Bryophytes and lichens in different types of forest set-asides in ... - SLU

Forest Ecology and Management 242 (2007) 374–390
www.elsevier.com/locate/foreco
Bryophytes and lichens in different types of forest
set-asides in boreal Sweden
Karin Perhans a, *, Lena Gustafsson a , Fredrik Jonsson b , Ulrika Nordin b , Henrik Weibull c
a Department of Ecology, Swedish University of Agricultural Sciences, Box 7002, SE-750 07 Uppsala, Sweden
b Ede 1400, SE-830 47 Trångsviken, Sweden
c Naturcentrum AB, Strandtorget 3, SE-444 30 Stenungsund, Sweden
Received 8 June 2006; received in revised form 20 November 2006; accepted 21 January 2007
Abstract
Setting aside forest land is practiced globally and is of vital importance to the preservation of forest biodiversity. As detailed species inventories
rarely precede the establishment of set-asides, empirical studies on the species content of different set-aside types are needed, in order to evaluate
their conservation relevance and to design efficient conservation strategies. Here, we compared the biodiversity value per unit area of three types of
set-asides, common in boreal Europe: nature reserves, key habitats and retention groups on clear-cuts, and we also included old managed forests as
reference sites. We surveyed bryophytes on all substrates and lichens growing on spruce in spruce-dominated forests in middle boreal Sweden and
in total 252 bryophytes and 176 lichens were found. For bryophytes the number of species, as well as the number of red-listed species, per unit area
was higher in key habitats than in retention groups and old managed forests, while intermediate in nature reserves. As the complementarity between
areas was greatest in key habitats, the total number of bryophyte species found was by far the highest in this set-aside type. For lichens the patterns
were similar but the differences much less pronounced. Site selection methods based on linear programming demonstrated that key habitats also
played the most important role in representing the surveyed taxa effectively but that the value of nature reserves increased when multiple
representations of each species were required. The study shows that the richness and composition of bryophytes and lichens differ between the
three set-aside types and consequently, the outcome of large-scale conservation strategies based on anyone type would be different. At present, key
habitats stand out as core areas of high diversity in the forest landscape but it is important also to acknowledge the temporal dynamics of the forest
landscape and how the capacity of each set-aside type to host species can be expected to change over time.
# 2007 Elsevier B.V. All rights reserved.
Keywords: Nature reserve; Woodland key habitat; Retention group; Red-listed species; Indicator species; Conservation strategy
1. Introduction
The establishment of set-asides for conservation purposes is
practiced in all parts of the world, with 11% of the global forest
area designated for the preservation of biodiversity (FAO,
2006). There has long been a theoretical interest in ecology on
how the size and spatial configuration of patches affect
communities and populations (e.g. MacArthur and Wilson,
1967; Diamond, 1975; Hanski, 1999), and this has been linked
to nature conservation, with substantial practical implications.
Further, in the last 20 years there has been an increasing focus
on systematic conservation planning, i.e. how to select
protected areas in a way that captures biodiversity as efficiently
* Corresponding author. Tel.: +46 18 67 22 67; fax: +46 18 67 35 37.
E-mail address: karin.perhans@nvb.slu.se (K. Perhans).
as possible (e.g. Margules and Pressey, 2000). But, there is still
a lack of empirical studies in which comparisons regarding
biodiversity are made between different types of set-asides.
Because large financial resources are spent on biodiversity
preservation and protection of forest land, this research topic
warrants more attention.
In Scandinavia, intensification of forestry during the last 100
years has led to even-aged, more fragmented forests with
shorter rotation periods and a scarcity of dead wood (Esseen
et al., 1997; Östlund et al., 1997). This has resulted in small or
declining populations of almost 2000 forest-living species
(Gärdenfors, 2005) and in recognition of this, increasing
amounts of forest land have been set-aside during the last
decades and new management techniques are being employed
in order to satisfy the needs of the flora and fauna. Today 4% of
the productive forest land in Sweden is protected by law,
although only 1% of this is outside the mountain forest
0378-1127/$ – see front matter # 2007 Elsevier B.V. All rights reserved.
doi:10.1016/j.foreco.2007.01.055

K. Perhans et al. / Forest Ecology and Management 242 (2007) 374–390 375
(Svedlund and Löfgren, 2003). This is a low proportion
compared to other countries where forest is a dominating land
class (FAO, 2006). On the other hand, the protection regulations
are strictly obeyed and thus the set-aside areas are normally
completely allocated to biodiversity conservation.
There are several types of forest set-asides in Sweden as well
as in other parts of boreal Europe, differing in establishment
form, legal status and size. Nature reserves represent the most
common type, with about 2500 established from 1967 to 2003,
covering 850,000 ha of productive forest land (Svedlund and
Löfgren, 2003). The mean size of forested or partly forested
nature reserves is 130 ha, but the size varies from a few hectares
up to several thousands of hectares. Most reserves are
established to protect biodiversity, although there can also be
recreational purposes. A majority of nature reserves have a
strong legal protection and timber harvesting is normally
prohibited.
Lately, the so-called woodland key habitats (hereafter called
‘‘key habitats’’) have come to play an important role in
conservation. The Swedish Forest Agency initiated a survey in
1993 on all forest holdings 5000 ha, the
forest owner is responsible for carrying out the inventory,
supervised by the Swedish Forest Agency. According to the
definition, a key habitat is a forest area of great importance to
sensitive flora and fauna because of its structural, historical and
physical characteristics and should contain or be expected to
contain red-listed species (Nitare and Norén, 1992; Norén et al.,
2002). The inventory of key habitats is based mainly on
structural characteristics and on a set of indicator species of
bryophytes, lichens and fungi, supposed to indicate forests with
high conservation value. The key habitats are classified into 51
different biotope types, of which coniferous forest, coniferous
wetland forest and water-associated habitats are the most
common types in middle Sweden (Swedish Forest Agency,
2006). Similar inventories of key habitats have recently been
conducted in Denmark, Norway, Finland, Estonia, Latvia and
Lithuania. Key habitats are generally small with a mean size of
3.2 ha (Swedish Forest Agency, 2006). Until 2006, about
56,000 key habitats were found on private land, corresponding
to an area of 150,000 ha, or slightly more than 1%, of the
productive forestland. Key habitats are not protected by law, but
the forest owner must consult with the Swedish Forest Agency
before logging. According to the forest certification schemes
FSC (Forest Stewardship Council) and PEFC (Programme for
the Endorsement of Forest Certification schemes), forest
owners should voluntarily set aside 5% of their productive
forest land, and where key habitats are present they are
normally prioritized. Additionally, due to the above mentioned
certification schemes and a strong public opinion, timber from
key habitats is virtually unsalable. Therefore, here we classify
key habitats as set-asides.
Increasingly, small groups of living trees (hereafter called
‘‘retention groups’’) are systematically left on clear-cuts as a
way to reduce the negative impacts of clear-cutting on
biodiversity (Hazell and Gustafsson, 1999). On average
3.1% of the total clear-cut area was left as groups of trees
(0.6%), groups of young trees (0.5%), buffer zones (0.9%) and
sensitive habitats (1.1%) in Sweden between 1995 and 1997
(Swedish Forest Agency, 2006). Also in e.g. Finland (Vanha-
Majamaa and Jalonen, 2001) and the USA (Aubry et al., 1999)
retention of trees at final harvest is a common practice.
Retention groups are intended as short-term refugia for many
organisms over the regeneration phase and can become sources
of dispersal into the growing forest (Esseen et al., 1997;
Franklin et al., 2000). However, the capacity of retention groups
to support biodiversity is poorly known (but see Vanha-
Majamaa and Jalonen, 2001; Nelson and Halpern, 2005).
As large-scale species inventories are expensive, forest setasides
are generally identified on the basis of stand
characteristics or indicator species. Thus, to a large extent
the actual species diversity in the set-asides is normally
unknown. Several comparisons have been made between setaside
and non set-aside areas, e.g. for birds in Finland (Virkkala
et al., 1994), saproxylic beetles in Norway (Sverdrup-
Thygeson, 2002) and vascular plants in South Africa
(Shackleton, 2000). But, we know of no study, in Scandinavia
or elsewhere, in which extensive field inventory of species has
been made in different types of set-asides, for the purpose of
comparison and evaluation of their biodiversity quality (here
defined as total number of species and presence of species of
conservation interest). Further, there is a scarcity of full-range
data on less known but species-rich organism groups like
bryophytes and lichens. Many of the conservation-oriented
studies performed on these groups have focused on red-listed
species (e.g. Johansson and Gustafsson, 2001; Gustafsson et al.,
2004). Bryophytes and lichens have been given increasing
attention within conservation since they are considered
sensitive to forestry operations, and in northern Europe they
are used as indicators of forest sites valuable to biodiversity
(Hallingbäck and Weibull, 1996; Esseen et al., 1997).
The aim of this study, which is part of a research project on
cost-effective reserve strategies in boreal Sweden, was to
compare the biodiversity quality per unit area in three types of
forest set-asides: nature reserves, key habitats and retention
groups, as well as in old managed forests. Bryophytes and
lichens were used as assessment tools and we focused on
total species richness and on two sub-groups of species of
conservation interest: red-listed species and indicator species,
which are assumed to indicate forests with high conservation
value.
2. Methods
2.1. Study area
The inventory was made in the northern part of Gävleborg
county (Fig. 1), within the middle boreal vegetation zone (Ahti
et al., 1968) in Sweden. The study area is approximately
150 km 150 km, central position 61845 0 N, 16810 0 E. The
elevation ranges from 80 m above sea level in the eastern parts
to 490 m above sea level in the northwestern part. The forests,
which cover 80% of Gävleborg county, consist mainly of

