic (Jónsdóttir
et al. 1995). Although moss carpets have a low stand-
ing crop, they can prevent seedlings from establish-
ment (van Tooren et al. 1987), and hence retard suc-
cession towards Nardo-Galion saxatilis communities.
The spread of mosses during haymaking without
fertilization is related to a decrease in pH (Bakker
and Viola palustris flower very early, and they might
have shed their seeds in early July, and thus miss the
dispersal vector. The mismatch in the period of seed
ripening and cutting may explain the lack of dispersal
of species typical of nutrient-poor soil conditions from
the ‘old’ field to the adjacent ‘new’ field, even though
they are cut at the same time and the machine goes
directly from one field into the other. Galium saxatile,
Agrostis canina, Nardus stricta, Hydrocotyle vulgaris
and Viola palustris were present from the beginning in
the ‘old’ field, and they spread (vegetatively) into the
permanent plots (Table 1, col. 5).
The poor dispersal from the ‘old’ field into the
adjacent ‘new’ field makes dispersal from the local
reference at ca. 500 m distance into the target area
unlikely. The same holds to a stronger degree for the
distance to be covered (up to 50 km) from the regional
references elsewhere at the Drenthe Plateau. It seems
that the number of years that restoration management
was carried out cannot explain the stage of succession
of an individual field under way from fertilized grass-
land towards Nardo-Galion communities. The history
of individual fields, including the remnants of earlier
communities indicating nutrient-poor soil conditions
seems more important (Bakker & Olff 1995).
An interesting point with respect to dispersal is
shown by the orchid species Dactylorhiza majalis:
they mainly established in the regime cut both in July
and September. The above-ground standing crop is
low under this regime, but has similar low values in the
adjacent regimes (Fig. 6). The tiny seeds of orchids are
supposed to be wind-dispersed, and are likely to reach
the adjacent regimes within a distance of 50 m. Seeds
of the species are also found attached to haymaking
machinery (Strykstra et al. 1997), and this may be a
further vector of dispersal. Further research is needed
to answer this question.
Many species characteristic of communities of nu-
trient-poor soil conditions including Nardo-Galion
saxatilis have also been found in droppings of live-
stock such as Galium saxatile, Nardus stricta, Agrostis
capillaris, Festuca rubra, F. ovina, Luzula campestris,
L. multiflora, Molinia caerulea, Potentilla erecta,
Calluna vulgaris, Carex pilulifera. Also Arnica montana
and Succisa pratensis can be dispersed by livestock,
but it is not known whether by endozoochory or
ectozoochory (Bonn & Poschlod 1998). It can be ex-
pected that Nardo-Galion species were dispersed in
the past by livestock characteristic of common grazing
land. Dispersal by livestock was not limited by fences
and flowering period. Hence, restoration of Nardo-
Galion communities in individual isolated fields by
cutting in early July is likely to meet constraints.
1
989; Olff & Bakker 1991). This acidification may
prevent some Nardo-Galion species from establish-
ment. Species such as Pedicularis sylvatica, Danthonia
decumbens, Polygala serpyllifolia may need a certain
level of base saturation. Nardo-Galion communities
are characterized by slightly higher pH, higher base
cation concentrations and lower Al:Ca ratios, com-
pared to the soils of Calluna vulgaris stands. Experi-
ments with Arnica montana and Calluna vulgaris
showed that the first species is much more sensitive to
both increased Al:Ca ratios and high aluminium levels
(
de Graaf et al. 1997, 1998).
Seed bank and dispersal
For many species characteristic of Nardo-Galion
communities too few (less than 5) records were found
in the seed bank database (Thompson et al. 1997).
Thus, more seed bank studies are needed. Seed bank
data collected in 1981 in the ‘new’ field revealed that
species that increased in the permanent plots such as
Rhinanthus angustifolius, Anthoxanthum odoratum,
Plantago lanceolata, Rumex acetosella, Agrostis capil-
laris did not reveal a seed bank prior to their emer-
gence in the established vegetation. These are typi-
cally species with a low longevity index and have a
transient or short-term persistent seed bank (Bakker et
al. 1991). Moreover they flower in spring (van der
Meijden 1990) and hence have produced seeds at the
period of haymaking in early July. This increases their
chance to be transported by haymaking machinery
from elsewhere (Strykstra et al. 1997). Other recently
established species such as Leontodon autumnalis and
Hypochaeris radicata flower later, but may have a
chance to be dispersed by wind as they have a pappus.
