Environmental perturbation, grazing pressure and soil
wetness jointly drive mountain tundra toward divergent
alternative states
Patrick Saccone1
*, Tuija Pyykkonen1
, Anu Eskelinen1,2
and Risto Virtanen1
1
Department of Biology, University of Oulu, PO Box 3000, Oulu FI-90014, Finland; and 2
Department of Environmental
Science and Policy, University of California Davis, One Shields Avenue, Davis, California 95616, USA
Summary
1. Plant communities are structured by complex interactions between multiple factors, which veil
our understanding of the effects of environmental changes on communities and ecosystems. Besides
the relative role of biotic and abiotic factors as community-structuring processes, addressing how
they jointly affect the ecological resilience and resistance of plant communities is crucial to understand
better the long-term response of communities facing global changes.
2. Here, we used the results from a long-term (23 years) perturbation experiment set up in Fennoscandian
mountain tundra to test these mechanisms. The experiment consisted of a transplantation of
twenty blocks of Vaccinium myrtillus heath vegetation including upper soil layer from a lower elevation
tundra heath habitat to a snowbed habitat 150 m higher in elevation where V. myrtillus lies
at its upper limit. In the snowbed with contrasting levels of soil wetness, half of the transplanted
blocks were protected from mammalian herbivores.
3. Our results revealed that in addition to the important role of environmental conditions as a structuring
force, the joint effects of multiple drivers resulted in divergent patterns in both plant functional
composition and species diversity among transplanted communities. Under environmental
perturbation (i.e. transplantation to snowbed), the heath vegetation was altered by grazing pressure
that reduced the cover of shrubs (especially V. myrtillus). In grazed dry snowbed, a species-rich
community with high functional type evenness and diversity developed. Reversely, in dry exclosures,
V. myrtillus gained high dominance associated with only few graminoids and forbs. In wet
snowbed conditions, shrubs tended to decline both in grazed plots and exclosures whereas bryophytes
attained high abundance. Grazing promoted species richness while soil waterlogging tended
to promote among-plot heterogeneity (b-diversity) which was highest in wet exclosures.
4. Synthesis. Our long-term experiment reveals that environmental perturbation, grazing and soil
wetness exhibit joint effects that induce divergent trajectories of tundra plant communities. We suggest
that a strong environmental perturbation triggers mountain tundra heath community to move
away from its equilibrium state. The outcome of this shift depends on the interplay between grazing
pressure and soil wetness that drive tundra plant communities toward divergent alternative states.
Key-words: alternative states, determinants of plant community diversity and structure, diversity,
herbivory, joint effects, long-term experiment, multiple drivers, species abundance distribution, tundra
shrubification
Introduction
The greening and shrubification of tundra are visible evidences
of the immediate consequences of global change in
arctic regions (Tape, Sturm & Racine 2006; Walker et al.
2006). The expected magnitude of tundra vegetation changes
in terms of functional composition and texture are suggested
to have strong implications for the regional and global feedback
with the climate and global changes (Cahoon et al.
2012; Loranty & Goetz 2012). However, recent studies reveal
that the shrub expansion is also context-dependent and variable
in space and time, underlining the importance of local
environmental conditions that can mask or buffer the main
expansion pattern (Elmendorf et al. 2012; Tape et al. 2012;*Correspondence author. E-mail: patrick.saccone@gmail.com
© 2014 The Authors. Journal of Ecology © 2014 British Ecological Society
Journal of Ecology 2014, 102, 1661–1672 doi: 10.1111/1365-2745.12316
Epstein, Myers-Smith & Walker 2013). Several studies from
contrasting habitats also reveal that local conditions can mediate
the response of communities to global change (Brooker
2006; Saccone et al. 2010; Kaarlej€arvi, Eskelinen & Olofsson
2013; Liancourt et al. 2013), because community composition
and functioning emerge from a complex interplay of processes
involving both biotic and abiotic factors (Lortie et al.
2004). In this context, the evaluation of the long-term consequences
of global change on arctic ecosystems requires a better
understanding of the complex web of interactions and
processes structuring tundra plant communities.
In low-productivity environments such as mountain tundra,
environmental constraints are commonly considered as the
main drivers of the composition of plant communities (K€orner
2003). In cold tundra biomes, for instance, soil disturbances,
air temperature or short growing season are suggested to
function as strong filters of the species pool (Keddy 1992). At
the same time, the importance of biotic factors (i.e. their
impact relative to other processes) such as plant interactions
or grazing as community structuring forces is expected to
decrease along increasing environmental severity (Grime
2001; Brooker et al. 2005; Michalet et al. 2006; Lezama
et al. 2014). As a result, tundra communities are dominated
by stress-tolerant, long-lived species capable of withstanding
strong inter-annual environmental variation and perturbations
(Hudson & Henry 2010; Wahren et al. 2013).
On the other hand, there are theoretical studies which predict
that grazing pressure exerts strong impacts on plant community
composition (Noy-Meir 1975; Fretwell 1977; Oksanen
et al. 1981), and that plant functional composition of tundra
is primarily structured by grazing pressure (Oksanen 1990).
Several studies from tundra regions emphasize that grazing
pressure strongly affects plant community composition (Olofsson
et al. 2009), plant species performance (Eskelinen 2008)
and plant stoichiometry (Zamin & Grogan 2013). These studies
suggest that the development of vegetation states in tundra
could be sensitive to shifts in grazing pressure. In this vein,
van der Wal (2006) has proposed that grazing pressure drives
shifts of tundra plant communities between lichen-, moss- and
grass-dominated alternative states. Indeed, several studies
report alternative tundra vegetation states and propose that
they are produced by the levels of grazing (Oksanen & Virtanen
1995; Zimov et al. 1995; Olofsson 2006; van der Wal &
Hessen 2009). Moreover, some studies also highlight that the
effects of grazing on tundra plant communities cannot be considered
in isolation from several locally varying factors such
as soil nutrient status (Eskelinen, Harrison & Tuomi 2012)
and many abiotic conditions changing with an elevation gradient
(Speed, Austrheim & Mysterud 2013).
Instead of viewing tundra plant communities driven solely by
either abiotic or biotic factors, a more fruitful approach is to
address the joint effects of multiple forces that may manifest
themselves in different relative strengths depending on the context.
Houseman et al. (2008) showed that multiple perturbations
may have complex effects on community dispersion that
are difficult to predict based on analysis of single factors and
that lead to divergent alternative states. Such ideas are well
accommodated by the concept of alternative stable states (ASS;
see Beisner, Haydon & Cuddington 2003 for a review) which
proposes that a perturbation of community-structuring processes
could move a community from a stable state to another
by triggering a crossover a threshold point between two basins
of attraction. These transitions are likely to depend on interacting
abiotic and biotic forces. However, the responses of plant
communities to multiple perturbations need to be studied over
time scales where ecological processes have even a theoretical
chance to reach a steady state (Schr€oder, Persson & De Roos
2005), especially in tundra ecosystems where most species are
long-lived and harsh environmental conditions limit the growth
of plants (Sonesson & Callaghan 1991; Grime 2001). Longterm
experiments are the most appropriate tool to disentangle
the interplay between biotic and abiotic factors acting in plant
communities and to assess their consequences for multiple vegetation
states (Tilman 1989).
Here, we report results from a 23-year experiment, established
in 1989 in mountain tundra of Finnish Lapland, to test
the mechanisms underlying the long-term effects of environmental
perturbation and local processes on plant community
composition and diversity, and development of alternative
states. The experiment consisted of the transplantation of
blocks (including soil and vegetation) of low elevation Vaccinium
myrtillus heath vegetation to a snowbed site 150 m
higher in elevation. This transplantation simulates strong perturbation
of environmental conditions since the surrounding
species pool, climate and hydrology, are altered. The blocks
transplanted to the snowbed were assigned to a grazing pressure
treatment in contrasting conditions of soil wetness. As
environmental conditions are predicted to be important determinants
of tundra plant assemblages, we expected that the
environmental perturbation due to the transplantation to the
snowbed would impose a considerable change in the transplanted
communities. However, as local biotic and abiotic
conditions are also predicted to play an important role in driving
tundra community dynamics, we primarily hypothesized
that grazing and soil wetness conditions would jointly affect
the development of vegetation states and yield divergent community
trajectories following environmental perturbation.