376
K. Perhans et al. / Forest Ecology and Management 242 (2007) 374–390
habitats were randomly selected from all forest patches which,
according to wRESEx, fulfilled the above stated criteria.
A retention group was in this study defined as an island of
forest, larger than 25 m 25 m but not larger than 0.5 ha,
completely surrounded by clear-cut area. Due to errors in the
satellite interpretation of small areas, a sufficient number of
retention groups could not be identified in wRESEx. Instead,
we gathered information from the satellite map system of the
Swedish Forest Agency and through interviews with forest
personnel at the Swedish Forest Agency and at the forest
company Stora Enso. The time since creation of the retention
groups varied but was always less than 15 years. The old
managed forest sites were selected by randomly sampling
crossings of dividing lines on a 10 km 10 km grid covering
the study area, and from there searching in a random direction.
The first forest patch that fulfilled the criteria, and was not
within the boundaries of a nature reserve or a key habitat, was
selected as an old managed forest site.
Fig. 1. Location of field study in boreal Sweden. Magnified area shows northern
part of Gävleborg county and the distribution of the 80 study plots. (&) Key
habitat; (&) nature reserve; (*) old managed forest; (*) retention group.
Norway spruce (Picea abies) and Scots pine (Pinus sylvestris)
with about 10% deciduous trees, mainly birch (Betula pendula
and B. pubescens)(Swedish Forest Agency, 2006). Key habitats
constitute 0.7% (5500 ha) of the productive forest on private
land and nature reserves 0.6% (9100 ha) of the total area of
productive forest land in the study area (Swedish Forest
Agency, 2006).
2.2. Selection of study sites
We randomly selected a total of 80 sites (areas with similar
forest composition and characteristics), 20 in each of the 4
forest categories: nature reserves, key habitats, retention groups
and old managed forests (Fig. 1). Information on the location of
existing nature reserves, as well as areas in the process of
becoming nature reserves, was provided by the county
administrative board of Gävleborg, and information on key
habitats were obtained from the Swedish Forest Agency and
from the forest company Holmen Skog.
The selection of sites was made by using the satellite map
‘‘wRESEx’’ (Angelstam et al., 2003), which divides all forest
land in Gävleborg county into 33 classes according to dominant
tree species and stand age. The resolution of the satellite map is
25 m 25 m. To select as similar, and thus comparable, sites as
possible within the four forest categories, we only used the class
‘‘>70% spruce, >110 years’’. Further criteria for sites were that
they should be larger than 50 m 50 m (except for retention
groups), >1 km apart (for sites of the same forest category, to
avoid interdependence), >5 km from the coast (to avoid the
special climate and geology of the coast), 0.5 m.
2.3.2. Species inventory
The species inventory was carried out between August and
October 2004. The same study plot as for the collection of forest
characteristics was used. We made total species lists of
bryophytes on all substrates, i.e. the ground, living and dead
trees up to 2.5 m above ground, stumps, logs, boulders and
other substrates reachable from the ground, and of lichens
growing on spruce, both living and dead standing trees, logs and
stumps. Only presence of species was recorded, i.e. no measure
of abundance.
Species were classified as red-listed according to the
Swedish Species Information Centre (Gärdenfors, 2005) and as
indicator species of forest of high conservation value according
to the Swedish Forest Agency (Nitare, 2000).
Nomenclature for bryophytes follows Söderström and
Hedenäs (1998), for lichens Santesson et al. (2004) and for
trees Krok and Almqvist (2001).
2.4. Data analysis
2.4.1. Species richness
Tests of differences in mean species number between forest
categories were performed using SAS/STAT 8.02 (SAS, 2000).
Analyses of variance including Tukey’s test for multiple
comparisons were executed in Proc ANOVA, with Bonferroni
correction. Non-normality was investigated with Shapiro Wilks
statistics, skewness and kurtosis in Proc UNIVARIATE, and

K. Perhans et al. / Forest Ecology and Management 242 (2007) 374–390 377
was found to be predominant for red-listed and indicator
species. Due to this, we added the non-parametric Kruskal–
Wallis test as a complement to test of means, in Proc
NPAR1WAY. Species accumulation curves (sample-based
rarefaction curves; Gotelli and Colwell, 2001) were constructed
in EstimateS 7.5 (Colwell, 2005).
2.4.2. Species composition
The similarity between plots of the four forest categories
regarding species composition was explored with NMS (Quinn
and Keough, 2002). Three other ordination methods were also
used to test the robustness of the pattern: PCA, CA and DCA.
All analyses were performed in PC-ORD (McCune and
Mefford, 1999) with all species included and with default
options, except that down-weighting of rare species was not
applied. The similarity of plots within the same forest category
was investigated with Sørensen index, calculated for all pairs of
plots within each category (n = 190) in EstimateS 7.5 (Colwell,
2005). Indicator Species Analysis (PC-ORD) was used to
identify species mainly associated with one forest category.
Only species which occurred in 5 plots or more in at least 1
forest category were included and 1000 randomizations were
used in the Monte Carlo test.
2.4.3. Site selection
To investigate how efficiently each forest category could
capture biodiversity, we applied a site selection method based
on linear programming. The method is useful in this context as
it accounts for both species richness and composition.
Specifically, the method assesses the complementarity of sites
(Vane-Wright et al., 1991), which is a measure of how much
each site or group of sites contributes to an overall goal. Here,
the smallest set of plots including all species in at least
one plot each was sought, i.e., a set covering problem (SCP)
was formulated. The SCP is mathematically expressed as
(Possingham et al., 2000; Rodrigues et al., 2000a):
Min z ¼ Xn
subject to :
x j
j¼1
X n
j¼1
a ij x j 1; i ¼ 1; 2; ...; m (1)
x j 2f0; 1g; j ¼ 1; 2; ...; n (2)
where z = total number of plots, x j = 1 if plot j is selected for the
reserve system, 0 if not, a ij = 1 if species i exists in plot j, 0if
not, j = index for plots, i = index for species.
The analysis was performed in AMPL CPLEX System 9.1
(ILOG, 2005). We interpreted the proportions of the four forest
categories in the optimal solution as their relative contribution
to efficient species representation. Separate analyses were
made to find optimal combinations of plots for representing all
species, red-listed species and indicator species. This was done
for bryophytes and lichens separately and for the two taxa
together. Where multiple optimal solutions to a certain problem
existed, we computed an average of the 30 first obtained.
In the classical formulation of the SCP, the minimum
representation level on the right hand side of inequality (1) is set
to 1, i.e. all species must be included in at least one plot each
(Underhill, 1994; Rodrigues et al., 2000a). However, Rodrigues
et al. (2000b) showed that demanding three representations of
each species considerably increased the probability of species
long-term persistence in a reserve network. Therefore, we also
repeated the analyses with the representation level set to 3. In
the latter case all species occurring in less than three plots were
excluded, as their long-term persistence was judged to be
questionable.
3. Results
3.1. Forest characteristics
The mean proportion of spruce in the forest categories varied
between 73% and 88% (Table 1). The highest mean proportion
of deciduous trees was found in nature reserves (18%) and the
lowest in old managed forests (4%). Retention groups deviated
in age with a mean of 103 years compared to 116–130 years for
the other categories, and in volume of living trees: 201 m 3 ha 1
compared to 312–347 m 3 ha 1 . The volume of dead wood was
largely similar in the categories, with a range for spruce
between 12.1 and 18.9 m 3 ha 1 . The ground moisture class was
mesic in most plots and brooks and springs were found in 65%
of the key habitats but only in 20–30% in the other categories.
3.2. Species richness
In total, 428 species, of which 252 bryophytes and 176
lichens, were found in the 80 study plots (Appendix A). The
mean number of species per plot was 106 and varied between
71 (a retention group) and 162 (a key habitat). Thirty-six
percent of the bryophytes and 54% of the lichens were found in
all forest categories, and 8% of the bryophytes and 19% of the
lichens were found in all forest categories in at least half of the
plots in each category. Of the recorded species, 28 bryophytes
and 55 lichens were indicator species and 10 lichens and 10
bryophytes were on the Swedish Red List. Of the 80 inventoried
plots, 63 had at least one red-listed species. The most common
red-listed species was the lichen Bryoria nadvornikiana,
occurring in 44 plots, followed by the bryophyte Anastrophyllum
hellerianum (18 plots) and the lichen Cliostomum leprosum
(11 plots).
Key habitats and nature reserves had a significantly higher
mean number of red-listed bryophytes and indicator species of
bryophytes than the retention groups and old managed forests
(Table 2). For lichens, the only significant difference was
between key habitats and retention groups for red-listed
species. No significant differences between old managed forest
and retention groups could be seen for any of the species
groups. The species accumulation curves revealed that the total
number of bryophyte species was significantly higher in key
habitats than in the other forest categories (Fig. 2a) and that no
significant differences in total species number occurred for
lichens (Fig. 2b).

