Tree Rings

Fritz Hans Schweingruber

Tree Rings
Basics and Applications of Dendrochronology

Kluwer Academic Publishers

Dordrecht / Boston / London
Library of Congress Cataloging-in-Publication Data

Schweingruber, Fritz Hans.

[Jahrring. English]
Tree rings: basics and applications of dendrochronology / Fritz Hans Schweingruber.
p. cm.
Translation of: Der Jahrring.
Bibliography: p.
Includes index.
ISBN -13: 978-0-7923-0559-0 e-ISBN -13: 978-94-009-1273-1
DOl: 10.1007/978-94-009-1273-1
1. Tree-rings. 2. Dendrochronology. I. Title.
QK477.2.A6S39131987
582.16'0116 - dc19 87-25648
CIP

Reprinted in 1989

Originally published by Paul Haupt, Berne, Switzerland in 1983 in

German under the title Oer Jahrring. This first English edition is a
revised and expanded translation from the German.

Published by Kluwer Academic Publishers

P.O. Box 17, 3300 AA Dordrecht, Holland

Sold and distributed in the U.S.A. and Canada

by Kluwer Academic Publishers
101 Philip Drive, Norwell, MA 02061, U.S.A.

In all other countries, sold and distributed by

Kluwer Academic Publishers Group
P.O. Box 322, 3300 AH Dordrecht, Holland

All Rights Reserved. English Edition

© 1988 Kluwer Academic Publishers, Dordrecht, Holland
Reprint of the original edition 1988
No part of the material protected by this copyright notice
may be reproduced or utilized in any form or by any means,
electronic or mechanical, including photocopying, recording,
or by any information storage and retrieval system,
without written permission from the copyright owner.
Contents

Detailed Table of Contents VI

Preface XII

Preface to the English Edition XIII

Looking to the Future XIV

Introd uction 1
I Origin of the materials 5
II Analysis of the materials 39
III Tree-ring growth and the site 95
IV Applied dendrochronology 143
V History of dendrochronology 255
Bibliography 262
General Index 273
Table of Contents Characterization of the most important tree
species in dendrochronology: Bristlecone
pine, oak, larch, spruce 27
Pinus longaeva D. K. Bailey. Bristlecone Pine 27
I Origin of the materials 5 Ouercus robur L.I Ouercus petraea Wi lid.
Common Oak and Sessile Oak 28
The site and the tree 6 Larix decidua Miller/ Picea abies (L.) Karsten.
Larch and Norway Spruce 29
The tree as integrator 6 Origins of the historic wood 30
Reaction time 6 Origin of the sub-fossil wood 31
Transformation into structure 7
State of preservation of historic and prehistoric
Tree sites on the earth 8 wood 32
Location on the planet 8 Changes occurring with time 32
Location on the continent 9 Decomposition in the presence of air (aerobic
Location in relation to ocean currents 9 decomposition) 32
Elevation 9 In temperate climate zones 32
Influence of environmental factors on tree sites 10 In arid climatic zones 33
In arctic climatic zones 33
Influence of topography 10 Decomposition under low-oxygen conditions
Influence of mechanical movement 10 (anaerobic decomposition) 34
Influence of geological substrate 11 Decomposition by fire 35
Influence of light 11 Changes in fossil wood caused by pressure 36
The site history 12 conservation of fossil wood through
mineralization 36
Site changes in geological periods 12 Artificial preservation of anaerobically
Site changes in prehistoric times 13 decomposed wood 37
Site changes in historic times 14
The product of the changes 15
Selection of the sampling site 16
Measurement of climatic factors at the sampling
II Analysis of the materials 39
site 18
Anatomical techniques 40
Measurement of ecological factors over decades 18
The network of meteorological stations 18 Preparation methods 40
Density of network 18 Sections 40
Limitations 18 Microsections 40
Extrapolation of the data 18 Cutting 41
Measurement of ecological factors over periods Staining the sections . 41
of a few years: Measurements in the stand 19 The microstructure in relation to wood density 41
Ecophysiological measurements 19 Taking samples 42
Measurement of cambial activity 19
Taking cores from living trees 42
Site characterization 20 The use of a borer-holding device as an aid to
Climatic integrators 20 orientation 42
Location and topography, climate, soil,
The labelling and transport of the cores 43
Taking samples from buildings and from fossil
vegetation 20
wood 43
Site description 22 Mounting the cores 43
Sampling site 22 The importance of proper coring for
Samples 22 radiodensitometry 43
Climate 22 Damage resulting from coring 44
Vegetation 22
Sampled tree 22 Damage to living trees 44
Geological substrate 23 The intensity of the damage 44
Soil 23 Damage to wood buildings 45
Damage to the landscape through the excavation
The worldwide sampling network 24 of fossil trunks 45
With living trees 24 Making the tree rings visible 46
Limits of comparability 24
The compromise 25 Material in good condition 46
With historic and prehistoric trees 26 Soft, decomposed wood 46
Charcoal 46
Examining the surfaces 46
Counting tree rings 47 The tree-ring series and the meteorological
Recording the changes of growth patterns 48 series are compared 80
Checking the reconstructions 80
Pointer years and the associated characteristics of
tree rings 49 Basic statistical methods 81
Cross-dating using the skeleton-plot method 50 Characterising tree-ring chronologies 81
Common terms 81
Cross-dating using graphs 51 Arithmetic mean standard deviation 81
The measurement of the tree-ring width 52 Sensitivity 82
Gleichlaufigkeit (sign test) 83
Techniques of tissue analysis 54 Interval trend 83
Applications 54 Correlation, correlation coefficient 84
Methods of structure analysis 54 Standardization and indexing 85
Filters and smoothing functions 86
Photometric techniques 56 Running means 86
Fields of application of these methods 56 Band-pass filter, Polynomial, Straight line,
Calibration 57 Negative exponential, Hugershoff function 87
Polished surfaces 57 Modelling (response functions) 88
Microsections 57 Reconstruction 90
Using single sequences 90
Basic steps in radiodensitometry 58 An example from climatology 91
Sawing 60 An example from the environmental sciences 91
Characterizing the meteorological data 92
Removal of extractives 62 Descriptive statistics 92
Acclimatization 63 Interpolating missing values 92
Determining the homogenity of a series 92
Radiography 64
The radiographic-densitometric measurement of
wood density 66
The equipment 66
III Tree ring growth and the site 95
The principle of radiodensitometric determination
of wood density 68 Growth increments and tree rings in tree trunks 96
Density integrator 69 Growth fluctuations in wood 96
Measuring the density within the tree ring 70 Growth fluctuations in bark 97
Measuring the minimum and the maximum Growth fiuctuations in the rhytidiome 97
density 70 Sensitive tissues and growth processes 98
Measuring of early- and latewood widths 71
The cambium 98
Checking and correcting the data 72
Cell enlargement 98
Possible sources of error 72 Cell wall growth 99
Processing the data 73 Distribution and number of cells 100
Visual sifting and checking the raw data 74 Secondary changes 101
74
Genetically determined differences in reactions 102
Visual agreement between the curves
Conclusions 75 Physiological-ecological amplitude 102
76
Variability in form within a plant 103
Dating in practice
Variability of form within a climatic zone 104
Methods 76
Problems 76
Reactions of a species to climate and site 106
Dating 76 Climate 106
Successful dating 77 External characteristics 107
The accuracy and reliability of dates obtained Internal characteristics 107
dendrochronologically 77 Anatomy of the annual ring 107
Statistics and electronic data processing 78
Density diagram 107
Prerequisites for application 78
The reaction of a flora to climate and site 108
Methods 78 Climate 108
Identification, isolation and elimination of long- External characteristics 108
term factors 78 Internal characteristics 109
Calculation of the characteristic parameters of
a data series 79
Ecologically important features in wood 110 The tree ring in relation to climate 133
Tree-ring width and boundaries 110 Different tree-ring features in relation to climate 134
Species from alpine and arctic zones 110 Density and width values 134
Species from temperate zones Intra-annual density fluctuations 135
111
Species from semi-arid zones 111 The effect of climate and site on tree rings 136
Width of earlywood and latewood and their
The climatological analysis of individual years 137
demarcation 112
Cell size The climatological analysis of time-series 138
113
Evaluating the response functions 138
Cell wall thickness 114
The suitability of different types of site for
Minimum density 115
dendrochronological analysis 139
Difference between maximum and minimum
density 115 A model of the growth of a tree 140
Intra-annual density fluctuations 116
A tree registers both short-term and long-term
The density diagram 117
influences 140
Relationship between tree-ring features 118 The cell wall acts as the information store for
these internal and external stimuli 140
Technical relationships to the tree-ring anatomy 118
The surface of the cell wall has a key role 140
Statistical relationship to the tree-ring anatomy 118
Structure is the manifestation of all metabolic
Biological-ecological relationships 119
processes 141
Climatological-ecological factors 120
The ageing process 121
Relationships between tree-ring features in trees
on the same site 122 IV Applied dendrochronology 143
A. Within one species 122
Tree-ring research in the historical sciences 144
The site 122
Variations in absolute values 122 Important prerequisites for dating 145
Within one tree 122 Terminal tree ring 145
From tree to tree 123 The sapwood-heartwood boundary 146
Visual agreement of width and density curves 124 The relationship between felling date and the
Form and number of intra-annual density utilization of the wood 147
fluctuations 124 Re-use of old timber 148
Long-distance transport of wood 148
B. Between different species 125
The history of house-building in the Neolithic
Differences in absolute values 125
period in Switzerland 149
Characteristic curve patterns 125
Frequency of intra-annual density fluctuations 125 History of settlement in the New Stone Age and
Bronze Age in Switzerland 150
The relationship between individual tree-ring
features in trees on similar and on differing Relative chronologies 150
sites 126 Absolute chronologies 151
The climate in relation to the chronologies 151
The mean agreement between curves 126
Trees of the same species in dry regions 127 Prehistoric settlements in the southwest of North
Trees of the same species at lower elevations 127 America 152
Different species in different zones 127
The development of settlements in Betatakin and
Spatial representation of the agreement 128
Kiet Siel 153
Agreement over a given year 129
Supply of wood 154
Maximum densities 129
Wood in the settlement 155
The number of intra-annual density fluctuations 129
Agreement over longer periods of time 130 Historical buildings 156
Maximum densities 131
Trier Cathedral, Federal Republic of Germany 156
Agreement between the absolute values 131
The Wolf House, Arkansas 157
The agreement between the curves in relation to
distance 131 Sacred and secular buildings in Northern Germany 158
The effect of climate on the tree ring 132 Sacred buildings in Greece and Turkey 159
Tree-ring features 132 Village in an alpine area 160
Climatic variables 132
Material and dating 160
Methods for the comparing of tree-ring features Standard curves 161
with meteorological measurements 132
The pattern of radial growth in relation to weather
conditions 133
Dating old masters 162 The history of glaciers 194
Aims and objectives in art history 162 Age of living trees in the approaches to glaciers 194
The aims and objectives in dendrochronology 162 The effect of ice on living trees 195
Dendrochronological methods 162 The age of subfossil tree stumps 195
The material 163 Places where tree trunks are found 196
The limitation of the method 163 The history of the Great Aletsch glacier in
Dating individual works 164 Switzerland 197
Peter Paul Rubens: 'Child playing with a bird' 164 Slope movements 198
Dating a painted panel from a house in the east
of Switzerland 165 Tree-ring research and wind 200
Dating a series of paintings 165 The relationship to wind velocity 201
Criminal investigation and art dealing 166 The effect of wind direction on the tree 202
The effect of extreme wind conditions on the tree 202
Determining the species or type of wood 166 The effect of wind on environmental conditions 203
Establishing a time sequence 166
Dendroch ronology 166 Forest fires 204
The dating of stringed instruments 167 The effect of fire on the tree-ring pattern 205
The kidnapping of the Lindbergh baby 167 Dating forest fires 206
Dating illegal fellings of trees 167 The effects of forest management on the
Tree-ring research in climatology 168 incidence of fires 208
Conclusions 208
Ways of obtaining climatic information 169 The effect of weather conditions and vegetation
The first stage: construction growth maps 169 on the incidence of forest fires 209
The second stage: obtaining climatic
information from individual series 170 Volcanic activity 210
The third stage: the reconstruction of climatic The direct effects of the eruption of volcanoes 210
conditions over large areas 172 Examples 213
The search for cyclical patterns 174 Mt. St. Helens, Washington, U.S.A. 213
The fourth stage: climatic conditions in earlier Laacher-Volcano, Germany 213
times 174 Katmai, Alaska 214
Tree-ring research and geomorphology 176 Frost rings 214
The effect of ash rain on climatic conditions and
Reaction mechanisms in trees 176 annual ring growth 215
Types of damages 178
The location of wood finds 178 Earthquakes 216
Geomorphological processes which are reflected Giant waves 217
in the growth of a tree 179
Tectonic movements 218
Variability in the run-off of rivers 180
Ice jams and icings 219
Rio Negro, Argentina 180
Colorado River, U.S.A. 181 Tree-ring research on fossilized woods 220
More recent changes in the course and flow of What sort of organisms does the sediment
rivers 182 contain? 220
Where the fossilized woods are found 221
Changes in the course of a river 182 What proportion of the original vegetation do the
Construction of reservoirs 182 finds represent? 221
Woods that were once buried and have since
Seasonality 221
been exposed 183
Tree-ring research in entomology 222
Rivers in prehistoric times 184
Damage to tree-ring patterns 222
Oaks in the post-glacial time in Central Europe 184 Insect damage recurring at regular intervals 223
Vegetation in river valleys 184 Insect damage which does not recur at regular
Changes in rivers 185 intervals 224
Site changes 186 The extent of damage 225
Pines in the Late Ice Age in Switzerland 187 Relationship between a site and the climate 226
Tree finds in peat bogs 188 Relationship between defoliation, radial growth
and mortality rate 227
The finds 188 Conclusions 227
Where the trunks are found 188
Interpreting these findings 189
Changes in water table 190
The history of sea floods 192
The location and age of trees 193
The growth of trees 193
Tree-ring research in phytopathology 228
Fungal attacks 228
The defence mechanism in the tree trunk 228
Damage due to peeling 229
Condition of spruce forests 229
Growth inhibiting effect of Armillaria mellea
(honey fungus) 229
Mistletoe 230
Growth damage 230
Damage 231
Ways of keeping the damage under control 232
Tree-ring research in forestry 232
Thinning 232
Pruning 233
Fertilization and soil improvement 234
Study of yield in tropical forests 236
Methods of determining growth 236
Studying cross-sections of the trunk 236
Age determination of trees in plantations of
known age 237
Determination of growth by dendrometers and
markings 237
Tree-ring research in wood technology 238
Tree-ring research in environmental sciences 240
Methods 240
Dating of sudden changes in growth behavior 240
Regional growth reductions: United States and
Central Europe 241
Regional growth reductions: Wallis, Switzerland 242
Damage due to local environmental factors 244
Growth reduction related to natural
environmental factors 245
Relationship between reduction and recovery
and the biological characteristics of trees 246
The question of causality 247
Radiocarbon method 248
Long-term deviations 248
Short-term deviations 249
Reasons for the variations 250
Combustion of organic fossil fuels 250
Hydrogen bomb explosions 251
The consequences for historical research 251
Stable isotopes 252

V The history of dendro-

chronology (Europe and USA) 255

The first tentative steps 256

Tree ring research develops 257
Andrew Ellicott Douglass (1867-1962) 257
Dendrochronology in Europe 260
Bruno Huber (1899-1969) 260
Tree-ring research as a recognized science 260
Barriers erected between the different branches
of knowledge are at the root of many of our
problems. One specialized science is not able
to provide a complete global picture which, in the
complexity of our modern existence, would give
us something firm to hold on to. This is why we
are looking for a synthesis; we want a compre·-
hensive view.

C. F. von Weizsacker
Preface

At a meeting of dendrochronologists an American colleague described the

effects of volcanic eruptions on annual ring formation in bristlecone pines. I
knew very little about either volcanoes or American pines! At the same
meeting European scientists spoke on the dendrochronological dating of
lakeshore settlements and the effects of larch bud moth attack on trees in
the Alps. It is possible that American participants were not in a position to
fully appreciate these papers either. In other words, dendrochronology is an
extremely interdisciplinary science; its facets range from modern statistics
on wood anatomy to the history of art. It is difficult even for dendrochronol-
ogists to keep in touch with the whole spectrum, and even more difficult for
the layman to obtain an overall view of the many methods and fields of
application.
In recent times specialisation has begun to hinder communication be-
tween the various sectors. Archaeologists, for instance, set up their own
dendrochronological laboratories and construct independent chronologies to
serve their particular interests. The scientific institutions which previously
carried out such work are now turning more and more to strongly statistically
or biologically-oriented questions. The full wealth of information contained in
tree rings, however, will be revealed only when dendrochronologists make a
concerted effort to relate the findings of the different fields. In spite of
inevitable specialisation, it is necessary that the expert concern himself with
the work of his colleagues.
We should not lose sight of one trite but paramount fact; a tree ring is a
product of nature. Technical and statistical processes may clarify, quantify
or summarise our observations but they are not an end in themselves and
can not be divorced from our observation of natural development.
My aim in producing this book is to present, by means of simple texts
and numerous clear illustrations, a general survey of dendrochronology and
also to show, or at least remind, dendrochronologists themselves what a
broad path we are following. The simple presentation is intended to lead to
better communication between the various branches of the discipline and
counteract extreme specialisation. My own experience in the university
world has shown that a simple presentation of dendrochronological findings
often bridges the gap between arts and sciences. The specialist will need
to exercise forebearance; hardly any aspect is discussed down to the last
detail. After I had finished writing the book I realised that a solo attempt to
present such a vast field was bound to be somewhat subjective. The
description of radiodensitometric techniques, for instance, refers mainly to
methods and equipment in use at the Swiss Federal Institute of Forestry
Research. The work of Evertsen in 1982 indicated that substantial differences
exist between the results produced in different laboratories.
This book is intended to be used by teachers and students in intermediate and
advanced courses and those studying history, archaeology, geology, geography,
glaciology, climatology, biology, ecology, forestry and physics.
Preface to the English edition

Four years ago hardly anyone would have believed that the analysis of tree
rings would progress so rapidly. In Europe the last remaining gaps in a
7000 year-old chronology have been closed. Nearly every oak growing in
gravel or settlements in western and central Europe can now be dated
accurately. This has considerably simplified the work of archaeologists
correlating the chronological order of settlements.
Many pieces of work have been produced on the dying trees and forests
of Europe and we may assume that in the near future tree rings will provide
a lot of information about ecological changes in this century.
With the destabilization of our environment it has become clear that the
tree ring has recorded previous environmental changes infallibly. Changes
in the water table, river flow, atmospheric conditions, water balance and
climate can all be studied with the aid of tree rings. Only with hindsight can
we understand our situation today. Tree rings will tell us whether better
environmental-conditions have created better conditions for growth.
Only limited progress has been made in finding new methods of measur-
ing. Existing methods have been improved slightly - new equipment for
measuring tree-ring width, new densitometric apparatus - and a method
used widely in the U.S.A. - pointer year dating - has been used for
examining conifers in Europe. Alongside the technically demanding methods
of measuring, this simple technique based on observation will continue to
gain ground.
The number of dendrochronological laboratories has increased in all
corners of the world and it is to be expected that quite a lot of historical
and biological laboratories will be extended to provide space for tree-ring
analysis. This English edition closely follows the format of the German one.
My aim is the same - to present a wide, and in part complex, science in a
simple form.
Many colleagues have given me valuable support in the preparation of
this book. lowe a particular debt of gratitude to friends in the Laboratory of
Tree Ring Research in Tucson, U.S.A. where I spent six months and gained
insight at first hand into the problems of American dendrochronologists.
I extend my thanks to: Dr. O. U. Braker wrote most of the texts in the
statistical chapter. Prof. Dr. P. I. Kuniholm critically corrected the manuscript.
Doris Pichler, and M. Sebek made the final graphs and W. Schoch made
many of the drawings and helped me with the lay-out. Ernst Schar prepared
the historical chronologies.
The original german text was translated by Ena M. Hartland and Maxin Backus
to whom I would like to express my grateful thanks.
The Swiss National Science Foundation provided generous financial
support for the tree-ring research.
Looking to the future