Materials and methods
STUDY SITE
The study was carried out on Mt. Jehkats at Kilpisj€arvi in the northwestern
Finnish Lapland (69°010
N and 20°500
E). The mountain tundra
at altitudes of 600–800 m above sea level (a.s.l.) is characterized
by dwarf shrub heaths which are more fragmented above 800 m a.s.l.
where sedge-grass meadows and snowbed type communities occur
extensively (Oksanen & Virtanen 1995). Vaccinium myrtillus, a deciduous
dwarf shrub, dominates heath vegetation on slope sites with
intermediate snow persistence and it is usually absent from snowbed
sites, where a short growing season likely limits plant growth (Wijk
1986; Sonesson & Callaghan 1991). In the study area, V. myrtillus
heaths are common up to altitudes of about 700 m a.s.l. and occur
sporadically up to altitudes of about 900 m a.s.l. (Oksanen & Virtanen
1995).
© 2014 The Authors. Journal of Ecology © 2014 British Ecological Society, Journal of Ecology, 102, 1661–1672
1662 P. Saccone et al.
The gradient from the low elevation heath to the high elevation
heath-snowbed mosaic represents a substantial change in the characteristics
of physical environment. The aerial primary net production
on heaths rich in V. myrtillus is about 294 Æ 41 g mÀ2
yearÀ1
, while
that of snowbed sites rich in Salix herbacea is
65 Æ 23 g mÀ2
yearÀ1
(Kyll€onen 1988). In the snowbed site, snow
cover remains till late June which is about 10 days later than in the
heath site 150 m lower in the altitude (own observations). Soil nutrients
in the heath and snowbed habitats are at about equal levels (Virtanen
1998).
The snowbed habitat is grazed by arvicolids and reindeer (Rangifer
tarandus (Smith 1827)) that are the most important grazers of the
study area. The grey-sided vole (Myodes rufocanus (Sundevall 1846))
uses V. myrtillus as its main winter food (Kalela 1957). The moss
and graminoid specialist Norwegian lemming (Lemmus lemmus
(Linnaeus 1758)) exerts the strongest grazing impact on snowbed vegetation
(Virtanen, Parviainen & Henttonen 2002) and can mechanically
injure shrubs. Semi-domesticated reindeer induces a relatively
heavy grazing and trampling pressure during the summer season (Oksanen
& Virtanen 1995). The densities of arvicolids fluctuated
throughout the whole study period and two outbreaks of lemmings
occurred (in 1998–99 Virtanen, Parviainen & Henttonen 2002; and in
2010–11 H. Henttonen, unpubl. data). The grazing pressure on low
elevation heath habitat is lower than on the snowbed. On the one
hand, reindeer preferably visit snowbed areas richer in sedge and
grasses and, on the other hand, the density of arvicolids is lower in
the heath (H. Henttonen, unpubl. data).
EXPERIMENTAL DESIGN
In late June 1989, 30 40 9 50 cm blocks of V. myrtillus heath vegetation
including 5–10-cm thick layer of humus and mineral soil were
dug out from the natural heath habitat lying at an altitude of 660 m
a.s.l. and randomly allocated into different treatments. To be able to
test for the transplantation and environmental perturbation effects, 10
blocks were transplanted within the original V. myrtillus heath and
were used as control plots (C). The environmental perturbation consists
in the transplantation of the remaining 20 blocks to a snowbed
site of about 15 9 30 m, 150 m higher in elevation. As a result, the
blocks originating from the low elevation heath have been transplanted
in a new habitat with a shorter growing season, a lower productivity
level, harsher climatic conditions, higher grazing pressure,
different hydrologic and topographic conditions and a different local
species pool in the surrounding. Ten blocks within the snowbed site
were protected from herbivores (E for exclosure) using metal mesh
fences that were 60–70 cm in diameter, 60-cm high with 1-cm mesh
size, and dug into the soil. The remaining 10 blocks were left grazed
(G). Fences excluded small mammal herbivory and the clear contrast
in average height of fenced (15.9 Æ 1.04 cm) and grazed
(6.75 Æ 0.77 cm) communities attests that fences also strongly
reduced but not fully excluded reindeer grazing (some herbivory
marks on the top of taller plants) because of the open tops of the
fences. Within the recipient snowbed, meso-topography induces differences
in soil wetness which delimit relatively dry and wet soil
areas characterized by Kiaeria and Polytrichum spp., and Sphagnum
spp., respectively. We confirmed these recurrent field observations of
soil wetness by gravimetric soil moisture measurements in midAugust
2012 (see Appendix S1 in Supporting Information) and the
20 snowbed blocks were split in dry (soil water content 64.9 Æ 4.9%
dry soil mass) and wet (soil water content 199.4 Æ 19.9% dry soil
mass) according to the significant differences between both groups
(see Appendix S1). Finally, the experimental design contained five
treatments: control plots (C), grazed plots (G.dry) and exclosures
(E.dry) on dry snowbed, and grazed plots (G.wet) and exclosures
(E.wet) on wet snowbed.
COMMUNITY COMPOSITION AND DIVERSITY, AND
STATISTICAL ANALYSES
In the central part of each block, a 0.125 9 0.125 m area was delimited
and preserved from any other experimental manipulations during
the 23 years of the experiment. This size of experimental plots was
suitable given the small stature and number of vascular plants (2–
15 cm and up to 16 species) and the high number of extremely small
cryptogam species in the plots (1–2 mm and up to 22 species) that
are traditionally considered as an important component of arctic vegetation
and ecosystem functioning (e.g., van der Wal, van Lieshout &
Loonen 2001). Also, we recorded 88 different species in the 30 study
plots (see Fig. 2 and Appendix S2), representing a large proportion
(> 50%) of the local diversity, which suggest that the design captured
most of the spatial heterogeneity in the system. In mid-August 2012,
the abundance of each plant species was estimated using visual
estimation method (Kent 2012) that was found the only feasible
method to record very small cryptogam and many small vascular
plant species. The nomenclature of the vascular flora, bryophytes and
lichens follows H€amet-Ahti et al. (1998), Ulvinen, Syrj€anen & Anttila
(2002) and Vitikainen et al. (1997), respectively. Because different
aggregation levels of the plant data may highlight different processes
(Dorrepaal 2007), we also pooled species by functional type (PFT)
and the aggregation commonly used in the studies of arctic vegetation,
that is, graminoids, forbs, shrubs, lichens and bryophytes (based
on Bruun et al. 2006) was used. Species richness and Shannon diversity
index were calculated for each plot and b-diversity (Sørensen
1948) values were calculated to quantify divergence of community
composition within each treatment.
We analysed the divergence of communities transplanted in the different
treatments using Non-Metric Dimensional Scaling (NMDS)
ordination. The ordinations were based on Bray-Curtis distances with
Wisconsin double standardized and square-root transformed species
and PFT cover data. The respective effects of experimental treatments,
that is, environmental perturbation, grazing pressure, soil wetness
and grazing pressure 9 soil wetness interaction on community
dissimilarities were evaluated using Permutational Multivariate Analyses
of Variances Models (PERMANOVA, Anderson 2001) and their significance
were assessed with 999 permutations. The differences in bdiversity
among the five treatments were evaluated using Permutational
Multivariate Dispersion Analyses Model (PERMDISP; Anderson,
Ellingsen & McArdle 2006) and their significance was assessed
with 999 permutations. We used ‘adonis’, ‘metaMDS’ and ‘betadisper’
functions in the VEGAN package (Oksanen et al. 2012) in R software
(R Development Core Team 2013) to perform the PERMANOVA
analyses, NMDS ordinations and PERMDISP analysis, respectively.
