See discussions, st ats, and author pr ofiles f or this public ation at : https:www .researchgate.ne tpublic ation259798701 [626764]

See discussions, st ats, and author pr ofiles f or this public ation at : https://www .researchgate.ne t/public ation/259798701
Measu ring Forest Biodiversity at the Stand Scale: An Evaluation of Indicators
in Eu ropean Forest History Gradients
Article    in  Ecologic al Bulle tins · Januar y 2004
DOI: 10.2307/20113319
CITATIONS
57READS
85
2 author s, including:
Some o f the author s of this public ation ar e also w orking on these r elat ed pr ojects:
URBIS :r econnecting people with their long lost nat ural envir onment View pr oject
Value of T ransboundar y Nat ure Pr otected Ar eas Sit uated ne ar the EU Out er Bor ders (TR ANP AREA) View pr oject
Per Ang elstam
Swedish Univ ersity of Agricult ural Scienc es
257 PUBLICA TIONS    10,622 CITATIONS    
SEE PROFILE
All c ontent f ollo wing this p age was uplo aded b y Per Ang elstam on 04 Mar ch 2019.
The user has r equest ed enhanc ement of the do wnlo aded file.

305 Copyright © ECOLOGICAL BULLETINS, 2004Ecological Bulletins 51: 305–332, 2004
The dominant natural vegetation in Europe is forest and
woodland (Mayer 1984, Hannah et al. 1995, Ellenberg
1996). For long the major current threat to its biodiversityis the loss and severe alteration of once naturally dynamic
forests (Stanners and Bourdeau 1995, Hannah et al. 1995,
Peterken 1996, Smith and Gillett 2000, Anon. 2002) andMeasuring forest biodiversity at the stand scale – an evaluation
of indicators in European forest history gradients
Per Angelstam and Monika Dönz-Breuss
Angelstam, P . and Dönz-Breuss, M. 2004. Measuring forest biodiversity at the stand
scale – an evaluation of indicators in European forest history gradients. – Ecol. Bull. 51:
305–332.
We present a system for field measurement of compositional, structural and functional
forest biodiversity indicators at the scale of stands suitable for use in adaptive manage-
ment. T o evaluate the practical usefulness of the indicators we collected data on 21
groups of basic variables in five European case study landscapes representing hemibore-
al, mountain and lowland temperate forests. The stand scale survey sites were stratified
with respect to the land use history both within and among case studies. As a measure-
ment of the cost for monitoring all the variables in one management unit we estimated
that a total of 23–43 person-days were needed, divided into planning (ca 9%), data
collection (ca 70%), data management (ca 15%) and analysis (ca 6%).
We present a sub-set of results focussing on the occurrence of three compositional (spe-
cialised pendulous lichen thalli >20 cm, trees with >80 cm DBH, lying and standing
dead wood), structural (“special” trees, proportion of deciduous trees and old forest)
and functional (uprooted trees, wood-decaying bracket fungi, browsing) indicators of
biodiversity. In general the indicators reflected the trends in the history of forest and
land use both within and among the five case studies. T wo of the indicators stand out as
particularly interesting at the Pan-European scale. These are the amount of dead wood
and the frequency of occurrence of uprooted trees. “Special” trees, old forest and wood-
decaying bracket fungi also performed well, but not always with the same direct rela-
tionship to land use history. T rees with >80 cm DBH showed mixed results. Browsing,
by contrast, appeared to be related to more subtle changes at the regional scale such as
the extirpation of large carnivores and other factors that maintain a high density of large
browsing herbivores. Finally, the specialised species indicator and the proportion of
deciduous trees appeared to indicate the local, but not the regional situation.
T ogether with ecologically founded performance targets for different indicators of the
elements of biodiversity, monitoring results could be used to evaluate the extent to
which biodiversity policies are implemented in actual landscapes.
P . Angelstam (per.angelstam@smsk.slu.se), School for Forest Engineers, Fac. of Forest Scienc-
es, Swedish Univ. of Agricultural Sciences, SE-739 21 Skinnskatteberg, Sweden and Dept of
Natural Sciences, Centre for Landscape Ecology, Örebro Univ., SE-701 82 Örebro, Sweden.
– M. Dönz-Breuss, Dept of Wildlife Biology and Game Management, Univ. of Natural
Resources and Applied Life Sciences, Peter Jordan Str. 76, A-1190 Vienna, Austria.

306 ECOLOGICAL BULLETINS 51, 2004of pre-industrial cultural woodland (Kirby and Watkins
1998, Rackham 2003, Angelstam et al. 2003). Additional-
ly, global change is appearing as a new factor, although
with less predictable consequences (Watson et al. 2000).
Monitoring the status and progress of relevant indicators
of biodiversity is hence a basic prerequisite for the develop-
ment of active adaptive ecosystem management aiming at
implementing sustainable development in practise (Davis
et al. 2001, Meffe et al. 2002, Berkes et al. 2003). The
transition from the classic forest sustainability concept fo-
cussing on wood as a renewable resource, to ecological sus-
tainability based on forest ecosystem management requires
additional data collection of relevant indicators (Angel-
stam 1998a, b, Schlaepfer and Elliott 2000, Duinker
2001). Additionally, new tools for assessment and commu-
nication of these indicators to different stakeholder groups
are needed (Puumalainen et al. 2002, Uliczka et al. 2004,
Ullsten et al. 2004).
The Global Biodiversity Assessment (Heywood 1995),
a knowledge assessment linked to the convention on bio-
logical diversity (Anon. 1992), stressed the need for estab-
lishing monitoring systems for biodiversity. Such monitor-
ing systems can be developed for a variety of ecosystems
and spatial scales ranging from international to local (Lars-
son et al. 2001, Angelstam et al. 2004d). In Europe, the
need to use biodiversity indicators in forest monitoring
programmes has been formalised by the Ministerial Con-
ference on the Protection of Forests in Europe (Ramet-
steiner and Mayer 2004).
In spite of substantial efforts to derive indicators, many
of the broader indicators used at the international, regional
and national scale are not operationally useful at the scale
of the forest management unit (Puumalainen 2001, Angel-
stam et al. 2001, Franc et al. 2001, Larsson et al. 2001).
Duinker (2001) reviewed the problems and pitfalls related
to identification and naming, classification and evaluation,
all of which may hamper indicator development and its
application in practice. T raditionally, the scientific com-
munity has proposed detailed systems for different subsets
of biodiversity elements (e.g. Jonsell et al. 1998, Jonsson
and Jonsell 1999, Nilsson et al. 2001). However, such de-
tailed systems would be considered very costly to imple-
ment in management units, and do not always communi-
cate well to most land managers or to the general public
(Uliczka et al. 2004). On the other hand, some rapid as-
sessment systems are currently used in actual forest man-
agement. For example, Drakenberg and Lindhe (1999)
developed a system originally aimed at education of forest
field staff for rapid assessment of the conservation value of
forest stands. However, a major drawback with these sim-
ple systems is that they are not quantitative, thus often not
enabling a comparison of monitoring results with conser-
vation performance targets.
These are some reasons why practical tools to measure
elements of biodiversity at the scale of the forest manage-
ment unit are still not at hand. The challenge of introduc-ing such a system is to bridge the gap between detailed
scientific approaches to biodiversity monitoring on the
one hand, and the need for a cost-efficient tool that can be
applied and communicated without deep expert knowl-
edge on the other (Hambler 2004). We thus see the devel-
opment of practical biodiversity measurements as a process
where an initial step is to satisfy the need for social licence
to operate, before starting more complicated and costly
scientific approaches (Bunnell and Johnson 1998).
The biodiversity concept is complex, and different
ecoregions such as the boreal, temperate and mountain
forests of Europe have different natural disturbance re-
gimes with many developmental stages (Mayer 1984,
1992, Ellenberg 1996, Angelstam 2003). For application
in practice at the scale of the forest management unit, the
measurements used should be relevant, unambiguous,
and easy to communicate (Duinker 2001). To manage for
forest biodiversity, one needs to stratify forests using the
different natural and cultural disturbance regimes to
which native species are adapted. Therefore, we propose a
top-down approach for selecting several elements of bio-
diversity representing representative disturbance regimes
and forest types (e.g. Angelstam 1998a, b) within differ-
ent forest regions (e.g. Larsson et al. 2001). Additionally,
the data should be simple to collect in the field in a cost-
efficient way, the information should be understandable
for many stakeholders with a minimum of training, and
the method should be applicable throughout the snow-
free season. The sample size should be sufficiently large to
allow detection of differences in various elements of bio-
diversity both among stands at a given time and between
different points in time when repeated measurements
have been made. This would allow the detection of trends
over time and the evaluation of progress in policy imple-
mentation. To stress these aspects, Higman et al. (1999)
argues for a “SMART” selection of indicators that are Spe-
cific, Measurable, Action-oriented, Realistic, and Time-
framed.
Broadly speaking an increasing anthropogenic foot-
print on ecosystems eventually results in reduced species
richness (e.g. Mikusi ński and Angelstam 1998, T rauger et
al. 2003). In forest systems intensively managed for sus-
tained wood yield, even-aged stands of single tree species
dominate. In addition, the amount and quality of dead
wood (Siitonen 2001, Nilsson et al. 2002) and the number
of large trees (Nilsson et al. 2002) are reduced to a mini-
mum. Further, the foliage height diversity is often simpli-
fied to single layers, therefore altering the vertical structure
and thus the suitability of the stand as habitat for a wide
range of species (Brokaw and Lent 1999). Additionally, ec-
osystem processes are altered and ecosystem integrity may
be lost (Pimentel et al. 2000). T o describe the complex
changes, the biodiversity concept and its constituent ele-
ments are useful; for details see Larsson et al. (2001:11 et
seq.). Here, we follow the same logic by using elements
representing the composition, structure and function of

307 ECOLOGICAL BULLETINS 51, 2004biodiversity as outlined by Noss (1990) and later used by
Larsson et al. (2001) and Stokland et al. (2003).
The aim of this paper is two-folded. Firstly, we present a
monitoring system (see Appendix) at the scale of stands
within a landscape, which aims at communicating the
quantity of different elements of biodiversity by selecting
robust variables that can be measured in the field with a
minimum of training. Secondly, we evaluate this system by
testing the idea that management simplifies natural forest
ecosystems in a systematic way. This subsequent empirical
part of the study was conducted using replicated sampling
in land use history gradients in Scotland, W Austria, N
Italy, Poland and Russia.
Gradients in land use history among
and within case studies
The concept of naturalness
Although with an evolutionary perspective change is the
rule, policies such as those related to biodiversity of Euro-
pean forests and woodland make explicit reference to the
concept of naturalness (Anon. 2002, Rametsteiner and
Mayer 2004). Although we are aware of the ambiguity of
this concept (e.g. Balée 1998, Egan and Howell 2001), it is
obvious that forest biodiversity indicators should represent
elements found in naturally dynamic forests (Peterken
1996), or pre-industrial cultural landscapes with semi-nat-
ural woodland components (Kirby and Watkins 1998,
Rackham 2003).
The degree of naturalness of forest ecosystems reflects
the intensity of human interventions (Peterken 1996).
Different levels of utilisation intensity are characterised not
only by changed structures, but also by altered composi-
tion of species’ assemblages. The composition and struc-
ture interact with functional diversity and constitute to-
gether the biological diversity of an area. Forest and other
wooded land where natural processes and species have
been retained or restored have a high conservation value
that has been recognised at the policy level (Rametsteiner
and Mayer 2004). Such forests are also important for un-
derstanding basic ecological principles and can be used as
reference areas when setting up management priorities and
models for sustainable forest management (Lindenmayer
and Franklin 2002, Angelstam and Kuuluvainen 2004).
Both regional comparisons of the human footprint on
nature (Mikusi ński and Angelstam 1998, 2004, Siitonen
2001, Shorohova and T etioukhin 2004, Angelstam et al.
2004a, c) and local case studies (Östlund et al. 1997, Ax-
elsson and Östlund 2001) provide evidence of declines in
different elements of biodiversity following land-use inten-
sification. This suggests that time can to some extent be
replaced with space (Angelstam et al. 1995, Egan and
Howell 2001). In many regions these gradients in histori-cal impact on landscapes can be steep. In Austria, for ex-
ample, Grabherr et al. (1998) showed that 3% of the total
forest area can be classified as natural (without any human
impact), 22% as semi-natural, 41% as moderately altered,
27% as altered and 7% as artificial. These experiences
show that forests can be ranked with respect to their degree
of naturalness (Peterken 1996).
Description of the case studies
Data were collected in five case studies representing three
different European ecoregions north of the Mediterrane-
an, viz. the hemiboreal, mountain, and lowland temperate
ecoregions (Larsson et al. 2001). Following the macroeco-
nomic development from the centre to the periphery of
economic development in Europe (see Angelstam et al.
2004a) we ranked the case studies from the regionally least
to those most impacted by forest management (T able 1).
Within each ecoregion we conducted field studies in local
landscapes representing the scale of forest management
units to cover the range of historical land use types from
natural reference areas (Arcese and Sinclair 1997, Angel-
stam et al. 1997) to altered landscapes. Data collection was
made according to the methodology presented in the Ap-
pendix.
Hemiboreal forest
The human use of the hemiboreal forest has a long and
complex history (Peterken 1996, Kirby and Watkins
1998). The gradual spread of the industrial revolution in
Europe (e.g. Williams 2003), and gradual development of
intensive forest management practices, have resulted in
degradation of the properties of naturally dynamic systems
where the land use history is long (Angelstam et al. 1995).
T oday, intact forest landscapes remain only in the most re-
mote parts of Europe (Fali ński 1986, Yaroshenko et al.
2001). Here we report results from case studies in Scotland
and westernmost Russia.
Abernethy, Scotland – Abernethy (ca 57.2°N, 3.5°W) is
located on the northern slopes of the Cairngorm Moun-
tains in Scotland. In the last 250 yr, the Abernethy forest
experienced dramatic changes (Steven and Carlisle 1959,
Summers et al. 1999). From ancient times the forest cover
was heavily reduced until the 1830s. During the 1840s res-
toration of timber resources began with plantations and
the use of the shelterwood system started. This led to an
increase in forest cover until the 1870s. Later, the forest
cover remained fairly constant (O’Sullivan 1973). In
1866, the duty on imported timber was removed, which
resulted in uneconomic forest management in remote are-
as (Grant 1994). At that time the landowners started to
realise the economic potential of sport hunting, mainly of
red deer Cervus elaphus and roe deer Cervus elaphus and roe deer Cervus elaphus Capreolus capreolus

308 ECOLOGICAL BULLETINS 51, 2004(Dunlop 1997, Smout 1997). This resulted on the one
hand in an increase of forest cover, on the other hand in a
simultaneous increase in the number of deer and a strong
impact on tree regeneration success by deer. From 1850 to
1900, large areas were considered in need of regeneration.
Since some areas were difficult to regenerate naturally, one
had to rely on plantations (Dunlop 1997). In the 20th cen-
tury, forest fellings were mostly done during the two world
wars (O’Sullivan 1973) but also in the 1970s and 1980s
(Summers 1998). A total of 415 sampling plots subdivided
into seven different coarse landscape types were described.
Pskov, W Russia – For very long the forests of the Pskov
area were used for agricultural purposes. Forests were cut
and burned to gain farmland and meadows. In Russia, the
cessation of serfdom in 1861 gradually led to more inten-
sive forest use for local markets. The “farmer and land
bank” system, whereby new free farmers acquired land and
paid by logging, led to increasing forest harvesting between
1906 and 1914. However, logging was mostly selective.
After the revolution in 1917 forests were cut without regu-
lations and logging was generally concentrated to the easily
accessible parts of the landscape. Around 1935 all mature
stands ready for final-felling as well as older had been cut.
During the World War II the forestry activity declined and
harvesting was restricted to the vicinity of roads. After the
war mechanisation started. The first tractor was used in1949 and the first “friendship” chainsaw in 1954. In the
1950s central heating became popular in the urban areas
and villages, which reduced logging considerably and fa-
voured an increased amount of deciduous trees. Simulta-
neously the population on the countryside dwindled, and
fields and meadows were gradually abandoned. Reforesta-
tion after harvesting started only in the 1960s.
During the latter half of the 1990s commercial harvest-
ing increased again. T o demonstrate more nature-friendly
forest harvesting methods, and to advocate the need for
modified forest policies allowing structures to be left for
biodiversity conservation, a model forest project was devel-
oped in Pskov (<http://www.wwf.ru/pskov/eng/>). The
Pskov Model Forest area is located NE of Pskov around the
village Mayakovo (ca 50.10°N, 29.15°E). Our data collec-
tion was carried out in the actual model forest area (ca
18000 ha) as well as in its surroundings, totalling an area
of ca 45000 ha. A total of twenty 1-km2 squares subdivid-
ed into four coarse landscape types were sampled using
320 sampling plots. We ranked the survey plots on forest
land on wet sites as the most natural ones as they have been
traditionally the least accessible ones. Mesic, and in partic-
ular dry sites dominated by Scots pine Pinus sylvestris for- Pinus sylvestris for- Pinus sylvestris
ests, were historically the most important sites for forest
harvesting because of the value of the wood of this tree
species. Abandoned agricultural land was considered as the
most altered stratum.Table 1. Stratification of the land cover types among (rows), and within (columns) in the five European case studies
according to the regional and local forest history from the strata with a long history of use (top/very altered) to the strata
with little expected alteration (bottom/natural).
Forest ecoregion Study area Very altered Altered Near-natural Natural
Hemiboreal Abernethy (1)birch pine savannah, – –
woodland, pine remote,
pine centre, pine remote
Scots pine best
plantation,
exotic
plantation
Mountain Montafon forest centre, forest at the forest in –
cult. landscape periphery extensive use
Mountain Trudner Horn (2)apple orchard, coppice forest, private forest, –
vineyard, larch meadow public forest
agriculture
Hemiboreal Pskov poor site cultural mesic site rich site
landscape
Lowland NE Poland plantation encroaching conif. forest (3)Białowieża
temperate forest on decid. forest (4)national park
former cultural
land
(1) See Frid (2001) for details about the strata.
(2) See Appendix (Table 1 and Fig. 1) for details.
(3) The Knyszy ńska forest, see text and Angelstam et al. (2002).
(4) The Białowieża forest outside the National Park, see the text and Angelstam et al. (2002).

