Forgot your password?
Document Actions
Lemauviel Roze 2003
DOI: 10.1007/s00267-002-2813-5
Response of Three Plant Communities to Trampling
in a Sand Dune System in Brittany (France)
SERVANE LEMAUVIEL* used to compare three typical plant communities belonging to
FRANCOISE ROZE
¸ ´ the various landscape units: mobile dune, semifixed dune, and
Universite de Rennes 1
´ fixed dune. The strong contrasts between communities be-
Equipe “Dynamique des communautes” ´ longing to different successional stages reflect their ecological
U.M.R. 6553, Service d’Ecologie Vegetale
´ ´
functioning. The mobile dune and semifixed dune with their
Complexe scientifique de Beaulieu
low resistance contrasted with the fixed dune. Only the vege-
35042 Rennes cedex, France
tation cover of the semifixed dune benefited from long-term
ABSTRACT / Dunes that are protected because of their very trampling and had a very high resilience (134%). This re-
rich and diverse plant communities are often exposed to ex- sponse could be explained by a good balance of two opposite
cessive visitor pressure. The effects of trampling on the habitat factors: soil compaction increasing soil stability and moisture
must be known from a conservation viewpoint but also are content, and vegetation destruction. Because of their low re-
important for management. To determine the response of silience, trampling seems to be harmful for fixed dunes in the
plant assemblages to trampling by people, an experimental long term. The tourist pressure seems easier to integrate in to
study was conducted on the state-owned dunes at Quiberon the mobile dunes and the semifixed dunes if periods of recov-
(Brittany, France). Indices of resistance and resilience were ery are included in the management.
During the 20th century, human activities linked to Toullec and others 1999, Gallet and Roze 2001, 2002).
´
tourism have affected and damaged many coastal areas The coastline is affected and many authors have chosen
(Liddle and Greig-Smith 1975, Paskoff 1989, Van der sand dunes to study the effect of trampling on vegeta-
Meulen 1997, Olsauskas 1996). The interest in seaside tion (Liddle and Greig-Smith 1975, Slatter 1978, Hyl-
recreation and living has led to the construction of gaard and Liddle 1981, McDonnell 1981, Bowles and
seaside resorts, including hotels, apartments, campsites, Maun 1982, Andersen 1995, Lemauviel 2000).
and golf courses on the seafront (Doody 1989, Priest Excessive visitor pressure can damage dunes and
and others 1997). Dunes, reduced to small areas, have lead to the degradation of vegetation, which exposes
become habitats that are valuable not only because of the sand to rain and wind erosion (Doody 1989).
their rarity, but also because of their great biodiversity Moderate trampling can have a positive effect on
(Wanders 1987, De Raeve 1989, Dupont 1993). Manag- species diversity (Van der Maarel 1971). Disturbances
ers are now aware of the need to conserve sand dunes; such as that caused by grazing rabbits conserve ma-
they are protected from the most destructive activities ture dunes as small grasslands (Thomas 1960, ten
such as sand extraction, or motorcycling (Guilcher and Harkel and Van der Meulen 1996, Novo and Merino
Hallegouet 1991). They do, however, receive large
´ ¨ 1997). Trampling could also prevent scrub encroach-
numbers of tourist visitors, and trampling becomes a ment of coastal habitats (Goldsmith and others 1970,
problem as it represents the major disturbance affect- Andersen 1995) and retain the young sand dune
ing dune vegetation. It is therefore important to know successional stages.
how resistant dune vegetation is to trampling and its Trampling can be studied in two ways, from a
ability to regenerate. conservation viewpoint or as management tool, but
Trampling is an integral part of the problems of before drawing any conclusions, its effects on the
conservation management of natural areas that are at- habitat must be better understood. This paper de-
tractive for tourists (Gomez-Limon 1995, Toullec 1997, scribes the results of an experiment conducted on a
site that is heavily frequented in summer: the state-
owned dune at Quiberon (Brittany, France). The
KEY WORDS: Sand dune; Resilience; Resistance; Trampling; Vegeta-
tion
purpose of the research is to study the response to
trampling of three plant communities representing
*Author to whom correspondence should be addressed; email: the different dune landscape units: mobile dune,
servane.lemauviel@univ-rennesl.fr semifixed dune, and fixed dune.
Environmental Management Vol. 31, No. 2, pp. 227–235 © 2003 Springer-Verlag New York Inc.
228 S. Lemauviel and F. Roze
´
Materials and Methods in five contiguous 1-m-long permanent lines, in the
middle of each foot path. Each plant species in contact
The state-owned sand dune at Quiberon in the south
with each of the permanent points was recorded so as
of Brittany, France (47°30 N, 3°10 W) consists of a long
to estimate their frequency. The first measurement was
range of dunes extending from Gavres to Quiberon
ˆ
made immediately before trampling to determine the
(Guilcher and Hallegouet 1991). Recognized for its
´ ¨
initial state of the vegetation. Plants do not die imme-
great diversity of flora (Lahondere and Bioret 1997),
`
diately after trampling, so Cole and Bayfield (1993)
the dune has been the subject of a European project
suggested that the effects of trampling should be mea-
(LIFE) and forms part of the Natura 2000 network
sured several weeks later. The second measurements
(Romao 1997).
were therefore made two weeks after the disturbance.
Three vegetation communities are included in the
This timing is well adapted to herbaceous vegetation,
study. The mobile dune or yellow dune is an open tall
which changes very quickly at this season but is insuffi-
grassland dominated by Ammophila arenaria corre-
cient for the moss-lichen stratum, which was therefore
sponding to the phytosociological association Euphorbio-
excluded from our observations. The final measure-
Ammophiletum R. TX. 1945 (Gehu 1994). It develops on
´
ment was made one year after trampling to determine
the foredune where the dominant factor is sand mobil-
the extent that the plant cover had regenerated.
ity. When the relief decreases inland, the semifixed
The results were expressed as the relative frequency
dune also called a transition dune replaces this group-
(RF) compared to the control by using the equation of
ing. It is an open and short grassland belonging to the
Cole and Bayfield (1993):
Festuco-Galietum Gehu 1964 group (Gehu 1994). More
´ ´
inland, the fixed dune or grey dune colonizes the flat Frequency at time t
part of the landscape. It is a close and short grassland RF % cf 100
Initial frequency
very rich in species corresponding to the Roso-Ephedre-
tum (Ku ¨hnholtz-Lordat 1923, Vander Berghen 1958, where cf is the initial frequency in the control divided
Gehu 1994).
´ by the frequency at time t in the control.
The experimental protocol was based on that of As advocated Cole and Bayfield (1993), relative fre-
Cole and Bayfield (1993) and adapted to our study quency is based on the sum of the frequency of all
area. For each community, five simulated foot paths species. Relative frequency is more informative than
5 m long and 50 cm wide, were delineated in an ho- just a frequency of total vegetation as it integrates a
mogeneous area. Mobile dunes are characterized by possible superposition of the vegetation layers. Relative
their relief and a structure that changes from the beach frequencies were compared by the Kruskal-Wallis test
toward the rear dune. To minimize any effect of slope, (Scherrer 1984) in order to test the statistical effect of
foot paths were chosen parallel to the coastline along an increasing intensity of trampling. This nonparamet-
the top of a dune ridge. For the other communities, ric test was chosen because the data did not fit a normal
foot paths were chosen perpendicular to the coast line. distribution.
Five trampling treatments were conducted on each of To determine the response of plant communities to
the five foot paths. Control areas were selected in each trampling, two indices were chosen: resistance (Liddle
zone to be used as a baseline for comparing the effects 1975)— the ability of a community to not change after
of trampling. The intensity range of trampling treat- a disturbance—and resilience (Cole and Bayfield
ments varied, depending on the communities because 1993)—the ability to regenerate after a disturbance.
their immediate responses were very different. The Each of the graphs plotting the relative frequency
highest number of passages was first determined so as against the trampling intensity were fitted to a polyno-
to cause more than 50% reduction of the vegetation mial regression. Analyses of variance (Sokal and Rohlf
cover. Then, intermediate values were assigned to the 1981) were used to compare calculated data with ob-
other foot paths. Intensity of trampling varied from 0, servations.