378
K. Perhans et al. / Forest Ecology and Management 242 (2007) 374–390
Table 1
Differences between forest categories regarding forest characteristics
Key habitats (n = 20) Nature reserves (n = 20) Old managed forests (n = 20) Retention groups (n = 20) p-value a
Altitude (m a.s.l.) 242 87 (115–390) a 340 82 (216–490) b 273 90 (84–415) ab 326 67 (175–459) b 0.001 (A)
Living trees
Stand age (years) 130 35 (98–251) a 119 19 (92–175) ab 116 19 (87–161) ab 103 25 (45–142) b 0.065 (K)
Total tree volume (m 3 ha 1 ) 323 125 (128–570) a 347 104 (172–555) a 312 86 (145–473) a 201 89 (69–384) b
0.5 m (number)
18 23 (0–87) a 21 15 (0–53) a 16 17 (0–62) a 15 16 (0–50) a 0.331 (K)
Mean + S.D. (standard deviation) and range presented. Forest categories with same letters do not differ significantly according to Tukey’s test for multiple
comparisions.
a Derived from ANOVA (A), Kruskal–Wallis (K) or Chi-square test (C). Significant results in bold type.
b II = mesic, III = mesic-moist, IV = moist.
3.3. Species composition
None of the different ordination analyses, NMS, PCA, CA,
DCA, revealed any distinct groupings for either bryophytes or
lichens (Fig. 3; only NMS presented). Although most plots
were positioned in a large cluster, the bryophyte community of
some key habitats and retention groups seemed to differ slightly
from the other forest categories. The Sørensen values showed
that, in particular, nature reserve plots were more similar in
composition than key habitat plots. Mean Sørensen value
(average proportion of shared species between two plots)
within these two types was for bryophytes 0.66 and 0.53,
respectively, and for lichens 0.76 and 0.73. Intermediate values
were obtained for old managed forests and retention groups.
If species occurring in all of the four forest categories were
excluded, 25–31% of the found species remained for nature
reserves, old managed forests and retention groups, but 46% of
the species for key habitats, indicating the presence of more
uncommon species in this category. There was a similar pattern
for the number of species which were unique for one forest
category (species which were not found in any other category):
76 such species (61 bryophytes and 15 lichens) occurred in key
habitats compared to 9, 14 and 23 species in nature reserves, old
managed forests and retention groups, respectively.
According to the Indicator Species Analysis some species
were rather strongly associated ( p < 0.01) with one forest
category (Appendix A). For nature reserves these species were
the bryophytes Cephalozia leucantha, Hylocomiastrum umbra-
Table 2
Differences between forest categories regarding species number
Key habitats (n = 20) Nature reserves (n = 20) Old managed forests (n = 20) Retention groups (n = 20) p-value a
Mean Total Mean Total Mean Total Mean Total
All species
Bryophytes 59.7 19.7 a 219 47.1 13.2 ab 140 38.1 9.9 b 129 40.7 14.5 b 156

K. Perhans et al. / Forest Ecology and Management 242 (2007) 374–390 379
Fig. 2. Species accumulation curves (sample-based; Colwell, 2005). (a) Bryophytes:
confidence intervals (95% CI) given only for key habitats and retention
groups. (b) Lichens: no forest category differed significantly from any of the
others and therefore no confidence intervals are given.
tum, Lophozia ascendens, and for key habitats the bryophytes
Cephalozia bicuspidata, Pellia neesiana, Plagiothecium laetum,
Sphagnum centrale, and the lichens Arthonia leucopellaea,
Lecanactis abietina. At lower significance ( p < 0.05),
some species were also indicative of retention groups; the
bryophytes Ceratodon purpureus, Pogonatum urnigerum,
Sphagnum angustifolium and the lichens Bryoria furcellata,
Cladonia gracilis, Lecanora hypoptella, Pycnora sorophora
and Xylographa vitiligo. No species was significantly
associated with old managed forests.
3.4. Site selection
In all site selection analyses except for red-listed lichens,
key habitats constituted the largest part of selected plots in the
optimal solutions (Fig. 4). The pattern was especially
pronounced for indicator species, and for bryophytes compared
to lichens. Retention groups were of least importance for redlisted
and indicator species and generally, nature reserves and
old managed forests appeared to contribute about equally to
species representation for all species groups. When at least
three representations of each species were required, the nature
reserves constituted a larger fraction of the optimal proportions
(Fig. 5) and became as important as key habitats to represent the
different subsets of bryophyte species. For lichens, key habitats
were still the most important forest category. The number of
Fig. 3. NMS (McCune and Mefford, 1999). (a) Bryophytes and (b) lichens. Plus
signs = key habitats, stars = nature reserves, filled standing quadrates = old
managed forests and circles = retention groups.
optimal solutions varied between 1 and >30 but the alternative
solutions were generally very similar.
4. Discussion
4.1. Comparison of forest categories
Despite the fact that we applied a sampling scheme that
should guarantee that the forest composition and site conditions
were largely similar among the forest categories, we found

380
K. Perhans et al. / Forest Ecology and Management 242 (2007) 374–390
Fig. 4. Set covering problem (SCP) with one representation of each species. (a)
Bryophytes and lichens, (b) bryophytes and (c) lichens. Bars show optimal
proportions of forest categories for representing all species in the four species
groups.
Fig. 5. Set covering problem (SCP) with three representations of each species.
(a) Bryophytes and lichens, (b) bryophytes and (c) lichens. Bars show optimal
proportions of forest categories for representing all species in the four species
groups with species occurring in less than three plots excluded.
important differences in species richness and composition in
the four categories.
Key habitats contained the largest number of species in all
surveyed species categories. In the site selection analyses they
generally constituted the largest part of selected plots,
indicating apart from a high species richness also a high
complementarity between plots. The most important explanation
for this is probably that key habitats are small hotspots that
do not contain buffers, as opposed to nature reserves which
usually are considerably larger and contain a mixture of high
and low quality parts. Presence of brooks or springs was also
comparatively more common in the key habitats than in the

