Virgin forests as a knowledge source for central European silviculture: reality or myth?

Virgin forests as a knowledge source for central European silviculture: reality or myth?
For. Snow Landsc. Res. 79, 1/2: 19–32 (2005)                                                       19

Virgin forests as a knowledge source for central European
silviculture: reality or myth?

Peter Brang

WSL Swiss Federal Research Institute, Zürcherstrasse 111, CH-8903 Birmensdorf, Switzerland.

In central Europe, the few remnants of virgin forest left have often been studied for reasons based
on the following line of argument: Such studies increase our understanding of forest dynamics
under a natural disturbance regime and are therefore an important basis for close-to-nature silvi-
culture. This paper reviews the history of this idea, examines its merits and drawbacks focusing on
selected silvicultural issues, and explores the potential for using virgin forest research as a source
of ideas for developing silviculture in future.
    Research on virgin forests has had important implications for managed forests in the case of
maintaining nurse logs as a seedbed for tree seedlings, of knowing the maximum age and size of
trees, and, to a minor extent, of identifying the habitat requirements of species which are found
mainly in virgin forests. However, past research on virgin forests has not provided any major
direct contributions to defining close-to-nature silviculture. The main obstacles to knowledge
transfer seem to be: (i) the scarcity of virgin forest remnants in central Europe, (ii) differences
between the stand histories and current stand structures in virgin and managed forests, and (iii)
the better opportunities to study phenomena of interest directly in managed forests, with easier
access and possibilities of using manipulative approaches (field experiments) with proper repli-
cation. Current trends in central European forests and silviculture, such as the creation of forest
reserves, the increase in growing stock, the shift to continuous-cover forestry and the increasing
importance of forests as habitats for endangered species, suggest there is still much to learn from
virgin forest remnants, and to justify further research. However, virgin forests should only be
studied when an issue can best be addressed by using them as the object of study. Otherwise,
managed forests should be preferred, or a combination of virgin and managed forests.

Keywords: virgin forests, forest reserves, protected areas, close-to-nature silviculture, central
Europe, forest dynamics, research history

1       Introduction

In 1905, Julius Fröhlich started his career as a trainee in the forestry department of Bosnia.
At this time, about 50 % of the total forest area of Bosnia of two million ha was considered
virgin forest (FRÖHLICH 1954, p. 9), i.e. forest that has not had any human-induced treat-
ment (LUND 2002). Fröhlich surveyed these forests and planned their exploitation. Today,
most of these forests have lost their virgin status. Twenty-seven “strict forest reserves”
totalling 3125 ha in area remain in Bosnia (PARVIAINEN et al. 2000). Only five of these
reserves can be considered as “virgin forest reserves” (PINTARIÇ 1999). The total area of
virgin forest remnants in central Europe is difficult to estimate since this area depends on
the definition of the notion “virgin forest” (LUND 2002). Strictly protected forest areas cover
about 100 000 ha in Austria, Belgium, the Czech Republic, France, Germany, Hungary, The
Netherlands, Poland, Slovakia and Switzerland (PARVIAINEN et al. 2000). However, only a
minor part of these forests can be considered to be virgin forests. In Scandinavia and eastern
Europe, larger patches with little human impact still exist (PARVIAINEN et al. 2000).
20                                                                                        Peter Brang

    Forest scientists and managers have often stated that forest management could profit
from a better understanding of virgin forest dynamics. For example, RUBNER (1920) claimed
that we could learn three things from virgin forests: Whether stands are naturally mixed or
mono-specific; whether they are uniform in structure or heterogeneous; and how natural
regeneration occurs. Rubner was an early pioneer in promoting virgin forests as objects of
study. Since about 1950, a large body of scientific studies and papers dealing with virgin forest
remnants in Europe has developed. Indeed, the reserve status of many of these remnants
was promoted and secured by forest managers and scientists.
    This interest in virgin forests has often been justified by the following line of argument:
These forests have not been subject to human impact, and studying them therefore offers us
opportunities to advance our understanding of forest structure and dynamics (LEIBUNDGUT
1959). This is particularly interesting given the current trend towards a more nature-oriented
silviculture, which largely relies on using natural processes to reach silvicultural goals (OTTO
1995; SCHÜTZ 1999, 2004). It is even possible to completely rely on natural processes, with-
out any silvicultural interventions, which is appealing on steep-slope protection forests
where interventions are costly (HILLGARTER 1978; BRANG et al. 2004b).
    In this paper, I will first review the history of scientific interest in virgin forests in Europe.
Second, I will examine the conceptual foundation of this interest, using examples in which
research results obtained in virgin forests have been applied to forest management. Third, I
will explore the potential for future contributions from research on virgin forests to designing
silvicultural systems.

2        History of scientific interest in virgin forests

Interest in virgin forests is linked to the veneration of nature in general, as inspired by the
Swiss-French philosopher Rousseau (1712–1778). In the 19th century, interest in natural
forests increased partly as a counter-movement to a more and more rational and technical
world, and also to man-made uniform forests which often rather resemble arable land. This
emotional appeal of natural forests can be illustrated by a statement by WESSELY (1853),
who described an expedition into a virgin forest in Austria: “… intruding into this wilderness
where man had apparently not laid his foot had an unutterable appeal, which no one was
able to withstand. It was a feeling which may have been the same as that of the great sailors
when they discovered new continents1.” Also today, at least three different virgin forests are
called “Cathedral Grove” (Internet search, 2 December 2004), which illustrates how these
forests are valued.
   GÖPPERT (1868) was one of the earliest foresters to study virgin forests in Europe.
GAYER (1898), BIOLLEY (1901), GURNAUD (1886) and ENGLER (1900) promoted a type of
naturalistic silviculture that was in contrast to the common practice of large-scale clearcut-
ting. Their silvicultural thinking prepared the terrain for a new interest in virgin forests,
which were, during their life, still being harvested and thus transformed into managed
forests in eastern and south-eastern Europe (FRÖHLICH 1954).
   CERMAK (1910), in his paper on “Silvicultural issues in virgin forests”, described the forest
composition and structures as well as disturbance regimes in virgin forests in Bosnia-
Herzegovina. By “silviculture”, he still meant the sustainable use of the remaining virgin
forests. One of the early citations attributing a particular value to virgin forests for silviculture
in managed forests can be found in a paper by RUBNER (1920), with the title “Silvicultural

