Content uploaded by Sonja Wipf
Author content
All content in this area was uploaded by Sonja Wipf on Oct 30, 2017
Content may be subject to copyright.
Journal of Applied
Ecology
2005
42
, 306–316
© 2005 British
Ecological Society
Blackwell Publishing, Ltd.
Effects of ski piste preparation on alpine vegetation
SONJA WIPF*†, CHRISTIAN RIXEN*†, MARKUS FISCHER‡,
BERNHARD SCHMID† and VERONIKA STOECKLI*
*
WSL Swiss Federal Institute for Snow and Avalanche Research SLF, Flüelastr. 11, 7260 Davos Dorf, Switzerland;
†
Institute of Environmental Sciences, University of Zurich, Winterthurerstr. 190, 8057 Zurich, Switzerland; and
‡
Institute of Biochemistry and Biology, University of Potsdam, Villa Liegnitz, Lennéstr. 7a, 14471 Potsdam, Germany
Summary
1.
Ski resorts increasingly affect alpine ecosystems through enlargement of ski pistes,
machine-grading of ski piste areas and increasing use of artificial snow.
2.
In 12 Swiss alpine ski resorts, we investigated the effects of ski piste management on
vegetation structure and composition using a pairwise design of 38 plots on ski pistes
and 38 adjacent plots off-piste.
3.
Plots on ski pistes had lower species richness and productivity, and lower abundance
and cover of woody plants and early flowering species, than reference plots. Plots on
machine-graded pistes had higher indicator values for nutrients and light, and lower
vegetation cover, productivity, species diversity and abundance of early flowering and
woody plants. Time since machine-grading did not mitigate the impacts of machine-
grading, even for those plots where revegetation had been attempted by sowing.
4.
The longer artificial snow had been used on ski pistes (2–15 years), the higher the
moisture and nutrient indicator values. Longer use also affected species composition by
increasing the abundance of woody plants, snowbed species and late-flowering species,
and decreasing wind-edge species.
5.
Synthesis and applications.
All types of ski piste management cause deviations from
the natural structure and composition of alpine vegetation, and lead to lower plant species
diversity. Machine-grading causes particularly severe and lasting impacts on alpine
vegetation, which are mitigated neither by time nor by revegetation measures. The
impacts of artificial snow increase with the period of time since it was first applied to ski
piste vegetation. Extensive machine-grading and snow production should be avoided,
especially in areas where nutrient and water input are a concern. Ski pistes should not
be established in areas where the alpine vegetation has a high conservation value.
Key-words
:artificial snow, biodiversity, functional groups, machine-grading, snow
ecology, Switzerland
Journal of Applied Ecology
(2005)
42
, 306–316
doi: 10.1111/j.1365-2664.2005.01011.x
Introduction
The structure of alpine vegetation is affected by altitude,
aspect and inclination. Some alpine habitats are extraor-
dinarily rich in species, while communities in the highest
and most extreme regions consist of only a few specialists
(Ellenberg 1988). Furthermore, most communities are
characterized by specific proportions of functional
groups, such as dwarf shrubs, grasses and herbs. The
communities can consist of species from special eco-
logical groups, for example snowbed species, indicating
depressions with a long-lasting snow cover and a very
short vegetation period, or wind-edge species, indi-
cating a long vegetation period and extreme winter
temperatures (Körner 1999).
Alpine ecosystems are sensitive and susceptible to
changes in land use and climate (Chapin & Körner 1995;
Fischer & Wipf 2002; Laiolo
et al
. 2004). Perturbations
of the alpine habitats and changes in snow cover char-
acteristics, for example through skiing, are likely to have
impacts on species composition, diversity and produc-
tivity of alpine vegetation, which in turn may have negative
effects on ecosystem functioning and stability (Tilman
1996; Mulder, Uliassi & Doak 2001). As ski tourism is
Correspondence: Sonja Wipf, Swiss Federal Institute for
Snow and Avalanche Research SLF, Flüelastr. 11, 7260 Davos
Dorf, Switzerland (fax + 41 81 417 01 10; e-mail wipf@slf.ch).
307
Ski piste
preparation and
alpine vegetation
© 2005 British
Ecological Society,
Journal of Applied
Ecology
,
42
,
306–316
one of the most important economical factors in
European alpine regions (Abegg
et al
. 1997; Elsasser &
Messerli 2001), the area affected by ski pistes or by con-
structions related to tourism is still increasing.
Skiing and ski piste preparation by snow-grooming
vehicles are likely to cause mechanical damage to the
vegetation and the soil (Cernusca
et al
. 1990; Rixen
et al.
2004). Further more, the winter preparation of ski
pistes leads to a thin and compressed snow cover with
decreased temperature insulation capacity. Soils and
vegetation under these ski pistes may experience
temperatures lower than
−
10
°
C (Rixen, Haeberli &
Stoeckli 2004), while under undisturbed snow-packs,
temperatures rarely fall below 0
°
C (Rixen 2002; Rixen,
Stoeckli & Ammann 2003). As a result, plant species
with insufficient cold hardiness and plants sensitive to
mechanical stress may be damaged, resulting in shifts
among functional groups and a higher proportion of
unvegetated ground. In summer, piste construction
measures, such as machine-grading, are carried out to
remove obstacles like trees and rocks or to level rough
or bumpy soil surfaces. In the process of machine-
grading, the upper soil layers and the vegetation are
removed or heavily damaged (Pröbstl 1990; Haeberli
1992; Bayfield 1996; Titus & Tsuyuzaki 1999; Ruth-
Balaganskaya & Myllynen-Malinen 2000). Seed mix-
tures are often applied to accelerate revegetation, but
the success of revegetation measures declines with
height above sea level (a.s.l.) (Urbanska 1997).
In response to ongoing climatic changes, artificial
snow production is increasingly used in most ski resorts
of the world. Climate-change scenarios predict changes
in seasonal snowfall patterns, with the snow season
beginning later and ending earlier, and a rise in the snow
line (IPCC 1996, 1998, 2001). As a consequence, the
minimum altitude termed snow-secure for winter sports
in the European Alps would rise from 1200 m today to
1500 m a.s.l. within the next 30 years (Abegg 1996).
Winters with little snow have already been a major
concern of winter sport resorts since the mid-1980s
(Seilbahnen Schweiz 2001). As a strategy to minimize
the dependence on natural snowfall, ski areas invest in
systems for artificial snowing.