309 ECOLOGICAL BULLETINS 51, 2004Mountain forest
While variation in the use of the hemiboreal and lowland
temperate forests are generally related to differences in ac-
cessibility due to longitude and latitude, local gradients in
the naturalness of mountain forests are often related to alti-
tude (Grabherr et al. 1998).
T rudner Horn, N Italy – South Tyrol (Alto Adige) is locat-
ed in the north Italian Etsch river valley south of the Alps.
The region has a distinct vegetation zonation from low to
high altitude, and an associated variation from higher to
lower intensity of past and present land use (Peer 1995). At
present, 42% of South Tyrol’s land area is forested. Fifty-
two percent of the forest area is owned by private landown-
ers with an average size of 10 ha. Private companies own
16% of the forest area, 29% belong to public bodies (e.g.
villages), 2% to the church, and 1% is county forest (Plon-
er pers. comm.). Other important land cover types are or-
chards and vineyards, fields and meadows, as well as larch
meadows and alpine pastures. Norway spruce Picea abies
(62%), larch Larix decidua (18%), Scots pine (11%), stone Larix decidua (18%), Scots pine (11%), stone Larix decidua
pine Pinus cembra (5%) and silver fir Pinus cembra (5%) and silver fir Pinus cembra Abies alba (3%) are Abies alba (3%) are Abies alba
the economically most important tree species (Ploner pers.
comm). The proportion of deciduous trees is low (1%).
The nature park T rudner Horn forms the core of the
study area and is located ca 20 km south of Bolzano (ca
46.6°N, 11.3°E). Established in 1980 it covers an area of
69 km2. The park hosts the most diverse and species-rich
area in the region, and ranges from sub-mediterranean to
alpine vegetation. It includes a wide range of land use
forms from vineyards and ancient coppice forest to agricul-
ture, managed forests and larch meadows. T o cover the
main landscape types of the region additional sampling
plots were located outside the nature park. Based on to-
day’s characteristic combination of natural and anthropo-
genic factors we subdivided the study area into 7 different
sampling strata excluding tree line forest, which is not re-
ported here (see Appendix for details). We collected data in
21 different km2 plots, the total number of plots surveyed
was 290.
Montafon, W Austria – The Montafon valley is located in
the southern part of Vorarlberg, the westernmost province
of Austria (ca 47.1°N, 9.9°E). The valley consists of 10
municipalities with an area of 563 km2 and a total popula-
tion of 18000 inhabitants. About 50% of the area is cov-
ered by alpine meadows, 23% by forest, 20% is alpine hab-
itat above the tree-line and 7% agricultural and urban
land. The forests reach up to ca 1800 m a.s.l., and cover a
total area of 16000 ha. Winter tourism is the main source
of income. In historical times, mining was one of the most
important local industries. Due to the high demand of
timber for mining virtually no naturally dynamic forest is
left (Grabherr et al. 1998). However, due to the dramatic
topography altered landscapes as settlements at the bottom
of the valleys alternate with near-natural areas such as pro-tected forest in steep terrain. T oday, >80% of the forest
fulfil a protective function for the site itself and for settle-
ments in the valleys. Approximately 33% of the area is in
slopes steeper than 45°.
In the Montafon valley, the distribution of forest types
is mainly determined by altitude. In the valley floor, and
up to 1000 m a.s.l., the forest is dominated by deciduous
(beech Fagus silvatica, maple Acer pseudoplatanus, lime
Tilia cordata, ash Fraxinus excelsiorash Fraxinus excelsior ash ) and mixed forests Fraxinus excelsior ) and mixed forests Fraxinus excelsior
(Norway spruce, beech and silver fir). Above 1000 m
spruce forests predominate. Larch and stone pine can only
be found close to the tree-line. The total number of sam-
pling plots was 324 (25 km2 plots) distributed among four
coarse landscape types (of which the stratum tree line has
been excluded for the analyses).
Lowland temperate forest
From the Atlantic Ocean to the Ural Mountains, central
Europe is a lowland plain. Due to a benign climate for
agriculture and easy access only ca 0.2% of the once wide-
spread forest can still be considered intact (Hannah et al.
1995). Reference areas are thus hard to find. An exception
can be found in NE Poland, where the Bia ł can be found in NE Poland, where the Bia ł can be found in NE Poland, where the Bia owieża forest is
located (Fali ński 1986).
NE Poland – We sampled five different landscape types in
NE Poland. These were the Bia ł NE Poland. These were the Bia ł NE Poland. These were the Bia owieża National Park,
managed forest in Bia ł managed forest in Bia ł managed forest in Bia owieża outside the national park,
managed coniferous forest in the Knyszy ń managed coniferous forest in the Knyszy ń managed coniferous forest in the Knyszy ska forest, pine
plantations in the Biebrza valley and the Knyszy ń plantations in the Biebrza valley and the Knyszy ń plantations in the Biebrza valley and the Knyszy ska for-
est, and encroaching deciduous forest on former agricul-
tural land (for details see Angelstam et al. 2002). The total
number of sampling plots was 402.
Due to its remote location, the Bia ł Due to its remote location, the Bia ł Due to its remote location, the Bia owieża forest
(53.1°N, 23.5°E) has undergone much less dramatic
changes than other forests of NE Poland. The area has
been used as a hunting ground since the 15th century, set
aside for Lithuanian dukes, Polish kings, Russian tsars and
German nazis (J ędrzejewska and J ędrzejewski 1998). The
core area of the Polish section of this 1250 km2forest area
was declared as a national park in 1921 (Fali ński 1986,
Vera 2000). The Polish part covers 580 km2 of which 47
km2 is strictly protected with tourism and research as the
only permitted activities (J ędrzejewska and J ędrzejewski
1998). The forest types range from fresh to wet and from
coniferous to mixed and deciduous stands. Riparian forests
and alder Alnus glutinosa swamps are also present in some Alnus glutinosa swamps are also present in some Alnus glutinosa
parts. The deciduous “gr ąd” forest (oak-lime-hornbeam
Quercus spp. – Quercus spp. – Quercus Tilia spp. – Tilia spp. – Tilia Carpinus betulus ) dominates inside Carpinus betulus ) dominates inside Carpinus betulus
BiałBiałBiaowieża National Park (Fali ński 1986). The old-growth
deciduous forest is characterised by a multi-layered structure
from small plants and seedlings to trees up to 40 m tall.
Most of the forest outside the National Park has been
exploited or partially cleared in historical times. Neverthe-
less, the forest as a whole has undergone less dramatic

310 ECOLOGICAL BULLETINS 51, 2004changes than other forests of NE Poland (Fali ński 1986,
Jędrzejewska and J ędrzejewski 1998, Vera 2000). As early
as in the third century, birch Betula spp. and hornbeam
were used for charcoal production. Forest management
started in the 16th century in the forest fringes. At the end
of the 17th century the exploitation of the forest increased.
Massive timber exploitation occurred during World War I
both by Germans and the British European Timber Cen-
tury Corporation. The intensity of logging has since then
slowed down although harvesting is still carried out in
most of the forest. The same forest types can be found as in
BiałBiałBiaowieża National Park, but in different proportions
and with structural differences mainly due to past and on-
going logging activities.
The Knyszy ń The Knyszy ń The Knyszy ska Forest situated between the Biebrza
marshes and Bia ł marshes and Bia ł marshes and Bia owieża forms a large continuous forest
block (Sokolska and Leniec 1996) north and east of
BiałBiałBiaystok (53.3°N,łystok (53.3°N, ł 23.2°E). Large parts of the forest grow
on acidic sandy soils, which are not favourable to decidu-
ous trees. Scots pine makes up 70% of the standing vol-
ume in that forest. Another 10% consist of Norway spruce
and ca 20% are deciduous tree species, mostly birch, oak,
and alder. The Knyszy ń and alder. The Knyszy ń and alder. The Knyszy ska Forest has been an important
timber growing area for a long time. Pine has been export-
ed to the Netherlands and U.K. since the 16th century.
The most extensive felling occurred in 1915–1918. Refor-
estation is mainly done in pure pine stands while natural
regeneration is allowed on other sites.
The Biebrza marsh (ca 53.4°N, 22.6°E) is an ancient
cultural landscapes gradually being abandoned. Encroach-
ing deciduous forest ( Salix spp., Salix spp., SalixBetula spp., Betula spp., BetulaAlnus spp.) isAlnus spp.) isAlnus
common on old formerly mowed and grazed grasslands.
The pine plantations in Biebrza were mainly established
after World War II. These plantations are characterised by
their even age and are in most cases pine monocultures.
Natural regeneration is either absent or very poor and un-
dergrowth is missing. The pine plantations in Knyszy ń dergrowth is missing. The pine plantations in Knyszy ń dergrowth is missing. The pine plantations in Knyszy ska
Forest are generally older than those in Biebrza and contain a
higher proportion of other trees species, mainly spruce.
Stratification of regional and local gradients
Incorporating both the regional economic history among,
and the local history within the different case studies, we
stratified the data into different groups ranging from the
centre (very altered landscapes) to the periphery (near-nat-
ural landscapes) of economic development (see above and
Table 1).
Methodology
Based on a review of literature, interviews with forest man-
agers and field trials we propose a system of measurements
of major elements of biodiversity for boreal, hemiborealand central European coniferous forests (see Appendix).
The elements are chosen based on the idea that the devel-
opment of biodiversity measurements must proceed from
existing forest management data (e.g. tree species composi-
tion, wood volume and site type), by gradually adding new
variables and measurements (Angelstam 1998b) shown to
be characteristic of natural reference areas (Mayer 1984,
Faliński 1986, Peterken 1996). Such indicators should in-
clude elements representing composition, structure and
function of forests (Larsson et al. 2001).
The comprehensive system presented in the Appendix
is a synthesis of several existing approaches applied in con-
temporary practical forest management. These include
forest taxation (Reed and Mroz 1997), indicator species
for high conservation value forests (Thompson and Angel-
stam 1998, Nilsson et al. 2001, Norén, et al. 2002) and
evaluation of high conservation value by careful observa-
tion of compositional and structural elements of biodiver-
sity (Drakenberg and Lindhe 1999). Additionally we in-
troduce some new indicators regarding processes affecting
the maintenance and renewal of forest habitats. By and
large the approach follows that of the EC-funded BEAR
project on the development of biodiversity evaluation tools
proposed by Larsson et al. (2001). The target user is the
manager of the local landscape in the form of the small
forest owners in a village, or a management unit of a com-
pany that wants to practice adaptive management and thus
start collecting information about the status, and if repeat-
ed trends, of indicators of different biodiversity elements.
T o provide an idea of the cost of applying our methodolo-
gy in a management unit we estimated the total number of
working days it took to carry out the five different case
studies, respectively. In this paper we report on three indi-
cators for each of the three groups of biodiversity elements
(T able 2).
According to the Appendix our method should result in
a sample size that is a multiple of the 16 survey plots in
each 1 km2 square. This is, however, not always the case in
our data. For example, low forest cover in Abernethy
meant that we adjusted the spatial pattern of the survey
plots, and in some of the other study areas some sites were
simply not accessible. T o test for statistical differences
among strata in the different case studies we used χ2-tests
for data based on the frequency of occurrence of different
elements of biodiversity, and t-tests and ANOVA for com-
parisons of the basal area of dead wood and the proportion
of deciduous trees. In the analyses we consider each plot to
be an independent sample. However, given that the plots
are clustered and separated by only 250 m within a cluster,
there is a risk for spatial pseudoreplication if the forest
stands are large. However, because the stands in hemibore-
al, mountain and nemoral forest are usually only a few hec-
tares in size, and the main aim of this paper is to present
the methodology as such and an overview of the results of
a small subset of the data, we do not consider spatial pseu-
doreplication a problem.

311 ECOLOGICAL BULLETINS 51, 2004Results
Composition
The frequency of occurrence of long (>20 cm) pendulous
lichen thalli varied significantly among the different strata
in T rudner Horn, Montafon and Pskov (Table 3, Fig. 1).
However, the direction of change differed and was not
consistent with respect to the degree of naturalness. In Ab-
ernethy there was no significant difference between thetwo strata, and in NE Poland long lichens were not ob-
served at all. In the two mountain forest case studies (T rud-
ner Horn and Montafon) the frequency of occurrence in-
creased with increasing naturalness. In Pskov there was no
clear trend.
Regarding trees with a DBH >80 cm the frequency of
occurrence in the plots was generally low (<10%), except
for the “altered” strata in Abernethy and the “near-natural”
and “natural” strata in NE Poland. The difference among
the strata was significantly different in both these case stud-
Table 3. Results regarding compositional elements of biodiversity in five European case studies. Note that in the case
studies of Trudner Horn and Montafon there was no natural stratum, and that in Abernethy the near-natural and natural
strata did not exist.
Abernethy Trudner Horn Montafon Pskov NE Poland
(n) (n) (n) (n) (n)
Lichens >20 cm (occurrence, %)
very altered 1.9 (206) 2.6 (76) 17.4 (161) 14.6 (96) 0.0 (80)
altered 5.3 (209) 11.9 (67) 62.9 (89) 0.0 (80) 0.0 (82)
near-natural – 19.1 (147) 51.4 (74) 3.1 (64) 0.0 (160)
natural – – – 8.9 (80) 0.0 (80)
χ2= 3.3 χ2= 12.0 χ2= 51.3 χ2= 15.9 –
p = 0.07 p = 0.003 p<0.001 p = 0.001 –
Trees >80 cm DBH (occurrence, %)
very altered 1.9 (206) 2.6 (76) 5.0 (161) 0 (96) 0.0 (80)
altered 22.5 (209) 7.5 (67) 9.0 (89) 3.8(80) 0.0 (82)
near-natural – 7.5 (147) 6.8 (74) 4.7 (64) 15.6 (160)
natural – – – 1.3 (80) 57.5 (80)
χ2= 40.6 χ2= 2.3 χ2= 1.5 χ2= 5.3 χ2= 122.5
p<0.01 p = 0.32 p = 0.46 p = 0.15 p<0.001
Sum dead wood (basal area, m2 ha–1)
very altered 0.7 (206) 0.1 (76) 0.8 (161) 2.6 (96) 0.3 (80)
altered 1.9 (209) 0.9 (67) 2.7 (89) 4.6 (80) 1.4 (82)
near-natural – 1.0 (147) 4.1 (74) 6.0 (64) 2.6 (160)
natural – – – 6.5 (80) 9.7 (80)
t = 6.5 F = 7.1 F = 42.1 F = 16.4 F = 119
p<0.001 p<0.001 p<0.001 p<0.001 p<0.001Table 2. List of variables representing different biodiversity indicators analysed in this study.
Biodiversity Variable Description and unit for the survey plot data.
element
Composition Lichens >20 cm Occurrence (%) of pendulous lichen thalli >20 cm
Trees >80 cm DBH Occurrence (%) of trees with >80 cm DBH
Dead wood Basal area of standing and lying dead wood >10 cm DBH
Structure Special trees Occurrence (%) of moss and lichen-covered, bent, damaged,
hollow and forked trees
Deciduous trees >10 cm Proportion of living deciduous trees of all living trees >10 cm
Old forest stands Occurrence (%) of stands with “ageing” or “old-growth” age
classes
Function Uprooting Occurrence (%) of uprooted trees
Wood-decaying bracket fungi Occurrence (%) of wood-decaying bracket fungi
Browsing Occurrence (%) of browsing by ungulates

312 ECOLOGICAL BULLETINS 51, 2004studies (T able 3 and Fig. 1). In NE Poland the stratum con-
sisting of the Bia ł sisting of the Bia ł sisting of the Bia owieża National Park was outstanding with
a frequency of occurrence of trees with a DBH > 80 cm of
58%
The total basal area of standing and lying dead wood
was significantly different among the strata in all five case
studies (Table 3). In addition, in all study areas, the basal
area increased consistently with increasing naturalness as
expected from the local history of each stratum (Fig. 1).
The range of basal areas among the strata in the different
case studies was 30 to 100-fold, from 0.1 m2ha–1ha–1ha in the
very altered strata in T rudner Horn and 0.3 m2ha–1ha–1ha in for-
est plantations in NE Poland, to 9.7 m2ha–1ha–1ha in the
BiałBiałBiaowieża National Park.Structure
The frequency of occurrence of special trees (trees with cer-
tain structures such as different microhabitats typical for
natural forests) as defined in Table 2 and the Appendix was
significantly different among the strata in all five case stud-
ies (T able 4). The frequency of occurrence increased con-
sistently with increasing naturalness as expected from the
local history of each stratum (Fig. 2). In the “near-natural”
and “natural” strata the frequency of occurrence was gener-
ally ca 80% or higher.
The proportion of living deciduous trees of all living
trees varied significantly among the strata in all five case
Fig. 1. Differences in the three compositional indicators of biodi-
versity: frequency of occurrence of pendulous lichen (top), trees
with >80 cm DBH (middle) and basal area of standing and lying
dead wood (bottom) in the four different strata (see T able 1) in
five European case studies. Note that in NE Poland pendulous
lichens were not found at all.
Fig. 2. Differences in the three structural indicators of biodiversi-
ty: frequency of occurrence of special trees (top), proportion of
living deciduous trees of all living trees with >10 cm DBH (mid-
dle) and biologically old stands (ageing and old-growth) (bot-
tom) in the four different strata (see T able 1) in five European
case studies.