1    German citations have been translated by the author.
For. Snow Landsc. Res. 79, 1/2 (2005)                                                          21

conclusions from virgin forests”: “With increasing culture and more sophisticated tech-
niques, man will not detach himself from nature, but will increase his ties with it and govern
it, by knowing it and subordinating himself to its restrictions”. BASELER (1932, p. 39)
reviewed early research on virgin forests. According to him, the scientists investigating virgin
forests thought that these forests “should become an educational aid for managing cultural
forests in a close-to-nature manner”. Later, AUBRÉVILLE (1938) published a book on
African rain forests. In his account of forest dynamics, he already developed the mosaic
cycle theory, which was later refined by REMMERT (1991).
    After World War II, the Iron Curtain made it difficult for scientists from western countries
to study virgin forest remnants in eastern Europe. In spite of this, the uniqueness of these
remnants and their scarcity in western Europe stimulated international cooperation. Several
scientists had already started to study the structure and dynamics of virgin forests in the
1950s (and even 30s, ZLATNIK 1938). It is no coincidence that these scientists (PINTARIÇ
1959; MLINSEK 1978, 1993; LEIBUNDGUT 1982; MAYER et al. 1989; KORPEL’ 1995) advocated
a type of silviculture which was partly in contrast to the contemporary practise. They were
against schematic clearcutting and in favour of small-scale interventions taking the particu-
larities of each stand, and even of each tree, into account. The title of a paper by VÉZINA
(1959) clearly conveys the link between virgin forest studies and close-to-nature silviculture:
“Contribution to the study of virgin forests as a basis for developing a silvicultural system
which is more close-to-nature.” MLINSEK (1978) hoped to find the knowledge basis for
close-to-nature silviculture in untouched forests. Research and observations in virgin forests
were intended to stimulate the conversion of forests heavily influenced by humans into
more natural forests.
    The idea of conversion was actually present much earlier when the question of whether
virgin forests more closely resemble even-aged or uneven-aged forests was first raised. In
my opinion, it is not clear why this should matter so much since there is no need for managed
forests to closely resemble virgin forests. At the beginning of the 20th century, however, this
was a matter of debate. While some authors (WESSELY 1853; KOPEZKY 1895) compared virgin
to single-tree selection forests (plenter forests, “Plenterwald” in German), others (CERMAK
1910; RUBNER 1920; GEHRHARDT 1923) claimed that they more closely resemble even-aged
forests. BASELER (1932) provides a good overview of the different opinions. It becomes
clear that the answers vary according to site conditions, which influence the occurrence of
tree species.
    Leibundgut advocated an approach to silviculture with a free conduction of cuts (“freie
Hiebsführung” in German, SCHÜTZ 1994), which made full use of the silvicultural “toolbox”.
He considered forests not as tree farms, but as dynamic ecosystems, although he clearly
focused on the tree layer and not on the whole forest ecosystem. This viewpoint was there-
fore already ecosystem-based, although today it would still seem narrow. According to this
ecosystem-based view, virgin forests are an interesting object of study. This argumentation
continued with, e.g., FABIJANOWSKI (1978), PRUSA (1985), ZUKRIGL (1990), and BERNADZKI
et al. (1998).
    In spite of the apparent stress laid on the importance of studying virgin forests, interest in
virgin forests at the time was of a rather general nature. The main agenda of the scientists
mentioned above was strengthening the scientific basis of “close-to-nature” or “nature-
oriented” silviculture. The real contributions of virgin forest studies to this agenda were not
clear. It rather seems that virgin forests were a fascinating object of study on their own, and
that these studies had, with “nature”, a common denominator with nature-oriented silvicul-
ture, but no direct link. Therefore, the conceptual justifications for research on virgin forests
seem, in retrospect, to have been quite vague until about 1990. The statement of objectives
of a monitoring program in forest reserves illustrates this vagueness: “Application to silvicul-
22                                                                                     Peter Brang

ture and forest management by deriving knowledge for treating production forests and for
conservation of species and habitats in forests” (ALBRECHT 1988). MAYER et al. (1972)
claimed that research on virgin forest remnants would deliver the basis for biological ration-
alisation: “A closer insight into the natural development dynamics will allow us to realize the
combined productive and social functions in our managed forests with a minimum of regu-
lating measures.” This is in line with the opinion of LEIBUNDGUT (1982, p. 16): “Virgin forest
research should deepen our knowledge about the characteristics of our tree species and
their coexistence in stands, and thus improve the fundamentals of a real close-to-nature silvi-
culture, so that we can achieve objectives at minimum expense.” KORPEL’ (1995, p. 1) had
the same point of view: “A close-to-nature forest management will essentially involve con-
sidering the regular processes and characteristics which take place in natural forests, without
human interference.”.
    Recently, these vague arguments favouring virgin forest studies have been augmented by
the concept of mimicking natural forest dynamics to preserve biodiversity, which was devel-
oped in North America (LERTZMAN et al. 1996; SPIES and TURNER 1999; FRANKLIN et al.
2002). This approach is also increasingly advocated in Europe (ANGELSTAM 1998;
PETERKEN 1999; BENGTSSON et al. 2000). To implement such an approach, however, a better
understanding of virgin forest dynamics is required. The following citation nicely illustrates
both justifications for virgin forest research, the old vague justification and a recent more
precise one: “It is generally accepted that natural forests are the basic model for the
realisation of nature-oriented silviculture. In strict forest reserves the development cycle of
natural forests can be observed, elucidated and understood, and these findings subsequently
mimicked in production forests” (PARVIAINEN et al. 2000).
    A recent attempt to apply virgin forest research to silviculture is the establishment of
paired plots in forest reserves and nearby managed forests (MEYER et al. 2004). If pairs have
similar site conditions, developmental stages and stand histories, they can directly be
compared, and the management influence on stand structure and habitat characteristics
    In line with the idea of mimicking disturbance regimes of virgin forests in production
forests, the interest in natural and virgin forests in Europe has considerably increased during
the last two decades because of their importance as habitats for virgin forest dependent
species (SCHUCK et al. 1994; DIACI 1999). In many countries, a network of forest reserves has
been established, and the development of these reserves is being studied (PARVIAINEN et al.
2000). In line with this interest in biodiversity, there is a clear shift away from pure studies of
stand structure, stand dynamics and regeneration to studies of virgin forests as a habitat for
diverse fauna and vegetation, including endangered species. The shift in research focus from
stand dynamics to habitat characteristics is coupled with a shift in scale, from studies at the
tree and stand levels to studies at the landscape level, according to the habitat requirements
of the species in question.
    In terms of research methods in virgin forests, the earliest accounts were based on
descriptions of the stand structures in extended areas, although the exact area described is
rarely given (WESSELY 1853). Sometimes, stem distributions of a few temporary plots that
were subjectively selected are presented to illustrate the key findings (SEIDEL 1848;
FRÖHLICH 1954). Today, such an approach would not be scientifically acceptable anymore. It
is fascinating, though, to read the purely descriptive account of RUBNER (1920) on virgin
forests, and to notice that most of his statements are in agreement with the results of recent
research, which were gained with much more sophisticated and time-consuming methods.
    Starting about in 1950, many permanent plots were established, and the developmental
phases mapped (KORPEL’ 1995). Recently, this still subjective approach has been com-
plemented with representative inventories at the landscape scale (Projektgruppe Natur-
For. Snow Landsc. Res. 79, 1/2 (2005)                                                          23