Effects of ski piste manipulation in single resorts and
single effects of artificial snowing or machine-grading
have been investigated in earlier studies (Urbanska 1997;
Kammer 2002). In our comprehensive study, sites covered
a wide range of locations, altitudes and aspects in 12
Swiss ski resorts. We investigated plots on ski pistes
and adjacent control plots beside pistes. With a total of
76 vegetation records arranged in a factorial design,
we addressed the following specific questions. (i) Do
productivity, species richness and composition of the
vegetation on ski pistes differ from the vegetation off
pistes? (ii) What are the impacts of machine-grading on
productivity, species richness and composition of ski
piste vegetation? (iii) How does the use of artificial
snow modify environmental conditions on ski pistes and
how do productivity, species richness and composition
of the vegetation respond? (iv) Do the effects of artifi-
cial snow differ between graded and ungraded ski
pistes?
Materials and methods
We studied the vegetation of plots on and next to ski
pistes in 12 ski resorts in the Swiss Alps (see Appendix
1), ranging in altitude between 1750 and 2550 m a.s.l.
and representing a wide range of vegetation types (mostly
alpine grasslands on acidic or calcareous bedrock and
dwarf shrub heath), time since grading and years of
application of artificial snow. The study sites were chosen
on machine-graded or ungraded ski pistes with either
natural or artificial snow, resulting in four different ski
piste types: natural snow/ungraded (ns), artificial snow/
ungraded (as), natural snow/graded (nsg) and artificial
snow/graded (asg). Sites were selected in consultation
with the local ski piste managers to ensure that the
pistes had received the same management continuously
in previous years.
At each study site, we randomly chose a plot on a ski
piste and an undisturbed control plot in the nearest off-
piste area to allow for pairwise comparisons. The dis-
tance between piste and control plot was 15–50 m and
the pair of plots did not differ significantly in altitude,
inclination, aspect and management. All ski pistes were
prepared by snow-grooming vehicles and were used
annually by skiers from approximately mid-November/
early December to Easter (mid-April). Artificial snow
had been produced in all ski resorts for 2–15 years (
n
=
21 pistes, mean 7·1 years). On 17 pistes in nine resorts,
machine-grading (i.e. construction of ski pistes during
which vegetation and topsoil are removed) had been
carried out 4–30 years ago (
n
= 17 pistes, mean 18·7
years). On 10 machine-graded plots, revegetation had
been encouraged by sowing seed, but details of seed
mixtures and sowing techniques were not available.
From 1 July to 24 August 2000, we visited each site to
collect vegetation records. In 4
×
4-m plots, we noted the
presence of each vascular plant species and estimated
its cover in seven classes according to Braun-Blanquet
(Müller-Dombois & Ellenberg 1974). We assigned a
cover value to each class for further calculations (0·01%
to ‘
r
’, 0·1% to ‘+’, midpoint to other classes). To compare
site conditions between plots, we calculated weighted
indicator values for moisture (F), soil reaction (R),
nutrients (N) and light (L) per plot according to Landolt
(1977). For each plot, we also estimated the percentage
area not covered by vascular plants.
To assess productivity, we randomly selected three
patches of 0·2
×
0·2 m plot
−
1
, harvested the biomass 3 cm
above ground and pooled the samples. In cases where
plots had been grazed, we randomly sampled ungrazed
308
S. Wipf
et al.
© 2005 British
Ecological Society,
Journal of Applied
Ecology
,
42
,
306–316
areas with similar vegetation within 5 m distance from
the plot. Three grazed plots with no undisturbed veget-
ation nearby were omitted from biomass sampling.
We measured annual productivity as the total biomass
excluding litter, dead biomass and woody shoots older
than 1 year. The biomass was dried at 70
°
C and weighed
to the nearest 0·01 g.
For each plot we calculated species number and
Shannon’s index of diversity as
H
= –
Σ
P
i
ln
P
i
, where
P
i
is the abundance class of species
i
. We calculated
the abundance of four functional groups: graminoids
(Poaceae, Cyperaceae, Juncaceae), woody plants, leg-
umes and remaining forbs (including Pteridophyta), to
obtain taxonomical and plant functional characteris-
tics of the plots. According to their habitat preference
in relation to snow cover and length of growing season
(Ellenberg 1988; Wilmanns 1998; Delarze, Gonseth &
Galland 1999), we classified 26 of 322 species as inhab-
iting snowbeds and 30 as inhabiting wind-edges (Wipf
et al
. 2002). Depending on their phenology, we also
classified 67 species as early flowering and 94 as late-
flowering, according to the flowering period indicated
in Lauber & Wagner (2001). For the four functional,
two ecological and two phenological groups, we calcu-
lated the proportions relative to total species number
and total vegetation cover. See Appendix 2 for a list of
species and their classification.
We analysed our data with hierarchical analysis of vari-
ance (
) models. Differences in the type of bed-
rock, aspect, slope and hours of sunshine between pairs
were taken into account by the pairwise design of piste
and off-piste plots. Moreover, we considered altitude as
a covariate in all analyses. To mirror the study design
with pairs of plots on and next to pistes, we fitted the type
of plot pairs, i.e. the type of piste treatment of the piste
plot in the pair. Next we fitted the factor piste, which
indicated differences between plots on and next to pistes
within the pairs, across all types of piste treatment. Then,
we fitted the interactions of type of pair (machine-
grading or years of artificial snow) and piste, which
indicated whether the difference between plots on and
next to the piste depended on the type of piste treatment.
We considered machine-grading, which is applied only
once, as a binary factor, and artificial snowing as a con-
tinuous variable (time since start of application of arti-
ficial snow) because of its potentially cumulative impacts.
Three-way interactions between resort, type of plot
pair and piste indicated whether effects of machine-
grading and artificial snow differed between resorts. To
analyse the effect of succession and revegetation meas-
ures on the vegetation of machine-graded ski pistes, we
performed the same analysis on a reduced data set
(pairs of graded piste and control plots) and included
years since machine-grading as a continuous variable
and sown as a binary factor. If necessary, dependent
variables were transformed prior to analysis to reach
normality. Residuals were checked visually for normal
distribution. Analyses were carried out with SPSS 10·0·5
(SPSS 1999).
Results
.
