Davos, Switzerland

Preface Abstracts Addresses

Swiss Federal Institute for Forest, Snow and Landscape Research

Binding Foundation

Swiss National Science Foundation

Top

 The scientific workshop 'Structure of Mountain Forests - Assessment, Impacts, Management , Modelling' provides a platform for the presentation and discussion of the state of knowledge on the significance of mountain forest structure for natural hazards, biodiversity, nature conservation, stand dynamics, silviculture, forest management and on the methodology of structure assessment.

The horizontal and vertical structure are key factors of the multiple functions of mountain forests. The typical structural features of mountain forests vary greatly from those of lowland forests. Typical high altitude forests of temperate zones contain relatively few species and consist of open stands with gaps. The trees are arranged in clusters, have short stems and long crowns.

The stand structure is one of the important determinants for prevention of natural hazards, such as avalanches, rock fall, erosion, debris flow, land slides or floods. Stand structure is also closely correlated with biodiversity and habitat richness, thus significant for nature conservation and wildlife. Both the stand structure and the stand stability are dynamic, and stands often do not regenerate easily. Silvicultural systems must be adapted to the specific conditions and needs in high altitude. The economic and the technical situation make forest management much more difficult in mountainous areas. Assessing, analysing and modelling stand structures are therefore prerequisites for understanding and managing mountain forests with multiple functions.

The workshop is oriented towards scientists willing to contribute actively to the workshop and interested in mountain forest structures, protective function of mountain forests, stand stability, mountain forest dynamics, nature conservation in mountainous regions, mountain forest silviculture and management, assessment and modelling of stand structures.

The workshop focuses on the following topics:

1. Structure and stand dynamics: including stand regeneration, longterm stability
2. Structure and protection against natural hazards: avalanches, rock fall, erosion, land slides, floods, etc.
3. Structure and biodiversity: nature conservation, habitat for plants, animals and humans
4. Structure and management: silviculture, minimal and optimal tending measures
5. Structure analysis and modelling: methods, case studies, applications

 

Walter Schönenberger, Chairman of the workshop

 

Organising Committee

 

Workshop chairman: Walter Schoenenberger

Co-chairman: Werner Frey

Advisory council: Gregorio Montero, Pietro Piussi

Local arrangements: Werner Frey

Field trips: Werner Frey, Peter Brang, Peter Bebi, Walter Schoenenberger, Veronika Stoeckli, Ueli Wasem

Technical programme: Ueli Wasem, Peter Bebi, Peter Brang, Erica Zimmermann

On Site Secretariat: Erica Zimmermann, Christine Berni

Scientific committee: The task of the scientific committee is to review the final versions of voluntary papers submitted for publication. Each member of the scientific committee takes the responsibility for about 5 papers, each to be reviewed by two reviewers of his choice. Review work is due in October 1999. Members of the scientific committee are:

Walter Schoenenberger (chairman, guest-editor), Peter Brang, Switzerland, Harald Bugmann, USA, Christophe Chauvin, France, Alan Mitchell, Canada, Gregorio Montero, Spain, Luigi Portoghesi, Italy, Erik Valinger, Sweden

 

 

 

 Abstracts of PAPERS / POSTERS

Top 

 

Monitoring silviculture in mountain forests

Chauvin Christophe and Mermin Eric

A traditional approach to monitor silviculture in irregular stands is the "curves method", comparing a general diameter inventory, at the scale of a parcel, to some norm (Liocourt) or simply to previous inventories on the same parcel (Gurnaud, Biolley). Although seeming to provide a mathematical security, this approach bears several disadvantages: the same curve can represent very different parcels, concerning the spatial repartition of trees and stands, or their vitality, their dynamics, and finally their future. There are many scales of irregularity in the structure of mountain forests: tree, bunch, site fertility, they all require a closer description.

Then, another approach has been developed from a naturalistic background: mapping silvogenetic phases (Meyer, Leibundgut, Oldeman). But this holistic approach, widely intuitive, meets problems of reproductibility, for the identification of the phases and their limits in the field.

An intermediate, statistical approach has been developed in the French Alps, around the concept and structure typology, for spruce stands: the stands are measured and described by point measurements from a systematic grid. The stand type is determined in the field from easily assessed or measured criteria: the coverage of respective relative-height classes (strata), and the relascope basal-area. These criteria proved relevant by a general data analysis on sample parcels, as closely correlated to other identified criteria for the stability and the dynamics of the stands.

This typology of structures is now widely in use for forest management (mapping) and silviculture (guides), being a common and objective language between various specialists of different regions. It is also the basis for further scientific developments, such as modelling of dynamics and studies of interactions between forests and natural hazards: permanent reference plots have been installed all over the northern Alps, for a better characterisation of the internal structure of the types, for the monitoring of their dynamics and for training purposes.

 

Use of a trajectory calculation model and a GIS for better taking into account the forest role for the protection of roads against rockfall

Berger Fred and Mermin Eric

Forest has played a brake role as regards rockfall. This role is obvious but difficult to quantify because this phenomenon is very random: we do not know to model departures, sizes and rebounds of rocks, and shocks against the trees. The structures of protection are necessary and essential, but this role allows to decrease their dimension and their cost.

This study uses a stochastic method of simulation already used in an expertise not taking account of the forest. It is based on a model deterministic. The model simulates one hundred falls of blocks for each zone and varies by chance the parameters at each rebound inside a certain range, representing the irregularity of the ground.

This model took into account the forest stands by addition of a friction coefficient. To improve this integration of the forest role, essential mechanisms (rupture of wood and trees) have been studied by means of experimental field devices. We better described forest stands (cover, structure, basal area) by using the typology of structure of coniferous forest stands (50 m steps) and we have assessed the probabilities of shocks according the structure of stands. The name of this model is "Azimi-Cattiau" model.

The simulations using this "stand model" have shown the essential effect of the forest about the stopping distances and the heights of the rebounds. For example, if the forest above the Val d' Isère road disappeared, we assess that 71% of rocks will reach the road instead of 18%.

In a next step, we will create a determinist model taking into account the behaviour of trees with respect to rockfall. This model will be tested in some forest stands representative of the structural typology. Each tree of these stands will be mapped and we will simulate rockfall in these stands. This model will allow to validate statistically the "Azimi-Cattiau model" and to improve it. It will assess the most effective and durable types of structure of forest stands.

Mapping works were required to complete this study. Further than the help it provides to the scientific research domain, this software may be taken as the backbone of a real management assistance tool and therefore help preventing rockfall on infrastructure such as roads.

 

Structures of mountain forests as a basis for the assessment of different forest functions

Bebi Peter

Stand structure was studied in 200 systematically selected plots in subalpine forests, and a typology of the subalpine, spruce-dominated forests in Switzerland was made through the statistical evaluation of these plots. A key for the stereoscopical assessment of 21 forest structure types was set up and a forest area of 4 km² near Davos was mapped and terrestrially verified with the aid of that key and aerial photographs (IR, 1: 9000).

The same area was also mapped using aerial photographs from 1930. A geographical information system (GIS) was used to calculate the transition-probabilities between the different types in the period from 1930-1997. The area was divided into different site-strata in order to examine and model the dynamic of forest under different site conditions (e.g. avalanche activity, elevation, cattle grazing, exploitation).

In order to increase the objectivity of the forest-structure maps, attempts were made to automate the forest-structure mapping-process through image processing. Promising results have been obtained to identify single trees and to automatically assess the most important structure types.

Two GIS-based examples of applications in the study area show that the structure-mapping methods can be a very useful and practical tool in combination with a 25 m digital terrain model for the assessment of different forest functions. In the first example, the risk of avalanche starting-zones below the timberline is modelled and mapped on the basis of the statistical evaluation of data of stand structure and terrain of 150 real avalanche starting-zones. In the second application, habitat-preference maps of Capercaille (Tetrao urogallus) and Black Grouse (Tetrao tetrix) were modelled on the basis of published expert knowledge.

 

Models to assess the risk of damage from snow and wind

Valinger Erik and Fridman Jonas

Each year, damage to forests caused by snow and wind causes high economic losses. In Sweden, approximately 4 million m³ are damaged annually by snow and wind, roughly corresponding to a value of 150 million USD, and in Europe, the damage amounts to hundreds of millions of USD each year. To help to reduce these losses, tools for risk assessment within forest management have been developed.

Predictions were developed of the risk of damage from snow and wind to Scots pine (Pinus sylvestris L.), Norway spruce (Picea abies (L.) Karst.) and Birch (Betula spp. L.) plots using tree, stand, and site characteristics. The data was obtained from 6756 permanent sample plots within the Swedish National Forest Inventory, which were inventoried twice at five-year intervals between 1983 and 1992. Input data for model development used measurements from the first inventory of tree characteristics for the largest sample tree, stand, and site data, and records of snow and wind damage from the second inventory.

Models were developed for three different regions for pine and spruce dominated sites, while models for the whole country were developed for birch sites. The estimated proportion of damaged plots was in general highly overestimated (31.7–56.2%), compared with the observed proportion of 3.4-11.9%. The models for Norway spruce comprising tree, stand, and site data show the best predictability of damaged plots, with 60.6 to 67.6% of plots correctly classified.

It is concluded that the models developed can be used to detect sites with high probability of damage from snow and wind, and thus be used as tools to reduce future damage and costs in practical forestry.

 

Survey and interpretation of stand structure in the 2nd Swiss National Forest Inventory (NFI)

Brändli Urs-Beat

The Swiss National Forest Inventory (NFI) aims at providing information about the state and development of the Swiss forest. Special emphasis is put on the ability of forest stands to fulfil the different functions in a sustainable way. Therefore, detailed data about stand structures are needed.

The NFI database of the second inventory (1993-1995) and the methods to collect data about stand structure are discussed. Results for the lowlands as well as for the mountainous regions will be shown in order to characterise the Swiss mountain forests. The NFI model, which assesses the capacity of forests to protect against avalanches, is presented. The results show that today, two thirds of these protection forests are in good or very good condition. On the other hand, there is a lack of regeneration in protection forests and stand stability is more critical than in other forests. Methodological aspects concerning the monitoring of stand structure in a third national inventory are discussed.

 

Assessment, analysis and interpretation of regeneration in the 2nd Swiss National Forest Inventory (NFI)

Duc Philippe and Schwyzer Andreas

The forest structure is the key factor for the fulfilment of the multiple functions of mountain forests. The condition of a permanently stable structure is a sustainable regeneration of forests. That is why the assessment, analysis and interpretation of the actual regeneration is important. In this context, we are interested in the following questions:

1. What means sustainable regeneration of mountain forests?
0. What about the distribution and development of regeneration in mountain forests?
0. What are our conclusions for the assessment, analysis and interpretation of regeneration?

a) Starting from forest structure and from the aim of forest management in mountain forests, two possibilities how to control the sustainability of forest regeneration are shown:
- with a stem-oriented method based on the model of selection forest
- with a surface-oriented method based on knowledge of group-structured selection forests.

b) The survey of regeneration by the NFI allows indications about canopy and stem-number density on sample plots. Because of changes of the survey method, no indications can be given about the temporal development on identical plots since the first NFI. On the other hand, we can examine the correlation between factors describing station or stand and canopy or stem-number density on the base of survey data from NFI. Our results are completed with results of research on the spatio-temporal development of regeneration in the subalpine altitudinal zone.

c) Based on knowledge about distribution and development of regeneration, we first draw some conclusions for the assessment of regeneration as a part of forest structure. We secondly analyse if the present data on forest regeneration can be used for the control of sustainable regeneration. And we thirdly give an outlook of future research activities to improve the control of sustainable regeneration.

