Environ Monit Assess (2016) 188: 452
DOI 10.1007/s10661-016-5457-2
Changes in vegetation cover and composition in the Swedish
mountain region
Henrik Hedenås & Pernilla Christensen &
Johan Svensson
Received: 14 December 2015 / Accepted: 27 June 2016 / Published online: 7 July 2016
# Springer International Publishing Switzerland 2016
Abstract Climate change, higher levels of natural re-
source demands, and changing land use will likely
lead to changes in vegetation configuration in the
mountain regions. The aim of this study was to
determine if the vegetation cover and composition have
changed in the Swedish region of the Scandinavian
Mountain Range, based on data from the long-term
landscape biodiversity monitoring program NILS
(National Inventory of Landscapes in Sweden).
Habitat type and vegetation cover were assessed
in 1740 systematically distributed permanent field
plots grouped into 145 sample units across the
mountain range. Horvitz–Thompson estimations
were used to estimate the present areal extension
of the alpine and the mountain birch forest areas
of the mountain range, the cover of trees, shrubs,
and plants, and the composition of the bottom
layer vegetation. We employed the data from two
subsequent 5-year monitoring periods, 2003–2007
and 2008–2012, to determine if there have been
H. Hedenås (*) : P. Christensen
Department of Forest Resource Management, Swedish University
of Agricultural Sciences, Skogsmarksgränd, 901 83 Umeå,
Sweden
e-mail: henrik.hedenas@slu.se
H. Hedenås
e-mail: henrik.hedenas@gmail.com
J. Svensson
Department of Wildlife, Fish, and Environmental Studies,
Swedish University of Agricultural Sciences, 901 83 Umeå,
Sweden
any changes in these characteristics. We found that
the extension of the alpine and the mountain birch
forest areas has not changed between the inventory
phases. However, the total tree canopy cover in-
creased in the alpine area, the cover of graminoids
and dwarf shrubs and the total cover of field
vegetation increased in both the alpine area and
the mountain birch forest, the bryophytes de-
creased in the alpine area, and the foliose lichens
decreased in the mountain birch forest. The ob-
served changes in vegetation cover and composi-
tion, as assessed by systematic data in a national
and regional monitoring scheme, can validate the
results of local studies, experimental studies, and
models. Through benchmark assessments, monitor-
ing data also contributes to governmental policies
and land-management strategies as well as to di-
rected cause and effect analyses.
Keywords Alpine vegetation . Climate change . Land
use . Monitoring . Mountain birch forest . Vegetation
change
Introduction
It has frequently been argued that climate change,
higher levels of natural resource demands, a greater
diversity of stakeholders, and changing land-use
policies will lead to changes in vegetation cover and
composition (e.g., Körner et al. 2005). Mountain vege-
tation in particular is anticipated to change to a
452 Page 2 of 15
significant degree (Theurillat and Guisan 2001;
Callaghan et al. 2013; Field et al. 2014). Vegetation
models and heating and snow manipulation experi-
ments indicate that large-size graminoids, shrubs,
and trees will be favored and expand in the mountains
on the expense of small-size graminoids, lichens, and
mosses, but the results of these experiments are not
unequivocal (Arft et al. 1999; Cairns and Moen 2004;
Moen et al. 2004; Kaplan and New 2006; Walker et al.
2006; Bokhorst et al. 2009; Olofsson et al. 2011;
Pearson et al. 2013). However, the responses of vegeta-
tion to climate change might be modified by herbivores
such as reindeer, voles, lemmings, moose, hares, and
geometrid moths (Wielgolaski 2005; Jepsen et al. 2008;
Olofsson et al. 2009; Van Bogaert et al. 2011; Callaghan
et al. 2013). For example, grazing by reindeer may
hamper the expansion of shrubs and trees in the alpine
tree-line ecotone (e.g., Olofsson et al. 2009). Further,
intense reindeer grazing might also induce a transforma-
tion from dwarf shrub-dominated heaths to graminoid-
dominated field vegetation (Olofsson et al. 2001). The
magnitude of outbursts of geometrid moths might in-
crease with a warmer climate, which may limit the
expansion and growth of trees and shrubs (Jepsen
et al. 2008; Callaghan et al. 2013). However, the mag-
nitude and regularity of vole cycles might decline in
response to warmer winter climate (Hörnfeldt 2004; Ims
et al. 2008), which in turn may release the growth of the
vegetation (Olofsson et al. 2009; Callaghan et al. 2013).
Hence, the main characteristics and predictions of
changes in vegetation cover and composition are com-
plicated owing to multiple and interacting drivers and
feedback responses as well as to stochastic events
(Callaghan et al. 2013).
It is also generally assumed that changes in the dis-
tribution and composition of the mountain vegetation
should be expected due to changes in land use over the
past century (Emanuelsson 1987; Theurillat and Guisan
2001; Körner et al. 2005; Tasser and Tappeiner 2009).
One example is the mountain birch forests adjacent to
settlements and seasonal mountain holding, which his-
torically were more open due to a relatively large out-
take of firewood and to more stationary grazing by
domestic animals. As these types of land use decrease
those lands naturally may undergo forest regeneration
succession (Kullman 1979; Olsson et al. 2000; Bryn and
Daugstad 2001; Karlsson et al. 2007; Östlund et al.
2015). It has also been suggested that many plants might
expand into and colonize higher altitude areas in the
Environ Monit Assess (2016) 188: 452
absence of grazing livestock (Speed et al. 2012).
Further, higher pressure from tourism and its as-
sociated infrastructure might lead to both direct
vegetation changes through disturbance of ground
surface vegetation, and to indirect changes due, for
example, to altered grazing pressure by reindeer as
reindeer might avoid areas with frequent tourism activ-
ities (Emanuelsson 1984; Van Bogaert et al. 2011;
Callaghan et al. 2013).
Expected changes in vegetation cover and composi-
tion might generate major impacts on ecological func-
tions, ecological processes, and ecosystem services in
the mountain regions (ACIA 2005; Körner et al. 2005).
For example, the mountain birch forest—which is an
important land-cover type in the reindeer husbandry
system—might become denser and for that reason
avoided by reindeer (Wielgolaski 2005), and the forag-
ing qualities in the alpine area might decrease owing to
decreases of palatable, nutritious plants that occur in
snow-bed vegetation (Edenius et al. 2003; Hedenås
et al. 2011). This will directly affect the prerequisites
for a sustainable livelihood for local indigenous people
and the long-term survival of the Sámi culture in the
northern half of Sweden (cf. Sandström 2015). Changes
in the vegetation cover might also negatively influence
the amenity values of the Swedish mountains for tourists
who commonly are attracted by the magnificent and
continuously open alpine areas (Bäck 2002; Heberlein
et al. 2002). Thus, ongoing and potential future changes
in the mountain vegetation are of great policy concerns
for governments (Beniston 2003), as emphasized by the
Government of Sweden in its environmental objective
‘A Magnificent Mountain Landscape’ (Government
Offices of Sweden 2004; SEPA 2014).
Until now, most evidence for changes in vegetation
cover and composition in alpine areas have been based
on re-inventories of old vegetation studies on a very
limited geographical range (e.g., Walther et al. 2005;
Moen and Lagerström 2008; Van Bogaert et al. 2011;
Wilson and Nilsson 2009; Kullman 2010; Virtanen
et al. 2010; Hedenås et al. 2011, 2012; Rundqvist et al.