The indicator values of the vegetation composition
showed an increased supply in moisture and nutrients
available for plants (indicator value F,
P
= 0·098; N,
P
= 0·003; Tables 1 and 2 and Fig. 1) and a tendency for
higher base content of the soil (R,
P
= 0·059; Tables 1
and 2) on ski pistes. Despite the favourable environ-
mental conditions for growth, plant productivity on ski
pistes was lower than beside the pistes (Tables 1 and 3
and Fig. 2). The mean number of species per plot was
11% lower on piste plots than on off-piste plots (
P
= 0·047;
Tables 1 and 3 and Fig. 2). The pistes thus had a neg-
ative effect on the species richness in alpine grassland
and dwarf shrub vegetation.
Woody plants covered 24·3% of unmanipulated con-
trol plots, but were significantly less frequent on ski
pistes (10·5% cover,
P
= 0·003; Tables 1 and 4 and Fig. 3),
mainly because of their low abundance on machine-
graded pistes (1·7% cover, see next section). The same
pattern was found in the proportion of woody species
(Tables 1 and 5 and Fig. 3). Legume species were more
abundant and had higher cover on piste plots than on
control plots. Both cover and species number of early
flowering species were lower on piste than on control
plots (
P
= 0·005; Tables 1 and 6 and Fig. 4). Thus, the
snow compaction changed the plot composition of
functional and ecological groups.
-
The proportion of ground not covered by vegetation
was almost five times higher on graded than on ungraded
piste plots (
P
= 0·008; Tables 1 and 3 and Fig. 2) but
was unaffected by revegetation measures (sowing) or
time since machine-grading. Corresponding with the
high proportion of ground not covered by vegetation,
the light input for the plants (expressed as indicator
value L) was higher on graded piste plots than on
ungraded ones (
P
= 0·002; Tables 1 and 2 and Fig. 1).
Nutrient availability to the vegetation (expressed as
indicator value N), which was generally increased on
ski pistes, reached even higher values on graded pistes
than on ungraded ones (
P
= 0·07; Tables 1 and 2 and
Fig. 1). Whereas the productivity on ungraded piste plots
was similar to that of off-piste plots, it was reduced 4·6
times on graded piste plots (
P
= 0·011; Tables 1 and 3
and Fig. 2). Machine-graded pistes that had been re-
vegetated by sowing supported fewer species (
P
= 0·098).
The Shannon index was lower for all graded piste plots
(
P
= 0·03; Tables 1 and 3 and Fig. 2).
Machine-grading also affected the vegetation com-
position: woody plants, which were generally less common
309
Ski piste
preparation and
alpine vegetation
© 2005 British
Ecological Society,
Journal of Applied
Ecology
,
42
,
306–316
Table 1. The effects of ski piste treatments on the parameters describing the vegetation in 12 Swiss ski resorts: Predicted means
and slopes of the correlation with years of artificial snow. The n in parentheses refer to the biomass measurements (n lower because
of missing values)
Total
n = 76 (73)
Mean
Off-piste
n = 38 (36)
Mean
Piste
n = 38 (37)
Mean
Piste graded
n = 17 (16)
Mean
Piste ungraded
n = 21
Mean
Correlation with
years of artificial snow
Slope
Indicator values
F* 2·85 2·83 2·87 2·88 2·87 0·000
R† 2·55 2·50 2·59 2·61 2·58 0·007
N‡ 2·46 2·38 2·54 2·56 2·53 0·003
L§ 3·80 3·76 3·85 4·03 3·71 0·01
Diversity and productivity
Species number 36·4 38·5 34·3 31·7 36·5 0·30
Shannon index 3·6 3·8 3·4 2·8 3·8 −0·02
Annual productivity (g m−2) 158·0 173·1 143·3 46·9 216·7 −0·25
Proportion of cover (%)
Grasses 41·4 38·8 44·0 38·7 48·4 −0·93
Forbs 24·2 26·5 22·0 20·5 23·1 0·13
Woody plants 17·4 24·3 10·5 1·6 17·7 0·38
Legumes 5·1 2·9 7·4 10·6 4·8 0·15
Snowbed species 8·4 7·3 9·6 14·0 6·0 1·36
Wind-edge species 7·5 7·7 7·4 8·0 6·9 −0·27
Early flowering species 35·6 40.8 30·5 22·9 36·6 −1·26
Late-flowering species 30·5 28·0 32·9 34·1 32·0 0·03
Bare ground 11·8 7·0 16·1 28·6 6·0 0·26
Proportion of species (%)
Grasses 27·9 27·9 27·8 27·2 28·2 −0·13
Forbs 60·2 59·8 60·6 63·2 58. 5 0·09
Woody plants 11·9 13·0 10·8 6·9 14·0 −0·23
Legumes 7·8 6·6 9·1 10·9 7·6 0·21
Early flowering species 33·4 35·7 31·1 25·5 35·7 −0·71
Late-flowering species 31·0 29·9 32·2 32·5 31·9 0·25
*F = 2, medium dryness; 3, medium humidity.
†R = 2, acid soils; 3, neutral or weakly alkaline soils.
‡N = 2, nutrient-poor soils; 3, medium- to nutrient-poor soils.
§L = 3, half-shaded conditions; 4, full-light conditions.
Table 2. table for the indicator values of the vegetation for moisture (F), soil acidity (R), nutrient (N) and light availability
(L) measured on 38 ski piste plots and 38 corresponding off-piste plots in 12 Swiss ski resorts. The type of piste treatment
(machine-grading and years of artificial snowing) was fitted for pairs of piste/off-piste plots. The factor ski piste indicates the
difference between piste and off-piste plots. The dependent variables were (ln(x + 1))-transformed prior to analysis. Three plots
were excluded from biomass sampling. The sample size for productivity is therefore 73, and the degrees of freedom are indicated
in parentheses. MS, mean squares; F, variance ratio; subscripts refer to source of variation (*)P < 0·1, *P < 0·05, **P < 0·01, ***P
< 0·001
Source of variation
Skeleton analysis
MS F d.f.
Variance ratio
F-value R-value N-value L-value
Altitude MSEMSE/MSRest 1 3·14 5·87(*) 70·23** 37·19**
Resort MSRMSR/MSRest 11 3·34 4·87(*) 4·43(*) 1·83
Type of plot pair MSTMST/MSRT 3 0·16 0·50 0·25 0·75
Piste MSPMSP/MSRP 1 3·27(*) 4·42(*) 14·64** 9·46*
Grading × piste MSLP MSLP/MSRLP 1 0·32 2·68 4·51(*) 24·36**
Duration of snowing × piste MSAP MSAP /MSRAP 1 3·68(*) 4·80(*) 3·26(*) 0·13
Grading × duration of snowing × piste MSLAP MSLAP /MSRest 1 1·02 0·03 0·03 0·06
Resort × type of plot pair MSRT MSRT/MSRest 23 1·42 1·52 3·97(*) 0·87
Resort × piste MSRP MSRP/MSRest 11 0·51 0·45 0·93 0·24
Resort × grading × piste MSRLP MSRLP/MSRest 7 0·37 0·44 1·31 0·18
Resort × duration of snowing × piste MSRAP MSRAP/MSRest 11 0·34 0·51 0·44 0·15
Rest MSRest 4
Total 76
310
S. Wipf
et al.