313 ECOLOGICAL BULLETINS 51, 2004studies (Table 4). The case studies in regions with a long
management history located in hemiboreal (Abernethy) or
mountain forest (T rudner Horn and Montafon) had gen-
erally lower proportions of deciduous trees than the case
study in lowland temperate forest zone (NE Poland), and
Pskov where forest management has a short history (Fig.
2). Within the individual case studies the highest propor-
tions of deciduous trees were found in the “very altered” or
“altered” strata, which were usually located at lower alti-
tude on richer soils.
The proportion of plots having check-marks indicating
the presence of ageing or old-growth stands varied signifi-
cantly among the strata in Abernethy, T rudner Horn,
Montafon and NE Poland, but not in Pskov (Table 4). In
Pskov the data indicate that there was generally a low oc-
currence of older stands. In the other case studies the fre-
quency of occurrence of older stands increased consistently
with increasing naturalness as expected from the local his-
tory of management of each stratum.
Function
The frequency of occurrence of both uprooted trees and
bracket fungi varied significantly among the strata in all
five case studies (Table 4, Fig. 3). Except in Abernethy, thevalues increased consistently with increasing naturalness as
expected from the local history of each stratum.
Finally, the incidence of signs of browsing damage by
ungulates indicated large differences among the five case
studies. In Abernethy the browsing pressure could not be
assessed because of the absence of young trees due to a very
long history of intensive browsing by red deer and sheep
(Watson 1983, Smith 1993). In the two mountain forest
case studies the browsing damages increased with increas-
ing naturalness of the stratum (T able 5). While there was a
generally low level of browsing (ca 30% or less), and no
significant difference among the strata in Pskov, there was
a significant difference among strata and a very high level
of browsing (>85%) in NE Poland.
Costs
With 5–7 strata within each forest management unit and a
large (>50) sample in each stratum, the total number of
person-days for applying the complete version of system of
biodiversity indicators (as described in the Appendix) was
estimated at 23 (flat terrain in Abernethy) to 43 (steep ter-
rain in Montafon) working days. This total effort can be
divided into planning (ca 9%), data collection (ca 70%),
data handling (ca 15%) and analysis (ca 6%) (Table 6).Table 4. Results regarding structural elements of biodiversity in five European case studies. Note that in the case studies of
Trudner Horn and Montafon there was no natural stratum, and that in Abernethy the near-natural and natural strata did
not exist. Regarding the n-values see Table 3.
Abernethy Trudner Horn Montafon Pskov NE Poland
Special trees (occurrence, %)
very altered 55.3 26.3 75.8 57.3 45.0
altered 73.7 88.1 97.8 77.5 45.1
near-natural – 94.5 97.3 78.1 60.6
natural – – – 78.8 95.0
χ2= 15.3 χ2= 132.8 χ2= 33.4 χ2= 14.5 χ2= 56.3
p<0.001 p<0.001 p<0.001 p = 0.002 p<0.001
Decid. trees >10 cm DBH (proportion of all living trees, %)
very altered 27.9 26.3 30.2 27.5 6.2
altered 14.6 34.7 6.0 74.4 84.7
near-natural – 11.7 13.4 66.7 45.1
natural – – – 58.8 66.5
F = 9.6 F = 13.7 F = 19 F = 41.1 F = 92.8
p<0.001 p<0.001 p<0.001 p<0.001 p<0.001
Old forest stands (occurrence, %)
very altered 20.9 17.1 32.3 7.3 1.3
altered 81.3 46.3 56.2 8.8 7.3
near-natural – 60.0 54.0 15.6 49.4
natural – – – 8.8 96.3
χ2= 152 χ2= 37.0 χ2= 17.4 χ2= 3.4 χ2= 197.0
p<0.001 p<0.001 p<0.001 p = 0.34 p<0.001

314 ECOLOGICAL BULLETINS 51, 2004Note that in this paper only a small subset of all the indica-
tors described in the Appendix are reported.
Discussion
Patterns in biodiversity indicators within and
among case studies
We found large differences in the state of indicators of dif-
ferent elements of biodiversity both within and among the
five European case studies. In general the results from ana-
lysing a limited subset of 9 different indicators confirm the
general pattern of impoverished forest biodiversity with
increasing human impact. Broadly speaking, this is related
to both clearing of productive sites for agricultural purpos-
es throughout history, and the local and international de-
mand for timber. This has resulted in several frontiers ofuse, ranging from high-grading to intensive silviculture
spreading from the central to peripheral areas of Europe
(Angelstam et al. 1995, Yaroshenko et al. 2001). Such his-
torical gradients can be found at several spatial scales. First,
with an economic history perspective, western Europe has
a longer and more intensive history of land use than east-
ern Europe (Gunst 1989). Second, in the lowland temper-
ate ecoregion clearing for agricultural purposes has created
large variations in forest cover both among and within re-
gions due to a combination of site characteristics and so-
cio-economic changes (Darby 1956). Finally, in mountain
forests physical inaccessibility and altitude has gov erned the forests physical inaccessibility and altitude has gov erned the forests physical inaccessibility and altitude has gov
intensity of forest use making virtually every valley into a for-
est history gradient illustrating the complex interaction be-
tween nature and man (Grabherr et al. 1998).
The classification of indicators as compositional, struc-
tural and functional (Noss 1990, Larsson et al. 2001) is
appealing, but always open for discussion. For example, a
certain tree species or “special” trees can be viewed both as
a component with the basal area recorded, and as a struc-
tural element and the proportion of all trees and frequency
of occurrence calculated, respectively. Similarly, bracket
fungi can be viewed as a compositional element if individ-
ual species are recorded, and as a functional element if their
role as wood decaying organisms is stressed. These prob-
lems are discussed in detail in the Appendix.
Composition
Species are probably the best known elements of biodiver-
sity. More detailed studies of the occurrence of different
species groups such as mosses and lichens have demon-
strated clear differences in the frequency of occurrence of
specialised species in relation to the naturalness (often ap-
proximated as stand age) of forest stands (Gustafsson
2002). Such studies have, however, generally been made in
local study areas as regular research projects, and not over
large regions including considerable biogeographic varia-
tion in the distribution of different species. The highly var-
iable frequency of occurrence of tall pendulous lichen thal-
li among the five case studies indicates that the use of even
such a coarse species indicator is not feasible at the Pan-
European scale. Specialised lichen indicator species groups
such as pendulous lichens of the genera Alectoria ,Bryoria
andUsnea , or species with cyanobacteria as photobionts,
are adapted to a light and moist microclimate (see Hedenås
2002). Their local occurrence can thus be dependent on
presence of microclimatic features such as forest interior
old-growth stands and steep shaded ravines, but also on a
generally moist macroclimate due to an oceanic climate or
high altitude (Rolstad et al. 2001). Thus a high frequency
in one landscape may reflect the same degree of naturalness
as a low frequency in another.
Analyses of the habitat characteristics of endangered
forest species show that they often depend on natural forest
components such as large trees and dead wood (Berg et al.Fig. 3. Differences in the three functional indicators of biodiversi-
ty: frequency of occurrence of uprooted trees (top), bracket fungi
(middle) and browsing by ungulates (bottom) (y-axis) in the four
different strata (see Table 1) in five European case studies. Note
that in Abernethy, Scotland there were no shrubs or young trees
based on which the browsing pressure could be estimated.

315 ECOLOGICAL BULLETINS 51, 20041994). These features are characteristic for old forests and
for natural disturbance regimes in a wide range of forest
ecosystems (Peterken 1996, Nilsson et al. 2002), but have
had to give way to even-aged single-species stands in man-
aged systems (Esseen et al. 1997). The results of this study
confirm this. Regarding large trees (>80 cm DBH), Aber-
nethy and NE Poland stand out as being the only two case
studies having strata including large nature reserves. Com-
pared to this, the strata in the remaining case studies have a
very small proportion of large trees.
The sum of the basal area of standing and lying dead
wood appears to be one of the indicators that performs
most consistently in relation to both the local stratificationwithin case studies as well as among them. NE Poland is a
particularly clear example. The basal area measurements
suggest about a 30-fold difference between the forest plan-
tations on the one hand and the Bia ł tations on the one hand and the Bia ł tations on the one hand and the Bia owieża National Park
on the other. A simplified estimate of the total wood vol-
ume suggests that the difference is even larger. Using the
mean tree height of the forest in the two strata the volume
of dead wood in plantation forests can be estimated (vol-
ume = basal area × height × trunk form factor; Hamilton
1982) to ca 170 m3ha–1ha–1ha in the Bia ł in the Bia ł in the Bia owieża National Park
and to ca 3 m3ha–1ha–1ha in forest plantations in NE Poland,
suggesting a 60-fold difference in the total amount of dead
wood. This is consistent with the estimates of Siitonen
Table 6. Number of strata, plots and person-days for application of the collection of data as described in the Appendix in
five European case studies. Because this study focuses on strata where the natural vegetation is productive forest, the tree
line strata in Trudner Horn (see Appendix) and Montafon were excluded from the analyses. Note also that some of the
original strata have been merged (Table 1).
No. of strata Plots per Number of person days
(see Table 1) day
planning field work enter data analyses # days total
Abernethy 7 32 3 13 5 2 23
Trudner Horn 7 12 3 24 5 2 34
Montafon 6 10 3 33 5 2 43
Pskov 4 15 3 22 5 2 32
NE Poland 5 14 3 28 5 2 38Table 5. Results regarding functional elements of biodiversity in five European case studies. Note that in the case studies
of Trudner Horn and Montafon there was no natural stratum, and that in Abernethy the near-natural and natural strata did
not exist. Regarding the n-values see Table 3.
Abernethy Trudner Horn Montafon Pskov NE Poland
Uprooted trees (occurrence, %)
very altered 7.8 5.3 32.3 37.5 21.3
altered 17.2 31.3 62.9 47.5 39.0
near-natural – 36.7 74.3 78.1 53.8
natural – – – 82.5 86.3
χ2= 8.5 χ2= 25.8 χ2= 43.6 χ2= 50.8 χ2= 73.3
p<0.04 p<0.001 p<0.001 p<0.001 p<0.001
Bracket fungi (occurrence, %)
very altered 13.6 2.6 38.5 54.2 47.5
altered 2.9 16.4 39.3 75.0 67.1
near-natural – 22.5 66.2 84.4 80.7
natural – – – 86.3 98.8
χ2= 15.9 χ2= 14.8 χ2= 17.4 χ2= 29.0 χ2= 60.6
p<0.001 p<0.001 p<0.001 p<0.001 p<0.001
Browsing (occurrence, %)
very altered – 15.8 58.4 14.6 87.5
altered – 68.7 93.3 21.3 95.1
near-natural – 81.6 73.0 31.3 98.8
natural – – – 27.5 97.5
– χ2= 93.6 χ2= 34.2 χ2= 7.4 χ2= 16.7
p<0.001 p<0.001 p = 0.06 p<0.001

316 ECOLOGICAL BULLETINS 51, 2004(2001) and Nilsson et al. (2002) suggesting a decline of
dead wood and large trees by up to two orders of magni-
tude when comparing the historic range of variation with
managed forests.
Structure
The results from the five case studies show that even if the
frequency of “special” trees increased with decreasing his-
toric human footprint, they were relatively common in
most strata. However, neither of the study areas can be
characterised as having intensive forest management,
which is the praxis for example in Sweden’s and Finland’s
boreal forests. Therefore, “special” trees have not been re-
moved earlier in the succession during pre-commercial
and commercial thinning.
Agriculture creates pressure on the natural forest envi-
ronment, but also plays an important role in maintaining
cultural woodland and semi-natural habitats (Tappeiner et
al. 2003). For some forest species depending on early and
un-managed stages in the natural succession of deciduous
tree species, the abandonment of fields and wooded grass-
lands may provide increased amount of habitat
(Mikusiński et al. 2003). This was evident for example in
NE Poland in the altered stratum consisting of abandoned
previously managed grasslands in the Biebrza area. Here
encroaching deciduous trees maintain several woodpecker
species otherwise only found in naturally dynamic forests
(Angelstam et al. 2002). In Abernethy, Pskov and NE Po-
land this was associated with recent abandonment of
fields, meadows or pastures. In the case study T rudner
Horn the results also reflect the local land-use history
where abandoned coppice forests are dominating in the
surroundings of settlements. This is also true for the case
study Montafon where collection of forest litter was made
close to the villages. The strata altered and near-natural
were located on higher elevation and deciduous trees are
therefore less common in these strata.
Regarding old forest stands, except for Pskov the fre-
quency of occurrence of old forest stands increased with
decreased alteration as predicted by the local forest and
land use history. The much lower amounts of old forest
stands in Pskov is probably also due to the intensive war-
fare during the World War II that took place in the area in
the early 1940s. According to Barraclough (1993) the
front line between German and Russian troops was located
just NE of the Pskov study area. Still today, the area shows
many signs of the war. In the study area Montafon and
T rudner Horn, old forest stands are mostly located far
away from the settlements on higher elevations. Especially
in Montafon, old forest stands remain in locally remote
areas, today playing an important role for biodiversity
(Dönz-Breuss et al. 2004). In Abernethy, the frequency of
occurrence of old forest stands was very high in the altered
stratum (80%). This can be explained by the existing Na-
ture Reserve in this stratum.Function
Uprooting is an important process whereby micro-scale
primary succession can take place on the exposed root
plates (Read 2000). While strong winds in principle may
cause uprooting of trees and create root plates anywhere,
the harvesting of trees precludes this. In our case studies,
uprooted trees showed a pattern of occurrence that was
very similar to the total amount of dead wood with lower
frequency in the very altered strata and an increase in the
frequency of occurrence in the more natural strata. Still,
however, uprooted trees occurred throughout all strata and
case studies. In the case studies in the Alps (T rudner Horn
and Montafon) avalanches can also cause uprooting.
In general the frequency of occurrence of wood-decay-
ing bracket fungi paralleled both the amount of dead wood
and the occurrence of uprooted trees. This is quite natural
as all three properties are closely linked. However, Scotland
was an exception with a decreasing frequency with less in-
tensive history of management. Note that the overall fre-
quency of occurrence was much lower than in the other
case studies.
The browsing level varied more among than within the
case studies. Scotland was one extreme with such a long
history of browsing on young trees that they were absent
and this indicator could not be surveyed. In the two
mountain forest case studies high browsing intensity by
wild ungulates and livestock was associated with the occur-
rence of relatively large forest tracts, high densities of ungu-
lates and a low density of human population and transport
infrastructure. In the Alps functional populations of large
carnivores are absent (Breitenmoser 1998) and hunting
regulates ungulate populations. The relatively low brows-
ing level in the Pskov case study is associated with the pres-
ence of viable populations of large carnivores, which limits
the density of large herbivores (cf. Lindén et al. 2000). In
NE Poland the browsing pressure was high in all strata.
Studies of the interaction between populations of large
mammals thus requires macroecological approaches where
landscape-scale studies are replicated (Angelstam et al.
2000, Berger et al. 2001, Ripple and Beschta 2003).
Selection of indicators and monitoring to
learn
Monitoring of the state and trends of different elements of
biodiversity can be made at several levels ranging from the
policy level to the level of forest management units (An-
gelstam et al. 2004d). In addition, research projects are
needed to validate the links between cause and effect. This
three-phased approach to monitoring thus encompasses
implementation monitoring, effectiveness monitoring
and validation monitoring, respectively (Busch and T rex-
ler 2003). This study focuses on effectiveness monitoring,
where monitoring is used to learn rather than learning to
monitor (Gunderson 2003). The target group for the