waldreservate 1993; Anonymus 2000), with paired plots of managed and unmanaged forest
as described above (MEYER et al. 2004), and with modelling (RADEMACHER et al. 2001).

3       Knowledge transfer from virgin to managed forests

3.1     Methods and examples used

In this section, I will ask in what respects studying virgin forests, and extrapolating the silvi-
cultural knowledge gained to managed forests, has been a successful approach. I will try to
answer this question using examples focusing on the following selected aspects of silviculture:
– Long-term competitive interactions between tree species
– Natural regeneration under canopy cover
– Maximum age and size of trees
– Forest dynamics without management operations
– Natural disturbance regimes
– Effects of tending and thinning measures on stem quality
– Habitat requirements of species which depend on virgin forests
As extrapolations in time and space can be problematic (BRANG et al. 2004a; KÖHL et al.
1995), I will first briefly review the conditions that should be met for valid extrapolations.

3.2     Extrapolation from virgin forests to managed forests

Forests vary greatly with respect to site factors (mainly climate, geology and soil), disturb-
ance regimes (e.g., frequency and extent of windthrow and snow break) and forest history
(e.g., current developmental stage, plant migration after the Ice Age). They are therefore
also highly diverse in terms of species composition and stand structure. For instance, the
mortality, establishment and growth rates of trees and disturbance regimes often depend on
site factors. Therefore, such rates can only be extrapolated with caution from one site to
another and should only be applied to a management context if the sites are sufficiently
similar. If research on virgin forests and nature reserves, which are in most cases studied as
single case studies, should result in silvicultural applications, this site-dependency could be a
barrier to knowledge transfer.
    Nevertheless, relatively static site factors such as climate, geology and soil may be quite
similar, even with distant sites. This similarity often shows up in the occurrence of similar
plant communities, of similar altitudinal sequences of plant communities, and in similar
height-diameter relationships among the trees. If we find a similar herb layer in a Fagus
sylvatica forest in the Ukrainian Transcarpathians and in Switzerland, and similar height
growth-stem diameter relationships (COMMARMOT et al. this issue), we can safely assume a
large similarity in site factors, and, since site factors are related to disturbance regimes, even
in the latter.
    Silvicultural decisions always have a long-term perspective and must therefore cope with
large uncertainty with respect to, e.g., future social needs. In the light of this uncertainty, it
seems unreasonable to set unrealistically high standards for knowledge transfer from
unmanaged to managed forests on similar sites. In research, however, in which generally
valid patterns are sought, similarity standards must be higher. For instance, for paired plots
of managed and unmanaged forests, plot selection criteria for ensuring sufficient similarity
must be strictly applied (MEYER et al. 2004).
24                                                                                  Peter Brang

   Historically, the inference made from research in virgin forests decreased over time,
paralleled by increasing knowledge about virgin forests and more strict standards in science
in general. Pure observations and temporary plots were used in the 19th and at the beginning
of the 20th century to draw general conclusions. Later, in the second half of the 20th century,
longitudinal case studies on small permanent plots were the dominant method, and the
conclusions were more site-specific. The case study character of this method, however, was
not always considered, and conclusions were sometimes made without sufficient replication.
Since about 1970, inventories in virgin forests with plots in systematic grids have increasingly
been used. In central Europe, however, we have to accept that the virgin forest remnants are
so scarce, and the establishment of many nature reserves so recent, that, for the majority
of the sites, we will not be able to quantitatively describe natural disturbance regimes with
satisfying certainty for several decades. The representativity gained with inventories remains
representativity for case studies, and not for site types.

3.3    Long-term competitive interactions between tree species

Many of the early reports on virgin forests mention competitive interactions between and
within tree species. For instance, the observed shade tolerance of seedlings and saplings of
Abies alba Mill., Picea abies (L.) Karst. and Fagus sylvatica L. was used to explain their
ability to outcompete light-demanding species (CERMAK 1910; RUBNER 1920). Although the
competition-induced mortality of light-demanding species in the presence of shade-tolerant
species has not yet been demonstrated with longitudinal data from virgin forests, it is
supported by practical experience and by evidence from pollen analysis (BURGA 1988).
Accordingly, Fagus sylvatica virgin forests are often poor in other tree species (FRÖHLICH
1947; KORPEL’ 1995; MEYER et al. 2002; COMMARMOT et al. this issue). These facts were
known to silviculturists, but could be confirmed in virgin forests, where stand structures
often differ from those found in managed forests. However, even if we had ample scientific
evidence about the competitive abilities of tree species from virgin forest remnants, it seems
difficult to transfer general patterns to managed forests since such abilities heavily depend
on site factors, or may be due to a particular stand history in the case studies. For instance,
SCHÜTZ (2004) found, in contrast with other studies, a competitive advantage of Fraxinus
excelsior L. over Acer pseudoplatanus L. in a small, man-made gap in a managed forest in
the Swiss central Plateau.