The main aim of this study is to analyse the assessment method of stand stability in the Swiss National Forest Inventory (NFI). The NFI collects various tree, stand and site variables on sample plots. The stand stability is an important criterion to evaluate forest condition. It is assessed by expert judgement. The NFI defines stand stability as the probability of considerable damages to the stand or even its breakdown occurring within a fixed interval of time. Therefore, the assessment of stand stability can be considered a risk evaluation. It follows a specific protocol: Of the assessed site factors, nine are weighed under the aspect of potential hazards to the stand («hazard profile») and eleven stand parameters describe its resistance potential («resistance profile»). The variable «stand stability» is derived from these two profiles. In spite of this methodical proceeding, the stand stability as assessed by the NFI still is an expert opinion and thus is quite sensitive to the assessor’s subjectivity. This hampers the analysis and makes the interpretation of changes difficult.

For this reason, we examine how the assessment of stand stability reflects reality by searching the relation between stability at the beginning of an observed period (first NFI) and the following evolution of forest condition parameters (between the first and the second NFI).

Furthermore, we explore if stand stability can be derived from more objective and reproducible measurements and assessments. By searching for those factors that have the greatest influence on stand stability, we aim at understanding to what extend this variable can be regarded as objective . Among a great number of assessment parameters (assessor, daytime and season of survey, etc.), site factors (exposition, altitude, etc.), stand and tree characters, we identify those variables which provide the best explanation for stand stability.

The structure and texture of young stands in the mountainous region of Arasbaran were studied and compared with the old stands. Arasbaran is located in East-Azarbeidjan (North-West Iran). It is a transitional zone with sub-humid climate between the humid Caspian zone in the North and the sub-humid Zagros zone in West Iran.

The forests in this region are protected forests and mainly formed with oak (Quercus macranthera Fisch & Meyer, Q. komarovii Camus), hornbeam (Carpinus betulus L., C. orientalis Mill.), maple (Acer campstre L., A. monspessulanum L.) elm (Ulmus glabra Huds.), wild cherry (Cerasus avium (L.) Moench), ash (Fraxinus excelsior L.) and juniper (Juniperus communis L., J. foetidissima Willd, J. excelsa M.B.). The main forest associations are pure oak (Quercetum), pure hornbeam (Carpinetum), mixed oak-hornbeam (Querco-Carpinetum), mixed broad-leaved and needle-leaved species.

In order to study the structure of young stands, the experimental forest "Sutanchay" with a total area of 2400 ha was selected. 139 sample plots, each 100 m², were set up and all species with less than 7.5 cm dbh were studied.

The results showed that the most frequent regeneration system was coppice with standards. High stands was also recognised in some parts of the studied area as well as coppice forests in other parts. Only 28.7% of regeneration was produced from seed. In coppice form, each stump produced 6 stems in average with a mean height of 3.7 m. The ability in producing coppice sprouts were highest among oak and hornbeam individuals.

The portion of species within the young stands were: hornbeam (31.7%), oak (29.3%), maple (8.2%), wild cherry (1.4%) and other species (29.4%). The studied stands mainly showed a two-storeyd structure, with standards in the upper layer and coppices in the under storey. Juniper stands show mountainous collective structure, trees are arranged in clusters and the stands do not regenerate easily.

Mountain forests in the European Alps and in the Rocky Mountains share many similarities. For example, stands are typically dominated by one to a few tree species, growth is strongly constrained by climatic factors, regeneration occurs irregularly, and trees tend to occur in clusters. However, forests in the two regions are subject to strongly different climatic regimes, relating mainly to the seasonality and variability of temperature and precipitation. Additionally, the frequency of natural disturbances like wind, insect infestations, and fire is much higher in the Rocky Mountains as compared to most of the European Alps.

What brings about the similar appearance of these forests in spite of the large differences in the environmental forcing factors? There is an intimate linkage between the structure and dynamics of mountain forests, with multiple biotic and abiotic influences and feedbacks, and the current forest structure is the consequence of processes operating on time scales of decades to centuries. Thus, models of these processes are required to synthesise the existing knowledge, and to make long-term assessments of both past and future forest dynamics.

In this study, I present a systematic analysis of the abiotic and biotic factors that determine the long-term development of mountain forests along environmental gradients in the Swiss Alps and in the Colorado Front Range, based on a quantitative model of tree population dynamics that includes disturbances at the scale of the individual tree (stress-induced mortality) and the stand (fire). The relative importance of various environmental factors for determining stand structure in the two regions is analysed, and the feedback between the existing stand and its regeneration dynamics are assessed.

The model is then used to provide projections of the possible future structure and composition of mountain forests in the two study areas under several scenarios of climatic change as anticipated for the end of the 21^st century.

There is enormous pressure to come up with answers to questions asked by politicians and the public concerning the development of our environment and the potential risks our society might be confronted with. As the environmental and social context itself is rapidly changing, it is unknown what uses of a forest will appear in the future. The changing social and ecological context under which mountain forestry operates is therefore calling for an appropriate management mode to deal with uncertainties.

There is a need to act, monitor the results, learn from the past, adapt to new conditions through planning and to accept a philosophy of managing an ecosystem with the purpose of reducing potential future socio-ecological and environmental risk by understanding potential problems before they arise.

In order to evaluate the likelihood that adverse ecological effects may occur as a result of exposure to one or more stressors, long-term monitoring data, information, assumptions and uncertainties need to be systematically evaluated and analysed in order to help understand and predict the relationships between stressors and ecological effects in a way that is useful for environmental decision-making. This paper will discuss an approach and strategy for performing ecological risk assessments in mountain forests.

Mountain forests present a spatial, dynamic and functional heterogeneity which makes their description and the knowledge of the ecological processes difficult. The recognition of the basic units of the vegetation organisation is necessary to describe vegetation structure and diversity and to understand its dynamics and its functioning (Watt 1947, Pickett and White 1985, Remmert 1991).

A description of the spatial and temporal organisation of the vegetation was made in mountain spruce forest in order to understand its functioning. This study was carried out in accordance with integrated synusial phytosociology which is notably based on the distinction of different spatio-temporal levels closely imbricated (Gillet et Gallandat 1996). Only the first two integration levels were taken into account in this work: (1) the basic units named synusiae (elementary vegetation communities) and (2) the complexes of synusiae called phytocoenosis. A typology of this communities and a research of their ecological determinism were performed. For each vegetation unit some richness and diversity indices were calculated to analyse the spatio-temporal distribution of the vegetation diversity.

A description of the forest stand structure was also carried out in order to compare this ecological perception mode with the one of forest managers. The particular aim of this comparison was to establish a link between different objects defined by different methods: structure units and vegetation units. The structure description was realised in 50 m square plots. It was based on the relative cover of vertical layers and the presence of horizontal discontinuities of forest cover (Chauvin, Renaud, Rupé, Leclerc, 1994). This comparison allows us to determine and compare diversities associated with each type of structure.

Green alder (Alnus viridis D.C.) is a very common shrub species in the Alps. It is quite dynamic on relatively wet and drained soils at the subalpine and mountain belt, especially on abandoned meadows and pastures. Moreover, this taxon is considered as a keystone species due to its fundamental impact on environmental variables, making them change drastically and become more homogeneous.

The principle of this study is to draw up a balance sheet of the biodiversity changes according to Alnus evolution at different spatial scales, concerning groups directly linked to hazel grouse (Tetrao tetrix L.). We compared three succession stages of the Alnus dynamic trajectory, namely an abandoned pasture, an alder mosaic, and a thick alder stand. The results show that richness in vegetation and arthropods is strongly affected by shrub expansion, leading to the drop of Tetrao tetrix, a threatened alpine bird, important in terms of conservation in Western Europe. Actually, arthropods are essential for the young birds from birth to a couple of weeks, and the lack of vegetation under Alnus canopy exposes the hazel grouses to predation. However, these results are pondered according to the area of Alnus stands, and to which extent it can be considered as homogeneous or heterogeneous for Tetrao tetrix.

As a consequence, these data provide new support to debate about the costs and benefits of land management relating to green alder. Nevertheless, its effects on snowslides and water retention have to be taken in account as well, in order to suggest an adapted management for each case.

In some French mountain areas, the land use changes started around 1850 have resulted in a complete abandonment of any agro-pastoral activity by 1950. Thus, the landscape suffered important modifications of structure and functioning. Tree invasion does not take place homogeneously in space nor time. After interpretation of aerial photographs and mapping of recolonisation areas at and above the forest boundaries, we analysed the spatial pattern and age structure of sample plots established in the regeneration zones above 2000 m a.s.l. Two main forms of forest boundary dynamics have been noted:

(1) European larch (Larix decidua Miller) forms very dense colonisation patches alongside the forest margins, nearby the ancient chalets. The invasion seems limited to the meadows at the lower part of forested slopes and related to colluvial soil conditions. It took place immediately after abandonment of the pastures fifty years ago. Now stone pine is regenerating in these young growth larch woods.

(2) Stone pine (Pinus cembra L.) regenerates massively at tree line, forming dense seedling patches, mostly on convex relief forms.

The recolonisation pathways do not seem to be influenced by the adjacent forest stands but by the local edaphic and microclimatic conditions. Moreover, the potentialities of boundary movements depend on the spatial characteristics of the vegetation mosaic. Thus, understanding of the landscape structure is very important for the study and prediction of timberline fluctuations.

In some mountain areas, exploiting forest is of low profitability and foresters tend to reduce management to a minimum. They are even tempted to give up the management of some forests which are already dangerously unstable because of high capitalisation, old-growth stands and windthrown woods. These forests cannot carry out their functions of wood production, protection against natural hazards, etc, anymore.

In the context of mountain forest, more than anywhere else, regeneration success is a crucial step for the forest durability, by conditioning its renewal and its future structure. Thus, a good knowledge of natural processes influencing regeneration is essential. This is especially true for some coniferous altitude forests (fir and spruce), the renewal of which is known as being slow, sparse and uncertain. To explain these phenomena, numerous biotic or abiotic reasons have been put forward, at various spatial and temporal scales. Depending on the occurrence of different field characteristics and possible interactions with each other, it can ensure natural recruitment or not.

In the northern French Alps, a study was started in 1997, in two different mountain Norway spruce forests, to identify some microhabitats favouring natural seedling settlement. It focuses on the correlation existing between the renewal success and some field parameters such as:

• forest stand structure,
• support of regeneration (organic and mineral soil, decayed wood, …),
• light (direct and diffuse) availability at the ground surface,
• snow-cover distribution (amount and persistence),
• vegetation types,
• humus types (morphological, physical and chemical characters).

The aim of this study is :

• to predict favourable sites for natural seedling settlement under mountain spruce forest,
• to estimate (probabilities) its occurrence,
• to suggest management rules in order to allow, favour or maintain the renewal of these forests and consequently to secure their stability.