© 2005 British
Ecological Society,
Journal of Applied
Ecology
,
42
,
306–316
on piste plots, were reduced by 91% on graded piste plots
in comparison with ungraded ones (P = 0·03; Tables 1
and 4 and Fig. 3). Whereas only 6·9% of the species on
graded piste plots were woody species, they accounted
Fig. 2. Differences in productivity and diversity between
ski piste plots and corresponding control plots. (a) Mean
differences in productivity, species number, Shannon index
and the proportion of open ground for the four ski piste
treatments (ns, ungraded pistes with natural snow; as,
ungraded pistes with artificial snow; nsg, graded pistes with
natural snow; asg, graded pistes with artificial snow). (b) The
relationship between the differences in annual productivity
and the time since conversion to artificial snow.
Fig. 3. Differences in species composition between ski piste
plots and corresponding control plots. Mean differences in
the percentage cover of the four functional groups for the four
ski piste treatments (ns, ungraded pistes with natural snow; as,
ungraded pistes with artificial snow; nsg, graded pistes with
natural snow; asg, graded pistes with artificial snow).
Fig. 1. Differences in indicator values between ski piste plots
and corresponding control plots. (a) Mean differences in indicator
values for the four ski piste treatments (ns, ungraded pistes
with natural snow; as, ungraded pistes with artificial snow;
nsg, graded pistes with natural snow; asg, graded pistes with
artificial snow; F, moisture indicator value; R, reaction (soil
acidity) indicator value; N, nutrient indicator value; L, light
indicator value). (b–d) The relationship between the differences
in (b) F (moisture), (c) R (soil acidity) and (d) N (nutrient)
indicator values and the time since conversion to artificial snow.
311
Ski piste
preparation and
alpine vegetation
© 2005 British
Ecological Society,
Journal of Applied
Ecology, 42,
306–316
for 14% of the species on ungraded piste plots (P = 0·02;
Tables 1 and 5). The reactions of forb and grass species
were not uniform: in six resorts the proportion of forb
species was higher on graded compared with ungraded
pistes (resort × grading × piste, P = 0·02). The propor-
tion of grass species depended on the time since machine-
grading: compared with ungraded control plots, the
proportion was higher in recently graded pistes, and lower
the older the graded pistes were (P = 0·016). Species
that flower early in the year were reduced in abundance
by 28% (P = 0·018; Tables 1 and 6) and covered 37%
less ground on graded piste plots than on ungraded
ones (P = 0·004; Tables 1 and 6).
The longer artificial snow had been applied on ski pistes,
the higher the moisture and nutrient availability, and
the lower the soil base content (indicator values F, N
and R: P = 0·081, 0·099 and 0·051, respectively; Tables 1
and 2 and Fig. 1). However, we could not detect any
effects of artificial snow on plant productivity (Fig. 2).
Table 3. table for the proportion of bare ground, annual productivity, species number and Shannon index measured on 38
ski piste plots and 38 corresponding off-piste plots in 12 Swiss ski resorts. The type of piste treatment (machine-grading and years
of artificial snowing) was fitted for pairs of piste/off-piste plots. The factor ski piste indicates the difference between piste and off-
piste plots. Three plots were excluded from biomass sampling. The sample size for productivity therefore is 73, and the degrees
of freedom are indicated in parentheses. (*)P < 0·1, *P < 0·05, **P < 0·01, ***P < 0·001
Source of variation
Skeleton analysis
MS F d.f.
Variance ratio
Bare
ground (%) Productivity Species
Shannon
number
Altitude MSEMSE/MSRest 1 254·41*** 79·96** 5·59(*) 0·58
Resort MSRMSR/MSRest 11 9·53* 2·13 5·06(*) 8·00*
Type of plot pair MSTMST/MSRT 3 3·18 4·32(*) 1·33 1·62
Piste MSPMSP/MSRP 1 8·81* 9·78** 5·02* 4·21(*)
Grading × piste MSLP MSLP/MSRLP 1 13·64** 10·71* 0·35 7·27*
Duration of snowing × piste MSAP MSAP /MSRAP 1 0·02 0·01 0·23 0·26
Grading × duration of snowing × piste MSLAP MSLAP /MSRest 1 2·02 1·83 9·81* 9·99*
Resort × type of plot pair MSRT MSRT /MSRest 23 7·74* 1·01 1·90 2·21
Resort × piste MSRP MSRP/MSRest 11 8·18* 0·90 1·09 2·68
Resort × grading × piste MSRLP MSRLP/MSRest 7 6·05(*) 0·93 1·16 1·85
Resort × duration of snowing × piste MSRAP MSRAP/MSRest 11 (9) 2·74 1·60 0·85 2·27
Rest MSRest 4 (3)
Total 76 (73)
Table 4. table for the proportion of the four functional groups, grasses, woody plants, forbs and legumes, at the plot area
measured on 38 ski piste plots and 38 corresponding off-piste plots in 12 Swiss ski resorts. The type of piste treatment (machine-
grading and years of artificial snowing) was fitted for pairs of piste/off-piste plots. The factor ski piste indicates the difference
between piste and off-piste plots. Three plots were excluded from biomass sampling. The sample size for productivity therefore
is 73, and the degrees of freedom are indicated in parentheses. (*)P < 0·1, *P < 0·05, **P < 0·01, ***P < 0·001
Source of variation
Skeleton analysis
MS F d.f.