50, 150, 500 to 1500 passages for the mobile dune, from The indices of resistance and resilience of the
25, 50, 75 to 125 passages for the semifixed dune, and plant communities were calculated from the polyno-
from 250, 500, 750 to 1000 passages for the fixed dune. mial regression curves. The resistance index of
The foot paths were trampled on a single date in June Liddle (1975) is read on the after-trampling curve. It
1998. is the trampling intensity leading to a 50% reduction
The vegetation was monitored using permanent in the relative frequency of the vegetation. The resil-
lines (Daget and Poissonet 1971, Forgeard and Touffet ience index is calculated, according to the definition
1980). Twenty permanent points are spaced 5 cm apart of Cole and Bayfield (1993), with the help of the
Response of Sand Dune Communities to Trampling 229
Table 1. Initial floristic description of the three communitiesa
Species RF
Mobile dune Semifixed dune Fixed dune
Ammophila arenaria 96.8 (1.4)
Elymus farctus 4.2 (2.0)
Atriplex hastata 0.2 (0.2)
Crithmum maritimum 0.2 (0.2)
Calystegia soldanella 0.8 (0.4) 0.6 (0.3)
Vulpia membranacea 1.2 (1.2) 3.8 (0.9)
Festuca rubra ssp. arenaria 8.2 (3.2) 2.6 (1.1) 0.4 (0.3)
Galium arenarium 5.6 (2.1) 0.4 (0.3) 1.2 (0.5)
Sedum acre 28.8 (3.0) 2.6 (1.0)
Mibora minima 9.8 (1.6) 3.4 (0.9)
Plantago coronopus 4.4 (1.5)
Plantago lanceolata 2.6 (1.1) 0.6 (0.3)
Leontodon taraxacoides 1.6 (0.5) 0.2 (0.2)
Carex arenaria 1.6 (1.2)
Herniaria ciliolata 1.6 (1.1)
Geranium molle 1 (0.5)
Ononis repens 0.2 (0.2)
Rosa pimpinellifolia 85 (2.9)
Ephedra distachya 68.6 (7.0)
Homalothecium lutescens 41.8 (7.8)
Cladonia sp. 40.4 (8.1)
Tortula ruralis ssp. ruraliformis 2 (0.7) 9.2 (3.9)
Erodium cicutarium 5.2 (1.4) 16.2 (6.8)
Cerastium diffusum 5.8 (1.4)
Arenaria serpyllifolia 4.6 (1.7)
Euphorbia portlandica 2.2 (0.6)
Desmazeria marina 3.6 (1.0)
Linaria arenaria 1.8 (0.8)
Scleranthus annuus 0.2 (0.2) 1.2 (0.7)
Agrostis capillaris 0.8 (0.4)
Helichrysum stoechas 0.8 (0.5)
Eryngium campestre 0.2 (0.2) 0.6 (0.6)
Phleum arenarium 0.4 (0.3)
Asparagus officinalis 0.4 (0.4)
Bromus hordeaceus 0.2 (0.2)
Medicago littoralis 0.2 (0.2)
Torilis nodosa 0.2 (0.2)
Cynodon dactylon 0.2 (0.2)
Sum of all species RF 117.2 (3.9) 70.6 (5.5) 292.6 (16.9)
Species richness 2.32 (0.2) 5.6 (0.3) 5.84 (0.3)
a
Means (and standard errors) correspond to all footpaths’ permanent lines before trampling.
1-year-after trampling curve. It is the percent of where RF1 is the relative frequency 2 weeks after tram-
change in relative frequency that occurs during 1 pling and RF2 is the relative frequency 1 year after
year following a 50% reduction in frequency caused trampling.
by trampling.
Resistance (number of passages) : Intensity Results
leading to RF1 50% The initial vegetation composition of the three com-
munities in all foot paths is given in Table 1. The
Resilience (%)
species compositions of the mobile dune and fixed
[RF2 (at the intensity of Resistance) 50] dune are quite different, while semifixed dune inte-
100
50 grates many species existing in one of the other two
230 S. Lemauviel and F. Roze
´
Table 2. Effect of trampling on relative frequencya
Mobile dune Semifixed dune Fixed dune
T 15 days T 1 year T 15 days T 1 year T 15 days T 1 year
H 16.26 11.05 12.00 12.16 13.37 10.46
P 0.005 0.05 0.05 0.05 0.01 0.05
a
T: trampling; H: statistic of the Kruskal-Wallis test.
Figure 1. Relative frequency (RF) of the vegetation of the mobile dune (A), the semifixed dune (B), and the fixed dune (C) 2
weeks after trampling and 1 year later.
communities. Ammophila arenaria dominate the mobile weeks and 1 year after trampling are compared in
dune community with a mean soil cover of 96.8%, while Figure 1A. The total relative frequency of the vegeta-
the frequency of other species are 10%. In the semi- tion two weeks after the disturbance fell to 25.2% after
fixed dune community, all species have low cover values just 50 passages, but as trampling increased, the relative
except Sedum acre at 28.8%. The fixed dune community frequency decreased only slightly from 17% at 250
is codominated by four species: two woody plants, Rosa passages to 5% at 1500 passages. One year after the
pimpinellifolia (85%) and Ephedra distachya (68.6%); a disturbance, the relative frequency had returned to
moss, Homalothecium lutescens (41.8%); and a lichen, high values, but recovery was greatest in places where
Cladonia sp. (40.4%). Species richness increases from trampling was the least.
the mobile dune to the semifixed dune and to the fixed The results of the effect of trampling on the semi-
dune communities. The sum of the species frequency fixed dune are shown in Figure 1B. The total relative
distinguishes the semifixed dune with a low value from frequency of the vegetation shortly after the trampling
the mobile dune and then the fixed dune with the decreased slightly more at higher trampling intensities.
highest value, 292.6% which reveals the important veg- One year later, on the 25-passages trail, the relative
etation density. frequency decreased, but it increased on trails with
In the three communities, trampling has a signifi- more passages, reaching 181.5% on the trail with 125
cant effect on the relative frequency of the vegetation 2 passages.
weeks after trampling as well as 1 year later (P 0.05) The effects of trampling on the total relative fre-
(Table 2). The relative frequencies of all the vegetation quency of the fixed dune are shown in Figure 1C. Two
of the mobile dune in relation to trampling intensity, 2 weeks after the trampling, the relative frequency fell to
Response of Sand Dune Communities to Trampling 231
Table 3. Resistance and resilience calculated from regression modelsa
Community Date Model r2 P Resistance Resilience
5 2
Mobile dune T 15 days 66.9731 0.181365x 9.37.10 x 0.542 0.0002 95.7
T 1 year 96.856 0.0843392x 3.42.10 5 x2 0.444 0.0016 78.2
Semifixed dune T 15 days 102.288 0.943138x 3.88.10 3x2 0.396 0.0039 85.1
T 1 year 92.0719 0.646688x 1.10.10 2 x2 0.355 0.0081 133.6
Fixed dune T 15 days 92.0107 0.0959693x 4.19.10 5x2 0.583 0.000 585.1
T 1 year 91.4864 0.0516813x 2.63.10 5x2 0.135 0.0784 40.54
a
The closeness-of-fit of the regressions was estimated by an ANOVA between the model and the experimental values.
resilience values well below 100%; the fixed dune had
the lowest value followed by the mobile dune.
Discussion
The resistance and resilience clearly distinguished
the different communities. The landscape units—the
fixed dune, the semifixed dune, and the mobile dune—
showed clearly different responses.
Our results can be compared with other published
studies on the effects of trampling on dune vegetation.
Among the experimental studies of the effects of tram-
pling, Bowles and Maun (1982) compared two plant
Figure 2. The relationship between resistance and resilience communities in the dunes of Pinery Provincial Park on
to trampling of three communities—a mobile dune, a semi- the shore of Lake Huron, one an open grassland dom-
fixed dune, and a fixed dune.
inated by Calamovilfa longifolia and the other a grass-
land with abundant lichens of the genus Cladonia.
These authors recorded a low resistance in the Calam-
almost half after 250 passages, but as the number of ovilfa longifolia grassland, which suffered heavy damage
passages increases, the relative frequency shows no with only 50 passages, compared to the lichen-rich
change. One year after the disturbance, the relative grassland, which survived the same trampling intensity
frequencies were higher than they were at 2 weeks, but despite some yellowing. Similarly, in communities more
no trail recovered to its pretrampling condition. similar to those of French coasts, Liddle (1975) re-
The curves showing variations in the relative fre- ported a resistance of 288 passages for a mobile dune
quency in relation to trampling intensity were modeled and 344 passages for a dune grassland on the English
using polynomial regressions (Table 3). The curves coast. The results of our study were not entirely the
were all fitted to equations of the type: Y a bx same, since the resistance obtained for the mobile dune
cx2 . All the polynomials were second order and were was 96 passages and 564 –585 passages for the fixed
significantly correlated with the original data (P dune. These studies do agree, however, in terms of the
0.1). low resistance of the mobile dune and the higher resis-
The polynomial equations were used to calculate the tance of the fixed dune.
resistance (Liddle 1975) and resilience indices (Cole Cole (1995a) established a negative correlation be-
and Bayfield 1993) for each plant community (Table tween resistance and resilience, we also found this in
3). The values obtained for the three communities are the Quiberon dune landscape. The fixed dune, with
compared in Figure 2. The mobile dune and semifixed high resistance and low resilience, contrasted with the
dune were rather similar with a very low resistance semifixed dune characterized by low resistance and
index, i.e., values of 100 passages. In contrast, the higher resilience. These contrasts suggest that the var-
fixed dune community had a much higher resistance of ious landscape units function differently, even though
nearly 600 passages. The semifixed dune, which had they are spatially close. Some of the vital attributes of
the lowest resistance value, also had the highest resil- these ecosystems (Aronson and others 1993, 1995)
ience value of 133.6%. The two other communities had clearly differentiate them from one another and can
232 S. Lemauviel and F. Roze
´
explain their responses to disturbance. These include short grassland with a high cover. The flat topography
the species richness of plants, the total cover estimated and the presence of woody species give the fixed dune
by the total frequency, and the presence or absence of a resistance to mechanical disturbance such as tram-
key species such Ammophila arenaria in the mobile dune pling. The woody stratum, once damaged, can only
or Rosa pimpinellifolia and Ephedra distachya in the fixed regenerate slowly. Furthermore, soil movements that
dune. can result from trampling have an adverse effect on the
The mobile dune zone has the form of a tall and plants of the fixed dune, which are maladapted to an
relatively dense grassland, almost exclusively domi- unstable substrate.