Our final paper will present the latest results of these investigations.

Long-term series are often not available to determine the natural dynamics and stability of forests. In our study, we used current quantitative and qualitative information on tree conditions, stem decay and root rot. Although the period of observation was only three years, it was possible to estimate spatial relationships between tree conditions and past and future stand dynamics. The unmanaged pure mountain pine forest in the Swiss National Park was established after clear-cutting and subsequent grazing. As the site had been unmanaged and undisturbed since the establishment of the National Park in 1913, it was ideal for this analysis. After the establishment of a 2 ha plot in 1995, the first follow-up measurement in 1998 gave an annual mortality rate for standing trees of 1.3% and annual fall-down rates of 0.7%. Assessments of crown condition and subsequent mortality rates showed increased mortality rates with increased defoliation. In 1998, 20% of all trees of at least 12 cm dbh were standing dead and 5% were lying dead (however, not all lying dead trees were identified). Fall-down rates of standing dead trees were 3% per year. Given the current mortality, fall-down and expected ingrowth rates, the percentage of standing and lying dead trees with a stem diameter of at least 12 cm is increasing. An analysis of the diameter distribution showed that the trees that had died during the last three years were significantly smaller than the surviving trees, but slightly larger than the standing trees that had died previously. Of a sample of 34 trees that had died during the last 2 years, 91% had root rot (armillaria and/or heterobasidium). Distances to nearest neighbours were significantly smaller for dead trees than for living trees. However, dead trees were more likely to be located next to dead trees, while living trees were more likely to stand next to other living trees. This may be an indication that the usual competition-driven mortality is currently changing to a pathogen-driven mortality (although predisposition due to competition may still play a major role). This was confirmed by comparison with the results from long-term study plots and dendrochronological studies close to our study sites. Our study and other studies showed that with increased tree age, mortality and fall-down rates on site increased, leading to decreasing stand density that cannot be explained by competition or tree age alone. We expect that for our study plots this tendency will continue, leading eventually to a succession of the current overstory layer. Future studies are planned to quantify the effects of both armillaria and heterobasidium root rot in this process. Our study suggests that pathogens play an important role in the stability of natural mountain pine ecosystems.

The paper discusses the diversity and distribution pattern of tree species of Sitapahar Forest Reserve at Chittagong Hill Tracts (South) Forest Division, Bangladesh. Recently, these hilly forests were subject to severe degradation due to illegal felling, encroachment and over-exploitation. The present Sitapahar reserve is the remnant of the biologically rich hill forests of the country. Five permanent sample plots of the size of 100 m x 100 m (1 ha) were set up for tree vegetation sampling. A total of 992 stems (>10 cm dbh) of 54 species were recorded. The family Euphorbiaceae and Moraceae dominate, containing 5 species each, followed by Anacardiaceae with 4 species. The average stem number per hectare was 198, of which Dipterocarpus turbinatus represents the highest percentage (16.98%) followed by Quercus spp. (11.69%), Aphanamixis polystachya (9.88%), Bischofia javanica (7.46%), Artocarpus chaplasha (4.84%) and Myristica linifolia (4.84%). The percentage distribution of each individual species in different diameter classes showed the highest percentage (21.97%) of individuals in the 10-20 cm dbh range, and the number of species were highest (23.29%) in the 20-30 cm dbh range. Similarly, the percentage distribution of each individual species in different height classes showed the highest percentage of individuals (26.0%) in the 5-10 m height range followed by the 15-20 m height range (22.4%), and the number of species were highest (43 species) in the 15-20 m height range. The percentage distribution of individual stems of different species in different merchantable height classes showed that the highest percentage of individuals (29.4%) belongs to the 3-8 m height range followed by 23.3% in the 8-13 m height range and 21.7% stems in the 13-18 m height range. The Importance Value Index (IVI) was highest (63.26%) for D. turbinatus followed by Quercus spp. (28.98%), Bischofia javanica (24.05%), Aphanamixis polystachya (23.72%) and Artocarpus chaplasha (22.68%).

Though the hill forests in Bangladesh are managed by the clear-felling system followed by artificial regeneration, for Sitapahar forest reserve, the selection-cum-improvement management system is suggested in order to maintain the high level floral diversity of these natural forests.

The dark coniferous Taiga (Picea obovata Ledeb. and Abies sibirica Ledeb.) is the basic component of a vegetative cover of the western slope and dividing range in the middle Ural mountains. It has large hydrological, soil-protective importance and many other valuable qualities. The modern structure of forests has been developed under the influence of long-term anthropogenic impacts. Except for separate territories of primeval forest, the prevailing part of stands has arisen from felling or fire and now, they are at different stages of reproduction changes or age of development.

Methods of studying the reproduction and age dynamics are based on:

a) Use of a genetic approach to classification of forest types, reflecting the geographic and spatial ecological differentiation of forest-vegetation conditions, and also the dynamics in time of forest communities combined into forest types;

b) Use of the probability-statistical method of revealing the regenerative and age dynamics of forest communities with respect to different types of forest-vegetation conditions, and obtaining the characteristics of stands in each age class, printed out by a computer in the form of growth-trend tables for modal stands after processing forest inventory data (Shikhov, Smolonogov, 1984; Smolonogov and Shikhov, 1987).

The analysis of the received material has shown that all the processes of reproduction and age dynamics can be divided into two periods, six phases, and ontocenogenesis stages of forest-forming trees can be distinguished.

The first period: The restoration on final cutting areas of all forest-forming trees of this type of forest-vegetation conditions occurs in this period. The deciduous species (birch, aspen) prevail in the structure of formed stands. The deciduous species pass almost the whole cycle of vital development and determine speciality of endogenous changes of environment and structure of communities. The potential opportunities of coniferous species as main edificators of the following period are realised. The period is subdivided into three phases. The first phase continues for approximately 20 years and comes to an end by the formation of deciduous young growth with the participation of the undergrowth of preliminary generation of dark coniferous species. The second phase proceeds till 35-40 years and is characterised by the intensive growth of deciduous species. During the third phase, the upper deciduous story is stabilised and its degradation begins. Spruce and fir grow in the understories of the communities. It proceeds till 80 years.

The second period: The intensive destruction of the upper deciduous story and formation of dark coniferous forests structure occurs in this period. The fourth phase proceeds from 80 till 120 years. During this phase, the upper story, which is dominated by dark coniferous Taiga species, is formed. The fifth phase is characterised by the stabilisation of economically mature spruce and fir forest, high seed production of all forest-forming species, and the appearance of second-growth generations of dark coniferous Taiga trees. It proceeds from 120 till 160 years. The sixth phase begins after disintegration of trees of old-growth generations and formation of young generations on account of undergrowth. The unequal age of trees in a stand is increased. Endogenous (change of generations) and exogenous (selective felling, windfalls) reasons can influence the degree of non-uniformity of tree-age in a stand (Pozdeyev, 1992).

The revealed laws of structure and dynamics of mountainous dark coniferous forests allow developing scientific principles for forest management.

Nitrogen flow and storage of ecosystems were investigated in a subalpine ecotone in the Bavarian Alps in order to balance nitrogen budgets, to detect variations in nitrogen saturation and to estimate future nitrate outputs into groundwater after climate change. The following ecosystems occurring in a mosaic on a northerly slope at 1700 m a.s.l. were investigated: Spruce clusters (fragment of Adenostylo glabrae-Piceetum), dwarf pine krummholz (Rhododendro-Pinetum mughi), both occurring on thick acidic humus layers overlying limestone, mat-grass meadow on acidic humus overlying limestone (Geo montani-Nardetum) and calcareous meadow on rendzina developed from limestone debris (Caricetum ferrugineae). We measured bulk precipitation, throughfall, litterfall and seepage water in quantity and chemical composition for important bioelements, pH and DOC, fortnightly in summer and monthly in winter. Air humidity and temperature, soil matrix potential and temperature were recorded in order to explain inter-annual variability in nitrogen fluxes and derive a water balance model. Soil and biomass inventories were performed to characterise nutrient pools of the ecosystems.

Spatial and temporal differences in nitrogen in- and output are discussed considering structure, site variability and ecosystem succession.

In the investigated spruce cluster ecosystem, nitrogen input was elevated by 30 kg N*ha^-1*a^-1 as compared to bulk precipitation. 6,6 t*ha^-1*a^-1 are stored in the soil, total nitrogen storage amounts to 26 kg*ha^-1*a^-1 due to continued biomass accumulation.

Under pine krummholz, the nitrogen balance results from two components of contrasting behaviour: We detected low nitrogen outputs and high nitrogen storage in the upper horizons and inverse relations in the deeper horizons of the humus layer. This corresponds to a marked vertical gradient of soil chemistry, the upper horizons having C/N-ratios of 33 and pH of 3,8, the lower C/N-ratios of 18 and pH of 5,1. The low content of pine pollen (Weber, pers.com. 1998) indicates that the lower portion of humus was formed in a succession stage with Carex firma and Dryas octopetala vegetation.

In the mat-grass meadow growing on acidic humus comparable to the conditions found under spruce clusters and krummholz, we detected a lower yearly storage rate of 6 kg N*ha^-1*a^-1. High soil nitrogen store of 22 t*ha^-1, a relatively narrow C/N-ratio of 20 and biomass production between 2,3 to 3,2 t*ha^-1 with no current accumulation indicate N-saturation. Mat-grass (Nardus stricta) appears to enhance decomposition rather than contribute to humus layer accumulation. The humus accumulations found under mat-grass are assumed to stem from a previous krummholz stage as old pine roots can be found in the meadows.

The calcareous subalpine meadow, as the only community growing on mineral soil, stores nearly the whole nitrogen input from precipitation. At a soil nitrogen storage of 11 kg*ha^-1*a^-1, this system is clearly far removed from nitrogen saturation.

The evolution of the regeneration of natural forests is connected with the openings in the canopy cover (gaps). In the last few decades, the attention of silviculturists has been devoted to verifying treatments of low environmental impact in accordance with the principle of sustainable silviculture, conservation of biodiversity and modern concepts of forest management in ecosystemic terms. Studies to determine the role of artificial gaps on the establishment of tree species have therefore been carried out in many countries. In Italy, experimental areas were located in the Camaldoli Forest (southern slope) and in the Campigna Forest (northern slope) of the central Apennines.

In 1982, at the end of summer, six circular gaps ranging from 463 to 1243 m² were created. Three permanent transects were located inside each area, parallel to each other, and each of these was subdivided into subplots of 2 m² and characterised according to its within-gap position. All the seedlings were recorded in 1986, 1991 and 1997. Silver fir (Abies alba Mill) is not able to self-regenerate adequately in the Casentino National Park and seedlings occur above all in the gap-edge. In the Campigna Forest, the R.I. (Regeneraration Index, according to Magini,1967) currently reaches about 50-80% of the values considered to be full regeneration, while in the Camaldoli Forest, silver fir has disappeared in some areas in the gap-centre and partially in the gap-edge. This decline is closely related to competition between the woody species and deer browsing. At the same time, favourable conditions for the regeneration of beech (Fagus sylvatica L.) are found in the gap-edge. The natural regeneration of sycamore (Acer pseudoplatanus L.), established soon after the gap opening, is widespread in all study areas of the Campigna Forest both in the gap-edge and above all in the gap-centre. The rapid juvenile growth of the sycamore allows the seedlings to survive even when in competition with large sized herbs. Sorbus aucuparia L. seedlings are frequent in the gap-edge of the Campigna Forest area and, like sycamore, can survive competition with large-sized herbs. Other species sporadically occur, among them Salix caprea, Quercus cerris, Prunus avium, Castanea sativa, Sorbus aria, Laburnum anagyroides and Corylus avellana.