3.4    Natural regeneration under canopy cover

The study of natural regeneration processes was often mentioned as a justification for
research on virgin forests (HILLGARTER 1978; LEIBUNDGUT 1959, 1982). Long-term sup-
pression and release, the overlap between tree generations, the impact of browsing ungulates
on regeneration, and regeneration on nurse logs were of particular interest.
   The capacity of shade-tolerant tree species to grow vigorously after long-term suppression
has been well documented in and outside of virgin forests (e.g., SCHÜTZ 1969; CROSSLEY
1976; SEIDEL 1977). In central Europe, there is a sufficient area of single-tree selection
forests to study the phenomenon. Findings from virgin forests therefore only confirmed
results from managed forests.
   Many remnants owe their reserve status to their use as a hunting reserve (MAYER et al.
1989; e.g., Bialowieza: FABIJANOWSKI 1978; Neuwald: MAYER et al. 1972). Changes in tree
species composition as a result of long-term browsing pressure are often assumed, but are
For. Snow Landsc. Res. 79, 1/2 (2005)                                                          25

difficult to prove since longitudinal data on both species composition and browsing pressure
are unavailable. Studies from virgin forests confirm the general patterns of browsing pref-
erence for particular tree species. However, the interactions among ungulate population
densities, their feeding behaviour and seasonal migrations, and habitat suitability are
complex. Establishing generally valid patterns is difficult in managed forests alone, and,
therefore, extrapolations from virgin forests to managed forests are questionable as well.
   The case of regeneration on nurse logs is different (Fig. 1). In a recent literature review,
MAI (1998) summarizes 274 publications related to natural tree regeneration on dead wood.
Some species, in particular Picea abies, depend strongly on this substrate in cool-wet
environments such as many high-elevation forests. From MAI’s list, it becomes clear that
early accounts of regeneration on nurse logs are mostly based on observations in virgin
forests (GÖPPERT 1868; ENGLER 1904; CERMAK 1910; but see MOREILLON 1910). This
phenomenon could rarely be observed outside virgin forests since virtually all timber was
harvested in managed forests at that time, sometimes even including the stumps, and no
nurse logs were normally left. Regenerating subalpine Picea abies forests using stumps
(MOREILLON 1910) and nurse logs has often been proposed. At the beginning of the 20th
century, however, while this phenomenon was known (PILLICHODY 1910), it was not yet
translated into management recommendations (BAVIER 1910). We therefore know that
some species regenerate on nurse logs because we can observe it in virgin forests. Even
today, this phenomenon can rarely be observed outside forest reserves since the practice of
purposefully leaving coarse woody debris in stands is too recent.

                                                                      Fig. 1. Picea abies saplings
                                                                      on a downed log in the vir-
                                                                      gin forest remnant Scatlé,

3.5     Maximum age and size of trees

Most of our knowledge about the maximum age and size that trees of different species can
reach comes from virgin forests (e.g., KORPEL’ 1995). So far, this knowledge has not been
directly applied in managed forests since most trees are cut many decades before their bio-
logical death would naturally occur. This knowledge has indirectly been used for modelling
forest succession under the assumption of no management (BUGMANN and SOLOMON
2000), and for deriving minimum levels of regeneration in mountain forests based on the
remaining lifespan that current stands are expected to have (BRANG and DUC 2002).
26                                                                                 Peter Brang

However, there is no hard evidence for the assumption that maximum tree and stand ages
are identical in managed and unmanaged forests. In contrast to many virgin forests, only few
trees exhibit an early suppression phase in many managed forests, which may reduce the
maximum age of trees in the latter case. Predictions about the remaining lifespan of managed
forests are therefore uncertain. Knowledge about maximum tree age has also been used to
derive the duration of tree and whole stand generations, and to estimate the proportion of
different stages of development (KORPEL’ 1995), which has implications for the habitats
available historically for different organisms and thus for biodiversity.

3.6    Forest dynamics after abandonment of silvicultural operations

Our understanding of the dynamics of virgin forests has sometimes been used to assess the
implications of a no-intervention silvicultural strategy in mountain forests. This would be an
attractive option since silvicultural operations in protection forests (BACHOFEN and ZINGG
2001) on steep mountain slopes are costly. The difficulty is, however, that even if we
conclude that natural forest dynamics are unlikely to impair the protective effect of such
forests, this conclusion cannot be directly extrapolated to forests with a long history of
human impact. The starting conditions are different. For instance, the stand structures are
much more homogeneous in managed than in virgin forests under similar site conditions
(BRANG 2003).

3.7    Natural disturbance regimes

Early virgin forest definitions excluded early developmental stages resulting from large-
scale disturbances (BASELER 1932, p. 40), which are not likely to occur very often in central
European virgin forests (FRÖHLICH 1954; ZUKRIGL 1990). Knowledge of the frequency,
extent and intensity of natural disturbances of different types (e.g., windthrow) would be
helpful for designing anthropogenic disturbances (silvicultural interventions) which either
resemble natural disturbances, or minimize their occurrence (by enhancing resistance) or
their impact (by enhancing elasticity, BRANG 2001). However, the small size of the remaining
virgin forest remnants makes it very difficult to characterize natural disturbance regimes,
especially if large-scale disturbances from agents such as storms prevail. Therefore, designing
silvicultural operations to resemble natural disturbance regimes is not a promising approach
in central Europe. Moreover, one effect of the long-term management of the European
landscape has been an increase in overall species richness (WOHLGEMUTH et al. 2002).
Therefore, restoring virgin forests and natural disturbance regimes would increase the over-
all species richness only if it were restricted to certain parts of the forest area.