2011; Callaghan et al. 2013; Wipf et al. 2013) and on a
few standardized monitoring programs (Pauli et al.
2012). One problem with re-inventories of old vegeta-
tion studies is that these plots have been selected for a
specific purpose and hence do not represent random or
systematic sample points. For example, many study
plots are biased towards sites where changes in vegeta-
tion were expected to be evident, and thus any
Environ Monit Assess (2016) 188: 452
extrapolation of such data over larger areas might
overestimate the actual vegetation change (Chytrý
et al. 2014). In this respect, it has been argued that
long-term data series provided by large-scale moni-
toring schemes have an important role for baseline
assessments and as background data for policy de-
velopment and as input to regional and national
environmental evaluation frameworks (Lovett et al.
2007; Lindenmayer and Likens 2010). The National
Inventory of Landscapes in Sweden (NILS) was
designed to provide possibilities for baseline assess-
ments in habitat and landscape changes (Ståhl et al.
2011). NILS covers all terrestrial land-cover types in
Sweden. The field inventory protocol encompasses a
wide range of variables and groups of variables that
provide valuable information about the status of the
biophysical mountain landscape and how it might
change over time. The NILS program covers data
collected from 2003 and onwards. With a 5-year
phase for a complete national sample, the data series
now allows for change analysis between two subse-
quent time periods (2003–2007 and 2008–2012).
The main aim of this study was to assess
regional-scale vegetation conditions and changes in
the alpine area and mountain birch forest in the
Scandinavian Mountain Range in Sweden. The data
used in the study are based on field inventory data
from the NILS program.
This study addresses the following four key ques-
tions: (1) What are the areal extensions of the alpine
area and mountain birch forest in Sweden? (2) Has
the areal extension of the alpine area and mountain
birch forest changed between the initial NILS inven-
tory phase (2003–2007) and the first re-inventory
phase (2008–2012)? (3) What are the compositions
and coverage of the tree, shrub, field, and bottom
layers in the alpine area and mountain birch forest?
(4) Has the composition and coverage of the vege-
tation in the alpine area and mountain birch forest
changed between the two inventory phases? To our
knowledge, this is the first study that compiles and
analyzes data from permanent field plots in a sys-
tematic monitoring design over such a large geo-
graphic area, to report on the status and change of
the areal extension of main habitat types in the
Scandinavian mountain range and their vegetation
characteristics. We foresee that the results will pro-
vide a benchmark for strategic and operational land-
use policies in relation to the national and regional
Page 3 of 15 452
environmental quality goals and also contributes
with generic baseline information for directed cause
and effect analyses.
Methods
NILS design
NILS is a sample-based monitoring program that covers
all terrestrial land-cover types in Sweden. Its core pur-
pose is to assess conditions and changes in landscape
biodiversity and to provide land-use data as background
information for the evaluation of environmental objec-
tives, international environmental reporting, applied re-
search, and land-use policy development (Ståhl et al.
2011). In the NILS system, the land surface of Sweden
is divided into ten strata wherein a total of 631 sampling
units (5 × 5 km) were selected in a systematically pattern
based on a random start unit. The mountain region is
defined as stratum 10 with an eastern border that follows
the mountain boundary defined by the Swedish Society
for Nature Conservation (SSNC) (von Sydow 1988;
Fig. 1) and with a western extension to the Norwegian
border. In total, stratum 10 includes 145 systematically
distributed sampling units with 12 systematically placed
20-m radius circular field sample plots in each unit. The
sample plots are permanent and are divided into sub-
plots if there are clearly defined compartments with
different land use, vegetation structure, etc. (Esseen
et al. 2007; Christensen and Ringvall 2013). The com-
partment division points are registered along the periph-
ery of the circular plots which makes it possible to
calculate the area of the subplot. Since each subplot is
described separately, it is possible to make precise var-
ious statistical estimates, e.g., areal extension of a cer-
tain habitat type or vegetation coverage in a specific
habitat type. The variable list consists of 269 field
variables, and it is constructed to both document spatial
patterns on broader geographic scales as well as to
provide detailed habitat and species data for multiple-
scale analyses of natural and land use-induced changes
in landscape biodiversity (Fig. 1). A proportion of these
variables provide information about configuration on
habitat level, while others are more specifically oriented
towards the abundance of individual species or groups
of species (Ståhl et al. 2011).
The analyses in this study are based on a selection of
the data set from the first (2003–2007) and second
452 Page 4 of 15
Fig. 1 The map shows the location of stratum 10 (gray), i.e., the
Swedish mountain region, and the 145 systematically placed
5 × 5 km sample units (red). The 12 circular sample plots is in
the center of the sample unit with a distance of 250 m between the
centers of each plot and a distance of 2125 m between the centers
of each plot and the edge of the 5 × 5 km2. Each plot consists of
two concentric circular plots, and different variables are recorded
(2008–2012) 5-year NILS inventory phases for the
mountain region (stratum 10), i.e., each plot is re-
measured every 5 years, thus plots inventoried in 2003
were re-inventoried in 2008 and plots inventoried in
2004 were re-inventoried in 2009 etc. See Ståhl et al.
(2011) and Christensen and Hedström Ringvall (2013)
and the field manual by Esseen et al. (2007) for more
details on the NILS program.
Preparation and organization of the data variables
First, a total of 1917 field inventory plots and/or
subplots from the initial inventory phase and 1919
plots and/or subplots from the re-inventory phase
were classified in the field into one of the following
four main land-cover types: ‘terrestrial/semiaquatic’,
‘aquatic’, ‘glacier’, and ‘permanently snow-covered
Environ Monit Assess (2016) 188: 452
in these plots. Canopy cover is estimated in the 20-m radius plot,
and shrub cover, cover of field vegetation, and cover of the bottom
layer are estimated in the 10-m radius plot. The sample plot is
divided into subplots if the sample plot contains distinct areas of
different types of land use or land cover etc. (Esseen et al. 2007).
The figure also shows a list of the field variables used in this study
sorted size of the circular sample plots
land’. Each plot and/or subplot, henceforth collec-
tively termed ‘plots’, was treated as a separate unit
in the analyses. There were 1613 plots in the classes
‘terrestrial/semiaquatic’, ‘glacier’, and ‘permanently
snow-covered land’, and these were grouped as ter-
restrial plots.
In the second step, the terrestrial plots were clas-
sified as ‘alpine’, ‘mountain birch forest’, or ‘other’
(Fig. 2). In this study, we refer to these classes as
‘habitat types’. The alpine area was defined as the
area above the alpine forest line, defined as the
border for trees higher than 2 m with a canopy cover
of more than 10 %. This definition corresponds to
the border for alpine vegetation zone as defined by
Sjörs (1967, 1999) and Ahti et al. (1968), which
thus includes several different vegetation types such
as heaths, meadows, wetlands, substrate ground (barren
Environ Monit Assess (2016) 188: 452
Fig. 2 Photographs of the alpine
and mountain birch forest
habitats. Photographs were taken
by NILS field staff
ground), and glaciers (Nordiska ministerrådet 1984;
Carlsson et al. 1999; Gardfjell and Hagner 2012)
within the alpine zone at high altitudes. The moun-
tain birch forest has traditionally been considered as
a part of the subalpine vegetation zone (Sjörs 1967)
but has more recently been considered as a part of
the northern boreal vegetation zone (Sjörs 1999).