nated by Ammophila arenaria. This dune-building plant Two contradictory factors seem to regulate the re-
is the key species of the community in terms of geo- sponse of vegetation to trampling. The first effect is soil
morphology (Ku ¨hnoltz-Lordat 1923, Jungerius and Van compaction, which increases the soil density, decreases
der Meulen 1997) and in terms of biomass and species porosity, and increases the soil moisture content
richness (Lemauviel 2000). The capacity of the com- (Liddle and Greig-Smith 1975, Blom 1976, Blom and
munity to resist trampling, therefore depends mainly others 1979, Maun 1993). In a dry soil, compaction can
on the physical properties of marram grass. It is a lead to greater plant growth because of both the higher
perennial Poaceae with robust erect leaves and stems moisture content and also the improved rooting
and an extensive rhizome system. Erect herbs are the (Liddle and Greig Smith 1975). Soil compaction can
morphological type least resistant to trampling (Cole also favor seed germination because of the damper
1995a). The tufts of this tall grass break under the feet conditions. This is the case with some species of the
of walkers, so the immediate impact of the disturbance genus Plantago (Blom 1976). In contrast, trampling
is a major destruction of the plant cover. Yorks and adversely affects the vegetation by partially or totally
others (1997) studied the effects of the structure of the destroying it. This destruction takes place in the short
root system in the response of plants to trampling and term, but trampling can also influence the vegetation
did not find that rhizomes were an advantage in terms in the long term. The destruction of perennial species
of resilience. However, the rhizomes of marram grass will obviously have effects in subsequent years. The
are very long and ramified and this property surely same is true for all those plants that are damaged
allows the plant to persist even if the aboveground parts before they can produce seeds.
are destroyed. In the mobile dune, any change to the soil would
The semifixed dune, also called the transition dune, only be of short duration because of sand movements.
is covered with a short grassland similar to that of the More over, soil compaction cannot be considered as
fixed dunes, but is similar to the mobile dune in that it favorable on the mobile dune since the dominant spe-
has a low resistance. The term semifixed dune in par- cies, Ammophila arenaria, is not only adapted to the
ticular describes a very open dune community with an instability but also grows faster with sand burying
extremely low biomass compared to the mobile dune (Hutchings and de Kroon 1994).
and the fixed dune (Lemauviel 2000). The plant com- The semifixed dune is characterized by a rudimen-
munity is very diverse and includes both annual and tary soil that could benefit greatly from soil compaction
perennial species of short stature. The shallow rooted produced by heavy trampling intensities. The increased
vegetation is easily uprooted at low trampling intensi- moisture content and more stabilized soil could coun-
ties. On the other hand, the species of the semifixed terbalance the loss of plants. The semifixed dune veg-
dune are “stress tolerant ruderals” according to Grime’s etation cover benefited from the disturbance with its
(1979) definition. They are well adapted to distur- resilience of 134%. Soil compaction is much less of an
bances such as being buried by sand and have a great advantage in the fixed dune. This plant community,
capacity for regeneration. The initial relative frequen- with its high cover, has a deeper soil that retains mois-
cies were very low and resulted in high resilience values. ture more effectively. The major destruction of the
The fixed dune or grey dune is a relatively flat part vegetation would therefore mask any beneficial aspects
of the landscape on a well-stabilized soil with a high of compaction.
organic matter content. It is a mature plant community There is a curvilinear relation between trampling
(De Raeve 1989, Bonnot 1975). The flora is not exclu- and the response of the vegetation. The three plant
sively coastal (Vanden Berghen 1964) and includes a communities studied all fitted the same model, Y a
large number of species with diverse ecological niches. bx cx2, which is in agreement with the modeling of
The fixed dune is the richest zone of the landscape. In Cole (1995b), but the curves differed in their appear-
addition to herbaceous species, there is a dense carpet ance. It is interesting to note that for the semifixed
of mosses and lichens and woody species, forming a dune, the curve of the relative frequency, 1 year after
Response of Sand Dune Communities to Trampling 233
the disturbance, increased rapidly with increasing tram- Acknowledgments
pling intensity. The point of inflexion of the curve
We thank the O.N.F. and the C.E.L.R.L. for their
could therefore be an additional indicator for describ-
financial support. We are also grateful to Fanny Dubeau
ing the response of the vegetation 1 year after the
for her participation.
trampling. For a model of the type Y a bx cx2, the
point of inflexion occurs near to the value 0 of the
derivative, i.e., Y b 2cx, and therefore to the
Literature Cited
value b/2c of x. This index b/2c would tend toward zero
in classical curves such as those of the mobile dune and Andersen, U. V. 1995. Resistance of Danish coastal vegetation
fixed dunes, whereas it would approach a value of 1 in types to human trampling. Biological Conservartion 71:223–
230.
communities favored by trampling, such as the semi-
fixed dune. Aronson, J., C. Floret, E. Le Floc’h, C. Ovalle, and R. Ponta-
nier. 1993. Restoration and rehabilitation of degraded eco-
Goldsmith and others (1970) found that trampling systems in arid and semi-arid lands. I. A view from the south.
could be beneficial in coastal habitats, in the absence of Restoration Ecology 3:8 –17.
grazing, since it maintained an open vegetation. Aronson, J., C. Floret, E. Le Floc’h, C. Ovalle, and R. Ponta-
Andersen (1995) also recommended a reasonable in- nier. 1995. Restauration et rehabilitation des ecosystemes
´ ´ `
tensity of trampling for maintaining mobile dunes in a degrades en zones arides et semi-arides. Le vocabulaire et
´ ´
young stage. Nevertheless, trampling can be a danger- les concepts. Pages 11–29 in R. Pontanier, A. M’Hiri, N.
Akrimi, J. Aronson and E. Le Floc’h (eds.), L’homme
ous disturbance for an environment that has already
peut-il refaire ce qu’il a defait? John Libbey Eurotext, Paris.
´
been exposed to major stress (Slatter 1978). For exam-
Blom, C. W. P. M. 1976. Effects of trampling and soil compac-
ple, Maschinski and others (1997) found that tram- tion on the occurrence of some Plantago species in coastal
pling, combined with adverse weather conditions, sand dunes. Oecologia Plantarum 11:225–241.
could lead to the complete disappearance of popula- Blom, C. W. P. M., L. M. F. Husson, and V. Westhoff. 1979.
tions of sensitive species. The state-owned dunes at Effects of trampling and soil compaction on the occurrence
Quiberon are heavily grazed by rabbits, which could of some Plantago species in coastal sand dunes. IVa. The
explain the strong immediate impact of trampling on vegetation of two dune grasslands in relation to physical soil
factors. Proc K Ned Akad Wetensh Ser C 82:245–259.
the various landscape units. The resistance obtained for
the mobile dune and the semifixed dune were very low. Bonnot, E. J. 1975. Sur la place et le role des Bryophytes dans
ˆ
la vegetation des dunes. Pages 149 –158 in J. M. Gehu (ed.),
´ ´ ´
It was much higher for the fixed dune. The study site is Colloques phytosociologiques I. La vegetation des dunes
´ ´
very heavily visited, especially in July and August, and maritimes. J. Cramer, Vaduz.
tourists concentrate in strategic areas close to the car Bowles, J. M., and M. A. Maun. 1982. A study of the effects of
parks that serve the beach. In the long term, trampling trampling on the vegetation of Lake Huron sand dunes at
would seem to be very harmful for the fixed dune, Pinery Provincial Park. Biological Conservation 24:273–283.
which has a low resilience. Trampling will affect these Cole, D. N. 1995a. Experimental trampling of vegetation II.
communities for a long time. Andersen (1995) has Predictors of resistance and resilience. Journal of Applied
described the greater vulnerability of fixed dunes com- Ecology 32:215–224.
pared to mobile dunes. This part of the landscape is Cole, D. N. 1995b. Experimental trampling of vegetation. I.
Relationship between trampling intensity and vegetation.
therefore the community that should be given priority
Journal of Applied Ecology 32:203–214.
protection.