3.8    Effects of tending and thinning measures on stem quality

Interestingly, studies of virgin forests have rarely been used for scientifically underpinning
the need for tending and thinning measures in managed forests. If such measures are effec-
tive, the timber quality should be higher, and the species composition of the tree layer more
diverse in managed than in virgin forests. There are, however, several obstacles to such
comparisons. First, genetic differences can cause phenotypic variability among provenances,
and thus between virgin forests and remote managed forests. Second, the suppression phase
is usually much shorter in managed than in virgin forests. Therefore, the stem shape and
For. Snow Landsc. Res. 79, 1/2 (2005)                                                        27

branchiness of young trees in central European virgin forests is strongly influenced by com-
petition from overstory trees, while young trees in managed forests respond to competition
from trees of the same cohort (MLINSEK 1967). For Fagus sylvatica and Quercus petraea
(Matt.) Liebl., it has been observed that very long durations of suppression will impair the
ability of the trees to successfully release (KORPEL’, 1995). This is much less the case with
Abies alba and Picea abies (SCHÜTZ 1969).
   The obvious influence of tending and thinning measures on species composition is dif-
ficult to separate from an influence of the patch sizes of young stands and thus light
availability, which is usually greater in managed than in virgin forests. The early descriptions
of virgin forests often compare them to a particular type of single-tree selection forest: They
have a high growing stock and a closed canopy since gaps are filled with regeneration which
has reached the upper canopy and are thus dark (BASELER 1932).
   Thus, the potential for learning about tending and thinning from virgin forests is limited.
This and the excellent returns from forestry during most of the 20th century, which created
no need for reducing the costs in tending young stands, may explain the absence of com-
parative studies between tended young stands and virgin forests. The timber quality of virgin
forests mainly became an issue when they were exploited (FRÖHLICH 1954).

3.9     Habitat requirements of species which depend on virgin forests

In comparison to managed forests, virgin forests usually contain a higher proportion of large
trees (NILSSON et al. 2002; COMMARMOT et al. this issue) and larger volumes of dead wood
(coarse woody debris, HARMON et al. 1986; KORPEL’ 1997; SANIGA and SCHÜTZ 2002).
Some species depend on these characteristics, in particular birds (ANGELSTAM and
MIKUSINSKI 1994; WESO»OWSKI and TOMIA»OJC 1995; BÜTLER et al. 2004) and saproxylic
insects (GROVE 2002).
    Much of our knowledge about the range in quantities of dead wood in forests comes from
research on virgin forests or nature reserves. However, these levels and their fluctuations
(SANIGA and SCHÜTZ 2002) cannot be directly used as standard levels for managed forests
since the relationship between dead wood volumes (and quality) and habitat requirements
needs further study. Moreover, our knowledge about the habitat requirements of species
which appear to depend on virgin-forest type habitats originates less from virgin than from
managed forests (SCHIEGG 2000; BÜTLER et al. 2004). The importance of large trees as nesting
sites for rare birds was not first noted in remote virgin forests, but in the habitats of these
birds in managed forests.

3.10    Summary: Examples of knowledge transfer

In this section, I will summarize the examples given in the sections 3.3 to 3.9. Research on
virgin forests has had important implications for managed forests in the case of nurse logs as
a seedbed for tree seedlings, the maximum age and size of trees, and, to a minor extent, habitat
characteristics required for virgin forest dependent species. These topics concern objects
which have partly or almost fully disappeared in managed forests, and could therefore be
studied only in virgin forests. Research on virgin forests has been of minor importance for
managed forests in the cases of competitive interactions between tree species, natural regen-
eration under canopy cover, forest dynamics after abandonment of silvicultural operations,
natural disturbance regimes and the effects of tending and thinning measures on stem
quality. The main obstacles to knowledge transfer seem to be: (i) the scarcity of virgin forest
28                                                                                    Peter Brang

remnants in central Europe, which restricts the potential for site-specific replication and
thus generalisation, (ii) differences between the stand histories and current stand structures
of virgin and managed forests, and (iii) the opportunity to study phenomena of interest
directly in managed forests, with easier access and the possibility to use manipulative
approaches (field experiments) with proper replication.
   Past research on virgin forests did not, in contrast to earlier claims, provide direct con-
tributions to defining important characteristics of close-to-nature silviculture, such as a
prevalence of natural regeneration, reliance on tree species which naturally occur on a site,
use of shade as a principle for regulating intra-specific competition, reliance on opportunity
rather than deterministic decisions, and imitation of natural successional processes (SCHÜTZ
2004). These characteristics of close-to-nature silviculture were mainly developed in man-
aged forests, based on failures of silvicultural techniques far removed from natural processes
(OTTO 1995).

4      Outlook: Future silvicultural research in virgin forests

Virgin forests cannot be the unique reference point for managed forests. In landscapes with
a dense human population, the local demand for diverse forest products and services should
rather be the main guide to management.
   However, there are indeed trends in central European forests and silviculture which
might benefit from scientific findings from studying virgin forests since these trends make
managed forests more like virgin forests. The first trend is the creation of forest reserves and
national parks. Research on forest dynamics in virgin forests can help us predict the conse-
quences of this trend in terms of stand structure, species composition, habitat characteristics,
and corresponding influences on forest services such as protection against natural hazards
on steep slopes. In particular, research on virgin forests could contribute to a better under-
standing of stand disintegration and reinitiation phases, which were in the past short-circuited
by management. A second trend is the increase in growing stock, which has amounted to
about 1 % each year since about 1950 (Eidg. Forschungsanstalt WSL und Bundesamt für
Umwelt, Wald und Landschaft 2001, p. 77), and its consequences for forest products and
services. A similar increase can be observed in average stand age. A third trend is the trend
to continuous-cover forestry (“Dauerwald” in German), which is likely to reduce the mosaic
of developmental stages in central European forest landscapes, and will lead to stand
structures with a finer grain, more similar to that of virgin forests. Moreover, this trend is
also likely to favour shade-tolerant tree species (VON LÜPKE 2004), another characteristic of
central European virgin forests. A fourth trend is the increasing importance of forests as
habitats for endangered species. Large virgin forest remnants could help us to understand
spatial patterns of habitat use on different scales.
   In retrospect, there seems to be a gap in the history of research on central European
virgin forests between a strong motivation and a less strong argumentation. The virgin forest
label is appealing for research on silvicultural topics, but, on its own, insufficient as justifi-
cation. Virgin forests should be studied when an issue can best be addressed by using them
as objects of study. Otherwise, managed forests should be studied, or a combination of virgin
and managed forests. This principle applies to all issues where studying virgin forests is an
For. Snow Landsc. Res. 79, 1/2 (2005)                                                         29

I wish to thank Jean-Philippe Schütz and Andreas Zingg for helpful comments on an earlier version
of this paper, and Silvia Dingwall for improving the English.