The two dominating vegetation types in this zone
are heath and meadow forested habitats types dom-
inated by mountain birch (Betula pubescens ssp.
czerepanovii) forest (Nordiska ministerrådet 1984;
Carlsson et al. 1999). The mountain birch forest is
defined as a non-productive forest woodland (i.e.,
forest with a forest stock growth less than 1 m3 per
year and hectare and where the trees have reached
2 m in height and have a canopy cover exceeding
10 %). According to the definition applied in the
NILS program (Esseen et al. 2007), occasional
Page 5 of 15 452
predominantly shrub-formed trees or stumps of co-
niferous trees (Picea abies and Pinus sylvestris)
must be very few (at least 50 m between individual
trees and stumps).
The degree of vegetation cover (in percent) is
recorded with 1 % resolution based on a visual
estimate, which was selected as standard instead of
estimates into classes. Two cover estimations were
used in the NILS system—diffuse cover and strict
cover. Diffuse cover estimations include all parts
of, for example, a tree within the outer perimeter
of the tree canopy. Strict cover considers the veg-
etation according to a strict vertical projection, i.e.,
sections within, for example, a tree’s canopy area
that are not covered by flowers, leaves, branches,
or trunk are not included. The degree of diffuse
cover is therefore higher than the degree of strict
cover. NILS provided training for field staff in
452 Page 6 of 15
estimating cover in order to reduce the variation
between observers. This was done both in pre-
inventory training with practical exercises and
computer calibration programs with true cover cal-
culated from digital images (Gallegos Torell and
Glimskär 2009) and during the field inventory
with a more experienced person in the field team
together with a less experienced person.
The total canopy cover of living trees, regard-
less of height, was visually estimated as diffuse
cover within the 20-m radius circular sample plot.
In addition, the canopy cover of each tree species,
i.e., birches (Betula spp. mainly Betula pubescens
subsp. czerepanovii), all other deciduous trees, Norway
spruce (Picea abies), Scots pine (Pinus sylvestris), and
all other conifers, was estimated separately.
The total cover of living shrubs, regardless of height,
was visually estimated as diffuse cover within the 10-m
radius circular sample plot. In addition, the cover of
each shrub species (i.e., dwarf birch (Betula nana),
juniper (Juniperus communis), willows (Salix spp.),
and other shrub species) was estimated separately.
Here, we only present data for shrubs for the
2008–2012 re-inventory phase because the stan-
dard to estimate diffuse cover also on shrubs was
established after the 2003–2007 phase when strict
cover was used.
The total cover of the field vegetation was visu-
ally estimated as strict cover of all living herbs,
dwarf shrubs (Ericaceous species), graminoids,
ferns, and fern allies (i.e., Equisetum spp. and spe-
cies belonging to the class Lycopodiopsida) within
the 10-m radius circular sample plot (Esseen et al.
2007). In addition, the cover of the above species
groups was estimated separately for each group.
Data on field vegetation were lacking for 42 plots
in 2003. To allow a complete data set we incorpo-
rated the data from the same plots in 2008, thus
creating proxy data indicating a no-change plot.
The species Marsh Labrador tea (Rhododendron
tomentosum) was noted as a shrub in 2003 but
was included in the field vegetation as a dwarf
shrub in the NILS monitoring standard from 2004
and onwards. For the 2003 data, the cover of this
species was added both to the total cover of the
field vegetation and to the cover of dwarf shrubs
prior to the analysis.
The cover of five bottom layer species groups and
four substrate groups is in NILS estimated separately
Environ Monit Assess (2016) 188: 452
within a 10-m radius circular sample plot (Esseen et al.
2007). (1) ‘Foliose lichens’ included both foliose lichens
on the ground and those on stones, boulders, and bed-
rock. (2) ‘Reindeer lichens’ consisted of Cladonia
arbuscula, Cladonia rangiferina, Cladonia stellaris,
and Cladonia stygia. (3) ‘Other fruticose lichens’ in-
cluded all other forest-floor fruticose lichens except the
reindeer lichens. (4) ‘Sphagnum spp.’ consisted of all
Sphagnum species. (5) ‘Other bryophytes’ consisted of
all bryophytes except the Sphagnum spp. (6) ‘Humus
and peat’ included exposed humus or peat surfaces. (7)
‘Mineral soil and gravel’ included exposed mineral soil,
sand, or gravel with grain size ≤20 mm. (8) ‘Stones,
boulders, and flat rocks’ (grain size >20 mm) included
surfaces with no humus or plant cover, but crustose
lichens might occur. (9) ‘Other’ included water, litter,
branches, and logs as well as ground that deters plant
growth, e.g., asphalt, gravel, or concrete.
Cover of the above-mentioned species or groups
of species was estimated separately for each plot.
If trees, shrubs, or any of the components of the
field vegetation were present in small quantities,
i.e., with a maximum cover of 0.4 %, the total
vegetation cover, the total cover of shrubs, and the
total cover of the field vegetation, respectively,
were always noted separately as 0.2 %. Similarly
for each individual species and substrate group, if
a group was present in small quantities, defined as
with a maximum cover of 0.4 %, the cover value
was always noted as 0.2 %. The five bottom layer
groups were noted as 0 % if they were absent but
also if their observed cover did not exceed 0.4 %.
Vegetation data were not recorded in the field
inventory for plots on glaciers and permanently
snow-covered ground. For such plots, vegetation
cover was set to 0 % prior to analysis. Similarly,
tree and shrub covers were set to 0 % after a
visual control in aerial photos for plots located
on too steep or impassable terrain.
The taxonomy follows DYNTAXA (2015).
Analyses
We assessed changes based on the two 5-year inventory
phases in vegetation cover and composition in the alpine
area and mountain birch forest separately. These two
habitat types were found only in stratum 10 and not in
other NILS strata.
Environ Monit Assess (2016) 188: 452
Only plots classified as terrestrial (including the
NILS main habitat types of terrestrial/semiaquatic,
glacier, and permanently snow-covered land) were
included in the analyses. Thus, water surfaces
were not included. As a consequence, estimates
of the cover of the vegetation in a certain habitat
type in the mountain were calculated as the mean
cover of the vegetation on the terrestrial area in
that habitat type.
The total terrestrial areas of the alpine vegetation
and mountain birch forest were obtained separately
in two steps. First, the Horvitz–Thompson estimator
was used to estimate the mean ratios, i.e., the ratio
between the area of the focal mountain habitat type
in the sample units (i.e., the sample size was 145
separate 5 × 5 km2 in the mountain region) and the
sample unit area with its variance (Horvitz and
Thompson 1952; Särndal et al. 2003; Christensen
and Ringvall 2013). The Horvitz–Thompson estima-
tor is widely used to analyze design-based monitor-
ing programs (e.g., Magnussen et al. 2007; Bafetta
et al. 2011; Fridman et al. 2014). Second, the total
terrestrial areas for alpine vegetation and mountain
birch forest were obtained by multiplying the esti-
mated ratios with the total land area in stratum 10.
Similarly, Horvitz–Thompson estimations were used
to estimate the cover of trees, shrubs, and plants as
well as characteristics in the bottom layer. The var-
iances of the estimations are presented in the tables
as a standard error (SE) and a relative standard error
(RSE). The estimations were performed using the R
statistical package (version 2.15.1) and the svyratio
function in the Survey package (version 3.29;
Lumley 2004, 2010).