Cole, D. N., and N. G. Bayfield. 1993. Recreational trampling
The resistance and resilience indices defined by of vegetation: standard experimental procedures. Biological
Liddle (1975) and Cole and Bayfield (1993), respec- Conservation 63:209 –215.
tively, have been used in recent works as tools for Daget, P., and J. Poissonet. 1971. Principes d’une technique
characterizing plant communities in terms of their re- d’analyse quantitative de la vegetation des formations her-
´ ´
sponses to trampling (Cole 1995a, 1995b, Toullec and baces, Documents CEPE - CNRS 56:85–100.
´
others 1999, Toullec 1997, Gallet and Roze 2001, ´ De Raeve, F. 1989. Sand dune vegetation and management
2002). These indices are of value for managing the dynamics. Pages 99 –109 in F. Van der Meulen, P. D. Junge-
opening of sensitive areas to the public. The results rius and J. Visser (eds.), Perspectives in coastal dune man-
agement. SPB Academic Publishing, The Hague.
obtained for the mobile dune and especially for the
Doody, P. 1989. Conservation and development of the coastal
semifixed dune suggest that tourist pressure could eas-
dunes in Great Britain. Pages 53– 67 in F. Van der Meulen,
ily be included in site management. Nevertheless, the P. D. Jungerius and J. H. Visser (eds.), Perspectives in
concept of resilience only has meaning if the distur- coastal dune management. SPB Academic Publishing, The
bance is halted and management includes periods of Hague.
recovery. Dupont, P. 1993. La vegetation des dunes et de la lagune de
´ ´
234 S. Lemauviel and F. Roze
´
la Belle Henriette (Vendee). Interet, evolution, conse-
´ ´ ˆ ´ ´ Liddle, M. J. 1975. A theoretical relationship between the
quences de la fermeture de la lagune en 1971 et de la primary productivity of vegetation and its ability to tolerate
tempete de fevrier 1990. Bulletin de la Societe Botanique du
ˆ ´ ´´ trampling. Biological Conservation 8:251–255.
Centre Ouest 24:9 – 40. Liddle, M. J., and P. Greig-Smith. 1975. A survey of tracks and
Forgeard, F., and J. Touffet. 1980. La recolonisation des lan- paths in a sand dune ecosystem. I. Soils. Journal of Applied
des et des pelouses dans la region de Paimpont. Evolution
´ Ecology 12:893–908.
de la vegetation au cours des trois dernieres annees suivant
´ ´ ` ´ Maschinski, R., R. Frye, and S. Rutman. 1997. Demography
l’incendie. Bulletin d’Ecologie 11:349 –358. and population viability of an endangered plant species
Gallet, S., and F. Roze. 2001. Resistance of Atlantic heathlands
´ before and after protection from trampling. Conservation
to trampling in Brittany (France): influence of vegetation Biology 11:990 –999.
type, season and weather conditions. Biological Conservation Maun, M. A. 1993. Dry coastal ecosystems along the Great
97:189 –198. Lakes. Pages 299 –316 in E. Van der Maarel (ed.), Ecosys-
Gallet, S., and F. Roze. 2002. Long-term effects of trampling
´ tems of the world. 2B: Dry coastal ecosystems. Elsevier,
on Atlantic heathland in Brittany (France): resilience and Amsterdam.
tolerance in relation to season and meterological condi- McDonnell, M. J. 1981. Trampling effects on coastal Dune
tions. Biological Conservation 103:267–275. vegetation in the Parker river national wildlife refuge, Mas-
Gehu, J. M. 1994. Schema synsystematique et typologique des
´ ´ ´ sachusetts, USA. Biological Conservation 21:289 –301.
milieux littoraux Francais Atlantiques et Mediterraneens.
¸ ´ ´ Nova, F. G., and J. Merino. 1997. Pattern and process in the
Pages 183–212 in J. M. Gehu (ed.), Colloques phytoso-
´ dune system of the Donana national park, south-western
ciologiques XXII. La syntaxonomie et la synsystematique
´ Spain. Pages 453– 468 in E. Vander Maarel (ed.), Ecosys-
Europeennes comme base typologique des habitats. J. Cra-
´ tems of the world 2C: Dry coastal ecosystems. Elsevier,
mer, Vaduz. Amsterdam.
Goldsmith, F. B., R. J. C. Munton, and A. Warren. 1970. The Olsauskas, A. 1996. Influence of recreation on plant cover in
impact of recreation on the ecology and amenity of west Lithuania. Klaipedos Universiteto Leidykla, 210 pp.
semi-natural areas: methods of investigation used in the Paskoff, R. 1989. Les dunes du littoral. La Recherche 212:888 –
Isles of Scilly. Biological Journal of Linnean Society 2:287– 895.
306. Priest, S., E. Van der Maarel, and M. B. Usher. 1997. Recre-
Gomez-Limon, F. J. 1995. Recreational activities and loss of ational use of dry coastal ecosystems. Pages 519 –529 in E.
diversity in grasslands in Alta Manzanares Natural Park, Van der Maarel (ed.), Ecosystems of the world 2C: Dry
Spain. Biological Conservation 74:99 –105. coastal ecosystems. Elsevier, Amsterdam.
Grime, J. P. 1979. Plant strategies and vegetation processes. Romao, C. 1997. Natura 2000 —Manuel d’interpretation des
´
Wiley, Chichester, UK, 222 pp. habitats de l’Union Europeenne Version EUR 15, Commis-
´
Guilcher, A., and B. Hallegouet. 1991. Coastal dunes in Brit-
´ ¨ sion Europeenne, 103 pp.
´
tany and their management. Journal of Coastal Research Scherrer, B. 1984. Biostatistiques. Gaetan Morin, Boucher-
7:517–533. ville, 850 pp.
Hutchings, M. J., and H. de Kroon. 1994. Foraging in Slatter, R. J. 1978. Ecological effects of trampling on sand
plants: the role of morphological plasticity in resource dune vegetation. Journal of Biological Education 12:89 –96.
acquisition. Pages 159 –238 in M. Begon and A. H. Fitter Sokal, R. R., and F. L. Rohlf. 1969. Biometry. Freeman and
(eds.), Advances in ecological research. Academic Press, Company, New York, 859 pp.
London. ten Harkel, M. J., and F. Van der Meulen. 1996. Impact of
Hylgaard, T., and M. J. Liddle. 1981. The effect of human grazing and atmospheric nitrogen deposition on the vege-
trampling on a sand dune ecosystem dominated by Em- tation of dry coastal dune grasslands. Journal of Vegetation
petrum nigrum. Journal of Applied Ecology 18:559 –569. Science 7:445– 452.
Jungerius, P. D., and F. Van der Meulen. 1997. Aeolian dy- Thomas, A. S. 1960. Changes in vegetation since the advent of
namics in relation to vegetation in a blowout complex in myxomatosis. Journal of Ecology 48:287–307.
the Meijendel dunes, The Netherlands. Journal of Coastal Toullec, H. 1997. Dynamique des systemes perturbes: degra-
` ´ ´
Conservation 3:63–70. dation, restauration et gestion des landes pietinees. Ex-
´ ´
Ku
¨hnholtz-Lordat, G. 1923. Les dunes du Golfe du Lion. emple du site megalithique de Carnac-Bretagne. These
´ `
These d’etat. Universite de Montpellier, Montpellier,
` ´ ´ d’universite. Rennes 1, Rennes, France, 136 pp.
´
France, 307 pp. Toullec, H., S. Diquelou, F. Roze, and J. C. Gloaguen. 1999.
´ ´
Lahondere, C., and F. Bioret. 1997. Quelques aspects de la
` Reponse d’une lande atlantique au pietinement experimen-
´ ´ ´
vegetation littorale du Morbihan continental. Bulletin de la
´ ´ tal. Compte Rendu d’Academie des Sciences de la vie 322:809 – 815.
´
societe botanique du centre ouest 28:305–314.
´´ Van der Maarel, E. 1971. Plant species diversity in relation to
Lemauviel, S. 2000. Les dunes grises des cotes Atlantiques:
ˆ management. Pages 45– 63 in E. Duffey and A. S. Watt
fonctionnement, dynamique potentielle, principes de ges- (eds.), The scientific management of animal and plant
tion conservatoire et processus de restauration. These ` communities for conservation. Blackwell Scientific Publica-
d’universite: Universite de Rennes 1, Rennes, France, 264
´ ´ tions, Oxford.
pp. Van der Meulen, F. 1997. Dune water catchment in the Neth-
Response of Sand Dune Communities to Trampling 235
erlands. Pages 533–556 in E. Van der Maarel (ed.), Ecosys- Meulen, P. D. Jungerius and J. Visser (eds.), Perspectives in
tems of the world 2C: Dry coastal ecosystems. Elsevier, coastal dune management. SPB Academic Publishing, The
Amsterdam. Hague.
Vanden Berghen, C. 1964. La vegetation terrestre du littoral
´ ´ Yorks, T. P., N. E. West, M. R. J., and S. D. Waren. 1997.
de l’Europe occidentale. Les Naturalistes Belges 45:251–280. Toleration of traffic by vegetation. Life form conclusions
Wanders, E. 1989. Perspectives in coastal dune management and summary extracts from a comprehensive data base.
towards a dynamic approach. Pages 141–148 in F. Van der Environmental Management 21:121–131.