During my research in the course of writing this book certain central ques-
tions crystallized which should be taken up in future dendrochronological
research, particularly concerning biological-climatological relationships.
These are as follows:

all sampling should be carried out according to strict ecological guide-

lines so that the relationship between tree rings, climate and site
become clearer.
the fact that tree rings contain valuable ecological material should be
accorded wider recognition and dendrochronology should become a
permanent part of archaeological, historical and eco-physiological re-
search programmes.
the existing networks of tree ring series in USA, USSR and Europe
should be extended world-wide so that sufficient material becomes avail-
able for a global analysis of climate during the last tree centuries.
it has become clear that in future tree-ring analysis research will play an
indispensable part in work on pollution.
by correlating tree-ring series with historical, hydrological or climato-
logical information it will be possible to see our present-day situation in
an environmental perspective.
previous research into tree rings was limited to the analysis of com-
paratively few problems which however were studied intensively. In areas
suffering from environmental damage large samplings can be taken at low
cost which will help research workers to find the common factor in tree
damage and lead to damage analysis being carried out in wider areas.
 Introduction

F. H. Schweingruber, Tree Rings

© Kluwer Academic Publishers, Dordrecht, Holland 1988
Introduction

These two pages show a few of the many cases of
rhythmic growth increments occurring in nature. No
process of development is continuous and no organism
grows without interruption. Sooner or later - it may be
seconds or years - growth is limited by internal and
external factors.
Fluctuations in growth are usually irregular and only
seldom do they correspond to fixed astronomic cycles
such as days or years.
The organisms and structures illustrated here are all
manifestations of periodic or aperiodic growth. The topic
of periodic and aperiodic growth increments in trees is
dealt with throughout this book.
Growth processes can often only be detected through
measurements and are rarely reflected in structures.
Organisms with growth structures are as a rule short-lived
and seldom survive for more than a few years. Trees with
annual growth increments are a major exception to this.
Fortunately - for without them our particular science
would not be possible - trees are widely distributed over
the earth. We must bear in mind that the term 'annual
ring' is not absolutely accurate. It is true that usually only
one ring is produced in anyone year; its formation occurs
only within the vegetation period, however, which in many
areas lasts for barely four months.

E
~
Ol

o
o
z
8' f - - - - - '
h

Growth curve of a culture of bacteria (Nultsch, 1977).

Even under optimal conditions growth is not continuous. It de-
creases with exhaustion of nutrients and concentration of metabolic
products.

Diurnal rings in cotton (Frey-Wyssling, 1959). Annual increments in the horns of an ibex (Baumann, 1949).

2
Aperiodic increments in agate. Varves. More or less annual increments in the sediment of a lake.

Increments, probably annual, in coral of the Tertiary period. Increment layers in the shell of a fresh-water mussel. In the course
of one year several layers are formed in response to changes in the
environment.

Increment layers in the shell of a tortoise. Increment layers in a tooth of a mammoth.

3
 Origin of the materials

Massive pine (Pinus sy/vestris) with birches growing beneath, near the tree-line in
the east of Scotland. The maximum density in the annual rings of the three-hun-
dred year-old trees reflects to a great extent the summer temperature.

F. H. Schweingruber, Tree Rings

© Kluwer Academic Publishers, Dordrecht, Holland 1988
The site and the tree

The tree as integrator

A tree is a stationary living thing. Its crown, trunk and

roots are capable of reacting to environmental factors;
some parts of the tree receive signals from the surround-
ings while others react to them. The long-lived Pinus
longaeva integrates these ever-changing stimuli over
5000 years, the short-lived Populus alba only 50. This air
ability to integrate is reflected in many different kinds of
tree characteristics, such as geographical distribution, tree
crown and tree rings. In the course of evolution plants
have developed on every site, being best adapted to the
particular conditions obtaining there.

Reaction time

Measurements of ehvironmental conditions and their Q .

effects on tre~s have shown that a tree responds to . mOisture 0
changes immediately, i.e., within a few minutes of the
event. Such extremely short-term individual adjustments ()
I)

are scarcely reflected in the tree rings. The sum of all soli wa rmth
o o I:> '.>
these slight .changes, however, is expressed in a very
complex way in the annual ring, which reflects events
both in the current growth year and in the past. A tree is an integrator. It reacts to environmental facfors.

?
:iF ~ climate year I

30
u
20
'"
'""
'g 10
~ soil at depth of 5 em

variation in stem sI ze
15

E
u
10
B-no data
05

6AM 12 N 6PM lime 6AM 12N t- l I' 1 IH

28.JlJIV I 66 17.Aug. 1966

The reactions of trees are measurable. With the help of appropriate The reactions of the tree can be seen in the annual ring. The
eqUipment it can be shown that gas exchange, transpiration and condition of the tree in the previous year affects the developing tree
growth of the trunk are all related to precipitation, temperature and ring; the weather during the formation time, however, will affect the
other environmental factors (Fritts, 1976). ring of the following year (Fritts, 1976).

6
Transformation into structure

Most of the activities of a tree i.e. its physiological

processes, are manifested in its structure. A physiological

7
process initiated by a change in the environment even- LOW
PREC IPITATION----...,. Less
HIGH
More ~TEMPERATURE
tually becomes directly measurable as a growth change, \~ clo'" > ,o~~~~:o;- ~ ~
or evident in the variation of such features as length of '\ ~ I ~ Increased
Low soil Low vapor ----?- Rapid plant
shoots and needles, tree ring width, cell size or cell wall Decreased mO l slure ~pressure ~ e "ap oro t l on -E tempera tur es
activity In IN CREASE D Decreased Less
thickness. le~r:~~~~t=s
...
E WATE R STRESS ~trQnSplrall o n ~ e v(J poratlve
IN TRE E _____ dUring midda y cooling

De creased rates of L ess current ReduC%d net \ ~ I n~eosed

apica l growth ond--+ foliage photosynthesis stomatal closu re
leaf el ongation ~ dunng doy
Br.JI1ch R II1q
p.(,~lIes 72 73 74 75 76 \
~ Less translocat ion Increased

""dth Decreo.sed ~ - grO~lfhf~~~u~~I~rs respiration

co ncenlrcl!ons of ~
n m mn-o mn growt h re9u~o r s Less prod uction (Fr om B )
10 ,,0 ~ of cel l ports L ES S POTEN T IA L
Less cell ond substances FOR RAPID
expansion Short,er _______' /C AMBIAL GRO w TH
~ 9rowl~ ~ V
30 so ~ season ~
Radially nar row ~
Reduced ,r?l.es of
cell di VISion
Iracheids ~ Fewer )(ylem ~
formed ~ cells d,Z!.erenlioled
20 40 3 A NARROW RING
IS FORMED

10 30 2

Relationship between tree-ring width in twigs, length of shoots and The reaction mechanisms in trees are very complicated. Chain
needles in Douglas Firs. The same environmental factors seem to reactions are set up between the influencing environmental factors
affect these three growth characteristics in the same way (Shane and the parts of the tree which react to these influences. The
and Harper, 1979). diagram here gives only a glimpse of the whole (Fritts, 1976).

50 56
I I I Iii I

Relationship between needle length of a Bristlecone Pine on a site Differing cell sizes and cell wall thickness reflect frequent changes
on the tree-line in California and summer temperatures. In a cold in the environment. Spruce shoot 40 : 1 Cross section.
summer (1965) short needles are formed, in a warm one (1966) long
ones (LaMarche, 1974).

Fritts 1976 7
Tree sites on the earth

All tree sites and in particular their boundaries are influenced by their position on the earth. Important factors are:

Location on the planet

Because of the tilting of the earth's axis, the vegetation
periods in the two hemispheres are different; in tem-
perate zones in the northern hemisphere it lasts from
June to September; in the southern hemisphere it lasts
from December to March. The earth's shape and its
revolution round the sun have given rise to different
climatic zones; there are regions hostile to trees, regions
with seasonal changes due to differences in precipitation
or temperature, and regions with no pronounced seasons.
The sum totals of temperature and precipitation and the
distribution of these important factors for tree growth vary
with latitude. It must also be noted that, in high latitudes,
day and night do not follow a 24-hour rhythm.

parallel sun's rays

Kiruna (505 m)
oc Prnm

30 60
North Scandinavia 20 40
60 N
0

arctic 10 20
6·
2l.June 2 1. December o
Movement of the earth round the sun, side view. Position on 21st
June and 21st December.

Casablanca (55 m)
oc Pmm

30 60

Sahara.30· N 20 40
arid
10 20

Quesso (345 m)

oc Pmm

40 80
Beginning of the wheat harvest in both hemispheres; June to precipitation
September in the northern, November to January in the southern. Congo basin, Equator 30
Tree rings are formed in these months (Heyer, 1977). 60
tropical I----------~
20 40
temperature
10 20

o
Climatic diagrams from the Arctic circle, the Tropic of Cancer and
the Equator (Lieth and Walter, 1967), For explanation see p. 20.

8 Bar, 1977; Walter and Lieth, 1967; Strahler, 1975; Walter, 1964168; Weltforstat/as.
Location on the continent

The distribution of tree sites over the land-water mosaic

is mainly influenced by regional climatic conditions. The
extent to which the climate is continental or oceanic has a
very great effect on tree growth.

C
20

10 - ....
..
C"')

o
10

20
Temperature amplitudes between January and July in three different
climatic zones at the same latitude (Bar, 1977).
"C
o
Location in relation to ocean currents

20r
to

Warm and cold ocean currents play an important part in

the arrangement of vegetation belts and determine the
northern timber line to a great extent. 30~

40

Average January temperature at latitude 60'N. The temperature is

low over the continent but high in the region of the Gulf Stream (Bar,
1977).

Elevation

The effects of altitude vary greatly with geographical

position. This is particularly noticeable in timber line
regions, where usually a single factor determines the
presence or absence of fairly large perennial plants, so
that the principle of limiting factors is most clearly seen.
As an integrator of weather, vegetation is capable - as
long as man does not interfere too much - of reproduc-
ing all climatic zones faithfully. Maps showing the distribu-
tion of various species of trees indicate the differences
between the contours clearly. An example is shown
below; the vegetation changes with decreasing tempera-
ture and increasing precipitation and characteristic vege-
tation belts are formed at each level.

Leysin (1350 m)
Martigny (478 m)

precipitation

temper,lIure

valley mountainOUS area

Climatic diagrams from low and high altitude within a uniform Map of the vegetation of the Tyrol, a dry alpine valley (Schiechtl,
climatic area. Valais, Switzerland (Lieth and Walter, 1967). 1973).

9
Influence of environmental factors on tree sites

The effects of site factors on tree growth differ according variation and the inter-correlation of the many factors
to the position of the site on the earth's sphere and within involved. Two further examples show the effects of geo-
the climatic zone. The example given below illustrates this logical substrate and light relations.

Influence of topography

climatic zone subpolar 60' N semi-arid 30' N tropical 0'

53° 30>
..- /'

/
I

exposure and isolation angle at noon

on 21 st June and 23rd September. 30° N S 30° N S

consequences for the tree of the varying favorable too cold too warm favorable no difference
isolation:

soil temperature reI. high reI. low too high normal normal
soil moisture favorable too high too low favorable favorable too high
biological activity in soil high low low high high
growth conditions in root zone good poor poor good good
evapotranspiration normal too low too high normal normal
vegetation period reI. long too short too short reI. long all-year

consequences for tree growth of

environmental changes:

steeper slope positive negative negative positive neutral

higher precipitation negative negative positive positive negative
low temperatures negative negative positive positive neutral

The influence of topography on soils with normal drainage and on tree growth in different climatic zones.

Influence of mechanical movement

Changes in physical equilibrium initiate the formation of To satisfy its need to grow vertically a tree reacts
reaction wood; conifers produce compression wood while quickly to stimuli such as wind or earth movement. The
broadleaf trees form tension wood. reaction can be particularly clearly seen in the root collar.

soil movement
explosive growth

constant influence of wind

continual formation of compression wood

root collar; abruptly changing ring width

10
Influence of geological substrate

substrate hard rock, acid to basic soft stone, basic to acid,

e.g. granite, limestone e.g. marl
soil ranker, rendzina brown earth
weatherabil ity low high
penetrability low high
nutrient supply poor good
drainage good good to limi ted
water supply very weather- good over long periods,
dependent soil acts as reservoir.
tree growth good to poor usually good
tree-ring pattern sensitive compl acent

Infuence of geological substrate on tree growth in temperature zones

Influence of light

The vitality of a tree is substantially determined by the activity, so that narrow tree rings are formed. If the tree is
supply of light during the vegetation period. A tree grow- released after such a period of suppression, it is instantly
ing beneath a dense canopy receives only poor illumina- in a position to perform better.
tion, which reduces photosynthesis and thereby cambial

release

shade
!I light
suppression

I light
~ 1 lHWIII !
- - - - - - - - - - - -....- shade

light shade I ight shade light shade light shade

Reactions of tree and cambium to changes in light conditions

Trendelenburg and Mayer-Wegelin, 1955; Kittredge, 1948. 11

The site in history

Every site is the product of its substrate, time and the
eternally changing climate. The duration, type and inten-
sity of the influencing factors are responsible for the
character of the present-day site.
Site changes in geological periods
Immense climatic and geological changes have influenced
vegetation and sites fundamentally. Two examples serve
to illustrate this:
(1) Palaeozoic Era - about 300 million years ago.
At the end of the Palaeozoic era the present-day con-
tinents formed a large, practically complete landmass. In
the charred, silicified tree trunks from the Carboniferous
swamps in the north of the continent hardly any annual
rings are to be found. The climate may have been the
same the whole year through. In the south in Gondwana-
land, wood from the same period shows clear growth
areas - they are perhaps annual rings. Moreover, as in
the region of present-day South America, South Africa,
Australia and the Antarctic traces of glaciation have been
discovered, one must assume that the polar regions of
that time had a seasonal climate.
(2) Tertiary Period - about 70 to 5 million years ago.
From the evidence provided by fossilized plant and animal
remains it has been clear for a long time that a change
from tropical to temperate climate occurred in Europe in
the Tertiary Period. But only after the discovery of the
outer layers of micro-organisms in ocean-boring samples
was it possible to plot the temperature course for the
whole time period.
Conifer tree trunks from the Tertiary Era (miocene) in
central Europe have tree rings.

1 em annual growth
• prominent growth rings
.& faint growth rings
• no growth (ings

In the mid Carboniferous period (approx. 320 my BP) most of the

carbonised trees grew in tropical swamps (shaded area) and
growth rings are therefore absent. In woods growing in northern
latitude of the old continent distinct growth rings are present.
(Creber and Chaloner 1987).

3',
u30
~ 25
::J
§ 20

E 15
0>

'" 10
g
"
c.. " ~P~al~eo-c-e-ne~I~Eo-c-en-e--'I~O~I~ig-o-te-n-e-'1--7M~io-c-en-e--~I~P~lioc--en~e
I I
MID. a 50

Reconstruction of vegetation in the Miocene epoch in the area Probable temperature changes in northern Europe during the
around Lausanne, Switzerland. In areas where oak forests and vines Tertiary Period i.e. in the last 65 million years. (Bucharat 1978)
grow today, palms, laurels and acacias grew 15 million years ago
(Heer,1865).

12 Creber and Chaloner, 1987.

Site changes in prehistoric times

During the Ice Age prehistoric man experienced drastic neanderthalensis, on the other hand, lived in tundra with
changes in climate, ecology and flora. The Homo erectus creeping willow and dwarf birches. Homo sapiens in
of the early Quaternary Period in Europe was familiar with central Europe saw the arrival one after another of our
magnolias, sequoias and Sciadopitys and hemlock. Homo present-day forest trees.

Zurich at the time of the last Ice Age (Heer, 1865). Possible July temperature curve since the last interglacial time in
northern Europe (Grootes, 1977) and the probable glacial stages on
the northern boundary of the Alps (Welten, 1978, 1981).

Pollen CharCOilI Milmmals Anofact.

Appearance of main tree genera in northern Europe in the last 12000 Appearance and disappearance of various key fossils in prehistoric
years (Straka, 1970). caves in the upper Donau valley. The plant, animal and human
associations have changed fundamentally in the last 12000 years
(von Koenigswald et 81., 1979).

13
Site changes in historic times
Since the fading-out of the Ice Age, the climate has
fluctuated only slightly. In recent times, climatically caused
site changes have obviously occurred only in extreme
areas. Over the last three millennia man has altered the
landscape by clearing forests and in the last fifty years he
has gone a long way towards destroying the balance of
In mountain regions
In cold periods glacier tongues extended far down into
the valleys, in warmer times they retreated to greater
heights.
Rhone glacier in 1850
In arid regions
In moist phases some areas of the Sahara desert were
covered with vegetation; farmers grazed their cattle on
areas which today are desert.
Herd of cattle in a cliff drawing in the present-day desert.
14
nature through over-utilisation and pollution.
Plant distributions and species combinations are the
products of all these changes.
Different species areas have arisen as a result of
differing genetic predispositions of individual species and
different species combinations on ecologically similar
sites.
The Rhone glacier (Switzerland) in 1850 (left) and 1970
(right). In the course of 120 years, the glacier has sur-
rendered more than 2 km of valley floor.
Rhone glacier in 1970
At tree sites
Sites change even within the life-span of a tree. A site
can become impoverished through erosion of the organic
surface layers or enriched through accumulation of humus
(Thenius, 1977).
Exposed roots of a 4000-year-old Bristlecone pine. During this time
the soil has settled, the site has changed (Muench and Lambert,
1972).
Firbas, 1949/52; Gothan and Weyland, 1973; Heer, 1865; Straka, 1970.
The product of the changes

In the course of time, several separate species or sub-
species have arisen from one basic species through
isolation, and become more or less widely distributed.
This explains, for instance, why alpine flora under similar
ecological conditions form plant communities with the
same composition of genera, but with different species.
If the species areas are very distant from each other
and the different species clearly distinguishable, they
were probably isolated at an early stage. If the differences
in area and form between two species are only slight,
separation probably occurred during the Ice Age.

The widely-separated areas of Anemone nemorosa (1), amurensis (2), quinquefolia (3) and altaica (4) indicate early isolation (Straka, 1970).

Atlas and Pyrenees Southern Europe South-eastern Europe Eastern Europe and Asia Minor
mauretanica laricio nigra pallasiana
clusiana italica dalmatica banatica
salzmannii calabrica gocensis pindica
balcanica
caramanica
fenzlii

The geographical varieties of Pinus nigra (Straka, 1970).

These fifteen varieties originated in stands limited to the south in the cold periods, having been separated by fragmentation of the area and
subsequently developed in different directions.