K. Perhans et al. / Forest Ecology and Management 242 (2007) 374–390 381
other categories. Consequently, more bryophytes connected to
moist and wet conditions were found in the key habitats,
contributing to their high species richness and their prominent
role in the site selection analyses.
We deliberately located the study plots randomly in the
nature reserves as well in the other categories (within forest
sites fulfilling our selection criteria), meaning that plots could
sample both higher and lower quality parts of the reserve area.
Together with key habitats, nature reserves nevertheless
contained a higher mean number of red-listed and indicator
species of bryophytes than old managed forests and retention
groups, suggesting a generally high conservation value. In the
site selection analyses where three representations of each
species were required, a demand expected to increase the
probability of species long-term persistence, nature reserves
gained in importance relative to the other forest categories and
constituted almost as large proportions in the optimal selections
of plots as the key habitats. There could be several explanations
for this. A plausible one is a combination of the comparatively
high between-plot similarity of nature reserves, with a large
share of species occurring in many plots, the rather high mean
species number and the fact that the most uncommon species
were excluded in this analysis, affecting key habitats the most
as many of the species in this category were found in only one
or two plots.
The retention groups did not differ significantly from the old
managed forests in any of the analyses, and were of poorer
quality, as assessed by presence of red-listed and indicator
species, than key habitats and nature reserves. This is not
surprising given that selection of retention groups most likely
often is based on aspects of timber quality and conditions
affecting logging and logistics, and not only on assumed quality
for biodiversity which generally is the selection criterion for
key habitats and nature reserves.
The old managed forests were in many aspects similar to the
key habitats and the nature reserves. It might be unexpected that
these forests have about the same biodiversity quality as setasides,
but earlier studies (e.g. Gustafsson et al., 2004) have
shown that in this part of boreal Sweden, there are large areas of
biodiversity-rich old managed forests with a high number of
red-listed bryophytes and lichens. Detailed studies on forest
history in the region are lacking, but it is likely that the old
managed forests here, as in many other parts of boreal Sweden,
never have been clear-cut but instead have had a continuous tree
cover due to only selective fellings in the 19th and the
beginning of the 20th century (Östlund et al., 1997). Thus, the
old managed forests are first-generation managed forests with
substantial remains of qualities from the earlier natural forest
landscape. The conditions for biodiversity in future managed
forests will be very different, since they will regenerate after
clear-cutting and be part of landscapes with large areas of
young homogeneous, intensively managed forests.
Larger differences between the four forest categories were
found for bryophytes than for lichens, shown by the rarefaction
curves (Fig. 2) and by the fact that more species were found in
all four categories for lichens than for bryophytes. This most
likely was much due to the fact that bryophytes were surveyed
on all substrates, whereas lichens were only recorded on spruce.
The presence of a brook or spring, a few large boulders or old
deciduous trees in the inventoried plots would influence the
richness and composition of bryophytes greatly, but not to the
same extent lichens. However, as epiphytic and epixylic lichens
constitute the main part of the lichen flora in old-growth spruce
forests, the lichens recorded here most likely well represent the
lichen community. But, if deciduous trees had also been
surveyed, this might have altered the lichen species richness
and composition considerably.
As the goal with set-asides is to maintain biodiversity over
time, an important aspect to consider is the temporal dynamics,
i.e. how the set-asides can be predicted to change over time. It
has been proposed by Berglund and Jonsson (2005) that an
extinction debt (sensu Tilman et al., 1994) might be present in
the key habitats and that they in course of time will lose some of
their species. To an even greater extent the same applies to the
retention groups, being even smaller and heavily impacted by
edge effects. The proportion of the total area in each set-aside
type that could be affected by edge effects from adjacent clearcuttings
is very different in the three set-aside types. Assuming
mean sizes of retention groups, key habitats and nature reserves
of 0.2, 3 and 100 ha, respectively, circular shape of the setasides
and the edge effect extending 30 m into the set-asides
(Chen et al., 1995), the area of remaining undisturbed interior
habitat would be 0% in retention groups, 48% in key habitats
and 90% in nature reserves. As many bryophytes and lichens
are dependent on habitat with interior characteristics (Hilmo
and Holien, 2002; Hylander, 2004), this suggests that a larger
proportion of species run the risk of local extinction in the
retention groups than in the nature reserves, with key habitats
being in between.
4.2. Conservation implications
The high species richness in the key habitats, including many
uncommon species, show that they are core areas for biodiversity
in the present forest landscape. This observation is consistent
with studies of Götmark and Thorell (2003), who found that the
density of large trees and dead wood decreased with increasing
reserve size, and of Ranius and Kindvall (2006) who conclude,
based on modelling, that setting aside many small reserves is a
better strategy in a managed, highly fragmented landscape than
preserving a few large. The pattern is probably caused by the fact
that in managed forest landscapes, such as in Scandinavia, the
areas of highest quality are often small and scattered. This
changes the conditions for choosing among different conservation
strategies compared to other parts of the world where
forestry has been less intense. It should be borne in mind though,
that a conservation strategy based on setting aside many small
areas, such as key habitats, has inherent potential disadvantages
in terms of increased vulnerability to stochastic events and edge
effects. Also, large protected areas might be essential for species
with large home ranges and species depending on large
undisturbed tracts of old-growth forest.
As the conditions for biodiversity in future, more intensively
managed forests will be different from today, the value of

382
K. Perhans et al. / Forest Ecology and Management 242 (2007) 374–390
preserving more intact habitats for species in set-asides will
increase over time. However, as set-asides are intended to
function as sources of dispersal for species into the surrounding
matrix, a certain quality of the managed forests is needed for
this to be possible. Recently, it has been pointed out that a
reverse relationship also exists (Bengtsson et al., 2003). If the
set-asides are small and scattered, the matrix is an important
buffer for recolonization of species lost in a set-aside after a
disturbance, such as fire or a storm, or other stochastic events
often affecting small areas.
Even though the retention groups did not differ much from
the old managed forests in terms of their species content, there
are still reasons why they are interesting in a long-term
perspective. Over time, their species composition is likely to
change more compared with the other forest categories due to
their small size and their location on clear-cuts, which causes
tree death and a large exposure to wind and sun. Initially, local
species extinctions of forest interior species are to be expected
but successively colonisations of new species will occur.
Already now there were signs of disturbance in the retention
groups, as early successional bryophytes like C. purpureus and
P. urnigerum and light-preferring lichens like B. furcellata, C.
gracilis and X. vitiligo were indicative of this category. In other
organism groups, such as saproxylic beetles, there are species
that are strictly confined to the type of sun-exposed habitat
found in large amounts in retention groups (Kaila et al., 1997)
and for such species the retention groups likely constitute an
important complement to other set-asides types. The retention
groups are also the most widely dispersed of the three types of
set-asides, and could thus potentially function as stepping
stones for many groups of species in the managed forest
landscape. However, the capacity of retention groups to
function as stepping stones and as sources of dispersal of
species into the regenerating forest is still fairly unknown and
should preferably be studied by long-term monitoring.
We believe that nature reserves, key habitats and retention
groups have different functions and should therefore be used in
parallel as conservation tools in the forest landscape. Such a
conservation strategy, operating over multiple spatial scales,
also implies a risk-spreading approach as different species and
habitat types most likely are best conserved by different
measures. Therefore, we think that in designing efficient forest
conservation policies, most can be gained by optimizing
between areas within each set-aside type, rather than between
different types.
In this study, we have consistently worked only with the
quality per unit area of each forest category and therefore the
results should only be interpreted as such. In order to increase
the practical applicability, in subsequent studies within this
research project we intend to model the biodiversity quality in
the actual mean sizes of each forest category based on the
species data from this study, and also include the different
economic costs of setting aside each forest category. The
relation between species and environmental variables will also
be further explored.
Acknowledgements
We are very grateful to Sten Nordlund and Jan ten Hoopen
for recording many of the stand characteristics in the study plots
and to Claes Kindstrand for calculating volumes of living trees.
The county administrative board of Gävleborg, the Swedish
Forest Agency, Holmen Skog and Stora Enso kindly provided
information and assistance in the selection of study sites. We
also thank Bo Söderström, Sofie Wikberg, PhD students at the
department and two anonymous reviewers for valuable
comments on the manuscript. The study is part of the research
programme ‘‘Biodiversity and Economy’’, financed by the
Swedish Research Council for Environment, Agricultural
Sciences and Spatial Planning.
Appendix A. Species list
Species
Indicator
(IND) a
Red-list Frequency (number of plots) p-value ISA c
category b Key habitats Nature reserves Old managed Retention
n =20 n =20
forests n =20 groups n =20
Lichens
Absconditella lignicola 1 0 0 0
Alectoria sarmentosa IND 14 17 17 16
Arthonia incarnata VU 1 2 1 1
Arthonia leucopellaea IND 9 4 2 0 0.004 K
Arthonia mediella 8 10 7 4
Arthonia vinosa IND 4 3 2 2
Bacidia beckhausii 0 0 0 1
Biatora chrysantha 4 6 2 5
Biatora efflorescens 20 20 20 20
Biatora fallax DD 2 1 1 1
Biatora helvola 20 20 20 20
Biatora meiocarpa 1 0 0 1
Biatora ocelliformis 2 4 0 1
Biatora sphaeroidiza 8 10 8 4
Biatora vacciniicola 0 1 1 0
Biatora vernalis 0 2 0 1
Bryoria capillaries 20 13 20 20
Bryoria fremontii 2 2 2 1