All areas in the Campigna Forest are considered to be well regenerated: values of RI > 100 are considered the limit of good regeneration. In the Camaldoli Forest, the species dynamics is still at a shrub phase, as confirmed by the low values of RI < 40. The tested artificial gaps are able to change species composition and structure in 8-10 years in pure silver fir stands.

Dark coniferous forest is widely dispersed over China. It is one of the chief forest types in the Changbai Mountain Area and it is found between 1100-1900 m a.s.l. in the subalpine area of Changbai Mountain. The study object of this research was the dark coniferous forest on the northern slope of Changbai Mountain, which is mainly distributed between 1100-1850 m a.s.l. Twelve quadrates were investigated at the elevation of 1200 m, 1300 m, 1400 m, 1500 m, 1600 m, and 1700 m a.s.l.

The investigation results showed that the chief tree species in the dark coniferous forest are Picea jezoensis, Pinus olgensis, and Abies nephrolepis. At the lower elevation, there are Picea koraiensis, Pinus speciosa, and Pinus koraiensis. The main sub-canopy species are Acer ukurunduense, Betula castata, Sorbus pohuashanensis, and Populus ussuriensis. There are A. tegmentosum, A. mono, A. tschonoskii at the lower elevation and Betula ermanii at the higher elevation. The variance analysis of species in 12 quadrates showed significant differences between tree species, shrub species, and herb species at different elevations. For the total number of species in the quadrates, the difference was not significant.

The code index of the tree layer and the shrub layer at different elevations were all stable in lower value, but the code index of shrubs was a little higher. The code index of the herb layer varied with the variation of elevation, and it was highest at 1400 m a.s.l. For the total number of species, because most of them were herb species, the code index of total species had a varying tendency similar to that of the herb layer.

The analysis of the community coefficients between every two adjacent quadrates at different elevations could reveal the influence of elevation variation on the structure of the dark coniferous forest. The community coefficient between quadrates at 1200 m and 1300 m a.s.l. was highest, indicating that the dark coniferous forests at 1200 m to 1300 m a.s.l. belonges to the same forest type. The community coefficient between quadrates at 1300 m and 1400 m a.s.l. was lowest, indicating that the dark coniferous forests at an elevation higher than 1400 m and lower than 1400 m a.s.l. belonges to different forest types.

The cluster analysis of 12 quadrates showed that the 4 quadrates at an elevation lower than 1400 m a.s.l. should belong to one type, which was lower hilly dark coniferous forest. The other quadrates could also be divided into two types: middle hilly dark coniferous forest (lower than 1600 m a.s.l.) and higher hilly dark coniferous forest (from 1600-1700 m a.s.l.).

Taking the community coefficient as distance of quadrates, the polar ordination of 12 quadrates was also discussed in this research. The result was mostly similar to the cluster analysis.

Within the frame of a Pan-European Programme for Continuos Monitoring of Forest Ecosystems, 20 so-called Level II Plots have been installed quite well distributed in Austria. Data assessed regularly within this programme were used to quantify stand structure and biodiversity. Tree data (species, dimensions and location) where determined on the whole plot with 0.25 ha, ground vegetation was assessed on several small sub-plots as well as on one larger sub-plot.

The calculations were based on species composition, spatial pattern and diameter or height distribution.The prerequisites and the expressiveness of each index is discussed. A comparison of the different indices shows that high plant species diversity does not correlate necessarily with the diversity of stand structure. A method of combining different indices is presented.

In this study, 35’990 grids of 1km*1km cells are used as the Taiwan raster-base map in a geographic information system (GIS) database. In the 3rd Taiwan Forest Inventory, 4’002 ground-plot data-sets are distributed systemically by 3km*3km map points. The Taiwan raster-base map in GIS is used to extract 12 habitat factors of each plot. The co-ordinated sampling plots are used as the geographic locations which integrated the data of habitat factors, stand structure and growing stock. The 12 habitat factors are: elevation, slope, aspect, mean temperature per year, mean high temperature of July, mean low temperature of January, total rainfall per year and season, soil class, soil texture, soil-effective depth. These factors were derived from point scale (sample plot) to surface scale by interpolation methods. These interpolation methods included: Thinness polygon, inverse-distance weighting, spline, Kriging, and mono-multinomial trend. Cluster analysis was used to classify distinct groups and to solve problems of interaction between variables. The results showed that there are 948 habitat clusters in Taiwan (R-square=0.95) with these 12 habitat factors. The Weibull probability density function (pdf) is used to describe the diameter at breast height (dbh) distribution. The parameters of Weibull pdf are used to explain the location, scale and shape of dbh distribution.

Two methods are used to obtain the growing stock per hectare. The first one is to sum up the individual volumes in the plot (SUV), the second one is the diameter distribution method (DDM). The Weibull parameter, bio-diversity index, growing stock per hectare are then integrated with the habitat factors to develop spatial stand structure models and growing stock estimating models. The ecological range and habitat of the main forest types are concluded. The spatial stand structure models and growing stock estimating models are developed to get the whole characteristics of the forests in Taiwan. The GIS is used to store, overlay, interpolate, analyse and display the biodiversity, stand structure and growing stock of any forest types in the area of interest. The spatial distribution of forest types, diversity, growing stock and stand distribution are clearly shown by GIS. The interaction of stand structure and habitat clusters are shown in a mathematical model, a graph and a spatial map.

Whereas beech forms more or less pure or mixed dense stands in most parts of its natural range, this is not the case in higher mountainous regions. Thus, in the Harz Mountains (Germany), this species forms only scattered stands with only few individuals per hectar at about 800 m a.s.l.

We based our studies on the hypothesis that this different structure of beech stands, i.e. their occurence as small groups of trees or even single trees, will have an impact on the crossing system of this normally outcrossing species. As a consequence, influences on the genetic structures were assumed.

For the assessment of the genetic structures, beech stands from the above mentioned regions were studied by methods of isoenzyme analysis. Significant deviations from previously observed patterns of genetic variation have been found. This has been explained by the phenomenon of genetic drift.

No significant excess of homozygotes, indicating a tendency to self fertilisation, has been observed. This lack of homozygote excess may be explained by the observed high proportion of empty seeds, indicating a strong viability selection against homozygotes in early ontogenetic stages and thus prohibiting them to be detected by genetic methods.

The consequences of genetic drift as well as increased self-fertilisation rates are discussed in the context of silvicultural activities as well as gene conservation in mountainous regions.

The alpine protection forests are a substantial basis for the protection of the infrastructure in the Bavarian Alps. Due to increasing dense settlement and intensive development of tourism, intact and function-suited protection forests are more important than ever before. By a multiplicity of different factors, the functional fitness of the mountain forest is no longer ensured or strongly threatened in many areas.

It is one of the main demands of the Bavarian state forest administration to protect the mountain forests and maintain or rebuild these forests to fulfil their protective functions. To reach these targets, the forest administration started a reconstruction planning in these protection forests. For this, it is necessary to detect disturbances in the forest structure in time, in order to keep the high costs of complex measures of technical constructions as low as possible. So far, the data needed for this planning were raised by means of terrestrial surveys and visual interpretation of aerial photographs. Due to the extreme topography, many areas are very badly accessible, so that a frequent monitoring of the forests in the Bavarian Alps is always connected with high financial costs and high personnel effort.

A new possibility to get fast and efficient data for a wide-ranged evaluation of the status of the protection forest could be given by the application of the announced high-resolution satellite systems and their integration into a geographical information system (GIS). The satellite remote sensing can supply digital data in short time and by low costs for a big area. These data will be integrated in a GIS and combined with further information, e.g. data of the topography or information from the forest management planning. By an analysis of important parameters and the evaluation of modifications in the mountain forest on the basis of a multitemporal change detection by satellite data, critical stands can be identified, where the protection forest is threatended to loose its functions. These areas have to be examined more detailed by a terrestrial survey. The application of such a system could help to reduce the costs of the terrestrial surveys to a minimum. The monitoring of the protection forest could be based, thereby, on an efficient procedure, repeatable in short distances, in order to recognise critical developments in the mountain forests in time and to optimise the efforts of the protection forest reconstruction planning.

Five mixed Picea-Fagus stands in the montane belt of the northern Limestone Alps near Garmisch-Partenkirchen/Bavaria were chosen from a larger monitoring network to represent a representative cross -section of the regional forest communities. 10 trees per species were assessed for crown transparency. Within the crown projection of each tree, individual soil properties (profile morphology, topsoil pH) and floristic composition of the ground layer were recorded in two 1 m² quadrates. Multiple regression and trend-corrected correspondence analysis were used to detect the importance of microsites for vegetation.

In all examples, considerable variation of soil depth, profile morphology and topsoil pH were found which can be attributed to the patchy nature of soil-forming processes and disturbance. Thus, mountain forest soils should always be sampled in replicates. Topsoil pH depends mostly on forest floor depth, which is particularly variable on stony calcareous parent materials. Crown transparency of individual trees could not be predicted from the microsite data and it is hypothesised that our method of sampling is not sufficient to characterise rooting zones.

Ground vegetation species richness and cover are clearly reduced at microsites with deep litter layers, partly also at acid microsites with deep forest floor. Small-scale richness and total community richness show no clear relationship in our data. Species composition beneath beech and spruce trees was very similar in all stands, which corresponds to a week relationship between canopy and microsite. Litter quality at microsites (coniferous vs. deciduous), which is only weakly associated to the position of canopy species, is a much better predictor of species composition.

The gradients of species composition within stands are not in all cases related to the major regional community gradients. It appears from this study that variation in humus quality produces similar gradients at the community and at the stand level.

According to our results, mixed mountain forests can neither be regarded as a mosaic scaled by the distribution of coniferous and deciduous trees nor as a patchwork of well-defined microsites. The need to include temporal variation and the immense effort for data collection constrains the development of models integrating the system compartments.

A study area covering approximately 100 ha was set up in the subalpine Norway spruce forest of Paneveggio in the Eastern Italian Alps in 1992. Norway spruce forest is the most widespread forest type in the Central-Eastern Alps and is of great importance for timber production, catchment, protection, and nature conservation value. The area is divided into two parts. The first one is dedicated to silvicultural research (silvicultural reserve) and the second one is used as a reserve from which human activity is banished (wilderness area). In the part devoted to silvicultural research, several studies concerned with natural regeneration and ecological stability have been carried out. In the wilderness area, the aim is to monitor forest stands, and the first step of this research has been the description of the structure and the study of the origin and the types of disturbance that these stands have undergone. Disturbance history and forest structure were studied in three subalpine Norway spruce stands using different approaches. The presence of past intense human activity necessitates investigation of multiple lines of evidence, and the examination of all kinds of written evidence. Only by using all these sources of information is it possible to delineate the most important features of the disturbances that affected the origin and subsequent growth of a forest stand and to apply a close-to-nature silviculture that uses natural regeneration and mimic natural disturbances.