The Horvitz–Thompson estimator was also used
to estimate the change in areal extension of the
two focal habitat types, as well as the change
between the inventory phases in cover of trees
and plants and the characteristics of the bottom
layer. The magnitude of the change was deter-
mined as above with the svyratio feature in the
Survey package, which calculates the change as
the difference between the mean value of vegeta-
tion coverage from the first inventory and the
mean value of vegetation coverage from the re-
inventory. The variance was estimated as the sum
of the respective estimation variance −2 multiplied
by the covariance using the svycontrast function in
the Survey package. Based on the variance, a
Page 7 of 15 452
95 % confidence interval was calculated. Differ-
ences were, thus, considered significant at a 5 %
error level.
Results
The areal extension of the alpine area was estimat-
ed to be 3.19 Mha for both inventory phases and
to be 1.09 and 1.10 Mha for the mountain birch
forest for the first and the second inventory
phases, respectively. No significant changes in area
estimates could be detected between the 5-year
intervals. The standard errors of the estimates were
±0.34 Mha for the alpine area for both inventory
phases and ±0.17 Mha for the mountain birch
forest for both inventory phases.
The total canopy cover in the alpine area in-
creased significantly between the initial inventory
and the re-inventory from 0.55 ± 0.16 % to
0.71 ± 0.19 % (Fig. 3). The total canopy cover in
the mountain birch forest showed no significant
change, ranging from 37.28 ± 2.13 % to
38.25 ± 1.73 %. Birch (mainly Betula pubescens
subsp. czerepanovii) was the most abundant tree
species in both the alpine area and in the mountain
birch forest. There was no significant change be-
tween the initial inventory and re-inventory in can-
opy cover of individual tree species in either the
alpine area or the mountain birch forest (Fig. 4).
The total shrub cover was 17.9 ± 2.4 % in the alpine
area and 24.7 ± 2.8 % in the mountain birch forest
(Fig. 3). Dwarf birch was the most abundant shrub
species in both the alpine area and in the mountain birch
forest (Table 1).
There was a significant increase in field vege-
tation cover between the initial inventory and the
re-inventory for both the alpine area and the
mountain birch forest (Fig. 3). The total cover of
the field vegetation increased from 35.98 ± 2.22 %
to 43.30 ± 2.71 % in the alpine area and from
47.97 ± 2.56 % to 53.57 ± 2.40 % in the moun-
tain birch forest. Dwarf shrubs dominated the field
vegetation in both the alpine area and the moun-
tain birch forest (Fig. 5), and both graminoids and
dwarf shrubs increased significantly in the alpine
area and mountain birch forest.
Bryophytes dominated the bottom layer in the two
mountain habitat types (Table 2). Bryophytes and
452 Page 8 of 15
(a) Alpine area
(b) Mountain birch forest
Fig. 3 The estimated total canopy cover, total cover of shrubs, and
total cover of field vegetation, during the initial inventory (2003–
2007) and the re-inventory (2008–2012) in the alpine area (a) and
in the mountain birch forest (b). We only present data for shrubs
for the 2008–2012 re-inventory phase because the methods esti-
mating cover of shrubs was changed between the inventory
humus and peat decreased significantly in the alpine
area, while foliose lichens decreased significantly in
the mountain birch forest.
Discussion
Until now, most evidence on changes in vegetation
cover and composition in the Swedish mountains
has been based on re-inventories of old vegetation
studies that are representative for a specific loca-
tion or only cover a limited geographical range
(e.g., Moen and Lagerström 2008; Van Bogaert
et al. 2011; Wilson and Nilsson 2009; Kullman
Environ Monit Assess (2016) 188: 452
(c) Alpine area
(d) Mountain birch forest
phases. The estimates of cover are presented with a 95 % confi-
dence interval. (c) shows the change in cover in the alpine area and
(d) shows the change in cover in the mountain birch forest between
the two inventory phases. The estimates of the change of cover are
presented with a 95 % confidence interval. Note the different
scales on the axes
2010; Virtanen et al. 2010; Hedenås et al. 2011, 2012;
Rundqvist et al. 2011; Wehn and Olsson 2015). This
study, which is based on a regional set of a systemati-
cally sampled national monitoring data, is the first com-
plete compilation of the areal extent and vegetation
status—as well as changes in these aspects—in the
alpine area and mountain birch forest on the Swedish
side of the Scandinavian Mountain Range. The results
presented here can be used as a baseline for vegetation
in the Scandinavian mountain range, towards which
observed vegetation changes in specific studies can be
compared, as input data in forecast models based on
anticipated change, as background for land-use and
environmental policy development, and also to
Environ Monit Assess (2016) 188: 452
(a) Alpine area
(b) Mountain birch forest
Fig. 4 The estimated canopy cover of birch (Betula spp., mainly
Betula pubescens subsp. czerepanovii), other deciduous trees
(other dec), pine (Pinus sylvestris), spruce (Picea abies), and other
coniferous trees (other con) during the initial inventory (2003–
2007) and the re-inventory (2008–2012) in the alpine area (a) and
in the mountain birch forest (b). The estimates of canopy cover are
generate hypothesis in directed cause and effect
oriented studies.
Areal extension of alpine vegetation and mountain birch
forest
We found that the areal extension of these two
habitat types has not changed significantly be-
tween the two inventory phases, which might be
expected with only a 5-year time span. Previous
studies that have detected changes in the areal
distribution of mountain birch all covered longer
time spans (Kullman 2002; Van Bogaert et al.
2011). Our measure of the areal extension of the
alpine area in Sweden was 3.2 Mha, which
Page 9 of 15 452
(c) Alpine area
(d) Mountain birch forest
presented with a 95 % confidence interval. (c) shows the change in
canopy cover in the alpine area and (d) shows the change in
canopy cover in the mountain birch forest between the two inven-
tory phases. The estimates of the change of canopy cover are
presented with a 95 % confidence interval. Note the different
scales on the axes
corresponds well with the 3.3 Mha of alpine vegetation
reported by Carlsson et al. (1999). However, our
estimated area of mountain birch forest (1.1 Mha)
was about half as large as the mountain birch
forest area presented by Carlsson et al. (1999).
Our estimated area of mountain birch was also
lower than the estimated 1.4 Mha of mountain
birch forest reported as part of the 2013 Natura
2000 annex I habitat types article 11 report (Eionet
2014; Eide 2014). The difference between our
estimated area of mountain birch forest and the
area reported in the other studies is probably ex-
plained by the different definitions of mountain
birch forest. The definitions applied in the other
studies allow a higher occurrence of occasional
452 Page 10 of 15 Environ Monit Assess (2016) 188: 452

Table 1 Total cover of shrubs and cover of shrub species in the (although there is a strong trend) over the 5-year time
alpine area and in the mountain birch forest
span reported here. The discrepancy between the esti-
Estimation of the cover mates showing an increase in the total tree crown cover,
Inventory 2008–2012 but not in the crown cover of mountain birch, might be
explained by the fact that these estimates are based on
Cover SE RSEa
Shrub species % % % data collected separately from each other in the field
inventory. Thus, it is possible that this is a method-
Alpine ological artifact in the monitoring scheme that will
Betula nana 12.7 1.9 15.0 be adjusted for over time as a longer time series of
Juniperus communis 1.2 0.3 23.4 data becomes available. This finding implies, how-
Salix spp. 3.9 0.6 15.1 ever, that care has to be taken before far-reaching
Other shrub species 0 0 - conclusions are made. Our results indicate, though,
Total shrub cover 17.9 2.4 13.6 that the pattern of increasing canopy cover in the
Mountain birch forest alpine area and in the mountain birch forest that
Betula nana 10.4 2.1 20.0 have been observed in local studies covering longer
Juniperus communis 7.8 1.1 14.6 time spans might be indicative of a general pattern
Salix spp. 6.5 1.2 17.8 over the whole Swedish mountain range.