Response of Three Plant Communities to Trampling
in a Sand Dune System in Brittany (France)
SERVANE LEMAUVIEL* used to compare three typical plant communities belonging to
FRANCOISE ROZE
¸ ´ the various landscape units: mobile dune, semifixed dune, and
Universite de Rennes 1
´ fixed dune. The strong contrasts between communities be-
Equipe “Dynamique des communautes” ´ longing to different successional stages reflect their ecological
U.M.R. 6553, Service d’Ecologie Vegetale
´ ´
functioning. The mobile dune and semifixed dune with their
Complexe scientifique de Beaulieu
low resistance contrasted with the fixed dune. Only the vege-
35042 Rennes cedex, France
tation cover of the semifixed dune benefited from long-term
ABSTRACT / Dunes that are protected because of their very trampling and had a very high resilience (134%). This re-
rich and diverse plant communities are often exposed to ex- sponse could be explained by a good balance of two opposite
cessive visitor pressure. The effects of trampling on the habitat factors: soil compaction increasing soil stability and moisture
must be known from a conservation viewpoint but also are content, and vegetation destruction. Because of their low re-
important for management. To determine the response of silience, trampling seems to be harmful for fixed dunes in the
plant assemblages to trampling by people, an experimental long term. The tourist pressure seems easier to integrate in to
study was conducted on the state-owned dunes at Quiberon the mobile dunes and the semifixed dunes if periods of recov-
(Brittany, France). Indices of resistance and resilience were ery are included in the management.
During the 20th century, human activities linked to Toullec and others 1999, Gallet and Roze 2001, 2002).
´
tourism have affected and damaged many coastal areas The coastline is affected and many authors have chosen
(Liddle and Greig-Smith 1975, Paskoff 1989, Van der sand dunes to study the effect of trampling on vegeta-
Meulen 1997, Olsauskas 1996). The interest in seaside tion (Liddle and Greig-Smith 1975, Slatter 1978, Hyl-
recreation and living has led to the construction of gaard and Liddle 1981, McDonnell 1981, Bowles and
seaside resorts, including hotels, apartments, campsites, Maun 1982, Andersen 1995, Lemauviel 2000).
and golf courses on the seafront (Doody 1989, Priest Excessive visitor pressure can damage dunes and
and others 1997). Dunes, reduced to small areas, have lead to the degradation of vegetation, which exposes
become habitats that are valuable not only because of the sand to rain and wind erosion (Doody 1989).
their rarity, but also because of their great biodiversity Moderate trampling can have a positive effect on
(Wanders 1987, De Raeve 1989, Dupont 1993). Manag- species diversity (Van der Maarel 1971). Disturbances
ers are now aware of the need to conserve sand dunes; such as that caused by grazing rabbits conserve ma-
they are protected from the most destructive activities ture dunes as small grasslands (Thomas 1960, ten
such as sand extraction, or motorcycling (Guilcher and Harkel and Van der Meulen 1996, Novo and Merino
Hallegouet 1991). They do, however, receive large
´ ¨ 1997). Trampling could also prevent scrub encroach-
numbers of tourist visitors, and trampling becomes a ment of coastal habitats (Goldsmith and others 1970,
problem as it represents the major disturbance affect- Andersen 1995) and retain the young sand dune
ing dune vegetation. It is therefore important to know successional stages.
how resistant dune vegetation is to trampling and its Trampling can be studied in two ways, from a
ability to regenerate. conservation viewpoint or as management tool, but
Trampling is an integral part of the problems of before drawing any conclusions, its effects on the
conservation management of natural areas that are at- habitat must be better understood. This paper de-
tractive for tourists (Gomez-Limon 1995, Toullec 1997, scribes the results of an experiment conducted on a
site that is heavily frequented in summer: the state-
owned dune at Quiberon (Brittany, France). The
KEY WORDS: Sand dune; Resilience; Resistance; Trampling; Vegeta-
tion
purpose of the research is to study the response to
trampling of three plant communities representing
*Author to whom correspondence should be addressed; email: the different dune landscape units: mobile dune,
servane.lemauviel@univ-rennesl.fr semifixed dune, and fixed dune.
Environmental Management Vol. 31, No. 2, pp. 227–235 © 2003 Springer-Verlag New York Inc.
228 S. Lemauviel and F. Roze
´
Materials and Methods in five contiguous 1-m-long permanent lines, in the
middle of each foot path. Each plant species in contact
The state-owned sand dune at Quiberon in the south
with each of the permanent points was recorded so as
of Brittany, France (47°30 N, 3°10 W) consists of a long
to estimate their frequency. The first measurement was
range of dunes extending from Gavres to Quiberon
ˆ
made immediately before trampling to determine the
(Guilcher and Hallegouet 1991). Recognized for its
´ ¨
initial state of the vegetation. Plants do not die imme-
great diversity of flora (Lahondere and Bioret 1997),
`
diately after trampling, so Cole and Bayfield (1993)
the dune has been the subject of a European project
suggested that the effects of trampling should be mea-
(LIFE) and forms part of the Natura 2000 network
sured several weeks later. The second measurements
(Romao 1997).
were therefore made two weeks after the disturbance.
Three vegetation communities are included in the
This timing is well adapted to herbaceous vegetation,
study. The mobile dune or yellow dune is an open tall
which changes very quickly at this season but is insuffi-
grassland dominated by Ammophila arenaria corre-
cient for the moss-lichen stratum, which was therefore
sponding to the phytosociological association Euphorbio-
excluded from our observations. The final measure-
Ammophiletum R. TX. 1945 (Gehu 1994). It develops on
´
ment was made one year after trampling to determine
the foredune where the dominant factor is sand mobil-
the extent that the plant cover had regenerated.
ity. When the relief decreases inland, the semifixed
The results were expressed as the relative frequency
dune also called a transition dune replaces this group-
(RF) compared to the control by using the equation of
ing. It is an open and short grassland belonging to the
Cole and Bayfield (1993):
Festuco-Galietum Gehu 1964 group (Gehu 1994). More
´ ´
inland, the fixed dune or grey dune colonizes the flat Frequency at time t
part of the landscape. It is a close and short grassland RF % cf 100
Initial frequency
very rich in species corresponding to the Roso-Ephedre-
tum (Ku ¨hnholtz-Lordat 1923, Vander Berghen 1958, where cf is the initial frequency in the control divided
Gehu 1994).
´ by the frequency at time t in the control.
The experimental protocol was based on that of As advocated Cole and Bayfield (1993), relative fre-
Cole and Bayfield (1993) and adapted to our study quency is based on the sum of the frequency of all
area. For each community, five simulated foot paths species. Relative frequency is more informative than
5 m long and 50 cm wide, were delineated in an ho- just a frequency of total vegetation as it integrates a
mogeneous area. Mobile dunes are characterized by possible superposition of the vegetation layers. Relative
their relief and a structure that changes from the beach frequencies were compared by the Kruskal-Wallis test
toward the rear dune. To minimize any effect of slope, (Scherrer 1984) in order to test the statistical effect of
foot paths were chosen parallel to the coastline along an increasing intensity of trampling. This nonparamet-
the top of a dune ridge. For the other communities, ric test was chosen because the data did not fit a normal
foot paths were chosen perpendicular to the coast line. distribution.
Five trampling treatments were conducted on each of To determine the response of plant communities to
the five foot paths. Control areas were selected in each trampling, two indices were chosen: resistance (Liddle
zone to be used as a baseline for comparing the effects 1975)— the ability of a community to not change after
of trampling. The intensity range of trampling treat- a disturbance—and resilience (Cole and Bayfield
ments varied, depending on the communities because 1993)—the ability to regenerate after a disturbance.
their immediate responses were very different. The Each of the graphs plotting the relative frequency
highest number of passages was first determined so as against the trampling intensity were fitted to a polyno-
to cause more than 50% reduction of the vegetation mial regression. Analyses of variance (Sokal and Rohlf
cover. Then, intermediate values were assigned to the 1981) were used to compare calculated data with ob-
other foot paths. Intensity of trampling varied from 0, servations.
