Content uploaded by Urs-Beat Brändli
Author content
All content in this area was uploaded by Urs-Beat Brändli on Nov 17, 2018
Content may be subject to copyright.
A Swiss-Ukrainian Scientific Adventure
Inventory of the Largest
Primeval Beech Forest
in Europe
Editors: Brigitte Commarmot, Urs-Beat Brändli, Fedir Hamor, Vasyl Lavnyy
A Swiss-Ukrainian Scientific Adventure
Inventory of the Largest
Primeval Beech Forest
in Europe
Editors: Brigitte Commarmot, Urs-Beat Brändli, Fedir Hamor, Vasyl Lavnyy
Publisher:
Swiss Federal Institute for Forest, Snow and Landscape Research WSL, Birmensdorf
Ukrainian National Forestry University, L’viv
Carpathian Biosphere Reserve, Rakhiv
2013
Citation: Co mmar mo t , B.; Br ä n d l i , U.-B.; Hamo r , F.; l av n y y v. (eds) 2013: Inventory of the
Largest Primeval Beech Forest in Europe. A Swiss-Ukrainian Scientific Adventure.
Birmensdorf, Swiss Federal Research Institute WSL; L’viv, Ukrainian National Forestry
University; Rakhiv, Carpathian Biosphere Reserve. 69 pp.
Citation of chapters: Autors of chapter, 2013: Titel of chapter. In: Co mmar mo t , B.; Br ä n d l i ,
U.-B.; Hamo r , F.; l av n y y v. (eds) 2013: Inventory of the Largest Primeval Beech Forest in
Europe. A Swiss-Ukrainian Scientific Adventure. Birmensdorf, Swiss Federal Research
Institute WSL; L’viv, Ukrainian National Forestry University; Rakhiv, Carpathian Biosphere
Reserve. pp. x–y.
Editors:
Brigitte Commarmot, Swiss Federal Research Institute WSL, Birmensdorf, Switzerland.
e-mail: brigitte.commarmot@wsl.ch
Urs-Beat Brändli, Swiss Federal Research Institute WSL, Birmensdorf, Switzerland.
e-mail: urs-beat.braendli@wsl.ch
Fedir Hamor, Carpathian Biosphere Reserve, Rakhiv, Ukraine. e-mail: cbr-rakhiv@ukr.net
Vasyl Lavnyy, Ukrainian National Forestry University, L’viv, Ukraine. e-mail: lavnyy@gmail.com
Language revision and copy editing: Silvia Dingwall
Ukrainian translation: Vasyl Lavnyy
Maps: Martina Hobi and Meinrad Abegg
Layout: Jacqueline Annen, WSL
ISBN 978-3-905621-53-2
Copyright © 2013 by WSL, Birmensdorf
Printed in Ukraine by Colir PRO, L’viv
Available online: www.wsl.ch/publikationen/pdf/12494
Acknowledgements
We wish to thank the field teams and all the researchers and helpers of CBR, UNFU and WSL
listed in Appendix 1 for their dedicated work. Without them, the inventory would not have been
possible. We would also like to thank Peter Brang and Caroline Heiri for their useful comments
on this report, and Silvia Dingwall for improving the English. The project was supported by the
State Secretariat for Education, Research and Innovation SERI, Switzerland.
Cover photo: Uholka-Shyrokyi Luh protected massif. Photo M. Brüllhardt.
3
Foreword
Beech forests would, without human intervention, cover large parts of the nemoral zone in
Europe as climax vegetation. The beech (Fagus sylvatica) is one of the most successful tree
species in post-glacial vegetation history, with a distribution ranging from the mountains of the
Mediterranean region in the South to southern Scandinavia in the North, from the Atlantic Ocean
in the West to the eastern foothills of the Carpathians and the Crimean peninsula in the East.
Beech could potentially dominate most of the natural forest types within this extensive range,
from sea-level to the lowlands and up to the montane belt, where the temperate climate suits
them. In some areas it might even reach as far as the upper forest line. The proportion of beech
forests in the current forest cover of Europe has, however, been dramatically reduced through
millennia of land use. Untouched, old-growth beech forests mostly remain only in small patches
in a very few inaccessible areas or sites where the historical circumstances are in some way
special.
An international consensus on the preservation and sustainable management of forests was
reached by the time of the 1992 World Summit in Rio de Janeiro. In this context Europe has a
particular responsibility for the protection of the few remaining primeval and ancient beech for-
ests. The largest primeval beech forest in Europe is the Uholka-Shyrokyi Luh protected massif
in the Ukrainian Carpathians. This region has had a turbulent political history, which has left its
mark on land use, nature conservation and forest research. The first network of protected forest
areas was established in the region during the first decades of the 20th century under the Austria-
Hungarian and Czechoslovakian governments. Today, the primeval forest of Uholka-Shyrokyi
Luh is part of the Carpathian Biosphere Reserve (CBR), which was certified by UNESCO in
1992. In 2007, the forests of Uholka-Shyrokyi Luh, together with nine smaller primeval forest
remnants in Ukraine and Slovakia, were added under the name “Primeval Beech Forests of the
Carpathians” to the World Heritage List. This was the beginning of the current process of draw-
ing up a complete serial transnational list of the primeval and ancient beech forests in Europe to
be included in the World Heritage List.
Primeval forests are particularly interesting objects for forest research as they provide excel-
lent and necessary conditions for studying and understanding the ecosystem processes in
forests where no human intervention has occurred for a long time. The Carpathians are a kind
of locus classicus for virgin beech forest studies in Europe as they include large areas of prime-
val forest and have a long tradition of forest research. The Uholka-Shyrokyi Luh inventory project
is part of the Swiss-Ukrainian scientific cooperation on primeval forest research that started
fifteen years ago. It is an outstanding example of successful bilateral cooperation and scientific
teamwork. The inventory of 10 000 ha of primeval beech forest, which was carried out under
adventurous conditions in real wilderness, is unique. It provides for the first time information that
is really representative of a large area of primeval forest. The study is not only a valuable con-
tribution to our understanding of natural forest dynamics, but the resulting data also provide
much-needed reference values for nature conservation and natural forest management.
The research findings described here confirm the outstanding and universal value of the pri-
meval beech forests of the Carpathians. They also emphasize how the integrity of forest eco-
systems depends on the length of time they are able to develop naturally without disturbance,
and thus how urgent it is to protect the few remnants of primeval and ancient beech forests in
Europe. I hope this publication will encourage others to carry out similar investigations using
comparable methods in other remnants of primeval forest.
Prof. Dr. Hans D. Knapp, International Academy for Nature Conservation, Isle of Vilm
4Inventory of the Largest Primeval Beech Forest in Europe
Summar y
International conventions and resolutions on biological diversity, sustainable forest manage-
ment and climate change have led in recent decades to an increasing interest in having refer-
ence values from forests undisturbed by man. An outstanding example of such an undisturbed
forest is the primeval forest of Uholka-Shyrokyi Luh within the Carpathian Biosphere Reserve
(Ukraine). It is approximately 9000 ha (90 km2) in area and is thought to be the largest primeval
forest of almost pure European beech (Fagus sylvatica L.).
In 2010, the Swiss Federal Institute for Forest, Snow and Landscape Research WSL, the
Ukrainian National Forestry University UNFU and the Carpathian Biosphere Reserve CBR car-
ried out a sampling inventory of the Uholka-Shyrokyi Luh forest (survey perimeter 10282 ha) to
obtain representative data for the main forest parameters. Given the remoteness of the area,
long walking distances and difficult terrain, careful planning and organisation were required, as
well as the logistic support of the local forest service. The field work was carried out by six mixed
teams of Swiss and Ukrainian students and scientists, guided by three survey leaders from
Switzerland and Ukraine. Two teams together shared a leader and a cook, and lived in decentra-
lized camps, which were moved every week to minimize the walking needed to reach the sample
plots. The collaboration between the Ukrainians and Swiss worked very well and was enriching
for both sides. During the two-month sampling period, the teams assessed 314 sample plots
laid out on a systematic grid. All living and standing dead trees ≥6 cm DBH (diameter at 1.3 m
above ground) within the 500 m2circle plots were measured and assessed for features relevant
for biodiversity. Lying deadwood was assessed with line-intersect sampling (3 lines each 15 m
long per plot), and small trees (≥10 cm height and < 6 cm DBH) were surveyed on subplots
consisting of three concentric circles 5, 10 and 20 m2in area. The stand structure and any traces
of anthropogenic use were assessed on a circular interpretation area of 2500 m2around the
sample plot centre.
The primeval forest of Uholka-Shyrokyi Luh shows all the typical features of an old-growth
forest shaped by small-scale disturbances. The structure was mainly three-layered, and most of
the gaps encountered were not larger than the crown of a canopy tree. The growing stock per
ha was 582 (± 14) m3(mean ± standard error) and the deadwood volume 163 (± 8) m3. The ratio
of standing to lying deadwood was 1:5. The maximum DBH measured was 150 cm, and 10
trees per ha had a DBH of at least 80 cm. The density of habitat trees, i.e. living trees with fea-
tures such as cracks, holes, bark damage or similar that provide microhabitats, was 150 (± 8)
per ha (35% of the living trees). Of all the trees recorded, 97% were beech, although 14 other
tree species were identified. All species found in the tree population ≥6 cm DBH were also
present in the regeneration.
Traces of human presence were encountered on 19% of the assessed plots (interpretation
areas), mainly in the buffer and regulated protection zone of the protected massif. Most of these
traces do not affect the integrity and pristine character of the forest. Nevertheless, they imply a
certain pressure exerted from the nearby settlements and from the mountain pastures.
The data obtained provide good reference values for old-growth beech forests and a valuable
basis for more detailed analyses and comparisons with other old-growth and managed forests.
The inventory was carried out and documented in a replicable way, and can thus be repeated if
desired. The plots may also be used for other non-destructive studies, e.g. on fungi.
Keywords
Fagus sylvatica, virgin forest, old-growth forest, forest structure, deadwood, reference values,
Carpathian Biosphere Reserve
5
Резюме
ІНВЕНТАРИЗАЦІЯ НАЙБІЛЬШОГО БУКОВОГО ПРАЛІСУ ЄВРОПИ
Швейцарсько-українські наукові висліди
Міжнародні конвенції і резолюції щодо біологічного різноманіття,ведення сталого лісового
господарства та змін клімату спричинили в останні десятиліття все більший інтерес до вивчення
пралісів.Яскравим прикладом такого незачепленого людською діяльністю лісу є Угольсько-
Широколужанський праліс в Карпатському біосферному заповіднику (Україна). Він займає
площу близько 9000 га (90 км2)і вважається найбільшим пралісом,що сформувався з майже
чистого бука лісового (Fagus sylvatica L.).
У2010 році Швейцарський федеральний інститут досліджень лісу,снігу та ландшафтів (WSL),
Національний лісотехнічний університет України (НЛТУ України)і Карпатський біосферний
заповідник (КБЗ)організуваливибіркову інвентаризацію Угольсько-Широколужанського пралісу
(площа за периметром 10282 га), щоб отриматирепрезентативні дані про його основні параметри.
Враховуючи віддаленість району,довгі піші переходи і складний рельєф місцевості,велику
увагу було приділено ретельному плануванню та організації інвентаризації із залученням для
допомоги місцевих працівників КБЗ.Польові роботи проводилися шістьма змішаними групами з
швейцарських та українських студентів під керівництвом трьох польових менеджерів з Швейцарії
та України.Дві групи,кожна зі своїм менеджером і кухарем,жили у тимчасовому таборі,який
переміщувався щотижня у інше місце,щоб звести до мінімуму необхідні переходи до місць
закладання пробних площ.Між українцями і швейцарцями була хороша співпраця,що приносила
користь для обох сторін.Протягом двох місяців інвентаризації таксаційні групи заклали 314
кругових пробних площ на основі систематичної сітки.На цих пробних площах розміром 500 м2
були заміряні всі живі й мертві стоячі дерева з діаметром ≥ 6см на висоті 1,3 м над поверхнею
ґрунту для визначення лісівничо-таксаційних показників деревостану і біорізноманіття.Лежача
мертва деревина вимірювалася на лінійних трансектах (3 лінії на пробі довжиною 15 м кожна),
а невеликі дерева (≥10 см заввишки і <6см за діаметром на висоті 1,3 м)були обстежені на
трьох концентричних кругах площею 5, 10 і20 м2.Структура деревостану і можливі сліди
антропогенного впливу оцінювалися на кругові розміром 2500 м2довкола центру пробної площі.
Угольсько-Широколужанський праліс має всі характерні риси старовікового лісу з наявністю
невеликих прогалин у наметі.Структура деревостану в основному триярусна,а більшість з
прогалин за площею не перевищує розмір крони пануючих дерев.Запас ростучих дерев
становить 582 (±14) м3на гектарі (середнє значення ±стандартна помилка), а об’єм мертвої
деревини – 163 (± 8) м3/га.Співвідношення між стоячою та лежачою мертвою деревиною було
1:5. Максимальний діаметр стовбура дерев на висоті 1,3 м дорівнював 150 см,а10 дерев на
гектарі малидіаметр на висоті 1,3 м щонайменше 80 см.Щільність дерев-біотопів,тобто дерев з
тріщинами,дуплами,пошкодженнями кори або іншим середовищем існування живих організмів
становила 150 (± 8) шт./га (35% живих дерев). Серед деревних порід 97% всіх дерев становив
бук лісовий,хоча загалом у пралісі були виявлені дерева 14 різних видів.Всі види дерев з
діаметром на висоті 1,3 м≥6см також були присутні у складі самосіву і підросту.
Сліди антропогенного впливу були виявлені на 19% закладених пробних площ,в основному
в зоні антропогенних ландшафтів та в буферній зоні заповідного масиву.Більшість з цих слідів
не впливає на цілісність і незайманість пралісу.Однак вони свідчать про певний тиск на праліс
з навколишніх населених пунктів і високогірних пасовищ.
Отримані дані інвентаризації є добрим порівнянням для інших старовікових лісів бука та
цінною основою для більш детального аналізу та порівняння з іншими старовіковими і
господарськими лісами.Інвентаризація була проведена і задокументована таким чином,щоб
можна було знайти всі пробні площі у майбутньому.Тому при необхідності вона може бути
повторена через певний проміжок часу.Крім того,закладені пробні площі можуть бути
використані для інших неруйнівних досліджень,наприклад,для вивчення грибів.
Ключові слова
Fagus sylvatica, праліс,структура деревостану,еталонне значення,мертва деревина,
міжкультурне співробітництво,Карпатський біосферний заповідник.
6Inventory of the Largest Primeval Beech Forest in Europe
Abbreviations
a.s.l. above sea level
BFH Berne University of Applied Sciences
CBR Carpathian Biosphere Reserve
DBH diameter at breast height (measured 1.3 m above ground)
D7 upper stem diameter (measured 7 m above ground)
ETHZ Swiss Federal Institute of Technology Zürich
ha hectare (= 10000 m2or 0.01 km2)
HAFL School of Agricultural, Forest and Food Sciences
N number (e.g. number of trees or plots)
NFI National Forest Inventory
SE standard error
UNESCO United Nations Educational, Scientific and Cultural Organization
UNFU Ukrainian National Forestry University
WSL Swiss Federal Institute for Forest, Snow and Landscape Research
ZHAW Zurich University of Applied Sciences, School of Life Sciences and Facility
Management
Not e o n t he terms “ p rim eval” , “ virg in” an d “ o ld-g row th” f ores t:
We use the terms “primeval” and “virgin forest” as synonyms for a “forest undisturbed by
man”, i.e. where there has been no known significant human intervention, or where the last
significant human intervention was so long ago that the natural species composition and
processes have re-established (MCPFE 2007). In contrast, the term “old-growth forest” may
include forests previously managed but which have been left to develop naturally. They thus
show some old-growth characteristics, such as mixed tree ages and development phases
with senescent and dead trees, as well as standing and lying deadwood in all decay stages.
MCPFE Ministerial Conference on the Protection of Forests in Europe 2007: State of
Europe’s forests 2007. The MCPFE report on sustainable forest management in Europe.
MCPFE Liaison Unit Warsaw, Poland.
7
Contents
Foreword 3
Summary 4
Резюме 5
Abbreviations 6
1 Introduction 9
2 The Uholka-Shyrokyi Luh protected massif – an overview 13
2.1 Location and site conditions 13
2.2 Main plant associations 14
2.3 History of land use 15
2.4 Management of the protected massifs 17
3 The inventory – aims, methods and sampling design 19
3.1 Aims of the inventory 19
3.2 Inventory method and sampling design 19
4 Planning and management of the field survey 27
4.1 Pilot inventory 2009 27
4.2 Main inventory 2010 27
5 Data management and statistical evaluation 35
5.1 Storage and handling of the data 35
5.2 Evaluation routines 35
5.3 Volume estimation 37
6 Main results 41
6.1 Presentation and statistical interpretation of the results 41
6.2 Topography and anthropogenic traces 42
6.3 Tree species diversity and forest structure 45
6.4 Habitat structures 50
6.5 Conclusions 56
7 Outlook 59
Appendix
Appendix 1: Project organisation and division of tasks 61
Appendix 2: Survey equipment 62
Appendix 3: Distribution of anthropogenic traces 63
Appendix 4: List of tree species assessed 66
Appendix 5: Distribution of tree species 67
9
The Uholka-Shyrokyj Luh protected massif – an overview
European forests have been used and altered by
humans for thousands of years, with the most rapid
changes occurring during the Middle Ages (Küs t e r
1998). The expansion of human settlements not only
led to the forest area diminishing fast, but also to more
intensive use of the remaining forest. Wood continued
to be the main resource for heating, energy and con-
struction far into the 19th century, and the increasing
demand was met by exploiting and clear-cutting forests
even in remote areas. Forests were also used for graz-
ing, leaves were cut as fodder and litter was collected
as bedding for livestock and humans.
Only scattered relicts of primeval forest, also known
as virgin or primary forest, have survived in mountain-
ous areas, mainly in the geographic regions of the
Carpathians, the Balkans and the Alps (leiBUn d g Ut
1982; may e r et al. 1989; Ko r p el ’ 1995; p rů š a1985;
dia Ci 1999; g iUr g iU et al. 2001; Br ä n d l i a n d dowH-
a n y t s CH 2003; Ha mo r et al. 2008). These virgin forest
relicts have a high value for biodiversity conservation
(Pail l et et al. 2009), but they are also unique objects
for ecological and forest research as they provide
unique opportunities for studying the complex natural
structures, processes and ecosystem functions of for-
ests undisturbed by man.
The value of such old-growth forests was already
recognized in the 19th century, when the first forest
reserves were established in Poland and the Czech
Republic (Zie l o n y 1999; Ho r t et al. 1999). Since then,
most European countries have protected and set aside
near-natural forests as reserves where old-growth
structures can, in the long run, develop again. It may
take centuries, however, before such formerly man-
aged forests become like virgin forests again and pro-
vide the same kind of ecosystem functions as the
long-lost primeval forests.