!im Plcea mariana

0 Plcea sitchensis

Plcea engelmannii

0 Picea rubens

Distinctly separated areas of Individual species. Within the area

of the genus Plees, P. sitehensls occurs at low elevations and P.
engelmsnnll at greater heights in western North America.
The species area of P. rubens in the east Is very distinctive
(Harlow and Harrar, 1950).

Straka, 1970. 15
Selection of the sampling site

The selection of sampling sites is a key factor in dendro-
climatological research. In order to build up sampling
networks containing similar climatic information, it is
expedient to choose samples from boundary situations,
where growth is limited mainly by one factor, although not
from extreme outposts, where the information stored in
the tree ring pattern does not reflect the normal weather
conditions. It must also be borne in mind that natural
boundaries can be considerably displaced by human
activities such as animal grazing, felling and fire.

limitation by cold polar timberlines alpine timberlines

limitation by drought timberlines near deser ts lower Ilmberl ines

limitation by moisture maritime areas fog zones in mountains

. Timberlines.

Ecologically determined limits of distribution dry pine normal wet pine

beech

normal beech

Beech grows well on deep soils with normal permeability,

but cannot grow on shallow soils subject to periodic
drought or waterlogging. Pine succeeds on ecologically
beech
• ... beech
.. • -.;)
__
special sites e.g. steep slopes, bogs. On normal sites pine
is suppressed by beech.
L______--=~.==~p=rn=e==~.=--_ _ _--=:~==p:i::ne==~:J

16
Annual rings from varying sites within a restricted area of
a uniform climatic zone reflect different climatic events.
The delimitation of a site, however, is not easy. In every
case a sound knowledge of ecology and botany is neces-
sary for the selection of a sampling site on biological
grounds. The selection process is as follows:
Getting to know the site
Every objective, whether dendroclimatological, ecological
or pathological , requires that the researcher obtain a
comprehensive knowledge of the site by familiarizing
himself with the whole spectrum of conditions in the field .
The following a(ethe foremost criteria :
variability of the site conditions . Which plant societies
are typical?
evaluation of the probable ecological conditions as
shown by indicator plants.
identification of possibly suitable tree species (age,
form , annual ring width , etc.)
Deciding on the sampling site:
The following example from a temperate zone illustrates
the selection and delimitation of sites in relation to the
research goals.

Selection of trees in relation to the sites.

Geological substrate recent alluvium alluvium terrace volcanic tuff , bedrock

volcanic ash and granite coarse-grained volcanic ash
Soil brown earth and virgin soil brown earth rubble over virgin soil
loose ash
Conspicuous indicator grasses and large-leaved grasses - dwarf shrubs
plants plants
Dominant trees Nothofagus pumilio Nothofagus Nothofagus Araucaria araucana
Araucaria araucana Araucaria
Tree height up to 30 m Araucaria 12-30 m 8-12 m
Nothofagus up to 18 m
Disturbances tree felling , floods in tree felling landslides wind
the past fire fire
Choice of species and Araucaria and Nothofagus Araucaria and - Araucaria
research objectives Nothofagus
hydrology, especially dating - local climatology
river run-off forest fires
regional
climatology

Practical example of site definition for selection of sampling trees in an Araucaria araucana/Nothofagus pumilio region in the southern
Andes on the alpine timber line at 1700 m.

Ellenberg, 1978; Kuchler, 1966; Meusel, 196511978; Walter, 1964/ 1968. 17

Measurement of climatic factors at the sampling site

Measurement of ecological factors over decades: the network of meteorological stations

Density of the network Limitations

Dendroclimatology needs the long-term records of the
meteorological stations in order to relate tree-ring para-
meters to weather . Biologically, the most important factors
are temperature and precipitation . Air pressure does not
directly influence tree growth. Comparison of data pre-
sents considerable difficulties, particularly when values
from all parts of the world have to be compared .
The meteorological stations are not evenly distributed.
In highly developed areas of Europe, a dense network has
been in existence for over 100 years, but in other densely
wooded areas - the northern timber line regions or the
Himalayas for instance - only a few stations have been
operating for longer than 30 years .
The reliability of the data is very variable. The dendro-
climatologist must check them carefully before using them
in the calibration of tree ring data. This process is very
time-consuming. It spans the reconstruction of the history
of the station, the question of calibration of the instru-
ments, and the comparison of the data series with those
of other stations. Collaboration with meteorologists is
therefore essential. There are, however, some simple and
not very time-consuming methods which allow the den-
droclimatologist to form an estimate of the homogeneity
of the data series.

F
250

200

150

cumulatIve temperature dIfferences

100
December - February

50

0
1910 1930 1950 1970

Meteorological stations measuring air pressure around 1850 Cumulative temperature differences as a function of each year.
(Lamb, 19n, Vol. 2). The series is homogeneous when the points form a straight line.
The kink in the curve shown here may have been due to are-siting
of the station (Fritts, 1976).

Extrapolation of the Data

The official measuring station is seldom in the immediate
vicinity of the sampling site. Cons iderable differences
between conditions prevailing at the station and at the site
are therefore to be expected.

Relation between daily minimum temperature (top) and net photo-

synthesis at two stations in a dry alpine valley. Upper line: 600 m,
valley floor. Lower line: 1900 m, mountainous area (Larcher, 1973). o N 0 M A M A

18 Geiger, 1961; Lamb, 1977; Larcher, 1973; Tranquil/ini, 1979.

Measurement of ecological factors over periods of a few years: measurements in the stand

Meteorological data series from the stand under investiga-
tion provide, without doubt, the best values for compari-
son with annual ring parameters; tree site and measuring
station are in one and the same place. Only in rare
cases is it possible to continue the measurements over
decades. The physiological reactions of the tree can be
measured at the same time as the meteorological factors.
Measurements of gas-exchange processes in trees indi-
cate physiological activity. Technical problems limit inves-
tigations more or less to small trees. Dendrometers can
be used to register fluctuations in trunk thickness, which
form one of the expressions of physiological activity.

Ecophysiological measurements Measurement of cambial activity

Measuring chamber. Twig of Stone pine (Pinus cembra) with

measuring chamber for the registration of ecophysiological data.
Haesler, unpublished.
Stlllberg
.J:: 4
PN ne photosyntheSIS 13.8 1975 E
~ 3
2
..
t
't)

I
0
0
u - I
E 0 2 4 6 8 10 12 14 16 18 20 n 24

Two dendrometers on a conifer trunk.

120L'"LJ,,"n""~
'00
•
Top: a needle resting on the bark registers changes in thick-
80 ness at one point on the trunk.
~ 60 Bottom: a belt round the stem registers changes in circumference.
40 Both instruments send impulses to a recording unit (Tranquillini,
20 unpublished).
o
o 2 4 6 8 10 I 2 14 16 18 20 22 24
24 T. dir temperature
20 %
16 100
u
'2
8 80

0 ,, 60
- ..-- pIne
------ IIr
0 2 4 6 8 10 12 14 16 18 20 22 24 - - - - 1I1'h4'
- - - • .ofl)("P

IrdflSplrdtlof
40 - - be<,h

~
.
E
20

0 , O~~~~_.----_r----.-----._----r----.----
0 2 4 6 8 10 12 14 16 18 20 22 24 M o
.J::

Ecophysiological data for one day recorded on a Pinus mugo Daily radial growth of various central European tree species in
(Haesler, 1982). southern Germany (Mitscherlich, 1970-75).

19
Site characterization

Dendroclimatological investigations aim basically at obtain-
ing information on two points ; climatic changes over large
areas and weather conditions in particular localities. In
the end, however, the findings from all regions of the
earth possessing annual-ring-forming trees must be fitted
together like a puzzle and related to each other. For this,
site descriptions are of the greatest importance, even
if they seem unimportant to the researcher at the time
of sampling. Only by comparing the site descriptions is
it possible to determine which tree ring-sequences fit
together to form a unit. The formulation of regional den-
droclimatological objectives seems to me possible only by
means of good site descriptions.
There are two basic ways in which a site can be
described :
description of site integrators i.e. those elements
whose behaviour depends on climate.
measurement of the main site factors over a given
number of years.
The first method has the advantage that values from
widely differing areas can easily be compared. The
second allows direct comparison with tree-ring growth,
but usually permits only a poor characterization of the site,
since the biologically decisive extreme years are not
registered. Comparison of measurements from different
regions is, for chronological reasons, simply not possible .

Climatic integrators

The site is characterised by descriptions of position, Soil

topography, soil , vegetation and climatic-ecological condi-
tions. These parameters reflect the course of the weather Soil is an outstanding integrator of climatic and vegetation-
over many years, including extreme events. historic conditions over very long periods. The decoding
of the stored information however requires extensive
Location and topography investigations, as it extends beyond the description of the
soil profile and rooting conditions. Important points are
Information on geographical conditions elevation, relation the depth to which the tree roots extend and the pre-
of site to timber line, substrate etc. expresses the regional sence of roots of other species. Plants indicative of
and local character of the site. factors important for tree growth must be identified, and
information on nutrient supply and drainage must be
Climate obtained.
Climatic diagrams provide in concentrated form the best
su rvey of average conditions obtaining at any given area.
The type used by Lieth and Walther is very compact and
characterises each area very well. It must be borne in
mind that the climate of the sampling site or even the tree a b d e
sOle height above sea level mean mean
stand is not identical with that of the measuring station. temperature preCIpitation
C No of years of observation
In year In year

absolute maXimum )( 120

x
mean daily
maximum
'" " \ // '00

. u 40
"\ ~"", ~K"'""~ "I" 80 E
E

~30
8.'" humid
. \ •
/ 60g
c
!!!
0.
E
~
.>:~:-;./.
!!!
20 40 a

.'.
/
• drought
10
.
/ mean mon Iy temperature .........
20

mean dally mlnlmum .r .,

In he coldest month
absolute minimUm
y
Y
1ii!!~IIIIfE~~r:~-'-'ED"~?JI
•
• •t o
monlhs WIth dallv mean months With absolute
mlOlmum mperalure minimum under OOC
un r OOC
Explanation of a climatic diagram (Walter and Lieth, 1967).

20 Ellenberg, 1967; Walter and Lieth, 1967

Vegetation
In plant sociology there are several systems for describ-
ing and classifying sites. That of Braun-Blanquet has
proved very useful for the delimitation of small units,
especially in Europe, and dendroclimatological work in
temperate zones has shown that it is also well-suited for
site characterisation. Equally, it has been found that in
many cases simplifications of the system are permissible
while in others refinements would be advantageous. The
system requires a knowledge of plant species and their
significance as ecological indicators, although generally
we are far from being able to relate the reactions of trees
directly to a few indicator species or species groups.

General characteristics
Stand
Geological substrate limestone debris limestone limestone alluvium recent alluvium
boulder debris bedrock terrace
Soil Rendzina Rendzina with virgin soil Brown earth Clay
thick surface
layer of humus
Drainage good very good very good good good to
limited
Dominant trees Picea Pinus Pinus Picea Picea, Alnus
Conspicuous indicator Cyperaceae Ericaceae Gramineae Gramineae
plants Ericaceae Mosses Alnus and large-
leaved herbs
Tree heights to 20 m to 18 m to 8 m to 25 m to 30 m
Characterisation of fairly undisturbed ditto steep bank pine forest in ground water
the stand normal stand pine forest zone
extremely dry
Objectives of sampling regional climatology ditto local climate hydrology, particularly river
precipitation run-off
Picea abies/Pinus sylvestris forests at 1000 m in the lower alpine regions in the central Alps.

Common species in a, b, and c Differentiating species Stand

b a c
Picea abies indicating dry Carex ha/leri x
Pinus sy/vestris conditions Limodorum abortivum x
Cory/us ave/lana
Prunus maha/eb indicating Comus sanguinea x
Carex alba moist conditions Hy/ocomium s/endens x
Erica camea
Carex omithopoda indicating wet A/nus incana x
conditions Petasites alba x
Phytosociological division with differentiating species and characterisation with overall species composition.

21
Site description

A large number of easily recorded features can be used for the characterization of a site and the comparison of
different sites. The following list indicates those appropriate.

Sampling site Vegetation

date Tree cover as % tree height in m

place shrub cover % shrub height in m
locality, nearby town herb cover %
region, valley moss cover %
district, canton, country, stClte lichen cover %
country vegetation belt
latitude vegetation unit
longitude
height a.s.1. in m list of species: trees
map no. shrubs
slope inclination herbs
exposure moss
elevation of timber line in the area lichens

Samples Braun-Blanquet's system of rating species abundance

Name of collector 5 = covering more than 75% of the area

address 4 = any number of individuals covering
tree species 50~75% of the area
no. of trees sampled 3 - any number of individuals covering
no. of samples 25-50% of the area
2 - very numerous, or covering at least
5% of the area
Climate 1 - plentiful but of small cover value
+- sparsely or very sparsely present,
nearest meteorological station cover very small.
climatic diagram
average January temperature
average July temperature
average annual temperature
average precipitation in January E_ w N ___ S
position
average precipitation in July
average yearly precipitation
climatic zone

Sampled tree
faul ts (give diagram)
height (estimated)
sociological rank: dominant oblique
co-dominant crooked
dominated spiral grow h .,
"

rJj)
suppressed knobs, knots
crown: diameter ridges

lEY
foliage normal frost cracks
foliage scanty rotten spots, scars
strong ~ 2/3 of tree branched fire scars
damage - - - -_____
medium ~ 1/2 of tree branched bOring
Insect damage
weak ~ 1/3 of tree branched
woodpecker holes
eccentric
trunk: excellent quality, no faults
fluted
excellent, faultless to 20 m
etc
good, faultless to 5~ 10m
poor, faults to stem base

22 Assmann 1961; Ellenberg 1967; Fritts 1976.

Geological substrate

Rock type
Weathered or unweathered
Weatherability of the rock

Soil

so il pro file (Duchaufour 1970 )

sy mbols:

cm profile sketch tree roots profile sketch

cm tree roots
0 Aoo
Ao
10

20 AI

30

40
C
50

60 podzol + + + 60 brown rendzina

Symbols :
Surface layers:
organic layer with little
decomposition (li tter).
b;AJVVl iron enrichment
111111111

layer of decom position ~ rust stains

1111 III II~
Mineral soil : humus
V77T1/~ hed rock silicate
1+ + +1
carbonate
1--==1 carbonate
~ I I::J
bleached horizon I:0.'.0 . ..J
eo · : • • •• •
horizon boundaries: sharp

diffuse ---
loam-clay
I -=--1

Assessment of water reg ime

'Interception loss: approx. 20% for
closed stands, correspond ingly less
Preci pi tation annual
for more open stands.
during vegetation period
Water balance in relation to relief
Interception loss
cupola inflow only
Topographically determined inflow upper slope moderate runoff
runoff lower slope moderate inflow
G round water level at ... cm plateau inflow = runoff
Impervious layer at.. ..... cm - basin inflow only

Duchaufour, 1970; FitzPatrick, 1980; Muckenhausen, 1975; Scheffer and Schachtschnabel, 1970. 23
The worldwide sampling network

With living trees

In the absence of meteorological records going back over
a long period information on weather conditions in earlier
times can be supplied by annual rings. Since weather and
climate are global phenomena, the aim of dendroclimato-
logical research is to investigate climatic interplay on a
worldwide basis; to this end a worldwide sampling net-
work should be built up, within which valid comparisons
can be made.
This is possible only if the radio-densitometric method
is used, at least for temperate and boreal zones, as only
density values can provide sufficient climatological infor-
mation for these regions.
Hardly any information on precipitation can be obtained
from tree rings in arctic regions, as temperature is the
limiting factor even on local dry sites. In semi-desert
areas, on the other hand, the low precipitation even in
deep valleys with a relatively high ground water level
determines tree-ring width and density. The periods for
which most climatic information can be obtained are
June-September in the northern hemisphere and
November-February in the southern hemisphere. Re-
gions with marked topographical divisions offer the best
conditions for weather reconstructions. In mountain areas
tree rings contain information mainly on summer tem-
perature and in dry valleys mainly on precipitation.

Limits of comparability
The comparability of tree-ring data is restricted by the fact
that tree growth is controlled and limited by different
factors in different climatic zones. In many places the
climate prohibits tree growth. Man has stripped large
areas of forest, so that the original stands of trees have
disappeared. In the tropics trees do not usually form
regular annual rings. Dated chronologies cannot be con-
structed for these areas.

Areas without forest (arctic and dry deserts) and without trees which
form annual rings (tropics) .

.'
" ••
'¥, Aielsch forest
• nlJ..._"LJIor..J..~u..j....:.~...,t..V-~.IA.Jl._"L-\ll...JV--1w.p .......a."f---+--I" __-lf~~_rfW'r centr af afps
~ l~m
;;;
.,c:
'0
••

Maximum density curves for trees growing on the upper timberline (top) and those growing on dry sites in an inner-alpine valley, Valais,
Switzerland (Schweingruber ef al., 1979).

24
Forested areas for which summer temperature could be recon-
structed.
The compromise
The multiplicity of sites and tree reactions theoretically
permits the construction of a sampling network with an
almost arbitrary amount of climatic information; in practice,
however, this is not the case, since the means of
research are limited and worldwide dendroclimatological
research has only just begun. Consequently, the objec-
tives must be constrained and research is restricted to
sites on which the trees are exposed to similar environ-
mental factors. One of the research goals for the northern
hemisphere is the reconstruction of summer tempera-
tures. Since the maximum densities of conifers at the
northern and subalpine timberline reflect average summer
temperatures, they could provide more or less homo-
geneous climatic information for the last 300-500 years
over a wide-meshed sampling network. Such an inves-
tigation would also help to define the areas for which data
series for historic and prehistoric times could be con-
structed. See p. 128.
Distribution of years with abnormal high growth (white areas) and abnormal low ring-width growth (dotted areas) in trees from cool-moist
sites near timberline in Europe (Schweingruber, 1987).
Hughes et a/. (edit), 1980; Schweingruber, 1987.
In the south-west of North America a sampling network
of trees from dry sites on the lower forest limit was
studied (Stockton, personal communication). In these
areas the width of tree-ring growth was limited by defi-
ciency in precipitation. As the trees on the site were very
old, information was provided about precipitation condi-
tions in these arid regions for the last 500 to 700 years.
In Europe a sampling network of conifers from cool,
moist sites in summer on northern and subalpine timber-
lines was studied (Schweingruber, 1985). Cambial activity
is limited from several factors which change from year to
year. Thus the maps express mainly areas of high and low
growth and less climatic information.
A similar network has been established from Russian
dendrochronologists over large Russian and Asian regions
(Bitvinskas, personal communication).
. .
~
~
25
With historic and prehistoric trees

The stated aim of dendrochronological research is to
construct chronologies covering hundreds and thousands
of years. Technically, cross-dating makes this perfectly
possible, but the success of this process depends on the
coincidence of various favourable factors:
a chronology can quickly be built up from long-lived
trees Pinus /ongaeva, Fitzroya cupressoides and Se-
quoia gigan tea , with ages of over 2000 years are
suitable. The wood must be durable by nature or given
favourable conditions for preservation . Fitzroya, for
example, contains heartwood substances which resist
decomposition in a moist-warm climate. Pinus lon-
gaeva has decomposition-inhibiting resins and grows
in a desert climate in which decomposition scarcely
takes place. Oak also possesses decomposition-in-
hibiting tannins, and fossil stems are often present in
sediments with no aeration; decomposition cannot
occur under these circumstances.
the tree ring sequence must contain much climatic
information. Trees from sites on ecological boun-
daries, e.g. arctic or arid timber lines, are more
suitable than those from sub-tropical regions .
It is a lucky fluke when all these conditions are fulfilled.
Up to now it has been possible to build an unbroken
chronology only with the extremely long-lived Pinus
/ongaeva and the European oak (Quercus roburlQ.
petraea). The Norway spruce and European larch chronol-
ogies are still floating earlier than 900 A.D.
Each of these chronologies has a particular impor-
tance:
Pinus /ongaeva served for calibration of the C14
method.
Quercus sp. The sequences from central Europe pro-
vided an outline of river and landscape history in this area.
Currently, they are being used in the dating of prehistoric
lake dwellings in the region of the Alps, and also for
calibration of the radiocarbon method.
Larix decidua, Picea abies. The material comes from
the alpine timber line and is analysed radiodensitometri-
cally. It affords an insight into the annual weather condi-
tions of the last 8000 years.