Variance ratio
Cover
grass (%)
Cover woody
plants (%)
Cover
forbs (%)
Cover
legumes (%)
Altitude MSEMSE/MSRest 1 40·49** 0·30 3·28 0·44
Resort MSRMSR/MSRest 11 6·37* 3·76 6·70* 9·95*
Type of plot pair MSTMST/MSRT 3 1·02 1·58 0·61 2·38
Piste MSPMSP/MSRP 1 1·76 13·87** 2·47 5·39*
Grading × piste MSLP MSLP/MSRLP 1 0·72 7·67* 0·00 1·54
Duration of snowing × piste MSAP MSAP /MSRAP 1 29·86*** 9·20* 0·69 0·20
Grading × duration of snowing × piste MSLAP MSLAP /MSRest 1 0·00 0·12 0·99 0·81
Resort × type of plot pair MSRT MSRT/MSRest 23 5·21(*) 1·01 2·69 4·66(*)
Resort × piste MSRP MSRP/MSRest 11 3·23 0·70 1·77 5·30(*)
Resort × grading × piste MSRLP MSRLP/MSRest 7 1·53 1·19 1·02 2·83
Resort × duration of snowing × piste MSRAP MSRAP/MSRest 11 0·26 0·40 0·54 4·14(*)
Rest MSRest 4
Total 76
312
S. Wipf et al.
© 2005 British
Ecological Society,
Journal of Applied
Ecology, 42,
306–316
The impact of artificial snow on diversity was ambig-
uous. While the use of artificial snow had negative
effects on species number and Shannon index on graded
pistes, the opposite was found for ungraded pistes (three-
way interaction between piste, grading and duration of
snowing, P = 0·034; Tables 1 and 3).
The application of artificial snow affected the rela-
tive proportion of plant functional groups. The longer
artificial snow had been used, the more it had reduced
the negative effect of the ski piste treatment on the
cover of woody plants per plot (P = 0·011; Tables 1 and
4). However, the proportion of a plot covered by grass
species declined the longer a ski piste had been treated
with artificial snow (P < 0·001; Tables 1 and 4). The
effect of artificial snow on the species proportion of
forbs and the cover of legumes was not uniform but dif-
fered between ski resorts (three-way interaction resort
× artificial snow × ski piste, P = 0·014 and P < 0·083,
respectively; Tables 5 and 6). The longer artificial snow
had been applied to a ski piste, the higher the cover of
Table 5. table for the proportion of the four functional groups, grasses, woody plants, forbs and legumes, of the total
species number per plot measured on 38 ski piste plots and 38 corresponding off-piste plots in 12 Swiss ski resorts. The type of
piste treatment (machine-grading and years of artificial snowing) was fitted for pairs of piste/off-piste plots. The factor ski piste
indicates the difference between piste and off-piste plots. The dependent variables of were (ln(x + 1))-transformed prior to
analysis. Three plots were excluded from biomass sampling. The sample size for productivity therefore is 73, and the degrees of
freedom are indicated in parentheses. (*)P < 0·1, *P < 0·05, **P < 0·01, ***P < 0·001
Source of variation
Skeleton analysis
MS MS d.f.
Variance ratio
Grass
species (%)
Woody
species (%)
Forb
species (%)
Legume
species (%)
Altitude MSEMSE/MSRest 1 2·72 0·08 22·07** 8·59*
Resort MSRMSR/MSRest 11 11·33* 2·01 18·79** 2·44
Type of plot pair MSTMST/MSRT 3 0·29 0·45 0·26 1·99
Piste MSPMSP/MSRP 1 0·30 5·72 * 0·24 7·31*
Grading × piste MSLP MSLP/MSRLP 1 1·28 10·18 * 0·31 3·19
Duration of snowing × piste MSAP MSAP /MSRAP 1 0·02 0·01 0·04 0·08
Grading × duration of snowing × piste MSLAP MSLAP /MSRest 1 0·04 0·09 3·22 0·01
Resort × type of plot pair MSRT MSRT/MSRest 23 3·03 2·36 11·63* 1·11
Resort × piste MSRP MSRP/MSRest 11 6·93* 0·90 6·21 0·36
Resort × grading × piste MSRLP MSRLP/MSRest 7 4·19(*) 0·58 17·73** 1·13
Resort × duration of snowing × piste MSRAP MSRAP/MSRest 11 3·44 0·79 11·84* 0·53
Rest MSRest MSE/MSRest 4
Total 76
Table 6. table for the proportion of snowbed and wind-edge, as well as early flowering and late-flowering species, of the
total vegetation cover in a plot (%) measured on 38 ski piste plots and 38 corresponding off-piste plots in 12 Swiss ski resorts. The
type of piste treatment (machine-grading and years of artificial snowing) was fitted for pairs of piste/off-piste plots. The factor ski
piste indicates the difference between piste and off-piste plots. Three plots were excluded from biomass sampling. The sample size
for productivity therefore is 73, and the degrees of freedom are indicated in parentheses. (*) P < 0·1, *P < 0·05, **P < 0·01, ***P
< 0·001
Source of variation
Skeleton analysis
MS MS d.f.
Variance ratio
Snowbed
species (%)
Wind-edge
species (%)
Early
flowering (%)
Late-
flowering (%)
Altitude MSEMSE/MSRest 1 85·42*** 34·45** 51·38** 5·29(*)
Resort MSRMSR/MSRest 11 3·45 3·79 2·45 1·93
Type of plot pair MSTMST/MSRT 3 2·18 1·04 1·24 0·61
Piste MSPMSP/MSRP 1 3·14 0·07 13·90** 3·92(*)
Grading × piste MSLP MSLP/MSRLP 1 0·29 0·01 17·93** 2·12
Duration of snowing × piste MSAP MSAP /MSRAP 1 13·19** 6·68* 5·68* 0·09
Grading × duration of snowing × piste MSLAP MSLAP /MSRest 1 0·43 0·84 7·35(*) 1·21
Resort × type of plot pair MSRT MSRT/MSRest 23 4·09(*) 0·83 2·87 2·13
Resort × piste MSRP MSRP/MSRest 11 0·79 0·22 1·68 0·74
Resort × grading × piste MSRLP MSRLP/MSRest 7 2·77 2·28 0·86 0·72
Resort × duration of snowing × piste MSRAP MSRAP/MSRest 11 1·56 0·55 1·20 1·74
Rest MSRest MSE/MSRest 4
Total 76
313
Ski piste
preparation and
alpine vegetation
© 2005 British
Ecological Society,
Journal of Applied
Ecology, 42,
306–316
snowbed species (P = 0·004; Tables 1 and 6 and Fig. 4).