In recent decades, various international conventions
and resolutions on biological diversity, sustainable forest
management and climate change have been passed
and have led to an increasing demand for reference
values from undisturbed forests. The Ukrainian Car-
pathians harbour some of the largest remnants of
primeval forest of European beech (Fagus sylvatica L.).
If humans had not interfered, beech forests would today
cover extensive areas of Central Europe, from the Alps
“The endless horizons, unspoilt nature, clear streams
and the beech trees standing guard over the breath
taking natural world were well complemented by the
warmth and friendship that the Ukrainians emanated.
Perhaps it is the philosophy of life – ‘everything in due
course’ – that has allowed the forests to last so long.”
Daniel Oertig, student HAFL, Switzerland
1 Introduction
Brigitte Commarmot, Urs Beat Brändli, Fedir Hamor, Vasyl Lavnyy
Typical appearance of the primeval beech forest
of Uholka-Shyrokyi Luh, with uneven-sized beech
trees, lying deadwood, regeneration, and a
multilayered structure. Photo V. Chumak.
10 Inventory of the Largest Primeval Beech Forest in Europe
within the CBR, 8800 ha of which are thought to be pri-
meval forest (Br ä n d l i and d o w Ha n y t s CH 2003; Br ä n d l i
et al. 2008; sUKHar y UK 2005). To our knowledge, this is
the first systematic inventory of such a large virgin forest
area in Europe. In this report, we describe the sampling
design and the parameters assessed, the planning and
organisation of the field work, and the management and
analysis of the data collected in the terrestrial survey.
We also present findings about basic forest characteris-
tics, habitat structures, site factors and anthropogenic
traces. The report is intended to give an overview of the
inventory methods and basic calculations we used and
to serve as a reference and basis for more thorough
analyses and inventories in future. Not included in this
report are detailed structural analyses and comparisons
with other reserves or managed forests, as they are the
topics of separate scientific papers.
The report is intended for scientists working with
data from this inventory, forest ecologists, biologists,
conservationists and other people interested in refe-
rence data from virgin forests. Researchers planning to
carry out a similar inventory should also find it useful.
References
Br ä n d l i, U.-B.; d o w Han y t s CH, J., 2003: Urwälder im Zentrum
Europas. Ein Naturführer durch das Karpaten-Biosphären-
reservat in der Ukraine. Birmensdorf, Eidg. Forschungsan-
stalt WSL; Rachiw, Karpaten-Biosphärenreservat. Bern,
Stuttgart, Wien, Haupt. 192 pp.
Br ä n d l i, U.-B.; d o v Ha n y CH, J.; Co mmar mo t , B., 2008: Virgin
Forest of Uholka. Nature Guide to the Largest Virgin Beech
Forest of Europe. A UNESCO World Heritage Site. Bir-
mensdorf, Swiss Federal Research Institute WSL, Rakhiv,
Carpathian Biosphere Reserve. 24 pp.
Co mmar mo t , B.; Ba CHo Fe n , H.; BUn d Zia K, y.; Bü r g i, a .;
ramP, B.; s HPar y K, y.; s UKHar iUK, d.; v it e r , r .; Zi n g g ,
a., 2005: Structures of virgin and managed beech forests
in Uholka (Ukraine) and Sihlwald (Switzerland): a compar-
ative study. For. Snow Landsc. Res. 79, 1/2: 45–56.
Co mmar mo t , B.; Br a n g , P.; Br än d l i , U.-B., 2008: Monitoring
natural forest dynamics in Switzerland and Ukraine. In
HAMOR, F. (ed) Rozvytok zapovidnoï spravy v Ukraïni i
formuvannia pan’evropeis’koï ekolohichnoï merezhi. Mate-
rialy mizhnarodnoï konferentsiï. Ukraïna, Rakhiv, 1–13
lystopada 2008 r. Rakhiv, Carpathian Biosphere Reserve,
223–228.
Co mmar mo t , B.; Ha mo r , F. (eds) 2005: Natural Forests in the
Temperate Zone of Europe – Values and Utilisation. Confer-
ence 13–17 October 2003, Mukachevo, Ukraine. Proceed-
ings. Rakhiv, Carpathian Biosphere Reserve; Birmensdorf,
Swiss Federal Research Institute WSL. 485 pp.
dia Ci, J. (ed) 1999: Virgin forests and forest reserves in Cen-
tral and East European countries. History, present status
and future development. Department of Forestry and
Renewable Forest Resources, Biotechnical Faculty, Lju-
bljana. 171 pp.
Fig. 1.1. Sign with information about the 10 ha research plot of CBR
and WSL in the Uholka administrative unit. Photo U.-B. Brändli.
across the lower mountain ranges down to the low-
lands. The Carpathian relicts of primeval beech forest
are therefore of special interest for research. In 1998,
WSL initiated a co-operation with the Carpathian Bio-
sphere Reserve (CBR) and other Ukrainian institutes to
study the structure and diversity of virgin forests. In
2000, a 10 ha research plot (Fig. 1.1) was established
in the primeval forest of Uholka (Co mmar mo t et al.
2005). Since then, detailed measurements have been
carried out every five years to follow the natural forest
dynamics. This research plot allows insights into the
small-scale spatial structures and their dynamics, and
also into the interactions between individual trees and
different species, but it cannot be considered represen-
tative of the approx. 14000 ha of primeval beech forest
still preserved in the Krasna Massif of the Ukrainian
Carpathians (Hamo r et al. 2008). Large-scale syste-
matic (random) sampling was therefore needed to
obtain representative data from these forests.
At the conference “Natural Forests in the Temperate
Zone of Europe – Values and Utilisation”, which was
jointly organised by the WSL and the CBR in 2003
(Co mmar mo t and Hamo r 2005), the idea was born to
carry out such a large-scale virgin forest inventory. Dur-
ing the next few years, this idea was further developed
and discussed with possible partners and sponsors in
Ukraine and Switzerland, and presented at a confe-
rence at CBR in 2008 (Co mmar mo t et al. 2008). With
financial support from the State Secretariat for Educa-
tion, Research and Innovation SERI, Switzerland, it
was at last possible to start the project. In 2010, WSL,
CBR and the Ukrainian National Forestry University
(UNFU) carried out a sampling inventory over the whole
forest area in the Uholka-Shyrokyi Luh protected massif
11
Introduction
Giu r Giu , V.; Do n it ă, n .; Bâ n Diu , C.; r a Du , S.; Cenuş ă, r .;
DiSSe SCu , r .; St o i Cu l eSCu , C.; Bir iş, i.-a., 2001: Les
forêts vierges de Roumanie. asbl Forêt wallone, Louvain-
la-Neuve, Belgique. 206 pp.
Hamo r , F.; d o v Ha n y CH, y.; Po Ky n CHer ed a , v.; sUKHar y UK,
d .; BUn d Zya K, y.; Be r Kel a, y.; v o l o s HCHUK, m.;
Ho d o v a n et s , B.; Ka Bal , m., 2008: Virgin forests of Trans-
carpathia. Inventory and management. Carpathian Bio-
sphere Reserve, Royal Dutch Society for Nature Conser-
vation, Rakhiv. 79 pp.
Ho r t , l .; t eSař, V.; VršKa, t ., 1999: Forest reserve research
network. The Czech Republic country report. In: Diaci, J.
(ed). Virgin forests and forest reserves in Central and East
European countries: history, present status and future
development. Proceedings of the invited lecturers’ reports
presented at the COST E4 manangement committee and
working group meeting in Ljubljana, Slovenia; Ljubljana,
25–28. April 1998. Department of Forestry and Renewable
Forest Resources, Biotechnical Faculty, Ljubljana. 25–44.
Ko r p el ’, š., 1995: Die Urwälder der Westkarpaten. G.
Fischer, Stuttgart, Jena, New York. 310 pp.
Kü s t e r , H., 1998: Geschichte des Waldes. Von der Urzeit bis
zur Gegenwart. Beck, München. 266 pp.
le iBUn d g Ut , H., 1982: Europäische Urwälder der Bergstufe.
Verlag Paul Haupt. Bern. 308 pp.
may e r , H.; ZUKr i g l , K.; s CHr e mPF, w .; s CHl a g er g ., 1989:
Urwaldreste, Naturwaldreservate und schützenswerte
Naturwälder in Österreich. 2. Auflage. Universität für
Bodenkultur, Institut für Waldbau, Wien. 970 pp.
Pail l et , y.; Be r g è s , l .; HJä l t é n , J.; Ód o r , P.; av o n , C.;
Ber nHar d -r ö mer ma n n , m.; BiJl s ma , r .-J.; d eBr Uy n , l .;
FUHr , m.; g r a n d in , U.; Ka n Ka, r .; l Und i n , l .; l Uq Ue, s .;
mag Ur a , t .; ma t es a n Z, s .; mé s Zá r o s , i.; s e Bas t i á, m.-t .;
sCHmid t , w .; s t a n d o v á r , t .; t Ót Hmé r és Z, B.; Uo t il a , a.;
val l ad e r e s , F.; v el l aK, K.; v i r t a n e n , r ., 2009: Biodiver-
sity differences between managed and unmanaged for-
ests: Meta-analysis of species richness in Europe. Con-
serv. Biol. 24: 101–112.
prů š a, e., 1985: Die böhmischen und mährischen Urwälder
– ihre Struktur und Ökologie. Academia, Praha. 578 pp.
sUKHa r y UK, D.D. 2005: Pryrodni lisy i pralisy Karpats’koho
biosfernoho zapovidnyka. Ïkh znachennia ta zakhody
shchodo zberezhennia. In: Bioriznomanittia Ukraïns’kykh
Karpat. Materialy nauk. konf., prysviachenoï 50 richchiu
Karpats’koho vysokohirnoho statsionaru L’vivs’koho
natsional’noho universytetu im. Ivana Franka (30 lypnia –
3 serpnia 2005 roku). L’viv, 182–186.
Zie l o n y , r ., 1999: Natural forests and forests protected by
law in Poland. In: d ia Ci, J. (ed) Virgin forests and forest
reserves in Central and East European countries: history,
present status and future development. Proceedings of the
invited lecturers’ reports presented at the COST E4
manangement committee and working group meeting in
Ljubljana, Slovenia; Ljubljana, 25–28. April 1998. Depart-
ment of Forestry and Renewable Forest Resources,
Biotechnical Faculty, Ljubljana. 45–66.
“The work in the forest was fascinating, physically
demanding, but very satisfying. The view into the
forest was enthralling, especially in the more inacces
sible areas where there was a lot of deadwood and an
untouched natural environment. The work was
physically challenging due to the topography, with very
steep slopes and gullies, and because of the time it
took to walk there.”
Jonas Stillhard, student ZHAW, Switzerland
View from the “polonina” (mountain pasture)
on top of Menchul looking towards the South-
West. Photo L. Mini.
2 The Uholka-Shyrokyi Luh protected
massif – an overview
Fedir Hamor, Urs Beat Brändli
2.1 Location and site conditions
The Uholka-Shyrokyi Luh massif is one of the eight
protected massifs1united in the CBR. It is situated in
central Transcarpathia, the south-westernmost region
of Ukraine (Fig. 2.1). It belongs to the beech forest belt
on the southern slopes of the Polonyny Carpathians
(Krasna mountain range), and comprises the upper
basin of the Luzhanka, Velyka Uholka and Mala Uholka
rivers at altitudes ranging from 400 to 1400 m a.s.l.
(Fig. 2.2). The Uholka-Shyrokyi Luh massif covers an
area of around 16000 ha, of which 10400 ha are under
direct management of the CBR, while the other 5600
ha are managed by the state forest enterprises. The
massif is divided into two parts of similar size: the
administrative unit of Uholka, adjacent to the villages
Velyka Uholka and Mala Uholka, and the unit of Shy-
rokyi Luh, 12 km north of the Shyrokyi Luh village.
The relief of the Uholka-Shyrokyi Luh massif is very
fragmented and divided by several narrow valleys
formed by mountain streams. The massif consists
mainly of flysch formations of the Cretaceous and
Paleogene periods, with Jurassic limestone, calcareous
conglomerates, marls and sandstone. A unique land-
scape feature of the Uholka part of the massif is the
limestone ridge, which is part of the great tertiary lime-
stone range that stretches from the West to the South
Carpathians. Limestone rocks form up to 60 m high
cliffs and contain numerous karst caves, the longest of
which is over 1 km long. The Shyrokyi Luh area is
practically free of limestone. Clastic sedimentary rocks,
such as alevrolits (siltstone), sandstone and conglom-
erates, dominate, sometimes forming high cliffs. The
topsoil of the massif consists of acidic brown soils of
variable granulometric composition and depth.
The Uholka-Shyrokyi Luh massif lies in the Atlantic-
continental climatic region of the Ukrainian Carpathians,
with inflowing Atlantic air masses. The mean annual
temperature measured at the meteorological station of
CBR in Uholka at 430 m altitude is 7.7°C, the mean July
temperature 17.9°C, and the mean January tempera-
ture –2.7°C (averages for the years 1990–2010). The
1The term “massif” is used for the CBR protected areas and
does not correspond with the geological term.
14 Inventory of the Largest Primeval Beech Forest in Europe
absolute minimum measured was –25.1°C (on January
13, 1987), and the maximum 35.3°C (on July 15, 2001).
The mean annual precipitation from 1980–2010 was
1134 mm, 50–60 % of which fell during the vegetation
period (May to October). The snow cover is usually
40–60 cm, in some places reaching as much as 100 cm.
2.2 Main plant associations
Over 96% of the Uholka-Shyrokyi Luh protected massif
is covered with forests. Natural (old-growth) forests make
up 9980 ha, 8800 ha of which are primeval forests
(Br ä n d l i and d o w Han y t s CH 2003; s UKHa r y UK 2005).
Most of these are more-or-less pure beech forests.
Beech forms a continuous forest belt from 400 m a.s.l. up
to the upper timber line (1250–1350 m). The characteris-
tic herbaceous plants of the massif are ephemeral and
early spring species, such as Anemone nemorosa, A.
ranunculoides, Leucojum vernum, Dentaria bulbifera, D.
glandulosa, Galantus nivalis, Isopirum thalictroides, Cory
dalis solida, C. hallery and Heleborus purpurascens.
The forest vegetation has been classified into 10
forest formations (alliances) and 77 plant associations
(st o JKo et al. 1982; classification according to s Hel iaH-
so s o nKo 1991). Moist and humid pure beech forests
growing on mega- and mesotrophic soils are the most
widespread forest formations. Over 70% of the primeval
Fig. 2.2. Valley of the Luzhanka river. Photo R. Iseli.
Fig. 2.1. Location of the Uholka-Shyrokyi Luh massif. Source:
ArcGIS Map Service http://www.arcgis.com, “World Shaded
Relief” and “World Imagery”.
!
!
!
!
!
!
!
!
Kalyny
Drago vo
Vil' shany
Kolo chava
Shyro kyi Luh
Velyka Uh olka
Khust
Dolgoye
Uholka-Shyrok yi
Luh M assif
0 3 6 9 12
km
Romania
Hungary
Slovakia
Poland
UKRAINE
Uholka-Shyrok yi
Luh M assif
0 50 100 150 200
km
Uholka-Shyrok yi
Luh M assif
°
°
15
The Uholka-Shyrokyj Luh protected massif – an overview
Fig. 2.3. Cardamine glanduligera, characteristic species of the
Fagetum dentariosum. Photo B. Commarmot.
forest sites are Fagetum dentariosum (Fig. 2.3) or
Fagetum asperulosum, with the latter the most pro-
ductive of the Ukrainian Carpathian beech formations.
On sites where beech is less competitive, mixed com-
munities are formed, such as Querceto petraeae
Fagetum, Carpineto Fagetum or Acereto Fagetum.
In the administrative unit of Shyrokyi Luh, where the
climate is somewhat colder than in Uholka, the meso-
trophic associations Abieto Fagetum and Abieto Piceeto
Fagetum also occur in a few areas at high altitudes, as
do the humid associations Fageto Abietum and Piceeto
Fageto Abietum on the rocky slopes of the Tatry and
Yalynkovatyi areas. The monodominant association
Рiceеtum myrtilosum is preserved in several small
islands as well. In addition to the dominant beech for-
ests, associations such as Fageto Aceretum pseudo
platani, Ulmeto Fraxineto excelsioris Aceretum pseu
doplatani, Fraxinum excelsioris or Betuletum pendulae
may also be found.
Due to the region’s specific ecological conditions,
many relict and endemic plant species, as well as rare
associations, such as the Caprineto Fagetum spiraeoso
mercurialidosum with a significant share of thermophile
species, are preserved on the limestone cliffs in the
Uholka unit of the massif. Ulmeto Fraxineto Aceretum
occurs at the foot of the cliffs, with relict species such
as Lunaria rediviva and Phylitis scolopendrium. Many
rare associations are found in the Grebin and Mala
Kopytsia areas. These are the only sites in Ukraine
where the Fageto Tilietum platyphyllae sesleriosum
heuflerianae is protected (Fig. 2.4). Fagetum taxoso
hederosum, Fagetum taxoso sesleriosum and Fage
tum taxoso myrtilosum forest sites can be found spo-
radically distributed on limestone slopes. Uholka is the
second largest area in Ukraine containing the Tertiary
relict species Taxus baccata, and the only site in the
East Carpathians where Juniperus sabina is found.
2.3 History of land use
Transcarpathia’s borders have changed so frequently
over the centuries that industrialisation and intensive for-
est use began relatively late. Much of the mountain forest
remained untouched until the 18th century. Some forests
were not used for timber until even later because they
were kept as hunting grounds for the aristocracy or
because they contained no suitable water stretches for
rafting or other forms of transport for getting the timber
out of the forest. The remote beech forests in the Uholka-
Shyrokyi Luh area are one such example (Br ä n d l i et al.
2008). The intense livestock pasturing practised for cen-
turies on the mountain meadows had, however, a nega-
tive impact on the forest ecosystems and depressed the
upper forest line by 100–200 m in altitude in some places
(Fig. 2.5). Thus, the present forest line seems to be man-
Fig. 2.4. Tilia platyphyllos Scop. on a limestone cliff in Uholka.
Photo L. Mini.
16 Inventory of the Largest Primeval Beech Forest in Europe
Fig. 2.5. The upper forest line is 100–200 m lower than it would be without the intense livestock pasturing practised for centuries on the
mountain meadows. Photo M. Brüllhardt.
Fig. 2.6. Traditional subsistence farming in the anthropogenic landscape zone of the Uholka-Shyrokyi Luh protected massif. Photo O.
Nadyeina.
17
The Uholka-Shyrokyj Luh protected massif – an overview
made. Today, the Uholka-Shyrokyi Luh massif is still
interlaced with footpaths leading from the villages close-
by through the forests up to the mountain meadows.