The mountain ecosystems and landscapes research unit of Cemagref, Grenoble (France) has improved a ranking forest method according to their protective role against snow avalanches. This method takes into account the importance of natural hazards (avalanches, rockfall, etc.), the structure of stands, the rate of forest cover and the vulnerability of the threatened sites. The comparison of such information with forest data shows that forest maintain these phenomena at a level of potential activity. Forest intervention zones having priority are determined in this way.

But the understanding of this natural hazard mastership by the forest needs a better understanding of the forest cover influence on snow cover. An experimental site has been set up to study the snow distribution in the forest depending on the forest stand structures. It is located at Le Chornais, Arêches-Beaufort, Savoie (France).

This experimental site contains of a network of 51 snowpoles set up between 1400 and 1600 m a.s.l. The forest consists of even-aged spruce (Picea abies) stands. The bottom part is 60 years old and the top part 150 years old. For the winter 1994/95, each week, the height of snow has been read on each pole. Trees and their crowns in this experimental site have been mapped. Values of forest cover around each pole and for several values of radians, circles (0.5 m to 10 m) have been calculated with GIS Arc-Info.

First results allowed us to build a model of snow interception by the forest canopy. This model takes into account a tree effect and a stand effect. So it can be used to appreciate the effect on snow distribution by every kind of forest stands.

One of the applications in forest management is to determine the silvicultural need to maintain or increase the protective role against snow avalanches of forest stands and their durability.

The paper suggests a clustered arrangement of trees in high altitude afforestation in order to take into account the variety of favourable and unfavourable microsite conditions and to create well-structured stands. Long-term studies on the ecology of subalpine afforestations showed that the young trees were threatened for decades by fungal diseases, browsing ungulates like chamois and deer, adverse climatic conditions, and snow movements. The intensity of the threats, however, varied greatly between microsites and tree species. The pattern of snow disappearance within a potential afforestation area provided important clues for the evaluation of favourable and unfavourable microsites. Trees generally survived well on sites with early disappearance of snow cover in spring, whereas losses were high or even total on sites with long-lasting snow cover. On cold sites with thick layers of raw humus, trees performed poorly. On sites prone to snow gliding or avalanches, the saplings were often uprooted or the stems split or broken. Based on these ecological findings an adapted system of cluster arrangement of the afforestation trees in the terrain was developed and tested ("cluster afforestation", in German: Rottenpflanzung). Prior to planting the terrain was carefully assessed in terms of microsite conditions. In contrast to the conventional large-scale regular planting pattern, only the promising sites were planted, while unfavourable sites remained unplanted. In a cluster afforestation 3 to 6 "small collectives" were densely planted with 20 to 30 trees of one species. This way, the crowns of the trees within these small collectives will close within 5 to 10 years. After a few decades the small collectives should merge into a larger "tree cluster" (in German: Rotte). The distance between the tree clusters was large enough to keep them separate for their whole life. Such stands are expected to become highly structured and more resistant against strong winds and heavy snow loads and less threatened by insects than uniform stands. They can permanently contain extensive inner margins, and are therefore ecologically rich habitats pervaded by light and warmth. Cluster planting on favourable sites minimises sapling mortality and is the best strategy for afforestation at higher altitudes. It closely imitates natural regeneration leading to highly structured, resistant stands. Cluster planting further reduces expenses for tending during the early stages of the afforestation.

Different factors affecting tree growth of Scots pine mountain stands have been analysed. We have found some differences of individual tree growth between even-aged, uneven-aged and mixed stands with oak (Quercus pyrenaica). There is also a positive correlation between growth and tree diameter. Even-aged stands showed higher growth values than uneven-aged stands when trees surpassed 10 cm dbh, and stands from artificial plantations showed the highest growth rates. Pines from mixed forests showed the lowest rates.

An ecological stability property is a selected ecosystem attribute that characterises the behaviour in time of the ecosystem, including its dynamic reaction to disturbance. If the ecological context is clearly stated, stability properties, such as the constancy, resistance and resilience of various ecosystem attributes can be described and quantified. These attributes include structural attributes such as the spatial distribution of gaps, trees and microsites suitable for regeneration and coarse woody debris. Mountain forests that provide protection against natural hazards are managed to enhance ecosystem resistance and/or resilience, depending on the developmental stage. This is mostly done by manipulating the structure of the tree layer. However, manipulations of the stand structure usually affect other ecosystem attributes and related stability properties, including properties that are crucial for long-term effective protection against natural hazards, such as the availability of nutrients and water in the soil, the resistance of the ground vegetation against tree invasion, or the occurrence of heart rot in trees. The ecological stability properties that are crucial for long-term effective protection need therefore to be identified, monitored with suitable indicators, and integrated into sustainable forest management. A concept to achieve this integration at catchment scales is presented.

In a research project of the ETH Zurich, growth was examined in research areas located in six Norway spruce (Picea abies Karst) afforestations. The afforestations, densely planted and spatially separated tree islands (Rotten), are located in the belt of closed forests in the high-montane and sub-alpine zones of the Swiss Alps. They have reached the thicket and the pole stages in their development and until now, they have not been tended at all or else they have received only minor tending.

The horizontal co-ordinates of a total of 3259 stems in 62 collectives were surveyed. In addition, the annual growth of the leading shoots for the past 10 years, total tree height and the current dbh were recorded. Complete collectives were cut on one of the research areas to enable stem disc analyses. In this way, both the height as well as the diameter development could be reconstructed for the past 20 years.

The results show that the conical structure of natural tree islands to be noticed near the tree line, basically can not be achieved by even-aged afforestations on open ground in montane forests. Naturally formed tree islands radiate from very few individuals, gradually expanding around the edges by natural seeding or layering with the result that uneven-aged cohorts are formed. In contrast to the conical structure of natural tree islands, the planted tree collectives of all the examined areas are characterised, regardless of their size and the elevation, by the fact that the trees bordering the margins of the collectives are higher and, first of all, thicker than the trees situated inside the collectives. This advantage is already indicated in the thicket stage and it is significant in the pole stage.

The stems at the edge of the collectives are usually stable as a result of this development (height/diameter-ratio < 80). The inner sectors, on the other hand, become frailer with the advancing development stages. The stems have small crowns and a high height/diameter-ratio, the dense, concave canopy promotes the accumulation of heavy snow loads. The fragile inner sectors can possibly be avoided by reducing the stem number so that by the time the pole stage is reached, the tree collective only consists of marginal trees. Otherwise, early and effective tending is necessary at all the studied elevations in order to preserve the mechanical stability inside the collectives.

The results do not confirm the expectations expressed in the relevant literature. They allow the conclusion that the planting of tree collectives with 5-20 plants on very small areas is an economical method of quickly creating storied, stable stands which require little tending. The development of uneven-aged stands is promoted by the possibility of natural regeneration in the gaps between the collectives as well as the gradual and safe removal of individual collectives with subsequent regeneration. Tree collectives on small areas can basically be planted everywhere. Their advantages are particularly shown in silviculturally unsuitable sites and in slopes prone to avalanche fractures.

Calabria is a typical Mediterranean region. Geographically, it is the long tip of the boot-shaped Italian peninsula and extends from north-east to south-west and is surrounded on three sides by the Tyrrhenian and Ionic seas, which in the widest point of the region are less than 100 km apart. It is a prevailingly hilly and mountainous territory whose orography is formed by a series of ranges parallel to the cost. The climate has the well-known characteristics of the Mediterranean regimen, that is heavy rainfall in autumn and winter and a hot, dry summer, which also extends to its mountain belt, though only in July and August. Nevertheless, especially on the Tyrrhenian side, the average annual rainfall exceeds 600 mm also on the plain and makes the growth of a varied forest vegetation possible. About 30% of the region is covered by dense forests.

The mountain belt extends between 1000 m a.s.l. and the timberline, which on Mount Pollino (northern Calabria) reaches 2150 m a.s.l. The forests in this area are dominated by beech with silver fir, Calabrian pine and other hardwoods. Along the highest slopes of Mount Pollino, there are stands formed by Bosnian pine (Pinus leucodermis), which show a remarkable resistance to the dry calcareous soils and severe climate conditions. Just as in other Italian Mediterranean mountainous areas, the structure of forests originating from natural regeneration is quite composite, that is to say it is formed by a mosaic of small stands with different characteristics. This is the result of the variety of site conditions in small areas and heterogeneous silviculture treatments.

The excessive exploitation until the fifties compromised the important functions which these forests have, first of all soil protection and the regulation of water flow in watersheds. Many ruinous floods have occurred in Calabria during this century. Things have been gradually changing over the last decades. The productive function of the forest still plays a role in the regional economy, but forest management increasingly tends towards multifunctionality as local development in southern Italy’s mountains will depend more and more on tourism, which requires nice landscape, nature conservation areas, protection from natural hazards, etc.

This paper shows the recent parts of research conducted on Calabria’s mountain forest structure and the relationship between stand structure and forest efficiency, as well as soil conservation. These kind of studies represent one of the bases of a sustainable silviculture which is also respectful of intrinsic and functional forest values.

Investment in the development and testing of silvicultural alternatives to clearcutting in coastal montane British Columbia, Canada, has been increasing over the past ten years to address concerns over regeneration, sustainability, visual aesthetics, wildlife habitat and biodiversity. The objectives of this operational trial of montane alternative silvicultural systems (MASS) are; to test silvicultural alternatives, document costs and feasibility and to study the biological impacts of different harvesting intensities. In this study, growth and physiology of naturally regenerating and planted Abies amabilis and Tsuga heterophylla were used to compare Green Tree (25 trees/ha), Patch Cut (1.5 ha) and Shelterwood (25% ba retained) treatments with a 69 ha clearcut and the virgin forest. Five years after harvesting and planting, growth in the clearcut was not significantly different than in the silvicultural alternatives, except in the shelterwood where growth was less. The different levels of overstorey retention were not sufficient to modify foliage development and therefore the above-ground light environment of the Clearcut was not significantly different than the other systems. Neither light capture and use nor photosynthetic capacity were significantly different between the Clearcut and the other silvicultural systems. There were no seasonal soil water deficits or tree water stress in any of the silvicultural alternatives and the Clearcut was the same as the other systems. As indicated by the post-planting fertilization and herbicide treatments, growth limitations were primarily below-ground associated with poor nutrient availability. Overstorey shading did not limit growth but competition from herbaceous vegetation did affect seedling growth and physiology.

In the Swiss mountains, there are numerous Norway spruce timber stands which have never or have not received any silvicultural treatment for a long time. It is feared in many places that, should these stands be left to themselves, the stand stability and the stand structures will decline to such a degree that the protective function against natural hazards is reduced.

A new project is presented: On selected stands, the following will be monitored and analysed: a) selected silvicultural treatments to maintain and improve the stability of the stands b) the subsequent development of the stands. The effects of the chosen treatments on stand stability and structure will be examined in the long term by comparisons to adjacent stands with similar structures and similar stability which have received no treatment and with stands which have been managed "regularly" for a long time (experimental plots of the Swiss Federal Institute for Forest, Snow and Landscape Research (WSL)). The economical components will be considered with the support of models.