Some species characteristic of Nardo-Galion and other
communities of nutrient-poor soil conditions flower in
spring such as Nardus stricta, Carex nigra, C. ovalis,
Luzula multiflora, and may have a chance to be dis-
persed in early July by haymaking machinery. How-
ever, many species start flowering late in spring or in
summer, such as Galium saxatile, Juncus squarrosus,
Potentilla erecta, Molinia caerulea, Agrostis canina,
Hydrocotyle vulgaris, Succisa pratensis (van der
Meijden 1990). Therefore, it seems unlikely that they
can be dispersed into the target area by haymaking
machinery when it is cut in early July. Luzula campestris

1
18
BAKKER, J.P. ET AL.
Restoration perspectives
Berendse 1999). The practice of two annual cuttings
and removal of most nutrients, initially showed the
highest species richness (Fig. 5), but the number of
species decreased afterwards. Moreover Berendse et
al. (1992) found a decline in species number despite a
sharp decrease in dry matter yield of a meadow on
sandy soil by nutrient export through removal of hay.
However, it is not just species number that must be
considered, but also the number of species characteris-
tic of a target community. The present study demon-
strates that succession from fertilized grasslands to-
wards communities harbouring Nardo-Galion saxatilis
species is possible, but only to a certain extent. Annual
cutting and removal of the swath result in a decrease of
Nardo-Galion saxatilis communities occur on rela-
tively acid and nutrient-poor soils with a low plant
production capacity (Schaminée et al. 1996). It is well
known that the maintenance of Nardo-Galion commu-
nities needs grazing by livestock (Rodwell 1992;
Schaminée et al. 1996) or mowing (Schaminée et al.
1
996). Cessation of the regular management practices
results in bush encroachment and/or a dense layer of
dead material. This can be counteracted by cutting and
removal of the litter layer. More drastic interventions
are small-scale sod removal followed by grazing and/
or cutting and removal of the swath (Schaminée et al.
1
996).
Little is known about the possibilities of restoration
standing crop, but after 25 years not below a level of
–2
2
00 -300 g dw m , which is more than the 100-150 g
of Nardo-Galion communities from fertilized arable
fields or grasslands. The present study showed a strong
decrease of yield after 15 yr by annual cutting of
grasslands. The removal of nutrients by grazing sheep
is not efficient (Marrs et al. 1989). It is likely that in the
succession during cutting N is limiting in early succes-
sional stages, later followed by K, and that eventually
co-limitation of N, P and K takes place (Pegtel et al.
–2
dw m of the target community. Gradually the re-
moval of nitrogen is balanced by the input through
atmospheric deposition. Moreover, atmospheric depo-
–1
–1
sition (about 40 kg N ha yr ) in The Netherlands
–1
–1
surpasses the critical load (20 -25 kg N ha yr ) for
plant communities on nutrient-poor soils (Bobbink et
al. 1998). Atmospheric deposition and acidification
may result in a dense carpet of Rhytidiadelphus
squarrosus that prevents the establishment of new
species. Moreover, acidification may prevent the es-
tablishment of Nardo-Galion saxatilis species that need
a somewhat higher base cation concentration. Charac-
teristic Nardo-Galion species do not occur in the soil
seed bank and seem to have a low dispersal capacity.
It is clear from this study that haymaking alone will
not result in restoring the target community. Only
experiments including the introduction of target species,
can reveal causes of standstill of further succession, be
it acidity of the soil, high moss cover, or lack of
dispersal of propagules from elsewhere.
1
996). Nitrogen application in grassland cut for hay on
sandy soil to get faster depletion of K and P, resulted
after 17 yr in limitation of yield by K (van der Woude
et al. 1994). A period of 7 yr of only N application on
arable fields was supposed to deplete P. However, the
amount of extractable P showed large annual fluctua-
tions (Marrs et al. 1998). Schaminée & Jansen (1998)
assumed that Nardo-Galion communities can be re-
stored from arable fields by top soil removal and
subsequent grazing. This process may take up to 50 yr.