50, 150, 500 to 1500 passages for the mobile dune, from The indices of resistance and resilience of the
25, 50, 75 to 125 passages for the semifixed dune, and plant communities were calculated from the polyno-
from 250, 500, 750 to 1000 passages for the fixed dune. mial regression curves. The resistance index of
The foot paths were trampled on a single date in June Liddle (1975) is read on the after-trampling curve. It
1998. is the trampling intensity leading to a 50% reduction
The vegetation was monitored using permanent in the relative frequency of the vegetation. The resil-
lines (Daget and Poissonet 1971, Forgeard and Touffet ience index is calculated, according to the definition
1980). Twenty permanent points are spaced 5 cm apart of Cole and Bayfield (1993), with the help of the
Response of Sand Dune Communities to Trampling 229
Table 1. Initial floristic description of the three communitiesa
Species RF
Mobile dune Semifixed dune Fixed dune
Ammophila arenaria 96.8 (1.4)
Elymus farctus 4.2 (2.0)
Atriplex hastata 0.2 (0.2)
Crithmum maritimum 0.2 (0.2)
Calystegia soldanella 0.8 (0.4) 0.6 (0.3)
Vulpia membranacea 1.2 (1.2) 3.8 (0.9)
Festuca rubra ssp. arenaria 8.2 (3.2) 2.6 (1.1) 0.4 (0.3)
Galium arenarium 5.6 (2.1) 0.4 (0.3) 1.2 (0.5)
Sedum acre 28.8 (3.0) 2.6 (1.0)
Mibora minima 9.8 (1.6) 3.4 (0.9)
Plantago coronopus 4.4 (1.5)
Plantago lanceolata 2.6 (1.1) 0.6 (0.3)
Leontodon taraxacoides 1.6 (0.5) 0.2 (0.2)
Carex arenaria 1.6 (1.2)
Herniaria ciliolata 1.6 (1.1)
Geranium molle 1 (0.5)
Ononis repens 0.2 (0.2)
Rosa pimpinellifolia 85 (2.9)
Ephedra distachya 68.6 (7.0)
Homalothecium lutescens 41.8 (7.8)
Cladonia sp. 40.4 (8.1)
Tortula ruralis ssp. ruraliformis 2 (0.7) 9.2 (3.9)
Erodium cicutarium 5.2 (1.4) 16.2 (6.8)
Cerastium diffusum 5.8 (1.4)
Arenaria serpyllifolia 4.6 (1.7)
Euphorbia portlandica 2.2 (0.6)
Desmazeria marina 3.6 (1.0)
Linaria arenaria 1.8 (0.8)
Scleranthus annuus 0.2 (0.2) 1.2 (0.7)
Agrostis capillaris 0.8 (0.4)
Helichrysum stoechas 0.8 (0.5)
Eryngium campestre 0.2 (0.2) 0.6 (0.6)
Phleum arenarium 0.4 (0.3)
Asparagus officinalis 0.4 (0.4)
Bromus hordeaceus 0.2 (0.2)
Medicago littoralis 0.2 (0.2)
Torilis nodosa 0.2 (0.2)
Cynodon dactylon 0.2 (0.2)
Sum of all species RF 117.2 (3.9) 70.6 (5.5) 292.6 (16.9)
Species richness 2.32 (0.2) 5.6 (0.3) 5.84 (0.3)
a
Means (and standard errors) correspond to all footpaths’ permanent lines before trampling.
1-year-after trampling curve. It is the percent of where RF1 is the relative frequency 2 weeks after tram-
change in relative frequency that occurs during 1 pling and RF2 is the relative frequency 1 year after
year following a 50% reduction in frequency caused trampling.
by trampling.
Resistance (number of passages) : Intensity Results
leading to RF1 50% The initial vegetation composition of the three com-
munities in all foot paths is given in Table 1. The
Resilience (%)
species compositions of the mobile dune and fixed
[RF2 (at the intensity of Resistance) 50] dune are quite different, while semifixed dune inte-
100
50 grates many species existing in one of the other two
230 S. Lemauviel and F. Roze
´
Table 2. Effect of trampling on relative frequencya
Mobile dune Semifixed dune Fixed dune
T 15 days T 1 year T 15 days T 1 year T 15 days T 1 year
H 16.26 11.05 12.00 12.16 13.37 10.46
P 0.005 0.05 0.05 0.05 0.01 0.05
a
T: trampling; H: statistic of the Kruskal-Wallis test.
Figure 1. Relative frequency (RF) of the vegetation of the mobile dune (A), the semifixed dune (B), and the fixed dune (C) 2
weeks after trampling and 1 year later.
communities. Ammophila arenaria dominate the mobile weeks and 1 year after trampling are compared in
dune community with a mean soil cover of 96.8%, while Figure 1A. The total relative frequency of the vegeta-
the frequency of other species are 10%. In the semi- tion two weeks after the disturbance fell to 25.2% after
fixed dune community, all species have low cover values just 50 passages, but as trampling increased, the relative
except Sedum acre at 28.8%. The fixed dune community frequency decreased only slightly from 17% at 250
is codominated by four species: two woody plants, Rosa passages to 5% at 1500 passages. One year after the
pimpinellifolia (85%) and Ephedra distachya (68.6%); a disturbance, the relative frequency had returned to
moss, Homalothecium lutescens (41.8%); and a lichen, high values, but recovery was greatest in places where
Cladonia sp. (40.4%). Species richness increases from trampling was the least.
the mobile dune to the semifixed dune and to the fixed The results of the effect of trampling on the semi-
dune communities. The sum of the species frequency fixed dune are shown in Figure 1B. The total relative
distinguishes the semifixed dune with a low value from frequency of the vegetation shortly after the trampling
the mobile dune and then the fixed dune with the decreased slightly more at higher trampling intensities.
highest value, 292.6% which reveals the important veg- One year later, on the 25-passages trail, the relative
etation density. frequency decreased, but it increased on trails with
In the three communities, trampling has a signifi- more passages, reaching 181.5% on the trail with 125
cant effect on the relative frequency of the vegetation 2 passages.
weeks after trampling as well as 1 year later (P 0.05) The effects of trampling on the total relative fre-
(Table 2). The relative frequencies of all the vegetation quency of the fixed dune are shown in Figure 1C. Two
of the mobile dune in relation to trampling intensity, 2 weeks after the trampling, the relative frequency fell to
Response of Sand Dune Communities to Trampling 231
Table 3. Resistance and resilience calculated from regression modelsa
Community Date Model r2 P Resistance Resilience
5 2
Mobile dune T 15 days 66.9731 0.181365x 9.37.10 x 0.542 0.0002 95.7
T 1 year 96.856 0.0843392x 3.42.10 5 x2 0.444 0.0016 78.2
Semifixed dune T 15 days 102.288 0.943138x 3.88.10 3x2 0.396 0.0039 85.1
T 1 year 92.0719 0.646688x 1.10.10 2 x2 0.355 0.0081 133.6
Fixed dune T 15 days 92.0107 0.0959693x 4.19.10 5x2 0.583 0.000 585.1
T 1 year 91.4864 0.0516813x 2.63.10 5x2 0.135 0.0784 40.54
a
The closeness-of-fit of the regressions was estimated by an ANOVA between the model and the experimental values.
resilience values well below 100%; the fixed dune had
the lowest value followed by the mobile dune.
Discussion
The resistance and resilience clearly distinguished
the different communities. The landscape units—the
fixed dune, the semifixed dune, and the mobile dune—
showed clearly different responses.
Our results can be compared with other published
studies on the effects of trampling on dune vegetation.
Among the experimental studies of the effects of tram-
pling, Bowles and Maun (1982) compared two plant
Figure 2. The relationship between resistance and resilience communities in the dunes of Pinery Provincial Park on
to trampling of three communities—a mobile dune, a semi- the shore of Lake Huron, one an open grassland dom-
fixed dune, and a fixed dune.
inated by Calamovilfa longifolia and the other a grass-
land with abundant lichens of the genus Cladonia.
These authors recorded a low resistance in the Calam-
almost half after 250 passages, but as the number of ovilfa longifolia grassland, which suffered heavy damage
passages increases, the relative frequency shows no with only 50 passages, compared to the lichen-rich
change. One year after the disturbance, the relative grassland, which survived the same trampling intensity
frequencies were higher than they were at 2 weeks, but despite some yellowing. Similarly, in communities more
no trail recovered to its pretrampling condition. similar to those of French coasts, Liddle (1975) re-
The curves showing variations in the relative fre- ported a resistance of 288 passages for a mobile dune
quency in relation to trampling intensity were modeled and 344 passages for a dune grassland on the English
using polynomial regressions (Table 3). The curves coast. The results of our study were not entirely the
were all fitted to equations of the type: Y a bx same, since the resistance obtained for the mobile dune
cx2 . All the polynomials were second order and were was 96 passages and 564 –585 passages for the fixed
significantly correlated with the original data (P dune. These studies do agree, however, in terms of the
0.1). low resistance of the mobile dune and the higher resis-
The polynomial equations were used to calculate the tance of the fixed dune.
resistance (Liddle 1975) and resilience indices (Cole Cole (1995a) established a negative correlation be-
and Bayfield 1993) for each plant community (Table tween resistance and resilience, we also found this in
3). The values obtained for the three communities are the Quiberon dune landscape. The fixed dune, with
compared in Figure 2. The mobile dune and semifixed high resistance and low resilience, contrasted with the
dune were rather similar with a very low resistance semifixed dune characterized by low resistance and
index, i.e., values of 100 passages. In contrast, the higher resilience. These contrasts suggest that the var-
fixed dune community had a much higher resistance of ious landscape units function differently, even though
nearly 600 passages. The semifixed dune, which had they are spatially close. Some of the vital attributes of
the lowest resistance value, also had the highest resil- these ecosystems (Aronson and others 1993, 1995)
ience value of 133.6%. The two other communities had clearly differentiate them from one another and can
232 S. Lemauviel and F. Roze
´
explain their responses to disturbance. These include short grassland with a high cover. The flat topography
the species richness of plants, the total cover estimated and the presence of woody species give the fixed dune
by the total frequency, and the presence or absence of a resistance to mechanical disturbance such as tram-
key species such Ammophila arenaria in the mobile dune pling. The woody stratum, once damaged, can only
or Rosa pimpinellifolia and Ephedra distachya in the fixed regenerate slowly. Furthermore, soil movements that
dune. can result from trampling have an adverse effect on the
The mobile dune zone has the form of a tall and plants of the fixed dune, which are maladapted to an
relatively dense grassland, almost exclusively domi- unstable substrate.