Preventive improvement and in-time regeneration of mountain forests on steep slopes with direct protective functions for settlements, roads and railways is an important public task in Austria.In-time regeneration on steep slopes often requires femel-felling (group felling) in mature stands, which may lead to some problems:

Destabilisation of the remaining stand; clearings in snowy regions that are too small will hinder or inhibit regeneration, especially of intolerant tree species; whereas clearings that are too large may develop to avalanche starting-zones.

There is no doubt that there have been large-sized die-back phases in natural mountain forests before the beginning of forest management, but regeneration took place nevertheless. The essential difference to our time most probably was the fact that the trees were not removed but remained on the clearing. Dead, still standing and cross-wise lying trees and trunks formed a dense "entanglement" with positive influences on regeneration.

The report deals with experiences made in the western part of Tyrol (Austria) with artificial "entanglements" to enhance regeneration on steep-slope forest stands. Such "entanglements" are formed by felling along the contour line in sufficient density and by leaving the stumps as high as the expected snow pack. These entanglements have proved to be a good support for natural and artificial regeneration of mountain forests on steep slopes.

In mountain forests, uneven-aged stands are usually favoured because they combine short-term stability (resistance), granted by sturdy trees or groups of trees, as well as long-term stability (resilience), granted by a permanent renewal of forest cover. Yet, uneven-aged silviculture is complex because it has to refer not only to a "mean tree" of the stand but to all its diversity in term of tree population, spatial organisation and life events (regeneration, growth and death). Furthermore, the only uneven-aged silviculture which has been theorised, the selection system, is usually unsuited to mountain conditions because of climatatic conditions (necessity of spots of light and heat for regeneration in subalpine conditions), heterogeneity of microhabitats (cliffs, rocks, variations in geomorphology) and difficulty of exploitation (easier in small patches). Fundamental work on uneven-aged forests dynamics and management is then needed.

Models of forest dynamics can be of great help in this context, because they give a frame to formalise our knowledge on processes, to study how they interact and give birth to the dynamics of a stand, and to simulate the effect of thinning strategies, cutting down lengthy field experiments. We present a distance-dependent growth model for uneven-aged stands of Norway spruce. Competition between trees is modelled by the available light at 90% and 60% of the height of the trees. For each subject tree, the incident radiation is first spatially divided into elementary beams. The interception of each beam is calculated using a geometrical description of the crown of the surrounding trees and the available (non intercepted) light is summed up over all beams. This light submodel was validated comparing predicted light with observations made on hemispherical photographs. Height and diameter increment is predicted from tree age and light availability by empirical functions calibrated on two experimental plots in the French northern Alps. The model has been integrated in the software Capsis 3.0, which allows to test silviculture scenarios easily, selecting trees on an interactive map of the stand.

Such a model can help to understand how the structure of a stand influences its dynamics. We try to show this with three applications. The first example shows how the growths of three theoretical stands composed of the same trees depend on their spatial distribution. A second application illustrates the use of the model to compare silvicultural strategies: we show how concentrating a thinning on a limited amount of trees or groups of trees influences growth differently from a diffuse thinning. The last example shows how openings of different sizes result in different light patterns on the ground. Perspectives and future developments are then discussed with special emphasis on regeneration and death modelling.

Software has been developed which meets the needs of foresters, and runs on a PC or Macintosh computer. The approach adopted is one generally used in micro-engineering production. The software enables trees, groups of trees and dense forest to be recognised on different types of digitised aerial photographs (black-and-white, colour, or infrared) and using photographs at different scales. The programme calculates areas and locations through a link with a digital terrain model. There are numerous possible applications: calculating the percentage of forest cover, identifying the edges of the forest, estimating the stability of mountain forest, and running diachronic studies of forest dynamics.

The conservation of biological diversity should be given an unprecedented attention. It is recognised that biodiversity at all levels – gene pool, species and biotic community – is important for many reasons and that it is being rapidly diminished by habitat destruction and other damaging influences resulting from human population growth, pollution and economic expansion. No one seems to think that we can do anything effective to control the root causes soon enough to provide breathing space for the biota. Humanity is the main cause of the destruction of the ecosystem. The maintenance and structure of diversity could be achieved if there is a co-existence between Homo sepciens and other species. The interest of humanity should not override the interest of the other living organisms.

We investigated the nature of the resource limitation for natural regeneration of tree seedlings in a montane temperate old-growth rainforest with a prevailing gap-disturbance regime in the Coast Mountains of south-western British Columbia. Over 11 years, more than 22'000 seedlings were followed in an experiment involving i) different light levels inside and outside small gaps, ii) the exclusion of the root competition of canopy trees by trenching, and iii) exclusion of competing herbs and shrubs by weeding. Most tree seedlings under investigation were Abies amabilis, Tsuga heterophylla and T. mertensiana.

While seedling emergence was almost even in all experimental variants, survival and height growth were different. Survival was much higher in Abies than in Tsuga seedlings, and slightly higher in gaps than under canopy. Height growth was higher in gaps, and on trenched and on weeded plots, but some effects were only significant for seedlings smaller than 20 cm. Height growth was very slow, and 42% of the Abies seedlings even had, within three years, at least one missing internode on the terminal leader. Competing vegetation further impaired successful establishment, probably by reducing light levels and by smothering, while competition of canopy trees for nutrients and water was of minor importance.

Mountain regions are of special importance for global change research. Due to their strong altitudinal gradients, many mountain regions provide unique opportunities to detect and analyse global change processes and phenomena because of the following reasons:

Climatological, hydrological, cryospheric and ecological conditions change remarkably over relatively short distances; thus biodiversity tends to be high, and characteristic sequences of ecosystems and cryospheric systems are found along mountain slopes. The boundaries between these systems may experience shifts due to environmental change and thus may be used as indicators of such changes.

The higher parts of many mountain regions are not directly affected by human activities. These parts include many national parks and other protected environments, and thus may serve as locations where the environmental effects of climate changes alone, including changes in atmospheric chemistry, can best be studied.

The global distribution of mountain regions allows us to perform comparative regional studies to better understand the regional differentiation of the ongoing environmental change processes.

Integrated interdisciplinary research activities on global change in mountain regions are suggested to be implemented in a co-ordinated way to understand, model and predict environmental changes and their impacts in mountain regions and, to develop proposals towards sustainable resource management. The required research is structured around four interlinked activities:

Activity 1: Long-term monitoring and analysis of indicators of environmental change in mountain regions

Activity 2: Integrated model-based studies of environmental change in different mountain regions

Activity 3: Process studies along altitudinal gradients and in associated headwater basins

Activity 4: Sustainable land use and natural resource management.

A number of specific tasks are suggested under these four activities, which will be described briefly. Moreover, suggestions will be made for the implementation and international co-ordination of the research.

Applications of a growth model for Scots pine (Pinus sylvestris L.) stands in Spain for prevention of snow and wind damage are presented. The used model is a Parameter Recovery Model (PRM) based on the Weibull function. First, this model estimates stand growth and attributes; and in a second stage, the stand structure can be simulated. The model is valid for predicting thinned and non-thinned stand evolutions.

In this work four treatment are simulated in order to examine the effects of thinning on stand stability. The treatments are three low thinning regimes, with different intensities and initial ages and one control treatment (unthinned). Stand attributes, diameter distributions and h/d ratio by diameter class are estimated for each treatment.

According with results obtained in a study about damage caused by a snowstorm in Scots pine stands in a mountain area of Spain (Río et al., 1997), the critical value of h/d ratio above which the majority of the trees were damaged ranges from 70 to 90. Thus three degrees of susceptibility of a tree to snow and wind damage have been established: low susceptibility, with a h/d ratio below 70; middle with a h/d between 70 and 90; and high with a h/d above 90. For the four simulated treatments, the numbers of trees per hectare in each class of susceptibility are calculated.

The results clearly show the effect of thinning on the risk of snow and wind damage. On unthinned stands, there are many trees with high and middle susceptibility to damage and the stand density is also high. Thus, if a snowstorm or a strong wind occurs, damages are likely to happen. In the other treatments, the stand density and the number of trees with middle and high susceptibility are lower. These kinds of trees decrease with the heavier intensity of the thinning regime. The first thinning age is very important to assure the stability of a stand, since the effect of this thinning on the stand structure is not reached with a later heavy thinning. In order to reduce the risk of this kind of damages in Scots pine stands early, low and stark thinnings should be made.

TANAP, the Tatra National Park, was established in order to protect untouched mountainous nature and to reconstruct the parts changed by human activities in the Tatra Mts. From the total area of 110 000 ha, 75% are covered by forests, which are only protective and special purpose forests. Diversity in species, vertical structure and ages has been designated as a target for forest management since the set up of the national park in 1949. This concept was derived from the experience that only a heterogenous stand can fulfil long term stability of those mostly spruce stands.

The intentions to rebuild non-natural stands are permanently disturbed by regular wind and insect calamities. Foresters are forced by law to give priority on working-out the incidental fellings. This has led to a limited realisation of the thinning prescribed by the Forest Management Plan. During the last three decades, the percentage of incidental fellings was between 70-90 %. This facts made us analyse the stand structure and assess the perspective of its development according to current management practices and influencing factors.

According to the age, stand and site conditions, we selected 8 types of larch-spruce forest. Using 50 circle plots and aerial photographs, we quantified tree species composition, vertical and horizontal structure, crown projections, and natural regeneration. As influencing factors, we assessed the type of relief, the stand naturalness, the historical use (grazing, clear-cutting, fires), incidental fellings and thinnings.

The results have confirmed a good correlation between relief and stand heterogenity. Stands with diversity in species, ages and vertical structure were mostly found on rocky relief with ranker soils. Pure spruce stands are dominating on regular, flat relief with deeper and more loamy kambisoils. Easy access to the flat oldest moraines led to an economical use in ancient times. Consequently, forest stands are mostly not natural there. Areas where grazing took places are today covered by forests with heterogenous structures with remarkable stability and aesthetic value. Conversely, artificially established stands on former clear-cuts or burned places are very unstable, mostly even-aged monocultures. Current silviculture practices (thinning from below) are increasing their unnatural character.

By the method of critical loads, we assessed the ability of both natural and modified ecosystems to buffer S and N impute. High acid deposition is probably a reason of high Al concentration in soil waters which is demonstrated on some localities by limited growth of the root system. This could be one reason of regular uprootings on some localities. But the main reason is the wind. The south side of the Tatra Mts. is said to be the most affected by strong wind in the whole Carpathian ridge. Windfalls with the extend of natural disaster are repeated in 80-100 year cycle. By this time, a volume of 550-600 m³.ha^-1 is created in a stand with a high number of trees, short crown and a highly located centre of gravity.

Windfalls allow succession of larch on flat surfaces of moraines with deeper, loamy soils, which are optimal for spruce under "normal" conditions. Climax larch-spruce forest occurs on ranker soils with open tree canopy, suitable for permanent presence of larch. Using a GAP model, the potential impact of climate changes on larch-spruce forest has been assessed.