Long chronologies of the world.

A.D (anno domini 1

7000 6000 5000 4000 1000 0 1000 or absolute dated chronologlesl

-- - 6008 BC I I I' I Pmus longaeva California. USA

Irland/England
·Germany
pine
Northern Central Europe
12200 BP 11500 BP 973 AD
- f- Alps
pine pine
BP 400 AD

I
Pinus sylvesrris Scandinavia
---- l -----~--- - --
,
12000, 11000 10000 9000 8000 7000 6000 5000 4000 3000 2000 1000 o 1- 19501 for floallng chronologies
The longest chronol09ieS (State, 1987).

26
Characterization of the most important tree species in dendrochronology: Bristlecone pine, oak,
larch, spruce

Pinus longaeva D. K. Bailey, Bristlecone Pine

This evergreen conifer is found in continental alpine
zones. The average January temperature lies below O°C,
the lowest temperatures around -29°C, with average
summer temperatures between 10-15°C. In the rain
shadow of the Sierra Nevada precipitation is slight and
very variable; for the years 1949-1964, the precipitation
for the period December to March lay between 34-250
mm. At any time of year, there may be a whole month
without precipitation. The distribution of this species is
limited to a few high mountain areas in the south-west of
North America. It forms the timber line at 3000-3800 m.

The chronology
In 1953 E. Schulman discovered "the oldest living
things", the more than 4000-year-old bristlecone pines.
After Schulman's early death in 1958, W. Ferguson con-
tinued his work and constructed a continuous chronology
of 8700 years from fossil wood. Synchronizations were
made using tree-ring widths.
Fritts, 1969, investigated the ecological conditions in
relation to cambial activity of bristlecone pines in the
White Mountains over three vegetational periods. The
sequence is chiefly important for calibration of the radio-
carbon method, since no immediate conclusions about
weather can be drawn from this study.

Pinus longaeva. Twigs from good and poor sites, and a cone (R.
Hirzel, unpublished).

Geographical distribution of Pinus longaeva.

Pinus longaeva, wood, 40X and 100x.

Resin canals and small differences between earlywood and latewood
can be clearly seen in this cross-section.

Munch and Lambert, 1972; Schulman, 1958; Ferguson, 1968; Fritts, 1969. 27
Quercus robur, Quercus petraea - Common Oak and
Sessile Oak
These deciduous broadleaf trees with porous ring wood
cannot be clearly distinguished anatomically. The species
are light-demanding but have a broad climatic tolerance
within their area of distribution, occurring under very
varying condition, e.g.:

Western Europe: Southern Russia:

July temperature 22-24°C
January temperature -14°C
Precipitation 300 mm

Both species thrive best on deep, rather loose soil.

Quercus robur forms forests in the oceanic zone, often in
association with beech. Quercus petraea, on the other
hand, forms forests on dryer soils. It often occurs together
with hornbeam and Scots pine. Both species are widely
distributed between 40-60 o N in Europe. Quercus robur
forms forest at low and medium altitudes (0-400 m), 0.
petraea at medium altitudes (300-400 m). They may
reach an age of 500 years but in fact are seldom older
than 200.

The chronologies
In 1940 B. Huber of Munich, began the central European
tree-ring width oak chronology. Today, after 40 years of
work, dozens of chronologies of varying lengths exist,
some of local, others of wider significance. The longest
chronology of oaks, pertaining to the whole of Europe, Quercus robur (top) with pedunculate fruits and sessile leaves.
spans over 9000 years (see page 26). All these chro- Quercus petrses (bottom) with sessile fruits and pedunculate leaves
nologies are of historic importance but contain little (Hess etsl., 1967-1972).
climatological information.

Geographic distribution of Quercus robur and Quercus petrses. Quercus robur

(Hatched area) Ring-porous wood with multi-seriate rays and flame-like groups of
pores in latewood 40x.

28 Hegi, 1957.
Larix decidua Miller, Picea abies (L.) Karsfen, Larch and
Norway spruce
The deciduous larch and the evergreen Norway spruce
form belts in high mountain areas. Where annual precipi-
tation is over 1000 mm they tolerate an average January
temperature of -10°C, minimum temperatures of -35°C
and cool summers with average July temperatures of
10°C. Together with Pinus cembra they occur at the
alpine timber line. Larch tends to grow in the continental
and Siberian mountain zones, Norway spruce rather in the
suboceanic regions. Both species form forests at high
altitudes, 1500-2000 m in the Alps. Larches may live for
1000 years but Norway spruce seldom reaches an age of
more than 300 years. The distribution of larch is limited to
the mountains of central Europe. Norway spruce covers
the same area but in the north also forms stands along
the timber line, where Picea abies overlaps with Picea
obovata.

The chronology
At the Swiss Federal Institute for Forestry Research,
Birmensdorf, work is being done with radiodensitometric
methods on a Holocene conifer chronology. Much historic
wood is available, as the timber line zone in Switzerland is
covered with material from moraines, bogs and lakes of
earlier millennia, all containing fossil wood. Synchroniza-
tion is made easier by the availability of relatively old trees
with 300-500 tree rings (see page 26).

Plcu .ble. (right) with four-sided n"dle. and rather large, pendent
cone ••
L.rlx decidUa (left) with tuft. of needle. and ,mall cone, (He.. et
al., 1987-72, Hirzel, unpublished).

Geographical distribution of Picea abies. Region of validity of the Larix decidua (left). Conifer wood with pronounced latewood.
chronologies (Meusel, 1965-1978). Picea (right). Conifer wood often with weak latewood and gradual
early-Iatewood transition 40x.

29
Origins of the historic wood

The availability of wood for dendrochronological research o of sarnplonQS per year 'Ies German oak chronology
is not the same for all periods of time or for all regions. 150
Sland 1973
The following presents some thoughts - not exhaustive average _ maximum
127
- on facets of this problem. length of curves from individual
Human activity throughout the ages has determined Silmple
the availability of material today: 100

The settlement density is lower in regions with sensi-

tive tree-ring series, i.e. in cold or dry areas, than in
areas with more favourable climates. In spite of this,
50
much material is available in such areas because of
the durability of wood. The desert regions with Indian
settlements in south-west America are a prominent
example of this. o
Different areas were settled at different times; in yedr 500 500 1000 1500 2000
Europe, intensive settlement began 3000 years ago, in Hal std' period Ror"an ,rres early late modern
America 300 years ago. Varying modes of life result in La rene per od middle ages times
differing amounts of wooden relics. Sedentary peoples
leave more than nomadic groups and more building is
Number of samples in the oak chronology of E. Hollstein. Frequency
undertaken when economic conditions are favourable of finds coincides with settlement activity (Hollstein, 1980).
than in times of crisis.
The mode of construction is also a determining factor
for the present availability of wood suitable for re-
search. In stone buildings, for example, there is little
wood and in log constructions often only young
timber. Different types of wood of various dimensions
were selected according to function in the construc-
tion, and techniques also varied. In timberline areas
the only wood available was often irregularly shaped
and not of a particularly suitable size.
The position of the timber within the construction
has determined whether it has been preserved.
Material from the inside of well constructed buildings
is better preserved than that exposed to weather. In
temperate and boreal zones wood has often been
preserved for centuries in water, e.g. posts of pre-
historic lake dwellings around the Alps, or posts of
houses in medieval towns such as Haihabu on the
North Sea coast or Venice on the Adriatic .
Usually more material is to be found in regions with
a history of political stability than in those often dis-
rupted by wars. In Europe, innumerable buildings have
been wantonly destroyed. At the time of the Thirty
Years War (1618-1648), certain areas of the Alps
flourished while towns in other regions were being Log construction with halved and rough-hewn trunks. In cases such
as (b) sampling is often difficult and only seldom are the terminal
burnt down . rings present.
The supplying of building material varied from area
to area, so that in many northern European thickly
forested regions, buildings were exclusively of wood,
while in alpine areas with no forest, constructions were
almost entirely of stone.
The climate determined to a large extent the type of
building. In the Mediterranean region the buildings
were mainly of stone with few wooden parts, chiefly
for ecological reasons.

30 Fletcher et al., 1978; HoI/stein, 1980.

Charred beams in medieval ruins. Beams and posts in a neolithic settlement on a lake in
the the Swiss central plateau (Twann).

The origin of sub-fossil wood

The places where wood for the long chronologies was (a) under water
found give an indication of the conditions needed in order (b) under permanently dry conditions
that wood be preserved for thousands of years: (C) under permafrost conditions.

(a) Pine trunk of the late ice age found in a clay pit in
the Swiss central plateau (Diittnau, Zurich) (Kaiser,
1979).
(b) Fossil section of the bristlecone pine found in the
arid zone of the White Mountains, California.
(c) Sub-fossil tree trunks found in a moraine in the
Alps (Rothlisberger, 1976).

31
State of preservation of historic and prehistoric wood

Changes occuring with time

For investigations on both density and ring width in Growth of fungi, i.e. wood decomposition, is chiefly
historic and prehistoric chronologies it is important to be confined to the following environmental conditions:
able to recognize the mechanisms and organisms causing
decay, as well as the symptoms of decay, and to know
the different manners of preserving the wood. Biological Milieu wet damp dry
decomposition is dependent on the moisture content of Moisture content 30% 30% 36%
the wood, its oxygen content and the ambient tempera- Air content 15% 15% 15%
ture. In mesophilic zones decomposition is hindered by
fungicidal substances present in the wood. Thus some Fungal growth no yes no
wood of Fitzroya cupressoides in the rain-forest areas of
Chile and Argentina has survived for 4000 years.

Decomposition in the presence of air (aerobic decomposition)

In temperate climatic zones

Under normal conditions of moisture dead wood decays
rapidly, various rot fungi being involved. As long as the
widths of the tree rings can still be seen, partly decayed
wood can be used for dendrochronological purposes e.g.
criminological evidence. The trunk center of living trees,
especially of conifers, is often attacked by the fungus
Fomes annosus, which completely destroys the wood.
Therefore the inner parts, particularly of Norway spruce,
should not be used for dendrochronological purposes.
Wood decomposed in this way is not suitable for radio-
densitometric analysis. Both earlywood and latewood cell
walls are decomposed.

Soft rot, macroscopic Spruce, showing fungal infection macroscopic of open wound
(Bazzigher, 1973).

Soft rot, 800x. Fungal hyphae decompose the cell walls. Red rot, 800x. Cell walls broken by cracks.

32
In arid climatic zones
Arid conditions are unfavourable for the growth of wood-
decomposing organisms. The only potential wood-de-
stroyers are insects which produce water during the
metabolic break-down of cellulose (long-horned house
beetle and other wood-boring beetles). In the arid niches
of mesophilic zones e.g. in buildings, a small amount of
chemical decomposition takes place. In windy arid and
arctic areas, wood is eroded by sand and ice-crystal
blasting.
Longitudinal sectibn through a decomposed jOist. ,he cell wall
has diSintegrated through a physical-chemical process. This
happens more rapidly in the earlywood, where the cell walls are
thinner than in the latewood, and an outline of the density profile
is thus produced.
Conifer wood with terminal ring (Abies alba) 25x
left: interruption of growth in June. The earlywood cells are
clearly formed.
right: interruption of growth after the vegetation period. Close
under the bark, the last-formed tree ring has complete
latewood cells.
In arctic climatic zones
In arctic and alpine zones decomposition, i.e. fungal
growth, is inhibited by low temperatures. If a tree trunk in
high mountains is trapped in a glacier - whether through
landslides on lateral slopes or through being overtaken in
a glacial advance - it is embedded first in the ice itself
and later in sediments, and is excellently preserved. In
Hickin, 1973; Koenig, 1957; Liese (Edit), 1975; Rypa.cek, 1966; Schweingruber, 1978.
Wood in good condition from dry regions is suitable for
radiodensitometric investigations. There are often prob-
lems with the identification of the terminal ring, i.e. the
last ring formed by the tree. Only under a microscope is it
possible to see whether an outermost ring is incomplete
because biological conditions led to a stoppage of growth
or whether the latewood has become mechanically sepa-
rated from the earlywood.
Insect galleries. Damage occurs mainly in the relatively soft early-
WOOd.
Hardwood with terminal tree ring (Euonymus europea 250 X)
left: interruption of growth in June. Three rows of earlywood
vessel elements have already been formed.
right: interruption of growth after the vegetation period. The cells
adjacent to the cambium have a tangentially flattened
shape.
the arctic regions driftwood can be carried to permafrost
areas by floods.
On the banks of lakes, where the water-line is rising,
trees die and fall into the water and are thus preserved
for an indefinite period. Wood from such deposits is
eminently suitable for radiodensitometric analysis.
33
Decomposition under low-oxygen conditions (anaerobic decomposition)

In water and oxygen-poor sediments such as peat and

clay, decomposition progresses only slowly and is really
the first stage of carbonisation. Material from such de-
posits is practically always suitable for chronological, and
sometimes for radiodensitometric investigation.
The main agents of decomposition are bacteria and
actinomycetes. Decomposition can be recognized macro-
scopically by extreme shrinkage and loss in weight.

Bacteria in latewood cells of pine wood from the Bronze Age,

preserved in water (4000 years) 1200x.

Latewood cells of fir attacked by actinomy- Latewood cells of fir, preserved in water, In the final stage of decomposition the
cetes. The caries-like decomposition begins decomposed in a mosaic pattern. Actinomy- secondary walls swell and lose their struc-
in small patches and spreads over the whole cetes decompose practically only the secon- ture and stability. Fir, 1000x.
cell wa1l1000X. dary walls 400X.

4000 year-old oak trunk, preserved under water. The pronounced Detail from the decomposed zone. The latewood cells have lost all
shrinkage of the sapwood and outer heartwood indicates decom- stability and have collapsed when the wood dried out. Only in the
position by actinomycetes. earlywood areas is decomposition less advanced.

34
The sudden sharp decrease shown in the density diagram
indicates the position of the decomposition zone; in this
part only the ring widths can be used for synchronization.
In trunks from areas with low average temperatures, e.g.
the Arctic and the sub-alpine zones of mountains, only
the outer parts are decomposed. In samples of the same
age from lakes in warmer zones, e.g. the Swiss Central
Plateau, even the heartwood of slender oak trunks is
entirely decomposed. Decomposed wood can be found in
flooded coastal forest areas, fresh-water lakes, bogs,
sediments in meander zones of rivers and in alpine
moraines in practically every region of the earth.

glee
0.9 Density diagram of a Late Ice Age (13000 years old) pine trunk
~ ',wwood do",;ly from a clay pit near Diittnau, Zurich. In the well preserved section
on the left, both density and ring width can be used for dendro-
climatological analysis, but in the decomposed part on the right
0,8 only the ring widths are suitable (Kaiser, 1979).

0,7

- -+--+- badly dccomposc(l •

0,6

0,5

0,4 , I
I I

,
I
,
II
0.3
"'-.' \ '/7 '
dnnual rrllg wid 111

0,2 well preserved or _

only liglltly deCOIl1[losed

Decomposition by fire

A substantial part of the material is lost by carbonization. the density values do not correspond even proportionally
The remnants can only be analysed in term of ring widths ; to those of the original wood.