Appendix A (Continued )
Species
K. Perhans et al. / Forest Ecology and Management 242 (2007) 374–390 383
Indicator
(IND) a
Red-list Frequency (number of plots) p-value ISA c
category b Key habitats Nature reserves Old managed Retention
n =20 n =20
forests n =20 groups n =20
Bryoria furcellata IND 10 9 11 17 0.014 R
Bryoria fuscescens/implexa 17 19 19 20
Bryoria nadvornikiana IND NT 9 15 11 9
Bryoria simplicior 0 0 1 0
Buellia disciformis 0 0 1 0
Buellia schaereri 0 1 0 0
Calicium glaucellum 9 6 9 9
Calicium parvum IND 2 0 2 0
Calicium trabinellum 1 2 3 3
Calicium viride 2 1 3 2
Catinaria atropurpurea 0 0 0 1
Cetraria sepincola 1 0 1 2
Chaenotheca brunneola 0 0 1 1
Chaenotheca chrysocephala 19 20 19 16
Chaenotheca ferruginea 11 10 8 8
Chaenotheca furfuracea 12 9 10 6
Chaenotheca gracillima IND NT 3 0 1 0
Chaenotheca laevigata IND NT 0 1 1 0
Chaenotheca sphaerocephala VU 2 0 0 0
Chaenotheca stemonea 2 0 0 1
Chaenotheca subroscida IND 8 6 8 6
Chaenotheca trichialis 13 10 16 8
Chaenotheca xyloxena 1 0 4 1
Chaenothecopsis consociata 17 15 15 15
Chaenothecopsis debilis 0 0 1 0
Chaenothecopsis epithallina 1 0 3 0
Chaenothecopsis nana 4 3 2 1
Chaenothecopsis nigra 2 2 2 0
Chaenothecopsis pusilla 4 1 1 0
Chaenothecopsis pusiola 1 1 2 1
Chaenothecopsis savonica 4 0 1 0
Chaenothecopsis vainioana 1 2 0 1
Chaenothecopsis viridialba IND NT 0 2 1 1
Chaenothecopsis viridireagens 2 0 0 1
Cheiromycina flabelliformis NT 3 0 0 0
Chrysothrix candelaris 7 5 3 2
Cladonia arbuscula 3 1 1 4
Cladonia bacilliformis 12 18 15 19
Cladonia botrytes 3 4 4 3
Cladonia carneola 4 3 3 4
Cladonia cenotea 20 19 18 19
Cladonia chlorophaea 7 9 9 6
Cladonia coniocraea 20 20 20 20
Cladonia cornuta 3 6 5 6
Cladonia crispata 1 3 4 1
Cladonia deformis 2 1 1 5
Cladonia digitata 19 16 16 17
Cladonia fimbriata 17 11 16 18
Cladonia furcata 0 0 1 0
Cladonia gracilis 2 0 0 5 0.021 R
Cladonia macilenta 1 0 0 0
Cladonia norvegica 5 3 4 2
Cladonia ochrochlora 3 0 0 1
Cladonia pleurota 0 0 0 1
Cladonia pyxidata 2 3 4 4
Cladonia rangiferina 7 8 5 6
Cladonia squamosa 1 4 0 2
Cladonia sulphurina 1 3 5 3
Cliostomum leprosum NT 7 4 0 0 0.023 K
Cliostomum pallens 9 6 10 9
Dimerella pineti 8 10 10 4
Evernia prunastri 0 0 0 1
Fellhanera margaritella 9 5 7 10

384
Appendix A (Continued )
Species
K. Perhans et al. / Forest Ecology and Management 242 (2007) 374–390
Indicator
(IND) a
Red-list Frequency (number of plots) p-value ISA c
category b Key habitats Nature reserves Old managed Retention
n =20 n =20
forests n =20 groups n =20
Fellhanera subtilis 4 7 9 3
Fuscidea arboricola 12 7 6 8
Gyalideopsis piceicola 2 2 2 1
Hypocenomyce friesii 5 1 1 2
Hypocenomyce scalaris 3 1 3 1
Hypogymnia farinacea 4 3 3 0
Hypogymnia physodes 20 20 20 20
Hypogymnia tubulosa 20 20 20 20
Hypogymnia vittata IND 2 2 0 2
Icmadophila ericetorum IND 2 0 0 0
Imshaugia aleurites 9 4 9 9
Japewia subaurifera 20 20 20 20
Japewia tornoënsis 14 18 16 17
Lecanactis abietina 8 4 0 0 0.006 K
Lecanora aitema 0 0 0 1
Lecanora albellula 1 0 0 0
Lecanora boligera 1 0 0 0
Lecanora circumborealis 5 3 8 9
Lecanora expallens 17 14 14 14
Lecanora hypoptella 12 13 9 18 0.011 R
Lecanora pulicaris 11 6 12 8
Lecanora symmicta 1 0 1 2
Lecidea albofuscescens 18 20 18 17
Lecidea betulicola 0 1 0 0
Lecidea leprarioides 13 13 14 14
Lecidea nylanderi 20 20 20 20
Lecidea pullata 19 18 20 18
Lecidea turgidula 18 20 20 18
Lepraria spp. 20 20 20 20
Lopadium disciforme IND 11 13 10 6
Loxospora elatina 18 20 19 15
Melanelia exasperatula 1 1 2 2
Melanelia fuliginosa 1 0 0 0
Melanelia olivacea 0 1 0 1
Melanelia subaurifera 0 0 0 2
Micarea anterior 0 0 1 0
Micarea contexta 1 2 2 1
Micarea denigrata 4 2 2 2
Micarea globulosella 20 20 19 19
Micarea melaena 6 7 7 5
Micarea misella 6 2 2 2
Micarea peliocarpa 1 0 2 0
Micarea prasina 20 20 19 19
Microcalicium ahlneri IND 1 0 0 0
Microcalicium disseminatum 4 4 2 2
Mycobilimbia hypnorum 0 0 1 0
Mycoblastus affinis 17 20 17 18
Mycoblastus alpinus 15 10 13 12
Mycoblastus fucatus 18 15 16 17
Mycoblastus sanguinarius 20 20 19 18
Mycocalicium subtile 3 0 2 2
Ochrolechia alboflavescens 4 4 7 5
Ochrolechia androgyna 19 19 20 19
Ochrolechia microstictoides 20 20 20 20
Ochrolechia pallescens 3 4 2 8
Parmelia saxatilis 1 1 1 1
Parmelia sulcata 19 20 20 19
Parmeliopsis ambigua 20 20 20 20
Parmeliopsis hyperopta 20 20 20 20
Peltigera canina 0 0 1 0
Peltigera degenii 0 1 0 0
Peltigera polydactyla 0 3 1 0
Pertusaria amara 17 15 11 6 0.037 K