To analyse the effects of environmental perturbation, grazing pressure,
soil wetness and grazing pressure 9 soil wetness interaction on
PFT, V. myrtillus and S. herbacea cover and on diversity indices we
used two-way Generalized Linear Models (GLM) with a Gaussian
error distribution and an identity link function. We used otherwise
similar but one-way GLMs to test the difference among the five treatments
(i.e. C, G.dry, G.wet, E.dry and E.wet) for each PFT cover and
for diversity indices. In all analyses, data were square-root transformed
to meet the assumptions of the homogeneity of variances and
normality of errors. To test pairwise differences among treatments,
© 2014 The Authors. Journal of Ecology © 2014 British Ecological Society, Journal of Ecology, 102, 1661–1672
Alternative states of mountain tundra 1663
one-way GLMs were followed by post-hoc tests using ‘ghlt’ function
in the MULTICOMP package (Horthon, Bretz & Westfall 2008) in R.
Results
In general, all three treatment factors, that is, environmental
perturbation (transplantation to the snowbed), grazing pressure
and soil wetness significantly affected the composition of
transplanted communities, considering both species and
functional type based comparisons (Table 1). Both NMDS
ordinations (Fig. 1) illustrate the PERMANOVA results (Table 1)
and outline the divergence between the communities in
response to the 23 years of experimental treatments.
THE EFFECT OF ENVIRONMENTAL PERTURBATION DUE
TO THE TRANSPLANTATION TO SNOWBED
The clear separation along the first axis in both NMDS
ordinations (Fig. 1) between the control plots (C) and those
transplanted to the snowbed (G and E, dry and wet)
corresponded to 17% and 29% of the total variation in species
and functional composition, respectively, explained by the
environmental perturbation treatment (Table 1).
The small value of b-diversity in the control communities
(Table 2) indicated a relatively high level of homogeneity in
this community with a dominant species (V. myrtillus), some
co-dominant bryophyte species (e.g. Dicranum sp.) and
shrubs (Empetrum hermaphroditum, Phyllodoce caerulea and
S. herbacea), and several rare species (Figs 1a and 2a). In the
snowbed site, V. myrtillus remained dominant or co-dominant
species of the transplanted communities and reached comparable
abundance to the original heath site (Fig. 2), while
S. herbacea remained abundant only in dry conditions (Fig. 2
and Appendix S3). Reversely, E. hermaphroditum, P. caerulea
and Vaccinium vitis-idaea were clearly negatively
impacted by the transplantation to the snowbed, suggesting
these species tolerate snowbed conditions poorly.
The environmental perturbation also affected the cover of
graminoids, as reflected by their ingress in communities
Table 1. Results of Permutational Multivariate Analyses of Variances
(permanova) performed with environmental perturbation, soil wetness,
grazing pressure and the soil wetness 9 grazing pressure interaction
as explanatory variables on community cover matrix at the plant species
level (a) and on community cover matrix with species pooled by
plant functional type (b, PFT)
(a) Community (species) (b) Community (PFT)
Source of
variation d.f. F R2
P d.f. F R2
P
Environmental
perturbation
1 6.78 0.17 0.001 1 16.43 0.29 0.001
Soil wetness
(W)
1 3.15 0.08 0.002 1 7.55 0.13 0.001
Grazing
pressure(G)
1 3.19 0.08 0.003 1 6.41 0.11 0.001
W 9 G 1 1.92 0.05 0.036 1 1.31 0.02 0.252
Residuals 25 0.62 25 0.44
Ant alp
Car big
Car lac
Car rot
Car vag
Des fle
Eri ang
Eri vag
Fes ovi
Jun trif
Nar str
Phl alp
Vah atr
Bis viv
Cer cer
Cer fon
Epi ana
Gna sup
Leo aut
Lin bor
Ped lap
Ran acr
Ran pum
Sib pro
Sol vir
Tar sp.
Tri eur
Tro eur
Ver alp
Vio bif
Bar bin
Bar kun
Bar lyc
Cep.zia
Dip tax
Lop obt
Lop opa
Lop sud
Lop wen
Pel sp.
Ple alb
Tri qui
Bet nan
Emp
her
Phy cae
Sal her
Sal phy
Sal pol
Vac myr
Vac vit
Cet eri
Cet isl
Cet cri
Cla amaCla bel
Cla coc
Cla fir
Cla gra
Cla elo
Cla sp.
Cla
sub
Cla unc
Pel ruf
Ste sp.
Bar ith
Bra ref
Dic maj
Hyl spl
Kia gla
Kia sta
Ple sch
Poh nut
Poh sp.
Pol alp
Pol sexPol com
Pol hyp
Pol jun
Pol stri
San unc
Sph ang
Sph rus
Cet del
Cla cri
Dic
flex
Bar flo
Cla mit
Dic sco
E. dry
E. wet
E. dry
E. wet
G. dry
G. wet
G. dry
G. wet
(a) (b)Based on species cover matrix Based on plant functional types cover matrix
NMDS1
NMDS2
Fig. 1. Non-Metric Dimensional Scaling (NMDS) ordinations of plant community dissimilarity based on species cover (a) and on plant functional
type (PFT) cover (b). Plot labels are drawn at the centroids of each group and plots are located at the end of segments. Abbreviated species
names (see Appendix S1 for complete names) and plant functional types are additively plotted on the left (a) and right (b) ordination spaces,
respectively. NMDS stress = 0.2 (a) and 0.09 (b), respectively.
© 2014 The Authors. Journal of Ecology © 2014 British Ecological Society, Journal of Ecology, 102, 1661–1672
1664 P. Saccone et al.
transplanted to the snowbed site compared to their almost
complete absence in the control communities (Figs 2 and 3b,
Table 3). Some graminoid species attained especially high
cover in the snowbed (Fig. 2), for example Deschampsia
flexuosa, which reached 12.5% of the total cover in the dry
exclosures. We found similar pattern for forbs with a strong
effect of environmental perturbation but also significant
effects of soil wetness and grazing pressure (Table 3). Reversely,
the total abundance of bryophytes was not affected by
the environmental perturbation (Table 3). Finally, despite the
low overall cover and subordinate status of lichens (only 7%
of the total cover at the maximum, Figs 2 and 3e), they
strongly characterised the control community in the NMDS
ordination figure (Fig. 1b) and significantly decreased when
transplanted from the heath to the snowbed (Table 3).
THE JOINT EFFECTS OF ENVIRONMENTAL
PERTURBATION, GRAZING PRESSURE AND LOCAL SOIL
WETNESS
The communities transplanted to the snowbed were scattered
along the second axis of both NMDS ordinations and showed
distinct patterns depending on whether the ordination was
based on species or functional type data. In the ordination
space for species cover data (Fig. 1a), the grazed wet plots
(G.wet) were clearly isolated from the three other treatments.
In contrast, in the ordination for PFT cover data (Fig. 1b), the
dry exclosures (E.dry) were clearly separated from the three
other treatments. Such discrepancy among both analyses
Table 2. Mean Æ 1 SE of plant community diversity indexes
Treatments Species richness Shannon index b-diversity
Control 11.4 Æ 0.64 b 1.62 Æ 0.06 b 0.51 b
Grazed plots
Dry 18.5 Æ 2.22 a 2.33 Æ 0.18 a 0.53 ab
Wet 21.25 Æ 3.28 a 2.25 Æ 0.23 ab 0.59 ab
Exclosures
Dry 9.33 Æ 1.36 b 1.58 Æ 0.23 b 0.57 ab
Wet 13.25 Æ 2.93 ab 1.74 Æ 0.25 ab 0.66 a
*** * *
Asterisks under species richness and Shannon index columns indicate
significant differences among treatments tested by one-way Generalized
Linear Models on square-root transformed data with Gaussian
error distribution and an identity link function (***P < 0.001;
*P < 0.05). Asterisk under b-diversity column indicates significant
differences among treatments tested by Permutational Multivariate
Dispersion Analyses (PERMDISP) model (*P < 0.05). Different
letters indicate significantly different groups based on post hoc tests
(P < 0.05). b-diversity index were calculated using the Sørensen
index of dissimilarity (Sørensen 1948) at the treatment level.