Wind-edge species showed the opposite reaction. They
covered less ground the longer a ski piste had been
treated with artificial snow (P = 0·024; Tables 1 and 6
and Fig. 4). The longer artificial snow had been applied,
the lower the cover of early flowering plants on piste
plots (P = 0·036; Tables 1 and 6 and Fig. 4). On the
other hand, the relative number of late-flowering spe-
cies tended to increase the longer artificial snow had
been used (P < 0·096; Tables 1 and 6 and Fig. 4). Over-
all, the potentially snow-sensitive groups were affected
most by the production of artificial snow.
Discussion
.
The changes in indicator values for soil nutrients, mois-
ture and reaction (base content) under ski pistes may
be considered beneficial for plant growth (in terms of
enhancing growth). However, other non-beneficial
factors, including disturbance, seem to prevail because
overall productivity and species richness on ski pistes
are decreased. The negative effects may be the direct
result of disturbance of the vegetation via snow
Fig. 4. Differences in ecological groups between ski piste plots and corresponding control plots. (a) Mean differences in the
proportion of plot covered by ecological groups for the four ski piste treatments (ns, ungraded pistes with natural snow; as,
ungraded pistes with artificial snow; nsg, graded pistes with natural snow; asg, graded pistes with artificial snow). (b–e) The
relationship between the differences in (b) percentage cover of snowbed species, (c) percentage cover of wind-edge species,
(d) proportion of early flowering species and (e) proportion of late-flowering species and the time since conversion to artificial snow.
314
S. Wipf et al.
© 2005 British
Ecological Society,
Journal of Applied
Ecology, 42,
306–316
compaction under skiing and via mechanical abrasion.
Although some types of disturbance can increase bio-
diversity by suppressing dominant species in productive
environments (Connell 1979; Vujnovic, Wein & Dale
2002), abiotic stress reduces productivity and domin-
ance in alpine habitats and therefore reverses any
positive effects of disturbance on plant diversity (Kammer
& Möhl 2002).
The changes in the cover of woody plants on the ski
pistes may also be explained by mechanical disturbance.
Woody plants (mainly dwarf shrubs of the Ericaceae)
are particularly sensitive to mechanical injuries because
of their size, longevity and vulnerability of their hiber-
nation buds (Körner 1999).
In contrast to woody plants, legumes were more
abundant on pistes than next to pistes, perhaps because
of decreased competition from shrubs. Legumes may
have been responsible for the increased nutrient indi-
cator values on piste plots as a consequence of their
ability to fix nitrogen. The reduced abundance of early
flowering species on pistes in our study was most prob-
ably the result of a narrowed temporal niche for those
species. In field surveys, the flowering time of early sea-
sonal species was postponed after soil frosts on pistes
with natural snow (Baiderin 1981; Köck et al. 1989). A
delay in flowering often decreases fecundity and, con-
sequently, can negatively affect the abundance of early
flowering species (Kudo 1993; Stanton & Galen 1997).
-
Vegetation responses were highly pronounced on graded
pistes, most probably because machine-grading of ski
pistes is a drastic measure where soil and plants are
removed in summer during active plant growth. There
was a strong compositional shift from woody plants to
grasses or forbs, the extent of which depended partly on
the resort and partly on the successional stage of the ski
piste. The increased indicator values for nutrients and
light may be the result of the disturbance itself (through
mobilized nutrients and opened vegetation cover), or
of characteristics of the following pioneer species,
which often grow with high light and nutrient levels
(Grime & Jarvis 1975; Grime 1977). Graded ski pistes
with high proportions of bare ground are particularly
prone to increased surface runoff and erosion during
heavy rain (Löhmannsröben & Cernusca 1987). Reveg-
etation processes at high altitudes are slow, even if
revegetation measures such as sowing and planting are
applied (Bayfield, Urquart & Rothery 1984; Urbanska
1995; Titus & Tsuyuzaki 1999; Fattorini 2001). In our
study, the proportion of vegetation cover on graded pistes
was negatively correlated with altitude, but positively
affected by neither sowing nor time since machine-
grading. This indicates how difficult it is to achieve
revegetation at high altitudes. Machine-grading there-
fore is a particularly damaging management activity,
the consequences of which are more severe and long
lasting at higher altitudes.
Application of artificial snow increased the moisture
indicator values of the vegetation on piste plots. Arti-
ficially snowed pistes contained twice as much snow
mass as normal pistes at our study sites in the winter of
1999–2000 (Stoeckli & Rixen 2000). Because water for
artificial snow making is taken from lakes, rivers or ground
water, it usually contains minerals and other chemical
compounds (Kammer & Hegg 1990; Kammer 2002),
and thus provides a nutrient input during spring melt-
ing for about 4 weeks in late spring. Accordingly, the
vegetation on piste plots with artificial snow had an
increased nutrient indicator value. Adding artificial snow
to the ski pistes appeared to mitigate the mechanical
disturbance of skiing because woody species became
more abundant the longer artificial snow had been
applied. Artificial snow is usually produced at the
beginning of the winter and lasts beyond the end of the
skiing season in spring.
The increase in snowbed and late-flowering species,
and the decrease in wind-edge and early flowering
species, on pistes with artificial snow probably reflect
the shortened growth period resulting from a delay in
snowmelt of 2–3 weeks (Newesely 1997; Stoeckli &
Rixen 2000). Similar to the effects on normal pistes with
their compacted natural snow, the narrowed temporal
niche in spring presumably conferred a competitive
advantage to the snowbed and late-flowering species
over wind-edge and early flowering species (Cernusca
et al. 1990; Newesely 1997; Rixen et al. 2001).
Plant diversity reacted differently to artificial snow
on the graded and ungraded pistes. The increase in
diversity on the ungraded pistes might be because of
the decreased mechanical disturbance. The decrease in
diversity on the graded pistes, on the other hand, might
be a consequence of a decelerated revegetation process
because of the shortened growing period. These effects
can be enhanced if ice layers in the artificial snow cover
decrease gas permeability (Newesely, Cernusca &
Bodner 1994).
Whether the impact of artificial snow on alpine veg-
etation is considered positive or negative depends on
the current state of the vegetation and the environmental
objectives of a specific ski resort. If mechanical distur-
bance through snow-grooming vehicles or ski edges is a
major problem, the increased protection afforded by
artificial snow can be considered beneficial. However,
in the case of endangered habitats poor in nutrients,
like oligotrophic fens or nutrient-poor grasslands, the
additional nutrients input by the melt water of artificial
snow is clearly a negative impact.