In 1936, the first forest reserve 1024.5 ha in area
was established in the Luzhanka river basin, thanks to
the efforts of the prominent Czech botanists A. Zlatnik
and A. Gilitzer, to conserve the primeval beech forests
and the relict Norway spruce (Picea abies Karst) associ-
ation. After the Transcarpathian region was separated
from Czechoslovakia and became part of the Soviet
Union, the Uholka reserve was established in 1958 and
the Shyrokyi Luh in 1964. They were incorporated in
the newly founded Carpathian Reserve in 1968 and
1979, respectively.
In 1992 the Carpathian Reserve was designated a
UNESCO Biosphere Reserve, the Carpathian Biosphere
Reserve CBR, comprising eight geographically sepa-
rate massifs and a total area of 53630 ha, 14600 ha of
which are primeval forests. Today, most of these forests
are included in the UNESCO World Heritage Site
“Primeval Beech Forests of the Carpathians and the
Ancient Beech Forests of Germany”.
2.4 Management of the protected
massifs
CBR has a 10-year management plan for all the mas-
sifs, which includes a detailed forest description with
data on forest taxation, maps and the division of the
massifs into functional zones (Table 2.1). The plan
defines the level of logistic, financial and staff support,
specifies conservation measures, defines the tolerated
pressure on ecosystems and restricts the use of natural
areas. It also specifies the sanitary measures (includ-
ing salvage logging) to be applied for particular forest
diseases and insect infestations, and outlines how
infrastructure should be developed.
Any activity that could have a negative impact on the
local natural and historical-cultural features is forbidden
within the protected massifs of CBR. The regime for
using the natural areas in the functional zones corre-
sponds to the Seville Strategy (UNESCO 1996) and the
Madrid Action Plan (UNESCO/MAB 2008) for biosphere
reserves. All potential human impact is restricted within
the core zone, where natural processes occur without
any intervention. The buffer zone allows some conser-
vation measures for restoring natural ecosystems and
protecting them from harmful effects. The anthropogenic
landscapes zone is the area where traditional land man-
agement is practiced (Fig. 2.6). The zone with regulated
protection comprises two 50 m wide buffers on both
sides of the main corridors leading through the core
zone, where measures to maintain the corridors, e.g. to
clear away fallen trees, are allowed. Today, every year
approximately 1400 m3of timber are logged in the
anthropogenic landscape zone and partially within the
buffer zone of the Uholka-Shyrokyi Luh massif, 25 tons
of hay are mown, and 700 cattle pastured (average for
the years 2007–2011).
References
Br ä n d l i, U.-B.; d o w Ha n y t s CH, J., 2003: Urwälder im Zentrum
Europas. Ein Naturführer durch das Karpaten-Biosphären-
reservat in der Ukraine. Birmensdorf, Eidg. Forschungsanstalt
WSL; Rachiw, Karpaten-Biosphärenreservat. Bern, Stuttgart,
Wien, Haupt. 192 pp.
Br ä n d l i, U.-B.; d o v Ha n y CH, J.; Co mmar mo t , B., 2008: Virgin
Forest of Uholka. Nature Guide to the Largest Virgin Beech
Forest of Europe. A UNESCO World Heritage Site. Bir-
mensdorf, Swiss Federal Research Institute WSL, Rakhiv,
Carpathian Biosphere Reserve. 24 pp.
Sh el iah -So Son k o , Yu.R., 1991: Prodomus roslynnosti Ukraïny.
Kyiv, Naukova Dumka. 272 pp.
st o JKo , s .m.; t a s Jen Ke v y CH, l .o .; mil Kin a , l .i.; mal y n o v s Ky J,
K.a.; t r e t JaK, P.r .; man Ko , m.P.; BeZUs Ko , l .g .; t s Ur y K,
Je.i.; mel n y K, a.s ., 1982: Flora i roslynnist’ Karpats’koho
Zapovidnyka. Kyiv, Naukova Dumka. 222 pp.
sUKHar y UK, D.D. 2005: Pryrodni lisy i pralisy Karpats’koho biosfer-
noho zapovidnyka. Ïkh znachennia ta zakhody shchodo zbere-
zhennia. In: Bioriznomanittia Ukraïns’kykh Karpat. Materialy
nauk. konf., prysviachenoï 50 richchiu Karpats’koho vysoko-
hirnoho statsionaru L’vivs’koho natsional’noho univer-sytetu im.
Ivana Franka (30 lypnia – 3 serpnia 2005 roku). L’viv, 182–186.
UNESCO, 1996: Biosphere reserves: The SevilleStrategy and the
Statutory Framework of the World Network. UNESCO, Paris
UNESCO/MAB 2008. Madrid Action Plan for Biosphere Re-
serves (2008–2013). http://unesdoc.unesco.org/images/0016/
001633/163301e.pdf
Table 2.1. The functional zoning of the Uholka-Shyrokyi Luh massif under CBR management (Proekt orhanizatsiï terytoriï Karpats’koho
biosfernoho zapovidnyka 2002).
Administrative unit Total area
[ha]
Area according to functional zones [ha]
Core zone Regulated
protection zone
Buffer zone Anthropogenic
landscape zone
Shyrokyi Luh 5654 3871 175 1436 172
Uholka 4729 3246 248 1065 170
Total 10383 7117 423 2501 342
“It was the first time I ever worked with foreign people.
I liked collaborating with the different members of the
team. It was interesting to compare the Swiss
inventory method we used in the Uholka Shyrokyi Luh
massif with the Ukrainian one. We got to know a lot of
new parameters. The instruments are very good and
well developed as they are small, lightweight and use
only a minimal amount of energy. We will introduce
our students to these methods and technologies. I was
surprised how well everything was organised. We are
not used to doing such intensive planning work. It was
a good experience for all participants.”
Serhiy Gavryliuk, assistant professor UNFU, Ukraine
Igor Cherniuk marking the position of a sample
plot centre. The sample plots are randomly
distributed over the study area. Photo M. Hobi.
3 The inventory – aims, methods and sampling design
Adrian Lanz, Urs Beat Brändli, Brigitte Commarmot, Christian Ginzler
3.1 Aims of the inventory
The general aim of the forest inventory was to obtain
representative estimates of the state of the forest of
Uholka-Shyrokyi Luh, which is thought to be the largest
relict of primeval European beech forest. More specif-
ically, the goal was to collect information about the
number of trees, tree dimensions, growing stock, vol-
umes of standing and lying deadwood, forest structures
and regeneration density, tree species diversity, fre-
quency of trees with microhabitats (habitat trees), and
tree ages. An additional aim was to record traces of
recent and former anthropogenic use or activities to
assess the integrity of the forest and identify potential
threats. The goal was to collect data for qualitative and
quantitative descriptions and scientific analyses of the
primeval forest, which could serve as a reference for
comparisons with managed beech forests and forest
reserves to enable the development of appropriate
approaches to biodiversity conservation and forest
management. Furthermore, such an inventory would
provide a description of the initial state of the forest as
a basis for monitoring its future development.
The inventory was planned as a joint project of WSL,
UNFU and CBR, which would provide training and work
experience for Ukrainian and Swiss students and foster
Swiss-Ukrainian collaboration and scientific and cul-
tural exchange.
3.2 Inventory method and sampling
design
Obtaining reliable and objective information about
various characteristics of the forests in the protected
massif required not only a randomised (not purposive)
sample so that generally acceptable statistical infer-
ences could be made, but also a broad set of measure-
ments and observations to allow conclusions about
many different aspects of the forests to be drawn. A
ground-based survey (terrestrial inventory) with field
data collection on sampling plots was therefore chosen.
Here the sampling plots are randomly distributed over
the entire study area.
20 Inventory of the Largest Primeval Beech Forest in Europe
Requirements and materials available for planning
According to the aims of the survey, the inventory
should fulfil the following requirements: it should be
compatible with large area inventory data, such as the
Ukrainian Forest Inventory (Ukrderzhlisproekt 2006;
State Forestry Committee of Ukraine 2010) and the
Swiss National Forest Inventory NFI (Br a s s el and
lis CHKe 2001; Br ä n d l i 2010), and with the ongoing
monitoring program in Swiss forest reserves (Br a n g
et al. 2008, 2011). Thus, existing methods and defini-
tions (Kel l er 2005, 2011; t in n e r et al. 2010) should
be used and, where necessary, adapted to the local
conditions. The inventory should concentrate on the
core and buffer zones of the reserve. The data collec-
tion should be strictly non-destructive and based on
probability sampling principles (ma n d al l a Z 2008).
Sampling plots should be installed to allow re-measure-
ment later so that future changes can be efficiently
monitored. This involved establishing permanent plots
to ensure the inventory is continuous. The field survey
should not take more than two months because of the
students’ summer holidays and should be feasible with
a maximum of 6 field teams. These limitations were
imposed to be compatible with resource availability in
general and the local capacity for accommodation of
the field teams (see 4).
The following documents and data were available as a
basis for planning:
– Topographic maps on a scale of 1: 100 000
(Kievskaia Voenno-Kartograficheskaia Fabrika BKF
2003, sheet 183 Khust and 164 Mezhhor’e)
– A GIS data layer of the Uholka-Shyrokyi Luh massif
from CBR with the functional zones, contour lines,
rivers and paths
– Forest maps of the administrative units Uholka and
Shyrokyi Luh with partition borders 1:10000 and
1:25000 (VO “Ukrderzhlisproekt” 2001, Irpin)
– Orthophotos with a ground resolution of 0.87 m
from 2008 (Transcarpathian Geodesic Centre)
– Data from the 10 ha research plot in Uholka
(Co mmar mo t et al. 2005, 2009).
Survey perimeter
The perimeter of the inventory surrounds 99% of the
area of the Uholka-Shyrokyi Luh protected massif man-
aged by the CBR (Table 2.1). It comprises the core and
buffer zones of the massif, as well as a few enclosed
areas assigned to the regulated protection and anthro-
pogenic landscape zone (see map in the inside back
cover). The small area under traditional management
(anthropogenic landscape) along the southern border
of the massif was excluded from the survey. In total, the
area within the inventory’s perimeter comprises
10282.3 ha.
Population and target variables
The study objects, i.e. the population elements in sta-
tistical terminology, were living and dead trees, coarse
woody debris, forest regeneration (seedlings and sap-
lings), traces of anthropogenic use and of natural dis-
turbances, the horizontal and vertical structure of the
forest, and the local topographic conditions. Definitions
of the population elements and assessed variables are
described in detail in the field manual (Co mmar mo t et
al. 2010), which largely relies on those of the Swiss NFI
(Kel l er 2005) and the Swiss forest reserve inventory
(t in n e r et al. 2009). A short overview is given below.
The population of trees includes:
– living trees, whether standing or lying, with a
minimum diameter at breast height (DBH) of 6 cm,
measured vertical to the stem axis at a height of
1.3 m above ground (or the root collar),
– dead standing trees and snags (dead stems broken
above a height of 1.3 m) with a DBH ≥6 cm,
– dead lying trees (complete trees with crown and
root-plate) with a DBH ≥6 cm, and
– stumps (remaining base parts of stems) with a height
between 0.5 m and 1.3 m and a minimum diameter
of 6 cm.
The main attributes of interest are tree species and
DBH (Fig. 3.1). Further variables indicate the horizontal
layer in which the tree’s crown is situated, the stem
form (several variables), the crown length, micro-
habitats (several variables, such as cavities, cracks,
broken crown and occurrence of polypores) and the
degree of wood decay (5 classes). The stem heights of
all snags were measured. Tree height, the height to the
first green branch of the crown and the upper stem
diameter 7 m above ground (D7) were measured on a
sub-sample of living trees.
Lying deadwood not only includes complete trees,
but also broken stems, tree fragments and broken-off
parts from standing trees. The volume of lying dead-
wood was defined as the total volume of lying dead-
wood pieces with a diameter of ≥7 cm (over bark).
Thus, a single piece of deadwood may have a section
(coarser than 7 cm) accounted for in the lying dead-
wood volume and a section (smaller than 7 cm) not
accounted for in the lying deadwood volume. Only
above-ground material is included in lying deadwood.
A line intersect sampling technique was used to
assess the lying deadwood (see section “sampling
units” below). The variables measured in the field were:
the diameter (crosswise measurements), the decay
class (5 categories) and the tree species group (broad-
leaves and conifers) of the deadwood piece at the
intersection with the transect line.
21
The inventory – aims, methods and sampling design
Assessing forest regeneration involved measuring
living seedlings and saplings with a minimum height of
10 cm and a maximum DBH of 5.9 cm. These were
classified into 3 height classes and 6 DBH classes.
Further variables were the tree species and damage to
the leading shoot, in particular due to browsing. Local
site and stand characteristics with a potential influence
on the establishment and growth of forest regeneration
were also registered and included: the occurrence of
rocks, stones and boulders, type of topsoil (3 catego-
ries), competing vegetation, and shading.
Root-plates and canopy openings (gaps) were chosen
as indicators for natural disturbances (Fig. 3.3). Root-
plates were categorised as: root-plates with soil mate-
rial, root-plates without soil material and decomposed
former root-plates (recognizable as small mounds).
Canopy gaps were classified into 6 size classes (esti-
mated).
The vertical and horizontal forest structures were
assessed with two categorical variables (expert judge-
ments): the degree of crown cover in the upper, medium
and lower layers of the stand, and the type and degree
of canopy closure (aggregation of tree crowns in the
upper canopy layer).
Any traces of anthropogenic use observed were
classified into 10 categories (see 6, Table 6.2). The
traces were not quantified, for instance, by the number
of occurrences of traces on plots or by their size and
relevance. Nevertheless, it is still possible to assess
and monitor the amplitude and spatial distribution of
anthropogenic use (and activities) from the data.
The site factors assessed in the inventory were the
topographic characteristics: altitude, slope, aspect, and
relief (5 categories).
Sampling units (sample plot design)
Data from the 10 ha forest research plot in Uholka was
used to evaluate the optimum size of the sampling units
(sample plots) for trees and forest regeneration. The
sample plot design chosen is shown in Fig. 3.2.
Trees (DBH ≥6 cm) and root-plates were sampled
on circular plots with a fixed size of 500 m2(horizontal
radius of 12.62 m). Slope correction was applied to
ensure a uniform horizontal plot area of 500 m2. The
expected average number of (living and dead) stems
per plot was 15 (based on the stem density observed in
the Uholka forest research plot). Larger plots were
considered too difficult and error-prone, in particular on
steep slopes. We did not consider varying tree inclu-
sion probabilities (angle count sampling or concentric
circles) for the sake of simplicity and robust data collec-
tion and estimation. Moreover, the stem volume was
not the predominant population parameter of interest.
A sub-sample of the trees (so-called tariff trees) was
selected for measuring the tree height and upper stem
diameter (Fig. 3.4). The sub-sample includes all trees
in the first quadrant of the plot (sector between the
directions North and East, i.e. 0 and 90 degrees or 0
and 100 gon), as well as trees with a DBH of at least
60 cm (except for trees with broken stems and crowns).
The volume of lying deadwood was assessed on
three transect lines, each 15 m in length. The lines start
1 m from the sample plot centre and run in the direc-
tions of 35, 170 and 300 gon.
Regeneration was sampled on three concentric cir-
cular plots located 10 m from the centre of the main plot
(to the West): 5 m2for saplings between 10 cm and
39.9 cm height, 10 m2for saplings with a height between
40 and 129.9 cm, and 20 m2for saplings with a mini-
mum height of 130 cm and up to a DBH of 5.9 cm.
A circular interpretation area of 2500 m2(concentric
with the main sample plot for trees) was used to assess
Fig. 3.1. Mykola Korol measuring the DBH of a large beech tree.
The DBH was measured at a height of 1.3 m above ground.
Photo R. Tinner.
22 Inventory of the Largest Primeval Beech Forest in Europe
Fig. 3.3. Root-plates were assessed as indicators of natural
disturbances, but also as habitat features. Photo M. Hobi.
Interpretation area: 2500 m2
Sample plot: 500 m2
Sample plot centre
Regeneration sub-plot:
concentric circles of 5, 10 and 20 m2
Deadwood transect: 15 m
Fig. 3.2. Sample plot design.
Fig. 3.4. Martin Brüllhardt measuring the height of a tree. The
height was measured on only a sub-sample of trees. Photo A.
Khomiuk.
23
The inventory – aims, methods and sampling design
the horizontal and vertical structure of the forest, as
well as the occurrence of anthropogenic traces and
topographic variables. If a canopy gap was located
directly above the centre of a sample plot (point deci-
sion), it was noted and its size recorded.
The centre of the main sample plot and the centre of
the regeneration sub-plot were marked with a small oak
pole. The co-ordinates were registered with a GPS
device (Trimble GeoXH or Juno SB). To facilitate the
relocation of sample plots in future surveys, photo-
graphs were taken from the centre of the sample plot
in four different directions and one from downslope
towards the centre. In addition, any eye-catching objects
close-by (e.g. a rock or a tree with a large canker, Fig.
3.5) were registered with polar coordinates (distance
and azimuth from the plot centre), as were all trees
measured on the plot.
Fig. 3.5. To facilitate the relocation of sample plots in future surveys, any eye-catching objects close-by, such as this tree with a large
canker, were documented with their distance and azimuth from the plot centre. Photo U.-B. Brändli.
Distribution of sample plots (sampling grid)
Based on long experience with the Swiss NFI in land-
scapes with similar topographic conditions, and taking
into account the remoteness of the area with difficult
terrain conditions and long walking distances, work
performance of two sample plots per team and day was
assumed to be possible. This led us to conclude that a
sample size of approximately 350 plots should be
planned.
The sampling design chosen was a non-stratified,
systematic cluster sampling. Each cluster consisted of
two sample plots, with a distance of 100 m between
the two plots in a cluster. Clusters were arranged on a
rectangular grid (systematic sampling), with side
lengths of 445 m and 1235 m (Fig. 3.6). This design
resulted in 353 sample plots. The starting point for the
grid was randomly chosen.
24 Inventory of the Largest Primeval Beech Forest in Europe
The decision about how to distribute the sample plots
was based on the following considerations:
– The administrative units of Uholka and Shyrokyi Luh
are of similar size, and the forest (structure) can be
expected to be basically homogeneous over the
whole study area. Thus, a pre-stratification was not
judged appropriate. This does not, however, exclude
the use of strata (and other auxiliary information) in
the estimation stage of the inventory (post-stratifica-
tion).
– The systematic distribution of plots leads, in general,
to a higher precision of the estimates and to lower
inventory costs than with independent random point
sampling because the plots are distributed better
over the study area. Other advantages of systematic
sampling are the shorter walking distances involved
and the faster location of plot coordinates. Walking
distances are shorter on rectangular grids than on
quadratic grids (sCHmid -Haas 1993). A relation of up
to 4:1 between the longer and shorter side of rectan-
gular grids is acceptable (dv o r a K 2000), as other-
wise correlations between plots may become an
issue and should be addressed when estimating
sampling errors.
– Cluster sampling obviously reduces the cost (walk-
ing distances) of the inventory compared to an
inventory with the same number of single plots.