5       References

ALBRECHT, L., 1988: Ziele und Methoden forstlicher Forschung in Naturwaldreservaten. Schweiz.
  Z. Forstwes. 139: 373–387.
ANGELSTAM, P.K., 1998: Maintaining and restoring biodiversity in European boreal forests by
  developing natural disturbance regimes. J. Veget. Sci. 9: 593–602.
ANGELSTAM, P.; MIKUSINSKI, G., 1994: Woodpecker assemblages in natural and managed boreal
  and semiboreal forests. A review. Ann. Zool. Fenn. 31: 157–172.
Anonymus, 2000: European Commission EUR 19550 – COST Action E4 – Forest reserves
  research network WG2 “Recommendations for Data Collection in Forest Reserves, with
  an Emphasis on Regeneration and Stand Structure”. Luxembourg, Office for Official
  Publications of the European Communities. 135–181.
AUBRÉVILLE, A., 1938: La forêt coloniale. Les forêts de l’Afrique Occidentale française. Annales,
  Académie des Sciences Coloniales, IX. Paris, Société d’Editions Géographiques, Maritimes et
  Coloniales. 244 pp.
BACHOFEN, H.; ZINGG, A., 2001: Effectiveness of structure improvement thinning on stand struc-
  ture in subalpine Norway spruce (Picea abies (L.) Karst.) stands. For. Ecol. Manage. 145:
BASELER, J., 1932: Urwaldprobleme in Nordanatolien. Mitt. Inst. ausl. Forstwirtsch. Tharandt 2:
  168 pp.
BAVIER, B., 1910: Welches sind die Ursachen des so häufigen Fehlens der natürlichen Verjüngung
  in alten Fichtenbeständen hoher Lagen, und wie kann dieser ungünstige Zustand beseitigt
  werden? Wie sind solche Bestände inskünftig zu behandeln? Schweiz. Z. Forstwes. 61: 145–152;
  195–201; 227–236.
BENGTSSON, J.; NILSSON, S.G.; FRANC, A.; MENOZZI, P., 2000: Biodiversity, disturbances, eco-
  system function and management of European forests. For. Ecol. Manage. 132: 39–50.
  positional dynamics of natural forests in the Bialowieza National Park, northeastern Poland. J.
  Veget. Sci. 9: 229–238.
BIOLLEY, H., 1901: Le jardinage cultural. J. forestier suisse 52: 67–104; 113–131.
BRANG, P., 2001: Resistance and elasticity: promising concepts for the management of protection
  forests in the European Alps. For. Ecol. Manage. 145: 107–119.
BRANG, P., 2003. Resistance and elasticity: a conceptual framework for managing secondary forest
  ecosystems in Switzerland. In: RAPPORT, D.J.; LASLEY, B.L.; ROLSTON, D.E.; NIELSEN, N.O.;
  QUALSET, C.O.; DAMANIA, A.B. (eds) Managing for healthy ecosystems. Boca Raton, Florida,
  USA, Lewis Publishers. 921–933.
BRANG, P.; DUC, P., 2002: Zu wenig Verjüngung im Schweizer Gebirgs-Fichtenwald: Nachweis mit
  einem neuen Modellansatz. Schweiz. Z. Forstwes. 153: 219–227.
BRANG, P.; SCHÖNENBERGER, W.; FISCHER, A., 2004a: Reforestation in central Europe: lessons
  from inter-disciplinary field experiments. For. Snow Landsc. Res. 78, 1/2: 53–70.
  dynamik unter Störungen und Eingriffen: Auf dem Weg zu einer systemischen Sicht. Forum für
  Wissen 2004: 55–66.
BÜTLER, R.; ANGELSTAM, P.; EKELUND, P.; SCHLAEPFER, R., 2004: Dead wood threshold values
  for the three-toed woodpecker, Picoides tridactylus. Biol. Conserv. 119: 305–318.
BUGMANN, H.K.M.; SOLOMON, A.M., 2000: Explaining forest composition and biomass across
  multiple biogeographical gradients. Ecol. Appl. 10: 97–114.
30                                                                                   Peter Brang