317 ECOLOGICAL BULLETINS 51, 2004present approach is thus the non-expert who wants a rap-
id assessment technique that is not too costly, neither in
terms of time to learn the approach nor to perform it.
Based on the discussion of the subset of 9 indicators of
different elements of biodiversity we conclude that most
of them did reflect the trends in the history of forest and
land use both within and among the five case studies. T wo
of the indicators stand out as particularly interesting as
forest naturalness indicators at the European scale. These
were the total amount of dead wood and the frequency of
occurrence of uprooted trees. “Special” trees, old forest
and wood-decaying bracket fungi also performed well in
principle, but not always in the same smooth and gradual
way in relation to land use history as the previous two
indicators. Very large trees (>80 cm DBH) showed mixed
results, probably due to regional differences in factors af-
fecting the size of trees. Similarly, browsing appears to be
related to regional scale changes such as the extirpation of
large carnivores and other factors that maintain a high
density of large wild or domestic herbivores. The special-
ised lichen species indicator appeared to indicate the local,
but not the regional situation. This indicator was thus
more restricted at the Pan-European level, even if it has
been shown to be very useful locally (e.g. Nilsson et al.
1995). Species are also important from an educational
point of view. This, however, does not necessarily require
that the species themselves are used as indicator, but
knowledge about species, especially those which are val-
ued and well studied, is an effective way of communicat-
ing knowledge to different stakeholders (Uliczka et al.
2004). Finally, the proportion of deciduous trees was
largely unrelated to the estimated degree of naturalness in
all case studies. Nevertheless, the proportion of deciduous
trees may be a relevant measurement of the local biodiver-
sity (Mikusi ński et al. 2003). This indicator thus suggests
a potential conflict between naturalness and cultural land-
scapes as baselines for conservation. Alternatively, it stress-
es the need to include processes such as land abandon-
ment as biodiversity indicators of disturbances favouring
natural succession without management intervention.
Semi-natural forests can thus retain certain characteristics
allowing natural dynamics and biodiversity levels close to
the natural ecosystems.
At the Pan-European policy level the Ministerial Con-
ference on the Protection of Forests in Europe has gradu-
ally developed new indicators of sustainable management.
Regarding the criterion biodiversity there are 9 different
indicators: tree species composition, regeneration, natu-
ralness, introduced tree species, dead wood, genetic re-
sources, landscape pattern, threatened forest species and
protected forests (Rametsteiner and Mayer 2004). The
present attempt to evaluate the behaviour of indicators at
the stand scale, and particularly dead wood and natural-
ness, are thus promising. For some other indicators, e.g.
tree species and threatened (lichen) species the generality
at the Pan-European level was limited. The geographicalrange of use of indicators should therefore be clearly de-
fined both with respect to regional ecological aspects as
well as values (Duelli and Obrist 2003). This requires
studies at multiple spatial scales of how biodiversity indi-
cators behave in gradients of land use change, how indica-
tors are related to each other (Roberge and Angelstam
2004), and how different actors perceive them (Uliczka et
al. 2004).
The indicators we used were deliberately chosen to be
simple to learn and apply. Such indicators need of course
to be evaluated in detailed studies where the score of the
indicators are compared with the presence, and ideally fit-
ness, of a suite of relevant species (Angelstam et al.
2004b). This is particularly important because forest
managers usually assess forest components such as stand
age and size rather than the presence of particular focal
species. Studies about habitat requirements of declining
or endangered lichens and mosses (Berg et al. 1994), in-
sects (Jonsell et al. 1998) and birds (Angelstam et al.
2004b) show that they are closely related to habitat struc-
tures such as standing and lying dead wood, large trees
and hole trees. However, only for a few of the stand scale
indicator species proposed and even used (Norén et al.
2002), it is actually shown what forest components they
require at different scales (e.g. Nilsson et al. 1995, Uliczka
and Angelstam 1999, 2000). Such studies are urgently
needed.
We also stress the need to include processes that main-
tain different compositional and structural elements of bi-
odiversity. This was pointed out by Norton (2003) who
noted the lack of explicit address to all the three main
groups of biodiversity elements among institutions work-
ing with the implementation of biodiversity policies. We
suggest that the least labour-intensive elements should be
chosen for practical application in monitoring, and that
specific research should aim at elucidating the detailed re-
lationships, and regional variation in the relationships
among different elements of biodiversity.
Developing measurements of elements of biodiversity
applicable in management is an act of balance. On the one
hand the implementation of knowledge and techniques
should be affordable, on the other hand it should be a
rigorous approach where indicators are used only if their
indicator value has been documented, and where indica-
tors are added or revised as new knowledge appears. We
argue that the present method approaches this balance.
Our experience from interacting with managers while de-
veloping this method over several years is that we were
able to communicate to local managers that existing forest
monitoring schemes need to be revised to address the new
paradigm of sustainable forest management. As an exam-
ple, the managers of the communal forests of Montafon
in Austria developed a new forest inventory methodology
which addresses the multifunctionality of forests (Maier
and Breuss 2002, Dönz-Breuss et al. 2004).

318 ECOLOGICAL BULLETINS 51, 2004Estimates of monitoring costs
An important factor for the applicability of a study is the
financial expenditure. Therefore, one of the goals of devel-
oping this methodology was its cost-efficiency. In the five
case studies the total amount of person days ranged from
23 (Abernethy) to 43 (Montafon) days. This difference
depended on differences in the complexity of the forests,
accessibility of sampling plots due to topography and
weather conditions. Note, however, that application of the
method in the field, and the communication of the results
to different stakeholders, also constitutes education efforts
in what biodiversity actually is.
Communication of results
Implementing policies concerning the maintenance of for-
est biodiversity in actual landscapes is a major challenge to
a wide range of actors. This requires co-operation among
actors with different interests, attitudes and competencies
on different levels of society (Lee 1993, Gundersson et al.
1995). As an example, the five case studies included differ-
ent land owner categories including state (Pskov), non-
governmental organisation (Abernethy), communal
(Montafon), non-industrial private (T rudner Horn) as
well as local, regional and national authorities in the actual
landscape as a whole. The process of policy implementa-
tion is about goal formulation, mobilisation of resources,
organisational solutions and concrete priorities concerning
land use (Clark 2002). Considering this complexity, is it at
all possible to implement forest biodiversity policies at the
scale of forest management units?
This is an urgent research issue, which needs to be ad-
dressed for moving towards a more sustainable ecological
future (Clark 2002, Angelstam and Törnblom in press).
This requires interdisciplinary, but also transdisciplinary
(Jakobsen et al. 2004) approaches, bridging the common
barrier between natural and social sciences (Berkes and
Folke 1998, Hammer and Söderquist 2001, Norton
2003). T o understand this complex reality, policy and im-
plementation research is of vital importance (Gunderson
et al. 1995, Clark 2002). It can help us in identifying, de-
scribing and interpreting troublesome obstacles as well as
vital bridges in the actual implementation processes. This
empirical knowledge is important for developing and im-
proving different kinds of recommendations (for instance
policy formulations, implementation strategies, organisa-
tional solutions and decision support systems). Policy and
implementation research is a classical field within social
sciences with a good potential for fruitful interdisciplinary
research (Sabatier 1999). When it comes to implementing
ecological sustainability it is essential with a close collabo-
ration between social and natural scientists.
After showing that a biodiversity indicator behaves in
the expected way, monitoring results could be comparedwith performance targets to assess the status of different
elements of biodiversity. A next step would then be to eval-
uate the implementation of biodiversity monitoring, as-
sessment and communication to managers. This includes
the mapping of so-called implementation structures, start-
ing from the “bottom” (i.e. from individual owners of
land) up to the landscape and its different actors and inter-
est groups (Sabatier 1999). Here, an important aspect to
understand are the priorities and strategies in the land-use
of the owners and what types of co-operative networks
they have (Clark 2002). How do the implementing actors
understand policies and directives, and the results from
monitoring? Do they have capacity and tools to act, that is
resources, competence etc.? Do the implementing actors
feel that they have the possibility to influence policies and
directives and do they want to act in line with those or are
they opposed to them?
Acknowledgements – Our approach to measure different elements Acknowledgements – Our approach to measure different elements Acknowledgements
of biodiversity has developed over several years. The starting
point was the insight that to maintain biodiversity in practice,
science must approach practise rather than the opposite. Börje
Pettersson at StoraEnso stimulated the first practical implemen-
tation efforts in 1998, which then was developed in a systematic
way during the work within different research programmes fi-
nanced by the EC, WWF and MISTRA in Sweden. The work in
the different case studies was made possible thanks to the work of
many colleagues and students. We thank Peter Ekelund, Henny
Frid, Linda Nilsson, Olga Widén, Rainer Ploner, Jean-Michel
Roberge, Mikael Stenberg, Daniel Thorell, Sergey Roshdestven-
skiy, Jevgenij Korostelev, Ylva Lenhed and Jonas Johansson for
their devoted efforts in the field. Local logistic and financial sup-
port was received from the RSPB (Ron Summers) for the case
study Abernethy, Scotland, Amt für Naturparke Südtirol, Italy
(Rainer Ploner) for the case study T rudner Horn, Stand Monta-
fon-Forstfonds in Austria (Hubert Malin, Bernhard Maier), the
staff at the Pskov Model forest (T atyana Popova, Boris Roman-
yok) and finally the Bia ł yok) and finally the Bia ł yok) and finally the Bia owieża forest in Poland (Bogumila
Jedrzejewska). We thank Rita Bütler, Jakob Heilmann-Clausen,
Emmanuel Menoni, Sven Nilsson, Börje Pettersson and Jean-
Michel Roberge for valuable comments on the manuscript.
References
Angelstam, P . 1998a. Maintaining and restoring biodiversity by
developing natural disturbance regimes in European boreal
forest. – J. Veg. Sci. 9: 593–602.
Angelstam, P . 1998b. T owards a logic for assessing biodiversity in
boreal forest. – In: Bachmann, P ., Köhl, M. and Päivinen, R.
(eds), Assessment of biodiversity for improved forest plan-
ning. Kluwer, pp. 301–313.
Angelstam, P . 2003. Reconciling the linkages of land manage-
ment with natural disturbance regimes to maintain forest bi-
odiversity in Europe. – In: Bissonette, J. A. and Storch, I.
(eds), Landscape ecology and resource management: linking
theory with practice. Island Press, pp. 193–226.
Angelstam, P . and Kuuluvainen, T . 2004. Boreal forest distur-
bance regimes, successional dynamics and landscape struc-
ture – a European perspective. – Ecol. Bull. 51: 117–136.

319 ECOLOGICAL BULLETINS 51, 2004Angelstam, P . and Törnblom, J. in press. Maintaining forest bio-
diversity in actual landscapes – European gradients in history
and governance systems as a “landscape lab”. – In: Marchetti,
M. et al. (eds), Monitoring and indicators of forest biodiver-
sity in Europe – from ideas to operationality. EFI Symp.,
Florence.
Angelstam, P ., Majewski, P . and Bondrup-Nielsen, S. 1995.
West-east co-operation in Europe for sustainable boreal for-
ests. – Water Air Soil Pollut. 82:3–11.
Angelstam, P . et al. 1997. Biodiversity and sustainable forestry in
European forests: how East and West can learn from each
other. – Wildl. Soc. Bull. 25: 38–48.
Angelstam, P . et al. 2000. Effects of moose density on timber
quality and biodiversity restoration in Sweden, Finland and
Russian Karelia. – Alces 36: 133–145.
Angelstam, P ., Breuss, M. and Mikusi ński, G. 2001. T oward the
assessment of forest biodiversity of forest management units
– a European perspective. – In: Franc, A., Laroussinie, O.
and Karjalainen, T . (eds), Criteria and indicators for sustain-
able forest management at the forest management unit level.
European Inst. Proc. 38: 59–74, Gummerus printing, Saari-
järvi, Finland.
Angelstam, P . et al. 2002. Effects of forest structure on the pres-
ence of woodpeckers with different specialisation in a land-
scape history gradient in NE Poland. – In: Chamberlain, D.
and Wilson, A. (eds), Proc. of the 2002 annual IALE (U.K.)
held at the Univ. of East Anglia, pp. 25–38.
Angelstam, P . et al. 2003. Assessing village authenticity with sat-
ellite images – a method to identify intact cultural landscapes
in Europe. – Ambio 33: 594–604.
Angelstam, P . et al. 2004a. Land management data and terrestrial
vertebrates as indicators of forest biodiversity at the land-
scape scale. – Ecol. Bull. 51: 333–349.
Angelstam, P . et al. 2004b. Habitat modelling as a tool for land-
scape-scale conservation – a review of parameters for focal
forest birds. – Ecol. Bull. 51: 427–453.
Angelstam, P ., Mikusi ński, G. and Fridman, J. 2004c. Natural
forest remants and transport infrastructure – does history
matter for biodiversity conservation planning? – Ecol. Bull.
51: 149–162.
Angelstam, P . et al. 2004d. Monitoring forest biodiversity – from
the policy level to the management unit. – Ecol. Bull. 51:
395–304.
Anon. 1992. Convention on Biological Diversity. – UNCED 5
June 1992, <www.biodiv.org>.
Anon. 2002. Decision No 1600/2002/EC of the European par-
liament and of the council of 22 July 2002 laying down the
Sixth Community Environment Action Programme. – Offi-
cial Journal of the European Communities.
Arcese, P . and Sinclair, A. R. E. 1997. The role of protected
areas as ecological baselines. – J. Wildl. Manage. 61:
587–602.
Axelsson, A.-L. and Östlund, L. 2001. Retrospective gap analysis
in a Swedish boreal forest landscape using historical data. –
For. Ecol. Manage. 147: 109–122.
Balée, W . 1998. Advances in historical ecology. – Columbia
Univ. Press.
Barraclough, G. 1993. The Times atlas of world history. – Times
Books.
Berg, A. et al. 1994. Threatened plant, animal and fungus species
in Swedish forests: distribution and habitat associations. –
Conserv. Biol. 8: 718–731.Berger, J. et al. 2001. A mammalian predator-prey imbalance:
grizzly bear and wolf extinction affect avian neotropical mi-
grants. – Ecol. Appl. 11: 947–960.
Berkes, F . and Folke, C. 1998. Linking social and ecological sys-
tems. Management practices and social mechanisms for
building resilience. – Cambridge Univ. Press.
Berkes, F ., Colding, J. and Folke, C. 2003. Navigating social-
ecological systems. – Cambridge Univ. Press.
Breitenmoser, U. 1998. Large predators in the Alps: the fall and
rise of Man’s competitors. – Biol. Conserv. 83: 279–289.
Brokaw, N. and Lent, R. 1999. Vertical structure. – In: Hunter,
M. L. (ed.), Maintaining biodiversity in forest ecosystems.
Cambridge Univ. Press, pp. 373–399.
Bunnell, F . L. and Johnson, J. F . 1998. Policy and practices for
biodiversity in managed forests: the living dance. – Univ. of
British Columbia Press.
Busch, D. E. and T rexler, J. C. 2003. The importance of moni-
toring in regional ecosystem initiatives. – In: Busch, D. E.
and T rexler, J. C. (eds), Monitoring ecosystems. Island Press,
pp. 1–23.
Clark, T . W . 2002. The policy process. A practical guide for nat-
ural resource professionals. – Yale Univ. Press.
Darby, H. C. 1956. The clearing of woodlands in Europe. – In:
Thomas, W . L. (ed.), Man’s role in changing the face of the
Earth. Univ. of Chicago Press, pp. 183–216.
Davis, L. S. et al. 2001. Forest management: to sustain ecological,
economic and social values. – McGraw Hill.
Dönz-Breuss, M., Maier, B. and Malin, H. 2004. Management
for forest biodiversity in Austria – the view of a local forest
enterprise. – Ecol. Bull. 51: 109–115.
Drakenberg, B. and Lindhe, A. 1999. Indirekt naturvärdes-
bedömning på beståndsnivå – en praktiskt tillämpbar me-
tod. – Skog & Forskning 2: 60–66, in Swedish.
Duelli, P . and Obrist, M. K. 2003. Biodiversity indicators: the
choice of values and measures. – Agricult. Ecosyst. Environ.
98: 87–98.
Duinker, P . 2001. Criteria and indicators of sustainable forest
management in Canada: progress and problems in integrat-
ing science and politics at the local level. – In: Franc, A., La-
roussinie, O. and Karjalainen, T . (eds), Criteria and indica-
tors for sustainable forest management at the forest manage-
ment unit level. European Forest Inst. Proc. 38: 7–27, Gum-
merus Printing, Saarijärvi, Finland.
Dunlop, B. M. S. 1997. The woods of Strathspey in the nine-
teenth and twentieth centuries. – In: Smout, T . C. (ed.),
Scottish woodland history. Scottish Cultural Press, pp. 176–
189.
Egan, D. and Howell, E. A. 2001. The historical ecology hand-
book. – Island Press.
Ellenberg, H. 1996. Vegetation Mitteleuropas mit den Alpen, 5
Auflage. – Eugen Ulmer, Stuttgart, in German.
Esseen, P . A. et al. 1997. Boreal forests. – Ecol. Bull. 46: 16–47.
Faliński, J. B. 1986. Vegetation dynamics in temperate lowland
primeval forests. – Geobotany 8, Dr W . Junk Publ.
Franc, A., Laroussinie, O. and Karjalainen, T . (eds) 2001. Crite-
ria and indicators for sustainable forest management at the
forest management unit level. – European Forest Inst. Proc.
38, Gummerus Printing, Saarijärvi, Finland.
Frid, H. 2001. T rends in components of biodiversity in a Scottish
land-use history gradient. – Diploma thesis at the Dept of
Conservation Biology, Swedish Univ. of Agricultural Scienc-
es.