earlywood

~~~~~ tree ring

Carbonized piece of willow from an Eskimo settlement on Banks

Island, Canada (200X). The cell walls have shrunk in thickness by
approximately one third. The tree must have been felled during the
growing season, since the last year's growth is incomplete. There
are no brick-shaped latewood cells in the first ring under the bark.

35
Changes in fossil wood caused by pressure

If anaerobically decomposed tree trunks in swamps or the
sediment of rivers in glacial areas were covered by ice-
masses and then by gravel during the Quaternary period,
th en their form - both the anatomical and chemical
properties - was irreversibly changed . The wood had
turned into splint coal in the second stage of carboniza-
tion . In such material tree-ring widths and densities are so
altered that there is no question of radio-densitometric
analysis. The material is of great chronological value ,
however, since it is suitable for radiocarbon dating .

Heavily compressed spruce branch from a Pleistocene gravel bed in the Aare
south of Berne.

Photomicrograph of a compressed zone in a spruce trunk. The less dense

earlywood has been pushed into folds by the pressure of the ice-mass -
approximately 600 m thick at the maximum state of the Aare glacier - while the
denser latewood (top) has resisted the pressure (200X).

Conservation of fossil wood through mineralization

Theoretically, mineralized wood can be dendrochrono-

logically analyzed , since the ring widths have not been
altered. The densities, however, have been changed by
mineral deposition. Such material stems mainly from pre-
Quaternary marine deposits in areas now dry. Chaloneer
and Creber, 1967, found the earliest tree rings in the
history of the earth in lignified, carbonized pieces of
Callixylon newberry, a member of the fossil order Cor-
dailes from the Deronian Period.

Pine wood with mineralised sUlphr and silicates from volcanic sediment in lipari,
southern Italy 200X (Kaelin 1983).

36 Zimmermann, 1959.
Artificial preservation of anaerobically decomposed wood
Since anaerobically decomposed wood shrinks greatly
when it dries, and also partly disintegrates, it must be
preserved artificially. Any dendrochronological investiga-
tions must be made before preservation. For short periods
storage in water at relatively low temperatures is suf-
ficient; fungal attack is not to be expected, since the
easily decomposed polysaccharides are no longer pre-
sent. If the wood is not further preserved, however, bac-
This 6000 year-old post from a lakeshore settlement in Switzerland
shrank greatly and irregularly on drying. The sketched-in outlines
show the original form of the entire post and its cross-section.
Cell structure preserved with synthetic resin (Lyofix). The decom-
posed secondary wall is preserved in its loose structure. The pre-
serving agent has penetrated the cell walls 1000x (Braeker et al.,
1979).
Braker et al., 1979; Coles, 1981; Mathieu, 1961; Morgan et al., 1981.
terial anaerobic decomposition sets in after a certain time.
Artificial drying with prevention of shrinkage is expedient.
At present there are several preserving methods, but
none are problem-free. Polyethylene glycol (PEG) is often
used. Drying techniques using alcoholic ether give good
results but unfortunately are slightly dangerous to use.
Freeze drying in combination with PEG is another good
possibility.
A similar post impregnated with synthetic resin (Lyofix) and dried.
The impregnating fluid did not reach the inner parts, so that the
wood there shrank and crumbled. The shape of the post has not
changed.
Freeze-dried wood. The form of the cells remains, but the structure
has changed. The decomposed, loose secondary wall has collapsed
1000x (Braeker et al., 1979).
37
 II Analysis of the materials

The edge of a weathered larchwood beam taken from a wooden house in the
Valais, Switzerland. The beam has weathered in such a way that the lighter early-
wood has eroded, leaving the heavier latewood. This has formed a natural
density profile.

F. H. Schweingruber, Tree Rings

© Kluwer Academic Publishers, Dordrecht, Holland 1988
Anatomical techniques
The anatomy of the tree ring is, among other things, an
expression of the physiological processes in the tree. The
form and dimensions of the cell walls, the number and
size of the cells and the proportions of the different cell
types all provide information about the environmental
conditions that are experienced by the tree.
Microscopic techniques have been used in the field
of wood research mainly for differential anatomy and in
Preparation methods
The tissue structures can be made visible by three basic
methods: polishing the sample surface, cutting a plane
surface or cutting a section. Polished and plane surfaces
are particularly important in the analysis of tree rings.
Microsections
The sledge microtome has proved invaluable in the pro-
duction of microsections. Using this equipment it is pos-
sible to obtain 10 to 20 micrometer-thick sections of
superb quality provided that the following conditions are
met:
the wood should be embedded in carbowax (poly-
ethylene glycol)
a top-quality blade must be used
Sledge microtome with the blade in position.
40
tree-ring biology, particularly in the investigation of cam-
bial activity. Until recently, however, the microscope has
been little used for work in dendrochronology. Today
microscopic techniques are being increasingly applied in
electronic tissue analysis, where the efficient production
of long sample series is required. In radiodensitometry
microscopic preparations provide the basis for the under-
standing of density curves.
Sections
Sections can profitably be used where there is a need to
produce photomicrographs of high quality. They are also
used in the examination of cell walls and in the determina-
tion of the vitality of cells.
the blade must be optimally sharp: sharpened with a
special knife-sharpener, for example
the sample must be positioned in the holding device
with great care
skilled handling of the blade is required.
In addition to these requirements a considerable amount
of experience is desirable.
Knife-sharpener: the friction of the blade
against the fine-grained whetstone produces
an optimal cutting edge.
Grosser, 1977; Schweingruber et al., 1978; Wagenfuhr, 1966.
Cutting

The best sections are obtained from fresh wood. Dry 'V K
wood must be softened in boiling water before cutting.
Conifer pieces can be embedded in carbowax before

'/ ' [ ;;.~/-

cutting .
Where a sample which has been radiodensitometrically
examined is aiso to be anatomically investigated, a part is
removed after radiography and glued to a small wooden /
___ ~
, 1 em
block. This allows fairly large transverse sections to be cut
without difficulty.
2cm

A piece from a radiodensitometric lath is mounted on a wooden

block prior to being clamped in the microtome.

Staining the sections

In dendroclimatology, the dimensions of the cell walls
must be accurately measured. There are several ways of
staining sections so that such measurements can be
carried out.
The first step is to free the contents with Eau de Javelle
(KOCI or NaOCI). After this the cell is stained red with
safran in, which enables the cell walls to be clearly seen.
The procedure is different for biological work on tree
rings. Here the cells themselves, their contents and the
degree of lignification are of interest. Pianeze IIIB stain
can be used. The lignified cellwalls will then become
green, the non-lignified ones colourless and the cell
contents red .

The microstructure in relation to wood density

Density analyses reproduce anatomical features in a more

general form. Density patterns, whether exposed naturally
by weathering or artificially using radiodensitometry, can
be understood only in the light of the cellular structure of
the wood.

0,8 ~

o
o
~ 0.6 1

'"c::
~ 0.4

0,2 -

o L

Here the natural density structure of the wood Is the result of Photomicrograph and density diagram: the natural density structure
weathering. The soft earlywood has been eroded away and the can be made clear either by measuring the cell walls or by using
denser latewood stands proud. light or X-ray techniques.

Schweingruber, 1978. 41
Taking samples
tn •
•
Dendrochronological work, especially in the area of radio-
densitometry, can be successful only where the samples
that are being used have been correctly obtained and
E
u
o
N
~m~//2
) ~----------- ~j--------
mounted. It is only with the invention of the increment I
borer that dendrochronology has been able to develop

properly.
Taking cores from living trees
Coring these days is done almost exclusively with Swed-
ish increment borers. These are available from forestry
suppliers. The 5 mm diameter borers are the most suit-
able for radiodensitometric work. The double-helix screw-
thread enables the tree trunk to be easily penetrated . The
wood outside the borer is pushed aside and only the non-
compressed core passes into the drill tube, which is
slighty wider towards the rear. When the desired depth of
penetration is reached , an extractor spoon is pushed from
the back of the borer between the core and the inner wall
until it is tightly wedged between the core and the top of
the borer. The core is broken off within the tree with an
anti-clockwise movement and then pulled out.
orientation borer bore-support
device
LJ
Diagram showing a bore-support.
trunk from a distance of between two and three meters.
Where the sampling has to be done by only one person a
holding device is used . (See photo.) This can be adjusted
in the axial direction by a peg and screw, and in the radial
direction by a joint.
If, after some time, the borer gets blunt or notched , it
can quite easily be sharpened parallel to the inner surface
with a whetstone. Since it is the outermost annual rings
which are decisive for dating, a bore support is used to
ensure that the borer penetrates the tree trunk in a
straight line.

An increment borer. The inner wall is conical and widens towards

the rear produced by And. Mattsons, Mechaniska AB, Box 9. 5-79201
Mora, Sweden.

Taking a sample from a living tree. The borer is held in the correct
position by means of holding and orientation devices. The bore-
support ensures that straight cores are obtained.

The use of a borer-holding device as an aid to orientation

The slope of the branches usually indicates the position of
the pith . In conifers on hillsides the pith often lies towards
the hill. For boring, it is the orientation to the tree-axis
rather than the radial orientation which is of vital impor-
tance. Where the sampling is carried out by two people,
one of them should operate the borer while the other A borer holding device, showing the borer, the joint and the screw
should indicate the perpendicular to the axis of the tree from above and from the side.

42
The labelling and transport of the cores

The cores should be labelled with a soft pencil while still

moist. Ball-point and felt-tip pens have proved unsuitable
for this. The cores can then be placed in specially-made
holders or in drinking straws.
When the cores have dried out or are beginning to dry
out, they can be taken out of the holder with a needle.
The labelling should then be checked against the entries
in the field-book, and the samples tightly bundled to-
gether and wrapped in newspaper.

A core holder made out of corrugated cardboard. On top a core

inside a drinking straw.

Taking samples from buildings and from fossil wood

It is important to obtain permission to take samples. Local

residents can often be of help in acquiring the necessary
permission. If it is impossible to cut a whole disc from a
trunk in a building, the dry wood must be cored with a
cutting borer. An American make (Tucson) is available.
In regions with relatively high air and wood humidity
the only types of borers which are suitable are power-
driven ones, e.g. Birmensdorf. Tucson B rime nsd ort
Discs are generally cut from fossil wood using a power-
saw. Borer for dry wood.

Mounting the cores

In order to cut the cores into strips or to polish their

surfaces they are first glued into a groove in a wooden
block using a water-soluble glue. They are then held
together until they have fully dried out.
The cores are glued vertically into the groove when the
annual ring surfaces are to be directly examined, and
horizontally when they are to be radiographically inves-
tigated.
The cores mounted on supports:
(a) with cotton wool wound around the support;
(b) held in position with sellotape;
(c) held against another support with rubber bands.

The importance of proper coring for radiodensitometry

For radiodensitometric work in dendroclimatology the u

coring technique is vitally important. The accuracy of the -;jlCo u
"""-
Cl
, ) fibre ~ngle
o·
b) densItometric curve

density values later obtained depends largely on this. E

1Ji
""
1,0j

' lit
.Y. /
!2 ....................... ./ ;/
.r::
E0 ,8
M
E <-~/
.. .............. "

5 0.6

\J
$l
'" 0 .4
.0
~
The influence of fibre angle on the densitogram. Where the cores Vi
c
have been taken obliquely to the tree axis misleading values for the -8 0.2
maximum density result. 16
0 10 12 14
"'"

43
Damage resulting from coring

Damage to living trees

Sample cores are obtained from living trees with an

increment or displacement borer. Damage can result
mechanically from the borer itself, physiologically through
the reaction of the tree or pathologically through foreign
organisms.
Mechanical damage is of economic importance only in
the case of valuable timber. A bore hole can reduce the
value of a whole stack of planks. The secretion of resin or
anomalous heartwood transformation following access to
oxygen results in discolouration. Many kinds of fungi
cause the wood to become discoloured or to rot. Callus-
ing around the bore hole reduces the value of timber.

Beech: uneven discolouration resulting from anomalous heart-

wood formation and subsequent decomposition by fungi around
the bore hole (Lenz, unpublished).
The intensity of the damage

Relatively little is known about this and the literature

varies on the subject. Schopfer (1962) recorded 88%
discoloration in spruce resulting from boring where the
brxe hole was not subsequently treated. Lenz and Oswald
filled bore holes that they had made with grafting wax
(1971). After six years the following percentage of dis-
coloration were obtained: spruce 7-12%, fir 29-57%
and beech 68-92% . After 12 years no increase in the
damage was found. See also the chapter on tree ring
research in phytopathology.
American dendrochronologists and foresters on the
other hand do not seal bore holes. Now, while it is
certainly true that resin-rich species, e.g. pines and some
species with a high tannin content, are only rarely attacked
by fungi after coring, there are nonetheless species in
which coring does cause a certain amount of damage.
In Switzerland an attempt is made to prevent damage
resulting from coring by taking the following steps: Spruce: discolouration by resins (Lenz, unpublished).
The bore holes are sealed with grafting wax (which is
available from garden centers and forestry suppliers).
This very simple procedure should always be followed.
In valuable timber stands or when samples are being
taken from broadleaves with no coloured heartwood,
e.g. beech, coring should be kept to the minimum. It
should in any case never be undertaken out of idle
curiosity.
Branches should never be pushed into the bore hole.
This almost certainly exposes the tree to fungal
spores.

Spruce: callousing around the bore hole (Lenz, unpublished).

44 Lenz and Oswald, 1971; Schopfer, 1962; Shigo, 1983, 1986.

Damage to wood in buildings

Any damage caused to historical wood by coring is

generally slight. As the wood is normally under cover,
fungal decomposition is unlikely to occur.
Where the ends of beams have been sawn off the
effects can be hidden by repainting. Bore holes should be
plugged with a suitable dowel. The load-bearing capacity
of beams is affected only in the case of narrow beams.

Repairing a bore hole in a beam with a plug.

Damage to the landscape through the excavation of fossil trunks

Where fossil trunks are to be examined whole discs are
generally sawn off. If the trunks are buried in peat bogs
or moraines they very often have to be freed using
machines. Sampling must be kept to a minimum where
the trunks are in deposits that are in protected areas.
Today the remains of trunks in arid areas are increas-
ingly endangered. One of the reasons for this is that the
use of 4-wheel-drive vehicles has led to the opening up
of a number of desert regions to tourism. This has meant
that in the process of clearing up the areas the wood that
was scattered about has been burnt. This wood is often
centuries old. In the legendary Bristlecone Forests and in
the Petrified forests it has become fashionable for tourists
to take a piece of wood home with them as a souvenir.
This practice is banned in many of the national parks.
In some cases permission must be obtained from the
authorities in the locality when it is desired to collect wood
for dendrochronological purposes.

Excavating a peat-bog containing fossil trunks. The damage caused is considerable. In this Taking discs from fossil trunks found in a
particular case it is not important, however, as the land is shortly to be drained and built on peat-bog near Saas Fee, Valais, Switzerland
(Bircher, unpublished). (Renner, unpublished).

45
Making the tree rings visible Charcoal

Transverse fractures produce the best results. Where

The surfaces of samples, whether these have been
obtained by sawing or by coring, are generally extremely
rough. The structure of the rings can be rendered visible
by making the surface smoother. There are various ways
of doing this, depending on the condition of the material.
necessary brittle pieces can be wrapped up as they are
collected, or stabilised by pouring hot paraffin over them.
In large fractures the tree rings can then easily be seen
and measured.
Examining the surfaces

Material in good condition
Cutting a plane surface: Transverse radial strips are cut
along discs or cores using a sharp blade, e.g. a multiple
use snap-off blade knife, a carpet cutter or industrial
razor-blades. Where the tree-ring boundaries are difficult
to see, as is the case in diffuse-porous broadleaves,
these boundaries can best be made visible by making the
cut at an angle of between 20 and 40 degrees to the
fibres. Here a satisfactory cut can be obtained only by
The surfaces can best be examined using a stereomicro-
scope with a spotlight or a hand lens. Small details, e.g.
single rows of cells, can be seen only in very good light.
On field trips the author has often used sun-light for this
purpose.
G lUlillUJ • \~

4 ([)(
drawing the blade cleanly along the surface (saw-cut), and

~
this takes practice. Cut sections often expose the impor-
tant latewood cells. Sections can prove to be an excellent
method of making the structure visible, provided that the
lighting is good.
Implements for cutting the cores: snap-off blade knife, carpet
Polishing: This process both smooths the surface and cutter, heavy industrial razor-blade.
pushes planing dust into the lumen. This means that the
cell wall, which is generally dark, can easily be distin-
guished from the lighter, filled, cell-cavity area. Sand-
paper of different grades (100, 200, 320, 400) is used for
the polishing. Power planers are generally necessary
where a whole disc is to be polished. Cores, on the other
Preparing cores which are to be examined in transmitted light.
hand, can as a rule be smoothed by hand. Particularly
good results can be obtained by cutting a section which is
then polished using fine sandpaper wrapped around a
rubber eraser.
For samples from oak the contrast between the cell
walls and the lumen can often be heightened by rubbing
chalk dust into the pores.
In the case of materials where dating is problematic,
whole discs should be examined wherever possible. This
facilitates the location of 'missing' rings or the identifica-
tion of false tree rings.
Microsections: Samples displaying little contrast be-
tween the tree-ring boundaries, e.g., poplar or rosaceae,
should be mounted between two blocks and cut to about
t mm. They can then be examined in transmitted light.
Staining: So far this method has not proved to be very
successful. Isolated cut samples stained with paper-dye
have exposed the tree ring structure in diffuse-porous
broad-leaves. This can be of particular use where trees in
cities are being studied. (Cartasol, red and blue, K-2B;
Sandoz, Basel, Switzerland)
Radiography: See page 64.

Soft, decomposed wood

Very sharp blades must be used to make cuts in the

surface of soft or damp wood, e.g. posts from lake dwel-
lings or trunks from peat-bogs or from alluvial deposits. 20. 200x

Photograph of a polished transverse section of conifer wood

(Pinus ponderosa) at different magnifications. The differing thick-
nesses of the cell walls can be seen at a magnification of 200x.
The structure of the cell walls cannot, however, be seen.

46
Counting tree rings

Dendrochronology is based on the counting of tree rings.

This process provides important information and enables
the age of the wood to be determined. In forestry, for
example, all yield tables are based on tree age. Vegeta-
tion experts and geologists often need to know how old a
tree is, i.e. when its first ring was formed, in order to
estimate the age of a moraine or river-bed.
For trees in temperate zones the enumeration of the Poplar: tree-ring sequence with narrow rings and unclear boun-
annual rings is generally straightforward. This is because daries between the rings.
the clearly defined early and latowood tissue make it easy
to see the rings.
In other cases enumeration is more difficult: e.g. in
species which tend to form less well-defined ring boun-
daries, or where the rings are very narrow, as is the case
in ring-porous woods.
Enumeration becomes extremely problematic where
false annual rings occur, or where the rings peter out or