BURGA, C.A., 1988: Swiss vegetation history during the last 18000 years. New Phytol. 110:
CERMAK, L., 1910: Einiges über den Urwald von waldbaulichen Gesichtspunkten. Ges.
   Forstwes. 36: 340–370.
   D.; VITER, R.; ZINGG, A. 2005: Structures of virgin and managed beech forests in Uholka
   (Ukraine) and Sihlwald (Switzerland): a comparative study. For. Snow Landsc. Res. 79, 1/2:
CROSSLEY, D.I., 1976: Growth response of spruce and fir to release from suppression. For. Chron.
   52: 189–193.
DIACI, J. (ed), 1999: Virgin forests and forest reserves in central and east European countries:
   History, present status and future development. Biotechnical Faculty, Department of Forestry
   and Renewable Forest Resources, Ljubljana, Proceedings of the invited lecturers’ reports
   presented at the COST E4 Management Committee and Working Groups meeting in
   Ljubljana, Slovenia, 25–28 April 1998. 171 pp.
Eidg. Forschungsanstalt WSL; Bundesamt für Umwelt, Wald und Landschaft (BUWAL) (eds),
   2001: Lothar. Der Orkan 1999. Ereignisanalyse. Birmensdorf, Bern, Eidg. Forschungsanstalt
   WSL, Bundesamt für Umwelt, Wald und Landschaft BUWAL. 365 pp.
ENGLER, A., 1900: Wirtschaftsprincipien für die natürliche Verjüngung der Waldungen mit beson-
   derer Berücksichtigung der verschiedenen Standortsverhältnisse der Schweiz. Schweiz. Z.
   Forstwes. 51: 264–274; 300–310.
ENGLER, A., 1904: Der Urwald bei Schattawa im Böhmerwald. Schweiz. Z. Forstwes. 55: 173–182.
FABIJANOWSKI, J., 1978: Geschichte des Bialowieza-Urwaldes und die daraus sich ergebenden
   waldbaulichen Folgerungen. Allg. Forst Z. Waldwirtsch. Umweltvorsorge 33: 709–711.
   Disturbances and structural development of natural forest ecosystems with silvicultural impli-
   cations, using Douglas-fir forests as an example. For. Ecol. Manage. 155: 399–423.
FRÖHLICH, J., 1947: Über Vorkommen, Zusammensetzung und Aufbau der südosteuropäischen
   Laubmisch-Urwälder. Schweiz. Z. Forstwes. 98: 165–169.
FRÖHLICH, J., 1954: Urwaldpraxis. 40jährige Erfahrungen und Lehren. Berlin, Radebeul. 199 pp.
GAYER, K., 1898: Der Waldbau. Vierte, verbesserte Auflage, Berlin, Parey. 626 pp.
GEHRHARDT, E., 1923: Über Urwaldungen in den Karpathen. Silva. 361 pp.
GÖPPERT, H.R., 1868: Skizzen zur Kenntnis der Urwälder Schlesiens und Böhmens. Dresden,
   Blochmann & Sohn.
GROVE, S.J., 2002: Saproxylic insects and the sustainable management of forests. Ann. Rev. Ecol.
   Syst. 33: 1–23.
GURNAUD, A., 1886: La sylviculture française et la méthode de contrôle. Besançon, Jacquin.
   121 pp.
   K.J.; CUMMINS, K.W., 1986: Ecology of coarse woody debris in temperate ecosystems. Adv.
   ecol. res. 15: 133–302.
HILLGARTER, F.W., 1978: Waldbauliche Lehren aus Untersuchungen im Fichtenwald von Scatlé.
   Allg. Forstz. 33: 698–699.
KÖHL, M.; SCOTT, C.T.; ZINGG, A., 1995: Evaluation of permanent sample surveys for growth and
   yield studies: a Swiss example. For. Ecol. Manage. 71: 187–194.
KOPEZKY, R., 1895: Über Bestandesmassenaufnahmen im Urwalde. Ges. Forstwes. 12:
KORPEL’, S., 1995: Die Urwälder der Westkarpaten. Stuttgart, Jena, New York, Gustav Fischer.
   310 pp.
KORPEL’, S., 1997: Totholz in Naturwäldern und Konsequenzen für Naturschutz und Forst-
   wirtschaft. Forst Holz 52: 619–624.
LEIBUNDGUT, H., 1959: Über Zweck und Methodik der Struktur- und Zuwachsanalyse von Ur-
   wäldern. Schweiz. Z. Forstwes. 110: 111–124.
For. Snow Landsc. Res. 79, 1/2 (2005)                                                          31

LEIBUNDGUT, H., 1982: Europäische Urwälder der Bergstufe: dargestellt für Forstleute,
   Naturwissenschafter und Freunde des Waldes. Bern, Stuttgart, Haupt. 308 pp.
LERTZMAN, K.P.; SUTHERLAND, G.D.; INSELBERG, A.; SAUNDERS, S.C., 1996: Canopy gaps and the
   landscape mosaic in a coastal temperate rainforest. Ecology 77: 1254–1270.
LUND, H.G., 2002: Coming to terms with politicians and definitions. IUFRO Occasional Paper 14:
MAI, W., 1998: Naturverjüngung auf Moderholz. Aktueller Stand des Wissens als Ergebnis einer
   Literaturstudie. Freising, Bayerische Landesanstalt für Wald und Forstwirtschaft.
MAYER, H.; SCHENKER, S.; ZUKRIGL, K., 1972: Der Urwaldrest Neuwald beim Lahnsattel. Ges. Forstwes. 89: 147–190.
MAYER, H.; ZUKRIGL, K.; SCHREMPF, W.; SCHLAGER, G. (eds), 1989: Urwaldreste, Naturwald-
   reservate und schützenswerte Naturwälder in Österreich, 2. Aufl. Wien, Institut für Waldbau,
   Bodenkundliche Universität.
MEYER, P.; TABAKU, V.; VON LÜPKE, B., 2002: Elemente naturnaher Buchenwirtschaft –
   Ableitungen aus albanischen Urwäldern. In: Anonymus (ed): Tagungsbericht der Sektion
   Waldbau im Deutscher Verband Forstlicher Forschungsanstalten vom 12.09.–14.09.2001 in
   Ludwigslust. Ministerium für Ernährung, Landwirtschaft, Forsten und Fischerei, Schwerin.
MEYER, P.; BÜCKING, W.; SCHMIDT, S.; SCHULTE, U.; WILLIG, J., 2004: Stand und Perspektiven der
   Untersuchung von Naturwald-Vergleichsflächen. Forstarchiv 75: 167–179.
MLINSEK, D., 1967: Verjüngung und Entwicklung der Dickungen im Tannen-Buchen-Urwald
   “Rog” (Slowenien: Proceedings, IUFRO World Congress, XIV/4: 436–442.
MLINSEK, D., 1978: Brauchen wir Urwald? Allg. Forstz. 33: 684–686.
MLINSEK, D., 1993: Research in virgin forests – for forestry and society. In: BROEKMEYER,
   M.A.E.; VOS, W.; KOOP, H. (eds) European forest reserves. Proceedings of the European
   Forest Reserves Workshop, 6–8 May 1992, The Netherlands. Wageningen, Pudoc Scientific
MOREILLON, M., 1910: Du rajeunissement de l’épicéa dans les forêts des régions élevées, et plus
   spécialement dans le haut Jura vaudois. Travail de concours primé par la Société des Forestiers
   Suisses en 1909. Bern, Büchler & Co.
   LJUNGBERG, H.; MIKUSINSKI, G.; RANIUS, T., 2002: Densities of large living and dead trees in
   old-growth temperate and boreal forests. For. Ecol. Manage. 161: 189–204.
OTTO, H.-J., 1995: Zielorientierter Waldbau und Schutz sukzessionaler Prozesse. Forst Holz 50:
   2000: Final report summary: mission, goal, outputs, linkages, recommendations and partners.
   In: European Commission (ed): EUR 19550 – COST Action E4 – Forest reserves research
   network. Luxembourg, Office for Official Publication of the European Communities. 9–38.
PETERKEN, G.F., 1999: Applying natural forestry concepts in an intensively managed landscape.
   Glob. Ecol. Biogeogr. 8: 321–328.
PILLICHODY, A., 1910: Du rajeunissement de l’épicéa dans les forêts des régions élevées. Travail
   de concours primé par la Société des Forestiers Suisses en 1909. Bern, Büchler & Co.
PINTARIÇ, K., 1959: Urwald in Jugoslawien. Schweiz. Z. Forstwes. 110: 163–168.
PINTARIÇ, K., 1999: Forestry and forest reserves in Bosnia and Herzegovina. In: Diaci, J. (ed)
   Virgin forests and forest reserves in central and east European countries: History, present
   status and future development. Biotechnical Faculty, Department of Forestry and Renewable
   Forest Resources, Ljubljana, Proceedings of the invited lecturers’ reports presented at the
   COST E4 Management Committee and Working Groups meeting in Ljubljana, Slovenia,
   25–28 April 1998. 1–15.
Projektgruppe Naturwaldreservate im Arbeitskreis Standortskartierung der Arbeitsgemeinschaft
   Forsteinrichtung., 1993: Empfehlungen für die Einrichtung und Betreuung von Naturwald-
   reservaten in Deutschland. Forstarchiv 64: 122–129.
PRUSA, E., 1985: Die böhmischen und mährischen Urwälder – ihre Struktur und Ökologie. Praha,
   Academia Verlag der Tschechoslowakischen Akademie der Wissenschaften. 577 pp.
32                                                                                  Peter Brang