320 ECOLOGICAL BULLETINS 51, 2004Grabherr, G. et al. 1998. Hemerobie österreichischer Waldöko-
systeme. – Universitätsverlag Wagner Innsbruck, in German.
Grant, E. 1994. Abernethy Forest. Its people and its past. – Ar-
kleton T rust, Nethybridge.
Gunderson, L. 2003. Learning to monitor or monitor to learn. –
In: Busch, D. E. and T rexler, J. C. (eds), Monitoring ecosys-
tems. Island Press, pp. xi–xv.
Gunderson, L. H., Holling, C. S. and Light, S. S. (eds) 1995.
Barriers and bridges to the renewal of ecosystems and institu-
tions. – Columbia Univ. Press.
Gunst, P . 1989. Agrarian systems of central and eastern Europe. –
In: Chirot, D. (ed.), The origins of backwardness in eastern
Europe: economics and politics from the Middle Ages until
the early twentieth century. Univ. of California Press, pp. 53–
91.
Gustafsson, L. 2002. Presence and abundance of red-listed
plant species in Swedish forests. – Conserv. Biol. 16: 377–
388.
Hambler, C. 2004. Conservation. – Cambridge Univ. Press.
Hamilton, H. 1982. Praktisk skogshandbok. – Sveriges Skogs-
vårdsförbund, Stockholm, in Swedish.
Hammer, M. and Söderqvist, T . 2001. Enhancing transdiscipli-
nary dialogue in curricula development. – Ecol. Econ. 38: 1–
5.
Hannah, L., Carr, J. L. and Lankerani, A. 1995. Human distur-
bance and natural habitat: a biome level analysis of a global
data set. – Biodiv. Conserv. 4: 128–155.
Hedenås, H. 2002. Epiphytic lichens on Populus tremula : impli-
cations for conservation. – Ph.D. thesis, Dept of Ecology and
Environmental Science, Umeå Univ., Sweden.
Heywood, V . H. 1995. Global biodiversity assessment. – Cam-
bridge Univ. Press.
Higman, S. et al. 1999. The sustainable forestry handbook. –
Earthscan Publ., London.
Jakobsen, C. H., Hels, T . and McLaughlin, W . J. 2004. Barriers
and facilitators to integration among scientists in transdisci-
plinary landscape analyses: a cross-country comparison. –
For. Policy Econ. 6: 15–31.
Jędrzejewska, B. and J ędrzejewski, W . 1998. Predation in verte-
brate communities. – Springer.
Jonsell, M., Weslien, J. and Ehnström, B. 1998. Substrate re-
quirements of red-listed saproxylic invertebrates in Sweden.
– Biodiv. Conserv. 6: 1–18.
Jonsson. B. G. and Jonsell, M. 1999. Exploring potential biodi-
versity indicators in boreal forests. – Biodiv. Conserv. 8:
1417–1433.
Kirby, K. J. and Watkins, C. 1998. The ecological history of Eu-
ropean forests. – CAB International, Wallingford, U.K.
Larsson, T .-B. et al. (eds) 2001. Biodiversity evaluation tools for
European forest. – Ecol. Bull. 50.
Lee, K. N. 1993. Compass and gyroscope. Integrating science
and politics for the environment. – Island Press.
Lindén, H. et al. 2000. Large-scale forest corridors to connect the
taiga fauna to Fennoscandia. – Wildl. Biol. 6: 179–188.
Lindenmayer, D. B. and Franklin, J. F . 2002. Conserving forest
biodiversity. A comprehensive multiscaled approach. – Is-
land Press.
Maier, B. and Breuss, M. 2002. Multifunktionale Waldinventur
am Beispiel Stand Montafon-Forstfonds. – Kleine Waldzei-
tung 3: 10–12, in German.
Mayer, H. 1984. Die Wälder Europas. – Gustav Fischer, Stutt-
gart, in German.Mayer, H. 1992. Waldbau auf soziologisch-ökologischer Grund-
lage. – Gustav Fischer, Stuttgart, in German.
Meffe, G. K. et al. 2002. Ecosystem management. Adaptive,
community-based conservation. – Island Press.
Mikusiński, G. and Angelstam, P . 1998. Economic geography,
forest distribution, and woodpecker diversity in central Eu-
rope. – Conserv. Biol. 12: 200–208.
Mikusiński, G. and Angelstam, P . 2004. Occurrence of mam-
mals and birds with different ecological characteristics in re-
lation to forest cover in Europe – do macroecological data
make sense? – Ecol. Bull. 51: 265–275.
Mikusiński, G., Angelstam, P . and Sporrong, U. 2003. Distribu-
tion of deciduous stands in villages located in coniferous for-
est landscapes in Sweden. – Ambio 33: 520–526.
Nilsson, S. G. et al. 1995. T ree-dependent lichens and beetles as
indicators in conservation forests. – Conserv. Biol. 9: 1208–
1215.
Nilsson, S. G., Hedin, J. and Niklasson, M. 2001. Biodiversity
and its assessment in boreal and nemoral forest. – Scand. J.
For. Res. Suppl. 3: 10–26.
Nilsson, S. G. et al. 2002. Densities of large living and dead trees
in old-growth temperate and boreal forests. – For. Ecol.
Manage. 161: 189–204.
Norén, M. et al. 2002. Handbok för inventering av nyckelbio-
toper. – Skogsstyrelsen, Jönköping, in Swedish.
Norton, B. G. 2003. Searching for sustainability. Interdiscipli-
nary essays in the philosophy of conservation biology. –
Cambridge Univ. Press.
Noss, R. F . 1990. Indicators for monitoring biodiversity: a hierar-
chical approach. – Conserv. Biol. 4: 355–364.
Östlund, L., Zackrisson, O. and Axelsson, A.-L. 1997. The histo-
ry and transformation of a Scandinavian boreal forest land-
scape since the 19th century. – Can. J. For. Res. 27: 1198–
1206.
O’Sullivan, P . E. 1973. Land use changes in the forest of Aber-
nethy, Inverness-shire (1750–1900 A.D.). – Scottish Geogr.
Mag. 89: 95–106.
Peer, T . 1995. Die natürliche Pflanzendecke Südtirols. – Amt für
Landschaftsplanung, Bozen, in German.
Peterken, G. 1996. Natural woodland: ecology and conserva-
tion in northern temperate regions. – Cambridge Univ.
Press.
Pimentel, D., Westra, L. and Noss, R. F . 2000. Ecological integ-
rity. Integrating environment, conservation and health. – Is-
land Press.
Puumalainen, J. 2001. Structural, compositional and functional
aspects of forest biodiversity in Europe. – Geneva timber and
forest discussion papers. United Nations, New York.
Puumalainen, J. et al. 2002. Forest biodiversity assessment ap-
proaches for Europe. – EUR Rep. 20423, Joint Resarch Cen-
tre, Ispra, European Commission.
Rackham, O. 2003. Ancient woodland. – Castlepoint Press, Col-
vend, Dalbeattie.
Rametsteiner, E. and Mayer, P . 2004. Sustainable forest manage-
ment and Pan-European forest policy. – Ecol. Bull. 51: 51–57.
Read, H. 2000. Veteran trees: a guide to good management. –
English Nature, Northminster House, Peterborough, U.K.
Reed, D. D. and Mroz, G. D. 1997. Resource assessment in for-
ested landscapes. – Wiley.
Ripple, W . J. and Beschta, R. L. 2003. Wolf reintroduction, pre-
dation risk, and cottonwood recovery in Yellowstone Nation-
al Park. – For. Ecol. Manage. 184: 299–313.

321 ECOLOGICAL BULLETINS 51, 2004Roberge, J.-M. and Angelstam, P . 2004. Usefulness of the um-
brella species concept as a conservation tool. – Conserv. Biol.
18: 76–85.
Rolstad, J. et al. 2001. Epiphytic lichens in Norwegian coastal
spruce forest: historic logging and present forest structure. –
Ecol. Appl. 11: 421–436.
Sabatier, P . A. (ed.) 1999. Theories of the policy process. – West-
view, Boulder, CO.
Schlaepfer, R. and Elliot, C. 2000. Ecological and landscape con-
siderations in forest management: the end of forestry? – In:
von Gadow, K., Pukkala, T . and T omé, M. (eds), Sustainable
forest management. Kluwer, pp. 1–67.
Shorohova, E. and T etioukhin, S. 2004. Natural disturbances
and the amount of large trees, deciduous trees and coarse
woody debris in the forests of Novgorod Region, Russia. –
Ecol. Bull. 51: 137–147.
Siitonen, J. 2001. Forest management, coarse woody debris and
saproxylic organisms: Fennoscandian boreal forests as an ex-
ample. – Ecol. Bull. 49: 11–41.
Smith, G. and Gillett, H. (eds) 2000. European forests and pro-
tected areas: gap analysis. T echnical report. – UNEP World
Conservation Monitoring Centre, Cambridge, U.K.
Smith, J. S. 1993. Changing deer numbers in the Scottish High-
lands since 1780. – In: Smout, T . C. (ed.), Scotland since
prehistory, natural and human impact. Scottish Cultural
Press, Edinburgh, pp. 89–97.
Smout, T . C. 1997. Highland land use before 1800: misconcep-
tions, evidence and realities. – In: Smout, T . C. (ed.), Scottish
woodland history. Scottish Cultural Press, pp. 5–23.
Sokolska, J. and Leniec, H. 1996. Puszcza Knyszynska. – Suprasl,
Poland, in Polish.
Stanners, D. and Bourdeau, P . 1995. Europe’s environment: the
Dobris assessment. – European Environment Agency, Co-
penhagen.
Steven, H. M. and Carlisle, A. 1959. The native pinewoods of
Scotland. – Hartnolls, Bodmin, Cornwall.
Stokland, J. et al. 2003. Forest biodiversity indicators in the Nor-
dic countries. Status based on national forest inventories. –
T ema Nord 514, Agriculture and forestry, Nordic Council of
Ministers, Copenhagen.Summers, R. 1998. The history and ecology of Abernethy forest,
Strathspey. – RSPB Rep., RSPB Inverness, Scotland.
Summers, R. W. et al. 1999. The structure of ancient native
pinewoods and other woodlands in the Highlands of Scot-
land. – For. Ecol. Manage. 119: 231–245.
Tappeiner, U. et al. (eds) 2003. The EU agricultural policy and
the environment. – Europäische Akademie, Bozen and
Blackwell.
Thompson, I. D. and Angelstam, P . 1998. Special species. – In:
Hunter, M. L. (ed.), Maintaining biodiversity in forest eco-
systems. Cambridge Univ. Press, pp. 434–459.
T rauger, D. L. et al. 2003. The relationship of economic growth
to wildlife conservation. – Wildl. Soc. T ech. Rev. 03-1, The
Wildlife Society, Bethesda, MD, USA.
Uliczka, H. and Angelstam, P . 1999. Occurrence of epiphytic
lichens in relation to tree species and age in managed boreal
forest. – Ecography 22: 396–405.
Uliczka, H. and Angelstam, P . 2000. Assessing conservation val-
ues of forest stands based on specialised lichens and birds. –
Biol. Conserv. 95: 343–351.
Uliczka, H., Angelstam, P . and Roberge, J.-M. 2004. Indicator
species and biodiversity monitoring systems for non-indus-
trial private forest owners – is there a communication prob-
lem? – Ecol. Bull. 51: 379–384.
Ullsten, O. et al. 2004. T owards the assessment of environmen-
tal sustainability in forest ecosystems: measuring the natural
capital. – Ecol. Bull. 51: 471–485.
Vera, F . W . M. 2000. Grazing ecology and forest history. –
CABI Publ., Wallingford, U.K.
Watson, A. 1983. Eighteenth century deer numbers and pine
regeneration near Braemar, Scotland. – Biol. Conserv. 25:
289–305.
Watson, R. T . et al. 2000. Land use, land-use change and forest-
ry. Intergovernmental Panel of Climate Change. – Cam-
bridge Univ. Press.
Williams, M. 2003. Deforesting the earth. – Chicago Univ.
Press.
Yaroshenko, A. Yu. et al. 2001. The intact forest landscapes of
northern European Russia. – Greenpeace Russia and the
Global Forest Watch, Moscow.

322 ECOLOGICAL BULLETINS 51, 2004Appendix
Indicators of forest biodiversity at the
stand scale – a field guide
The actual landscape managed by a forest company, the
private landowners in a village, common land, or combi-
nations thereof, is where policies about maintaining biodi-
versity and actual management meet. T o manage for forest
biodiversity one needs to measure its main elements, and if
needed, modify management so that the elements develop
in the desired direction. A first step in the monitoring
process is to stratify forests into different forest vegetation
types using the different natural and anthropogenic distur-
bance regimes to which species are adapted (e.g. Rackham
2003, Angelstam and Kuuluvainen 2004). Then, in order
to be cost-effective, one need to identify and select re-
sponse variables (i.e. indicators) representing different ele-
ments of biodiversity that are affected by management di-
rectly or indirectly.
Based on a review of literature and experiences from
practical application in a number of case studies (Höjer
1999, Drakenberg and Lindhe 1999, Frid 2001, Angel-
stam et al. 2002a, b) we outline a system for measuring
major elements of biodiversity at the scale of stands for
European forests and woodland. The elements, or indica-
tors, are chosen based on the idea that the development of
biodiversity measurements must proceed from existing
management data (e.g. tree species composition, wood
volume and site type; e.g. Jonsson et al. 1993), by gradual-
ly adding new variables representing elements of biodiver-
sity. As suggested by Larsson et al. (2001) such variables
should be derived from the composition, structure and
function of reference areas representing both the biodiver-
sity conservation visions of naturally dynamic forest (May-
er 1984, Fali ński 1986, Peterken 1996) and pre-industrial
cultural landscapes (e.g. Rackham 2003). Such variables
should thus include all living and dead tree species, indica-
tor species, the structure of stands and finally the impor-
tant processes that maintain biodiversity (Noss 1990, Nor-
ton 2003). Apart from the indicators’ relevance from an
ecological perspective, we also stress the need to ensure that
the selected indicators communicate well to the stakehold-
ers involved (e.g. Uliczka et al. 2004).
Hierarchical sampling design
When a forest management unit (FMU) has been selected
for monitoring, the first step is to make a detailed plan for
the work. Before starting with the fieldwork the landscape
should be stratified to cover all representative strata or
landscape types. In the following section, we describe how
field data plots can be clustered to allow multi-scale analy-
ses, and how the collection of data in the plots is made indetail for different elements of biodiversity. This allows sta-
tistical analyses based on a stratified sampling procedure
(e.g. Quinn and Keough 2002).
Information about the landscape
A landscape as well as a FMU can be defined in many ways
(Muir 1999, Davis et al. 2001). However, because current
knowledge indicates that very small areas can not maintain
local populations of most specialised and area-demanding
species (e.g. Angelstam et al. 2004a), we suggest a size of ca
100 km2. The first step is to collect information about the
area to be monitored. If not made before, this can be done
by means of a questionnaire consisting of two parts (Angel-
stam et al. 2004b). The first part concerns the history of
presence and breeding status of a selected set of indicator
species that represent the actual biodiversity conservation
vision for which there is good information. Often that re-
stricts the analyses to vertebrates. The second part refers to
the age class and tree species composition of the present
landscape, as well as to its history of land use development.
In addition to information about the local landscape, the
history of land use development at the regional level is
compiled by reviewing relevant literature.
Stratification
T o stratify a landscape or FMU, both natural abiotic and
biotic, as well as anthropogenic factors influencing the pat-
tern of land cover types need to be understood. Therefore,
stratification of the study area should be based on knowl-
edge about both the land use history and the present habi-
tat composition. Sampling plots should then be replicated
in each stratum. Forest stand maps and land use zones (e.g.
different silvicultural and other management regimes both
outside and within conservation areas) can also be used to
stratify the landscape to be sampled. As an example we de-
scribe the coarse landscape types (i.e., strata) identified for
the Naturpark T rudner Horn study area (Südtirol/Alto
Adige, N Italy), where this methodology was applied (An-
gelstam et al. 2002a; Table 1, Fig. 1). In this example the
stratification was made according to topography, land-
scape history, and forest management systems.
Clusters of plots
The size of a forest stand can range from <1 ha to hundreds
of hectares, depending on the tradition of forest manage-
ment, and on the other processes affecting landscape pat-
terns. Sampling at the stand scale must therefore be adapt-
ed to the structure of the landscape. T o cope with this is-
sue, a number of clusters of data collection plots should be
placed within each stratum. T o allow analyses of the data at