are missing entirely. It is impossible where the tree dis-
plays zones of irregular growth. In many cases the exact
Oak: tree-ring sequence with narrow rings. The earlywood vessels
age of the tree can be determined only by applying the are almost touching each other and the latewood is barely
dendrochronological technique of cross-dating. developed.

Tree-ring sequence displaying a lot of width variations and 'false' Tree-ring sequence where a number of the rings are missing and
rings. Enumeration here is problematic. others wedging out. Enumeration here is problematic.

Trees and shrubs in temperate and boreal regions

generally form one ring a year. The occurrence of clearly
marked seasons in these regions means that very little
variation in growth occurs within any given year.
There can, by contrast, be enormous variations in
growth in trees and shrubs in arid and semi-arid regions.
This is a result of the unevenly distributed precipitation.
A ring may be formed every year, or there may be no
growth whatsoever, or the tree may even appear to form
two or three rings a year. For samples in these regions
age can be determined only by cross-dating.
Enumeration is generally extremely difficult for samples
in tropical regions. Only very few species are able to react
morphologically to the very slight differences between the
seasons and so produce an annual ring boundary as an
expression of a break in growth. Very few species in
these regions produce annual rings. Most produce either
no ring at all or just a suggestion of a growth zone. (See Cross-section through a piece of tropical wood. This species
pages 109, 237). (Cassia) displays growth zones but no annual rings.

Tree-ring sequence of a tropical broadleaf (mahagony). No rings

are discernible. Ennumeration here is impossible.

47
Recording the changes of growth patterns

Every change in environmental conditions is reflected in identified and dated with the naked eye or with the help
some way in the annual ring structure. Relatively rapid of a stereo-microscope. For any given disc or core, both
changes sustained over a number of years occur as a the point at which the variations began to occur, and their
result of the following: duration can be identified. The results of the analysis of
individual samples can then be brought together in a
changes in the tree's position single diagram. It is important to determine both the date
changes in the light, water or temperature conditions and the duration of these changes or events. The sum-
damage to the tree's crown, root or trunk system
mary diagrams enable the identification of individual, local
chemical changes in the environment. and regional changes. This is proving to be of particular
The changes affect the number and size of the cells and value in our age of pollution.
the thickness of the cell wall. Such variations can often be

I I I I 1 I I I 11111111
Examples of changes in tree ring width. Reductions in growth of I I I I I I I I I11111111111
less than 40% and recoveries of less than 150% following periods
of reduced growth cannot be seen with the naked eye. I Ili);111111
The intensity of the change can be quantified by measuring the I I II 1llliilU!
annual rings and comparing the number of those showing an I I I Ilid! 11111 1I:!I1I F IJotIn IJ011~
increase or decrease in growth with the same number of previous
rings. I II I I I I I I I 11;;:11111111 1 ~ 1100, :xlO"

I! I i I I I I hili! I I I I v ,

1st dnnud'
Cores In 4 1
rl09 N
4 19O;~ I «

]
5 H)4~

1947 j «
6
1')'>4 4 ~

O""~'r 01 I t"lif'f\\

1I1n.)Ci1'l I 1.lI I demltv f'on

fQW1C'
ft"'lnduc t ~
"'~""IIDn ",.,

age trend; the rings 5 abrupt growth decrease

become continuously smaller and abrupt growth release
2 abrupt growth decrease 6 continuous growth release
3 stepwise abrupt growth 7 abrupt growth release Phase
decrease
8 compression wood
4 abrupt growth decrease o
and continuous release 9 special features ;Z;
2 ~

The most common patterns of change in the annual ring width.

The abrupt changes - cores 2 to 5 and 7 - can be seen with the
f------'==i======i====9 3 "
u

naked eye. Such changes can often be dated by counting the rings
or by using pOinter years.

Presentation of the beginning and duration of growth changes.

Diagram showing 4 cores with abrupt growth reductions. The BO
results are summarized in the event diagram, phase diagram and
the circles to present the number of cores with reductions in a
specific year or period.

'-'--'--"--' time scale onset of change Duration and intensity of change I tH ·~_au 1

first annual ring 0 sl ight 40-55% reduction >-----< 40-55% reduced ,/. 't. O\lt.'ft'O
JII',juo t'd
[;iii moderate 56-70% reduction

•
• at heartwood I===; 56-70% reduced
o on core great> 70% reduction ~ > 70% reduced

48 Kontic et al., 1986; Schweingruber et ai., 1986.

Pointer years and the associated characteristics of tree rings

The dating of a tree-ring series is based on the identifi- considered to be complacent.

cation of rings with characteristic features. Dendrochro-
nology is primarily concerned with those rings which are
particularly narrow. Where these occur singly the years to
which they relate are termed pointer years. Where they
occur in a characteristic grouping they are referred to as
signatures. Where there are a number of pointer years
close together the tree-ring series may be referred to as
Sensitive; where they occur in a loose sequence, it is
The term pointer years has been variously defined. In
the skeleton-plot method it is used to mean those narrow
rings which occur fairly often. In the context of measured
ring series and chronologies, it may refer either to those
tree rings which are on average narrow or to those
intervals in a chronology which tend to rise or fall.
The width of the tree rings is one of a number of
features displayed by the ring which may be of help in
dating.

1530 n n n n n
Famous signatures in tree-ring sequences. On the left the Pueblo signature in Conifers of the American southwest (Douglass). On the right

WIIII////I,.,fll/J IVI;))))))))))))))))1\\\\\\\\"
the Saegesignatur in oaks from southern Germany (Huber). (n = narrow)

sensitive annual ring widths complacent

\\\ \ \\\\ \\\" \ \\" \\ I II m "I H\\\, \\\ l\m

\ \, \\1/111\ sensitive late wood widths complacent

d,stlnct minimal
Intraannual vaflauons In densltv

resin ducts compreSSion wood trau matic tissue

Characteristics in tree-ring sequences.

~.~~~
, ""
'\ \\\\' "~\'\\\\\""'I'lll A'
lll lIlIllIlIJlIl
'" '"
A A" I ~
A
N·,~ .. ,re"" . . . ~'"".... ~ ' :I ~ . . ~~ ~
~ ~ ~ ~, .. ,............. ~ \ ~"~ ;' ~' ~ m
, .............. '\ I / / ,-

~ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \\ I III \\\ \n\\\ \\\\,1111,111111111 \I U1\ IIIl "'Ui JI

I , ' , ' , ) '" ( ( ,

Visual matching of two tree ring sequences using pOinter years. That for 1976 is narrow and displays very little latewood. Those for 1962
and 1964 are of normal width but have very little latewood. The samples were taken in the dry Rhone valley area the Valais Switzerland
(Kienast et a/., 1981). ' ,

49
 Cross-dating using the skeleton-plot method

American dendrochronologists have used this metrod for

almost 80 years dating conifers growing under all types of
climatic conditions. Before any actual measurements are
made, an outline or skeleton of the tree-ring sequence is
plotted. Those rings which are characteristically narrow
are recorded subjectively, as they occur, on strips of pa-
per with 2 mm divisions. The width of each ring is com-
pared with that of the rings on either side of it. Where a
ring is considerably narrower than its neighbours a long
vertical line is drawn . Where it is only slightly narrower a
shorter line is drawn.
Since age-trend and long-term changes in the ring
width are not recorded using this method, it can be used ,oS
to synchronize ring series of the same age but with dif-
ferent absolute ring widths. In this way samples from a
fairly narrowly defined site can be aggregated to form a
composite . Insofar as only those variations in width which
occur more than once in a series are taken into account,
these local series serve to highlight only those narrow
rings which are characteristic for a locality.
"
Master chronologies can be constructed from these
local series. Such chronologies show only those narrow
rings which characterize a region or a particular altitude or The skeleton-plots of individual radii have been aggregated
site. Samples can then be dated by matching these plots to form a Single plot. This can then be related to the master
with recent tree-ring series. chronology (Dean, 1978).
Experience has shown that plots produced indepen-
PIOI DW 1969
dently by researchers with sound basic training tend to be
similar. The method is in fact only at first sight subjective.
Cropper (1979) has illustrated this in that he obtained
similar results from skeleton-plots for given tree-ring
IIII I
PIOl881973
t II LI I IIII II III I" I III II II 111111
series and from calculations for these series.
The method can be used to obtain a considerable
amount of information . This leads one to wonder why I
I IIII I 1,,110 II I
I I
European dendrochronologists do not made use of it. The 1710 20 30 40 1750 60 70 80 90 1800
significance of sketeton-plots as a means of identifying
Skeleton-plots of the same ring sequence drawn by two different
ecological changes remains largely unrecognized . people at different times. The agreement between the plots can
clearly be seen (Courtesy of the University of Arizona).

The relationship between a wood sample and the skeleton-plot ~nd

the tree-ring width curve for this sample. The s.k.eleton-plot hIgh-
lights the very narrow rings (Dean. 1978). In addition to the annual
ring parameters other important information can be noted on the
skeleton-plot:
.
8K·' 4 8

a
pp: pit h present 8K~ 186
np: near pith
± : the pith cannot be located
B: bark present
G: beelle galleries on the surface
r: terminal ring present
vv : number of rings missing not known

50 Stokes and Smiley, 1968; Cropper, 1979; Dean, 1978.

Cross-dating using graphs

! 1
1 i11111111111[::-L..ULULLLL..Lll..lL.lU...~
III " 1111111111111111111111111111111

lilllllllllllill i 11111111111 i
I

1 II111111111
I
I
I
I i 1 : i i 1 OtJU'tr~~:~
, 1875
I
+f~~~~--~~~~4.-.~~ ~

h·ll,.ct IS -nuwn
1190 I BOO 1870 '8](1 11\40 11160 .810 '950 '960 1970
•
'980

A schematic representation of the use of the bridging technique. The irregular occurrence of wide and narrow tree rings enables the samples
to be dated . Matching the inner layers of living trees with the outer layers of beams in a building means that the samples of known and
unknown age can be arranged in chronological order.
4

The measurement data are presented in the form of

3
curves which are then optically aligned. This procedure is
known as cross-dating. E
.§
Ring width may vary consicerably within any given 2
sequence. The ring may, for example, be 4 mm wide in -5
"0
the center and 0.25 mm wide when the tree is a hundred ~
years old. The ring width values tend, especially in
Europe, to be expressed semi-logarithmically: The time-
axis (abscissa) is linear while the axis showing the width 0
5 10
values (ordinate) is logarithmic. This ensures that even in time (years)
short series all the characteristically narrow rings can be
Arithmetic presentation of an tree-ring sequence. The differences in
seen clearly . width appear equally great for both narrow and wide rings.
Because density varies only minimally it can be ex-
pressed on a linear scale. 4
3
Nowadays, the curves are reproduced in the form of 2
computer plots. But it is certainly worth emphasizing that E
plotting by hand is still a perfectly acceptable way of E 1
presenting the data. .J:.

It is essential that the coverage of the chronology be -0 0.4

.~ 0.3
indicated, either in the form of a histogram or as a table of 0.2
values on the diagram itself.
Where the material is to be expressed in summary 5 10
form it is often confined to the number of tree rings (Ime (years)
analysed for a single tree or chronology. Bar charts are Semi-logarithmic presentation of the same sequence. Differences in
used to express these values. width emerge more strongly for the narrow rings that for the wider
ones. (Aniol 1983)

51
The measurement of tree-ring width

The first step in the measurement of ring width is to locate Where modern equipment is used the values are
the rings and to choose suitable places to take the printed out and registered in digital form. In the latest sys-
measurements. Samples where the rings are relatively tems, the digitization unit is connected to a computer
wide and which have been taken from species with fairly which can be programmed to print out the curves and to
clearly-defined tree-ring boundaries are straightforward synchronize them with existing ones by means of appro-
to measure. Problems arise where the samples are of priate testing. Here all the steps, from sample preparation
narrow rings, particularly where there is little developed to the final dating, are carried out in one place.
latewood or where the species is ring-porous and so has
practically no latewood. Measurements should not be
taken where there are irregularities, e.g. pressure wood,
knots or hazel wood or where the ring boundaries are
displaced because of rays.
The measurement of both early and latewood in coni-
fers is problematic in all but a few species. And wherever
there are no clearly defined boundaries between the early
and latewood accurate measurement is impossible. Com-
paratively good results can be obtained when the early
and latewood boundaries are first identified and marked.
Any optical instrument which can reproduce tree rings
to an accuracy of 1 mm, i.e. render them capable of
being measured, can be used for dendrochronological
purposes. The simplest pieces of equipment that can
be used are hand-lenses or stereo-microscopes with a
calibrated eyepiece. Here the values are read off and then
recorded by hand.
In the earliest versions of the ring measuring equip-
ment the widths were fed into a counting device, either
directly or via a spindle. This device then printed out the
values and the curves were drawn manually. Hand-lens with a scale: the simplest optical measuring instrument.

optical part recording the data data analysis

. rn
stereo-microscope

sample fUl).)~
. 0 .. • O~
.pmmuawumlit!W~@~~~::;!~mllm@l!!lc14 '--_~_"" .
1----"
sledge :
.,
I .,.
~. ~ /

T interface ... -...: , :

1 -
i
5;.
'.

I simple_ve_r_sl_on_____________~

I the first measuring machines

I m~re recent measuring equipment

I modern measuring and calculating equipment

Diagram illustrating the stages in the development of tree ring measuring equipment.

52 Huber, 1970.
An early version of a tree-ring measuring machine. This model, An early version of a tree-ring measuring machine. The samples
which was made in the 1930s, measured and recorded the annual were placed on a stage and moved around under a monocular
ring sequences using a mobile telescope (Courtesy of the Labora- microscope (Courtesy of the Laboratory of Tree Ring Research,
tory of Tree Ring Research, Tucson). Tucson).

terminal
and
printer computer

Diagram illustrating the tree-ring measuring system in the Centre for

Forestry Research, Laurentide, Quebec. The camera can be moved
about above the trunk disc. Observations are made using a color
television screen.

ltVnfh.(r,-ln D "n1TT£U(UItV( IU SlAn.D 0!1C1l1l

Machine-made tree-ring measuring equipment with a counting ,..
device. For twenty years this Eckland machine was the most widely- I
used annual ring measuring equipment (Courtesy of the Laboratory
of Tree Ring Research, Tucson).

A modem tree-ring measuring system wHh integral recording and The tree-ring curves are printed out synchronously. Statistical
calculating facilities. The dendrochronological laboratory of the City information about the reliability of the synchronization is thus
of Zurich. obtained (Schweingruber and Ruoff, 1979).

53
Techniques of tissue analysis

In the wood structure of the tree the effects of external and internal conditions are manifested in the form of a tree
ring picture. The structure thus revealed can be investigated ecologically in one of two different ways: using structure
analysis or by measuring the width and the density of the tree ring. Since more elements can be measured with
structure analysis this method is preferable, but until recently the amount of time needed was prohibitive. Today, the
application of modern electronic techniques enables structure analysis to be used efficiently.

Applications

Structure analysis can be applied to many different fields: patterns of wood enable a picture of environmental con-
to differentiate between various types of cells in wood, for ditions to be built up, in growth, in dendroecology and
example, or since studies of changes in the growth dendroclimatology.

(mml
H)OO
(pm 2 )
c
0
800
'"~
~ 4800
ffi
'"
'"
"0
a.
U
Q)
GOO
'"E en
'"E '"Ci
z
a o'" 400

Percentage of vessels
Percentage of fibres
15.2
70.1
22.7
54.9 12400
vessel area
Percentage of ray tissue 10.9 15.7
Percentage of axial parenchyma 3.8 6.7 I
Vessel density (number in mm) 53 82 1200
Mean vessel diameter).J 56.6 65.7
1930 1940 1950 1960 1970

Vessel area in the analysed stem dependent
on the age of the cambium (Eckstein et al.,
1977).
Structure analysis of tree rings before
and after damage. Acer saccharinum
from the center of Hamburg affected
by the drainage water from road-salt-
ing (Eckstein et al., 1974).

Methods of structure analysis

There are a number of different ways in which the struc- this structure can then be calculated. The more sophisti-
ture of wood can be determined. The basis of most of cated models carry out image analysis proper.
these methods is microscopic examination. Both micro- It is clear from the work that has been done in this area
sections and polished sections are suitable for this sort of up to now that structure analysis and radiodensitometry
analysis. produce comparable results. The basic steps of structure
For over a hundred years it has been possible to do analysis are presented below.
microscopic measurements using calibrated eyepieces.
Such measurements do, however, require considerable
time, especially in the preparation of the material to be
analysed. Nowadays data aquisition is accomplished pri-
marily through the use of electronic image analysis. The Material Optical Digital isation Computer
principle behind such analysis is straightforward. The type Examination Plotter
of equipment available ranges from the relatively simple
to the very sophisticated. Microsection light trans· T.V. monitor ') Grey-level
In the case of the less sophisticated equipment a mittance or plate with - } analysis
microscope conducting
micrograph is projected on to a plate with a fine con- grid
ducting grid. The structures can then be traced using a
Pol ished su rface (identification - Evaluation
marking pen. This causes impulses to be triggered off by surface illumination of co-ordinates
the grid which enable the x/y coordinates of the outlined microscope and enumeration)
structure to be identified. The area and the dimensions of

54
 specimen (e.g. microscopic or
macroscopic preparation)

optical input
(e.g. photograph, microscope picture)
t
optical image

image-conversion
(via T.V. system)

"OC',"O;' im."
detection (according to grey-level);
prior shading correction where necessary.

di9i":im."
image modifi cation,
correction

corrected "ideal" image

,
measurement, analysis,
data output

data
An image analysis system: 0: analyzer 'Ouantimet' (with monitor); C: computer with
console K; M: microscope with mounted television camera (Sell, 1978). Diagram illustrating the stages in the pro-
cess of electronic image analysis (Sell,
1978).

Scots pine, Pinus silvestris Xanten. BRO. 30 m a.s.1. magnification x 25

Comparison of the results of the determina-

tion of wood-density using:
(a) image analysis - where the percentage
area of the strips of tissue are multiplied by
the value obtained for the density of the
cell-wall 1.5 g/ccm.
0,2 (b) radiodensitometry - where a value is
obtained from the optical density of the
X-ray film.

55
Photometric techniques

With the discovery of the close relationship between the

density of wood on the one hand and its biological qual-
ities and technological aspects on the other, research
work in a number of different areas, for example the
biological sciences and the pulp and paper industries, has
been devoted to the development of efficient photometric
techniques to replace the time-consuming microscopic 1969 1968 1964 1962 1961
techniques which were being used earlier. The resulting
methods, which still find their application today, can be
'~------------------------------~vr------~I
said to be the forerunners of radiodensitometry proper
(Green, 1965).
The principle behind these techniques is straight-
forward and is very similar to that followed in micro-
densitometry. The light passing through a given surface
of the microscope objective is measured by a light cell
and transformed into electrical impulses. Using this
method both microsections and polished surfaces can be
examined.

Fields of application of these methods

Individual pieces of wood from trees of the same species

or species group are suitable for analysis using this
method. The differences in density can be recorded
numerically or in graphical form and then compared.
Density profiles which are of as high a standard as those
produced radiodensitometrically can be obtained using
photometric techniques, provided that the equipment Photometric measuring equipment for the analysis of wood samples
used is good. under surface illumination - 1: microscope; 2: specimen table;
3: interface; 4: plotter; 5: impulse trigger; 6: plot (tracheogram)
(Terskow etal., 1978).

light transmittance
%

40

transmitted
30 light

20

1951
10
reflec ed
light

I I I
, I
2 4 6 8 10mm

Photometrically produced curves showing the light transmittance of Light transmittance curve for an X-ray film (densitometer) and a light