However, the cost reduction is achieved at the
expense of less precise estimates. The optimum
design is difficult to predict, even with extensive
pre-experience and data from pilot studies. The
design chosen has two sampling plots per cluster,
and is based on cost and population estimates
obtained from data collected earlier on the local
forest research plot (Co mmar mo t et al. 2005, 2009),
in a pilot inventory in 2009 and during a field visit. An
operational advantage was that two survey teams
could work within alarm distance of each other,
which would be important if an accident or emerg-
ency occurred (see 4).
References
BRaSSel , P.; l iSc h k e, h . (eds) 2001: Swiss National Forest
Inventory: Methods and Models of the Second Assess-
ment. Birmensdorf, Swiss Federal Research Institute
WSL. 336 pp.
BRän d l i , u.-B. (Red.) 2010: Schweizerisches Landesforstin-
ventar Ergebnisse der dritten Erhebung 2004–2006. Bir-
mensdorf, Eidg. Forschungsanstalt für Wald Schnee und
Landschaft WSL, Bern, Bundesamt für Umwelt, BAFU.
312 pp.
")
")
!
(!
(!
(!
(!
(!
(!
(
!
(!
(!
(!
(!
(!
(!
(
!
(!
(!
(!
(!
(!
(!
(
!
(!
(!
(!
(!
(!
(!
(
600
700
500
800
900
1
0
00
800
1000
600
700
500
800
900
1000
800
1000
21
23
22 24
16
20
17
29
12
353351345335333
390388386384382380378
352350340338336334
391389387385383381379
135
134
339337
354
698000 699000 700000 701000
5347000 5348000
°
UTM 34N
0 200 400 m
River
Forest road
Path
Countour lines
Perimeter of inventory
Partition border
!
(
Sample plots (n=353)
21
°
UTM 34N
River
Forest road
Path
Countour lines
Perimeter of inventory
Partition border
!
(
Sample plots (n=353)
21
Uholka-Shyrokyi Luh
0 200 400 m
°
UTM 34N
River
Forest road
Path
Countour lines
Perimeter of inventory
Partition border
!
(
Sample plots (n=353)
21
Fig. 3.6. Sampling design with clustered plots arranged on a rectangular grid. Extract from the overview map on a scale of 1:20000.
25
The inventory – aims, methods and sampling design
BRan g , P.; c o mma Rmo t , B.; Ro h Re R, l .; Bu g ma n n , h., 2008:
Monitoringkonzept für Naturwaldreservate in der Schweiz.
Birmensdorf, Eidg. Forschungsanstalt für Wald, Schnee
und Landschaft WSL; Zürich, ETH Zürich, Professur für
Waldökologie. Available from World Wide Web <http://
www.wsl.ch/publikationen/pdf/8555.pdf>. 58 pp.
BRan g , P.; h eiRi, c .; Bu g ma n n , h . (Red.) 2011: Waldreservate.
50 Jahre natürliche Waldentwicklung in der Schweiz. Bir-
mensdorf, Eidg. Forschungsanstalt WSL; Zürich, ETH
Zürich. Bern, Stuttgart, Wien, Haupt. 271 pp.
co mma Rmo t , B.; Ba c h o f e n , h .; Bu n d z iak , Y.; BüRg i , a.;
RamP, B.; Sh PaRYk , Y.; Su k h aRiu k , d .; Vit eR, R.; zin g g , a .,
2005: Structures of virgin and managed beech forests in
Uholka (Ukraine) and Sihlwald (Switzerland): a compara-
tive study. For. Snow Landsc. Res. 79, 1/2: 45–56.
co mma Rmo t , B.; Sh PaRYk , Y.; Su k h aRYu k , d .; Bü Rg i, a .;
zin g g , a., 2009: Entwicklung zum Urwald? Ein Vergleich
zwischen dem Zürcher Sihlwald und dem Buchenurwald
Uholka in der Westukraine. Mitt. Hess. Landesforstver-
walt. 47: 42–48.
co mma Rmo t , B.; t i n n eR, R.; BRan g , P.; BRä n d l i, u .-B., 2010:
Stichprobeninventur im Buchen-Urwald Uholka-Schyrokyj
Luh – Anleitung für die Inventur 2010. [published online
January 2012] Available from World Wide Web <http://
www.wsl.ch/publikationen/pdf/11562.pdf> Birmensdorf,
Eidg. Forschungsanstalt für Wald, Schnee und Landschaft
WSL. 65 pp.
dVo Rak , l ., 2000: Kontrollstichproben im Plenterwald. Doc-
toral Thesis. Swiss Federal Institute of Technology, Zurich.
Available from World Wide Web <http://e-collection.library.
ethz.ch/view/eth:24036>. 176 pp.
kel l eR, m. (ed) 2005: Schweizerisches Landesforstinventar.
Anleitung für die Feldaufnahmen der Erhebung 2004–
2007. Birmensdorf, Eidg. Forschungsanstalt für Wald,
Schnee und Landschaft. 393 pp.
kel l eR, m. (ed) 2011: Swiss National Forest Inventory. Manu-
al of the Field Survey 2004–2007 [published online March
2011]. Available from World Wide Web <http://www.wsl.
ch/publikationen/pdf/10919.pdf. Swiss Federal Research
Institute WSL Birmensdorf, WSL. 269 pp.
man dal l az , d ., 2008: Sampling techniques for forest invento-
ries. Chapman & Hall/CRC, Boca Raton, FL, USA. 256 pp.
Sc h mi d -h aaS, P., Bau m an n , e., WeRn eR, J., 1993: Forest In-
ventories by Unmarked Permanent Sample Plots: Instruc-
tions. Birmensdorf, Swiss Federal Institute for Forest
Snow and Landscape Research. 135 pp.
State Forestry Committee of Ukraine, 2010: Instrukzija pro
porjadok vedennja derzhavnoho lisovoho kadastru i per-
vynnoho obliku lisiv. Directive N 298, 01.10.2010.
tin n eR, R., St Reit , k ., c o mm aRmo t , B.; BRan g , P., 2010:
Stichprobeninventur in schweizerischen Naturwaldres-
ervaten – Anleitung zu Feldaufnahmen. Birmensdorf, Eidg.
Forschungsanstalt für Wald, Schnee und Landschaft
WSL. 43 pp.
Ukrderzhlisproekt 2006: Instruktsiia z vporiadkuvannia liso-
voho fondu Ukraïny. Irpin’. 75 pp.
4.1 Pilot inventory 2009
Preparations for the inventory started in winter 2008. A
stepwise procedure was adopted, with a pilot inventory
in July 2009 and the main inventory in July and August
2010 (Fig. 4.1). The project organisation and division of
tasks between WSL, UNFU and CBR are described in
Appendix 1.
During the pilot inventory in summer 2009, the two
main leaders of the 2010 field work tested the inventory
methods (in particular the field manuals and field
forms), the field equipment (e.g. the usability and accu-
racy of the GPS system), and the time needed for the
measurements and assessments. They also explored
the project area and inspected possible accommoda-
tion facilities, the condition of footpaths and river cross-
ings, and the available infrastructure. This was done in
close collaboration with the local forest officers. In total,
18 sample plots distributed within the whole perimeter
were localised and measured during the pilot inventory.
The results allowed the field manual and field forms to
be improved, as well as the detailed maps for orien-
tation. The time analysis showed that a reasonable goal
would be to assess 1.5 to 2 sample plots per day and
team if walking distances and logistics were optimised
by having decentralised camp sites and the logistic
support of locals.
4.2 Main inventory 2010
The logistics of organising the inventory posed a con-
siderable challenge due to the remoteness of the area,
the long walking distances involved and the difficult
terrain, about which we had very little information.
Access by car was only possible at three points on the
southern border of the survey perimeter in Uholka and
Shyrokyi Luh. Within the perimeter, there were only a
few small footpaths used by the rangers and the main
tracks to the mountain pastures, which were wide
enough for horses. Good organisation and preparation
were essential, as was the support of the local forest
service, without which it would not have been possible
to carry out the inventory.
The different preparatory tasks consisted of two main
types: methodological and logistic preparation (Table 4.1).
Transporting material to the campsite with a
horse-drawn sledge. Photo M. Brüllhardt.
4 Planning and management of the field survey
Martina L. Hobi, Brigitte Commarmot, Ruedi Iseli, Mykola Korol
“During the sometimes tough work, we gained insights
into how different people collaborate under such
special circumstances. We had the opportunity to
learn more about different traditions and ways of living
and about interpersonal relationships.”
Volodymyr Trotsiuk, student UNFU, Ukraine
28 Inventory of the Largest Primeval Beech Forest in Europe
Field equipment and orientation in the field
The field manuals (Co mmar mo t et al. 2010) were trans-
lated into Ukrainian by scientists at UNFU, and field
forms were prepared in German, English and Ukrainian.
As electricity was available at only a few places outside
the perimeter and the power supply was not reliable,
field computers could not be used. All the data collected
had to be written on paper forms (six different forms
were used) and entered into the computer afterwards.
For this an access data-base with entry masks based
on the field forms was created. The limited access to
electricity and the long walking distances also restricted
the choice of measurement instruments and equip-
ment. A list of all the field equipment used for field work
can be found in Appendix 2.
The available topographic maps were on a scale of
1:100000. They were useful to obtain a general orien-
tation, but more detailed maps were needed for reaching
the sample plots. Maps based on GIS-data from CBR
were therefore created for the whole area and printed on
water-resistant paper. Orthophotos with a ground resol-
ution of 0.87 m from 2008 provided background data.
The images were very useful when figuring out which
areas are covered with forest and for finding the way
along the forest boundary. Point symbols, such as the
sample plots, additional objects for orientation (e.g.
shelters and bridges), and campsites were mapped.
The line symbols plotted were rivers, forest paths, for-
est partition borders and the perimeter of the biosphere
reserve. Some information about topography was pro-
Table 4.1. Methodological and logistic preparation.
Methodological preparation Logistic preparation
Revision and translation of Àeld manuals and Àeld forms
Preparation of maps and orthophotos
Data handling, storage and quality
Instruments and equipment
Instruction of Àeld teams
Accommodation and infrastructure
Safety measures
Support of locals for transport, meals, etc.
Weekend program
Handling of Ànances
Recruitment of survey teams
Operational planning and coordination of survey teams
Fig. 4.1. Preparation of the pilot inventory. Scientists of WSL, UNFU and the Ukrainian Research Institute of Mountain Forestry with
satellite images and GPS, discussing how best to Ànd the way in the Àeld. Left to right: Vasyl Lavnyy, Yuriy Shparyk, Christoph Düggelin,
Christian Ginzler. Photo U.-B. Brändli.
29
Planning and management of the Àeld survey
vided by 100-m contour lines, but they proved to be not
very accurate. All the maps were based on the coordi-
nate system UTM 34N. The Universal Transverse
Mercator (UTM) grid uses a 2-dimensional Cartesian
coordinate system, where coordinates in each zone are
measured to the North and East in meters. Two differ-
ent versions of the maps were used in the field: an
overview map on a scale of 1: 20000 (Fig. 3.6), which
covered the whole perimeter, and 32 sub-maps on a
scale of 1:8000.
To find the sample plots, the survey teams mainly
used a Garmin GPS as a navigation instrument, on
which all the plot coordinates (that is a point 20 m south
of the sample plot centre) had been saved in advance.
Since the accuracy of the GPS system in the massif is
only up to 5–7 m, the last 20 meters had to be measured
with a compass (Wyssen MERIDIAN MI‐4007) and
measuring tape (Co mmar mo t et al. 2010). This ensured
that the exact location of the plot centre was randomly
selected and not chosen in the field by the survey
teams. The walking route to the sample plot needed to
be well planned, especially as not all the entries on the
map (e.g. footpaths) were up to date (Fig. 4.2). In some
cases access to the sample plots was hindered by
steep slopes, windthrow areas, blackberry thickets, or
creeks. This meant that indirect routes were inevitable.
The best help for orientation in the field was the system
of rivers and ridges. It was always easier to walk along
ridges since large amounts of deadwood tended to
accumulate on the valley floors. Once the plot centre
was found, the more accurate Trimble GPS was used
to measure the actual coordinates of the plot centre as
precisely as possible. The collected data was later
improved by post-processing.
Recruitment and instruction of field teams
As the field work was to be carried out by Swiss and
Ukrainian students, the survey period was limited to
two months (summer holidays). Based on the experi-
ence with the pilot inventory, it was decided to work with
six survey teams simultaneously. The Swiss students
were recruited from the Department of Environmental
Systems Science at ETHZ, the Institute of Natural
Resource Sciences at ZHAW in Wädenswil, and the
School of Agricultural, Forest and Food Sciences at
BFH in Zollikofen. All the Ukrainian students were
recruited at the UNFU. The students had not only to be
well versed in forest ecology and the identification of
tree species, but also to have good language skills, and
be physically and mentally healthy and robust, enthusi-
astic about working in remote areas, hard working and
precise, open-minded, tolerant and flexible.
Fig. 4.2. Survey team on the way to a sample plot. Photo A. Khomiuk.
30 Inventory of the Largest Primeval Beech Forest in Europe
Fig. 4.4. Campsite of two survey teams in the Shyrokyi Luh unit of the protected massif. Photo D. Oertig.
Fig. 4.3. Distributing material and equipment at the base camp in Mala Uholka. Photo B. Commarmot.
31
Planning and management of the Àeld survey
The field teams were instructed before the start of the
sampling inventory. Instruction days (one in Switzerland
and one in Ukraine) were organised to acquaint the par-
ticipants with the measuring methods and instruments.
These days were also important for the team members to
get to know each other. All the participants received the
field manual before the start of the survey. The first two
days in Uholka were spent with further training and organ-
ising the field equipment (Fig. 4.3). Most of the questions
and uncertainties could be discussed and solved during
these and the following few days, when all the teams
were staying at the base camp. After one month, some of
the field team members were replaced and the new par-
ticipants were instructed for one day in Shyrokyi Luh.
Safety measures
Due to the remoteness of the area and the lack of mobile
phone coverage within large parts of the survey perim-
eter, special safety precautions had to be taken. All the
participants had one day of training in first-aid with
professionals before the beginning of the inventory. Pos-
sible dangerous situations and rules of conduct in the
field were discussed with all the survey teams and also
given to them in written form. One of the most important
rules was to take no risks and to adapt behaviour to the
capabilities of the weakest team member. The sampling
team decided as a group whether a sampling plot could
be measured or had to be classified as inaccessible
because of steep slopes or dangerous rocks. All teams
were equipped with first-aid kits, emergency phone
numbers and alarm whistles to alert other teams work-
ing nearby if necessary to get help. Additional medical
equipment was stored at the camp sites. The local forest
officers and rangers repaired river crossings and built
some additional simple bridges. Luckily, no major acci-
dent happened during the inventory.
Accommodation and infrastructure
During the planning phase of the inventory, 19 possible
accommodation sites (huts and camp sites) suitable for
two to three survey teams were selected, and 17 were
finally used (see map in the inside back cover). A camp
site had to have water, be accessible with horses, have
a large enough flat and dry area to set up tents, and not
be at risk of getting hit by falling trees or branches in
windy weather. The three forest huts “Plesha”, “Shcher-
banova Poljana” and “Pidshchavna” were repaired in
advance so that they could be used as accommodation
(Fig. 4.4). Two base camps were set up. The one at the
Mala Uholka forest centre of CBR had a power supply
and several rooms that could be used for accommo-
dation and as offices for administration and data entry.
The one at the CBR control point in Shyrokyi Luh also
had some rooms we could use, but it did not have
electricity, so a generator had to be installed.
Operational planning and coordination of the
survey teams
It was decided to work first with all teams in Uholka,
before moving to Shyrokyi Luh during the second
month of the inventory. This procedure had the advan-
tage that sampling could be completed in at least one
of the two more-or-less secluded administrative units,
even if the conditions were such that not all the planned
353 sample plots could be assessed. Initially the plan
was to start with a thinned sampling grid, which could
be made more compact later if there was still enough
time. However, this strategy was rejected as the walk-
ing distances would have been too long and the accom-
modation would have had to be changed too often.
The research area was divided into sampling divi-
sions according to where accommodation was available,
and to the topography of the area. Walking distances
could thus be minimised and sampling plots revisited
on successive days if this was necessary to complete
the work. The number of sample plots within each
division corresponded to the estimated work capacity
of two teams per week. With this design, two teams
could each use the same accommodation or camp site
and work in the same area. Each of these pairs of team
groups was accompanied by one of the leaders or the
assistant leader of the survey. This person was then
responsible for coordinating the sampling within the
particular division and for supervising the work of the
two teams. In addition, the group leaders were also
responsible for coring one tree per plot for age analysis.
The survey teams consisted of one Ukrainian and
one Swiss student each. The common language of the
groups was either English or German. The two main
leaders of the inventory understood Ukrainian as well as
English and/or German. The groups stayed at a specific
camp site for one week, returning to the base camp on
Friday evenings with all their material. Local people
helped with horses to transport material to and from the
camp sites, and a cook stayed with the groups during
the week, taking care of food and guarding the personal
belongings of the students (Fig. 4.5). The local forest
officers were very involved in organising the camp sites,
improving the infrastructure, organising local helpers
and being on call for any unforeseen circumstances.
Work progress varied according to the topography,
forest structure and weather conditions. This meant
continuous planning was necessary and the sampling
divisions had to be adapted from week to week. Since
mobile phone reception was very poor, the coordination
of the field teams was very challenging. An exchange of
information between the field groups was only possible
on weekends. Thus, the whole program of all the teams
had to be fixed at the beginning of the week and later
changes were almost impossible. All the logistic sup-
port also had to be arranged over the weekends.
32 Inventory of the Largest Primeval Beech Forest in Europe
Data entry and data quality
During the whole inventory, two Ukrainians stayed at
the base camps, typing the data from the field forms
into the data-base. Thus, the completeness of the
forms and the data could be continuously checked and
unclear entries could be discussed with the field teams.
This, together with the intensive instruction and the
continuous supervision of the survey teams by the
field-work leaders, helped to ensure the data was of
high quality. Due to the limited time available, it was not
possible for a different team to measure a certain per-
centage of sample plots a second time, which would
have allowed a better estimate of the data accuracy.
The data quality could, however, be verified to some
extent on the 18 sample plots measured during the pilot
inventory in 2009 and re-measured in 2010.
Time management and work performance
The inventory period (July 5 to August 27, 2010)
included 39 working days. 12 of these days were not
spent on surveying the sample plots due to the instruc-
tion of the survey teams and the logistics of setting up
and changing camps, or because of bad weather with
heavy rain and thunderstorms (Table 4.2). This left only
27 days for measuring and assessing the sample plots.
Since there were six survey teams, this resulted in 162
working days for assessing the planned 353 plots.
Only two sample plots, located on the northern
border of the protected massif, could not be assessed
due to lack of time. A further 37 plots were either inac-
cessible or considered too dangerous for the survey
teams. These were mainly plots in steep ravines with
slopes of more than 100 %. Thus a total of 314 sample
plots were surveyed, i.e. on average approximately two
plots per day and team, which corresponded well with
the estimated time needed. Finding the 39 plots that
turned out to be inaccessible also consumed a large
amount of time.