Appendix A (Continued )
Species
K. Perhans et al. / Forest Ecology and Management 242 (2007) 374–390 385
Indicator
(IND) a
Red-list Frequency (number of plots) p-value ISA c
category b Key habitats Nature reserves Old managed Retention
n =20 n =20
forests n =20 groups n =20
Pertusaria borealis/pupillaris 20 20 20 20
Pertusaria ophthalmiza 3 3 1 0
Phlyctis argena 1 0 0 0
Physcia adscendens 0 0 0 1
Physcia tenella 0 0 1 0
Placynthiella dasaea 19 19 13 13
Placynthiella icmalea 3 4 1 2
Placynthiella uliginosa 0 0 0 1
Platismatia glauca 20 20 20 20
Pseudevernia furfuracea 15 17 14 16
Pycnora leucococca 17 18 17 18
Pycnora sorophora 0 0 2 5 0.025 R
Pyrrhospora cinnabarina IND 0 0 1 0
Ramalina thrausta IND EN 1 0 0 0
Rinodina degeliana 1 0 0 0
Rinodina septentrionalis 2 0 0 0
Ropalospora viridis/Fuscidea pusilla 20 20 20 20
Sarcosagium campestre 1 0 0 0
Scoliciosporum chlorococcum 9 9 12 16
Scoliciosporum sarothamni 1 0 0 0
Trapeliopsis granulosa 0 1 0 0
Tuckermanopsis chlorophylla 20 20 20 20
Usnea filipendula 20 20 20 19
Usnea glabrescens 0 1 2 2
Usnea hirta 5 0 4 4
Usnea lapponica 1 1 2 0
Usnea subfloridana 17 19 19 17
Varicellaria rhodocarpa 1 1 2 0
Vulpicida pinastri 20 20 20 19
Xylographa parallela 4 2 2 3
Xylographa trunciseda 0 0 1 2
Xylographa vitiligo 5 3 3 10 0.024 R
Bryophytes
Amblystegium serpens 0 3 1 1
Amblystegium subtile 0 1 0 0
Amphidium lapponicum 1 0 0 0
Amphidium mougeotii 3 0 0 0
Anastrophyllum hellerianum IND NT 8 7 1 2
Anastrophyllum michauxii IND NT 1 0 0 0
Anastrophyllum minutum 8 7 5 6
Anastrophyllum saxicola 2 0 0 0
Andreaea rupestris 8 13 12 11
Aneura pinguis 1 0 0 3
Antitrichia curtipendula IND 1 1 0 0
Atrichum tenellum 2 0 0 2
Atrichum undulatum 5 1 0 2
Aulacomnium palustre 5 2 2 3
Barbilophozia attenuata 18 20 14 17
Barbilophozia barbata 9 6 4 4
Barbilophozia floerkei 2 5 7 4
Barbilophozia hatchery 8 16 11 10
Barbilophozia kunzeana 1 0 2 1
Barbilophozia lycopodioides 14 18 16 16
Bartramia halleriana 0 1 0 0
Bartramia pomiformis 3 3 1 2
Blasia pusilla 2 0 0 0
Blepharostoma trichophyllum 18 19 15 15
Blindia acuta 1 0 0 0
Brachythecium erythrorrhizon 2 1 0 2
Brachythecium oedipodium 17 17 14 13
Brachythecium plumosum 3 0 0 0
Brachythecium populeum 1 0 0 0

386
Appendix A (Continued )
Species
K. Perhans et al. / Forest Ecology and Management 242 (2007) 374–390
Indicator
(IND) a
Red-list Frequency (number of plots) p-value ISA c
category b Key habitats Nature reserves Old managed Retention
n =20 n =20
forests n =20 groups n =20
Brachythecium reflexum 13 15 14 8
Brachythecium rivulare 1 0 0 0
Brachythecium salebrosum 9 11 5 7
Brachythecium starkei 19 19 14 16
Brachythecium velutinum 3 2 1 0
Bryum capillare 1 0 0 0
Bryum flaccidum 1 1 0 1
Bryum pseudotriquetrum 1 0 0 2
Bryum sp. 1 1 0 1
Buxbaumia viridis IND 4 1 0 0
Calliergon cordifolium 4 1 1 2
Calliergon giganteum 1 0 0 0
Calliergon richardsonii 0 0 1 1
Calliergonella lindbergii 2 0 0 0
Calypogeia azurea NT 1 0 0 0
Calypogeia integristipula 16 19 16 15
Calypogeia muelleriana 10 6 1 5 0.015 K
Calypogeia neesiana 11 10 7 6
Calypogeia sphagnicola 0 1 1 1
Calypogeia suecica IND VU 5 3 0 0
Campylium protensum 1 0 0 2
Campylium stellatum 1 0 1 1
Campylophyllum sommerfeltii 0 1 1 0
Cephalozia bicuspidata 15 8 6 6 0.006 K
Cephalozia connivens 0 1 1 1
Cephalozia leucantha 4 12 1 1 0.001 N
Cephalozia loitlesbergeri 1 1 0 0
Cephalozia lunulifolia/affinis 16 19 12 11 0.028 N
Cephalozia pleniceps 2 2 3 4
Cephaloziella sp. 10 8 3 7
Ceratodon purpureus 2 2 2 8 0.021 R
Chiloscyphus minor 1 0 0 1
Chiloscyphus pallescens/polyanthos 7 1 2 2
Chiloscyphus profundus 19 16 15 13
Cirriphyllum piliferum 6 3 3 2
Climacium dendroides 3 0 0 0
Cynodontium fallax NT 1 0 0 0
Cynodontium polycarpon 0 0 0 1
Cynodontium strumiferum 7 9 9 9
Cynodontium suecicum IND 1 0 0 0
Cynodontium tenellum 1 0 0 0
Dicranella cerviculata 4 1 1 4
Dicranella crispa 1 0 0 0
Dicranella grevilleana 0 0 1 0
Dicranella heteromalla 1 0 0 2
Dicranoweisia crispula 2 0 0 0
Dicranum bonjeanii 0 0 0 1
Dicranum drummondii 0 1 0 0
Dicranum flexicaule 19 20 19 17
Dicranum fuscescens 19 20 20 18
Dicranum majus 20 20 20 18
Dicranum montanum 15 17 16 17
Dicranum polysetum 6 12 12 12
Dicranum scoparium 20 20 20 20
Diplophyllum albicans 0 0 1 1
Diplophyllum taxifolium 3 2 2 0
Eurhynchium pulchellum 1 1 0 1
Fissidens adianthoides 1 0 0 0
Fissidens osmundoides 3 1 0 0
Fontinalis antipyretica 1 0 0 0
Frullania dilatata 1 0 0 0
Geocalyx graveolens IND 3 2 0 1
Grimmia hartmanii 1 0 0 0

Appendix A (Continued )
Species
K. Perhans et al. / Forest Ecology and Management 242 (2007) 374–390 387
Indicator
(IND) a
Red-list Frequency (number of plots) p-value ISA c
category b Key habitats Nature reserves Old managed Retention
n =20 n =20
forests n =20 groups n =20
Grimmia torquata 1 0 0 0
Gymnocolea inflata 1 0 0 0
Harpanthus flotovianus 5 2 2 3
Hedwigia ciliata 3 2 4 4
Helodium blandowii IND 1 0 1 0
Herzogiella seligeri IND 1 0 1 0
Herzogiella striatella IND 1 0 0 0
Heterocladium dimorphum 3 4 0 0
Homalia trichomanoides IND 1 1 0 0
Hygrohypnum montanum VU 1 0 0 0
Hygrohypnum ochraceum 2 0 0 0
Hylocomiastrum pyrenaicum IND 1 0 0 0
Hylocomiastrum umbratum IND 8 14 6 4 0.004 N
Hylocomium splendens 20 19 20 20
Hypnum andoi 3 1 2 0
Hypnum cupressiforme 5 2 3 1
Hypnum pallescens 4 1 2 1
Isopterygiopsis pulchella 3 1 1 1
Isothecium alopecuroides 3 3 2 1
Isothecium myosuroides 7 10 7 5
Jungermannia leiantha IND 6 1 1 1 0.015 K
Jungermannia obovata 1 0 0 1
Jungermannia pumila 1 0 0 0
Jungermannia sphaerocarpa 1 0 0 1
Kiaeria blyttii 1 0 0 0
Lejeunea cavifolia IND 1 0 0 0
Lepidozia reptans 19 18 14 19
Leptobryum pyriforme 0 0 0 1
Lophozia ascendens IND NT 1 6 0 0 0.005 N
Lophozia bicrenata 1 1 0 0
Lophozia ciliata d 7 11 4 2 0.030 N
Lophozia heterocolpos 0 1 0 0
Lophozia incisa 7 5 2 3
Lophozia longidens 17 19 15 16
Lophozia longiflora 6 11 6 1 0.023 N
Lophozia obtuse 13 14 12 9
Lophozia silvicola 15 19 16 16
Lophozia sp. 0 0 0 1
Lophozia sudetica 5 1 0 1 0.025 K
Lophozia ventricosa 3 5 3 4
Marchantia polymorpha 0 0 1 0
Marsupella emarginata 3 0 0 0
Marsupella sphacelata 1 0 0 0
Metzgeria furcata 2 2 1 1
Mnium hornum 4 0 0 1
Mnium stellare IND 3 3 1 1
Nardia geoscyphus 1 0 0 0
Neckera complanata IND 1 0 0 0
Neckera oligocarpa IND 1 1 1 2
Neckera pennata NT 0 0 1 0
Oncophorus virens 1 0 0 0
Oncophorus wahlenbergii 2 0 0 0
Orthotrichum gymnostomum 3 4 4 3
Orthotrichum obtusifolium 4 4 5 3
Orthotrichum speciosum 5 7 4 5
Oxystegus tenuirostris 1 1 0 0
Paraleucobryum longifolium 12 16 12 9
Pellia epiphylla 5 0 0 2 0.030 K
Pellia neesiana 6 0 0 0 0.002 K
Plagiochila asplenioides 12 9 6 3
Plagiochila porelloides 8 7 2 3
Plagiomnium affine 4 1 1 0
Plagiomnium ellipticum 5 0 1 2