Des fle
Emp her
Phy cae
Sal
her
Vac myr
Vac myr
Vac myr
Vac myr
Vac myr
Species rank (number)
Sal her
Des fle
Des fle
Sph ang
Bar lyc
Bar lyc
Dic fle
Sal her
Pol com
Bar lyc
Sph ang
(a)
(b) (c)
(d) (e)
Fig. 2. Species Abundance Distribution curves by treatment. Mean Æ 1 SE of plant cover ranked by order of dominance. Arrows represented the
number of species occurring at least once in the blocks of each treatment. Data symbols plotted at the mean values figured the plant functional
type (PFT) of the species concerned. The abbreviated names of the most abundant species were adjoined to the error bars of corresponding point
(see Appendix S1 for complete names).
© 2014 The Authors. Journal of Ecology © 2014 British Ecological Society, Journal of Ecology, 102, 1661–1672
Alternative states of mountain tundra 1665
could probably be attributable to the uneven distribution of
species among functional types (see Appendix S2) and among
treatments, and in particular to the number of bryophytes species
occurring in abundance in wet conditions. Moreover,
each functional type includes a wide range of species with
different moisture preference ranging from rich liverwort
assemblages to hygrophilic Sphagnum spp. in bryophytes or
the generalist D. flexuosa and the stress-tolerant Nardus
stricta in graminoids which lead to different ordinations of
communities. Nonetheless, both ordinations shared the same
treatment ranking along the second axis: G.wet, E.wet, G.dry
and E.dry. To some extent, the second axes of both ordinations
reflected grazing pressure and soil wetness gradients
which explained between 8% and 13% of the total variation
in species and functional composition respectively (Table 1).
Grazed plots on snowbed harboured richer communities
(Table 2, Fig. 2b,d and see also Appendix S3), and they
seemed to be more homogenous as shown by lower b-diversity
(Table 2). After 23 years in the snowbed site, the shrub
cover was strongly reduced by the grazing pressure and, to a
lesser extent, by the environmental perturbation and the wet
conditions (Fig. 3a, Table 3). Grazing reduced the dominance
of shrubs and other potential dominants, and this favoured
low-growing forbs, graminoids, lichens and bryophytes, leading
to high evenness of functional types (Fig. 2b,d). The species
abundance distribution (SAD) curves of grazed plots
were flat showing a long tail of species with low cover and
the absence of any strongly dominating species. Specifically,
there were 22 and 17 species covering more than 1% of the
plot for G.dry and G.wet plots, respectively, whereas there
was only nine species with >1% cover for C plots (Fig. 2).
The positive effect of the grazing pressure on diversity was
greater in dry conditions with G.dry and E.dry communities
having the highest and lowest Shannon index values, respectively
(Table 2). This matched with a strong negative effect
of grazing on V. myrtillus on the dry snowbed site (Fig. 2
and Appendix S3), implying that the dominance of V. myrtillus
was associated with low species diversity. Indeed,
V. myrtillus reached its highest cover in dry exclosures (37%,
see Fig. 2) and shared the dominance with bryophytes and
graminoids in both grazed and exclosures plots on the wet
snowbed (Figs 1 and 2). Moreover, in dry exclosures 12 of
the 13 species with a cover >1% were vascular plants, and
forbs represented half of the subordinate species covering
more than 0.5% of the plot (Fig. 2c).
The effect of the grazing pressure was less pronounced in
communities transplanted to the wet snowbed where there
was strong variation in the cover of most abundant species
(see length of SE bars in Fig. 2). This finding coincided with
patterns of b-diversity, which was highest in the wet snowbed
plots (Table 2). The cover of bryophytes was affected by the
interactive effects of grazing and soil wetness (Table 3): bryophytes
abounded in wet conditions, remained abundant under
grazing in dry conditions, and were very scarce in the dry exTable
3. Results of the two-way Generalized Linear Models with environmental perturbation, soil wetness, grazing pressure and soil wetness
9 grazing pressure interaction as explanatory variables, and the cover of each plant functional type as response variables
Source of variation Resid. d.f.
Graminoids Lichens
Resid. dev F P Resid. dev F P
Null 29 128.67 40.90
Environmental perturbation 28 44.71 50.05 < 0.0001 27.97 14.43 0.0008
Soil wetness (W) 27 43.76 0.56 0.4525 27.94 0.03 0.8644
Grazing (G) 26 43.37 0.23 0.6346 22.65 5.9 0.0226
W 9 G 25 41.94 0.85 0.3644 22.41 0.27 0.6103
Forbs Bryophytes
Resid. dev F P Resid. dev F P
Null 29 116.09 253.43
Environmental perturbation 28 41.49 66.56 < 0.0001 251.63 0.35 0.5601
Soil wetness (W) 27 37.00 3.99 0.0565 161.42 17.48 0.0003
Grazing (G) 26 31.20 5.18 0.0317 150.87 2.04 0.1652
W 9 G 25 28.02 2.84 0.1041 129.04 4.23 0.0503
Shrubs
Resid. dev F P
Null 29 110.54
Environmental perturbation 28 94.32 7.98 0.0092
Soil wetness (W) 27 85.81 4.19 0.0513
Grazing (G) 26 52.83 16.23 0.0005
W 9 G 25 50.80 1.00 0.3261
Bold P values indicate significant effects of the explanatory variables.
© 2014 The Authors. Journal of Ecology © 2014 British Ecological Society, Journal of Ecology, 102, 1661–1672
1666 P. Saccone et al.
closures (Fig. 3d). On individual species level, the ground
layer of the original heath community was dominated by Dicranum
spp. and Barbilophozia spp. which became largely
replaced by hydrophilic mosses in the wet conditions, while
Barbilophozia spp. persisted in the grazed dry plots (Figs 1
and 2). Specifically, Sphagnum angustifolium in the grazed
wet plots reached 17% cover on average and Polytrichum
commune in the wet exclosures reached 26% cover. Finally,
while the cover of forbs was also negatively affected by the
combined effects of grazing and high soil wetness (Fig. 3c),
the cover and occurrence of lichens were only affected by the
grazing pressure regardless of soil wetness (Table 3), with a
strong negative effect of fencing (Fig. 2).
Discussion
The results of our long-term perturbation experiment show
evidence for markedly distinct vegetation states developed in
response to interactions of different treatments. This finding
supports our hypotheses that the development of vegetation
states is driven by joint effects of environmental filtering,
local biotic and abiotic conditions. These joint effects of multiple
drivers contrast with simple traditional ideas that abiotic
constraints and environmental filtering alone would explain
the majority of vegetation patterns and states in tundra, and
they also do not suggest that mere top–down control by grazers
would drive vegetation states regardless of abiotic
conditions. Our results rather support propositions stating that
combinations of abiotic factors and biotic processes are crucial
in the development of tundra vegetation states (see van
der Wal 2006; Houseman et al. 2008). Below, we discuss the
effects of the transplantation to the snowbed as a general
environmental perturbation and how it interacts with local
grazing pressure and soil wetness to develop distinct assemblages.
Furthermore, we discuss how their joint effects influence
community trajectories and mechanisms underlying the
development of alternative states in tundra vegetation.
CHANGES IN TRANSPLANT COMMUNITIES AND
UNDERLYING PROCESSES
Seven years after the transplantation, Virtanen (1998) noted a
consistent decrease of the initially dominant shrub species
(V. myrtillus) when transplanted to the snowbed, which was
interpreted to reflect the importance of environmental conditions
as the driving force in the tundra, according to the traditional
view (Billings & Mooney 1968; Grime 2001). Other
shrubs and herbaceous plants showed variable and often nonsignificant
changes in response to the treatments which
emphasized relatively high resistance of tundra heath communities
to the 7-year treatment. Now after 23 years, a time span
that largely exceeds the life time of most genets and/or ramets
(Rixen et al. 2004; Nestby et al. 2011), this initial resistance
was not evident any more, and strong and often contrasting
(a) (b) (c)
(d) (e)
Fig. 3. Mean Æ 1 SE of plant functional type (PFT) cover in each treatment. Asterisks following PFT names indicate significant differences of
cover among treatments tested by Generalized Linear Models (GLM) on square-root transformed data with Gaussian error distribution and an
identity link function (***P < 0.001; **P < 0.01). Different letters indicate significantly different groups based on post hoc tests (P < 0.05).