On average, 198 minutes were needed to reach and
assess one sample plot. Approximately one third of this
time was needed to walk to the plot (Fig. 4.6), one third
for the measurements and assessments of the sample
trees, and one third for all the other assessments (lying
deadwood, regeneration, forest structure, topography,
anthropogenic traces), including marking the plot cen-
tre, taking photos and recording all relevant information
to facilitate relocation of the plot in the future. The cal-
culated average time per sample plot does not include
the time needed to return from the last plot of the day to
the camp. As it was almost impossible to take the same
route back, the teams had to ensure they had enough
time to reach the campsite before dark.
Whenever possible, the teams tried to finish work on
a plot before the end of the day so that they did not have
Fig. 4.5. Marija Bohdan preparing breakfast for the survey
teams. Photo D. Oertig.
Table 4.2. Distribution of working days.
Working on plots 27
Coordination and changing camps 7
Instruction of groups 3
Bad weather 2
Total number of working days 39
71
20
66
17
15
9
Location of
plot centre
Marking plot
centre, photos
Tree data
Transect data
(lying deadwood)
Regeneration
Plot data
Fig. 4.6. Average time needed (in minutes) to assess a sample
plot, including the time taken to walk to the plot from the campsite
or last plot. The time does not include getting back to the camp in
the evening.
33
Planning and management of the Àeld survey
to visit the same plot twice. This resulted in long working
days, which were not finished when back at the camp.
There, the material had to be checked and the field
forms copied by taking photos of them.The field work for
the next day also had to be planned and prepared. It was
not unusual for the teams to work for 10–12 hours a day.
Weekend program
Several joint weekend activities were organised to
provide a break from work and an opportunity for the
groups to meet and share experiences. These included
excursions in Transcarpathia to find out more about the
way of life of the people living in the region and about
some other intercultural projects going on, as well as
visits to the city of L’viv, the university town where the
Ukrainian participants in the inventory were studying.
As the workload was heavy and the field work physi-
cally strenuous, the weekends were almost too short,
and the need for a change of scene sometimes con-
flicted with the need for relaxation and recreation. Nev-
ertheless, the joint weekend activities were important
for socialising and team building.
Financial issues
A difficult task during the inventory was the manage-
ment of finances. All the payments to the local people
for meals, transport and other support had to be made
in cash (by each of the team leaders), as well as the
payments for weekend activities. This meant organis-
ing enough opportunities and time to obtain cash. This
form of payment also complicated the book keeping, for
which a separate member of the administrative support
group was responsible.
Intercultural collaboration
The collaboration between the Ukrainians and the Swiss
worked very well and was enriching for both sides. All
survey teams were highly motivated and committed to
meeting the goals of the inventory (Fig. 4.7). The different
cultural and scientific backgrounds of the team members,
the different habits and ways of organising things rarely
led to discord, but tended rather to be inspiring and have
a positive effect on the work. The support of the local
forest service and local families was encouraging and
crucial for the success of this project. All participants
contributed to creating a good atmosphere, which helped
everybody keep going even when conditions were diffi-
cult and hard. The close and intensive collaboration of
the Ukrainian and Swiss students has led to many main-
taining contact and even friendships with each other.
References
co mma Rmo t B.; t in n eR R.; BRan g P.; BRän d l i u .-B., 2010: Stich-
probeninventur im Buchen-Urwald Uholka-Schyrokyj Luh –
Anleitung für die Inventur 2010. [published online January
2012] Available from World Wide Web <http://www.wsl.ch/
publikationen/pdf/11562.pdf> Birmensdorf, Eidg. Forschungs-
anstalt für Wald, Schnee und Landschaft WSL. 65 pp.
Fig. 4.7. Field crew of the forest inventory in Shyrokyi Luh: survey teams and forest ofÀcers from Shyrokyi Luh and Uholka. Photo M. Brüllhardt.
5.1 Storage and handling of the data
The data of the inventory is stored and maintained at
WSL, but the Ukrainian partners also have a copy of
the data-base. The data is stored in two schemas
(applications) of a relational data-base. The first one
contains the raw data copied from the field forms (Fig.
5.1). Any modification of the raw data to correct errors
in the transcription from the forms is documented in a
table. After plausibility checks, the raw data was trans-
ferred to a second application used for data analysis.
All derived variables, such as the basal area or the
volume of trees, are saved in this second application,
again with all changes to the data documented in a
separate table. Several indicator variables for frequently
used subsets of trees and plots have been defined in the
tables to ease data extraction and analysis. Typical
examples of such subsets are sub-populations of trees,
such as living and dead trees, the sample of stumps or
the domain of accessed plots in the survey.
5.2 Evaluation routines
The statistical software R (R Development Core Team
2008) was used to develop evaluation routines. Two
functions were programmed, which can be easily
parameterised to produce basic result tables. One
function handles tables with one subgroup at the tree
level (e.g. diameter classes) and one subgroup at the
plot level (e.g. functional zones of the reserve). The
function is able to read and combine data from different
data-base tables, such as the plot table, the regenera-
tion table and the tree table. The other function reads
data from the regeneration table only and allows two
subgroups. Both functions produce point and error esti-
mates, which are stored in a separate result table. The
target variable (e.g. basal area), the tree population to
be analysed (e.g. living trees), the sub-populations
(e.g. diameter classes), the sub-domain of interests
(e.g. management zones) and the definition of forest
land under investigation (e.g. accessible forest) can be
defined in a flexible way. The estimators are those
given in ma n d al l a Z (2008, p. 65–69) under a non-
Data cleansing. Photo G. Proyer.
5 Data management and statistical evaluation
Meinrad Abegg, Martina L. Hobi, Edgar Kaufmann, Adrian Lanz
“What a unique experience it was to wander through
the forest, and to see and feel the dynamics in action,
while at the same time becoming immersed in a new
culture! We returned home enriched with memories,
new friendships and scientific data.”
Luca Mini, student ZHAW, Switzerland
36 Inventory of the Largest Primeval Beech Forest in Europe
stratified, one-phase, one-stage cluster random sam-
pling scheme. The two adjacent sample plots in the
double rows of the inventory (Fig. 3.6) should be treated
as a cluster of two plots when estimating population
parameters. The estimators are as follows:
As a first step, tree variables are summarised to plot
values, the so-called local density Y(xl), which is always
standardised to a per hectare value. In formula [1] below,
the sum is over the N(xl)trees in plot lof cluster x.fiis the
extrapolation factor to obtain hectare values (in our case
usually 20), and X
iis the value of the target variable Xfor
tree i. X can be e.g. the tree volume or take the value of
1 for the estimation of the number of stems.
[1]
The number of plots per cluster M(x) within the area of
interest is then calculated according to formula [2], and
the average number of plots per cluster M
–
2(usually 2 in
this inventory) according to formula [3], where IFis an
indicator variable indicating whether the plot xllies in
the domain of interest F, and n2the number of clusters
used for the estimation.
[2]
The local density for the cluster Y
c(x) is simply the arith-
metic mean of the plot densities [4]:
[4]
Finally, the following estimators are used to calculate
the estimate for the (unknown) mean spatial density Y
ˆc
of the target variable Xin the area of interest F[5] and
the variance of the estimate Var(Y
ˆc)[6]:
[5]
[6]
Local density at plot level.
Tree (1 … N).
Value of target variable for tree i.
Extrapolation factor of the ith tree to obtain
hectare values.
Cluster (1 … n2).
Plot lin cluster x(1 … M).
Fig. 5.1. Vasyl Lutsyshyn and Volodymyr Savchyn in the CBR check-point ofÀce in Shyrokyi Luh, typing data from the Àeld forms to the
data base. Photo B. Commarmot.
!=! !
! ! !
!! !
=! !
!
!! !
!=!
!!( )
!!!
!=!!!!! ! !
!
!! !
! !
!=! ! ! !(! )
!!!
! ( ! )
!!!
!=!
!!(! !! ! )
! (! )
!!
!
!( ) −!
!
!!!
!:
:
!:
!
:
:
!
:
[3]
37
Data management and statistical evaluation
Number of plots nin cluster x.
Indicator variable, indicating whether the plot xllies
in the domain of interest (F).
Average number of plots per cluster falling into
the domain of interest (F).
Local density at the cluster level.
Estimated local density in the domain of interest
(point estimate).
Variance (error) of the estimated local density.
Set of clusters (sample) used for the estimation.
Number of clusters used for the estimation.
Confidence intervals are based on the standard error of
the estimate, which is the square root of the variance.
To test for differences in the estimators between two
sub-domains of interest, a two sample t-test with une-
qual variances was used.
The total of the target variable X, i.e. the total timber
volume in the reserve, is computed by multiplying the
above mean spatial density of the variable (and likewise
the standard error) with the area of interest (F).
The point estimator for the ratio of two variables X
and Z(e.g. the volume of beech in relation to the total
volume of all tree species) is simply the ratio of the two
estimates. The formula for estimating variance is
slightly more complicated (ma n d al l a Z 2008, p. 68).
5.3 Volume estimation
Volume of living trees
To create a tariff function to estimate the volumes for all
trees with an intact stem (not broken), we computed, in
a first step, the individual stem volumes V1of the 1054
trees with additional measurement of the stem height H
according to formula [7] (Kaufmann, unpublished):
[7]
The volumes V2of the 520 trees were computed with
additional measurements of the stem height H and the
upper stem diameter D7 according to formula [8], which
was developed for the Swiss NFI (Ka UFman n 2001, p.
163):
[8]
In a second step, the volumes of these 1574 trees were
used to calibrate a function that predicts the tariff vol-
ume (TV) for all trees measured [9]. The tariff volume is
defined as the stem volume from the ground to the
stem top, including the bark. The explanatory variables
are the DBH of the stem, the altitude of the plot (ALT, m
a.s.l.), the presence of a stem bifurcation (BF, an indi-
cator variable) and the crown length (CL, categorical
variable in three classes):
[9]
The coefficients were estimated by nonlinear regres-
sion analysis.
Finally, to calculate the total volume including branch
wood with a minimum diameter of 7 cm, a certain per-
centage was added to the stem volume (Fig. 5.2). This
percentage was assumed to be equal to the proportion
observed on average in Swiss forests (dUC et al. 2010).
Thus, the branch volume of beech trees was estimated
to be 17% of the stem volume. For all other deciduous
tree species, the value 7% for sycamore (Acer sp.) was
chosen, as this species accounts for the largest share
of the basal area of the admixed species. Branch
volumes were only added to trees with a complete
crown.
:
! !
:
!
:
!( ):
!
:
!:
!:
Fig. 5.2. The volume of a tree includes the stem volume from
the ground to the stem top and branch wood with a minimum
diameter of 7 cm. Photo L. Denzler.
!= 0.03427 + 0.35690 ∗!∗–0
∗–0.02497 ∗!∗
!= 0.002542 + 2.56612 ∗!−3.67034 ∗
34 ∗7!+ 0.39446 ∗7!∗+ 0.03567 ∗!∗
=−9.88133 + 3.03787 ∗−
−0.002725617 ∗
!
−0.000387604 ∗
−0.11263 ∗−0.044796 ∗
17 ∗
!: Va
38 Inventory of the Largest Primeval Beech Forest in Europe
Volume of dead standing trees, snags and stumps
To calculate the volume of dead standing trees, four
cases were distinguished:
a) Dead trees with intact stems and crowns (Fig. 5.3):
The volume was calculated according to the tariff
function TV (see formula [9] above), plus the addi-
tion for branch volume.
b) Dead trees with an intact stem but only stubs of
branches: The volume was calculated according to
the tariff function TV [9] (without branch wood).
c) Snags (stem broken above 1.3 m height, Fig. 5.4):
The snag volume V
s[10] was calculated according to
Co mmar mo t et al. (2005) as a cylinder, using the
measured snag height Hsand the modeled diameter
at half the snag height D0.5Hs. Based on the trees
where both the DBH and the upper stem diameter
D7 were measured, a simple linear model was calcu-
lated to predict the D7mod from the DBH [11]. The
DBH and D7mod were then used to estimate the
diameter decrease per m tree height DDm[12], and
to calculate the diameter at half the snag height D0.5Hs
[13].
[10]
[11]
[12]
[13]
Volume of the snag [m3].
Measured height of the snag [m].
Modelled diameter at half the snag height [m].
Measured diameter at 1.3 m above ground [m].
Modelled diameter decrease per meter [m].
Modelled diameter of the tree at a height of
7 m [m].
This approach to volume estimation is based on the
assumption that the diameter decrease per m from
the stem base to the top of the stem is constant and
equal to that between 1.3 m and 7 m height. As the
mean snag height was 5.25 ± 0.01 m (see 6.4), we
Fig. 5.3. Dead tree with intact stem and crown. The volume of
such a tree was estimated in the same way as that of a living tree.
Photo U.-B. Brändli.
Fig. 5.4. The volume of a snag was calculated as a cylinder,
using the measured snag height and the modeled diameter at
half the snag height. Photo B. Commarmot.
!=!! .! ! !
!
!
!
7! " # = 0.8834 ∗−0.019122
!=! " # ! ! ! ! " #
! .!
! .! ! !=−!!!
!−1.3
!:
!:
! .! ! !:
:
!:
7! " # :
39
Data management and statistical evaluation
think that any error arising from this assumption is
negligible. Another simplification was to apply the
cylinder formula, which slightly underestimates the
stem volume (Kr amer and aKç a 1995).
d) Stumps < 1.3 m height: The volume was computed
according to formula [10] as a simple cylinder, with
the height of the stump and its diameter halfway up,
which were both measured in the field.
Volume of lying deadwood
The volume of lying deadwood assessed with line inter-
sect sampling (3.2) was calculated according to Bö Hl
and Br ä n d l i (2007):
[14]
Estimated lying deadwood [m3/ha] on sample
plot xl.
Number of transects on sample plot xl.
Horizontal length of the kth transect [m].
Diameter of deadwood piece i[cm] measured
crosswise
Inclination of the deadwood piece [gon].
Number of deadwood pieces on the kth
transect line.
The volume was stored as a plot variable (represented
volume of deadwood per ha), and in a separate table
as volume per ha represented by each piece of dead-
wood. This meant the lying deadwood could be classi-
fied for the evaluation.
References
Bö h l , J.; BRä n d l i, u .-B., 2007: Deadwood volume assess-
ment in the third Swiss National Forest Inventory: meth-
ods and Ärst results. Eur. J. For. Res. 126: 449-457.
Co mmar mo t , B.; Ba CHo Fe n , H.; BUn d Zia K, y.; Bü r g i, a .;
ramP, B.; s HPar y K, y.; s UKHar iUK, d.; v it e r , r .; Zi n g g ,
a., 2005: Structures of virgin and managed beech forests
in Uholka (Ukraine) and Sihlwald (Switzerland): a compar-
ative study. For. Snow Landsc. Res. 79, 1/2: 45–56.
du c , P.; BRä n d l i, u .-B.; h e Rol d Bo n aRd i, a .; Rö Sl eR, e.;
th ü Rig , e.; u l meR, u .; f Ru t ig , f .; Ro SSet , c .; kau f ma n n ,
e., 2010: Holzprodunktion. In: BRä n d l i, u .-B. (Red.)
Schweizerisches Landesforstinventar. Ergebnisse der
dritten Erhebung 2004–2006. Birmensdorf, Eidg. For-
schungsanstalt für Wald, Schnee und Landschaft WSL.
Bern, Bundesamt für Umwelt, BAFU. 143–184.
kau f ma n n , e., 2001: Estimation of Standing Timber, Growth
and Cut. In: BRaSSel , P.; l iSc h k e , h . (eds) Swiss National
Forest Inventory: Methods and Models of the Second As-
sessment. Birmensdorf, Swiss Federal Research Institute
WSL. 162–196.
kRame R, h.; ak ç a, A., 1995: Leitfaden zur Waldmesslehre.
J.D. Sauerländer’s. Frankfurt am Main. 251 pp.
man dal l az , d ., 2008: Sampling techniques for forest invento-
ries. Chapman & Hall/CRC, Boca Raton, FL, USA. 256 pp.
R Development Core Team, 2008: R: A language and environ-
ment for statistical computing. R Foundation for Statistical
computing, Vienna, Austria.
(!)
:
Es
(!) = !
!!
!!
! ! !
! ! !! ! ! !
!
!
! ( ! )
!! !
!
! " #!
!!
! ! !
ℎ!
:
N
!
:
H
( ): N
!
1!
,
2!: Dia
: Incl
!:
Crown of old beech tree covered with mosses
and ferns. Photo U.-B. Brändli.
6 Main results
Brigitte Commarmot, Meinrad Abegg, Urs Beat Brändli, Martina L. Hobi, Mykola Korol, Adrian Lanz
6.1 Presentation and statistical inter-
pretation of the results
In this chapter, we present some basic findings about
the current status of the primeval forest of Uholka-
Shyrokyi Luh. All results relate to the areas accessible
within the perimeter (assessed plots; see 4), given
separately for the two administrative units Uholka and
Shyrokyi-Luh, as well as for the entire massif. In some
cases, they are also presented separately for the differ-
ent functional zones of the protected massif.
The results summarised in the following tables are
statistical estimates of unknown population parameters,
such as the average volume of living and dead trees per
hectare or the total number of stems per diameter class.
The estimates are subject to sampling errors, i.e.
sample-to-sample variations, which originate in the ran-
domised sample selection. For this reason, all estimates
are given with standard errors so that confidence inter-
vals for the unknown population parameter can be
computed at any desired level. For a 95% confidence
level, for instance, the lower and upper confidence limits
are X– 2 * SE(X) and X+ 2 * SE(X), while X–SE(X) and
X+ SE(X) are the confidence limits at the 68% confidence
level (Xdenotes the estimate, and SE(X) its standard
error). The correct interpretation of such a confidence
interval is: Assuming repeated samples have been ran-
domly collected according to the same sampling design,
the unknown population parameter would be within the
above mentioned confidence interval limits in 95% (68%)
of these samples (Co CHr a n 1977; s ä r n d a l et al. 1992).
Confidence intervals can be used to test whether the
observed difference between the estimates for two
different populations (e.g. the tree population of Uholka
and the tree population of Shyrokyi Luh) is statistically
significant. If the confidence intervals of the two esti-
mates do not overlap, we conclude that the difference
between the two population parameters is significant at
the given probability level, and that otherwise the differ-
ence is not significant.
In this survey, 314 sample plots 500 m2in size were
assessed, with a total surface area of 15.6 ha. Thus,
the sampling intensity in this survey is 0.15% (the entire
study area measures 10282.3 ha). On average, each
sample plot represents an area of 29.1 ha.
“This was a very good project. The steep slopes,
regeneration, deadwood and windthrow areas presen
ted difficulties. I thought such an inventory would not be
possible – but we did it.”