388
Appendix A (Continued )
Species
K. Perhans et al. / Forest Ecology and Management 242 (2007) 374–390
Indicator
(IND) a
Red-list Frequency (number of plots) p-value ISA c
category b Key habitats Nature reserves Old managed Retention
n =20 n =20
forests n =20 groups n =20
Plagiomnium medium IND 2 1 0 0
Plagiothecium curvifolium 20 20 20 20
Plagiothecium denticulatum 11 12 11 4
Plagiothecium laetum 13 8 7 0 0.004 K
Plagiothecium latebricola 0 1 0 0
Plagiothecium piliferum 4 0 0 1
Plagiothecium ruthei 1 0 0 0
Plagiothecium undulatum IND 0 0 1 0
Pleurozium schreberi 20 20 20 20
Pogonatum dentatum 1 0 0 1
Pogonatum urnigerum 3 1 1 7 0.029 R
Pohlia bulbifera 3 0 0 1
Pohlia cruda 6 6 2 7
Pohlia drummondii 1 0 0 0
Pohlia nutans 19 18 18 19
Pohlia proligera 3 0 0 0
Pohlia sp. 1 0 1 1
Pohlia sphagnicola 1 0 0 0
Pohlia wahlenbergii 0 0 0 1
Polytrichastrum longisetum/formosum 8 4 3 1 0.022 K
Polytrichum commune 14 8 11 13
Polytrichum juniperinum 8 7 8 13
Polytrichum piliferum 0 0 0 1
Polytrichum strictum 1 0 0 0
Pseudobryum cinclidioides IND 4 0 0 1
Pseudotaxiphyllum elegans 1 1 0 1
Pterigynandrum filiforme 4 3 1 0
Ptilidium ciliare 14 14 11 8
Ptilidium pulcherrimum 20 20 20 20
Ptilium crista-castrensis 18 19 18 18
Pylaisia polyantha 6 2 2 1
Racomitrium aciculare 3 0 0 0
Racomitrium fasciculare 2 0 0 0
Racomitrium heterostichum 0 1 2 1
Racomitrium microcarpon 6 6 5 8
Radula complanata 8 4 1 1 0.012 K
Radula lindenbergiana 1 0 0 0
Rhizomnium magnifolium 1 0 0 0
Rhizomnium pseudopunctatum 9 3 3 4
Rhizomnium punctatum 12 8 3 4 0.022 K
Rhodobryum roseum 11 4 5 5 0.043 K
Rhytidiadelphus subpinnatus IND 8 2 2 2 0.020 K
Rhytidiadelphus triquetrus 11 11 4 3
Riccardia chamedryfolia 1 0 0 0
Riccardia latifrons 5 1 2 1
Riccardia multifida 0 0 0 1
Riccardia palmate 2 0 0 0
Sanionia uncinata 19 19 18 17
Scapania curta 1 0 0 1
Scapania irrigua 6 0 1 4
Scapania lingulata 5 1 2 0
Scapania mucronata 5 2 0 0 0.024 K
Scapania scandica 6 2 2 2
Scapania subalpina 1 0 0 0
Scapania umbrosa 5 2 1 1
Scapania undulata 6 1 1 1 0.024 K
Schistostega pennata 1 0 0 0
Scorpidium revolvens 0 0 0 2
Sphagnum angustifolium 2 0 1 6 0.028 R
Sphagnum aongstroemii 2 0 0 0
Sphagnum capillifolium 4 0 2 1
Sphagnum centrale 9 1 1 3 0.003 K
Sphagnum fallax 5 1 0 6

Appendix A (Continued )
Species
K. Perhans et al. / Forest Ecology and Management 242 (2007) 374–390 389
Indicator
(IND) a
Red-list Frequency (number of plots) p-value ISA c
category b Key habitats Nature reserves Old managed Retention
n =20 n =20
forests n =20 groups n =20
Sphagnum fimbriatum 2 0 0 0
Sphagnum flexuosum 0 0 1 0
Sphagnum fuscum 0 0 0 1
Sphagnum girgensohnii 13 5 9 7
Sphagnum inundatum 0 0 0 1
Sphagnum magellanicum 2 1 0 0
Sphagnum majus 0 0 0 1
Sphagnum papillosum 1 0 0 2
Sphagnum quinquefarium IND 8 6 6 5
Sphagnum riparium 3 1 1 2
Sphagnum russowii 12 5 8 7
Sphagnum sp. 0 0 0 1
Sphagnum squarrosum 8 2 3 3
Sphagnum subfulvum 0 0 0 1
Sphagnum subnitens 1 0 1 3
Sphagnum teres 2 0 1 3
Sphagnum warnstorfii 1 0 1 1
Splachnum ampullaceum 2 0 0 0
Splachnum luteum 0 0 0 1
Splachnum rubrum 3 2 0 2
Splachnum sphaericum 1 1 0 0
Straminergon stramineum 5 1 0 3
Tayloria tenuis IND NT 1 0 0 0
Tetralophozia setiformis 1 0 0 0
Tetraphis pellucida 20 20 16 18
Tetrodontium ovatum VU 0 2 1 0
Timmia austriaca IND 1 1 0 1
Tortella tortuosa IND 1 0 0 0
Tritomaria quinquedentata 6 2 1 2
Ulota crispa IND 1 0 0 0
Ulota curvifolia 1 0 0 0
Warnstorfia exannulata 1 0 1 3
Warnstorfia sarmentosa 0 0 0 1
Zygodon rupestris 1 0 0 0
a Gärdenfors (2005). EN = endangered, VU = vulnerable, NT = near-threatened, DD = data deficient.
b Nitare (2000).
c ISA = Indicator Species Analysis (McCune and Mefford, 1999). Only shown for p < 0.05 and for species that occurred in at least five plots in any forest category.
The letter indicates to which forest category the species is associated: K = key habitats, N = nature reserves, R = retention groups.
d L. ciliata Damsholt, Söderström and Weibull.
References
Ahti, T., Hämet-Ahti, L., Jalas, J., 1968. Vegetation zones and their sections in
north-western Europe. Ann. Bot. Fenn. 5, 169–211.
Angelstam, P., Mikusinski, G., Eriksson, J.A., Jaxgård, P., Kellner, O., Koffman,
A., Ranneby, B., Roberge, J.M., Rosengren, M., Rystedt, S., Rönnbäck, B.-
I., Seibert, J., 2003. Gap analysis and planning of habitat networks for the
maintenance of boreal forest biodiversity–a technical report from the RESE
case study in the counties of Dalarna and Gävleborg. Department of Natural
Sciences, Örebro University and Department of Conservation Biology,
Swedish University of Agricultural Sciences.
Aubry, K.B., Amaranthus, M.P., Halpern, C.B., White, J.D., Woodard, B.L.,
Peterson, C.E., Laqoudakis, C.A., Horton, A.J., 1999. Evaluating the effects
of varying levels and patterns of green-tree retention: experimental design
of the DEMO study. Northwest Sci. 73, 12–26 (Special).
Bengtsson, J., Angelstam, P., Elmqvist, T., Emanuelsson, U., Folke, C., Ihse, M.,
Moberg, F., Nyström, M., 2003. Reserves, resilience and dynamic landscapes.
Ambio 32, 389–396.
Berglund, H., Jonsson, B.G., 2005. Verifying an extinction debt among lichens
and fungi in northern Swedish boreal forests. Conserv. Biol. 19, 338–348.
Chen, J., Franklin, J.F., Spies, T.A., 1995. Growing season microclimatic
gradients from clearcut edges into into old-growth douglas-fir forests. Ecol.
Appl. 5, 74–86.
Colwell, R.K., 2005. EstimateS: statistical estimation of species richness and
shared species from samples, version 7.5. Available at http://www.purl.oclc.org/estimates.
Diamond, J.M., 1975. The island dilemma: lessons of modern biogeographic
studies for the design of natural reserves. Biol. Conserv. 7, 129–146.
Esseen, P.-A., Ehnström, B., Ericson, L., Sjöberg, K., 1997. Boreal forests. Ecol.
Bull. 46, 16–47.
FAO, 2006. Global Forest Resources Assessment 2005. Progress towards
sustainable forest management. FAO Forestry Paper 147. Food and Agriculture
Organization of the United Nations, Rome.
Franklin, J.F., Lindenmayer, D., MacMahon, J., McKee, A., Magnuson, J.,
Perry, D.A., Waide, R., Foster, D., 2000. Threads of continuity. Ecosystem
disturbance, recovery, and the theory of biological legacies. Conserv. Pract.
1, 8–16.
Gärdenfors, U. (Ed.), 2005. The 2005 Red List of Swedish species. Swedish
Species Information Centre, Swedish University of Agricultural Sciences,
Uppsala.