© 2014 The Authors. Journal of Ecology © 2014 British Ecological Society, Journal of Ecology, 102, 1661–1672
Alternative states of mountain tundra 1667
species-specific responses of transplant communities to the
treatments occurred. Some shrub species from the original
heath virtually disappeared in the snowbed habitat, while the
abundance of the some other species was mainly determined
by the level of grazing and was also negatively affected by
soil wetness. These findings demonstrate the transient status
of the first results (Virtanen 1998) and that short and longterm
outcomes of ecological processes can markedly differ
(Tilman 1989; Schr€oder, Persson & De Roos 2005). Moreover,
this strong role of grazing and the fact that its effects
were more pronounced in dry snowbed conditions suggest
that mere environmental filtering mechanisms (Keddy 1992),
regarded as predominant in arctic and alpine systems (Grime
2001; K€orner 2003) are not sufficient to exclude shrubs from
snowbeds or to limit tundra shrubification. In contrast, our
findings highlight the importance of grazing as an efficient
biotic filter restricting the spread of dwarf shrubs to mountain
tundra snowbeds (see also Olofsson et al. 2009), and that the
outcome also depends upon soil wetness conditions and species
concerned. Thus, environmental constraints, grazing and
soil wetness may jointly filter shrub species from tundra
snowbeds.
The second long-term consequence of the transplantation of
communities to the snowbed was the strong contribution of
graminoids and forbs in the new assemblages. Because graminoids
and forbs are well established in the recipient site surrounding
the transplant communities while they are nearly
absent in the original V. myrtillus heath, such a result was
expected. However, we found also that they established relatively
equally in transplanted plots exposed to different treatments,
even in exclosures with high shrub cover. This finding
suggests that they are not strongly suppressed by shrub competitors
in harsh snowbed conditions, which is consistent with
ideas that the importance of competition could be low under
severe environmental conditions (Brooker et al. 2005; see
also Virtanen 1998). The graminoid PFT includes generalist
species exhibiting life strategies which combine good competitive
abilities with a tolerance to grazing pressure and wet soil
conditions (e.g., D. flexuosa, CSR, Grime 1988 or grazing
tolerant-exploitative-competitive species, Oksanen 1990). In
contrast, forbs present in the snowbed area may be less tolerant
to soil waterlogging and may also become negatively
impacted by thickening of the moss carpet (Gornall et al.
2011). It is also evident that the responses of forbs to grazing
reflect marked species-specific differences. Relatively tall
herbs such as Solidago virgaurea had fairly high cover in exclosures,
whereas prostrate herbs such as Epilobium anagallidifolium
were more characteristic in grazed plots. This is
compatible with previous studies showing that tall forbs are
sensitive to grazing whereas prostrate herbs are favoured by
grazing because they can avoid it (Lavorel et al. 1997; Kaarlej€arvi,
Eskelinen & Olofsson 2013).
The finding that bryophytes were hampered by the reduction
of grazing under dry conditions, where shrubs strongly
dominated, while they benefitted from the reduction of grazing
in wet conditions, is generally consistent with other studies
showing a generally negative relationship between tundra
shrubs and bryophytes (Pajunen, Oksanen & Virtanen 2011;
Pajunen, Virtanen & Roininen 2012). But this also suggests
that bryophyte abundance in tundra snowbeds is not merely
determined by direct negative grazing impacts (Moen, Lundberg
& Oksanen 1993; Virtanen, Henttonen & Laine 1997).
Indeed, competition with vascular plants and soil wetness act
in concert, and grazer-mediated impact on competition with
vascular plants may be an important force dictating bryophyte
responses.
Reversely to forbs and graminoids, lichens decreased in
response to the transplantation to the snowbed which is in
accordance with their well-known sensitivity to prolonged
duration of snow and soil waterlogging (Benedict 1990; Bruun
et al. 2006). Moreover, the additional decrease of lichens
inside exclosures, where especially shrubs and forbs were
abundant, agrees with previous studies suggesting that the
development of dense vascular plant cover may suppress
lichens (Cornelissen et al. 2001). However, in the high arctic
tundra, an opposite result, that is increase of lichens and
Fig. 4. Conceptual diagram based on the
Non-Metric Dimensional Scaling ordinations
on species and plant functional type (PFT)
matrixes (Fig. 1) and convert in the way of
the traditional topographic analogy of
alternative stable states. We suggest that (i)
environmental perturbation triggers the
community to shift away from its original
alternative stable state (solid dark grey
arrow), then (ii) the level of grazing pressure
determines the alternative basin of attraction
(dotted dark grey arrows) and (iii) the soil
wetness level locates the community within
the basin (dotted light grey arrows).
© 2014 The Authors. Journal of Ecology © 2014 British Ecological Society, Journal of Ecology, 102, 1661–1672
1668 P. Saccone et al.
decrease of deciduous shrubs, was observed in long-term exclosures
in dry tundra sites (Johnson et al. 2011), which may
indicate that the interactions between lichens and vascular
plants are highly dependent on site conditions.
The positive effect of grazing on species richness (a-diversity)
in our study matches with other evidences from mountain
tundra habitats (Austrheim & Eriksson 2001; Speed,
Austrheim & Mysterud 2013). It is noteworthy, however, that
other studies have shown contrasting results that vary from
negative impact of grazing on diversity in nutrient-poor habitats
(Proulx & Mazumder 1998) to positive impact in nutrient-rich
conditions (Eskelinen, Harrison & Tuomi 2012) and
sometimes the effect is negligible (Mayer et al. 2009; Johnson
et al. 2011). Here, in accordance with the results reported
by Olofsson (2006), the increase in species richness is associated
with simultaneous decreases of shrub cover and b-diversity.
Interestingly, such homogenization tendency was
associated with high species richness, while the grazing
homogenization hypothesis advocates that grazing promotes
the dominance of some browsing-tolerant species (low species
richness) which, in turn, reduces b-diversity (Rooney 2009).
The shift in the shapes of the SAD curves from exclosures to
grazed plots, that is, from short-tailed steep curves to longtailed
flat curves, resembles those encountered, for example,
in very long-term nutrient addition treatments (May &
McLean 2007). However, this latter example of niche differentiation
dynamic (Silvertown 2004) was due to the increase
of interspecific competition in resource-rich conditions while
our results reflect the outcomes of processes where grazing
limits any species from gaining dominance and subsequent
niche release which may indicate that grazing functions as a
stabilising force in the community (Adler, HilleRisLambers &
Levine 2007).
Community heterogeneity (b-diversity) tended to be higher
in wet snowbed conditions than in dry conditions which could
reflect an engineering effect of the bryophyte mat in wet conditions
able to harbour a large pool of subordinate species
that are susceptible to vary in cover and identity between the
plots, thus promoting heterogeneity. Thick bryophyte mats in
cold and slightly disturbed tundra environments reduce soil
temperatures and retain wetness, and this way slow down
nutrient cycling and further promote development of mossy
tundra (van der Wal, van Lieshout & Loonen 2001; Michel
et al. 2012). As a result, the developing moss may modify
the environmental conditions for subordinate species (Gornall
et al. 2011) and in particular vascular plants. Indeed, we also
found that high soil wetness had a negative effect on shrub
abundance especially inside exclosures, which could act,
together with moss engineering effects, as niche differentiation
processes for subordinate species and result in greater
community heterogeneity.
LONG-TERM DIVERGENCE AS AN EVIDENCE FOR
ALTERNATIVE STATES?