A comparative study in 9 sites on the effects of top
soil removal started in The Netherlands in the period
1
989-1995. Most sites were fenced in with adjacent
areas harbouring communities of nutrient-poor soil
conditions, including Nardo-Galion communities.
Grazing livestock may enhance dispersal of target
species into the target areas. In the first 2 yr after top
soil removal four target species of Nardo-Galion com-
munities established, namely, Festuca ovina and Luzula
campestris/multiflora appeared in three sites, whereas
Molinia caerulea and Potentilla erecta were found at
one site. It is likely that these species were dispersed
from nearby source populations in the fenced-in areas
Acknowledgements. We gratefully acknowledge the permis-
sion of the State Forestry Commission to work in the
Stroomdallandschap Drentsche A, and to carry out the cutting
regimes. We thank Latzi Fresco for his contribution to the data
analysis. Dick Visser made the figures. We acknowledge the
help of Nina Smits of ALTERRA, Wageningen to provide the
relevés of the Drenthe Plateau.
(
unpubl. data R. Verhagen). After 9 yr no new Nardo-
Galion species had established, although almost all of
4 target species are present in the local species pool
1
within the same or adjacent block of 5 km ¥ 5 km
(
Verhagen et al. 2001).
The restoration of ecological diversity of grasslands
and heathlands suffers from constraints (Bakker &

-
EFFECTS OF LONG-TERM CUTTING IN A GRASSLAND SYSTEM: PERSPECTIVES FOR RESTORATION OF PLANT COMMUNITIES- 119
References
Erisman, J.W. & Draaijer, G.P.J. 1995. Atmospheric deposi-
tion in relation to acidification and eutrophication.
Elsevier, Amsterdam, NL.
Everts, F.H. & de Vries, N.P.J. 1991. De vegetatieontwikkeling
van beekdalsystemen. Een landschapsoecologische ana-
lyse van enkele Drentse beekdalen. Historische Uitgeverij,
Groningen, NL.
Bakker, J.P. 1989. Nature management by grazing and cut-
ting. Kluwer Academic Publishers, Dordrecht, NL.
Bakker, J.P. & Berendse, F. 1999. Constraints in the restora-
tion of ecological diversity in grassland and heathland
communities. Trends Ecol. Evol. 14: 63-68.
Bakker, J.P. & Olff, H. 1995. Nutrient dynamics during
restoration of fen meadows by haymaking without ferti-
liser application. In: Wheeler, B.D., Shaw, S.C., Foyt,
W.J. & Robertson, R.A. (eds.) Restoration of temperate
wetlands, pp. 143-166. Wiley, Chichester, UK.
Bakker, J.P. & de Vries, Y. 1985a. The results of different
cutting regimes in grassland taken out of the agricultural
system.In: Schreiber, K.F. (ed.). Sukzession auf Grün-
landbrachen. Münster. Geogr. Arb. 20: 51-57.
Grootjans, A.P. 1980. Distribution of plant communities
along rivulets in relation to hydrology and management.
In: Wilmanns, O. & Tüxen, R. (eds) Epharmonie, pp.
143-170. Cramer, Vaduz, LI.
Hill, M.O. & Carey, P.D. 1997. Prediction of yield in the
Rothamstead Park grass experiment by Ellenberg indica-
tor values. J. Veg. Sci. 8: 579-586.
Jongman, R.H.G., ter Braak, C.J.F. & van Tongeren, O.F.R.
1987. Data analysis in community and landscape ecol-
ogy. Pudoc, Wageningen, NL.
Jónsdóttir, I.S., Callaghan, T.V. & Lee, J.A. 1995. Fate of
added nitrogen in a moss-sedge Arctic community and
effects of increased nitrogen deposition. Sci. Total Envir.
160: 677-685.
Kahmen, S., Poschlod, P. & Schreiber, K.F. 2002. Conserva-
tion management of calcareous grasslands. Changes in
plant species composition and response of functional
traits during 24 years. Biol. Conserv. 104: 319-328.