Vasyl Kostyshyn, PhD student UNFU, Ukraine
42 Inventory of the Largest Primeval Beech Forest in Europe
6.2 Topography and anthropogenic
traces
Local distribution and topographic characteristics
of the plots studied
The 314 assessed sample plots are distributed almost
equally over the administrative units of Uholka (46%)
and Shyrokyi Luh (54%). 71% are in the core zone of
the massif (Table 6.1).
The distribution of plots according to altitude, slope,
aspect and relief (Fig. 6.1) reflects differences in topo-
graphy and site conditions between the two administrat-
ive units, which may influence the growth conditions
and species composition of the forest. On average, the
terrain in Uholka is at lower altitudes, less steep and
more south exposed than that in Shyrokyi Luh. In
Uholka, the average altitude of the plots is 778 ± 22 m
and the average slope 46 ± 2 %, and in Shyrokyi Luh it
is 908 ± 18 m and 55 ± 1 %, respectively. It should be
noted that the steepest plots were classified as inac-
cessible and are therefore not represented in these
values. The lowest sample plot assessed is in Uholka
at 460 m a.s.l., and the highest in Shyrokyi Luh at 1270
m. In both territories, 80 % of the plots are on the mid-
dle slopes, and only a few are on hilltops or ridges
(upper zone of slope with water and nutrient run-off), or
on the foot slopes, where water and nutrients are most
plentiful. Whereas in Shyrokyi Luh the plots are almost
evenly exposed to the different aspects, only a few
plots in Uholka are north exposed (north-west to north-
east).
Traces of human activities
The anthropogenic traces assessed include all kind of
traces from discarded cigarette packets to traces of
recent or former logging. Traces of human activities
were found on 19% of all plots (interpretation area,
2500 m2in size). They were three to four times more
Table 6.1. Number of plots assessed in the administrative units
and functional zones.
Uholka Shyrokyi
Luh
Whole study
area
Core zone 98 124 222
Buffer zone 36 36 72
Other zones 11 9 20
All zones 145 169 314
Fig. 6.1. Relative frequency of a) altitude, b) slope, c) relief and d) aspect of the assessed plots in the administrative units Uholka and
Shyrokyi Luh. Total number of plots = 314.
Shyrokyi Luh
0 10 20 30 40 50
≤ 600
601–800
801–1000
> 1000
Number of plots [%]
Altitude [m]
Shyrokyi Luh Uholka
0 10 20 30 4
0 5
0
≤ 20
21–40
41–60
61–80
81–100
> 100
Number of plots [%]
Slope [%]
0 20 40 60 80 100
Plain, flat area
Foot slope, depression
Middle slope
Hilltop, upper slope
Not applicable
Number of plots [%]
0
5
10
15
20 N
NE
E
SE
S
SW
W
NW
Uholka
a)
c)
b)
d)
Number of plots [%]
43
Main results
Table 6.2. Anthropogenic traces found in the different functional zones. Number of plots (2500 m2in size) with anthropogenic traces.
Type of anthropogenic trace Core zone Buffer zone Other zones Whole study area
Timber cutting 7 6 4 17
Timber not removed 2 1 0 3
Timber removed 1 4 2 7
Not identiÀable 2 1 1 4
Not speciÀed 2 0 1 3
Roads or paths 10 19 3 32
Footpaths 9 6 1 16
Trampling tracks (e.g. by horses) 0 8 2 10
Wheel tracks 1 5 0 6
Livestock grazing, pasturing 0 12 0 12
Fire, traces of burning 1 2 0 3
Buildings or other constructions 1 0 0 1
Plantation 1 0 0 1
Litter, waste 6 10 1 17
Anthropogenic damage to trees 4 5 1 10
Research, monitoring 4 2 2 8
Other traces (gas pipeline) 0 1 0 1
Some anthropogenic traces 25 30 6 61
In % of all plots in respective zone 11 42 30 19
Total number of plots assessed 222 72 20 314
Fig. 6.2. Percentage of plots with anthropogenic traces found in the two administrative units. Total number of plots assessed = 314 (145 in
Uholka, 169 in Shyrokyi Luh). Plot size = 2500 m2.
0 42 6 8 10 12 14 16
Other (gas pipeline)
Research, monitoring
Damage to trees
Litter, waste
Plantation
Buildings
Soil cultivation, mining
Charcoal burning site
Fire, traces of burning
Grazing, pasturing
Roads and paths
Timber cutting
Plots with traces [%]
Shyrokyi Luh
Uholka
44 Inventory of the Largest Primeval Beech Forest in Europe
frequent in the buffer and regulated protection zone
than in the core zone (Table 6.2), where traces were
detected on only 11% of the plots. They also tended to
be more frequent in Uholka than in Shyrokyi Luh (Fig.
6.2). This is mainly due to the higher density of roads
and paths, in particular in the southern part of the
Uholka administrative unit.
“Roads and paths” were the most frequent anthropo-
genic traces found (Table 6.2). Paths consisted mainly of
small footpaths (Fig. 6.3), but in the buffer and regulated
protection zone there were also wider tracks used for
packhorses, horse-drawn vehicles or off-road trucks.
Traces of timber cutting were noted on 17 of the 314
plots, 7 of which were inside the core zone. This does
not necessarily mean, however, that timber was actu-
ally removed from the forest, although it clearly was
from one plot in the core zone. In other places, trees
that had fallen across a path had been sawn through to
facilitate access, with the timber left on site (Fig. 6.3).
Other anthropogenic traces found were waste, (anthro-
pogenic) damage to trees, and in the buffer zone also
traces of livestock grazing (mainly by sheep and goats,
Fig. 6.4).
Apart from paths and items to do with “research and
monitoring”, most anthropogenic traces were found
close to the settlements of Mala and Velyka Uholka,
along the upper forest line and along the main routes to
the mountain pastures in Shyrokyi Luh (see maps in
Appendix 3). All in all, the impact of anthropogenic use
– at least in the core zone of the reserve – appears to
Fig. 6.3. Small footpath used by locals and forest rangers in the
Uholka-Shyrokyi Luh massif. A fallen tree across the path has
been sawn through to facilitate access. Photo V. Chumak.
Fig. 6.4. Sheep and goats on the way to the mountain pasture. Photo B. Hasspacher.
45
Main results
be very small, although the traces found suggest that
the area is frequented quite often.
6.3 Tree species diversity and forest
structure
The current forest composition and structure allow
some inferences to be drawn about the forest dynamics
and (natural or anthropogenic) disturbances. The tree
species and structural diversity of a forest are also
important for biodiversity, as they provide different
habitats and diverse light conditions.
Tree species diversity
A total of 6779 living trees and 460 dead standing trees
and snags ≥6 cm DBH were assessed on the 314 sam-
ple plots. On average, there are 435 (±12) living trees
per ha in the study area (Table 6.4), most of which are
beech (Fagus sylvatica L.). The proportion of tree spe-
cies other than beech is less than 3% (Table 6.3), and
slightly higher in Uholka than in Shyrokyi Luh. In total,
15 tree species were identified on the plots, two of them
only in saplings less than 1.3 m tall (for the list of all
species recorded, see Appendix 4). The most frequent
admixed species are: sycamore (Acer pseudoplatanus
L.), silver fir (Abies alba Mill.), elm (Ulmus glabra Huds.)
and European hornbeam (Carpinus betulus L.), which
are all relatively shade tolerant, at least when young
(el l enBer g et al. 1992; ew al d 2007). Very light-
demanding species typically occurring in early suc-
cessional or pioneer phases, such as poplars (Populus
sp.) or willows (Salix sp.), were found very rarely, and
then only along the forest edge or close to a river.Apart
from a few Norway spruce (Picea abies Karst.) trees
planted beside huts, all species found are of natural
origin.
Sycamore has the widest distribution of all the
admixed species within the study area, occurring from
less than 500 m altitude up to the upper forest line (Fig.
6.5). Norway maple and elm were also found up to
altitudes of 1000 m or more, whereas the other admixed
broadleaved species were mainly found at lower alti-
Table 6.3. Tree species composition (percentage of basal area; living trees ≥6 cm DBH).
Uholka Shyrokyi-Luh Whole study area
% ± SE % ± SE % ± SE
Beech (Fagus sylvatica L.) 96.5 ± 1.0 97.9 ± 0.9 97.3 ± 0.7
Sycamore (Acer pseudoplatanus L.) 1.4 ± 0.6 0.5 ± 0.3 0.9 ± 0.3
Norway maple (Acer platanoides L.) 0.1 ± 0.1 0.2 ± 0.2 0.1 ± 0.1
European Hornbeam (Carpinus betulus L.) 0.8 ± 0.4 0.5 ± 0.4 0.6 ± 0.3
Elm (Ulmus glabra Huds.) 0.8 ± 0.7 0.0 ± 0.0 0.4 ± 0.3
Ash (Fraxinus excelsior L.) 0.1 ± 0.1 0.0 ± 0.0 0.1 ± 0.0
Silver Àr (Abies alba Mill.) 0.0 ± 0.0 0.8 ± 0.6 0.5 ± 0.3
Other species 0.3 ± 0.1 0.1 ± 0.1 0.2 ± 0.1
All species 100.0 ± 0.0 100.0 ± 0.0 100.0 ± 0.0
Fig. 6.5. Sycamore (Acer pseudoplatanus L.) saplings. Tree
species other than beech were more frequent in the regeneration
than in the tree population ≥6 cm DBH. Photo B. Reineking.
46 Inventory of the Largest Primeval Beech Forest in Europe
tudes in Uholka, and to a lesser extent also in Shyrokyi
Luh. Silver fir occurs only on a limited area in the north-
ern part of Shyrokyi Luh above 750 m a.s.l. Maps show-
ing the distribution of the sample plots with the most
widespread tree species can be found in Appendix 5.
All species present in the tree population ≥6 cm
DBH were also found in the regeneration (in total, 1277
saplings ≥10 cm tall and < 6 cm DBH were assessed).
The percentage of tree species other than beech
decreases from 17% in the regeneration class 1 to 7%
in class 2 and 4% in class 3 (Fig. 6.6). This finding
shows that beech is very competitive and may outlive
longer suppression periods than most of the other
species found. On average, there were more than
34000 seedlings and saplings per hectare (3 per m2),
69% of which were less than 40 cm tall.
Browsing damage to the leading shoot caused by red
deer (Cervus elaphus L.) or roe deer (Capreolus
capreolus L.) was very rarely found. Only 0.1% (± 0.1 %)
of all 10 to 129 cm high saplings had been browsed
during the previous year. Although saplings other than
beech were browsed about four times as often as
beech (0.4% ± 0.2% of the admixed species), the
current populations of red and roe deer do not seem to
have a negative impact on the regeneration density. It
appears that most of the admixed species are able to
maintain their albeit very low share in the distribution of
the tree population.
Forest structure
On average, the number of living trees in the virgin
forest of Uholka-Shyrokyi Luh was 435 (±12) per ha,
the basal area 37 (±1) m2, and the growing stock (vol-
ume of living trees) 582 (±14) m3(Table 6.4). Deadwood
(standing and lying) made up 22% of the total volume
Table 6.4. Main characteristics of forest structure. (Calliper limit = 6 cm DBH. The volumes include branch wood ≥7 cm diameter.)
SigniÀcant differences (t-test) between Uholka and Shyrokyi Luh are indicated with * (* P ≤0.05, ** P ≤0.01, *** P ≤0.001).
Population Parameter Uholka
Mean ± SE
Shyrokyi Luh
Mean ± SE
Whole study area
Mean ± SE
Living trees ≥6 cm DBH
Number of trees [N/ha] standing 411.0 ± 19.4 448.1 ± 15.3 431.0 ± 12.2
lying 3.6 ± 1.2 4.4 ± 1.1 4.0 ± 0.8
total 414.5 ± 19.4 452.5 ± 15.2 435.0 ± 12.2
Basal area [m2/ha]
standing 35.3 ± 1.2 37.3 ± 1.0 36.3 ± 0.8
lying 0.2 ± 0.2 0.3 ± 0.1 0.3 ± 0.1
total 35.5 ± 1.2 37.5 ± 1.0 36.6 ± 0.8
Volume [m3/ha]
standing 582.7 ± 21.5 574.6 ± 17.4 578.4 ± 13.6
lying 3.9 ± 2.8 3.7 ± 1.2 3.8 ± 1.5
total 586.6 ± 21.4 578.3 ± 17.2 582.1 ± 13.5
Deadwood
Number of trees [N/ha] standing1) 26.6 ± 3.8 32.2 ± 2.6 29.6 ± 2.2
lying2) 1.9 ± 0.8 1.0 ± 0.4 1.4 ± 0.4
Volume [m3/ha]
standing1) 30.4 ± 5.7 23.3 ± 2.9 26.6 ± 3.1
lying3) 150.1 ± 11.2 123.8 ± 10.0 135.9 ± 7.5
total 180.5 ± 12.9 * 147.1 ± 10.7 162.5 ± 8.4
1) Dead trees and snags
2) Only complete trees with crown and rootplate, and with DBH ≥6 cm
3) Total volume of lying deadwood (including the complete trees) with a diameter ≥7 cm, assessed with line intersect sampling
Fig. 6.6. Regeneration density according to height class and tree
species. Maple = Acer pseudoplatanus and Acer platanoides.
Total number of saplings (all height classes) < 6 cm DBH = 34642
(± 3859) per hectare.
0
5000
10000
15000
20000
25000
10–39 40–129 ≥ 130
Regeneration density (N/ha)
Height class [c
ther s ecies
a le
eech
47
Main results
(living and dead) of 745 m3per ha. These values lie
within the range of values given for primeval beech
forests in the north-western Carpathians of Slovakia
(KUCBel et al. 2012; Ko r Pe l ’ 1995). Tree density, basal
area and growing stock did not significantly differ
between the administrative units of Uholka and Shy-
rokyi Luh. The total deadwood volume was slightly
higher in Uholka than in Shyrokyi Luh, although the
difference is statistically of little significance (p = 0.05).
The horizontal and vertical forest structure can be
described in terms of the canopy closure, the degree of
crown cover in the different canopy layers and the
estimated gap sizes (see 3.2). In general, the virgin
beech forest of Uholka-Shyrokyi Luh is rather dense,
with only small gaps in the canopy (Fig. 6.7). The upper
canopy layer was mainly fairly loose (with few gaps the
size of a canopy tree) or closed (Fig. 6.8a). The canopy
was described as scattered on less than 20% of the
plots, which had several gaps large enough for more
than one canopy tree to fit. 61% of the sample plot
centres did not lie in a gap (Fig. 6.8b) and only 16%
were in gaps larger than 200 m2(corresponding to a
radius of 8 m). Some gaps larger than 5000 m2were
also found, indicating the albeit rare occurrence of
medium- to large-scale disturbances, e.g. caused by
wind. The vertical forest structure was mainly three-
layered (in two thirds of the sample plots), which means
that the degree of cover of the upper, medium and
lower canopy layer was ≥20% each. A one-layered
structure (where only one of the canopy layers had a
coverage of at least 20%) was found in only 7% of the
plots (Fig. 6.8c). The structural diversity was slightly
more pronounced in Shyrokyi Luh than in Uholka,
where there were fewer plots with a closed canopy and
the vertical structure was better developed (Fig. 6.8).
The diameter distributions, showing the number of
living trees per 4-cm diameter classes, differed slightly
between Uholka and Shyrokyi Luh, although the gen-
eral trend of the curves is similar (Fig. 6.9). The rotated
sigmoid type of diameter distribution is commonly
observed in primeval forests (Fig. 6.10). This type, where
the highest density of trees is in the smallest diameter
class, with a second small peak in the mid-diameter
range, is clearly visible in the curve for Uholka, and less
Fig. 6.7. Small-scale mosaic of multilayered forest with regeneration of beech growing in a small gap. Photo U.-B. Brändli.
48 Inventory of the Largest Primeval Beech Forest in Europe
Fig. 6.8. Structural chracteristics: a) Canopy closure (upper
layer); b) Size of gap above sample plot centre; c) Vertical
structure (number of canopy layers). Layers with a coverage of
less than 20% were not considered.
Shyrokyi Luh
Uholka
0
10
20
30
40
50
60
70
80
Not
applicable
One-
layered
Two-
layered
Three-
layered
Number of plots [%]
Shyrokyi Luh
Uholka
0
10
20
30
40
50
60
70
Closed Loose Scattered
Number of plots [%]
Shyrokyi Luh
Uholka
0
10
20
30
40
50
60
70
No gap
20–50
51–200
201–500
501–1000
1000–5000
> 5000
Not specified
Number of plots [%]
Estimated gap size [m2]
Shyrokyi Luh
Uholka
0
20
40
60
80
100
120
140
160
8 16 24 32 40 48 56 64 72 80 88 96 104 112 120 128 136 144 152
Living trees [N/ha]
DBH [cm]
Fig. 6.9. Diameter distribution of living trees ≥6 cm DBH in the administrative units Uholka and Shyrokyi Luh. The numbers on the
horizontal axis indicate the midpoints of the 4 cm DBH classes (e.g. 8 = 6.0–9.9 cm DBH class). The dashed lines show the distributions
predicted by a three-parameter Weibull function.
a) b)
c)
49
Main results
clearly visible in the curve for Shyrokyi Luh. It seems to
be a typical diameter distribution pattern for deciduous
old-growth forests without severe (stand-replacing)
disturbances (go FF and w es t 1975; w es t PHa l et al.
2006; al e s s an d r i n i et al. 2011).
Large and old trees
On average, 10 ± 1 trees per ha had a DBH of 80 cm or
more (Fig. 6.11). Such giant trees were more frequent
in Uholka (12 ± 1 per ha) than in Shyrokyi Luh (8 ± 1 per
ha), possibly due to the better growth conditions at the
lower altitudes. The largest tree measured was an elm
with a DBH of 150 cm, a height of 43 m and a volume2
of approx. 38 m3. The largest beech measured was 140
cm thick and had a volume of 28 m3. Both trees were
found in the Uholka administrative unit. The largest tree
encountered in Shyrokyi Luh, a beech, had a DBH
of 115 cm and a volume of 21 m3. In both territories,
trees of up to 50 m in height were found. The tallest tree
measured was a 53 m beech.
2Including branches ≥7 cm diameter.
A complementary study conducted in Uholka on 164
trees showed that beech in this area may reach an age
of 450 to 500 years, although trees older than 150
years are prone to stem rot (tr o t s i UK et al. 2012). Old
trees, however, are not necessarily very large, and the
largest trees are usually not the oldest ones. The oldest
beech where the tree rings could be reliably counted
was at least 464 years old and had a DBH of only 63 cm.
The tree-ring analyses revealed that beech trees may
survive long suppression periods of over 100 years.
This explains why only a weak relationship between
DBH and age was found, involving uncertainties of 100
to 200 years for a given DBH.
Fig. 6.10. Even within small areas, the tree diameters range widely. Photo U.-B. Brändli.