Despite the lack of a clear or generally accepted definition (Fukami
& Nakajima 2011), ASS represent a major conceptual
framework when addressing the relative importance of drivers
of plant community composition (Beisner, Haydon & Cuddington
2003). In his application of the ASS concept to boreal and
tundra ecosystems, van der Wal (2006) suggested that, along a
gradient of increasing grazing pressure, trampling and selective
feeding by reindeer induce a transition from lichen- to mossdominated
communities and that a transition from moss to
graminoid tundra occurs due to greater tolerance of graminoids
to repeated defoliation, enhancement of soil nutrients and
increase in soil temperature. Specifically for tundra heath communities,
he suggested that similar vegetation states resulting
from increasing grazing pressure can be identified on the basis
of replacement of lichens by mosses beneath shrubs, and by
replacement of shrubs by grasses (see Olofsson et al. 2001; Olofsson,
Stark & Oksanen 2004 for examples). In a recent study,
Olofsson, Moen & €Ostlund (2010) found no clear support of
the model proposed by van der Wal (2006) in boreal forest
floors, but they revealed that the local occurrence of the community
in a certain vegetation state may hinge on abiotic conditions
such as hydrological status or topography. Our results
support the idea that it may be overly simplistic to understand
alternative states without such connections among the multiple
drivers of plant communities.
Because our experiment tested the joint effects of three
main driving forces of the tundra system over a long-term
duration, we propose that placing our results in the theoretical
framework of ASS may reveal new insights about alternative
states of mountain tundra. van der Wal (2006) described alternative
states of tundra ecosystems mainly by the dominant
plant functional group. Here, our results suggest that the level
of species richness and the relative abundance of plant functional
types of a community could also reflect distinct attractors
(stable points) for mountain tundra vegetation. Indeed,
the strongest divergence appears in dry conditions between
the species-rich grazed communities with high evenness of
functional types and species-poor very lightly grazed (exclosures)
communities with strongly unbalanced representation
of PFTs. Grazing therefore appears to favour functional type
evenness and species diversity. Since niche differentiation
processes and forces favouring plant diversity and species
coexistence are considered as stabilising forces (Chesson
2000), grazing might enhance the resilience of an alternative
state. On the other hand, the unbalanced representation of
PFTs within the very lightly grazed communities could also
favour community stability because functionally poor communities
could have a limited potential to switch from a basin of
attraction to another (Didham, Watts & Norton 2005; Ives &
Carpenter 2007). Thus reduced grazing and resulting dominance
of competitive species might also form an alternative
state. As a result, we suggest that the contrasting communities
developed in grazed and ungrazed dry snowbed could represent
attractor points of our studied system.
Since wet bryophyte-dominated communities show a high
variability in species composition, they are still very likely to
be far from any new equilibrium. Additionally, they are less
divergent than the dry communities and harbour a large reservoir
of species from different functional types which might
© 2014 The Authors. Journal of Ecology © 2014 British Ecological Society, Journal of Ecology, 102, 1661–1672
Alternative states of mountain tundra 1669
facilitate community switches among multiple states. For
these reasons, we suggest that wet communities represent
transient alternative states (sensu Fukami & Nakajima 2011)
of one or the other grazing alternative basin. If bryophyte
dominance can characterize some tundra alternative states
(van der Wal 2006), this suggests that, in our system, soil
wetness conditions may reduce the resilience of communities
subject to contrasting level of grazing by maintaining high
compositional heterogeneity.
We summarize these interpretations in a conceptual model
(Fig. 4) where we suggest that the trajectories towards alternative
states are oriented at different levels by the combinations
of environmental perturbation, grazing pressure and soil
wetness conditions. We propose that environmental perturbation
triggers the community to shift away from the heath state
equilibrium. Similarly to Hidding, Tremblay & C^ote (2013)
who found that the presence of herbivores, even during a limited
period, can lead to alternative successional trajectories in
the boreal forest, here grazing pressure level then drives the
main trajectory toward high and low grazing alternative
basins of attraction. Finally, local soil wetness conditions
maintain alternative transient states of both low and high
grazing basins of attraction. Considering multiple drivers of
plant community states, the model highlights that the role of
each single factor is not necessarily separable from the other
factors and that their relative importance for community
trajectories is subject to variation. For instance, it is possible
that wet tundra snowbed could represent a basin of attraction
per se (the hollowed ‘wet’ grey cells in the conceptual
diagram, Fig. 4), but our transplant communities in wet conditions,
even after 23 years of perturbation treatment, lie still
in transient states with easy switching between alternative
states.
Since clear and indisputable evidence of ASS may not
really exist as stability is likely impossible to define (Ives &
Carpenter 2007), we do not suggest that our findings represent
the only possible outcome. Rather, we argue that these
patterns identified after more than 20 years of experimental
manipulations could help to better understand the role of the
multiple drivers of the mountain tundra vegetation and the
responses of these communities to environmental changes.
Our results highlight that local biotic and abiotic factors could
buffer or accelerate tundra responses to environmental change
and the resulting shifts in vegetation, for example, current
shrubification, and maintain alternative states.
Acknowledgements
This research was financially supported by the Academy of Finland (project #
259072 to R. Virtanen and project # 253385 to A. Eskelinen). We thank Kilpisj€arvi
Biological Station for logistic support during the maintenance of the
long-term experimental setups. We thank Lauren Sandhu, Matthew Heard and
two anonymous reviewers for their valuable comments on the early version of
the manuscript. The authors declare no conflict of interest.
Data accessibility
Data available from the Dryad Digital Repository (Saccone et al. 2014).
References
Adler, P.B., HilleRisLambers, J. & Levine, J.M. (2007) A niche for neutrality.
Ecology Letters, 10, 95–104.
Anderson, M.J. (2001) A new method for non-parametric multivariate analysis
of variance. Austral Ecology, 26, 32–46.
Anderson, M.J., Ellingsen, K.E. & McArdle, B.H. (2006) Multivariate dispersion
as a measure of beta diversity. Ecology Letters, 9, 683–693.
Austrheim, G. & Eriksson, O. (2001) Plant species diversity and grazing in the
Scandinavian mountains patterns and processes at different spatial scales.
Ecography, 24, 683–695.
Beisner, B., Haydon, D. & Cuddington, K. (2003) Alternative stable states in
ecology. Frontiers in Ecology and the Environment, 1, 376–382.
Benedict, J.B. (1990) Lichen mortality due to late-lying snow: results of a
transplant study. Arctic and Alpine Research, 22, 81–89.
Billings, W.D. & Mooney, H.A. (1968) The ecology of arctic and alpine plants.
Biological reviews, 43, 481–529.
Brooker, R.W. (2006) Plant–plant interactions and environmental change. New
Phytologist, 171, 271–284.
Brooker, R., Kikvidze, Z., Pugnaire, F.I., Callaway, R.M., Choler, P., Lortie,
C.J. & Michalet, R. (2005) The importance of importance. Oikos, 109, 63–
70.
Bruun, H.H., Moen, J., Virtanen, R., Grytnes, J.-A., Oksanen, L. & Angerbj€orn,
A. (2006) Effects of altitude and topography on species richness of vascular
plants, bryophytes and lichens in alpine communities. Journal of Vegetation
Science, 17, 37–46.
Cahoon, S.M.P., Sullivan, P.F., Shaver, G.R., Welker, J.M., Post, E. & Holyoak,
M. (2012) Interactions among shrub cover and the soil microclimate
may determine future Arctic carbon budgets. Ecology letters, 15, 1415–1422.
Chesson, P. (2000) Mechanisms of maintenance of species diversity. Annual
Review of Ecology and Systematics, 31, 343–366.
Cornelissen, J.H.C., Callaghan, T.V., Alatalo, J.M., Michelsen, A., Graglia, E.,
Hartley, a.E. et al. (2001) Global change and arctic ecosystems: is lichen
decline a function of increases in vascular plant biomass? Journal of Ecology,
89, 984–994.
Didham, R.K., Watts, C.H. & Norton, D.A. (2005) Are systems with strong
underlying abiotic regimes more likely to exhibit alternative stable states? Oikos,
110, 409–416.
Dorrepaal, E. (2007) Are plant growth-form-based classifications useful in predicting
northern ecosystem carbon cycling feedbacks to climate change?
Journal of Ecology, 95, 1167–1180.