323 ECOLOGICAL BULLETINS 51, 2004multiple spatial scales 1 × 1 km squares with 16 systemati-× 1 km squares with 16 systemati- ×
cally distributed survey plots in each are used (Fig. 2).
These 1-km2 squares are placed in a stratified random
manner in the study area, and are evenly distributed
among the coarse landscape types. The individual plot can
thus be viewed as a local patch, and the aggregation of
plots can be seen as clusters that represent stands of differ-
ent sizes in the landscape. T o be able to repeat the survey in
the future, with additional types of data if needed, each 1-
km2 square and plot should be marked on a detailed forestmap (e.g. 1:10000) with the plot ID number of the field
protocol.
Survey plots
Information about the occurrence of specific species (com-
position), habitat pattern (structure), as well as natural and
anthropogenic processes (function) is collected in each
plot. So far data on a total of 21 basic groups of variablesTable 1. Description of the coarse landscape types (strata) of the study area Naturpark Trudner Horn, N Italy as an
example for the application of the methodology for measuring elements of biodiversity (see Angelstam et al. 2002a and
Fig. 1).
Coarse landscape type Description of the coarse landscape type
Apple orchards and vineyards The bottom of the Etsch river valley is completely flat and the main land cover type
is apple orchard. The density of transport infrastructure is high with the major
highway from northern Italy via Brenner to Austria as well as a north-south
connection for railway. On the lower slopes above the Etsch river valley vineyards
are found from just over 200 to ca 600 m a.s.l. (Peer 1995). Management is
intensive. Several smaller villages with lush gardens are located in this stratum.
Coppice forest This habitat is the typical historical landscape type of Northern Italy’s broad-leafed
forest (Puumalainen 2001). The coppice forest is mostly privately owned and today
the forest is not managed any longer. Therefore, the forest is gradually getting older,
and today most of the forest is ca 50 yr old. The most common tree species
are oak ( Quercus spp.), ash ( Fraxinus spp.) and hornbeam Carpinus betulus .
Agricultural land This landscape type is located near the villages on intermediate altitude (1000–1800
m) and includes old cultural landscape remnants, farmland that is intensively
managed today and litter forest, i.e. forest that has been used for the collection of
litter. In some places, the forest is encroaching on the former agricultural landscape.
Inside the nature park Trudner Horn, the farmland is managed to maintain the old
cultural landscape.
Private forest The private forest is subdivided into small units of ownership and the management is
not very intensive. The forest is dominated by coniferous tree species interspersed
with some deciduous components (beech Fagus sylvatica, ash and birch Betula spp.)
and is located at altitudes from 1000 to 1800 m a.s.l.
Public forest This forest category is owned by the villages in the area and is more intensively
managed with distinct age-classes, small clear-cuts and plantation after final
harvesting. Norway spruce, fir and Scots pine dominate the forest. In the lower parts,
spruce and fir are dominating, whereas in the higher parts, pine is dominating on
dryer soils. The altitudinal range is between 1000 and 1800 m.
Larch meadows The character of open woodland is maintained by grazing and the presence of large
trees (e.g. larch). Today, in some areas hazel Corylus avellana and spruce are
invading the meadows, at some other places the meadows have been afforested. The
altitudinal range of meadows is between 1700 and 2100 m.
Tree line Today the tree line starts at 2100 m. Due to historical land-use (mostly grazing) the
tree line has been lowered by 200–300 m compared to its natural level. Above the
tree line, dwarf pine Pinus mugo can be found. Most of this area is publicly owned
but there is also some private land. Some of the sampling plots are close to a ski
resort, some near grazing land where people in the villages cut the forest historically
to create grassland for their cattle during summer.

324 ECOLOGICAL BULLETINS 51, 2004(T able 2) has been collected using one field protocol sheet
for each plot. The field data is collected in four steps: 1)
general information is recorded for the sampling plot; 2)
basal areas are measured for the living and dead trees of
different kinds; 3) small trees and shrubs are surveyed
within a circle with a radius of 4 m; and 4) all other meas-
urements are taken within a circle with a radius of 30 m.
Variables collected
First of all, general information (date, time, observer, loca-
tion etc.) is noted on the protocol sheet. Since elevation is a
potentially important factor for the local distribution of
species, the altitude of the sample plots is recorded using
an altimeter or information from the topographic map. In
addition, slope (in 5° steps), aspect (8 compass directions),
and position in the slope (valley, slope, ridge, or hilltop) are
recorded.Composition
For a whole system of forest vegetation types with different
disturbance regimes, such as boreal, temperate and moun-
tain forests, it is necessary to build a system that includes
several indicator species with different ecologies that repre-
sent the different vegetation types (Angelstam 1998a, b,
Thompson and Angelstam 1999, Nilsson et al. 2001, Ber-
glind 2004). T rees can be viewed as dominant species in
forests and should therefore form the starting point for de-
scribing the composition of a forest (Nikolov and
Helmisaari 1992). For the boreal forest, analyses of the
habitat characteristics of endangered forest species show
that most of them require old trees, deciduous trees, or
dead wood (Berg et al. 1994). As suggested by Anon.
(2002) and Nilsson et al. (2002) the size of trees is an addi-
tional attribute of interest.
The key role of trees is not limited to the period when
they are living. The legacies that trees leave after they have
Fig. 2. Principle for the distribution of sample plots in the study area. Every 1-km2 square includes 16 systematically distributed
sampling plots. A number of 1-km2 plots are distributed within each stratum in the landscape.Fig. 1. Schematic illustra-
tion of the general sequence
of landscape strata (i.e.,
coarse landscape types) in a
case study. This example is
for Naturpark T rudner
Horn, N Italy (Angelstam
et al. 2002a).

325 ECOLOGICAL BULLETINS 51, 2004died are of great importance for the maintenance of forest
biodiversity in both forest and woodland. A large number
of studies show that dead wood of different kinds (stand-
ing, lying, various diameter classes, decay stages, ground
contact, light exposure etc.) is a vital resource for many
endangered species and for important ecosystem functions
(e.g. Harmon et al. 1986, Siitonen 2001, Stokland 2001,
Jonsson and Kruys 2001). In addition, other forest ele-
ments at the scale of trees, which are of importance for
forest biodiversity (e.g. large old trees, hole-bearing trees,
hollow trees) need to be assessed as a complement to tradi-
tional forest management planning data (Nilsson et al.
2002). Such elements are characteristic under a range of
natural disturbance regimes in a wide variety of forest eco-
systems (Peterken 1996), but are uncommon in even-aged,
single-species stands created by intensive management (Es-
seen et al. 1997, Siitonen 2001). To locate remnant stands
of the natural boreal forest for conservation, species spe-
cialising on these natural forest elements can be used
(Norén et al. 2002).
Species with different life-history traits have different
levels of specialisation and spatial demands (Angelstam
1996, Mykrä et al. 2000). Restricting management con-
siderations only to species that do not have landscape-scale
requirements, such as those used to identify woodland keyhabitats (Hansson 2001, Norén et al. 2002), is insuffi-
cient. Incorporating species with landscape-scale require-
ments such as grouse (Swenson and Angelstam 1993),
woodpeckers (Mikusi ński and Angelstam 1998,
Mikusiński et al. 2001), overwintering passerine birds
(Jansson and Andrén 2003), herptiles (Berglind 2004) as
well as different groups of insects (Jonsell et al. 1998, Ehn-
ström and Axelsson 2002, Berglind 2004, Wikars 2004)
will increase our understanding of the status and trends of
compositional elements of biodiversity in a region. This
need for suites of indicator species representing the diversi-
ty of forest environments (Angelstam and Kuuluvainen
2004) is generally acknowledged (Angelstam 1998a, b,
Thompson and Angelstam 1999, Gustafsson et al. 1999,
Nilsson et al. 2001, Berglind 2004). Additionally, genes
(provenance) are important in the context of re-introduc-
tions of extirpated species and in afforestation.
While the use of vascular plants to indicate forest site
type and potential rates of tree growth has a long history
(Arnborg 1990), the suggested practical use of vascular
plant species to indicate conservation value (e.g. Karström
1992, Norén et al. 2002) has proven impractical in boreal
forest (Gustafsson 2001). However, vascular plants have
been shown to be useful indicators in other ecoregions
(Dumortier et al. 2002). By contrast, at the scale of patchesTable 2. List of the basic variables collected in the field for the present methodology to measure compositional, structural
and functional elements of biodiversity.
Description of basic variables
Composition basal area of all living tree species of different diameter classes
basal area of standing and lying dead wood of different decay stages and diameter classes
specialised pendant lichens (>20 cm) and conspicuous lichen species (e.g. Lobaria spp.; Evernia
divaricata ,Bryoria fremontii , Bryoria fremontii , Bryoria fremontii Alectoria sarmentosa )
insect specialist signs (exit holes) in standing/lying dead wood without bark
direct and indirect signs of specialised vertebrates (e.g. grouse, woodpeckers)
Structure canopy height
site type as determined by ground vegetation and its cover
stand structure
vertical layering
tree age structure
tree regeneration
shrub species
special trees with important microhabitats
trees with cavities
Function land management (fruit trees, pollarded trees)
land abandonment (signs of past land use such as harvesting stumps, land abandonment as
indicated by dead junipers, ruins, stone walls)
abiotic processes (fire, flooding, wind, snow break)
biotic processes (wood-living bracket fungi, bark beetle outbreaks, high stumps killed by wood
decaying fungi, snow or wind, uprooted trees)
damage by large mammal herbivores (browsing, bark peeling and velvet rubbing by ungulates)
predation (carnivore scats and corvid observations)
human disturbances

326 ECOLOGICAL BULLETINS 51, 2004of trees, the occurrence of wood-living bracket fungi,
mosses and epiphytic lichens are important indicators of
biodiversity (e.g. Bader et al. 1995, Esseen et al. 1996,
Gustafsson 2002). Monitoring of such indicator species
has thus been proposed for use both as tools for early detec-
tion of environmental changes and in follow-up assess-
ments of management activities (Angelstam 1998a, b,
Gustafsson 2002).
So far the lists of indicator species used for assessing the
conservation value of forest stands such as those listed by
Norén et al. (2002) have generally been compiled by spe-
cies’ experts based on their field experience. However, as a
rule the scientific validation of their indicator value (Gus-
tafsson 1999), nor over what geographical area a given
species is a relevant indicator, has been done. Nilsson’s et
al. (1995) study on the of lichen lungwort Lobaria pulmo-
naria and Berglind’s (2004) work on the sand lizard naria and Berglind’s (2004) work on the sand lizard naria Lacer-
ta agilis are important exceptions. Both studies indicate ta agilis are important exceptions. Both studies indicate ta agilis
that the presence of the focal species was associated with
that of several other taxa. Similarly, Mikusi ński et al.
(2001) showed that with the presence of three-toed wood-
pecker Picoides tridactylus and white-backed woodpecker Picoides tridactylus and white-backed woodpecker Picoides tridactylus
Dendrocopus leucotos , a wide range of other forest birds
were also found. Finally, Martikainen et al. (1998)
showed that the presence of the latter species was associat-
ed with a rich fauna of saproxylic insects. This supports
the umbrella species idea, that the presence of area de-
manding specialist species ensures the presence of other
species dependent on a particular habitat (Roberge and
Angelstam 2004).
The following requirements should be used for the se-
lection of indicator species: 1) the species should be a well
documented specialist of specified stand or landscape
properties, 2) the species should be easy to detect and iden-
tify in the field. T o allow species inventories throughout
the snow-free season, the specialised species selected in this
study were vascular plants (used only to determine the for-
est site type and indirectly the potential tree species com-
position), epiphytic lichens, bracket fungi, insects and res-
ident birds.
Basal area of living trees
T o assess the amount of different tree species in the sam-
pling plot, a relascope is used to measure the basal area of
all standing living trees with a diameter at breast height
(DBH) 10 cm. B ecause large trees have special impor-
tance for the fauna and the flora, living trees should be
further subdivided into two DBH classes: <40 cm and
40 cm. The occurrence of very large trees (DBH >80
cm) is also noted (cf. Nilsson et al. 2002). Some woody
species usually considered as shrubs, such as juniper Juni-
perus communis and hazel perus communis and hazel perus communis Corylus avellana , can even grow
to the height and DBH of regular trees. In this case the
basal area of those species is also noted on the field proto-
col.Basal area of dead wood
Coarse woody debris (CWD) is made up of snags and ly-
ing trees, and is one of the most important components of
the forest ecosystem enhancing biodiversity. Especially
CWD of higher DBH classes provides habitat for several
animal, plant and fungal species. In addition, large snags
are a key component for cavity nesting species. Standing
dead wood is measured with the relascope and divided into
the same DBH classes as living trees and into coniferous
and deciduous trees. To be considered for basal area meas-
urements, a snag must have a height of >1.3 m and a DBH
10 cm.
The basal area of lying dead wood is measured with the
relascope at breast-height (i.e., horizontal position corre-
sponding to breast height), measured from the thickest
end of the log. Lying dead wood is divided into the same
DBH classes as living trees. When possible, lying CWD
originating from coniferous and deciduous trees should be
noted. For all lying trees different decay stages should be
identified (Harmon et al. 1986). We used three stages
(hard, soft, and partly decayed logs where the log is no
longer straight but follows the local microtopography).
According to Stokland (2001) the latter corresponds to a
decay level of 50%.
Lichens
The use of epiphytic lichen species as bioindicators has a
long tradition. Some species’ sensitivity for air pollution
(Bates and Farmer 1992) and lack of continuity in micro-
site conditions of a forest (Tibell 1992) are good examples.
Large epiphytic lichen such as lungwort and witch’s hair
Alectoria sarmentosa have been shown to be good indicators Alectoria sarmentosa have been shown to be good indicators Alectoria sarmentosa
of old-growth forest at the stand scale (Esseen et al. 1996).
Lichens are used by bird species as food storage and forag-
ing substrate (Pettersson et al. 1995, Esseen et al. 1996).
On the protocol sheet, the occurrence of conspicuous spe-
cies such those mentioned above is noted. The occurrence
of hanging lichens (e.g. Usnea spp., Usnea spp., Usnea Alectoria spp., Alectoria spp., Alectoria Bryoria
spp.) longer than 20 cm and their frequency of occurrence
(single (1–2) trees, several (3–5) trees or abundant (>5)) on
trees are also noted.
Insects in dead wood
Wood-living insects are an important component of the
forest ecosystem as they are not only parts of the food-web
(important food source for e.g. woodpeckers) but also in-
dicators for natural disturbance processes. Dead wood
hosts a wide range of insect species. The occurrence of larg-
er insect species depends on dead wood of larger DBH-
classes (Zabranski pers. comm.). As insects themselves are
not easy to detect in the field, one can look for indirect
signs of insects in dead wood such as exit holes (Ehnström
and Axelsson 2002). The presence of exit holes of two size
categories (>5 or >10 mm) and three general shapes