Klapp, E. 1965. Grünlandvegetation und Standort. Parey,
Berlin, DE.
Kruijne, A.A., De Vries, D.M. & Mooi, H. 1967. Bijdrage tot
de oecologie van Nederlandse graslandplanten. Versl.
Landbouwkd. Onderz. 696. Pudoc, Wageningen, NL.
Londo, G. 1976. The decimal scale for relevés of permanent
quadrats. Vegetatio 33: 61-64.
Londo, G. 2002. Does Rhytidiadelphus squarrosus (Hedw.)
Warnst. increase in the Netherlands? Lindbergia 27: 29-
36.
Marrs, R.H., Rizand, A. & Harrison, A.F. 1989. The effects
of removing sheep grazing on soil chemistry, above-
ground nutrient distribution, and selected aspects of soil
fertility in long-term experiments at Moor House na-
tional reserve. J. Appl. Ecol. 26: 647-661.
Bakker, J.P. & de Vries, Y. 1985b. Über die Wiederherstellung
artenreicher Wiesengesellschaften unter verschiedenen
Mahdsystemen in den Niederlanden. Nat. Landsch. 60:
2
92-296.
Bakker, J.P., Bos, A.F., Hoogveld, J. & Muller, H.J. 1991.
The role of the seed bank in restoration management of
semi-natural grasslands. In: Ravera, O. (ed.) Terrestrial
and aquatic ecosystems. Perturbation and recovery, pp.
4
49-455. Ellis Horwood, New York, NY.
Bakker, J.P., Brouwer, C., van den Hof, L. & Jansen, A.
987. Vegetational succession, management and hydrol-
ogy in a brookland (The Netherlands). Acta Bot. Neerl.
6: 39-58.
1
3
Bekker, R.M., Bakker, J.P., Grandin, U., Kalamees, R.,
Milberg, P., Poschlod, P., Thompson, K. & Willems,
J.H. 1998. Seed shape and vertical distribution in the
soil: indicators for seed longevity. Funct. Ecol. 12: 834-
8
42.
Berendse, F., Oomes, M.J.M., Altena, H.J. & Elberse, W.T.
992. Experiments on the restoration of species-rich
1
meadows in the Netherlands. Biol. Conserv. 62: 59-65.
Bignal, D.I. & McCracken, E.M. 1996. Low-intensity farm-
ing systems in the conservation of the countryside. J.
Appl. Ecol. 33: 413-424.
Bobbink, R., Hornung, M. & Roelofs, J.G.M. 1998. The
effects of air-borne nitrogen pollutants on species diver-
sity and semi-natural European vegetation. J. Ecol. 86:
Marrs, R.H., Snow, C.S.R., Owen. K.M. & Evans, C.E.
1998. Heathland and acid grassland creation on arable
soils at Minsmere: identification of potential problems
and a test of cropping to impoverish soils. Biol. Conserv.
85: 69-82.
7
17-738.
Bonn, S. & Poschlod, P. 1998. Ausbreitungsbiologie der
Pflanzen Mitteleuropas. Quelle & Meyer, Wiesbaden,
DE.
Oberdorfer, E. 1978. Süddeutsche Pflanzengesellschaften.
Teil II. Fischer Verlag, Stuttgart, DE.
de Graaf, M.C.C., Bobbink, R., Verbeek, P.J.M. & Roelofs,
J.G.M. 1997. Aluminium toxicity and tolerance in three
heathland species. Water Air Soil Pollut. 98: 229-241.
de Graaf, M.C.C., Verbeek, P.J.M., Bobbink, R. & Roelofs,
J.G.M. 1998. Restoration of species-rich heaths: the im-
portance of appropriate soil conditions. Acta Bot. Neerl.
Olff, H. & Bakker, J.P. 1991. Long-term dynamics of stand-
ing crop, species richness and vegetation composition
after the cessation of fertilizer application to hay fields.
J. Appl. Ecol. 28: 1040-1052.
Oomes, M.J.M. & Mooi, H. 1981. The effect of cutting and
fertilizing on the floristic composition and production of
an Arrhenatherion elatioris grassland. Vegetatio 47: 233-
239.