Elmendorf, S.C., Henry, G.H.R., Hollister, R.D., Bj€ork, R.G., Bjorkman, A.D.,
Callaghan, T.V. et al. (2012) Global assessment of experimental climate
warming on tundra vegetation: heterogeneity over space and time. Ecology
Letters, 15, 164–175.
Epstein, H.E., Myers-Smith, I. & Walker, D.A. (2013) Recent dynamics of arctic
and sub-arctic vegetation. Environmental Research Letters, 8, 015040.
Eskelinen, A. (2008) Herbivore and neighbour effects on tundra plants depend
on species identity, nutrient availability and local environmental conditions.
Journal of Ecology, 96, 155–165.
Eskelinen, A., Harrison, S. & Tuomi, M. (2012) Plant traits mediate consumer
and nutrient control on plant community productivity and diversity. Ecology,
93, 2705–2718.
Fretwell, S.D. (1977) The regulation of plant communities by food chains
exploiting them. Perspectives in Biology and Medicine, 20, 169–185.
Fukami, T. & Nakajima, M. (2011) Community assembly: alternative stable
states or alternative transient states? Ecology letters, 14, 973–984.
Gornall, J.L., Woodin, S.J., Jonsdottir, I.S. & Van der Wal, R. (2011) Balancing
positive and negative plant interactions: how mosses structure vascular
plant communities. Oecologia, 166, 769–782.
Grime, J.P. (1988) Plant Evol. The C-S-R Model of Primary Plant Strategies –
Origins, Implications and Test (eds L.D. Gottlieb & S.K. Jain), pp. 371–393.
Chapman & Hall, Chichester, UK.
Grime, J.P. (2001) Plant Strategies, Vegetation Processes and Ecosystem Properties,
2nd edn. Wiley, Chichester, UK.
H€amet-Ahti, L., Suominen, J., Ulvinen, T. & Uotila, P. (eds.) (1998) Retkeilykasvio
(Summary: Field Flora of Finland), 4th edn. Finnish Museum of Natural
History, Botanical Museum, Helsinki.
Hidding, B., Tremblay, J.P. & C^ote, S.D. (2013) A large herbivore triggers alternative
successional trajectories in the boreal forest. Ecology, 94, 2852–2855.
Horthon, T., Bretz, F. & Westfall, P. (2008) Simultaneous inference in general
parametric models. Biometrical Journal, 50, 346–363.
Houseman, G.R., Mittelbach, G.G., Reynolds, H.L. & Gross, K.L. (2008) Perturbations
alter community convergence, divergence, and formation of multiple
community states. Ecology, 89, 2172–2180.
© 2014 The Authors. Journal of Ecology © 2014 British Ecological Society, Journal of Ecology, 102, 1661–1672
1670 P. Saccone et al.
Hudson, J.M. & Henry, G.H. (2010) High Arctic plant community resists
15 years of experimental warming. Journal of Ecology, 98, 1035–1041.
Ives, A.R. & Carpenter, S.R. (2007) Stability and diversity of ecosystems concepts
of stability. Science, 317, 58–62.
Johnson, D.R., Lara, M.J., Shaver, G.R., Batzli, G.O., Shaw, J.D. & Tweedie,
C.E. (2011) Exclusion of brown lemmings reduces vascular plant cover and
biomass in Arctic coastal tundra: resampling of a 50+ year herbivore exclosure
experiment near Barrow, Alaska. Environmental Research Letters, 6, 1–8.
Kaarlej€arvi, E., Eskelinen, A. & Olofsson, J. (2013) Herbivory prevents positive
responses of lowland plants to warmer and more fertile conditions at
high altitudes. Functional Ecology, 27, 1244–1253.
Kalela, O. (1957) Regulation of reproduction rate in subarctic population of
vole Clethrionomys rufocanus (Sund.). Annales Academiae Scientiarum Fennicae,
Series A IV, Biologica, 34, 1–60.
Keddy, P.A. (1992) Assembly and response rules: two goals for predictive
community ecology. Journal of Vegetation Science, 3, 157–164.
Kent, M. (2012) Vegetation Description and Data Analysis. A Practical
Approach, 2nd edn. John Wiley & Sons Ltd, Chichester, UK.
K€orner, C. (2003) Alpine Plant Life, 2nd edn. Springer, Berlin.
Kyll€onen, H. (1988) Alpine and Subalpine Vegetation at Kilpisj€arvi, FinnishLapland:
distribution of biomass and net production and annual variations in
biomass. Acta Universitatis Ouluensis Series A, 202, 1–78.
Lavorel, S., McIntyre, S., Landsberg, J. & Forbes, T.D.A. (1997) Plant functional
classifications: from general groups to specific groups based on
response to disturbance. Trends in Ecology & Evolution, 12, 474–478.
Lezama, F., Baeza, S., Altesor, A., Cesa, A., Chaneton, E.J. & Paruelo, J.M.
(2014) Variation of grazing-induced vegetation changes across a large-scale
productivity gradient. Journal of Vegetation Science, 25, 8–21.
Liancourt, P., Spence, L.A., Song, D.S., Lkhagva, A., Sharkhuu, A., Boldgiv,
B., Helliker, B.R., Petraitis, P.S. & Casper, B.B. (2013) Plant response to climate
change varies with topography, interactions with neighbors, and ecotype.
Ecology, 94, 444–453.
Loranty, M.M. & Goetz, S.J. (2012) Shrub expansion and climate feedbacks in
Arctic tundra. Environmental Research Letters, 7, 011005.
Lortie, C.J., Brooker, R.W., Choler, P., Kikvidze, Z., Michalet, R., Pugnaire,
F.I. & Callaway, R. (2004) Rethinking plant community theory. Oikos, 107,
433–438.
May, R. & McLean, A. (2007) Theoretical Ecology – Principles and Applications,
3rd edn. Oxford University Press, Oxford, UK.
Mayer, R., Kaufmann, R., Vorhauser, K. & Erschbamer, B. (2009) Effects of
grazing exclusion on species composition in high-altitude grasslands of the
Central Alps. Basic and Applied Ecology, 10, 447–455.
Michalet, R., Brooker, R., Cavieres, L.A., Kikvidze, Z., Lortie, C.J., Pugnaire,
F.I., Vialente-Banuet, A. & Callaway, R. (2006) Do biotic interactions shape
both sides of humped-back model of species richness in plant communities?
Ecology Letters, 9, 767–773.
Michel, P., Lee, W.G., During, H.J. & Cornelissen, J.H.C. (2012) Species traits
and their non-additive interactions control the water economy of bryophyte
cushions. Journal of Ecology, 100, 222–231.
Moen, J., Lundberg, P.A. & Oksanen, L. (1993) Lemming grazing on snowbed
vegetation during a population peak, Northern Norway. Arctic and
Alpine Research, 25, 130–135.
Nestby, R., Percival, D., Martinussen, I., Opstad, N. & Rohloff, J. (2011) The
European blueberry (Vaccinium myrtillus L.) and the potential for cultivation.
A review. The European Journal of Plant Science and Biotechnology, 5, 5–
16.
Noy-Meir, I. (1975) Stability of grazing systems: an application of predatorprey
graphs. Journal of Ecology, 63, 459–481.
Oksanen, L. (1990) Predation, herbivory and plant strategies along gradients of
primary productivity. Perspectives in plant competition (eds J. Grace & D.
Tilman), pp. 445–474. Academic Press, New York, NY, USA.
Oksanen, L. & Virtanen, R. (1995) Topographic, altitudinal and regional patterns
in continental and suboceanic heath vegetation of northern Fennoscandia.
Acta Botanica Fennica, 153, 1–80.
Oksanen, L., Fretwell, S.D., Arruda, J. & Niemel€a, P. (1981) Exploitation ecosystems
in gradients of primary production. American Naturalist, 118, 240–261.
Oksanen, J., Blanchet, F.G., Kindt, R., Legendre, P., Michin, P.R., O’Hara,
R.B., Simpson, G.L., Solymos, P., Stevens, M.H.H. & Wagner, H. (2012)
vegan: community ecology package. R package version 2.0-7, http://CRAN.