a microsection of red spruce (Picea rubens) (Green, 1965). reflection curve for heartwood from a Siberian larch (Larix sibirica)
(Vaganov and Terskow, 1977).

56
Calibration

Sapwood or heartwood of microsections or polished Polished surfaces

surfaces can be used for calibration. Volumetric-gravi-
Since the technique is based on the reflection of light,
metrically determined densities agree with photometrically
colour differences within a trunk are reproduced and
determined ones. An overall calibration, however, such as
differences between sapwood and heartwood and irregu-
the X-ray analytical process provides, seems unfeasible
larities in colour falsify the density values,
with polished surfaces and light transmission through
sections.
Microsections
Reliable density profiles can only be obtained from sec-
0.1 tions of absolutely uniform thickness. The preparation of
• large sections is difficult and often impossible. Coloured
substances in rays and other cell elements are irregularly
0.6 distributed over the transverse surface; they reduce light
transmission and reflection and consequently the density
value measured. The obvious answer is to use stained
05
sections, although Mueller-Stoll, 1947, reports that
0
absorption of stains varies from species to species. The
.!::!. results obtained using this method cannot be compared
Ol
with each other.
c- OA
'"<1/c
"0 heartwood sapwood

0.3

0.2 ~

0.1L-____ ____ ____ ______ ____ ____

~ ~ ~ ~ ~ ~

0.4 0.5 0.6 0,7 0.8 0 .9 1.0

proportion of cell wall

Relationship between wood density (ordinate) and photometrically Light reflection curve from a sap-heartwood transition zone. The
determined proportion of cell walls in stained microsections of Picea darker heartwood reflects less light than the paler sapwood
enge/mannii, Pseudotsuga doug/asii, Tsuga sp., Picea sitchensis (Vaganow and Terskow, 1977).
and Abies ba/samea (Green, 1965).

The relatively simple method i.e. that using radiodensi-

0,8 tometry, is appropriate when the relationship between the
values and not the absolute values themselves is investi-
gated. It was used by Mueller-Stoll (1947) in his work on
0,7
the problem of earlywood-Iatewood demarcation, and by
Vaganow and Terskow (1979) in their investigation of
0.6 ecological problems in relation to wood growth and tree
0 rings. It is possible that photometric techniques can
~ 0 ,5 produce usable density profiles from decomposed conifer
C- wood which is unsuitable for radiodensitometric analysis.
.;;;
c 0,4
QJ
"0

0 .3
.• •
0.2 •
0.1
50 60 70%
I igh t reflex ion

Relationship between volumetric-gravimetrically determined wood

density (ordinate) and light reflection from a surface of Siberian
larch. (Larix sibirica) (Vaganow and Terskow, 1977).

Green, 1955; Muller-Stoll, 1947; Terskow et al., 1978. 57

Basic steps in radiodensitometry

Obtaini n9 the samp le cores

from I iVlng trees

Cu tt ing out a lath

X - raying the lath

d'
I
I ' \\

1/ \
Removal of resins and heartwood substances
I \\
I

the density of the wood IS indicated by the degree

10 which he III m blac ens.

58
 Measu rement of density with the densitometer

visual check

Analysing
choice
the film

lepr senlalion of Ihe comple e registration of the main

optical analysIs
annual ring d nSlly curve 0,' annual ring parameters
0"
0' ,
\\\\\\\\\\ \\\\\\\\\U
iJIII

aper lure

f: ! : :
<==>
..
\

I' ,\ '
\t
'
" ,

lighl-source

Standardiz ing the annual ring parameters and

expressing them in graph form enables dating and
and climato logical analysis to be carried out.

, ..-...... .J../ I I \,,j"'......... ~ . . \v.. \_. ~~. .:" /,: ... " .~ ... ' .. '... '

.' ",' " \ , ' \' . : ,,;, .. , : ...\ , ...., , " , •• - .... ~ : • " .. j ~- •

. . ......... , r· ' .. ' .. '~ . . . ,' ........ /',...,/'~. ",' .. "~ . . . .".... .;.,
. , , ' ........ , ....- ........... _ .... ' .......... . ,'.
,~ ",
................................. ............ . .

The results can be compared with findings in

physics, biology, ecology, climatology,geology.
glaciology. histC\ry, archaeology. etc_
Q [.::~. . \:'" ..... u u • •, . \ ' . '~ •• ' U.U ,__ I
...............................................

Lenz et al.. 1976; Green, 1965; Parker, 1971. Paige, 1966, 59

Sawing

Specimens of uniform thickness are required for radio- X-ray picture will not be clear. The use of an orientation
graphy. For this reason the laths are cut from round cores device permits some correction to the orientation to be
or rectangular samples. The sample must be cut per- made, thus allowing such faults as oblique coring or spiral
pendicular to the direction of the fibres, otherwise the growth to be partly compensated for.

Sample core , 5mm In diameter, obtained using an increment borer.

Sample core , 10mm in diameter, obtained using a drywood borer.

Rectangular sample from a disc taken from a trunk.

venical Iine In the eyepIece

of the stereo· mIcroscope

( )
sample on the
movab le p a
l te

wood fIbres here t heannual ring bou ndaries

devia te from the perpendIcular
by about 50

Sample cores glued onto the wooden supports.
The sample has been glued to a wooden support. The fibers in the
sample are aligned with a vertical line in the eye-piece of the stereo-
Illicroscope. This is made possible through the use of a holding
device. This consists of a fixed and a movable plate, the latter
holding the support on which the wood samples have been placed.
The degree of variation from the perpendicular is noted on the
support.

scale on the fixed plate

eyepiece with vertical line

The holding device with a movable upper plate and a scale on the
The orientation equipment. fixed lower plate.

60
 With the sample still in position the holdfng device is
placed underneath a twin-bladed circular saw and a lath,
usually 1.25 mm thick, is cut out. The saw blades, which
are of the sort normally used only for metalwork, cut with
great precision (1-2% variation in width). In the case of
well-bored samples only a single cut is necessary while
for badly-bored ones two or three overlapping cuts may
be needed. After sawing, the unwanted wood at the side
of the lath is broken away with a rigid plastic strip. Inci-
sions are made with a sharp blade along the base of the
lath, which is in turn carefully detached from the support.
The samples are labelled with India ink: this does not
show up on the film. The thickness of the sample is then
measured with a micrometer, this measurement being
necessary for the subsequent density calculations. The
accuracy of the coring and the cutting determines the
clarity of the X-ray picture: for good samples the tissue
structure is clearly visible, and for poorly-prepared ones it
is blurred.

feed
backwards/forwards motor

feed right/left

Sawing equipment for the production of laths for radiography. Detail: twin-bladed Circular saw, with thrust strip in position, in
operation.

S S II II Core orientation in relation to the fibre angle

~! '!
I I I I

ii Zi :111>1
I I I I

II rl
!' I II E. EZ i :2
I I I I I
Well cored sample

Bad cored sample. Corrections have to be made by overlapping saw

cuts.

Detaching the cut lath from the support. The quality of the X-ray picture.
Top: Good - here the structure is clearly visible.
Bottom: Poor - here the structure is blurred.

61
Removal of extractives

Certain parts of wood contain substances which do not
directly belong to the tree ring structure itself. Where
radiodensitometric work is to be done on the wood these
substances must be removed because:
such substances generally enter the tissue after the
ring has been formed. Resins, for instance, are pro-
duced continually by the secretion cells of the resin
ducts; other substances appear in the cells during the
transformation of the sapwood to heartwood, and
water is distributed more or less locally throughout the
tree-trunk.
they vary enormously in the rate at which they absorb
X-rays: resins absorb relatively little compared to
cellulose: water absorbs slightly more; and certain
other cell substances display considerably higher
X-ray absorption.
Resins should be extracted in a Soxhlet extractor in
alcohol or acetone. This process takes several hours. The
heartwood substances of a number of species, including
larix, tsuga and fitzroya, are water-soluble and so can
be removed by boiling in water. Extractables may be
removed before or after lath-production.

Extractives present in
woods but not formed at
the same time as the tree
ring tissue.

Quercus sp. 400:1 Tetraclinis sp. 100:1 Pinus strobus ca. 400:1
Cross-section. Tannins in the Radial cut. Phenolic sub- Cross-section. Resin in the
wood cells. stances in the parenchyma resin ducts.
cells.

cellulose acelale
glce
kal'brauon malenal)
1.2

::-
.~

'"
"0
1,0 resin. spruce
t-------- water

E
Q; 0 C
E
:;
'"
gc- 0 .6
heanwood ~ubslances. farch ..
•••••••••••••••••
;;; wood : pone and larch
E t--- wood. spruce
-5 0 .4
:> wood: Slone,p,ne

0.2
t----- holocellulose
30 20 10 o 10 ?O 30 40 50%
X-ray pictures of larch heart- Extraction apparatus (Soxhlet).
de"'dl,on of lhe radlogrph IC from lhe wood; left: before the
-.plumel"c·gravi melfiC densllY extraction of heartwood
substances; right: after
The rates of X-ray absorption for different woods and wood sub- the extraction of these
stances in relation to cellulose acetate. substances.

62 Parker and Jozca, 1973; Schweingruber et al., 1978.

Acclimatization

Where samples are to be radiodensitometrically analysed

to obtain climatological information, the samples must be
dried. This is because the distribution of water in a tree- 1.1+-____________ __ ~ ~-- __ " ...... - .... ,
L---~--------~

trunk is irregular and water has a higher X-ray absorption

than cellulose. The samples must all have the same water
content if comparison of their absolute values is to be
1.0+---~----~~----------------~~~~~
valid. Since wood is hygroscopic, its moisture content can
be controlled by keeping it for some time at a fixed
maximum density
relative humidity. In the relatively damp European climatic
zone wood is radiographed in conditions in which it
may naturally reach hygroscopic equilibrium. In this case
the samples are acclimatised at 20°C and 50% relative
humidity. This gives a wood moisture content of approxi-
mately 9%. (The amount of water is taken as a proportion
of the dry weight of the wood). Air-dry, 1.25 mm thick
----- - wood mOisture content 12%
laths should reach hygroscopic equilibrium in one to two
hours. - - - - wood mOisture content 7%

mi nimum denSI y
O.3 +-------/7,~'-'~'~~----~~----~~~~--_1
Maximum and minimum density curves for conifer wood. The water .,.~ .......

content has been adjusted to 12% and 7% respectively, and can be

traced radiographically and quantified using densitometric tech-
niques (Schweingruber et al., 1978). 0.2 +-- - - - - - - - - - - - -- - - - -

Positive of a radiogram, showing the irregular distribution of water in wood. (The dark patches are water.) The distribution is noticeably
irregular, both in the individual cells and in larger areas. (5X) (Polge, unpublished).

63
Radiography

The X-ray equipment and how it is used:
Safety regulations require that X-ray proof walls and doors
be installed and that there are safety switches which
enable irradiation to take place only when the room is
sealed. Dosimeters must be used to check the amount of
radiation to which the operators are exposed. The type of
equipment most commonly used is produced by the
Belgian firm, Balteau.
The films used:
The most suitable of the many types available is the
Kodak RP/M. This has a density scatter of only 0.007 over
the whole film surface.
Developing the films:
Films which are designed for rapid processing take only
90 seconds in an automatic processor. The equipment
that large hospitals have at their disposal is always kept in
good condition and is therefore excellent for the develop-
ment of radiodensitometric films.

Samples in a cellophane carrier ready for radiation. A radiogram of these samples. Density 0 is taken to correspond
to the blackening of the underground of the film.
g/e m'

,.; I
A

;~. X·ray source : ~ 0.1

',0
sodium lamp J:.
I···· ~ 0 __. .~--~~--. .~------~------. .~
....
:;1: '"
J:.
B
/iI, ~ 0.1
, I! E
~ radiation chamber
1',!1
preparation room ~ 0 I C
N
~ 01
'0
/I : ~
:1 :
:

iii : I
IiI

samplesi 0.1
D
• Li' ••
sampleslfllms
~;;;;;;;;;;::=;:. 101
E

5 10 15 20 25
em

....::;~: .
~
Product EmulSion Vol· Time Scatter
coaton!l tage of the
mean
Plan of a radiation chamber. The radiation chamber is separate denSity
from the preparation and control rooms. on g/cm

A - Kodak RP/M double Sided 16 kV 10 min 0.007

B - 3M Medical double Sided 14 kV 4 min 0.012
X-Ray. Type S
C - Kodak PE 4006 double Sided 14 kV 4 min 0.017
D - Agfa CURIX dOuble Sided 13 kV 10 min 0.030
RPI·PE·FW
E - Kodak RP 'SU one Sided 19 kV 8 min 0.035

Density scatter in different types of film.

Film, coating, voltage, time, density scatter about the mean in
glcc (Lenz et al., 1976).

64
Irradiation:
The X-ray source is 2.5 m above the film. This means that
practically only parallel rays from the center of the X-ray
tubes penetrate the samples. This ensures that the
boundaries between the early and latewood are repro-
duced without blurring. The samples are in a cellophane
carrier which is placed directly on the film.
The center of the ray geometry is determined using a
Geiger counter and the film itself. The duration of irradia-
tion depends on the distance between the radiation
source and the film: the longer the period of irradiation
the sharper the contrast obtained. For this reason wood
samples that are only 1.25 mm thick are irradiated at 11
kV and 20 mA for 90 minutes.

glee
1,2
sample 1 2 3 2 3
3mm 2mm 1,25mm 2mm 1,25mm

.'"
C.
1,0

E
:Jj 0,8
.Y.
u
~
E 0.6
E
M
c:
0
~ 0,4 The influence of X-ray exposure time and sample thickness on
] radiographic contrast.
Group A: spruce samples 3 mm, 2 mm and 1.25 mm thick were
>- 11 11'1'
..
.~ 0,2
c:
"0
j ,

group A
\11
group 8 ----j
uniformly exposed for 6 minutes. The density contrast falls with the
decreasing thickness of the sample, The absolute density value
(scale) is valid only for the 3 mm thick samples.
Group B: samples 2 mm and 1.25 mm thick were exposed for 25 and
0
180 minutes respectively. By prolonging the exposure time the
contrast can be increased to approximate that obtained using the 3
0 5 '0 15 20 25 30 mm thick samples,

Ray geometry of the X-ray tube,

A field with uniform irradiation intensity (isodose lines). By arranging
the test films in a cross the areas receiving the same dosage of
irradiation can be determined. The blackening of the film is practi-
cally uniform from the centre to the edge of the rectangle represent-
ing the surface of the film,

Lenz et al., 1976 65

The radiographic-densitometric measurement of wood density

The optical density of the film is measured using a
microdensitometer. The film of the radiographed samples
rests on a movable table which passes over a light source
with a slit-shaped aperture. The light signals of varying
intensity are transformed by a potentiometer into electrical
impulses. These in turn activate a strip-chart recorder . A
magnetic storage unit registers selected values . The
densitometer table , the rolls of paper for the strip-chart
recorder and the tape-punch move in time with one
another in response to signals transm itted from the con-
trol mechanism to the stepper motor.
A new, complete radiodensitometric apparatus is pro-
duced by Kutschenreiter, Siccaburgstrasse 64, A-1100
Vienna, Austria.

The equipment

Measurement of denSltv wIth the deOSIiOmeter

visual chee

Analysing
the 101m

reprf'SCl1lauorl of the complete rt:<)f trilliOn 0 he rT1dll)

opU ,JI
.'lIlnu~il ring density CtHvc
Schematic representation of
a densitometric equipment
Iiln

Radiographic-densitometric equipment (System Birmensdorf 1983)

66
 41-----------_ photocell

I 4l---=-~
I
objective
I
I
!
<::P I

optical wedge
light measurong silt
- I- servomoter
I I
I I
i
I reference beam
measurong beam

: I
{r----@----~
light source

Diagram of a double-beam microdensitometer

Microdensitometer (Joyce Loebel MK III CS) The measuring and reference beams are switched alter-
(a) housing with optical wedge natively through a light at a frequency of 50 Hz. This cell
(b) eyepiece transforms the light into electric current.
(c) objective A servomotor moves the optical wedge in such a way
(d) potentiometer
(e) plotter and recording table as to ensure that the values of both beams are equal.
(f) specimen table The amount of displacement is then registered directly
(g) stepping motor. by a pen or a potentiometer. The density curves can
then be produced (Polge, 1966).

ddta sourc da ta acqUisi tlon data plolting

r'-'- '- '- '- '- '-'-'- '-'-'- '- '- '- '-'-'- '- '-'
r continuous recording I Ir~~~l

chart-reco rder

J
density peak values
Integrator (y)
(y) \ selective

"''"'::':-__u'~~~~~oc~~_=:I_j '1" "i:,::::1

:I IL _________ ~ ___ ": , : :,'-

_ _ _ _ _ _ _ _ _ _ _ J. +
I -
__ '---_ _ _I---;:===~--~
reroodlc record; ng
.. .
: :: : ,.:

x·values
COlli ler
(stepPing motor)

densitometer control deVice plJnch tape

Diagram showing the functions of a data acquisition system
Impulses are sent from the densitometer to the control
mechanism, the chart-recorder and the tape-punch.
The control mechanism can, in response to signals
from the stepper motor, register the data in a continuous,
periodic or selective form.
The chart-recorder continuously registers all the sig-
nals that it receives, in the form of a two-dimensional
graph (analog values). The tape-punch registers data in
digital form, and can do this continuously, periodically or
selectively.

67
The principle of radiodensitomeiric determination of wood-density

A cellulose acetate calibration wedge forms the basis for

each of the calibrations. This wedge has a sloping face of
continuously graded thickness and a stepped face. The
density of the wedge material (1.274 g/cc) and the thick-
ness of the steps are known. The wedge is radiographed
with the wood samples.
The calibration is based on the following principles:
Given that the optical density of the film does not
increase linearly it must be related in some way to the
density of the wedge. This approximation can be done
in one of two ways:
The optical approximation wedge (measuring wedge)
in the microdensitometer is so chosen as to match the
density progression on the film. In this approximation
wood densities rather than optical densities are mea-
sured.
Using mathematical methods the grey tones of the cailibration wedge
measuring wedge in the densitometer are related to
those of the cellulose wedge on the film. Using five
density levels of the calibration wedge along a section
of an ellipse, the wood density can be calculated from
the optical density.

- .;""'.:

wedge In the denSitometer

Cellulose acetate wedges and the X-ray photographs

q cc I rlevlatl on
g.'cc
r- III

1.0>-

r
0.8~ 1.99mm
-_ 0.006
j.

0,6- l ,48mm
M ~0.005~
c
o

0.41- : 0.006 ~
t-

Densitogram of a cellulose acetate calibration wedge irradiated at

0.2r -=- 0.005 ~
17 kV for 6 minutes. The calibration relates to wood that is 3 mm I-
thick. Here the deviation from the mean was not more than ± 0.006 -::- 0.005
g/cc. Where an optically non-linear wedge is inserted in the path of Omm
the reference beam of the densitometer or a mathematical transfor-
Ot- ~--------------- : 0.004
mation of the density curve is used, it is possible to obtain linear ..L
values for the wood density. 0 20 40 60mm

68 Echols, 1973; Lenz et aI., 1976; Schweingruber et al., 1978.

Density integrator
The values obtained using the densitometer must be
integrated over a certain distance. This enables the
density of the wood to be determined. Pieces of cellulose
acetate and of woods that are commonly investigated are
carefully shaped and then measured and weighed. The
mean density is thus obtained . A null density is obtained
by integrating the blank background of the film with the
densitometer.
Comparison of the known volumetric-gravimetric den-
sities of the cellulose acetate and of the wood with their
known , volumetric densities enables the differing absorp-
tion rates of the materials to be determined . The value for
the cellulose acetate can then be calculated. This means
that the radiographic and the volumetric-gravimetric den-
sities can be put on the same footing . The calculation of
the correction factor is performed only once for each type
of wood .

q Cc q r

0.6-
0.55

...., Q;
;;;
i:
~O,50 '0
c:
c
.tt ' 0' •
OJ
'0 '0 0'"
(J £: .I::
E. 0
u
0
u
0.
[<:' 'is 'is
C)
o ""inc: '"inc:
/ ~ 0.45 ~
~
:l
'0
:l

• "
u ~
.' "
0'"c:
~
x
'-'
~
x
o <ID '" '"
La" x decidua
Ponus cembra
•
•
•
(;)
0

0
33
0 AO
• @ 0

0,3 0,4 0,5 0,6 gee

0.40 0.45 0,50 0,55 q Cl
volumetr l grdv,metr r dens, y

The relationship between the radiographic-densitometric and volumetric-gravimetric densities.

For Scots pine, larch and Swiss stone pine the radiographic density For spruce the radiographic density is too high by 12.9% and a
is too high by 7.2%, and a correction factor of 0.933 is used to correction factor of 0.886 is used.
approximate the absolute density.

%
q r 25

20
,- ~'\
~ 0,9 \,
0
0;
!~ \
~ Pinus contfrta

I
l
'0 15
u I
-§.
'"
0.6
'~ "- ,f Pinus pon~erosa
~
li 1
10
-. - - ~
\ Pseudotsuga
.,.
. . .1 menziesii
'0
~
0.3 "
'-L'"
5 -...
----....:
0
~ o
~ --- ~
0 5 10 15mm 0.1 0,2 0,3 0,4 0,5 0.6 D) 0,8 0.9
density g. rc

The mean density above a given threshold can be calculated by
dividing all the values measured at a distance of 10 microns by the
number of steps. In this example, the mean density above a
threshold of a glcc is 0.42 g/cc. The mean density above the
threshold 0.6 glcc is 0.82 glcc (lenz et a/., 1976).
The percentage of different density levels in three types of wood (5
trees for each type). The average percentage of different levels can
be obtained by integrating the density above different thresholds
(Echols, 1973).

69
Measuring the density within thEYtree ring
Measuring the minimum and the maximum density
In order to do this a densitometer must be available of a