50 Inventory of the Largest Primeval Beech Forest in Europe
6.4 Habitat structures
Deadwood and old trees provide the most important
habitat features for thousands of wood-dwelling ani-
mals, fungi, mosses and lichen species (mül l e r and
Bü t l e r 2010; Br ä n d l i and Be r a n o v a 2011; l a s s a UCe
et al. 2011) (Fig. 6.12). As many as one third of Euro-
pean forest species depend on deadwood for their sur-
vival (Bo o d y 2001; s i it o n e n 2001). The uprooting of
trees creates not only deadwood, but also a pit-and-
mound microrelief with exposed root-plates, bare min-
eral soil and humus, which are also important micro-
habitats for many species (Ul a n o v a 2000; Bo Ug et and
dUel l i 2004; l õ HmUs et al. 2010). The inventory data of
the Uholka-Shyrokyi Luh primary beech forest may
serve as a reference for such habitat elements when
evaluating the ecological value of managed forests.
Deadwood, stumps and root-plates
The average number of dead standing trees was 30 ±
2N per ha. 70% of them were broken (snags), and only
18% still had their complete crown including fine
branches (Table 6.5). In addition to the dead standing
trees and snags, there were 3 ± 1 stumps (50–129 cm
high) per ha. The average height of stem breakage of
snags was 5.25 ± 0.01 m. The difference in the number
of dead standing trees and snags between Uholka and
Shyrokyi Luh is statistically not significant. In both
administrative units, the ratio of living to dead trees and
snags was the same (14:1).
The proportion of dead trees per diameter class was
more-or-less equal for trees up to a DBH of 60 cm
(6%), and twice as much for trees larger than 60 cm
DBH (12%) (Table 6.6). The three DBH classes
between 21 and 80 cm had all more-or-less the same
number of dead standing trees (4–5 N/ha), which
abruptly decreased to 1 N/ha in the DBH class 81–100
cm (Fig. 6.13). It seems that most of the beech trees
reach their natural life span before they are 80 cm thick.
As can be seen in Fig. 6.9, the number of living trees
larger than 80 cm DBH also decreases abruptly.
The total volume of deadwood was 163 m3per ha
(Table 6.4), of which 84 % was lying deadwood (Fig.
6.14). The ratio of standing to lying deadwood was the
Fig. 6.11. Survey team posing in front of a giant beech. From left to right: Volodymyr Trotsiuk, Martina Hobi, Igor Cherniuk, Luca Mini
and Jonas Stillhard. Photo M. Hobi.
51
Main results
same in both administrative units Uholka and Shyrokyi
Luh. Fresh and hard deadwood contributed 37% to the
total deadwood volume, and the proportion of more
advanced decay stages was 17% to 27% per stage
(Table 6.7). The decomposition rate depends mainly
on the substrate specific variables tree species and
size, and the environmental variables temperature and
precipitation (Har mo n et al. 1986; Her r mann and
Pr e s Co t t 2008; Zel l et al. 2009). In central Europe,
dead beech trees take about 30 to 60 years to decom-
pose completely (Büt l e r et al. 2005; CHr i s t en s e n et al.
2005; l o mBa r d i et al. 2008; mü l l e r -Us i n g and Bar -
t s CH 2009). The proportion of fresh and hard deadwood
was higher in the standing deadwood than in the lying
deadwood (Fig. 6.15). This may be because lying
deadwood usually has soil contact and thus decays
faster, while dead standing trees tend to break and fall
when decay is more advanced. The volume of dead-
wood, particularly of fresh and still hard deadwood, was
higher in Uholka than in Shyrokyi Luh (Table 6.7). This
indicates that the mortality in Uholka has increased in
the last few years due to endogenous (age) and/or
exogenous factors, such as the storm of March 23/24,
2007 or the heavy wet snow fall of October 14, 2009 in
the area. Even if the deadwood accumulation tends to
be temporally and spatially clustered due to natural
disturbances, the high average deadwood volume and
its more-or-less even distribution across early and
well-advanced decay classes (Table 6.7 and Fig. 6.15)
indicate habitat continuity over the whole study area.
On average, 11 root-plates from fallen trees were
found per ha in each of the two administrative units of
the protected massif (Table 6.8 and Fig. 6.16). They
were most frequent on hillsides with a slope of 21–40 %.
The pit-and-mound topography caused by the uproot-
ing of trees may be still recognizable when the root-
plates are already decomposed. The small mounds left
by the decomposed root-plates indicate former natural
disturbances, such as wind or landslides. 42 such
mounds per ha were identified in the surveyed area.
The similar densities of recent and former root-plates in
Uholka and Shyrokyi Luh suggest the natural distur-
bance regimes in the two administrative units have
been similar for a long time. Recent and decomposed
root-plates (mounds) seem to be very rare on steep
slopes, maybe due to erosion and gravity effects or
different soil properties.
Fig. 6.12. Deadwood and old trees provide important
habitats for many species. From top down: Rosalia alpina,
Bielzia coerulans, Lobaria pulmonaria, Hericium corraloides.
Photos: M. Brüllhardt, J. Bürgi, L. Mini.
52 Inventory of the Largest Primeval Beech Forest in Europe
Fig. 6.13. Number of dead standing trees and snags according to
DBH-class in the administrative units. Error bars = standard error.
Table 6.5. Number of dead standing trees, snags and stumps per ha (in total, 460 dead standing trees and snags and 49 stumps were
assessed). SigniÀcant differences (t-test) between Uholka and Shyrokyi Luh are indicated with * (* P ≤0.05, ** P ≤0.01, *** P ≤0.001).
Category Uholka
[N/ha ± SE]
Shyrokyi Luh
[N/ha ± SE]
Whole study area
[N/ha ± SE]
Complete trees with crowns including Àne branches 5.1 ± 1.3 5.2 ± 1.0 5.2 ± 0.8
Complete stems with remains of coarse branches 4.6 ± 1.6 3.2 ± 0.6 3.82 ± 0.8
Snags 16.9 ± 2.3 * 23.8 ± 1.9 20.6 ± 1.5
All dead standing trees and snags 26.6 ± 3.8 32.2 ± 2.6 29.6 ± 2.2
Stumps 2.8 ± 0.6 3.5 ± 0.7 3.2 ± 0.5
Table 6.6. Number of dead standing trees and snags according to
DBH-class and in % of all standing trees (living and dead).
DBH-class Whole study area
[N/ha ± SE] % of all trees
6–20 cm 16.4 ± 2.0 6.0
21–40 cm 4.5 ± 0.5 5.0
41–60 cm 3.2 ± 0.4 5.8
61–80 cm 4.4 ± 0.5 11.9
81–100 cm 1.0 ± 0.3 12.4
101–120 cm 0.1 ± 0.1 12.7
> 120 cm 0.0 ± 0.0 0.0
All DBH classes 29.6 ± 2.2 6.4 Shyrokyi Luh
Uholka
0
5
10
15
20
25
6–20 21–40 41–60 61–80 81–100 101–120
Dead standing trees and snags
[N/ha]
DBH-class [cm]
Fig. 6.14. Lying deadwood covered with Trichaptum biforme, a saprobic fungus commonly found on hardwood logs and snags. Photo M.
Brüllhardt.
53
Main results
Table 6.7. Total deadwood volume (standing1and lying2) according to decay stage. SigniÀcant differences (t-test) between Uholka and
Shyrokyi Luh are indicated with * (* P ≤0.05, ** P ≤0.01, *** P ≤0.001).
Decay stage Uholka Shyrokyi Luh Whole study area
[m3/ha ± SE] [m3/ha ± SE] [m3/ha ± SE] [%]
Fresh deadwood 22.2 ± 6.3 * 9.1 ± 2.2 15.1 ± 3.2 9.3
Hard deadwood 52.0 ± 7.2 37.6 ± 4.5 44.3 ± 4.1 27.2
Rotten wood 34.1 ± 5.7 29.8 ± 3.5 31.8 ± 3.2 19.8
Mouldering wood 42.9 ± 4.9 43.6 ± 5.6 43.2 ± 3.7 26.5
Mull wood 29.4 ± 4.1 25.5 ± 3.8 27.3 ± 2.8 16.7
Not speciÀed 0.0 ± 0.0 1.5 ± 1.1 0.8 ± 0.6 0.6
Total deadwood 180.5 ± 12.9 * 147.1 ± 10.7 162.5 ± 8.4 100.0
1Standing dead trees and snags ≥6 cm DBH, 2Fallen trees and coarse woody debris ≥7 cm diameter.
0 %
10 %
20 %
30 %
40 %
50 %
60 %
70 %
80 %
90 %
100 %
Standing1Lying2
27 m3/ha 136 m3/ha
% of deadwood volume
Fresh deadwood
Hard deadwood
Rotten wood
Mouldering wood
Mull wood
Not specified
1Standing dead trees and snags ≥6 cm DBH.
2Fallen trees and coarse woody debris ≥7 cm
diameter.
Fig. 6.15. Percentage distribution of decay stage in
standing and lying deadwood. Whole study area.
Fig. 6.16. Small windthrow area with uprooted and broken trees. The pit-and-mound relief with exposed root-plates provides a variety of
microhabitats. Photo M. Brüllhardt.
54 Inventory of the Largest Primeval Beech Forest in Europe
Table 6.9. Number, proportion and mean DBH of living trees with microhabitats. SigniÀcant differences (t-test) between Uholka and
Shyrokyi Luh are indicated with * (* P ≤0.05, ** P ≤0.01, *** P ≤0.001).
Type of microhabitat Uholka
[N/ha ± SE]
Shyrokyi Luh
[N/ha ± SE]
Whole study area Mean DBH
[N/ha ± SE] % of all trees [cm ± SE]
Crown with deadwood 82.3 ± 13.0 81.6 ± 10.1 81.9 ± 8.1 18.8 23.4 ± 1.0
Broken crown 38.2 ± 3.7 41.1 ± 2.9 39.8 ± 2.3 9.1 20.7 ± 0.9
Broken stem 11.5 ± 1.8 ** 5.8 ± 0.9 8.4 ± 1.0 1.9 20.7 ± 1.7
Polypores 1.3 ± 0.4 * 3.0 ± 0.6 2.2 ± 0.4 0.5 63.8 ± 6.2
Bark damage (bare wood core) 19.3 ± 2.5 *** 33.0 ± 2.7 26.7 ± 1.9 6.1 40.8 ± 1.6
Cracks in wood core 9.9 ± 1.5 9.7 ± 1.2 9.8 ± 0.9 2.2 41.8 ± 2.4
Hole 11.6 ± 1.7 15.3 ± 1.9 13.6 ± 1.3 3.1 35.3 ± 1.9
Cavity with mull wood 9.1 ± 1.2 10.7 ± 1.3 9.9 ± 0.9 2.3 43.8 ± 2.3
Hollow stem 2.8 ± 0.8 4.6 ± 0.8 3.8 ± 0.6 0.9 56.7 ± 4.2
Any type of microhabitat 142.6 ± 12.1 156.5 ± 9.6 150.1 ± 7.6 34.5 26.5 ± 0.8
Total number of trees assessed 414.5 ± 19.4 452.5 ± 15.2 435.0 ± 12.2 24.8 ± 0.5
Table 6.8. Number of recent root-plates and mounds (decomposed former root-plates) according to slope. SigniÀcant differences (t-test)
between Uholka and Shyrokyi Luh are indicated with * (* P ≤0.05, ** P ≤0.01, *** P ≤0.001).
Slope [%] Uholka [N/ha] Shyrokyi Luh [N/ha] Whole study area [N/ha]
Root-plates
≤20 2.0 ± 1.3 35.0 ± 35.0 7.5 ± 5.8
21– 40 14.0 ± 3.7 12.0 ± 3.2 13.2 ± 2.5
41– 60 3.3 ± 1.3 4.7 ± 1.3 4.0 ± 0.9
61– 80 4.2 ± 1.8 6.4 ± 1.2 5.6 ± 1.0
81–100 0.0 ± 0.0 0.0 ± 0.0 0.0 ± 0.0
All slopes 10.6 ± 2.2 10.4 ± 1.5 10.5 ± 1.3
Mounds
(former root-plates)
≤20 24.0 ± 7.9 6.0 ± 6.0 21.0 ± 6.9
21– 40 34.5 ± 6.5 30.1 ± 6.1 32.7 ± 4.6
41– 60 27.8 ± 4.7 35.1 ± 4.4 31.8 ± 3.2
61– 80 15.4 ± 3.6 23.7 ± 4.5 20.6 ± 3.2
81–100 – 5.0 ± 3.3 4.0 ± 2.7
All slopes 41.5 ± 5.2 41.8 ± 4.2 41.7 ± 3.3
Fig. 6.18. Cavity with mull wood. Photo U.-B. Brändli.
55
Main results
Habitat trees
Habitat trees are living trees with special features, such
as broken stems, cracks, holes or cavities, that provide
microhabitats for specialised animals and plants (Fig.
6.17). In both units of the massif, 35% of all living trees
featured at least one of the microhabitats listed in Table
6.9. Apart from bark damage and stem breakages,
most of the microhabitats assessed were similarly
frequent in Uholka and Shyrokyi Luh (Table 6.9). Bark
damage was much more frequent in Shyrokyi Luh,
whereas trees with broken stems were more numerous
in Uholka. The latter might indicate a natural disturb-
ance in recent decades with a core area in Uholka, e.g.
the wet snow event of October 2009, which caused
considerable damage at lower altitudes.
The most prevalent microhabitat type was deadwood
in the crown of trees, followed by broken crowns and
bark damage, whereas hollow stems or polypores
(bracket fungi) are rare. Only one tree in a hundred had
a hollow stem. Some of the microhabitats observed are
independent of tree DBH: crown deadwood, broken
crowns and broken stems. Trees with such microhabi-
tats had a mean DBH of 21–23 cm, which was even a
bit smaller than the average for the whole population
(25 cm). Bark damage, cracks, holes and cavities with
mull wood (Fig. 6.18) occurred more frequently in thicker
trees (mean DBH 35–44 cm), and hollow stems even
had a mean DBH of 57 cm and trees with polypores of
64 cm. While most types of microhabitats are related to
tree age and occur mainly in old trees, microhabitats,
such as broken crowns and stems and crown dead-
wood, may result from natural disturbances and can
thus also be found in younger trees.
Polypores, holes, cavities with mull wood and hollow
stem were assessed not only in living, but also in dead
standing trees and snags (Table 6.10). Whereas poly-
pores and holes were frequently found on dead trees
and snags (on 31% and 23%, respectively), cavities
with mull wood and hollow stems were quite rare (in
9% and 6%, respectively).
Habitat trees with m icrohabitats
1 Crown with deadwood ≥10 % of crown volume
2 Broken crown (branches or bole) ≥10 % of crown volume
3 Broken stem
4 Polypores
5 Bark damage (bare wood core) ≥300 cm2
6 Cracks in wood core ≥1 m length
7 Hole in wood core with a diameter ≥3 cm and a depth ≥5 cm
8 Cavity with mull wood at the stem base below 1.5 m in height
9 Hollow stem with a hollow diameter ≥50 % of tree diameter
Fig. 6.17. DeÀnitions of the microhabitats surveyed on living
trees. Drawing: Yvonne Rogenmoser.
Table 6.10. Number of dead standing trees and snags with microhabitats. SigniÀcant differences (t-test) between Uholka and Shyrokyi
Luh are indicated with * (* P ≤0.05, ** P ≤0.01, *** P ≤0.001).
Type of microhabitat Uholka
[N/ha ± SE]
Shyrokyi Luh
[N/ha ± SE]
Whole perimeter
[N/ha ± SE]
% of all standing dead
trees and snags
Polypores 6.9 ± 1.2 ** 11.3 ± 1.1 9.3 ± 0.8 31.4
Hole 5.1 ± 0.9 * 8.0 ± 1.1 6.7 ± 0.7 22.6
Cavity with mull wood 2.1 ± 0.5 3.0 ± 0.7 2.6 ± 0.4 8.8
Hollow stem 0.8 ± 0.3 * 2.4 ± 0.6 1.7 ± 0.4 5.7
Total trees and snags assessed 26.6 ± 3.8 32.2 ± 2.6 29.6 ± 2.2
2
8
9
1
3
5
6
7
4
56 Inventory of the Largest Primeval Beech Forest in Europe
6.5 Conclusions
The inventory provides estimates of the main forest
parameters with a precision of 5% (number of trees,
basal area and volume of living trees) to 10% (volume
of deadwood), at a confidence level of 95%. The esti-
mates thus serve as good reference points for old-
growth beech forests under similar conditions. The
estimates are only precise enough for beech and, to a
lesser extent, for sycamore. For other tree species and
for rare occurrences, such as hollow stems, the esti-
mates are not precise enough and the sample size
should be enlarged.
Many of the features we found are typical charac-
teristics of old-growth forests with gap dynamics, domi-
nated by a small-scale disturbance regime (BaUHUs
et al. 2009; w ir t H et al. 2009). They include a multi-
layered forest structure with mainly small canopy gaps,
a large tree diameter range (up to 150 cm dbh), the
presence of up to 500-year-old beech trees, high grow-
ing stock (582 m3/ha), a high amount of standing and
lying deadwood (163 m3/ha) of all decay classes, and a
high density of habitat trees (150/ha; 35% of the living
trees). The shape of the diameter distribution curve, the
area-wide occurrence of regeneration of all species
present in the overstorey and the absence of significant
differences in the main forest parameters between the
two administrative units of Uholka and Shyrokyi Luh
indicate that the forest as a whole is in a steady state
and likely to maintain its structure in the long run. The
tree species composition and the forest structure
observed both suggest that little timber apart from the
occasional tree has ever been cut in the core zone. It
seems, therefore, that the integrity and pristine charac-
ter of the forest in Uholka and Shyrokyi Luh have been
maintained. This fact, together with the vast size of the
area, makes the forest of Uholka and Shyrokyi Luh one
of the most valuable remnants of primeval forest in
Europe. The traces of human presence (mainly in the
buffer and transition zones, but in some places also in
the core zone) show, however, the pressure exerted
from the nearby settlements, on the one hand, and the
“poloninas” (mountain pastures) on the other. The latter
are not only used by the traditional shepherds, but
also by commercial berry pickers. Tracks suitable for
off-road trucks present a threat as they many also
facilitate illegal activities like poaching and logging. The
situation on the border of the reserve should be moni-
tored continuously to identify any negative impacts
early.
References
al eSSan d Rin i, a.; Bio n d i, f .; d ifil iPPo , a.; zia c o , e.; Pio VeSan ,
g ., 2011: Tree size distribution at increasing spatial scales
converges to the rotated sigmoid curve in two old-growth
beech stands of the Italian Apennines. For. Ecol. Manage.
262: 1950–1962.
BauhuS, J.; Pu et t ma n n , k.; meSSieR, c ., 2009: Silviculture for
old-growth attributes. For. Ecol. Manage. 258: 525–537.
Bo o d Y, l ., 2001: Fungal community ecology and wood de-
composition processes in angiosperms: from standing
tree to complete decay of coarse woody debris. Ecologi-
cal Bulletin 49: 43–56.