RADEMACHER, C.; NEUERT, C.; GRUNDMANN, V.; WISSEL, C.; GRIMM, V., 2001: Was charakte-
   risiert Buchenurwälder? Untersuchungen der Altersstruktur des Kronendachs und der räum-
   lichen Verteilung der Baumriesen in einem Modellwald mit Hilfe des Simulationsmodells
   Before. Forstwiss. 120: 288–302.
REMMERT, H., 1991: The mosaic-cycle concept of ecosystems. Ecological studies 85. New York,
   Springer. 168 pp.
RUBNER, K., 1920: Die waldbaulichen Folgerungen des Urwaldes. Naturwiss. Z. Forst- Land-
   wirtschaft 18: 201–214.
SANIGA, M.; SCHÜTZ, J.-P., 2002: Relation of dead wood course within the development cycle of
   selected virgin forests in Slovakia. J. For. Sci. 48: 513–528.
SCHIEGG, K., 2000: Effects of dead wood volume and connectivity on saproxylic insect species
   diversity. Ecoscience 7: 290–298.
SCHUCK, A.; PARVIAINEN, J.; BÜCKING, W., 1994: A review of approaches to forestry research on
   structure, succession and biodiversity of undisturbed and semi-natural forests and woodlands
   in Europe. Eur. For. Inst. Working Pap. 3: 64 pp.
SCHÜTZ, J.-P., 1969: Etude des phénomènes de la croissance en hauteur et en diamètre du sapin
   (Abies alba Mill.) et de l’épicéa (Picea abies Karst.) dans deux peuplements jardinés et une
   forêt vierge. Dissertation, ETH Zürich. Beih. Z. Schweiz. Forstver. 44.
SCHÜTZ, J.-P., 1994: Der naturnahe Waldbau Leibundguts: Befreiung von Schemen und
   Berücksichtigung der Naturgesetze. Schweiz. Z. Forstwes. 145: 449–462.
SCHÜTZ, J.-P., 1999: Neue Waldbehandlungskonzepte in Zeiten der Mittelknappheit: Prinzipien
   einer biologisch rationellen und kostenbewussten Waldpflege. Schweiz. Z. Forstwes. 150:
SCHÜTZ, J.-P., 2004: Opportunistic methods of controlling vegetation, inspired by natural plant
   succession dynamics with special reference to natural outmixing tendencies in a gap regen-
   eration. Ann. For. Sci. 61: 149–156.
SEIDEL, A., 1848: Beiträge zur Kenntnis des Urwaldes. Tharandter Forstliches Jahrbuch. 158 pp.
SEIDEL, K.W., 1977: Suppressed grand fir and Shasta red fir respond well to release. Res. Note
   USDA For. Serv. Pacific Northwest For. Range Exper. Station 288.
SPIES, T.A.; TURNER, M.G., 1999: Dynamic forest mosaics. In: HUNTER, M.L. (ed): Maintaining
   biodiversity in forest ecosystems. Cambridge, Cambridge University Press. 95–160.
TOMIA»OJC, L.; WESO»OWSKI, T., 2004: Diversity of Bialowieza Forest avifauna in space and time.
   J. Ornithol. 145: 81–92.
VÉZINA, P.-E., 1959: Contribution à l’étude des forêt vierges comme base pour le développement
   d’une sylviculture plus près de la nature. Essai d’application aux forêts résineuses de l’Est
   canadien. Schweiz. Z. Forstwes. 110: 135–149.
VON LÜPKE, B., 2004: Risikominderung durch Mischwälder und naturnaher Waldbau: ein Span-
   nungsfeld. Forstarchiv 57: 43–50.
WESO»OWSKI, T.; TOMIA»OJC, L., 1995: Ornithologische Untersuchungen im Urwald von Bia-
   lowieza – eine Übersicht. Ornithol. Beob. 92: 111–146.
WESSELY, J., 1853: Die österreichischen Alpenländer und ihre Forste. Wien, Braumüller.
WOHLGEMUTH, T.; BÜRGI, M., SCHEIDEGGER, C.; SCHÜTZ, M., 2002: Dominance reduction of
   species through disturbance – a proposed management principle for central European forests.
   For. Ecol. Manage. 166: 1–15.
ZLATNIK, A., 1938: Prozkum prirodzenych lesu na Podkarpatské Rusi. Dil prvni: Vegetace a
   stanovistè reservace Stuzica, Javornik a Pop Ivan. In: Sbornik Vyzk. Ust. Zemèdèl. Praha,
   244 pp.
ZUKRIGL, K., 1990: Naturwaldreservate in Österreich. Stand und neu aufgenommene Flächen.
   UBA-Monographie 21: 232 pp.

                                                                         Accepted May 23, 2005
You can also read
Next slide ... Cancel