We have shown that ski pistes in general, and machine-
grading and artificial snow production in particular,
cause deviations from natural plant species composition
and decrease plant species richness. Machine-grading
constitutes the most drastic vegetation disturbance on
ski pistes. It should be avoided wherever possible, as it
causes lasting damage that cannot be overcome even by
315
Ski piste
preparation and
alpine vegetation
© 2005 British
Ecological Society,
Journal of Applied
Ecology, 42,
306–316
revegetation measures, particularly at higher altitudes.
Impacts of ski pistes in general, and of artificial snow-
ing in particular, appear comparatively moderate, but
are by no means negligible. With the ongoing inten-
sification of ski resorts, the use of artificial snow will
become more prevalent and the vegetation will change
over an increasing area. Moreover, impacts of artificial
snowing are cumulative and will become even more
pronounced in the long term. In summary, mountain
regions with a high proportion of areas with extensive
outdoor recreation activities, like the European Alps,
are facing continuous change of their traditional
unique environment and vegetation. Therefore, we
recommend that environmental goals in ski resort
management should be established and respected. In
particular, we recommend carefully recording the veg-
etation in a specific area before any intensification of
use as ski piste, and complete avoidance of areas with
vegetation of particularly high conservation value.
Moreover, we recommend that large-scale machine-
grading should be avoided, and that long-term snow
production should be banned in areas where any
increase in the supply of nutrients and water is a con-
cern. Overall, the establishment of ski pistes should not
be allowed in areas where any changes in vegetation
composition or any decrease in plant species richness
cannot be tolerated.
Acknowledgements
The study was supported by the Swiss Agency for
the Environment, Forests and Landscape (SAEFL,
Switzerland) and the cantons of Valais and Grisons. We
thank the ski resort managers for their collaboration,
Katharina Schudel for help in the field, and Dominik
Kulakowski and the anonymous referees for comments
that helped improving the manuscript.
Supplementary material
The following material is available from
http://www.blackwellpublishing.com/products/
journals/suppmat/JPE/JPE1011/JPE1011sm.htm.
Appendix 1. List of study sites
Appendix 2. List of species
References
Abegg, B. (1996) Klimaänderung und Tourismus: Klimafolgen-
forschung am Beispiel des Wintertourismus in den Schweizer
Alpen. Vdf Hochschulverlag AG, Zürich, Switzerland.
Abegg, B., Koenig, U., Buerki, R. & Elsasser, H. (1997) Climate
impact assessment im tourismus. Die Erde, 128, 105–116.
Baiderin, V.V. (1981) Winter recreation and subnivial plant
development (translated from Russian). Ecologiya, 5, 3–8.
Bayfield, N.G. (1996) Long-term changes in colonization of
bulldozed ski pistes at Cairn Gorm, Scotland. Journal of
Applied Ecology, 33, 1359–1365.
Bayfield, N.G., Urquart, U.H. & Rothery, P. (1984) Colon-
ization of bulldozed track verges in the Cairngorm Mountains,
Scotland. Journal of Applied Ecology, 21, 343–354.
Cernusca, A., Angerer, H., Newesely, C. & Tappeiner, U. (1990)
Auswirkungen von Kunstschnee: eine Kausalanalyse der
Belastungsfaktoren. Verhandlungen der Gesellschaft für
Ökologie, 19, 746–757.
Chapin, F.S. III & Körner, C. (1995) Arctic and Alpine Bio-
diversity: Patterns, Causes and Ecosystem Consequences.
Springer, Berlin, Germany.
Connell, J.H. (1979) Intermediate–disturbance hypothesis.
Science, 204, 1345–1345.
Delarze, R., Gonseth, Y. & Galland, P. (1999) Lebensräume
der Schweiz. Ökologie, Gefährdung, Kennarten, 1st edn. Ott
Verlag, Thun, Switzerland.
Ellenberg, H. (1988) Veg etation Ecology of Central Europe,
4th edn. Cambridge University Press, Cambridge, UK.
Elsasser, H. & Messerli, P. (2001) The vulnerability of the
snow industry in the Swiss Alps. Mountain Research and
Development, 21, 335–339.
Fattorini, M. (2001) Establishment of transplants on
machine-graded ski runs above timberline in the Swiss
Alps. Restoration Ecology, 9, 119–126.
Fischer, M. & Wipf, S. (2002) Effect of low-intensity grazing
on the species-rich vegetation of traditionally mown sub-
alpine meadows. Biological Conservation, 104, 1–11.
Grime, J.P. (1977) Evidence for the existence of three primary
strategies in plants and its relevance to ecological and
evolutionary theory. American Naturalist, 111, 1169–
1194.
Grime, J.P. & Jarvis, B.C. (1975) Shade avoidance and shade
tolerance in flowering plants II. Effects of light on the
germination of species of contrasted ecology. Light as an
Ecological Factor II (eds G.C. Evans, R. Bainbridge &
O. Rackham), pp. 525 – 532. Blackwell Scientific Publications,
Oxford, UK.
Haeberli, W. (1992) Construction, environmental problems
and natural hazards in periglacial mountain belts. Perma-
frost and Periglacial Processes, 3, 111–124.
IPCC (1996) Climate Change 1995: Impacts, Adaptations, and
Mitigation of Climate Change. Scientific-Technical Analysis.
Contribution of Working Group II to the Second Assessment
Report of the Intergovernmental Panel on Climate Change.
Intergovernmental Panel on Climate Change, Cambridge
University Press, Cambridge and New York.
IPCC (1998) The Regional Impacts of Global Change.
An Assessment of Vulnerability. Special Report of IPCC
Working Group II. Intergovernmental Panel on Climate
Change, Cambridge University Press, Cambridge and
New York.
IPCC (2001) Climate Change 2001: Impacts, Adaptations and
Vulnerability. Intergovernmental Panel on Climate Change,
Cambridge University Press, Cambridge and New York.
Kammer, P.M. (2002) Floristic changes in subalpine grass-
lands after 22 years of artificial snowing. Journal for Nature
Conservation, 10, 109–123.
Kammer, P. & Hegg, O. (1990) Auswirkungen von Kunst-
schnee auf subalpine Rasenvegetation. Verhandlungen der
Gesellschaft für Ökologie, 19, 758–767.
Kammer, P.M. & Möhl, A. (2002) Factors controlling species
richness in alpine plant communities: an assessment of the
importance of stress and disturbance. Arctic, Antarctic, and
Alpine Research, 34, 398–407.
Köck, L., Holaus, K., Schnitzer, R. & Walch, A. (1989) Quan-
titative und qualitative Veränderungen der Vegetation durch
den Einfluss des Skisportes auf Pisten und Langlaufloipen.