BRän d l i, u .-B.; BeRan o Va, J., 2011: 3.4 Forest age. In: ch iRic i,
g .; Win t eR, S.; mcRo BeRt S, R.e. (eds) National Forest In-
ventories: Contributions to Forest Biodiversity Assess-
ments. Managing Forest Ecosystems 20: 56–60.
Bo u g e t , c .; d ue l l i, P., 2004. The effects of windthrow on for-
est insect communities: a literature review. Biol. Conserv.
118: 281–299.
Bü t l eR, R.; l a c h at , t .; Sc h l a ePf eR, R., 2005: Grundlagen für
eine Alt- und Totholzstrategie der Schweiz. Laboratorium
für Ökosystemmanagement, Ecole polytechnique fédé-
rale, Lausanne. Im Auftrag des Bundesamtes für Umwelt,
Wald und Landschaft. 100 pp.
ch RiSt en Sen , m.; h ah n , k .; mo u n t f o Rd , e.P.; o d o R, P.;
St an d o VaR, t .; Ro ze n BeRg a R, d.; d iac i , J.; WiJde Ven , S.;
meYeR, P.; Win t e R S.; VRSka , t ., 2005: Dead wood in Euro-
pean beech (Fagus sylvatica) forest reserves. For. Ecol.
Manage. 210: 267–282.
co c h Ra n , W.g ., 1977: Sampling techniques. 3rd Edition.
Wiley Series in Probability and Mathematical Statistics.
John Wiley & Sons, New York. USA. 428 pp.
el l en BeRg, h.; WeBeR, h.e.; d ül l , R.; WiRt h , V.; WeRn eR, W.;
Pau l iSSen , d ., 1992: Zeigerwerte von PÅanzen in Mittel-
europa. 2. verbesserte und erweiterte AuÅage. Scr. Geo-
bot. 18: 258 pp.
eWal d , J., 2007: Ein pÅanzensoziologisches Modell der
Schattentoleranzvon Baumarten in den Bayerischen Alpen.
Forum geobotanicum 3: 11–19.
go f f , f .g .; WeSt , d ., 1975: Canopy-understory interaction
effects on forest population structure. For. Sci. 21: 98–
108.
haRmo n , m.e.; f Ran k l i n , J.f.; SWan So n , f .J.; So l l inS, P.,
gReg o RY, V.; l at t in , J.d.; an d eRSo n , n .h .; c l in e, S.P.;
au men , n .g ., l i en k ae mPeR, g .W.; c Ro mac k , k.J.; c u mmin S,
k .W., 1986: Ecology of Coarse Woody Debris in Temper-
ate Ecosystems. Adv. Ecol. Res. 15: 133–302.
heRRma n n , S.; PReSc o t t , c .e., 2008: Mass loss and nutrient
dynamics of coarse woody debris in three Rocky Moun-
tain coniferous forests: 21 year results. Can. J. For. Res.
38: 125–132.
ko RPel ’, Š., 1995: Die Urwälder der Westkarpaten. G. Fischer,
Stuttgart, Jena, New York. 310 pp.
ku c Be l , S.; San ig a , m.; Ja l o Via R, P.; Ven c u Rik , J., 2012:
Stand structure and temporal variability in old-growth
beech-dominated forests of the northwestern Carpathi-
ans: A 40-years perspective. For. Ecol. Manage. 264: 125–
133.
laSSa u c e, a .; Pail l et , Y.; Ja c t el , h .; Bo u g e t , c ., 1011:
Deadwood as surrogate for forest biodiversity: Meta-
analysis of correlations between deadwood volume and
57
Main results
species richness of saproxylic organisms. Ecol. Indic. 11:
1027–1039.
lõ h mu S, P.; t u RJa, k .; l õ h mu S, a., 2010: Lichen communities
on treefall mounds depend more on root-plate than stand
characteristics. For. Ecol. Manag. 260: 1754–1761.
lo mB aRd i, f .; c h eRu Bin i, P.;l a SSeRRe, B.; t o g net t i, R.; maR-
c he t t i, m., 2008: Tree rings used to assess time since
death of deadwood of different decay classes in beech
and silver Är forests in the central Apennines (Molise, Italy).
Can. J. For. Res. 38: 821–833.
mü l l eR, J.; Bü t l eR, R., 2010: A review of habitat thresholds
for dead-wood: a baseline for management recommenda-
tions in European forests. Eur. J. For. Res. 129: 981–992.
mü l l eR-u Sin g , S.i.; Ba Rt Sc h , n ., 2009: Decay dynamics of
coarse and Äne woody debris of a beech (Fagus sylvatica
L.) forest in Central Germany. Eur. J. For. Res. 128: 287–
296.
SäRn d a l , c .e.; SWen SSo n , B.; WRet ma n , J., 1992:. Model As-
sisted Survey Sampling. Springer Series in Statistics.
Springer, New York, USA. 694 pp.
Siit o n e n , J., 2001: Forest management, coarse woody debris
and saproxylic organisms: Fennoscandian boreal forest as
an example. Ecological Bulletin, 49, 11–41.
tRo t Siu k , V.; h o B i, m.l .; c o mm aRmo t , B., 2012: Age struc-
ture and disturbance dynamics of the relic virgin beech
forest Uholka (Ukrainian Carpathians). For. Ecol. Manage.
265: 181–190.
ul an o Va, n .g., 2000: The effects of windthrow on forests at
different spatial scales: a review. For. Ecol. Manage. 135:
155–167.
WeSt Ph al , c .; t RemeR, n.; o h e imB, g .V.; h a n Sen , J.; g a d o W,
k . V.; h äRd t l e, W., 2006: Is the reverse J-shaped diameter
distribution universally applicable in European virgin
beech forests? For. Ecol. Manage. 223: 75–83.
WiRt h , c .; meSSieR, c .; Be Rg eRo n , Y.; f Ran k , d .; f a n k h än e l ,
a. 2009: Old-growth forest deÄnitions: a pragmatic view.
In: WiRt h , c .; g l e ix n eR, g .; h eima n n , M. (eds) Old-growth
forests. Springer, Berlin, Heidelberg. 11–33.
zel l , J.; k än d l eR, g .; han e Win k el , m., 2009: Predicting con-
stant decay rates of coarse woody debris – A meta-analy-
sis approach with a mixed model. Ecol. Model. 220: 904–
912.
In this publication, we presented the main results of the
2010 inventory in the Uholka-Shyrokyi Luh massif,
Ukraine. Further analyses will include comparisons
with other old-growth forests and managed forests, as
well as structural analyses of the data we collected on
various scales. In an ongoing PhD study, high-resol-
ution stereo satellite images are being used to calculate
a canopy surface model of the whole primeval beech
forest of Uholka-Shyrokyi Luh. This model will be com-
bined with point data from the terrestrial inventory to
provide a basis for characterising the forest structure,
identifying canopy gaps and analysing the large-scale
disturbance regime.
The inventory was carried out and documented in
such a way that it can be repeated, if desired. This
allows the development of the forest to be monitored
and any changes identified. Since the small-scale
dynamics of old-growth forests may have considerable
impact, the wooden poles marking the sample plot
centres should be replaced after 5 to 10 years even if
no second inventory is planned. Otherwise it could be
difficult to relocate the exact position of the sample plot
centres again even though their coordinates and the
positions of all trees are known.
The sample plots may also be used for other (non-
destructive) studies/inventories, which might benefit
from comparison with the forest data collected. For
example, a large-scale population study of the tree
lungwort Lobaria pulmonaria has already been made,
as has a floristic study of other lichens growing on trees
(ongoing project of WSL and the M.G. Kholodny Insti-
tute of Botany of the National Academy of Sciences of
Ukraine in Kyiv) on and around the sample plots.
We hope that this inventory has helped raise interest
in this primeval forest and that it will provide a useful
basis and impetus for further virgin forest research.
“This virgin beech forest may be eternal. It is a big,
wild and natural forest I did not know we had in
Ukraine. We used horse transport and lost our way at
times, but no such huge primeval (beech) forest can
be found anywhere else in Europe.”
Igor Cherniuk, student UNFU, Ukraine
Hollow beech tree. Photo L. Denzler.
7 Outlook
Brigitte Commarmot, Urs Beat Brändli, Fedir Hamor, Vasyl Lavnyy
61
Appendix
Appendix 1
Project organisation and division of tasks
Project management: Brigitte Commarmot, WSL
Coordination Ukraine: Vasyl Lavnyy, UNFU
Supervision: Peter Brang, WSL, Fedir Hamor, CBR
ScientiÀc work:
MeinradAbegg, WSL (data-base, data analyses)
Yuriy Berkela, CBR (GIS, maps)
Urs-Beat Brändli, WSL (inventory consulting, sampling method,
Àeld manuals, data analyses)
Brigitte Commarmot, WSL(sampling method, Àeld manuals,
Àeld forms, data analyses)
Christian Ginzler, WSL (remote sensing, GIS, maps)
Martina Hobi, WSL (GIS, maps, tree-ring data, data analyses)
Edgar Kaufmann, WSL (volume tariffs)
Mykola Korol, UNFU (Àeld manuals, data analyses)
Adrian Lanz, WSL (inventory design, statistical advice)
Raphaela Tinner (sampling method, Àeld manual)
Translation:
Victoria Gubko, CBR (English/Ukrainian)
Mykola Korol, UNFU (German/Ukrainian)
Vasyl Lavnyy, UNFU (German/Ukrainian)
Preparation of logistics:
Vasyl Pokynchereda, CBR
Dmytro Sukharyuk, CBR
Victoria Gubko, CBR
Vasyl Rehush, CBR (forest manager, Uholka)
Valeriy Feyer, CBR (forest manager, Shyrokyi Luh)
Vasyl Lavnyy, UNFU
Instruments and equipment:
Brigitte Commarmot, WSL
Christoph Düggelin, WSL
Martina Hobi, WSL
Vasyl Lavnyy, UNFU
Safety measures:
Victoria Gubko, CBR
Martina Hobi, WSL
Ruedi Iseli, Hasspacher & Iseli Gmbh
Vasyl Lavnyy, UNFU
Vasyl Pokynchereda, UNFU
Field-work manager: Ruedi Iseli, Hasspacher & Iseli Gmbh
Field teams:
Martin Brüllhardt, student ETHZ
Janine Bürgi, student ZHAW
Lukas Glanzmann, student ETHZ
Andrea Grimmer, student ZHAW
Silas Hobi, student ETHZ
Oliver Leisibach, student ZHAW
Luca Mini, student ZHAW
Daniel Oertig, student HAFL
Matthias Ofner, student ZHAW
Jonas Stillhard, student ZHAW
Igor Cherniuk, student UNFU
Serhiy Gavryliuk, assistant professor UNFU
Andriy Khomiuk, student UNFU
Vasyl Kostyshyn, PhD student UNFU
Nataliia Rehush, junior researcher CBR
Volodymyr Trotsiuk, student UNFU
Group leaders:
Ruedi Iseli, Hasspacher & Iseli Gmbh
Mykola Korol, lecturer UNFU
Martina Hobi, PhD student WSL
Instruction of Àeld teams:
Ruedi Iseli, Hasspacher & Iseli Gmbh
Mykola Korol, UNFU
Martina Hobi, WSL
Brigitte Commarmot, WSL
Raphaela Tinner, WSL
Meinrad Abegg, WSL
Data entry:
Vasyl Lutsyshyn, PhD student UNFU
Volodymyr Savchyn, PhD student UNFU
Administration, book-keeping:
Beate Hasspacher, Hasspacher & Iseli GmbH
Logisitic support Uholka:
Vasyl Rehush (administrative support)
Vitaliy Motriya (administrative support)
Mykhaylo Rehush (administrative support)
Vasyl Krychfalushy (transport with horses)
Mykhaylo Tanchynez (transport with horses)
Ivan Dobrunyk (transport with truck)
Vasyl and Magdalyna Semyanovskiy (cooking)
Tetyana Rehush (accommodation, cooking)
Mykhaylo Nemesh (transport, cooking)
Mariya and Vasyl Nemesh (cooking)
Tetyana Nemesh (cooking)
Logisitic support Shyrokyi Luh:
Valeriy Feyer (administrative support)
Vasyl Mula (administrative support)
Ivan Tanchynez (transport with horses)
Andriy Dudla (transport with horses)
Ivan Moskal (transport with truck)
Petro Pokovba (transport)
Ivan and Olena Moksal (cooking)
Volodymyr and Marija Bohdan (cooking)
Ivan Oleksiy (cooking)
Vasyl Drahun (transport, cooking)
Trametes versicolor. Foto M. Brüllhardt.
62 Inventory of the Largest Primeval Beech Forest in Europe
Appendix 2
Survey equipment (for each survey team)
Documents
1 Overview map on a scale of 1:20000 (based on GIS
data, with orthophotos from 2008 as background data)
covering the whole study area
32 Sub-maps on a scale of 1:8000
2 Field manuals (German and Ukrainian)
Sets of forms on normal and waterproof paper (6
forms per plot in German and/or English; reference
forms in Ukrainan)
3 Cards with codes for sample trees (German,
Ukrainian and English)
2 Cards with codes for tree and shrub species
2 User’s Manual with brief instructions for Vertex and
Transponder, Criterion RD 1000, GPS Garmin etrex
summit HC and GPS Trimble GeoXH/Juno SB
(German and English)
2 Safety cards with emergency phone numbers
Rules of conduct in the field
Stationery
Writing material
Marking chalk
Lumber crayon with holder
1 Calculator HP 10s
Batteries for Vertex/Transponder and Garmin GPS
Equipment and tools
2 Backpacks
2 Cruiser vests
1 Pocket knife (Victorinox Forester)
1 Club hammer
1 First aid kit
2 Alarm whistles
2 Oak poles (4 x 4 x 40 cm; one with and one without
marking) per sample plot
2 Mobile phones (personnel equipment)
Survey instruments and material
1 Clinometer SUUNTO (PM 5/400 PC)
1 Compass SUUNTO (KB 14/400 gon)
1 Compass Wyssen MERIDIAN MI-4007, 400 gon
1 Tripod for the Wyssen compass
1 Plumbline
1 Hypsometer Vertex IV (or Vertex III) (to measure
distances and tree heights)
1 Transponder for Vertex
1 Criterion RD 1000 (only 1 instrument for 2 teams)
1 Monopod Gitzo GM1130MT (for Criterion)
1 GPS Garmin etrex SUMMIT HC
1 GPS Trimble GeoXH or Trimble JUNO SB with
TerraSync Standard Edition Software
1 Calliper Haglöff MANTAX blue 80 cm
1 Diameter tape
1 Measuring tape (logger’s tape with automatic
rewind) 20 m (steel)
1 Measuring tape 50 m (fibreglass)
1 Digital camera, Sony Cyber-Shot DSC-WX1
1 Memory Stick, Pro Duo 4GB Sandisk
2 Plastic callipers for young trees (0 to 15 cm with 1
cm graduations)
2 Folding metre rules
2 Ranging poles (2 m)
63
Appendix
Appendix 3
Distribution of anthropogenic traces
The map below shows the distribution of plots where anthropogenic traces (one or more types) were found. The
number of different traces recorded may be used as an indicator of the degree of anthropogenic pressure and
impact. For the distribution of the different types of anthropogenic traces, see the maps on the next double page.
Number of different anthropogenic traces
No traces
1 type of traces
2 - 3 types of traces
4 - 5 types of traces
Inaccessible plot
Regulated protection zone
Core zone
CBR perimeter
0 2 4 km
°
Buffer zone
Anthropogenic landscapes
64 Inventory of the Largest Primeval Beech Forest in Europe
Timber cutting Roads and paths
Livestock grazing, pasturing Fire, traces of burning
Absence of traces
Inaccessible plot
Presence of traces
0 2.5 5 km
CBR perimeter
Regulated protection zone
Core zone
°
Buffer zone
Anthropogenic landscapes
65
Appendix
CBR perimeter
Regulated protection zone
Core zone
Litter, waste Anthropogenic damage to trees
Research, monitoring Other anthropogenic traces
Presence of traces
Absence of traces
Inaccessible plot
0 2.5 5 km
°
Buffer zone
Anthropogenic landscapes
66 Inventory of the Largest Primeval Beech Forest in Europe
Appendix 4
List of tree species assessed
Number of living trees assessed per species during the sampling inventory 2010. RC (regeneration class) 1:10–39 cm high; RC 2:
40–129 cm high; RC 3: ≥130 cm high and < 6 cm DBH.
Species Trees ≥6 cm dbh
(500 m2circle)
Regeneration (saplings)
RC 1 (5 m2circle) RC 2 (10 m2circle) RC 3 (20 m2circle)
Abies alba Mill. 16 1 2 4
Acer sp. 1
Acer platanoides L. 14 15 4 7
Acer pseudoplatanus L. 46 69 26 21
Betula pendula Roth 8
Carpinus betulus L. 104 1 4 1
Corylus avellana L. 19 2 1
Fagus sylvatica L. 6531 243 235 618
Fraxinus excelsior L. 7 2 1
Picea abies Karst. 1
Prunus avium L. 1 1
Quercus petraea Liebl. 1 2
Salix sp. 1 1
Salix caprea L. 9
Sambucus nigra L. 8
Sorbus aucuparia L. 1
Ulmus glabra Huds. 13 5 4 5
Total 6779 341 278 658
67
Appendix
Appendix 5
Distribution of tree species
The maps on this and the following double page show the number of tree species per sample plot and the distribution
of the main tree species found. “Presence” means that at least one living individual of the species was recorded on
the plot, either in the population of trees ≥6 cm DBH or in the regeneration.
Number of tree species per plot
0 2 4 km
°
Buffer zone
Anthropogenic landscapes
Regulated protection zone
Core zone
CBR perimeter
Inaccessible plot
5 - 7 species
2 - 4 species
1 species
68 Inventory of the Largest Primeval Beech Forest in Europe
CBR perimeter
Regulated protection zone
Core zone
Fagus sylvatica Acer pseudoplatanus
Acer platanoides Carpinus betulus
0 2.5 5 km
°
Buffer zone
Anthropogenic landscapes
Inaccessible plot
Absence of species
Presence of species
69
Appendix
Ulmus glabra Fraxinus excelsior
Abies alba Other living species
Regulated protection zone
Core zone
CBR perimeter
0 2.5 5 km
°
Buffer zone
Anthropogenic landscapes
Inaccessible plot
Absence of species
Presence of species
In 2010, the Swiss Federal Institute for Forest, Snow and Landscape Research WSL,
the Ukrainian National Forestry University and the Carpathian Biosphere Reserve carried out
an inventory of the Primeval Beech Forest of Uholka-Shyrokyj Luh in the Ukrainian Carpathians.
This report describes the sampling design and the parameters assessed, the planning and
organisation of the Àeld work, and the management and analysis of the data collected.
It presents Àndings about basic forest characteristics, such as species composition, tree
densities, growing stock, volumes of standing and lying deadwood, forest structure, regeneration
density and density of habitat trees, but also about the anthropogenic traces found within the
survey perimeter of 10282 ha. This report should be particularly useful for people interested in
reference values from primeval forests and researchers planning a similar inventory in remote
areas.