nated by Ammophila arenaria. This dune-building plant Two contradictory factors seem to regulate the re-
is the key species of the community in terms of geo- sponse of vegetation to trampling. The first effect is soil
morphology (Ku ¨hnoltz-Lordat 1923, Jungerius and Van compaction, which increases the soil density, decreases
der Meulen 1997) and in terms of biomass and species porosity, and increases the soil moisture content
richness (Lemauviel 2000). The capacity of the com- (Liddle and Greig-Smith 1975, Blom 1976, Blom and
munity to resist trampling, therefore depends mainly others 1979, Maun 1993). In a dry soil, compaction can
on the physical properties of marram grass. It is a lead to greater plant growth because of both the higher
perennial Poaceae with robust erect leaves and stems moisture content and also the improved rooting
and an extensive rhizome system. Erect herbs are the (Liddle and Greig Smith 1975). Soil compaction can
morphological type least resistant to trampling (Cole also favor seed germination because of the damper
1995a). The tufts of this tall grass break under the feet conditions. This is the case with some species of the
of walkers, so the immediate impact of the disturbance genus Plantago (Blom 1976). In contrast, trampling
is a major destruction of the plant cover. Yorks and adversely affects the vegetation by partially or totally
others (1997) studied the effects of the structure of the destroying it. This destruction takes place in the short
root system in the response of plants to trampling and term, but trampling can also influence the vegetation
did not find that rhizomes were an advantage in terms in the long term. The destruction of perennial species
of resilience. However, the rhizomes of marram grass will obviously have effects in subsequent years. The
are very long and ramified and this property surely same is true for all those plants that are damaged
allows the plant to persist even if the aboveground parts before they can produce seeds.
are destroyed. In the mobile dune, any change to the soil would
The semifixed dune, also called the transition dune, only be of short duration because of sand movements.
is covered with a short grassland similar to that of the More over, soil compaction cannot be considered as
fixed dunes, but is similar to the mobile dune in that it favorable on the mobile dune since the dominant spe-
has a low resistance. The term semifixed dune in par- cies, Ammophila arenaria, is not only adapted to the
ticular describes a very open dune community with an instability but also grows faster with sand burying
extremely low biomass compared to the mobile dune (Hutchings and de Kroon 1994).
and the fixed dune (Lemauviel 2000). The plant com- The semifixed dune is characterized by a rudimen-
munity is very diverse and includes both annual and tary soil that could benefit greatly from soil compaction
perennial species of short stature. The shallow rooted produced by heavy trampling intensities. The increased
vegetation is easily uprooted at low trampling intensi- moisture content and more stabilized soil could coun-
ties. On the other hand, the species of the semifixed terbalance the loss of plants. The semifixed dune veg-
dune are “stress tolerant ruderals” according to Grime’s etation cover benefited from the disturbance with its
(1979) definition. They are well adapted to distur- resilience of 134%. Soil compaction is much less of an
bances such as being buried by sand and have a great advantage in the fixed dune. This plant community,
capacity for regeneration. The initial relative frequen- with its high cover, has a deeper soil that retains mois-
cies were very low and resulted in high resilience values. ture more effectively. The major destruction of the
The fixed dune or grey dune is a relatively flat part vegetation would therefore mask any beneficial aspects
of the landscape on a well-stabilized soil with a high of compaction.
organic matter content. It is a mature plant community There is a curvilinear relation between trampling
(De Raeve 1989, Bonnot 1975). The flora is not exclu- and the response of the vegetation. The three plant
sively coastal (Vanden Berghen 1964) and includes a communities studied all fitted the same model, Y a
large number of species with diverse ecological niches. bx cx2, which is in agreement with the modeling of
The fixed dune is the richest zone of the landscape. In Cole (1995b), but the curves differed in their appear-
addition to herbaceous species, there is a dense carpet ance. It is interesting to note that for the semifixed
of mosses and lichens and woody species, forming a dune, the curve of the relative frequency, 1 year after
Response of Sand Dune Communities to Trampling 233
the disturbance, increased rapidly with increasing tram- Acknowledgments
pling intensity. The point of inflexion of the curve
We thank the O.N.F. and the C.E.L.R.L. for their
could therefore be an additional indicator for describ-
financial support. We are also grateful to Fanny Dubeau
ing the response of the vegetation 1 year after the
for her participation.
trampling. For a model of the type Y a bx cx2, the
point of inflexion occurs near to the value 0 of the
derivative, i.e., Y b 2cx, and therefore to the
Literature Cited
value b/2c of x. This index b/2c would tend toward zero
in classical curves such as those of the mobile dune and Andersen, U. V. 1995. Resistance of Danish coastal vegetation
fixed dunes, whereas it would approach a value of 1 in types to human trampling. Biological Conservartion 71:223–
230.
communities favored by trampling, such as the semi-
fixed dune. Aronson, J., C. Floret, E. Le Floc’h, C. Ovalle, and R. Ponta-
nier. 1993. Restoration and rehabilitation of degraded eco-
Goldsmith and others (1970) found that trampling systems in arid and semi-arid lands. I. A view from the south.
could be beneficial in coastal habitats, in the absence of Restoration Ecology 3:8 –17.
grazing, since it maintained an open vegetation. Aronson, J., C. Floret, E. Le Floc’h, C. Ovalle, and R. Ponta-
Andersen (1995) also recommended a reasonable in- nier. 1995. Restauration et rehabilitation des ecosystemes
´ ´ `
tensity of trampling for maintaining mobile dunes in a degrades en zones arides et semi-arides. Le vocabulaire et
´ ´
young stage. Nevertheless, trampling can be a danger- les concepts. Pages 11–29 in R. Pontanier, A. M’Hiri, N.
Akrimi, J. Aronson and E. Le Floc’h (eds.), L’homme
ous disturbance for an environment that has already
peut-il refaire ce qu’il a defait? John Libbey Eurotext, Paris.
´
been exposed to major stress (Slatter 1978). For exam-
Blom, C. W. P. M. 1976. Effects of trampling and soil compac-
ple, Maschinski and others (1997) found that tram- tion on the occurrence of some Plantago species in coastal
pling, combined with adverse weather conditions, sand dunes. Oecologia Plantarum 11:225–241.
could lead to the complete disappearance of popula- Blom, C. W. P. M., L. M. F. Husson, and V. Westhoff. 1979.
tions of sensitive species. The state-owned dunes at Effects of trampling and soil compaction on the occurrence
Quiberon are heavily grazed by rabbits, which could of some Plantago species in coastal sand dunes. IVa. The
explain the strong immediate impact of trampling on vegetation of two dune grasslands in relation to physical soil
factors. Proc K Ned Akad Wetensh Ser C 82:245–259.
the various landscape units. The resistance obtained for
the mobile dune and the semifixed dune were very low. Bonnot, E. J. 1975. Sur la place et le role des Bryophytes dans
ˆ
la vegetation des dunes. Pages 149 –158 in J. M. Gehu (ed.),
´ ´ ´
It was much higher for the fixed dune. The study site is Colloques phytosociologiques I. La vegetation des dunes
´ ´
very heavily visited, especially in July and August, and maritimes. J. Cramer, Vaduz.
tourists concentrate in strategic areas close to the car Bowles, J. M., and M. A. Maun. 1982. A study of the effects of
parks that serve the beach. In the long term, trampling trampling on the vegetation of Lake Huron sand dunes at
would seem to be very harmful for the fixed dune, Pinery Provincial Park. Biological Conservation 24:273–283.
which has a low resilience. Trampling will affect these Cole, D. N. 1995a. Experimental trampling of vegetation II.
communities for a long time. Andersen (1995) has Predictors of resistance and resilience. Journal of Applied
described the greater vulnerability of fixed dunes com- Ecology 32:215–224.
pared to mobile dunes. This part of the landscape is Cole, D. N. 1995b. Experimental trampling of vegetation. I.
Relationship between trampling intensity and vegetation.
therefore the community that should be given priority
Journal of Applied Ecology 32:203–214.
protection.
Cole, D. N., and N. G. Bayfield. 1993. Recreational trampling
The resistance and resilience indices defined by of vegetation: standard experimental procedures. Biological
Liddle (1975) and Cole and Bayfield (1993), respec- Conservation 63:209 –215.
tively, have been used in recent works as tools for Daget, P., and J. Poissonet. 1971. Principes d’une technique
characterizing plant communities in terms of their re- d’analyse quantitative de la vegetation des formations her-
´ ´
sponses to trampling (Cole 1995a, 1995b, Toullec and baces, Documents CEPE - CNRS 56:85–100.