327 ECOLOGICAL BULLETINS 51, 2004(round, oval, or half-moon) on barkless snags, logs, and
stumps were noted in each survey plot. Although very
coarse, this measurement captures some of the diversity in
wood-boring species. That such a short-cut can be useful
was confirmed by Oliver and Beattie (1993) who estimat-
ed the species richness of spiders, ants, polychaetes, and
mosses by dividing individuals into recognisable taxonom-
ic units and were able to show that there was little differ-
ence between classifications made by a non-expert and tax-
onomic specialist.
Vertebrate indicators
Forest dwelling grouse (Aves, Tetraonidae) and woodpeck-
ers (Aves, Picidae) are examples of species that are special-
ised on certain forest types (Swenson and Angelstam 1993,
Angelstam and Mikusi ński 1994, Storch 1999, Derleth et
al. 2000). Woodpeckers for example can be used as indica-
tor species for certain natural forest types (Mikusi ński and
Angelstam 1998) but also as indicators for the diversity of
forest bird species (Mikusi ński et al. 2001). Both direct
observations (sightings and calls) as well as indirect signs
(e.g. feeding signs of woodpeckers; droppings; tracks;
cones with feeding signs of red squirrel Sciurus vulgaris ) are Sciurus vulgaris ) are Sciurus vulgaris
noted during the period of data collection inside the plots.
To increase the vertebrate sample size, observations made
en route to the next sample plot are noted on the same
sheet in a specific field. The number of observations (calls
and sightings) of corvid birds (such as raven Corvus corax ,
crow Corone cornix , jay Garrulus glandarius , and magpie
Pica pica ) are noted. The reason is that a disruption of Pica pica ) are noted. The reason is that a disruption of Pica pica
predator-prey relationships such as increased abundance of
generalist predators may affect the breeding success of for-
est species such as capercaillie T etrao urogallus (Kurki et al. T etrao urogallus (Kurki et al. T etrao urogallus
2000).
Because of their important function as primary cavity
nesters, woodpeckers are considered as keystone species.
They build cavities used by secondary and weak-primary
cavity nesters (Martin and Eadie 1999). Indirect signs of
black woodpeckers Dryocopus martius noted on the proto- Dryocopus martius noted on the proto- Dryocopus martius
col sheet are holes in anthills and deep feeding excavations
at the base of spruce trees, in logs, or in stumps. The oc-
currence of the great spotted woodpecker Dendrocopos
major is identified by indirect signs such as piles of cones major is identified by indirect signs such as piles of cones major
emptied of seeds. Finally, occurrence of the three-toed
woodpecker is denoted by horizontal lines of small holes
in the bark of trees (Bütler et al. 2004). Fresh signs on
spruce show transparent, sticky, and flowing resin, or
transparent drops are present in the holes. Fairly fresh
signs show a whitish or yellowish resin that does not flow
anymore. Old signs show no resin, but only small holes. A
sign of long use is when the tree has formed bulges. Final-
ly, indirect signs of woodpeckers that cannot be attributed
surely to a specific species are noted on the form as “wood-
pecker”.Structure
The main components of a forest are the trees. Because
species have adapted to different developmental stages of
trees from seeds to the decayed wood and to the various
tree species, it is essential that the structure of trees at mul-
tiple spatial scales is monitored. Compared with dynamic
natural forests, managed forests have altered structural var-
iability at all spatial scales. Typical changes are truncated
age class distribution (Angelstam et al. 2004b), declines in
the amount of large forest areas (Mykrä et al. 2000, Aksen-
ov et al. 2002), altered tree species composition, increased
homogeneity in tree spacing, truncated diameter distribu-
tions favouring smaller trees (Nilsson et al. 2002), and re-
duced frequency of occurrence of standing live damaged
trees, standing dead trees, and fallen trees in different decay
stages (Siitonen 2001). Similarly, but at the landscape
scale, there has been a disproportionate loss of forests on rich
sites to agriculture and other uses (Angelstam et al. 2003a).
At the stand scale the distribution of different tree spe-
cies, decay stages of dead wood, basal area, height, as well
as horizontal and vertical layering constitute basic structur-
al attributes. Additionally, the variance in tree sizes needs
to be included as well as some measurement of horizontal
layering. At the landscape scale, the range of age classes
including biologically old stands should be estimated.
Such structural aspects can be illustrated by combining
different compositional elements as described above.
Canopy height
The average canopy height of the trees in the sampling plot
is measured with a height-meter to the nearest metre. In
the analyses the data is presented in 5-m classes.
Cover of ground vegetation and site type
Vascular plant species indicating abiotic qualities have
been used for a long time in agriculture and forestry in
many countries (Ellenberg 1996). In most cases the sys-
tems have been used to specify soil type or soil productivi-
ty, but also for habitat classification. Using the stand classi-
fication system of Hägglund and Lundmark (1987), each
sample plot is attributed to one of the following site types:
tall herb (TH), low herb (LH), ground without field layer
(GW), broad-leaved grass (BG), narrow-leaved grass
(NG), sedge-horsetail ( Carex spp./ Carex spp./ Carex Equisetum spp.) (SH),
bilberry Vaccinium myrtillus (BL), lingonberry Vaccinium myrtillus (BL), lingonberry Vaccinium myrtillus Vaccinium
vitis-ideae (LB), crowberry/heather ( Empetrum nigrum /
Calluna vulgaris ) (CR), bog (LE), lichen cover 25–50% Calluna vulgaris ) (CR), bog (LE), lichen cover 25–50% Calluna vulgaris
(WL), lichen cover >50% (LI). In those areas where the
classification system of Hägglund and Lundmark (1987)
can not be applied, characteristic site indicator plant are
noted for the specific stand. For the ground vegetation,
cover is recorded in 10% units. For bilberry (BL) and crow-
berry/heather (CR) the height is noted to the nearest dm.

328 ECOLOGICAL BULLETINS 51, 2004In various studies, bilberry has been shown to be a key
species affecting the suitability of habitat for a variety of
species (e.g. Suchant and Braunisch 2004). The flowering
part and leaves are used by Lepidoptera larvae (Baines et al.
1993), which in turn are foraged by for example birds
(Atlegrim and Sjöberg 1995, Storch 1999). Several herbiv-
ores also feed on the berries and stems of bilberry (Breuss
1999, Nordengren et al. 2003). Beside its function as for-
aging substrate, bilberry (as well as other ericaceous
shrubs) also provide cover to forest dwelling species (Storch
1999). Hence, the total cover of the green ground vegetation
in general and the cover of bilberry specifically are estimated.
Stand structure
Based on basal area data there are several ways of describing
the structure of forest stands. The identity and relative
abundance of tree species on different site types and the
relative amount of different types of dead wood are two
examples (Stokland 2001).
Vertical layering
Vertical structure is the bottom-to-top configuration of
above-ground vegetation within a forest stand (Brokaw
and Lent 1994). Variation in vertical structure depends on
soil, particular climate, tree species and other plant species,
but also on temporal aspects (succession). The vertical or-
ganisation of forest vegetation has various direct (food,
nesting, resting, perching, etc.) and indirect effects (e.g.
internal stand microclimate and distribution of animal
prey) on animals and plants. Species richness usually in-
creases with succession because species richness increases
with vertical complexity, which itself increases with stand
age (Brokaw and Lent 1994). The vertical cover of forest
vegetation in each plot is estimated in 10% units for three
height classes: 1.3–4 m, > 4–10 m, and >10 m. In addi-
tion, the relative proportion of deciduous and coniferous
foliage is noted for each layer.
T ree age structure
During the succession after natural or anthropogenic dis-
turbance a forest stand goes through a variety of stages that
affect the structural diversity (Oliver and Larsen 1996). In
a managed forest these age classes can be distinguished into
young forest, thicket, middle-aged and final harvest forest.
Age classes that are beyond this classification are usually
clumped into one “overmature” age class without further
distinctions. Yet, for a variety of animal and plant species
the older age classes (>>120 yr) constitute suitable habitat.
Therefore, a classification of the forest into more age class-
es (especially including the ageing as well as biological old
forest types) is of importance.
Using forest stand databases or field observations all
sample plots are classified into six stand age classes with thefollowing age intervals: old-growth (>150 yr), ageing forest
(110–150 yr), final felling (70–110 yr), middle-aged forest
(30–70 yr), young forest (5–30 yr), and clear-cut (<5 yr)
(cf. Angelstam and Kuuluvainen 2004). If the forest is
even-aged, the specific age class is noted. If the forest is
mixed-aged, all occurring age classes are noted. The specif-
ic age-class standards should of course be adapted to local
conditions. Indeed, the duration of a natural cycle in a nat-
ural forest is ca 350 yr, compared with 60–100 yr in man-
aged forest (Pennanen 2002).
Tree regeneration
The number of small trees with a height 1.3 m and a
DBH <10 cm is counted for each species within a 4-m
radius. The presence or absence of nurse logs and nurse
stumps, defined by the presence of tree seedlings (those
species that have been measured for the basal area) on fall-
en trees or stumps is recorded as a measurement of stand
age and occurrence of decay stages in a later developmental
phase. Additionally, the type of regeneration (natural or
artificial, i.e. planted) is specified.
Shrub species
Shrubs are an important component of forests as they con-
tribute to foliage height diversity, and are used as forage
(leaves, stems, fruits) and cover for a variety of animal spe-
cies. The occurrence of different shrub species is docu-
mented. If certain species have been severely browsed by
deer or livestock, this is noted in a separate data-field.
Special trees with important microhabitats
The presence/absence of living trees (DBH 10 cm) of
special types with certain structures such as different mi-
crohabitats typical for natural forests is noted. The types of
special trees recorded are: bent tree, naturally damaged
trees (broken top), forked trees (only large trees of com-
mercial species), trees with retarded growth (especially at
the tree line), very large trees (DBH >80 cm), hollow trees,
moss-covered trees, and lichen-covered trees. The presence
of such special trees indicates a low intensity in forest use.
Indeed, in commercially used forests these trees are gener-
ally cut already in the first thinning period due to their low
economic value.
Trees with cavities
The presence of different types of trees with woodpecker
(or other cavity nesting birds) holes of different sizes are
noted: 1) small holes (diameter < 5 cm) mostly used by tits,
2) large holes (diameter 5–10 cm) made by small to medi-
um-sized woodpecker species (e.g. great spotted wood-
pecker), and 3) very large (oval) holes (longest diameter
>10 cm) made by large woodpecker species (e.g. black

329 ECOLOGICAL BULLETINS 51, 2004woodpecker). Additionally, the occurrence of nest-boxes
for cavity nesters is recorded in order to be able to qualify
the availability of nesting habitat for cavity nesters.
Function
Species have adapted to certain elements and structures of
a landscape in an ecoregion. Most European landscapes
have a long history of human land use. As a result, special-
ised species on certain structures of natural forests and pre-
industrial landscapes become vulnerable or even locally or
regionally extinct. T o maintain viable populations of such
species, certain structural elements or vegetation types may
have to be continuously maintained or even restored.
Maintenance of biodiversity in the long term should there-
fore include also processes that affect habitat renewal at
different spatial scales (Larsson et al. 2001, Norton 2003).
Natural biotic and abiotic disturbances maintaining the
amount and types of dead wood, different microhabitats
and regeneration of tree species are important examples.
An example at the scale of patches of trees is the distribu-
tion of decay stages of dead wood and uprooted trees
(Stokland 2001). Signs of insect or fungal outbreaks and
other natural processes as well as signs of land abandon-
ment are useful additional measurements to demonstrate
the occurrence of habitat renewal at the stand scale.
Although processes that maintain functions are more
subtle elements of biodiversity than species and structures,
they are important for the maintenance of different distur-
bance regimes. Altered fire frequencies (Niklasson and
Granström 2000), hydrologic regimes (Bergquist 1999)
and air pollution causes leaching of nutrients such as nitro-
gen from sensitive soils and changes in vegetation in some
regions (Ellenberg 1996) are abiotic examples. Past fire
events, different forest practices and past agriculture activ-
ities may all influence the present situation in a stand.
Landscape changes favouring generalist predators that af-
fect breeding success of forest birds (Kurki et al. 2000) and
browsing by superabundant wild herbivores on certain de-
ciduous trees species (Angelstam et al. 2000) that modify
forest composition are two examples of altered trophic in-
teractions. Socio-economic changes in rural communities
followed by land abandonment constitute another exam-
ple (Angelstam et al. 2003b).
Land management and abandonment
If possible, the silvicultural system of management (e.g.
clear-cutting, shelterwood, etc.) and direct signs of timber
use such as wood-storage are recorded. Thus the age (fresh/
old) and spatial distribution of the stumps in the sampling
plot is classified as single, widespread, or clumped and is
noted in the field protocol. Similarly, we record evidence of
ongoing traditional management of trees by pollarding
and coppicing.The abandonment of woodland pastures and other
types of pre-industrial cultural landscapes are other major
processes affecting biodiversity. Almost all over Europe the
forest area is increasing, either because policies encourage
increased forest cover or because forest is encroaching on
former agricultural land (Nilsson et al. 1992, Carey et al.
2003). The effects on biodiversity are diverse and complex
including both loss of important habitat types in the cul-
tural landscape and initiation of secondary succession of
use for other species (Mikusi ński and Angelstam 1999,
Mikusiński et al. 2003).
Marks and Gardescu (2001) reviewed how site history
can be inferred from evidence that can directly be observed
from a careful search through a piece of forest or wood-
land. The presence of signs of local charcoal production,
stumps made by ax, chainsaw or a harvester is therefore
recorded. In remote areas in northern Scandinavia and
Russia, very old and high stumps (0.5–1 m) are signs of
logging made during winter very early in the forest history,
i.e. during the phase of high-grading. This phenomenon
can also be found in the Alps. There, high stumps are left as
a natural “hazard control system” to protect against ava-
lanches, land slides and rock falls. Note that the absence of
stumps in a plot may have two alternative causes: 1) the
stand has never been harvested or has not been harvested
for a very long time, or 2) the stand was established on
formerly open land, such as abandoned agricultural fields.
T rees that indicate a historically more open landscape
such as those having branches close to the ground, dead
lateral but living vertical branches, large crowns and old
trees (e.g. Juniperus ,Sorbus ,Quercus ) encroached by shade- Quercus ) encroached by shade- Quercus
tolerant tree species (Marks and Gardescu 2001, Rackham
2003) are also recorded. The presence of afforested or over-
grown fields, fences of stone walls, trees, or barbed wire,
juniper (live or dead), wide-crowned trees, nitrogen-loving
species such as elder Sambucus spp. and nettle Sambucus spp. and nettle Sambucus Urtica spp. Urtica spp. Urtica
are noted. Damaged or destroyed old farming houses and
stables, stone walls and other signs of human use are re-
corded as well.
Natural abiotic and biotic processes
Both abiotic and biotic processes are important for the
maintenance of biodiversity (e.g. Nolet and Rosell 1998,
Ulanova 2000). In areas with fire as a natural disturbance
regime the presence of fire scars in living trees and in
stumps is therefore noted to assess past stand events. Signs
of avalanches, trees broken by wind, uprooted trees, ero-
sion, beaver-dams as well as wood-decaying bracket fungi
and bark beetle Ips spp. (Coleoptera: Scolytidae) infesta-Ips spp. (Coleoptera: Scolytidae) infesta- Ips
tions in the sampling plots are therefore recorded. Uproot-
ed trees are trees with an exposed root system, an impor-
tant substrate for several mosses. Finally, trees broken due
to fungal attack and top-broken trees as primarily a result
of high snow pressure in stands of high altitude or latitude
are recorded.