Other shrub species <0.1 <0.1 97.7
Total shrub cover 24.7 2.8 11.2 Changes in coverage and composition of the shrub,
field, and bottom layer vegetation
a
Relative standard error (RSE)
Further, we observed an increase in the cover of
conifers (Carlsson et al. 1999; Gardfjell and Hagner graminoids and dwarf shrubs as well as an increase
2012) than the NILS definition (Esseen et al. 2007) that in the total cover of the field vegetation both in the
was used here. alpine area and the mountain birch forest, while the
cover of mosses and lichens decreased. This is in
Changes in vegetation coverage and composition congruence with previously reported warming ex-
periments that suggest that graminoids and dwarf
Changes in coverage and composition of the tree layer shrubs might increase with increasing temperatures
while lichens and bryophytes might decrease
We showed that the coverage and composition of both (Walker et al. 2006). Increased cover of graminoids
the tree and the field vegetation have changed for both might also be due to increasing reindeer grazing
the alpine area and the mountain birch forest. We ob- pressure (cf. Olofsson et al. 2001). However, such
served an overall increase in the total canopy cover in an interpretation is unlikely on the scale of the
the alpine area, which is in congruence with several results presented here for two reasons. First, the
observations over the last few decades (Kullman 2002; number of reindeer over the whole mountain region
Van Bogaert et al. 2011; Hedenås et al. 2011; Rundqvist fluctuates around a relatively constant number, al-
et al. 2011). The increase in canopy cover might be though there might be local changes in population
attributed to a combination of factors such as a delayed size (Moen and Danell 2003). Second, as an effect
re-expansion of trees following the BLittle Ice Age^ that of grazing, we should also have observed a decline
ended in the early twentieth century (see discussion in in dwarf shrubs along with the increase in
Rundqvist et al. 2011), stimulated growth and regener- graminoids (cf. Olofsson et al. 2001). Instead, we
ation of birch following recent climate warming also observed an increase in dwarf shrub cover.
(Callaghan et al. 2013), and/or a continued long-term
recovery from former land use such as large outtakes of Changing status of mountain vegetation: importance
firewood and grazing by livestock and reindeer (Östlund of long-term monitoring
et al. 2015; Sandström 2015). Our results show that the
cover of mountain birch, the absolutely dominating tree Long-term monitoring of essential functional compo-
species in the alpine area, has not increased significantly nents of ecosystems is important for detecting and
Environ Monit Assess (2016) 188: 452
(a) Alpine area
(c) Mountain birch forest
Fig. 5 The estimated cover of graminoids, herbs, dwarf shrubs,
and ferns and fern allies (Ferns) during the initial inventory (2003–
2007) and the re-inventory (2008–2012) in the alpine area (a) and
mountain birch forest (b). The estimates of the cover of field
vegetation are presented with a 95 % confidence interval. The
documenting the changing status of ecosystems (Lovett
et al. 2007; Keith et al. 2013). If the here observed
tendency of increasing tree crown cover in the alpine
area and changes in cover in functional groups in the
field layer becomes more and more pronounced, this
might have profound effects on ecosystem functions,
services, and biodiversity in mountain ecosystems
(Chapin et al. 2005; Keith et al. 2013). For example,
tree recruitment and growth might affect the diversity of
vascular plants and cryptogams (Walker et al. 2006;
Batllori et al. 2009), facilitate the thawing of permafrost
(Johansson et al. 2006), and influence the regulation of
CO2 (Heliasz et al. 2011), reindeer herding (Wielgolaski
2005), and tourism. Since tourists preferable are
attracted to extensive open alpine areas with vast areas
to overlook rather than by dense mountain birch forests
(Bäck 2002; Heberlein et al. 2002), more and denser
Page 11 of 15 452
(b) Alpine area
(d) Mountain birch forest
cover of graminoids and dwarf shrubs have increased significantly,
at the 5 % error level, both in the alpine area (c) and in the
mountain birch forest (d) between the two inventory phases. The
estimates of the change of canopy cover are presented with a 95 %
confidence interval. Note the different scales on the axes
vegetation will affect the amenity values of the
mountain landscape. Further, changes in functional
groups in the field layer might indicate a shift in
grazing pressure (Olofsson et al. 2001), land use
(Tasser and Tappeiner 2009), temperature (Walker
et al. 2006), or snow cover (Olofsson et al. 2011).
Changes in the composition of the bottom layer
might also change with climate, land use, and grazing
pressure (Walker et al. 2006; Callaghan et al. 2013).
We thus stress that the change in coverage and
composition of the tree, shrub, field, and bottom layers
are important attributes to monitor over the long term
when assessing the change in status of mountain
ecosystems. Christensen and Ringvall (2013) concluded
in their assessment of the statistical properties of the
NILS program that with the exception of stochastic
and cyclic events (Franklin 1989) and observation errors
452 Page 12 of 15 Environ Monit Assess (2016) 188: 452

Table 2 The cover of species and substrate groups in the bottom layer in the alpine area and in the mountain birch forest

Estimation of the cover Estimate of the differences

Inventory 2003–2007 Re-inventory 2008–2012 95 % CIc

Cover SE RSEb Cover SE RSEb Diff. Lower Upper

Species/Substratea % % % % % % % % %

Alpine
Foliose lichens 2.06 0.42 20.53 1.74 0.30 17.09 −0.31 −0.87 0.24
Reindeer lichens 3.24 0.50 15.35 4.03 0.53 13.20 0.79 −0.19 1.77
Other fruticose lichens 6.74 1.08 16.01 6.61 1.14 17.18 −0.13 −2.06 1.80
Sphagnum spp. 5.34 1.42 26.62 4.81 1.18 24.60 −0.53 −1.33 0.27
Other bryophytes 36.61 1.79 4.88 33.33 1.72 5.15 −3.28 −6.39 −0.17
Humus and peat 5.21 1.04 19.96 2.23 0.65 29.07 −2.98 −5.34 −0.63
Mineral soil and gravel 3.35 0.67 20.04 4.20 0.86 20.40 0.85 −0.74 2.44
Stones, boulders, and flat rocks 19.01 2.33 12.26 18.54 2.43 13.13 −0.47 −2.08 1.15
Other 18.66 1.95 10.44 22.13 2.06 9.30 3.47 −1.18 8.13
Mountain birch forest
Foliose lichens 1.72 0.32 18.42 1.02 0.18 17.96 −0.70 −1.26 −0.13
Reindeer lichens 2.41 0.51 21.31 2.82 0.79 28.06 0.41 −1.04 1.85
Other fruticose lichens 2.19 0.60 27.35 1.86 0.53 28.30 −0.33 −1.42 0.75
Sphagnum spp. 8.46 1.84 21.76 8.35 1.74 20.78 −0.11 −1.95 1.74
Other bryophytes 58.39 2.46 4.21 53.99 3.38 6.26 −4.40 −8.92 0.13
Humus and peat 0.77 0.25 31.77 0.44 0.12 28.31 −0.34 −0.84 0.17
Mineral soil and gravel 0.21 0.10 47.71 0.15 0.08 55.73 −0.06 −0.20 0.09
Stones, boulders, and flat rocks 5.22 1.48 28.46 4.99 1.33 26.57 −0.23 −1.25 0.79
Other 20.93 2.44 11.67 25.03 3.12 12.48 4.10 −1.92 10.11

The change in cover was significant, at the 5 % error level, when zero was not included in the confidence interval. Italic text denotes
significant changes
a
Foliose lichens consist of foliose lichens on the ground and on stones, boulders, and bedrock. Reindeer lichens consist of the lichens
Cladonia arbuscula, Cl. rangiferina, Cl. stellaris, and Cl. stygia. Other fruticose lichens consist of all other fruticose lichens on the ground
surface except the reindeer lichens. Sphagnum spp. consists of all Sphagnum species. Other bryophytes consist of all bryophytes except
Sphagnum spp. Humus and peat are exposed humus or peat. Mineral soil and gravel include exposed mineral soil, sand, or gravel with a
grain size ≤20 mm. Stones, boulders, and flat rocks include material with a grain size >20 mm with no humus or plant cover, but crustose
lichens might occur. Other includes water, litter, dense field layer components, branches, logs, etc.