4
7: 89-111.
de Vocht, A. 1999. Basishandboek SPSS 8 & 9 voor Win-
dows 95 & 98. Bijleveld Press, Utrecht, NL.
Ellenberg, H. 1979. Zeigerwerte der Gefässpflanzen Mittel-
europas. Scr. Geobot. IX. Goltze, Göttingen, DE.
Oomes, M.J.M. & Mooi, H. 1985. The effect of management
of succession and production of formerly agricultural
grassland after stopping fertilization. In: Schreiber, K.F.

1
20
BAKKER, J.P. ET AL.
Thompson, K., Hodgson, J.P., Grime, J.P., Rorison, I.H.,
(
ed.) Sukzession auf Grünlandbrachen. Münster. Geogr.
Arb. 20: 59-67.
Pegtel, D.M., Bakker, J.P., Verweij, G.L. & Fresco, L.F.M.
996. N, K and P deficiency in chronosequentional cut
summer-dry grasslands on gley podzol after the cessa-
tion of fertilizer application. Plant Soil 178: 121-131.
Rodwell, J.S. (ed.) 1992. British Plant Communities. Vol. 3.
Grasslands and montane communities. Cambridge Uni-
versity Press, Cambridge, UK.
Band, S.R. & Spencer, R.E. 1993. Ellenberg numbers
revisited. Phytocoenologia 23: 277-289.
Thompson, K., Bakker, J.P. & Bekker, R.M. 1997. The soil
seed banks of North West Europe: methodology, density
and longevity. Cambridge University Press, Cambridge,
UK.
1
van der Meijden, R. 1990. Flora van Nederland. 21st ed.
Wolters-Noordhoff, Groningen, NL.
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:
van der Woude, B.J., Pegtel, D.M. & Bakker, J.P. 1994.
Nutrient limitation after long-term nitrogen fertilizer ap-
plication in cut grasslands. J. Appl. Ecol. 31: 405-412.
van Tooren, B.F., During, H.J. & Lensink, M.J.G. 1987. The
influence of the bryophyte layer on the microclimate in
chalk grassland. Abstr. Bot. 9: 219-230.
Verhagen, R., Klooker, J., Bakker, J.P. & van Diggelen, R.
2001. Development of nutrient-poor plant communities
on formerly agricultural soils: a comparison between the
target vegetation and the actual vegetation after top soil
removal. Appl. Veg. Sci. 4: 75-82.
Wieniger, F.M. 1936. Nachtrag zur Schnellstickstoff-
bestimmungsmethode mit Silen als Beschleuniger.
Wochenschr. Brauerei 53:251-252.
Zar, J.H. 1996. Biostatistical Analysis. 3rd ed. Prentice-Hall
Inc, New York, NY, US.
2
25-244.
Schaminée, J.H.J. & Jansen, A. (eds.) 1998. Wegen naar
natuurdoeltypen. Rapport IKC Natuurbeheer 26: 1-319.
Schaminée, J.H.J., Stortelder, A.H.F. & Weeda, E.J. 1996.
De Vegetatie van Nederland, deel 3. Plantengemeen-
schappen van graslanden, zomen en droge heiden. Opulus
Press, Uppsala, SE.
Schiefer, J. 1981. Bracheversuche in Baden-Württemberg.
Beih. Veröff. Nat. schutz Landsch. pfl. Baden-Württem-
berg 22: 1-325.
Strykstra, R.J., Verweij, G.L. & Bakker, J.P. 1997. Seed
dispersal by mowing machinery in a Dutch brook valley
system. Acta Bot. Neerl. 46: 387-401.
ter Braak, C.J.F. 1988. CANOCO- a FORTRAN program for
canonical community ordination by partial detrended ca-
nonical correspondence analysis and redundancy analy-
sis. Agricultural Mathematics Group, Wageningen, NL.
Zobel, M., van der Maarel, E. & Dupré, C. 1998. Species
pool: the concept, its determination and significance for
community restoration. Appl. Veg. Sci. 1: 55-66.
Received 20 June 2001;
Revision received 7 December 2001;
Accepted 7 December 2001.
Coordinating Editor: R.H. Marrs.