390
K. Perhans et al. / Forest Ecology and Management 242 (2007) 374–390
Gotelli, N.J., Colwell, R.K., 2001. Quantifying biodiversity: procedures and
pitfalls in the measurement and comparison of species richness. Ecol. Lett.
4, 379–391.
Götmark, F., Thorell, M., 2003. Size of nature reserves: densities of large trees
and dead wood indicate high value of small conservation forests in southern
Sweden. Biodivers. Conserv. 12, 1271–1285.
Gustafsson, L., Hylander, K., Jacobson, C., 2004. Uncommon bryophytes in
Swedish forests—key habitats and production forests compared. For. Ecol.
Manage. 194, 11–22.
Hallingbäck, T., Weibull, H., 1996. En värdepyramid för naturvårdsbedömning
av ädellövskog. Svensk Botanisk Tidskrift 90, 129–140 (In Swedish with
English summary).
Hanski, I., 1999. Metapopulation ecology. In: Oxford Series in Ecology and
Evolution, Oxford University Press, Oxford.
Hazell, P., Gustafsson, L., 1999. Retention of trees at final harvest—evaluation
of a conservation technique using epiphytic bryophyte and lichen transplants.
Biol. Conserv. 90, 133–142.
Hilmo, O., Holien, H., 2002. Epiphytic lichen response to the edge environment
in a boreal Picea abies forest in central Norway. The Bryologist 105, 48–56.
Hylander, K., 2004. Living on the edge: effectiveness of buffer strips in
protecting biodiversity in boreal riparian forests. Doctoral Thesis, Department
of Ecology and Environmental Science, Umeå University, Sweden.
ILOG, 2005. CPLEX System 9.1. ILOG, Gentilly.
Johansson, P., Gustafsson, L., 2001. Red-listed and indicator lichens in woodland
key habitats and production forests in Sweden. Can. J. For. Res. 31,
1617–1628.
Kaila, L., Martikainen, P., Punttila, P., 1997. Dead trees left in clear-cuts benefit
saproxylic Coleoptera adapted to natural disturbances in boreal forest.
Biodivers. Conserv. 6, 1–18.
Krok, T.O.B., Almqvist, S., 2001. Svensk flora: Fanerogamer och
ormbunksväxter, 28th ed. Liber Utbildning, Stockholm (in Swedish).
Margules, C.R., Pressey, R.L., 2000. Systematic conservation planning. Nature
405, 243–253.
MacArthur, R.H., Wilson, E.O., 1967. The theory of island biogeography. In:
Monographs in Population Biology No. 1, Princeton University Press,
Princeton, New Jersey.
McCune, B., Mefford, M.J., 1999. PC-ORD. Multivariate Analysis of Ecological
Data, Version 4. MjM Software Design, Gleneden Beach, Orgeon, USA.
Nelson, C.R., Halpern, C.B., 2005. Short-term effects of timber harvest and forest
edges on ground-layer mosses and liverworts. Can. J. Bot. 83, 610–620.
Nitare, J. (Ed.), 2000. Signalarter. Indikatorer på skyddsvärd skog. Flora över
kryptogamer. Swedish Forest Agency, Jönköping (in Swedish).
Nitare, J., Norén, M., 1992. Woodland key habitats of rare and endangered species
will be mapped in a new project of the Swedish National Board of Forestry.
Svensk Botanisk Tidskrift 86, 219–226 (in Swedish with English summary).
Norén, M., Nitare, J., Larsson, A., Hultgren, B., Bergengren, I., 2002. Handbok
för Inventering av Nyckelbiotoper. Swedish Forest Agency, Jönköping (in
Swedish).
Östlund, L., Zackrisson, O., Axelsson, A.-L., 1997. The history and transformation
of a Scandinavian boreal forest landscape since the 19th century.
Can. J. For. Res. 27, 1198–1206.
Possingham, H., Ball, I., Andelman, S., 2000. Mathematical methods for
identifying representative reserve networks. In: Ferson, S., Burgman, M.
(Eds.), Quantitative Methods for Conservation Biology. Springer-Verlag,
New York, pp. 291–305.
Quinn, G.P., Keough, M.J., 2002. Experimental Design and Data Analysis for
Biologists. Cambridge University Press, Cambridge.
Ranius, T., Kindvall, O., 2006. Extinction risk of wood-living model species in
forest landscapes as related to forest history and conservation strategy.
Landscape Ecol. 21, 687–698.
RIS, 2005. Riksinventering av skog Fältinstruktion 2005. Department of Forest
Resource Management and Geomatics and Department of Forest Soils,
Swedish University of Agricultural Sciences (in Swedish).
Rodrigues, A.S.L., Orestes Cerdeira, J., Gaston, K.J., 2000a. Flexibility,
efficiency and accountability: adapting reserve selection algorithms to more
complex conservation problems. Ecography 23, 565–574.
Rodrigues, A.S.L., Gaston, K.J., Gregory, R.D., 2000b. Using presence-absence
data to establish reserve selection procedures that are robust to temporal
species turnover. Proc. R. Soc. Lond. 267, 897–902.
Santesson, R., Moberg, R., Nordin, A., Tønsberg, T., Vitikainen, O., 2004.
Lichen-forming and Lichenicolous Fungi of Fennoscandia. Museum of
Evolution. Uppsala University.
SAS, 2000. SAS OnlineDoc. Version 8. SAS Institute Inc., Cary, NC.
Shackleton, C.M., 2000. Comparison of plant diversity in protected and
communal lands in the Bushbuckridge lowveld savanna South Africa. Biol.
Conserv. 94, 273–285.
Swedish Forest Agency, 2006. Statistical Yearbook of Forestry, Jönköping.
Söderström, L., Hedenäs, L., 1998. Checklist of Swedish bryophytes. Myrinia 8,
58–90.
Svedlund, L., Löfgren, R., 2003. Protecting the Forests of Sweden: Legal
Protection in the Form of National Parks, Nature Reserves, Habitat Protection
Areas and Nature Conservation Agreements. Swedish Forest Agency
and Swedish Environmental Protection Agency, Stockholm.
Sverdrup-Thygeson, A., 2002. Key habitats in the Norwegian production forest:
a case study. Scand. J. For. Res. 17, 166–178.
Tilman, D., May, R.M., Lehman, C.L., Nowak, M.A., 1994. Habitat destruction
and the extinction debt. Nature 371, 65–66.
Underhill, L.G., 1994. Optimal and suboptimal reserve selection algorithms.
Biol. Conserv. 70, 85–87.
Vane-Wright, R.I., Humphries, C.J., Williams, P.H., 1991. What to protect?—
systematics and the agony of choice. Biol. Conserv. 55, 235–254.
Vanha-Majamaa, I., Jalonen, J., 2001. Green tree retention in Fennoscandian
forestry. Scand. J. For. Res. (Suppl 3), 79–90.
Virkkala, R., Rajasarkka, A., Vaisanen, R.A., Vickholm, M., Virolainen, E.,
1994. The significance of protected areas for the land birds of southern
Finland. Conserv. Biol. 8, 532–544.