This poster resumes the behaviour of four subspecies of varieties of Pinus nigra Arn. Data were obtained from three inventories developed in 1965 (34 years), 1986 (55 years) and 1998 (67 years) in a P. nigra afforestation in Covarrubias (Burgos).

The afforestation is based in a subspecies introduction experiments: subsp. Salzmannii (Dunal) Franco (hispanic subspecies), subsp. Nigra (Ritch) Schwz. (Austrian subspecies) and subsp. laricio (Poiret) Maire, the latter with both corsican an calabrian varieties. This is the only experiment with the four pines in Spain.

The behaviour is compared with Pinus sylvestris L. and P. pinaster Ait. in the same afforestation.

The poster presents natural Norway spruce Picea abies (L.) forests in karstic depressions on the Sneznik mountains. The Sneznik mountains are the central part of the Dinaric Karst in the south-western part of Slovenia. On the slopes of the highest peak Sneznik (1796 m), the upper beech timberline (Fagetum subalpinum) reaches the altitude 1500 - 1600 m a.s.l. On the surrounding plateau, there are geomorphologic shapes named doli, drage and kotli (kettles). These are one of the greatest karstic depressions with diameter between 0,5 and 1 km and depth from 100 to 300 m. Characteristic for them is also temperature and vegetation inversion. While the plateau is covered mostly with subalpine and mountain beech forests, natural Norway spruce sites occur in lower parts of depressions.

In large and medium size depressions, named also freezing ravines, mostly Piceetum montanum dinaricum sites are found, accelerated by both microclimatic and edaphic ecological factors. In smaller and deeper freezing ravines, more extreme factors are present, which results in Piceetum subalpinum dinaricum sites.

As some of the depressions are inside of protected forest reserves, we can observe natural development of stands, structure and natural regeneration not influenced by forest management. The poster presents some of these characteristics as well as possibilities for natural regeneration after windbreak (1993) on a small area in one of the freezing ravines.

Dendrochronological investigations of Norway spruce in several depressions show similarity, therefore, it was possible to construct a 167 year long spruce chronology for representative natural dinaric spruce forests on the Sneznik mountains.

In a case-study, the spatial structure and development of a forest which has been left undisturbed by human intervention was investigated, applying comprehensible quantitative methods. The study was applied to mountain pine forests (Pinus montana var. arborea) in the Swiss National Park in the lower Engadine. Based on a hierarchically structured taxonomy with various structural levels, the analysis started from the individual tree.

The following methods were applied in the structural analysis:

• Numerical evaluation without (number of trees, growth increment etc.) and with consideration of spatial constellations (crown cover, aggregation index, distance of neighbouring trees).
• Pictorial representations (profiles, ground plans and chart diagrams).
• Ordinary Dirichlet tessellation (area potentially available, identification of neighbouring trees).
• Cluster analysis (single linkage) considering various functions in relation to distance and subsequent generalisation in terms of geometrical primitives (lines).
• Bivariate kernel smoothing (fixed normal density kernel) with various kernel widths and attributes.

Structural indices alone conveyed only an abstract and incomplete insight into the structure of a forest. Methods based on immediate neighbours and their distance do not account for the clustering of trees.

Pictorial representations are a traditional means of documenting visual impressions. Computers enabled an area whose trees have been individually charted to be depicted as profile and ground plan in any spatial position or time frame. The representations render clear and realistic pictures, despite the relatively simple reconstruction of the tree architecture.

The ordinary Dirichlet tessellation was very descriptive in its display of graphical and numerical results. However, it did not account sufficiently for the complex spatial conditions and their mutual influences. The ordinary Dirichlet tessellation has to be extended in order to take into account, first of all, the vertical structure.

Cluster analysis visualised discrete linear textures. The assembled clusters were further analysed. An aggregation of the elements was feasible on hierarchically heightened levels.

Kernel smoothing allowed the forests to be viewed as continuous texture. The results were projected as three-dimensional surfaces. The option to use different attributes for the kernel smoothing (dbh, growth increment, etc.) resulted in varied perspectives. The width of the kernel was continuously adjustable without impairing objectivity. This allowed the choice of a number of scales.

The applied methods have further development potential. Progress made in EDP and computational geometry will open up new possibilities in the future.

The geographical information system Arc/Info was applied as an analytical instrument.

Three structural levels were classified in the study of these mountain pine stands:

• Clusters (trees growing in close proximity to each other, up to ten trees per square meter).
• Lines (80 to 85% of all trees grow within belts of 2,5 m); in places, the lines run from north to south predominantly.
• Patches (< 10 a); by comprising structural lines or individual trees cluster; patches made up of individual trees vary depending on the criteria applied.

The studied mountain pine stands demonstrated a fairly big spatial heterogeneity. Fallen dead trees were mainly facing south. The amassing of dead trees (Dieback patch) was particularly striking. The reasons behind the various spatial structures are not known.

On October 7, 1970, the first National Park of Germany was founded in the thickly wooded mountains of the Bavarian Forest along the border between Bavaria and the Czech Republic. In summer 1997 when the Bavarian Forest National Park was extended to 240 km², it became one of the largest non-fragmented conservation areas in Middle European forests.

According to the international criteria for nature reserves of Category II ("National Park") of the International Union for Conservation of Nature and Natural Resources (IUCN, 1994) it is the most important aim of the National Park to protect natural or semi-natural ecosystems within their natural dynamics. For this purpose, management zones were founded successively, wherein further impacts by man are stopped and the natural processes are allowed to take place. Today, the main part of the original park is in this so-called "nature zone".

In summer 1983 and autumn 1984, heavy storms released radical changes in the forests of the National Park for the first time: a lot of windthrow areas of a total size of 173 ha were created, especially spruce forest stands on swampy and weak soil. In the following years spruce forests were damaged by bark beetle attacks mostly in the neighbourhood of the windthrow areas, and until 1991, 210 ha were killed. During the next two years, the increase of disturbed forests was insignificant, but since 1994, a new bark beetle calamity is observed in an unexpected dimension: Until now, 2148 ha of spruce forests are damaged!

In the mountain spruce forests of the original National Park, over 1200 m a.s.l, where management by man was stopped already in 1971, the changes are particularly radical, because 58% of this area are disturbed meanwhile.

This development, which caused consternation and lack of understanding in great parts of the native population, gives us the unique opportunity in Middle Europe to investigate the reforestation of large disturbed forest areas without any human influence.

Since 1988, CIR-aerial photographs were taken annually to document the increase of areas with damaged forests. An inventory, which is made every two years, shows the state of natural regeneration of the mountain spruce forests. In 1992, permanent geobotanical plots were established for long-term investigations, in untouched areas as well as in cleaned but not reafforested windthrow areas. The measurements will be repeated every five years and provide detailed information about the undisturbed processes of reforestation.

Daming Mountain is located at 23°10′~23°38′ north latitude and at 108°10′~108°45′ east longitude, on the tropic of Cancer. The total area is about 1'100 km². It has its own positional characteristics. As an effect of economic activities by man, biodiversity has been reduced on a large scale in Daming Mountain district. Ecological and environmental studies of Daming Mountain have theoretical and practical significance on the conservation and utilisation of biodiversity, as well as understanding the laws of happening and succession of forest vegetation, restoring and rebuilding the degraded ecosystem.

The key to the recovery of evergreen broad–leaved forest-degraded ecosystems is the restoration and development of the species diversity. The development tendency and rate of species diversity in the vegetation restoration process in the degraded ecosystem in evergreen broad–leaved forest in Daming Mountain have been studied. The main results are as follows:

1. The restoration rate of the species diversity in the restoration process in a degraded ecosystem is very different between the mid-mountain area and the low-mountain-hills area in Daming Mountain. In the mid-mountain area, Cunninghamia lanceolata was planted after controlled burning. 178 species appeared on a 600 m² site after 20 years of natural restoration, including 58 species in the tree layer. In addition, the forest community which consist of light-demanding broad-leaved pioneer trees and mid-tolerance broad-leaved trees needs only about 20 years to form, whereas the development of mid-tolerance evergreen broad-leaved forest needs 60~100 years. The developmental rate of the species diversity in a restoration process in degraded ecosystems in evergreen broad-leaved forest is very slow in the low-mountain-hills area. In 20~30 years, in a restored community, 22 species appeared on a 400 m² site, including only 3 species in the tree layer. In addition, a Pinus massoniana-forest in this district was continually disturbed by human actions, so that was is difficult to recover, but the forest remains relatively stable.

2. The change in species numbers in a recovering process of a degraded ecosystem tends toward climax by quick increasing in the early stage (2~20 years), decreasing in the mid-stage (50~60 years), and keeping up a level in late-stage (over 150 years).

3. A change of individual numbers per unit area varies in the different layers of a community. The individual numbers of the tree layer are increasing in 1~20 years, and decreasing later. The individual numbers of the shrub layer tend to continuously increase with degraded ecosystem restoration. However, a change of individual numbers in the herb layer is non-linear.

4. By analysing the diversity index of communities in different time and space, it was found that the species composition of communities develop toward the community composition in the zonal evergreen broad-leaved forest.

In this paper, we present data and results of a thinning experiment in a natural regeneration stand of Pinus sylvestris L. in Burgos (Spain).

The experiment began in 1972, when the stand was 41 years old. It is now 68 years old. The statistical design of the experiment was a randomised complete block with three blocks and three treatments. Low thinning has been carried out with varying intensities in a ten-year rotation. The treatments were control, light and moderate thinning. The thinning weight was measured by using the Residual Basal Area (%) as parameter.

Since the first inventory (1972) three thinnings have been carried out (1972, 1987 and 1997). We have done five inventories: 1972, 1982, 1987, 1992 and 1997.

The results are represented by the main variables of the stand: number of trees by hectare, diameter, height, basal area, and volume before and after thinnings. We also analysed the evolution of current and mean increment in terms of the age and the thinning intensity.

We recommend applying low and strong thinning at early ages, beginning with the first thinning, in the first quality, at the age of 20-25 years, although we do not obtain any economic profit at that stage.

The impact of wild ungulates on rowan (Sorbus aucuparia L.) population structure has been studied in three locations in the Trento region (Italy). The heigth and size structures observed, affected by ungulates, were compared with the data from Damages Inventories previously performed and residuals from power function models were calculated. The distributions of residuals shows the strong impact of the ungulates between 75-175 cm heigth and between 1,5-5 cm of diameter at the ground level.

The analysis of the population structure results to be the best way to study, and to assess the ungulate impact on rowan and to forecast the effects of the ungulate damage.

The issue of sustainability in silviculture has been the focus of much interest in the field of forestry over the last few years. Alpine forests have been logged intensively or otherwise modified by man but they have always been subject to multiple rather than exclusive use, so that often wood production has been sacrificed in favour of other functions in order to obtain the greatest social and economic benefits. In the past, alpine forests were exploited for timber, grazing and protection, whereas now their main functions are protection and timber production along with tourism, recreation and nature conservation.

Protection is of vital importance to humans and their activities in the Alps, and the sustainability of non-intervention is therefore an important issue: in the short- to medium-term, the failure to manage these forests leads to intolerable risks for people who live in and make a living in alpine valleys.