b
Relative standard error (RSE)
c
Confidence interval (CI), Lower = lower confidence interval level and Upper = upper confidence interval level

(Milberg et al. 2008), even quite small changes in com-
mon variables such as total coverage of trees, shrubs,
and field vegetation are detectable through a systematic
sample of relatively few sample units that together is
representative for a larger geographic scales. In NILS, a
total of 145 sample units represent the mountain range.
However, to detect changes of less common variables
such as the range of aspen expansion in alpine areas
(cf. Van Bogaert et al. 2010; Rundqvist et al. 2011),
a higher sampling intensity or directed studies
would be necessary (cf. Christensen and Ringvall
2013). Monitoring should be sensitive to the societal
demand for environmental data and thus harbor
adaptive capacity, including both altered standards
as in the case of switching from strict to diffuse
cover of shrubs in NILS and adoption of new vari-
ables and methods. The NILS program, which is the
only national environmental monitoring program
with a full coverage of the Swedish side of the
Scandinavian mountain range, provide solid data
for generic biophysical data such as cover of main
species and species groups. For a more focused use
Environ Monit Assess (2016) 188: 452
in cause and effect analyses for climate change early
warning assessments, for example, the NILS data
need to be supported by directed research studies.
We foresee that this study will generate such future
opportunities for the Scandinavian mountain range.
Conclusions
Despite a relatively short-time span of 5 years, we found
large-scale changes in vegetation cover and composition
in the Swedish mountain region. The total canopy cover
increased in the alpine area. The cover of graminoids
and dwarf shrubs and the total cover of field vegetation
increased in both the alpine area and the mountain birch
forest, while lichens and bryophytes decreased. There
were, however, no changes in the areal extension of
either the alpine area or the mountain birch forest.
Although there are limitations in determining
the underlying causes of observed vegetation
changes, national and regional monitoring schemes
can confirm or disprove the results of local studies
and experimental studies or models. Through
benchmark assessments, monitoring also contrib-
utes to the evaluation of government policies and
strategic land-management frameworks.
Acknowledgments We would like to acknowledge the Swedish
Environmental Protection Agency and the county board of admin-
istration in Dalarna, Jämtland, Västerbotten, and Norrbotten both
for funding this study and for fruitful discussions. Thanks to the
NILS field staff for their assistance. We also thanks M. Hogg at the
Semantix for improving the language.
References
ACIA (2005). Arctic climate impact assessment. Cambridge:
Cambridge University Press.
Ahti, T., Hämet-Ahti, L., & Jalas, J. (1968). Vegetation zones and
their sections in northwestern Europe. Annales Botanici
Fennici, 5(3), 169–211.
Arft, A. M., Walker, M. D., Gurevitch, J., Alatalo, J. M., Bret-
Harte, M. S., Dale, M., et al. (1999). Responses of tundra
plants to experimental warming: meta-analysis of the inter-
national tundra experiment. Ecological Monographs, 69(4),
491–511.
Bäck, L. (2002). Friluftslivet i de svenska Lapplandsfjällen.
[Outdoor life in the Swedish Lapland mountains]. Utmark,
3. [In Swedish.]
Bafetta, F., Fattorini, L., & Corona, P. (2011). Estimation of small
woodlot and tree row attributes in large-scale forest
Page 13 of 15 452
inventories. Environmental and Ecological Statistics, 18(1),
147–167.
Batllori, E., Blanco-Moreno, J. M., Ninot, J. M., Gutiérrez, E., &
Carrillo, E. (2009). Vegetation patterns at the alpine treeline
ecotone: the influence of tree cover on abrupt change in
species composition of alpine communities. Journal of
Vegetation Science, 20(5), 814–825.
Beniston, M. (2003). Climatic change in mountains regions: a
review of possible impacts. Climatic Change, 59(1–2), 5–31.
Bokhorst, S., Bjerke, J. W., Tømmervik, H., Callaghan, T. V., &
Phoenix, G. K. (2009). Winter warming events damage sub-
Arctic vegetation: consistent evidence from an experimental
manipulation and a natural event. Journal of Ecology, 97(6),
1408–1415.
Bryn, A., & Daugstad, K. (2001). Summer farming in the subal-
pine birch forest. In: F. E. Wielgolaski. (Ed.), Nordic moun-
tain birch ecosystems. Man and Biosphere, vol. 27 (pp. 307–
316). New York: Parthenon Publishing group.
Callaghan, T. V., Jonasson, C., Thierfelder, T., Yang, Z.,
Hedenås, H., Johansson, M., et al. (2013). Ecosystem
change and stability over multiple decades in the
Swedish subarctic: complex processes and multiple
drivers. Philosophical Transactions of the Royal
Society B, 368, e20120488.
Carins, D. M., & Moen, J. (2004). Herbivory influences tree lines.
Journal of Ecology, 92(6), 1019–1024.
Carlsson, B., Karlsson, P. S., & Svensson, B. M. (1999). Alpine
and subalpine vegetation. Acta Phytogeographica Suecica,
84, 75–89.
Chapin III, F. S., Berman, M., Callaghan, T. V., Convey, P., Crepin,
A. S., Danell, K., et al. (2005). Polar systems. In R. Hassan,
R. Scholes, & N. Ash (Eds.), Ecosystems and human well-
being: current state and trends, vol. 1 (pp. 717–743).
Washington: Island Press.
Christensen, P., & Ringvall, A. H. (2013). Using statistical power
analysis as a tool when designing a monitoring program:
experience from a large-scale Swedish landscape monitoring
program. Environmental Monitoring and Assessment,
185(9), 7279–7293.
Chytrý, M., Tichý, L., Hennekens, S. M., & Schaminée, J. H. J.
(2014). Assessing vegetation change using vegetation-plot
databases: a risky business. Applied Vegetation Science,
17(1), 32–41.
DYNTAXA 2015. Taxonomic Database. The Swedish Species
Information Centre, SLU, Uppsala. [In Swedish.]
h t t p : / / w w w. s l u . s e / s v / c e n t r u m b i l d n i n g a r - o c h -
projekt/dyntaxa/search/. Accessed May 2015.
Edenius, L., Vencatasawmy, C. P., Sandström, P., & Dahlberg, U.
(2003). Combining satellite imagery and ancillary data to
map snowbed vegetation important to reindeer Rangifer
Tarandus. Arctic, Antarctic, and Alpine Research, 35(2),
150–157.