´
others 1999, Toullec 1997, Gallet and Roze 2001, ´ De Raeve, F. 1989. Sand dune vegetation and management
2002). These indices are of value for managing the dynamics. Pages 99 –109 in F. Van der Meulen, P. D. Junge-
opening of sensitive areas to the public. The results rius and J. Visser (eds.), Perspectives in coastal dune man-
agement. SPB Academic Publishing, The Hague.
obtained for the mobile dune and especially for the
Doody, P. 1989. Conservation and development of the coastal
semifixed dune suggest that tourist pressure could eas-
dunes in Great Britain. Pages 53– 67 in F. Van der Meulen,
ily be included in site management. Nevertheless, the P. D. Jungerius and J. H. Visser (eds.), Perspectives in
concept of resilience only has meaning if the distur- coastal dune management. SPB Academic Publishing, The
bance is halted and management includes periods of Hague.
recovery. Dupont, P. 1993. La vegetation des dunes et de la lagune de
´ ´
234 S. Lemauviel and F. Roze
´
la Belle Henriette (Vendee). Interet, evolution, conse-
´ ´ ˆ ´ ´ Liddle, M. J. 1975. A theoretical relationship between the
quences de la fermeture de la lagune en 1971 et de la primary productivity of vegetation and its ability to tolerate
tempete de fevrier 1990. Bulletin de la Societe Botanique du
ˆ ´ ´´ trampling. Biological Conservation 8:251–255.
Centre Ouest 24:9 – 40. Liddle, M. J., and P. Greig-Smith. 1975. A survey of tracks and
Forgeard, F., and J. Touffet. 1980. La recolonisation des lan- paths in a sand dune ecosystem. I. Soils. Journal of Applied
des et des pelouses dans la region de Paimpont. Evolution
´ Ecology 12:893–908.
de la vegetation au cours des trois dernieres annees suivant
´ ´ ` ´ Maschinski, R., R. Frye, and S. Rutman. 1997. Demography
l’incendie. Bulletin d’Ecologie 11:349 –358. and population viability of an endangered plant species
Gallet, S., and F. Roze. 2001. Resistance of Atlantic heathlands
´ before and after protection from trampling. Conservation
to trampling in Brittany (France): influence of vegetation Biology 11:990 –999.
type, season and weather conditions. Biological Conservation Maun, M. A. 1993. Dry coastal ecosystems along the Great
97:189 –198. Lakes. Pages 299 –316 in E. Van der Maarel (ed.), Ecosys-
Gallet, S., and F. Roze. 2002. Long-term effects of trampling
´ tems of the world. 2B: Dry coastal ecosystems. Elsevier,
on Atlantic heathland in Brittany (France): resilience and Amsterdam.
tolerance in relation to season and meterological condi- McDonnell, M. J. 1981. Trampling effects on coastal Dune
tions. Biological Conservation 103:267–275. vegetation in the Parker river national wildlife refuge, Mas-
Gehu, J. M. 1994. Schema synsystematique et typologique des
´ ´ ´ sachusetts, USA. Biological Conservation 21:289 –301.
milieux littoraux Francais Atlantiques et Mediterraneens.
¸ ´ ´ Nova, F. G., and J. Merino. 1997. Pattern and process in the
Pages 183–212 in J. M. Gehu (ed.), Colloques phytoso-
´ dune system of the Donana national park, south-western
ciologiques XXII. La syntaxonomie et la synsystematique
´ Spain. Pages 453– 468 in E. Vander Maarel (ed.), Ecosys-
Europeennes comme base typologique des habitats. J. Cra-
´ tems of the world 2C: Dry coastal ecosystems. Elsevier,
mer, Vaduz. Amsterdam.
Goldsmith, F. B., R. J. C. Munton, and A. Warren. 1970. The Olsauskas, A. 1996. Influence of recreation on plant cover in
impact of recreation on the ecology and amenity of west Lithuania. Klaipedos Universiteto Leidykla, 210 pp.
semi-natural areas: methods of investigation used in the Paskoff, R. 1989. Les dunes du littoral. La Recherche 212:888 –
Isles of Scilly. Biological Journal of Linnean Society 2:287– 895.
306. Priest, S., E. Van der Maarel, and M. B. Usher. 1997. Recre-
Gomez-Limon, F. J. 1995. Recreational activities and loss of ational use of dry coastal ecosystems. Pages 519 –529 in E.
diversity in grasslands in Alta Manzanares Natural Park, Van der Maarel (ed.), Ecosystems of the world 2C: Dry
Spain. Biological Conservation 74:99 –105. coastal ecosystems. Elsevier, Amsterdam.
Grime, J. P. 1979. Plant strategies and vegetation processes. Romao, C. 1997. Natura 2000 —Manuel d’interpretation des
´
Wiley, Chichester, UK, 222 pp. habitats de l’Union Europeenne Version EUR 15, Commis-
´
Guilcher, A., and B. Hallegouet. 1991. Coastal dunes in Brit-
´ ¨ sion Europeenne, 103 pp.
´
tany and their management. Journal of Coastal Research Scherrer, B. 1984. Biostatistiques. Gaetan Morin, Boucher-
7:517–533. ville, 850 pp.
Hutchings, M. J., and H. de Kroon. 1994. Foraging in Slatter, R. J. 1978. Ecological effects of trampling on sand
plants: the role of morphological plasticity in resource dune vegetation. Journal of Biological Education 12:89 –96.
acquisition. Pages 159 –238 in M. Begon and A. H. Fitter Sokal, R. R., and F. L. Rohlf. 1969. Biometry. Freeman and
(eds.), Advances in ecological research. Academic Press, Company, New York, 859 pp.
London. ten Harkel, M. J., and F. Van der Meulen. 1996. Impact of
Hylgaard, T., and M. J. Liddle. 1981. The effect of human grazing and atmospheric nitrogen deposition on the vege-
trampling on a sand dune ecosystem dominated by Em- tation of dry coastal dune grasslands. Journal of Vegetation
petrum nigrum. Journal of Applied Ecology 18:559 –569. Science 7:445– 452.
Jungerius, P. D., and F. Van der Meulen. 1997. Aeolian dy- Thomas, A. S. 1960. Changes in vegetation since the advent of
namics in relation to vegetation in a blowout complex in myxomatosis. Journal of Ecology 48:287–307.
the Meijendel dunes, The Netherlands. Journal of Coastal Toullec, H. 1997. Dynamique des systemes perturbes: degra-
` ´ ´
Conservation 3:63–70. dation, restauration et gestion des landes pietinees. Ex-
´ ´
Ku
¨hnholtz-Lordat, G. 1923. Les dunes du Golfe du Lion. emple du site megalithique de Carnac-Bretagne. These
´ `
These d’etat. Universite de Montpellier, Montpellier,
` ´ ´ d’universite. Rennes 1, Rennes, France, 136 pp.
´
France, 307 pp. Toullec, H., S. Diquelou, F. Roze, and J. C. Gloaguen. 1999.
´ ´
Lahondere, C., and F. Bioret. 1997. Quelques aspects de la
` Reponse d’une lande atlantique au pietinement experimen-
´ ´ ´
vegetation littorale du Morbihan continental. Bulletin de la
´ ´ tal. Compte Rendu d’Academie des Sciences de la vie 322:809 – 815.
´
societe botanique du centre ouest 28:305–314.
´´ Van der Maarel, E. 1971. Plant species diversity in relation to
Lemauviel, S. 2000. Les dunes grises des cotes Atlantiques:
ˆ management. Pages 45– 63 in E. Duffey and A. S. Watt
fonctionnement, dynamique potentielle, principes de ges- (eds.), The scientific management of animal and plant
tion conservatoire et processus de restauration. These ` communities for conservation. Blackwell Scientific Publica-
d’universite: Universite de Rennes 1, Rennes, France, 264
´ ´ tions, Oxford.
pp. Van der Meulen, F. 1997. Dune water catchment in the Neth-
Response of Sand Dune Communities to Trampling 235
erlands. Pages 533–556 in E. Van der Maarel (ed.), Ecosys- Meulen, P. D. Jungerius and J. Visser (eds.), Perspectives in
tems of the world 2C: Dry coastal ecosystems. Elsevier, coastal dune management. SPB Academic Publishing, The
Amsterdam. Hague.
Vanden Berghen, C. 1964. La vegetation terrestre du littoral
´ ´ Yorks, T. P., N. E. West, M. R. J., and S. D. Waren. 1997.
de l’Europe occidentale. Les Naturalistes Belges 45:251–280. Toleration of traffic by vegetation. Life form conclusions
Wanders, E. 1989. Perspectives in coastal dune management and summary extracts from a comprehensive data base.
towards a dynamic approach. Pages 141–148 in F. Van der Environmental Management 21:121–131.