Landesanstalt für Pflanzenzucht und Samenprüfung in Rinn
1939–89, pp. 59–79. Eigenverlag, Innsbruck, Rinn, Austria.
Körner, C. (1999) Alpine Plant Life. Springer Verlag, Berlin,
Germany.
316
S. Wipf et al.
© 2005 British
Ecological Society,
Journal of Applied
Ecology, 42,
306–316
Kudo, G. (1993) Relationship between flowering time and
fruit-set of the entomophilous alpine shrub, Rhododendron
aureum (Ericaceae), inhabiting snow patches. American
Journal of Botany, 80, 1300–1304.
Laiolo, P., Dondero, F., Ciliento, E. & Rolando, A. (2004)
Consequences of pastoral abandonment for the structure
and diversity of the alpine avifauna. Journal of Applied Ecology,
41, 389–397.
Landolt, E. (1977) Ökologische Zeigerwerte zur Schweizer
Flora. Ver öffentlichungen des Geobotanischen Institutes der
Eidgenössischen Technischen Hochschule, Stiftung Rübel,
Zürich, 64, 1–208.
Lauber, K. & Wagner, G. (2001) Flora Helvetica, 3 edn. Paul
Haupt, Bern, Switzerland.
Löhmannsröben, R. & Cernusca, A. (1987) Ursachen für
Erosionsprozesse im Skigebiet Meran 2000. Verhandlungen
der Gesellschaft für Ökologie, 15, 189–196.
Mulder, C.P.H., Uliassi, D.D. & Doak, D.F. (2001) Physical
stress and diversity–productivity relationships: the role of
positive interactions. Proceedings of the National Academy
of Sciences (USA), 98, 6704 –6708.
Müller-Dombois, D. & Ellenberg, H. (1974) Aims and Methods
of Vegetation Ecology. Wiley, New York, NY.
Newesely, C. (1997) Auswirkungen der künstlichen Beschneiung
von Schipisten auf Aufbau, Struktur und Gasdurchlässigkeit
der Schneedecke, sowie auf den Verlauf der Bodentemperatur
und das Auftreten von Bodenfrost. PhD Thesis. Leopold
Franzens Universität, Innsbruck, Switzerland.
Newesely, C., Cernusca, A. & Bodner, M. (1994) Entste-
hung und Auswirkung von Sauerstoffmangel im Bereich
unterschiedlich präparierter Schipisten. Verhandlungen der
Gesellschaft für Ökologie, 23, 277–282.
Pröbstl, U. (1990) Skisport und Vegetation. Stöppel, Weilheim,
Germany.
Rixen, C. (2002) Artificial snow and snow additives on ski
pistes: interactions between snow cover, soil and vegetation.
PhD Thesis. University of Zurich, Zurich, Switzerland.
Rixen, C., Casteller, A., Schweingruber, F.H. & Stoeckli, V.
(2004) Age analysis helps to estimate plant performance on
ski pistes. Botanica Helvetica, 114, 127–138.
Rixen, C., Haeberli, W. & Stoeckli, V. (2004) Ground temper-
atures under ski pistes with artificial and natural snow.
Arctic, Antarctic, and Alpine Research, 36, 403– 411.
Rixen, C., Stoeckli, V. & Ammann, W. (2003) Does artificial
snow production affect soil and vegetation of ski pistes? A
review. Pe rspectives in Plant Ecology, Evolution and System-
atics, 5, 219–230.
Rixen, C., Stoeckli, V., Huovinen, C. & Huovinen, K. (2001)
The phenology of four subalpine herbs in relation to snow
cover characteristics. Proceedings of the Sixth IAHS Sym-
posium 2001. Soil–Vegetation–Atmosphere Transfer Schemes
and Large Scale Hydrological Models (eds A.J. Dolman,
A.J. Hall, M.L. Kavvas, T. Oki & J.W. Pomeroy), pp.
359–362. Publication No. 270. IAHS, Maastricht, the
Netherlands.
Ruth-Balaganskaya, E. & Myllynen-Malinen, K. (2000) Soil
nutrient status and revegetation practices of downhill ski-
ing areas in Finnish Lapland: a case study of Mt Ylläs.
Landscape and Urban Planning, 50, 259–268.
Seilbahnen Schweiz (2001) Wirtschaftsbericht 2001. Paul
Haupt, Bern, Switzerland.
SPSS (1999) SPSS for Windows. SPSS Inc., Chicago, IL.
Stanton, M.L. & Galen, C. (1997) Life on the edge: adapta-
tion versus environmentally mediated gene flow in the snow
buttercup, Ranunculus adoneus. American Naturalist, 150,
143–178.
Stoeckli, V. & Rixen, C. (2000) Characteristics of artificial
snow and its effect on vegetation. Conference Proceedings of
the International Snow Science Workshop (eds K. Birkeland,
E. Adams & F. Johnson), pp. 468–471. American Avalanche
Association, Bozeman USA, Big Sky, Montana, USA, 1–6
October 2000.
Tilman, D. (1996) Biodiversity: population versus ecosystem
stability. Ecology, 77, 350–363.
Titus, J.H. & Tsuyuzaki, S. (1999) Ski slope vegetation of
Mount Hood, Oregon, USA. Arctic, Antarctic, and Alpine
Research, 31, 283–292.
Urbanska, K.M. (1995) Biodiversity assessment in ecological
restoration above the timberline. Biodiversity and Conserva-
tion, 4, 679–695.
Urbanska, K.M. (1997) Restoration ecology research above
the timberline: colonization of safety islands on a machine-
graded alpine ski run. Biodiversity and Conservation, 6,
1655–1670.
Vujnovic, K., Wein, R.W. & Dale, M.R.T. (2002) Predict-
ing plant species diversity in response to disturbance
magnitude in grassland remnants of central Alberta. Canadian
Journal of Botany–Revue Canadienne de Botanique, 80,
504–511.
Wilmanns, O. (1998) Ökologische Pflanzensoziologie, 6th edn.
Quelle und Meyer, Wiesbaden, Germany.
Wipf, S., Rixen, C., Freppaz, M. & Stoeckli, V. (2002) Ski piste
vegetation under artificial and natural snow: patterns in
multivariate analysis. Interdisciplinary Mountain Research
(eds R. Bottarin & U. Tappeiner), pp. 170–179. Blackwell
Science, Berlin, Germany.
Received 5 June 2004; final copy received 22 December 2004