R-project.org/package=vegan.
Olofsson, J. (2006) Short- and long-term effects of changes in reindeer grazing
pressure on tundra heath vegetation. Journal of Ecology, 94, 431–440.
Olofsson, J., Moen, J. & €Ostlund, L. (2010) Effects of reindeer on boreal forest
floor vegetation: does grazing cause vegetation state transitions? Basic and
Applied Ecology, 11, 550–557.
Olofsson, J., Stark, S. & Oksanen, L. (2004) Reindeer influence on ecosystem
processes in the tundra. Oikos, 105, 386–396.
Olofsson, J., Kitti, H., Rautiainen, P., Stark, S. & Oksanen, L. (2001) Effects
of summer grazing by reindeer on composition of vegetation, productivity,
and nitrogen cycling. Ecography, 24, 13–24.
Olofsson, J., Oksanen, L., Callaghan, T., Hulme, P.E., Oksanen, T. & Suominen,
O. (2009) Herbivores inhibit climate-driven shrub expansion on the tundra.
Global Change Biology, 15, 2681–2693.
Pajunen, A., Oksanen, J. & Virtanen, R. (2011) Impact of shrub canopies on
understorey vegetation in western Eurasian tundra. Journal of Vegetation Science,
22, 837–846.
Pajunen, A., Virtanen, R. & Roininen, H. (2012) Browsing-mediated shrub canopy
changes drive composition and species richness in forest-tundra ecosystems.
Oikos, 121, 1544–1552.
Proulx, M. & Mazumder, A. (1998) Reversal of grazing impact on plant species
richness in nutrient-poor vs. nutrient-rich ecosystems. Ecology, 79,
2581–2592.
R Development Core Team (2013) R: A language and environment for statistical
computing. R Development Core Team, Vienna, Austria.
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.
Rooney, T.P. (2009) High white-tailed deer densities benefit graminoids and
contribute to biotic homogenization of forest ground-layer vegetation. Plant
Ecology, 202, 103–111.
Saccone, P., Pages, J., Girel, J., Brun, J. & Michalet, R. (2010) Acer negundo
invasion along a successional gradient: early direct facilitation by native pioneers
and late indirect facilitation by conspecifics. New Phytologist, 187,
831–842.
Saccone, P., Pyykkonen, T., Eskelinen, A. & Virtanen, R. (2014) Data from:
Environmental perturbation, grazing pressure and soil wetness jointly drive
mountain tundra toward divergent alternative states. Dryad Digital Repository,
.doi:10.5061/dryad.1154f
Schr€oder, A., Persson, L. & De Roos, A.M. (2005) Direct experimental evidence
for alternative stable states: a review. Oikos, 110, 3–19.
Silvertown, J. (2004) Plant coexistence and the niche. Trends in Ecology &
Evolution, 19, 605–611.
Sonesson, M. & Callaghan, T.V. (1991) Strategies of survival in plants of Fennoscandian
tundra. Arctic, 44, 95–105.
Sørensen, T.A. (1948) A method of establishing groups of equal amplitude in
plant sociology based on similarity of species content, and its application to
analyses of vegetation on Danish commons. Kongelige Danske Videnskabernes
Selskab Biologisk Skrifter, 5, 1–34.
Speed, J.D., Austrheim, G. & Mysterud, A. (2013) The response of plant diversity
to grazing varies along an elevational gradient. Journal of Ecology, 101,
1225–1236.
Tape, K., Sturm, M. & Racine, C. (2006) The evidence for shrub expansion in
Northern Alaska and the Pan-Arctic. Global Change Biology, 12, 686–702.
Tape, K.D., Hallinger, M., Welker, J.M. & Ruess, R.W. (2012) Landscape heterogeneity
of shrub expansion in Arctic Alaska. Ecosystems, 15, 711–724.
Tilman, D. (1989) Ecological experimentation: strengths and conceptual problems.
Long-term studies in ecology (ed. G.E. Likens), pp. 136–157. SpringerVerlag,
New York, NY, USA.
Ulvinen, T., Syrj€anen, K. & Anttila, S. (eds) (2002) Suomen sammalet- levinneisyys,
ekologia, uhanalaisuus. (Abstract: Bryophytes of Finland: distribution,
ecology and red list status). The Finnish Environment, 560, 1–354.
Virtanen, R. (1998) Impact of grazing and neighbour removal on a heath plant
community transplanted onto snowbed site, NW Finnish Lapland. Oikos, 81,
359–367.
Virtanen, R., Henttonen, H. & Laine, K. (1997) Lemming grazing and structure
of a snowbed plant community – a long-term experiment at Kilpisj€arvi, Finnish
Lapland. Oikos, 79, 155–166.
Virtanen, R., Parviainen, J. & Henttonen, H. (2002) Winter grazing by the Norwegian
lemming (Lemmus lemmus) at Kilpisj€arvi (NW Finnish Lapland) during
a moderate peak. Annales Zoologici Fennici, 39, 335–341.
Vitikainen, O., Ahti, T., Kuusinen, M., Lommi, S. & Ulvinen, T. (1997)
Checklist of lichens and allied fungi of Finland. Norrlinia, 6, 1–123.
Wahren, C.H., Camac, J.S., Jarrad, F.C., Williams, R.J., Papst, W.A. & Hoffmann,
A.A. (2013) Experimental warming and long-term vegetation dynamics
in an alpine heathland. Australian Journal of Botany, 61, 36–51.
van der Wal, R. (2006) Do herbivores cause habitat degradation or vegetation
state transition? Evidence from the tundra. Oikos, 114, 177–186.
van der Wal, R. & Hessen, D.O. (2009) Analogous aquatic and terrestrial food
webs in the high Arctic: the structuring force of a harsh climate. Perspectives
in Plant Ecology, Evolution and Systematics, 11, 231–240.
© 2014 The Authors. Journal of Ecology © 2014 British Ecological Society, Journal of Ecology, 102, 1661–1672
Alternative states of mountain tundra 1671
van der Wal, R., van Lieshout, S.M.J. & Loonen, M.J.J.E. (2001) Herbivore
impact on moss depth, soil temperature and arctic plant growth. Polar Biology,
24, 29–32.
Walker, M.D., Wahren, C.H., Hollister, R.D., Henry, G.H.R., Ahlquist, L.E.,
Alatalo, J.M. et al. (2006) Plant community responses to experimental warming
across the tundra biome. Proceedings of National Academy of Sciences
of the United States of America, 103, 1342–1346.
Wijk, S. (1986) Performance of Salix herbacea in an alpine snow-bed gradient.
Journal of Ecology, 74, 675–684.
Zamin, T.J. & Grogan, P. (2013) Caribou exclusion during a population low
increases deciduous and evergreen shrub species biomass and nitrogen pools
in low Arctic tundra. Journal of Ecology, 101, 671–683.
Zimov, S.A., Chuprynin, V.I., Oreshko, A.P., Chapin, F.S. III, Reynolds, J.F.
& Chapin, M.C. (1995) Steppe-tundra transition: a herbivore-driven biome
shift at the end of the Pleistocene. American Naturalist, 146, 765–794.
Received 12 March 2014; accepted 12 August 2014
Handling Editor: Matthew Heard
Supporting Information
Additional Supporting Information may be found in the online version
of this article:
Appendix S1. Mean Æ 1 SE of soil water content (% dry soil mass)
in the three conditions of the experimental design: The original heath
and the wet and dry areas of the snowbed.
Appendix S2. List of species observed in the experimental design
with their abbreviated species names reported in the Figs 1a and 2,
their attributed Plant Functional Type.
Appendix S3. Results of the two-way Generalized Linear Models
(GLM) with environmental perturbation, soil wetness, grazing pressure
and soil wetness 9 grazing pressure interaction as explanatory
variables, and the species richness, Shannon index and the cover of
Vaccinium myrtillus and Salix herbacea (dominant shrubs of the system)
as response variables.
© 2014 The Authors. Journal of Ecology © 2014 British Ecological Society, Journal of Ecology, 102, 1661–1672
1672 P. Saccone et al.