Eide, W. (ed.) (2014). Arter och naturtyper i habitatdirektivet –
bevarandestatus i Sverige 2013. [Species and habitats in the
Habitats Directive - conservation status in Sweden in 2013].
Uppsala, SE: The Swedish Species Information Centre, SLU.
Eionet 2014. http://cdr.eionet.europa.eu/Converters/run_
conversion?file=se/eu/art17/envucgywg/SE_habitats_
reports-131223-141035.xml&conv=350&source=
remote#9040. Accessed 8 August 2014.
452 Page 14 of 15
Emanuelsson, U. (1984). Ecological effects of grazing and tram-
pling on mountain vegetation in northern Sweden. PhD
Thesis. Lund, SE: University of Lund.
Emanuelsson, U. (1987). Human influence on vegetation in the
Torneträsk area during the last three centuries. Ecological
Bulletins, 30, 95–111.
Esseen, P.-A., Glimskär, A., Ståhl, G., & Sundquist, S. (2007).
Field instruction for the national inventory of the landscape
in Sweden. Umeå: Department of Forest Resource
Management, Swedish University of Agricultural Sciences,
http://www.slu.se/Documents/externwebben/s-fak/skoglig-
resurshallning/Landskapsanalys_
publikationer/Publikationer2003-2009/NILS_manual_f%C3
%A4lt_2007_English.pdf
Field, C. B., Barros, V. R., Dokken, D. J., Mach, K. J.,
Mastrandrea, M. D., Bilir, T. E., et al. (2014). Climate
change 2014. Impacts, adaptation, and vulnerability.
Part a: Global and sectoral aspects. Contribution of
working group II to the fifth assessment report of the
intergovernmental panel on climate change. Cambridge:
Cambridge University Press.
Franklin, J. F. (1989). Importance and justification of long-term
studies in ecology. In G. E. Likens (Ed.), Long-term studies
in ecology (pp. 3–19). NY: Springer.
Fridman, J., Holm, S., Nilsson, M., Nilsson, P. Hedström Ringvall,
A., & Ståhl, G. (2014). Adapting National Forest Inventories
to changing requirements—the case of the Swedish National
Forest Inventory at the turn of the twentieth century. Silva
Fennica 48(3), doi: 10.14214/sf.1095
Gallegos Torell, A., & Glimskär, A. (2009). Computer-aided cal-
ibration for visual estimation of vegetation cover. Journal of
Vegetation Science, 20(6), 973–983.
G a r d f j e l l H . & Å . H a g n e r ( 2 0 1 2 ) . I n s t r u k ti o n f ö r
Habitatinventering i NILS and MOTH, 2012. [Manual for
habitat monitoring, in NILS and MOTH 2012.]. Umeå SE.
Department of Forest Resource Management, Swedish
University of Agricultural Sciences, [In Swedish.]
Government Offices of Sweden (2004). Environmental quality
objectives: A shared responsibility. Summary of government
bill 2004/05:150. Stockholm, SE: Government Offices of
Sweden,
Heberlein, T. A., Fredman, P., & Vuorio, T. (2002). Current
tourism patterns in the Swedish mountain region. Mountain
Research and Development, 22(2), 142–149.
Hedenås, H., Olsson, H., Jonasson, C., Bergstedt, J., Dahlberg, U.,
& Callaghan, T. V. (2011). Changes in tree growth, biomass
and vegetation over a 13-year period in the Swedish sub-
Arctic. Ambio, 40(S1), 672–682.
Hedenås, H., Carlsson, B. Å., Emanuelsson, U., Headley, A. D.,
Jonasson, C., Svensson, B. M., & Callaghan, T. V. (2012).
Changes versus homeostasis in alpine and sub-alpine vege-
tation over three decades in the sub-Arctic. Ambio, 41(S3),
187–196.
Heliasz, M., Johansson, T., Lindroth, A., Molder, M., Mastepanov,
M., Friborg, T., et al. (2011). Quantification of C uptake in
subarctic birch forest after setback by an extreme insect
outbreak. Geophysical Research Letters, 38, L01704.
Hörnfeldt, B. (2004). Long-term decline in numbers of cyclic
voles in boreal Sweden: Analysis and presentation of hypoth-
eses. Oikos, 107(2), 376–392.
Environ Monit Assess (2016) 188: 452
Horvitz, D. G., & Thompson, D. J. (1952). A generalization of
sampling without replacement from a finite universe. Journal
of the American Statistical Association, 47(260), 663–685.
Ims, R. A., Henden, J. A., & Killengreen, S. T. (2008).
Collapsing population cycles. Trends in Ecology &
Evolution, 23(2), 79–86.
Jepsen, J. U., Hagen, S. B., Ims, R. A., & Yoccoz, N. G. (2008).
Climate change and outbreaks of the geometrids
Operophtera brumata and Epirrita autumnata in subarctic
birch forest: Evidence of a recent outbreak range expansion.
Journal of Animal Ecology, 77(2), 257–264.
Johansson, M., Christensen, T. R., Akerman, H. J., & Callaghan,
T. V. (2006) What determines the current presence or absence
of permafrost in the Torneträsk region, a subarctic landscape
in northern Sweden?. Ambio 35(4), 190–197.
Kaplan, J. O., & New, M. (2006). Arctic climate change with a
2 °C global warming: Timing, climate patterns and vegeta-
tion change. Climatic Change, 79(3–4), 213–241.
Karlsson, H., Hörnberg, G., Hannon, G., & Nordström, E. M.
(2007). Long-term vegetation changes in the northern
Scandinavian forest limit: A human impact-climate synergy?
The Holocene, 17(1), 37–49.
Keith, D. A., Rodríguez, J. P., Rodríguez-Clark, K. M., Nicholson,
E., Aapala, K., Alonso, A., et al. (2013). Scientific founda-
tions for an IUCN red list of ecosystems. PloS One, 8(5),
e62111.
Körner, C., Ohsawa, M., Spehn, E., Berge, E., Bugmann, H.,
Groombridge, B., et al. (2005). Mountain systems. In R.
Hassan, R. Scholes, & A. N. (Eds.), Ecosystems and human
well-being: Current state and trends, vol. 1 (pp. 681–715).
Washington: Island Press.
Kullman, L. (1979). Change and stability in the altitude of the
birch tree-limit in the southern Swedish Scandes 1915-1975.
Acta Phytogeographica Suecica, 65, 94–121.
Kullman, L. (2002). Rapid recent range-margin rise of tree and
shrub species in the Swedish Scandes. Journal of Ecology,
90(1), 68–77.
Kullman, L. (2010). A richer, greener and smaller alpine world:
Review and projection of warming-induced plant cover
change in the Swedish Scandes. Ambio, 39(2), 159–169.
Lindenmayer, D. B., & Likens, G. E. (2010). The science and
application of ecological monitoring. Biological
Conservation, 143(6), 1317–1328.
Lovett, G. M., Burns, D. A., Driscoll, C. T., Jenkins, J. C.,
Mitchell, M. J., Rustad, L., et al. (2007). Who needs envi-
ronmental monitoring? Frontiers in Ecology and the
Environment, 5(5), 253–260.
Lumley T. (2004). Analysis of complex survey samples. Journal
of Statistical Software, 9(8).
Lumley T. (2010). Complex surveys. A guide to analysis using R.
Wiley, NJ.