330 ECOLOGICAL BULLETINS 51, 2004Large mammal herbivory
The effect of large herbivores on the forest at the landscape
scale is measurable through the intensity of browsing on
preferred shrubs (e.g. Anon. 1988, Angelstam et al. 2000).
Droppings of red deer Cervus elaphus , moose Alces alces
and other large herbivores are noted, as well as tracks and
game crossings. Note that moose bark peeling on young
pine and on middle aged spruce/aspen just above breast
height, or even higher made in winter can co-occur with
deer damage. Presence/absence of deer damage like brows-
ing, bark peeling, velvet rubbing as well as the presence of
livestock in the forest is identified by the observation of
livestock dung and recorded in a specific field.
Predation
Sightings and calls of corvids (see above) and large avian
predators, scats of red fox Vulpes vulpes and marten Vulpes vulpes and marten Vulpes vulpes Martes
sp. are noted. The relative occurrence of generalist preda-
tors indicates the predation pressure (e.g. Andrén et al.
1985).
Human disturbance
Information about the presence of roads, river/brook, ski-
slopes (winter tourism), and hiking trails show the accessi-
bility and therefore vulnerability of an area to human dis-
turbance (seasonal or year-around). This information can
also be extracted from the maps.
References
Aksenov, D. et al. 2002. Atlas of Russia’s intact forest landscapes.
– Global Forest Watch Russia, Moscow.
Andrén, H. et al. 1985. Differences in predation pressure in rela-
tion to habitat fragmentation. – Oikos 45: 273–277.
Angelstam, P . 1996. Ghost of forest past – natural disturbance
regimes as a basis for reconstruction of biologically diverse
forests in Europe. – In: DeGraaf, R. and Miller, R. I. (eds),
Conservation of faunal diversity in forested landscapes.
Chapman and Hall, pp. 287–337.
Angelstam, P . 1998a. Maintaining and restoring biodiversity by
developing natural disturbance regimes in European boreal
forest. – J. Veg. Sci. 9: 593–602.
Angelstam, P . 1998b. T owards a logic for assessing biodiversity in
boreal forest. – In: Bachmann, P ., Köhl, M. and Päivinen, R.
(eds), Assessment of biodiversity for improved forest plan-
ning. Kluwer, pp. 301–313.
Angelstam, P . and Mikusi ński, G. 1994. Woodpecker assemblag-
es in natural and managed boreal and hemiboreal forest – a
review. – Ann. Zool. Fenn. 31: 157–172.
Angelstam, P . and Kuuluvainen, T . 2004. Boreal forest distur-
bance regimes, successional dynamics and landscape struc-
tures – a European perspective. – Ecol. Bull. 51: 117–136.
Angelstam, P ., Breuss, M. and Gossow, H. 2000. Visions and
tools for biodiversity restoration in coniferous forest. – Proc.
from the International Conference on Forest Ecosystem Res-
toration, Vienna, pp. 19–28.Angelstam, P . et al. 2002a. T rends in forest biodiversity compo-
nents along an altitudinal gradient in the Etsch river valley in
South T yrol, N Italy. – Amt für Naturparke, Bozen, Italy.
Angelstam, P . et al. 2002b. Effects of forest structure on the pres-
ence of woodpeckers with different specialisation in a land-
scape history gradient in NE Poland. – In: Chamberlain, D.
and Wilson, A. (eds), Proc. of the 2002 Annual IALE (U.K.)
Conference, pp. 25–38.
Angelstam, P . et al. 2003a. Gap analysis and planning of habitat
networks for the maintenance of boreal forest biodiversity in
Sweden – a technical report from the RESE case study in the
counties Dalarna and Gävleborg. – Dept of Natural Scienc-
es, Örebro Univ., Dept of Conservation Biology, Swedish
Univ. of Agricultural Sciences, Länsstyrelsen Dalarnas län
Rapport 2003:26 and Länsstyrelsen Gävleborg Rapport
2003:12.
Angelstam, P . et al. 2003b. Assessing village authenticity with sat-
ellite images – a method to identify intact cultural landscapes
in Europe. – Ambio 33: 594–604.
Angelstam, P . et al. 2004a. Habitat modelling as a tool for land-
scape-scale conservation – a review of parameters for focal
forest birds. – Ecol. Bull. 51: 427–453.
Angelstam, P . et al. 2004b. Land management data and terrestrial
vertebrates as indicators of forest biodiversity at the land-
scape scale. – Ecol. Bull. 51: 333–349.
Anon. 1988. Schweizerisches Landesforstinventar. – Berichte Nr.
305, Eidg. Anstalt für das forstliche Versuchswesen, Bir-
mensdorf, Schweiz, in German.
Anon. 2002. Decision No 1600/2002/EC of the European par-
liament and of the council of 22 July 2002 laying down the
Sixth Community Environment Action Programme. – Offi-
cial Journal of the European Communities.
Arnborg, T . 1990. Forest types of northern Sweden. – Vegetatio
90: 1–13.
Atlegrim, O. and Sjöberg, K. 1995. Lepidoptera larvae as food
for capercaillie chick ( T etrao urogallus ): a field experiment. – T etrao urogallus ): a field experiment. – T etrao urogallus
Scand. J. For. Res. 278–283.
Bader, P ., Jansson, S. and Jonsson, B. G. 1995. Wood-inhabiting
fungi and substrate decline in selectively logged boreal spruce
forests. – Biol. Conserv. 72: 355–362.
Baines, D., Baines, M. M. and Sage, R. B. 1993. The importance
of large herbivore management to woodland grouse and their
habitats. – In: Lovel, T . and Hudson, P . (eds), Proc. of the 6th
International Grouse Symposium, pp. 93–100.
Bates, J. W . and Farmer, A. M. 1992. Bryophytes and lichens in a
changing environment. – Oxford Univ. Press.
Berg, Å. et al. 1994. Threatened plant, animal and fungus species
in Swedish forests: distribution and habitat associations. –
Conserv. Biol. 8: 718–731.
Berglind, S.-Å. 2004. Area-sensitivity of the sand lizard and spi-
der wasps in sandy pine heath forests – umbrella species for
early successional biodiversity conservation? – Ecol. Bull. 51:
189–207.
Bergquist, B. 1999. Impact of land use and buffer zones on
stream environments in woodland and agricultural areas – a
literature review. – Fiskeriverket Rapport 3: 5–118, in Swed-
ish.
Breuss, M. 1999. Untersuchungen zum winterlichen
Nahrungsspektrum des Haselhuhns Bonasa bonasia in den Bonasa bonasia in den Bonasa bonasia
Gailtaler Kalkalpen. – Diploma thesis at the Dept of Wildlife
Biology and Game Management, Univ. of Agricultural Sci-
ences Vienna, in German.

331 ECOLOGICAL BULLETINS 51, 2004Brokaw, N. and Lent, R. 1994. Vertical structure. – In: Hunter,
M. L. (ed.), Maintaining biodiversity in forest ecosystems.
Cambridge Univ. Press, pp. 373–399.
Bütler, R., Angelstam, P . and Schlaepfer, R. 2004. Quantitative
snag targets for the three-toed woodpecker Picoides tridacty-
lus. – Ecol. Bull. 51: 219–232.
Carey, P . D. et al. 2003. The multi-disciplinary evaluation of a
national agri-environment scheme. – J. Environ. Manage.
69: 71–91.
Davis, L. S. et al. 2001. Forest management. T o sustain ecologi-
cal, economic, and social values. – McGraw Hill.
Derleth, P ., Bütler, R. and Schlaepfer, R. 2000. Le pic tridactyle
(Picoides tridactylus ): un indicator de la qualité écologique de Picoides tridactylus ): un indicator de la qualité écologique de Picoides tridactylus
l’écosystème forestier du Pays-d’Enhaut (Préalpes suisses). –
Schweiz. Z. Forstwiss. 151: 282–289, in French.
Drakenberg, B. and Lindhe, A. 1999. Indirekt naturvärdes-
bedömning på beståndsnivå – en praktiskt tillämpbar me-
tod. – Skog & Forskning 2: 60–66, in Swedish.
Dumortier, M. et al. 2002. Predicting vascular plant richness of
fragmented forests in agricultural landscapes in central Bel-
gium. – For. Ecol. Manage. 158: 85–102.
Ehnström, B. and Axelsson, R. 2002. Insektsgnag i bark och ved.
– ArtDatabanken, Swedish Univ. of Agricultural Science,
Almkvist och Wiksell, Uppsala, in Swedish.
Ellenberg, H. 1996. Vegetation Mitteleuropas mit den Alpen, 5
Auflage. – Eugen Ulmer, Stuttgart, in German.
Esseen, P .-A., Renhorn, K.-E. and Pettersson, R. B. 1996. Epi-
phytic lichen biomass in managed and old-growth boreal
forests: effect of branch quality. – Ecol. Appl. 6: 228–238.
Esseen, P . A. et al. 1997. Boreal forests. – Ecol. Bull. 46: 16–47.
Faliński, J. B. 1986. Vegetation dynamics in temperate lowland
primeval forests. – Geobotany 8, Dordrecht, Dr W . Junk
Publ.
Frid, H. 2001. T rends in components of biodiversity in a Scottish
land-use history gradient. – Diploma thesis at the Dept of
Conservation Biology, Swedish Univ. of Agricultural Scienc-
es.
Gustafsson, L. 1999. Thoughts behind the use of indicator spe-
cies in practical forestry in Sweden. – Sv. Bot. Tidskr. 92: 89–
103.
Gustafsson, L. 2001. Red-listed species and indicators: vascular
plants in woodland key habitats and surrounding production
forests in Sweden. – Biol. Conserv. 92: 35–43.
Gustafsson, L. 2002. Presence and abundance of red-listed plant
species in Swedish forests. – Conserv. Biol. 16: 377–388.
Gustafsson, L., de Jong, J. and Norén, M. 1999. Evaluation of
Swedish woodland key habitats using red-listed bryophytes
and lichens. – Biodiv. Conserv. 8: 1101–1114.
Hägglund, B. and Lundmark, J.-E. 1987. Markvegetationsty-
per, skogsmarksflora. – Skogsstyrelsen, Jönköping, in Swed-
ish.
Hansson, L. 2001. Key habitats in Swedish managed forests. –
Scand. J. For. Res. Suppl. 3: 52–61.
Harmon, M. E. et al. 1986. Ecology of coarse woody debris in
temperate ecosystems. – Adv. Ecol. Res. 15: 133–202.
Höjer, O. 1999. Indicators of forest biodiversity – an evaluation
of specialist species and forest components in two boreal
landscapes. – Diploma thesis at the Dept of Conservation
Biology, Swedish Univ. of Agricultural Sciences.
Jansson, G. and Andrén, H. 2003. Habitat composition and bird
diversity in managed boreal forests. – Scand. J. For. Res. 18:
225–236.Jonsell, M., Weslien, J. and Ehnström, B. 1998. Substrate re-
quirements of red-listed saproxylic invertebrates in Sweden.
– Biodiv. Conserv. 7: 749–764.
Jonsson, B., Jacobsson, J. and Kallur, H. 1993. The forest man-
agement planning package. Theory and application. – Stud.
For. Suec. 189.
Jonsson, B. G. and Kruys, N. (eds) 2001. Ecology of woody de-
bris in boreal forests. – Ecol. Bull. 49.
Karström, M. 1992. Steget före i det glömda landet. – Sv. Bot.
Tidskr. 86: 115–146, in Swedish.
Kurki, S. et al. 2000. Landscape fragmentation and forest com-
position effects on breeding success in boreal forests. – Ecol-
ogy 81: 1985–1997.
Larsson, T .-B. et al. (eds) 2001. Biodiversity evaluation tools for
European forest. – Ecol. Bull. 50.
Marks, P . L. and Gardescu, S. 2001. Inferring forest stand history
from observational field evidence. – In: Egan, D. and How-
ell, E. A. (eds), The historical ecology handbook. Island
Press, pp. 177–198.
Martikainen, P ., Kaila, L. and Haila, Y. 1998. Threatened beetles
in white-backed woodpecker habitats. – Conserv. Biol. 12:
293–301.
Martin, K. and Eadie, J. M. 1999. Nest-webs: a community-
wide approach to the management and conservation of cavi-
ty-nesting forest birds. – For. Ecol. Manage. 115: 243–257.
Mayer, H. 1984. Wälder Europas. – Gustav Fischer, in German.
Mikusiński, G. and Angelstam, P . 1998. Economic geography,
forest distribution, and woodpecker diversity in central Eu-
rope. – Conserv. Biol. 12: 200–208.
Mikusiński, G. and Angelstam, P . 1999. Man and deciduous
trees in boreal landscape. – In: Kovar, P . (ed.), Nature and
culture in landscape ecology. Experience for the 3rd Milleni-
um. The Karolinum Press, Prague, pp. 220–224.
Mikusiński, G., Gromadzki, M. and Chylarecki, P . 2001. Wood-
peckers as indicators of forest bird diversity. – Conserv. Biol.
15: 208–217.
Mikusiński, G., Angelstam, P . and Sporrong, U. 2003. Distribu-
tion of deciduous stands in villages located in coniferous for-
est landscapes in Sweden. – Ambio 33: 520–526.
Muir, R. 1999. Approaches to landscape. – McMillan Press.
Mykrä, S., Kurki, S. and Nikula, A. 2000. The spacing of mature
forest habitat in relation to species-specific scales in managed
boreal forests in NE Finland. – Ann. Zool. Fenn. 37: 79–91.
Niklasson, M. and Granström, A. 2000. Numbers and size of
fires: long term trends in Swedish boreal landscape. – Ecol-
ogy 81: 1484–1499.
Nikolov, N. and Helmisaari, H. 1992. Silvics of the circumpolar
boreal forest tree species. – In: Shugart, H. H., Leemans, R.
and Bonan, G. B. A. (eds), Systems analysis of the global
boreal forest. Cambridge Univ. Press, pp. 13–84.
Nilsson, S., Sallnäs, O. and Duinker, P . 1992. Future forest re-
sources of western and eastern Europe. – The Parthenon
Publ. Group, Carnforth, U.K.
Nilsson, S. G. et al. 1995. T ree-dependent lichens and beetles as
indicators in conservation forests. – Conserv. Biol. 9: 1208–
1215.
Nilsson, S. G., Hedin, J. and Niklasson, M. 2001. Biodiversity
and its assessment in boreal and nemoral forest. – Scand. J.
For. Res. Suppl. 3: 10–26.
Nilsson, S. G. et al. 2002. Densities of large living and dead trees
in old-growth temperate and boreal forests. – For. Ecol.
Manage. 161: 189–204.

332 ECOLOGICAL BULLETINS 51, 2004Nolet, B. A. and Rosell, F . 1998. Comeback of the beaver ( Castor
fiber): an overview of old and new conservation problems. –fiber): an overview of old and new conservation problems. – fiber
Biol. Conserv. 83: 165–173.
Nordengren, C., Hofgaard A. and Ball, J. 2003. Availability and
quality of herbivore winter browse in relation to tree height
and snow depth. – Ann. Zool. Fenn. 40: 305–314
Norén, M. et al. 2002. Handbok för inventering av nyckelbi-
otoper. – Skogsstyrelsen, Jönköping, in Swedish.
Norton, B. G. 2003. Searching for sustainability. Interdiscipli-
nary essays in the philosophy of conservation biology. –
Cambridge Univ. Press.
Noss, R. F . 1990. Indicators for monitoring biodiversity: a hierar-
chical approach. – Conserv. Biol. 4: 355–364.
Oliver, I. and Beattie, A. J. 1993. A possible method for the rapid
assessment of biodiversity. – Conserv. Biol. 7: 562–568.
Oliver, C. D. and Larsen, B. C. 1996. Forest stand dynamics. –
McGraw-Hill.
Peer, T . 1995. Die natürliche Pflanzendecke Südtirols. – Amt für
Landschaftsplanung, Bozen, in German.
Pennanen, J. 2002. Forest age distribution under mixed-severity
fire regimes – a simulation-based analysis for middle boreal
Fennoscandia. – Silva Fenn. 36: 213–231.
Peterken, G. 1996. Natural woodland: ecology and conservation
in northern temperate regions. – Cambridge Univ. Press.
Pettersson, R. B. et al. 1995. Invertebrate communities in bore-
al forest canopies as influenced by forestry and lichens with
implications for passerine birds. – Biol. Conserv. 74: 57–
63.
Puumulainen, J. 2001. Structural, compositional and functional
aspects of forest biodiversity in Europe. – Geneva timber and
forest discussion papers. United Nations, New York and Ge-
neva.
Quinn, G. P . and Keough, M. J. 2002. Experimental design and
analysis for biologists. – Cambridge Univ. Press.Rackham, O. 2003. Ancient woodland. – Castlepoint Press, Col-
vend, Dalbeattie.
Roberge, J.-M. and Angelstam, P . 2004. Usefulness of the um-
brella species concept as a conservation tool. – Conserv. Biol.
18: 76–85.
Siitonen, J. 2001. Forest management, coarse woody debris and
saproxylic organisms: Fennoscandian boreal forests as an ex-
ample. – Ecol. Bull. 49: 11–41.
Stokland, J. 2001. The coarse woody debris profile: an archive of
recent forest history and an important biodiversity indicator.
– Ecol. Bull. 49: 71–83.
Storch, I. 1999. Auerhuhnschutz: Aber wie? – Wildbiologische
Gesellschaft München, in German.
Suchant, R. and Braunisch, V . 2004. Multidimensional habitat
modelling in forest management – a case study using caper-
caillie in the Black Forest, Germany. – Ecol. Bull. 51: 455–
469.
Swenson, J. E. and Angelstam, P . 1993. Habitat separation by
sympatric forest grouse in Fennoscandia in relation to forest
succession. – Can. J. Zool. 71: 1303–1310.
Thompson, I. D. and Angelstam, P . 1999. Special species. – In:
Hunter, M. L. (ed.), Maintaining biodiversity in forest eco-
systems. Cambridge Univ. Press, pp. 434–459.
Tibell, L. 1992. Crustose lichens as indicators of forest continuity
in boreal coniferous forests. – Nord. J. Bot. 12: 427–450.
Ulanova, N. G. 2000. The effects of windthrow on forests at dif-
ferent spatial scales: a review. – For. Ecol. Manage. 135: 155–
167.
Uliczka, H., Angelstam, P . and Roberge, J.-M. 2004. Indicator
species and biodiversity monitoring systems for non-indus-
trial private forest owners – is there a communication prob-
lem? – Ecol. Bull. 51: 379–384.
Wikars, L.-O. 2004. Habitat requirements of the pine wood-liv-
ing beetle T ragosoma depsarium (Coleoptera: Cerambycidae)
at log, stand, and landscape scale. – Ecol. Bull. 51: 287–294.
View publication statsView publication stats