The most important feature in a protection forest is its "stability". The significance of stability changes according to the purpose of the stand. In protection forests, the stability of a population is its ability to maintain its structure and vitality in the face of internal and external influences, and its ability to reliably and continuously carry out its functions. Since the activities necessary to maintain and improve stability are costly and demanding, an acceptable rather than the ideal degree of stability should be aimed at in order to assure the functions required of the protection forest over the next 20-50 years. Operations aimed at achieving this minimum degree of stability are collectively referred to as "minimal tending of protection forests". Examples of minimal tending measures in Aosta Valley (Italy) are presented.

During a couple of decades, the selective cutting system in mountain forests has gained increasing interest in practical forestry in Norway. The system is mainly performed in Norway spruce forests and a prerequisite is an uneven-aged stand structure. Both ecological and economical reasons can be found for increased use of the system. Depending on site conditions and stocking, e.g. occurrence of advance regeneration, some 40-80% of standing tree biomass will be removed in the cutting. Mainly dominating or co-dominating trees and badly hurt trees will be removed. Small and medium sized trees and all advance regeneration should be left. A crucial point in forest planning and management for these areas are the long-time forest production and prognoses with such a selective cutting system. Material from the national forest survey consisting of uneven-aged stands from selected regions in south-central Norway has been analysed for growth. Measured increment has been tested against increment functions developed for even aged stands. Although fairly good agreement was achieved, new functions were developed based on the forest survey material. These functions are based on standing volume, site index and volume-weighted age of the stand. Evaluation of the functions against independent material from long term plots and temporary plots, where stand growth history has been reconstructed, is in progress. So is also the evaluation of the gain of accuracy in using single tree models. The results seem to indicate a reduced volume increment in the long run at a magnitude of 20-40% compared to even aged stands.

The "Abetina di Fondo" is a reserve forest for the protection of the village of Fondo (Chiusella valley, province of Turin). It is one of the few relict forests mainly composed by silver fir situated between Sesia and Susa Valley, between 1250 and 1500 m a.s.l. with north aspect and more than 30° inclination. The 21.2 ha old stand (more than 200 years old) stays in the middle of an expansion area of about 130 ha where fir is usually younger than 40. The ownership is private in a common property of more than 50 people (the old families of the village).

The study was preceded by a historical analysis. The structure of the stand has been carried out analysing the old stand and the expansion area (classical dendrometric parameters in transect plots), regeneration (transects), and age structure (dendroecological methods).

We know from the historical research that this silver fir forest already existed in the 17th century. In 1700-1800 it was managed with two rotation periods: at about 24 (may be beech coppice, if it was present) and 85 years. The last cut documented was realised in 1927 (only 6 trees).

The old stand is quite pure (other sporadic species are, mainly in the clearing, Sorbus aucuparia, S. montana, Salix caprea; only few young Fagus sylvatica were found); the diameter distribution is typically uneven-aged with an elevated frequency of big trees; the dominant height is about 26 m, the maximum diameter found was 101 cm dbh; the basal area is 33,4 m² and the growing stock amounts to 355 m³/ha. All these parameters show that the stand structure is not normal with a relatively low density.

The stand is characterised by a high rate of necromass composed by both standing and fallen down dead trees. Some of this breakage is relatively recent: the cause is probably due to the hurricane "Vivian" (1990).

The very young regeneration (seedlings) was quite absent at the time of the first relief in 1997, while it was very abundant in 1999.

In conclusion, it is impossible to determine if the origin of the fir stand of Fondo in Chiusella Valley is natural or artificial; in any case, today’s structure, as it is very similar to the one of a virgin forest, suggests for this little stand in natural evolution, to create a natural reserve to monitor the future evolution. Meanwhile, the fir expansion area can be a good example of an advanced phase of secondary succession in fallow lands in the southern part of the Alps.

Brown bear (Ursus arctos arctos) is catalogued in Spain as an endangered species. The present distribution of the species in Spain is reduced to two sectors, both in the Cantabric range. The populations in the Pyrenees can be classified as almost extinct. The extension of the two Cantabric sectors is 3228 km² and both territories are not connected.

The occidental Cantabric population has an estimated census of 50 to 60 bears and the oriental sector counts with a population of 15 to 25 bears.

One of the critical factors for brown bear is the autumn alimentation, mainly based on accorn crop of beech, sessile oak and pedunculate oak.

Oak is a fundamental species for bear alimentation due to the high productivity of accorns and the continuity of production.

An important nucleus of brown bear is located in the "Fuentes Carrionas and Fuente del Cobre" Natural Park (Palencia province). The natural park holds important regular even aged forests of sessile oak. All the oak forests are included in Bear Management Nucleus. In many woods, the density is very high, indicating a reduced acorn production. The decrease of traditional uses such as extensive cattle raising and the high density of many stands renders possible the expansion of beech.

Beech is, from an acorn point of view less interesting for bear than oak. The expansion of beech can be considered negative for bear on a medium-long time.

In the presented study, 68 regular oak forest management units, highly used in autumn by bear are analysed. The dasometric structure and the density indexes of these stands is presented. Three forests are more intensively analysed. Based on 170 plots, the appearance of natural regeneration and of interesting vegetal species for bear alimentation, a more detailed analysis is made. The plots are grouped and evaluated on base of the complementary vegetation.

Based in the structure analysis and in the Silvicultural Norms for sessile oak in the Southern Cantabric Range, silvicultural prescriptions for bear conservation are presented.

The structure of Scots pine (Pinus sylvestris) mountain stands has been analysed in the Cercedilla and Navacerrada forests (Madrid, Spain). Two techniques of multivariate analysis have been applied together with six variables, considering 10 cm intervals of diameter class: non hierarchic cluster analysis and discriminant analysis. Both types of analysis were applied to data obtained from a forest inventory performed on the studied area. Moreover, a systematic sampling for selected circles was performed. Each circle represented a differentiated developmental structure: uneven-aged stand, two-storied high forest of Scots pine and oak (Quercus pyrenaica), a reforested stand, and an even-aged stand.

Ten different clusters were used, according to data obtained from the statistical analysis and the present situation of the studied stand. The discriminant analysis assigned a cluster to the systematic sampling plots, according to the maximum probability criteria of pertaining to a certain cluster. Maximum values were comprised between 0.3 and 0.4.

An estimation very close to the real situation of the forest could be obtained using the diameter at breast height, as well as a representation of the space variability of the structure inside the same circle, in spite of the difficulties inherent to the stands studied, which have a strong structural complexity.

On the territory of the Ukrainian Carpathians, we distinguish the following vegetative belts:

• A front range belt: forests with oak prevalence (Quercus robur L.), beech (Fagus sylvatica) and fir (Abies alba) forests
• A bottom belt: mountain forests with prevalence of a beech (Fagus silvatica L.)
• A transitional belt: with superiority of fir (Abies alba Mill.)
• A top belt: with prevalence of a spruce (Picea abies L.), subalpine and alpine of a belt high-mountain vegetation.

Spruce forests occupy 42% of the covered forest grounds, beech forests 36% accordingly, oak 10% and fir forest 7%. The common area of the land of state forest of the Carpathian region covers about 2 million ha.

Since 1964, the Ukrainian Research Institute of Mountain Forestry conducts systematic supervision on 15 permanent plots. Objects were put in old timber stands (9 in wild stands) all of basic wood species. All trees (not less than 200) on each plot are numbered and measured. Accordingly, research is carried out at all other levels of vegetation. The plots in spruce forests (7 plots) are now in the stage of technical ripeness, in fir (3 pieces) and oak (3 pieces), they are in the stage of unhinging the maternal timber-stand and forming a new generation, in beech (2 pieces), they have reached the stage of natural ripeness. By mathematical methods the parameters of century changes are set into these woody ecosystems. Dynamic structure models of the basic forest types of the Ukrainian Carpathians have been developed. Descriptions of herbaceous covering, undergrowth, brushwood and second tier are included in the exclusive taxational indexes of the basic stand of the model structure.

Within the "Biodiversity" programme in the mountain forests of the Ukrainian Carpathians, 6 permanent plots were installed in order to study the biodiversity. According to the Braun-Blanquet scale, the diffusion of all vegetation species on tiers are estimated.

In an experimental forest (Lusiwald) near Davos, the significance of snow for the regeneration of a typical subalpine protection forest is studied. Avalanche protection forests are mainly situated on steep and snow-rich slopes. In such sites, snow might be crucial for young trees in terms of load, isolation and occurrence of parasitic snow fungi.

In our experimental site Lusiwald, growth, damage and survival of seedlings and saplings of Norway spruce (Picea abies L. Karst. ) as well as winter snow conditions have been monitored for twenty years. In the years 1982-1987 and 199 -1999, occurrence and survival of seedlings was recorded on 104 permanent plots, situated in and around two newly cut forest openings. In 1989, 111 groups, each of 16 3-year-old spruce trees were planted, according to a set of different light conditions. Height-growth, damage and survival of trees was measured each summer. In winter, snow height was recorded on permanent snow-stakes while in spring, the ongoing snow melt was mapped.

Our results show that accumulation as well as melting of snow is irregular throughout the forest and its openings. The prevailing local snow conditions are corresponding to a damage-pattern of the trees caused by herbivores and pathogens. In the openings, where snow piles up to a height of 2.5 m and melting occurs only in late May, infection by snow mould, the parasitic fungus Herpotrichia nigra, is frequent and browsing by deer is marginal. The opposite was found under the tree crowns: snow height is minimal and snow-melt occurs as early as beginning of may, browsing by deer is intense but damage by snow mould is absent. Both browsing and snow mould do not seem to have an important effect on survival and growth of saplings. Overall height-growth of saplings is faster in the openings, due to enhanced light conditions in the summer. However, the earliest stage of regeneration, the seedlings, is negatively affected in the openings: germination and early survival of spruce seedlings is worst, due to damage by snow mould.

Contrasting demands of different life stages concerning snow and light conditions in the Lusiwald indicate that optimal regeneration starts under the early snow-free tree crown and proceeds in the snow-rich opening. As gap size is the governing factor of snow accumulation, future research has to relate occurrence of seedlings and saplings to the natural dynamics of the adult stage: dieback and gap dynamics in subalpine spruce forests.

MUFOMA is a Concerted Action (CA) within the FAIR-programme of the EU dealing with multifunctional forest management in mountainous regions. It brings together policy and silvicultural approaches and experiences from nine European countries with mountain forests.

The Concerted Action aims at improving the effectiveness and efficiency of silvicultural and forest policy means towards multifunctional management by

• assessing existing policy measures and silvicultural techniques used in different countries for the same objectives,
• identifying lacks in concepts as well as indicators for necessity or omission of silvicultural interventions,
• explaining success and failure of existing measures or programmes,
• developing recommendations for adaptation or improvement of silvicultural and forest policy measures or instruments.

The concept of the CA follows a ´bottom-up´ approach. The chosen method is a comparative analysis and an in-depth assessment of practical implementation, effects and outcome of existing means. Core of the CA are three workshops, where silvicultural and political situation at twelve ´demonstration stands´ will be analysed and assessed.

Topics covered at the 1st workshop:

• public perception of natural disturbances, silvicultural intervention, and risks
• national guidelines for mountain forest management
• controlling silvicultural measures in protective forests

Top 

Top

last update 5.10.99 by Walter Schoenenberger