Magnussen, S., Smith, B., & Sandoval Uribe, A. (2007). National
Forest Inventories in North America for monitoring forest
tree species diversity. Plant Biosystems, 141(1), 113–122.
Milberg, P., Bergstedt, J., Fridman, J., Odell, G., &
Westerberg, L. (2008). Observer bias and random var-
iation in vegetation monitoring data. Journal of
Vegetation Science, 19(5), 633–644.
Moen, J., & Danell, O. (2003). Reindeer in the Swedish
Mountains: An assessment of grazing impacts. Ambio,
32(6), 397–402.
Environ Monit Assess (2016) 188: 452
Moen, J., & Lagerström, A. (2008). High species turnover and
decreasing plant species richness on mountain summits in
Sweden: reindeer grazing overrides climate change.
Antarctic, and Alpine Research, 40(2), 382–395.
Moen, J., Aune, K., Edenius, L., & Angerbjörn, A. (2004).
Potential effects of climate change on treeline position in
the Swedish mountains. Ecology and Society, 9, 16.
Nordiska ministerrådet. (1984). Vegetationstyper i Norden. [Types
of vegetation in the Nordic countries.] Nordiska
ministerrådet. [In Swedish.]
Olofsson, J., Kitti, H., Rautiainen, P., Stark, S., & Oksanen, L.
(2001). Effects of summer grazing by reindeer on composi-
tion of vegetation, productivity and nitrogen cycling.
Ecography, 24(1), 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(11), 2681–2693.
Olofsson, J., Ericson, L., Torp, M., Stark, S., & Baxter, R. (2011).
Carbon balance of Arctic tundra under increased snow cover
mediated by a plant pathogen. Nature Climate Change, 1(4),
220–223.
Olsson E. G. A., Austrheim G., & Grenne S. N. (2000). Landscape
change patterns in mountains, land use and environmental
diversity, Mid-Norway 1960-1993. Landscape Ecology,
15(2), 155–170.
Östlund, L., Hörnberg, G., DeLuca, T. H., Liedgren, L., Wikström,
P., Zackrisson, O., & Josefsson, T. (2015). Intensive land use
in the Swedish mountains between AD 800 and 1200 led to
deforestation and ecosystem transformation with long-lasting
effects. Ambio, 44(6), 508–520.
Pauli, H., Gottfried, M., Dullinger, S., Abdaladze, O., Akhalkatsi,
M., Alonso, J. L. B., et al. (2012). Recent plant diversity
changes on Europe’s mountain summits. Science, 336(6079),
353–355.
Pearson, R. G., Phillips, S. J., Loranty, M. M., Beck, P. S. A.,
Damoulas, T., Knight, S. J., & Goetz, S. J. (2013). Shifts in
Arctic vegetation and associated feedbacks under climate
change. Nature Climate Change, 3(7), 673–677.
Rundqvist, S., Hedenås, H., Sandström, A., Emanuelsson, U.,
Eriksson, H., Jonasson, C., & Callaghan, T. V. (2011). Tree
and shrub expansion over the past 34 years at the tree-line
near Abisko, Sweden. Ambio, 40(S1), 683–692.
Sandström P. (2015). A toolbox for co-production of knowledge
and improved land use dialogues. The perspective of reindeer
husbandry. PhD Thesis. Umeå, SE: Swedish University of
Agricultural Sciences.
Särndal, C. E., Swensson, B., & Wretman, J. (2003). Modell
assisted survey sampling. NY: Springer-Verlag.
SEPA (2014). Förslag till en strategi för miljökvalitetsmålet
Storslagen fjällmiljö. Redovisning av ett regeringsuppdrag.
[Proposal for a strategy for the environmental quality objec-
tive BA magnificent mountain landscape^. Recognition of a
governmental commission.] Skrivelse 2014–06-05. NV 04–
173-13. Stockholm, SE: SEPA. [In Swedish.]
Sjörs H. (1967). Nordisk växtgeografi. [Nordic plant geography.]
2nd ed. Stockholm, SE: Svenska Bokförlaget, [In Swedish.]
Page 15 of 15 452
Sjörs, H. (1999). The background: Geology, climate, and zonation.
Acta Phytogeographica Suecica, 84, 5–14.
Speed, J. D. M., Austrheim, G., Hester, A. J., & Mysterud, A.
(2012). Elevational advance of alpine plant communities is
buffered by herbivory. Journal of Vegetation Science, 23(4),
617–625.
Ståhl, G., Allard, A., Esseen, P.-A., Glimskär, A., Ringvall,
A., Svensson, J., et al. (2011). National Inventory of
Landscapes in Sweden NILS-scope, design, and expe-
riences from establishing a multiscale biodiversity
monitoring system. Environmental Monitoring and
Assessment, 173(1–4), 579–595.
Tasser, E., & Tappeiner, U. (2009). Impact of land use
changes on mountain vegetation. Applied Vegetation
Science, 5(2), 173–184.
Theurillat, J.-P., & Guisan, A. (2001). Potential impact of climate
change on vegetation in the European alps: A review.
Climatic Change, 50(1–2), 77–109.
Van Bogaert, R., Jonasson, C., De Dapper, M., & Callaghan, T. V.
(2010). Range expansion of thermophilic aspen (Populus
tremula L.) in sub-Arctic Sweden. Arctic, Antarctic, and
Alpine Research, 42(3), 362–375.
Van Bogaert, R., Haneca, K., Hoogesteger, J., Jonasson, C., De
Dapper, M., & Callaghan, T. V. (2011). A century of tree line
changes in sub-Arctic Sweden shows local and regional
variability and only a minor influence of twentieth century
climate warming. Journal of Biogeography, 38(5), 907–921.
Virtanen, R., Luoto, M., Rämä, T., Mikkola, K., Hjort, J., Grytnes,
J.-A. & Birks H. J. B. (2010). Recent vegetation changes at
the high altitude tree line ecotone are controlled by geomor-
phological disturbance, productivity and diversity. Global
Ecology and Biogeography, 19(6), 810–821.
von Sydow, U. (1988). Gräns för storskaligt skogsbruk i fjällnära
skogar: Förslag till naturvårdsgräns. [Border for large-scale
forest management in montane forests: Proposal of a new
border for nature concern.] Stockholm, SE: Svenska
naturskyddsföreningen. [In Swedish.]
Walker, M. D., Wahren, C. H., Hollister, R. D., Henry, G. H. R.,
Ahlquist, L. E., Alatalo, J. M., et al. (2006). Plant community
response to experimental warming across the tundra biome.
Proceedings of the National Academy of Sciences, 103(5),
1342–1346.
Walther, G. R., Beissner, S., & Burga, C. A. (2005). Trends in the
upward shift of alpine plants. Journal of Vegetation Science,
16(5), 541–548.
When, S., & Olsson, W. G. A. (2015). Performance of the endemic
alpine herb Primula scandinavia in a changing European
mountain landscape. Annales Botanici Fennici, 52(3–4),
171–180.
Wielgolaski, F. E. (2005). Plant ecology herbivory, and human
impact in Nordic mountain birch forests. Berlin: Springer.
Wilson, S. D., & Nilsson, C. (2009). Arctic and alpine
vegetation change over years. Global Change Biology,
15(7), 1676–1684.
Wipf, S., Stockli, V., Herz, K., & Rixen, C. (2013). The oldest
monitoring site of the alps revisited: accelerated increase in
plant species richness on Piz Linard summit since 1835.
Plant Ecology and Diversity, 6(3–4), 447–455.