type where the form and position of the measuring slit can
be adjusted. The maximum and minimum densities can
be measured accurately only if the slit can be positioned
parallel to the tree ring boundary. Failure to orient the slit
correctly leads to inexactitudes in the results.
Where the tree rings are very narrow, the width of
the slit must be reduced to a minimum so that the early
and latewoods do not become integrated. Slit widths of
between 30 and 60 microns seem to be most suitable.
The maximum density is in fact the average of the den-
sities of many latewood cells. The more cells that are
integrated in a densitogram, the more representative is
the value obtained of the whole tree. For this reason the
slit should be as long as possible. A slit length of between
0.8 and 1.6 mm is generally used.
The control mechanism automatically registers both the
positive and the negative maximum values. This is partic-
ularly useful where the tree rings have an irregular density
Positive of an X-ray picture of spruce wood, showing straight and
structure. These maximum values are automatically iden- oblique light measuring slits over the latewood. Measuring using the
tified, in that only those values are recorded which fall angles shown results in incorrect values being obtained for the
short of or exceed a fixed amount of the total amplitude. maximum density. This is illustrated in the diagram below.

aJ core
O' 2 , 5'1 s' I 10° I I annual flng
l
i I i I boundary
bJ mea· . . I
sunng sl,t I
glee cJ densitometric curves
i I i
I I
I lIgh t
measuring slit
i!" 1.0
'u;
c
~ 0.8
<,>
.£
a.
~0.6
o
-0
~ 0.4 ,

2 4 6 8 10 12 14 16 mm

The effect of an oblique slit position on measurements of maximum

density. Deviations of more than 2S result in significantly false
values being obtained.

Photomicrograph of larchwood. The measuring slits are drawn in

over the latewood. The narrow slit integrates about 30 cells, the
wider one about 120.

70
Measurement of early and latewood widths
For densitometric purposes earlywood and latewood are
differentiated in one of two ways:
Delimitation according to a predetermined density
level (fixed threshold). This means that a particular
density threshold is set and that where values exceed
or fall below this threshold the boundary between the
two woods is deemed to occur.
Demarcation according to a density level that is related
to the total density amplitude within a tree ring (float-
ing threshold). Here the threshold varies from ring to
ring. It tends to lie around the mid-point between the
maximum and minimum density.
Both these methods provide an answer to the problem of
differentiating between early and latewood that is accept-
able and technically feasible. The second alternative
outlined here, that of demarcating in line with a floating
threshold, would tend to reflect the biology of the tree
better.
The diagrams and photographs illustrate measurement
along a fixed threshold.

Selective data acquisition for curves with:

glee gn
year
1.0 1.0
1945 44 43 42 41 40 39 38 1937
+ year
197069 68 67 66 65 64 63 1962

,i
2:" 0.8 2:" 0.8
in in
cQ) c
'0 '"
-0
•
'-' ,I
~ 0.6 -§. 0.6 '~
." ."

8-
.;:;
?
-0
\ \

,~ I
~ 0.4 ~ 0.4

0.2 0.2
0~------5--------1-0-------1~5-------2-0------~
25 0 5 10 Mm
1[, 20 25
mm

(a) regular density progression (Abies alba) (b) irregular density progression (Pinus strobus)
• =value higher than the threshold value - = lowest value
S = value lower than the threshold value o = non-registered value (on the ring for 1968)
+ = peak value (Lenz et al., 1976)

Tree-ring sequences illustrating regular and irregular density pro- the subalpine zone of the Alps. Below: irregular progression in
gression. Above: regular progression in larches (Larix decidua) from Scots pine (Pinus silvestris) from the lowlands of northern Germany.
Both x25.

Echols, 1973; Lenz et al., 1976; Schweingruber, 1978. 71

Checking and correcting the data

Correct synchronization is crucial to dendrochronology. tangential traumatic resin ducts

After data has been obtained using the densitometer it is irregular ring boundaries, e.g. in the case of hazel
extremely important that it be properly checked. The growth
values for the width and the density should be checked as reaction or compression wood
a matter of course. biological decomposition

Possible sources of error

There are many different problems which may be un-

avoidable during the preparation of the samples and
which may lead to unreliable results. All these possible
sources of error must be taken into account when check-
ing the data. The main sources of error are:

Technical errors
splits in the samples
overlaps in samples which have been cut obliquely Hazel growth. Irregularly-formed ring boundaries
where peak values have not been registered it ap-
pears as if an annual ring is missing
very narrow rings may not be registered when the
measuring slit is too wide
where the sample has been badly oriented the rings
may well come out blurred on the X-ray picture

Biological faults
missing rings (climatological and pathological)
double rings
non-extractable, strongly-reflecting substances, such Sooty mould deposit with high X-ray absorption. Crack in the
as fu ng i. sample. Insect traces.

Compression wood over a

VH/~jJJIIW\l'
period of 13 years.

,
Fungal hyphae with high X-ray
absorption (dark spots). - - : ' . 1 .

,
Decomposed wood : light I "
patches. (circled).
'.
~......,.. """' -I ~ !.... _ k .. 111 :. _ .... t.

Insect traces in earlywood.

Tree rings wedging out.

(Nogler, 1981).

72
Processing the data

Using a computer the data are processed as follows: manually connected, so producing curves for the ring
width and the maximum density.
The values obtained from the microdensitometer are The curves are checked on a light table . Should the
printed out. This allows certain manipulative and curve display any anomalies, it is checked against the
mechanical errors to be detected and corrected. X-ray film and the strip-chart diagram.
Change of format and the correction of further errors. The mean curves for the raw data are calculated and
The data are au tomati cally put into a form which printed out.
enables them to be easily surveyed and manipulated , Each individual curve is checked against the mean
and a special program arranges the data in the curve. By calculating the Gleichlaufigkeit and the [-
required order. The ring width values and the latewood values for the individual curves and the mean curve
percentages are calculated. Once the data have been any remaining errors can be pin-pointed.
printed out any remaining errors can be corrected.
The data are printed out in graph form using a plot or a The corrected data sets are then ready for further pro-
print-plot programme. The graph points are then cessing.

2 I I 0 34 46 104 24 70 343 3AR

2 III 34 34 102 Ib ~o 320 3Rg
2 112 31 37 104 I~ ~o ?bO 300
2 11 3 32 3R 105 1~ 51 255 3g 1
2 1 14 11 32 103 P 114 ?7~ 302
2 115 30 3~ 107 14 47 ?gR 303
2 II b 33 24 102 \? 3b 33~ 304
2 1I 7 32 3? 105 13 45 ?80 305
2 II R 32 35 102 15 50 300 30b
2 11 g 3b 0 03 116 ;> 0 550 ~O7
2 I?O ?g 2R og II 3g ?82 308
10 225 ?h 0 q ? II 182 532 0
19 26 q 71 I f2 5321
226 ? II
19 2n ?7 10 79 3 13 231 5322
lO 2,:>11 27 (> 3., (> 0 0 5323-
19 2?0 30 13 RS h 10 31b 532"*
19 230 36 4 70 3 7 429 "326
10 231 25 1'5 91 'i 20 250 53?7

Microfiche reader. Checking the data. Errors that have been located are noted. The information is
then fed back into the computer for correction.

........", c..... ,,, L .. , •

1 1 .... I (
1'1.... t
11 .. '

, ..
I ' ....
~
I I

1 ~'" I

. . . . . • . • . , ___ • _____ l • • • • • • • • • : . . ___ . __ . . . . . . . . . . . ; • . : . • ___ . , . __ : __ . ___ : _____ . _ : .. _________ J • • • • • : • • : . . . . . . . . . . . : • • • ; . : • • • , • • • • • • • • • ,

"I"! !

I ~ ,,. .. ~
'l'''".
t· II'
"'I"
, , .... ~:.
,
'I. ""'r" .. " "

."
Print-plot of a number of annual ring features obtained radiographically-densitometrically. The samples can be visually matched and
synchronized using such plots.

73
Visual sifting and checking of the raw data

When the raw data have been corrected, the curves tion of material can significantly affect the quality of the
drawn and the first Gleichlaufigkeit calculations made, the final results. What shape the scientific analysis will take is
material must be evaluated. This essentially visual assess- determined during this sifting stage and depends on the
ment is extremely important since the subsequent selec- characteristics of the curves produced.

Visual agreement between the curves

Good agreement both in the pattern and the Poor agreement In pattern and level over
level over a number of years. u a number of years. u

--
u
Cl
u
0,
~
Vi
c
OJ
"0

0.8

0.8

Agleement o f an II)(Jividual curve with the mean curve, expressed as Gleichlaufigkcll.

good aglc mcnt in 1I10st cases poor agreement In many cases

sumple sdrnple
C(Jre 2 3 ~ tl (} 7 8 9 10 core 2 3 ,1 :. 6 7 8 9 10

•
Glelchlauf IgkCl t
• • • 0
• •
I I
0 0 • 0 0 0

o 60 64.9 • 75 79.9
6!J 69.9 . 80 84.9
• 70 74.9 • 85:n and above

sensitive curve complacent curve

high variation /\ . low variation

V\/0/V/'·Vv1vv-
74
 Charac ens IC curves

even t rend, due f or example to ageing uneven trend, due for example 10 forestry measures,
damage or Ihe effecl of ell male

~-
__ i '. _ ......, .
/
\/
", < "" \<>~ \ .-;~~ .
\" / \
. . . - ,; V
...... _~ 6 I \
. "
1\ /\ J\ /\ J' - •\
\-7\( ' \.J V\ _T '=

abrupl change, due for example 10 fire and the

subsequenl Increase In Ilghl or 10 damage caused periodi ci ti es caused by regular attack by insects
by envlronm nlal factors

/\/\
.
/
.-.
/~./\.

here Ihere are a considerable number of large denslly suslaliled decrease III Wldlh or density, due for exam ple
lutluatlons; some of Ihese produce 'false' annual rings 10 alr·pollution or pollution of Ihe root·zone area

Conclusions

The ad hoc evaluation of the quality of the site which

was made at the time of sampling can now be looked at
again in conjunction with the results of the assessment of
the tree-ring characteristics for that site. The sampling
material, field notes, X-ray films, strip-chart diagrams and
the digital raw-value curves for the individual parameters
are all taken into account in considering the following;
Whether the cores are of sufficient quality to justify
further work being done with them. Where cores are
to be excluded from further analysis, the reason or
reasons for this, whether they are technical, biological,
statistical or ecological should be given. Consideration
must also be given to the question of whether the
whole site is to be excluded or fresh samples are to
be obtained.
Whether the site can be integrated into a fairly wide-
spread network. Do the species, the number of rings,
etc. meet the necessary requirements?
Whether there are any special features which would
indicate further areas for study.
Which data should receive particular attention during
further work.

75
Dating in practice
Baillie (1982) has given a full description of dating in
the European archaeological dendrochronology. I shall
attempt here to make only one or two comments on
dating in practice. Reliable dating requires the mastery of
the simple methods involved and considerable experi-
ence and patience. The following should be borne in
mind:
Methods
Every statistically produced date must be confirmed at the
light table. The calculation of Gleichlaufigkeit and of the
correlation values facilitates synchronization, especially in
the construction of chronologies covering thousands of
years, and here the computer is an invaluable aid. None-
theless, visual comparison remains the essence of den-
drochronology.
Basis of comparsion: 41001/496018 T ree-ri ng width
Gleichliiufigkeit values for
Position Length of overlap
(year no.) (in years) 95% 99% 99,9% certainty
515/173 175 58.4
515/125 124 65.9
515/ 72 71 84.5
Problems
Short ring sequences are generally less suitable for dating
than long ones. The shorter the sequence, the greater the
likelihood there is that the dating will be inaccurate. For
periods which give a very smooth curve and have very
few pointer years, it may not even be possible to date
series of 120 years or more, while 30 years may suffice
for a dating within a period with an irregular, characteristic
pattern. Missing rings falsify the statistical calculations.
Their location is possible only through visual comparison
with complete chronologies. They are common in conifer
wood, where the rings are very narrow. Ring-porous
species such as oak do not display this phenomenon.
Smooth curves are less easy to date than sensitive ones.
Three synchronized ring-width curves for fossilized larch from the
subalpine zone. The upper two curves are smooth but nonetheless
exhibit characteristic periods and pOinter years.
76
The density values for decomposed wood are not
reliable. This means that only the ring width can be used
for dating. In certain species, especially the five-needle
pines (haploxylon) it is practically impossible to date using
the maximum density measurements from just one site.
Dating
The shorter the time-span to which a sequence belongs,
the more straightforward and certain is its dating. It is, for
instance, easier to date Central European samples from
high altitudes - where trees first appeared 9000 years
ago - than samples from low-lying areas, where trees
have existed for 13 000 years. There are a number of
factors which narrow down the range of possible periods
through relative dating:
Typological features of buildings and artefacts
Geological features, such as the position of moraines,
their soil and vegetation
Radiocarbon dating can provide very useful informa-
tion. It is important, however, not to neglect possible
inaccuracies in the measurements, the distortion in the
time-scale and the position that the sample originally
had in the tree-trunk
The degree of preservation of the material cannot
usually be taken as an indication of its age.
How exactly the synchronization work should be carried
out depends on a number of factors: the importance of
the sample, the length of the series and the chronological
information available. As a first step, poorly-covered
chronologies with no missing rings are cross-dated visu-
ally and mathematically. The resulting mean curves form
the basis for further synchronizations. Sections from long
sequences can often be aligned mathematically to other
curves, and the subsequent synchronization of different
mean curves is often possible. As a final step, it often
proves possible to aggregate single sequences sharing a
common characteristic, e.g. sensitivity, into a single
chronology.
The lower curve, which is sensitive, could only be placed because all
the pOinter years in the older part are available. The alignment here
is not sufficiently good to enable the curves to be dated.
Successful dating

It is difficult to lay down general conditions for success in Here the Gleichlaufigkeit is high for both tree-ring width
this area. Observations made so far lead us to believe that and maximum density, even though the curves are
the chances of dating a sample accurately are good wrongly aligned (Renner, unpublished) .
where :
Length of o verl ap Gleichl ii ufigkeit values in %
the curves exhibit high sensitivity; samples with in y ears ring w idth max imum density
smooth curves are as a rule difficult to date ; 223 61.2 62.3
information is available on at least two tree-ring para- 255 57 .5 58.2
meters, e.g. ring width and maximum density. Where 273 55.5 58 .3
both these parameters exhibit high Gleichla.ufigkeit 339 59 .9 61 .5
and correlation values for the same period the match- 344 55.1 55 .8
ing will be confirmed by visually comparing the curves; 344 56.4 57.4
several samples covering the same period are avail- 344 54 .8 54 .5
able for analysis; 344 54 .7 55.7
relatively long curves are being analysed ;
the samples cover comparatively short time-spans. (after Renner, unpublished.)
It is always difficult to say in advance whether any given
samples will be datable. Dendrochronological laboratories Gleichla.ufigkeit values for two wrongly aligned curves.
in Europe and America are rarely able to date more than Here one parameter displays high values (Renner, un-
two-thirds of the samples they receive . published) .

Leng t h of ove rlap G leichl iiu figkeit values in%

in y ears ring w idth max imum density Length of overlap Gleichliiufig keit values in %
47 79.9 62.8 in years ring w idth max imum density
139 57.9 62.9 41 76.3
153 70 .6 65.4 59 67 .2
158 74.7 78 .2 70 70.0
210 84 .9 77.4 71 70.5
223 67.7 63.7 94 63.3
27 4 74.1 77 .2 71 72 .0
304 75 .8 65.4 79 70.9
360 55.7 61.9 91 68 .2
Gleichla.ufigkeit values obtained from visually aligned,
correctly dated curves from a uniform site at the Alpine (after Renner, unpublished .)
timberline (Renner, unpublished).

The accuracy and reliability of dates obtained dendrochronologically

. .. . . .., "',
Ideally all dating done dendrochronologically would be
.-./Iv .
""'V .
'..
. /\' 1\1\ .J
,/\ '. ,,_.,.\..'
/" .
\ ,.
correct. Since, however, it is practically impossible to find
a statistical criterion which enables a hypothesis to be .~( \ \ ,/"
. \.. . 1\/\ ,' j ...........
\1 '\ / "...'
' -:-,\.,.
. j".... V\ . \p\
i i\'j"V\'"" \\;;,\.t,\
verified with a hundred percent certainty, and since, on " .1 I'
the other hand, visual matching cannot claim to be an
•
\
. •• -,\
.1 /'_ .
' .2
.
•.•. • •
. ' \

. i "'" \", .
.~

. . "i"""\ 1\ j\
infallible method , the possibility of incorrect dating can ./\'\./" ...\ /• .•.•'.I"' .'-,.~\
' \ I;,
• ',
j\
; . ..... \ .......' ., . . \ \'. l\i' ./ \ ...:\
f" ~"\ I " • ;".,_-1 \\/1/ .' V \/v,¥ "'V\i . \,,~-..:/\
never be ruled out in difficult cases . Where , however,
there is a well- covered mean curve with clear pointer
years available for comparison, the likelihood of the wrong . \ ..... .\ ..../'.";' . .W •
date being given is very small provided that the curves
display high Gleichla.ufigkeit and correlation values for the Two synchronously placed fossil larch curves.
Top: tree-ring width.
same period , and have the same pointer years . Bottom: maximum denSity.
It is worthwhile bearing in mind that dendrochronology These can be dated with absolute certainty.

j. """'"f _, /.
is an empirical rather than a mathematical discipline, even ! ...
though it relies quite heavily on a number of mathematical j"' \ . ....... \
operations. This means that, however carefully measure- ".~
ments are made and calculations carried out, mistakes in ,,).......
dating can never be completely ruled out. ''''''.V ·-· .' • t'·'· " .\
".,.... ... ... ...... ..........."" \-.'.....~. .
i
. ",.," ,/'"
'\\ h,·,1 .....~, ,". . '-'\ / . ~... '
vii V . \ /\ .. ' . ," ,
Two tree ring width curves for fossil spruce, not synchronously
placed. Here the Gleichliiufigkeit is 70%.
,
\ .. j'" . '" •....J
\ ",
'I."

77
Statistics and electronic data processing

Prerequisites for application

Dendrochronological research without the aid of statistics
and electronic data processing is today unthinkable. When
tree ring analysis first began, relatively few ring se-
quences were available but now there are thousands in
every laboratory. This accumulation of data necessitated
the development of more appropriate and more powerful
methods of comparison. The accumulation and interpreta-
tion of large amounts of data has also been made practical
by electronic data processing. The two main tools may be
described as follows:
Statistics: a method for quantitative analysis and inter-
pretation or checking of data for the testing of assump-
tions, hypotheses or models. Statistical methods do not
produce new qualitative information, but serve to extract
from given data the answers to particular scientific ques-
tion (Sachs, 1978).
Electronic data processing (EDP): a method for the
rapid execution of mathematical operations on large quan-
tities of data and for the graphical representation of the
results.
Statistical methods can be successfully applied only if:
the statistical process used is suitable for the given
data set and properly applied.
a clearly-formulated hypothesis, based on biological
considerations, is proposed.
the data are good. The best statistical processes will
fail if the basic material is heterogeneous, i.e. if mean
curves are constructed from single curves from dif-
ferent sites. Statistics does not obviate the careful
selection of samples in the field.
a large quantity of data can be brought together and
analysed according to a set of clear principles and
recognised tendencies be quantified.
the results are critically appraised and interpreted from
a biological viewpoint. If the results are statistically
acceptable but not explicable in terms of biology or
climatology, they remain doubtful until confirmed for
other data sets.
Details of the methods may be found in standard text-
books e.g. Sachs 1978 and Taubenheim 1969. Fritts 1976
describes the application of these methods to dendro-
chronology.
It is clear that samples taken by different people with
differing research interests will be much more variable
than those taken by only one person. It is also widely
accepted that the quality of the samples - and hence of
the results obtained - can be greatly affected by such
factors as researcher fatigue, or bad weather.
These factors must be given due weight when a
decision is being made as to whether to commit funds to
carrying out costly calculations. It is also of considerable
importance that the dendrochronologist today be on his
guard against losing his feeling for the sampling site after
having taken his samples.

Methods

All dendrochronological calculations are based on se-

quential yearly measurements from tree-ring and meteor-
ological series.

Identification, isolation and elimination of long-term ~ 100

c
<D
factors "0 90
E
::J
These may reflect aging, climatic fluctuations, decom- E 80
x
position, etc. The use of smoothing functions shows both ro
E 70
the variations from year to year and the long-term trends. A
Statistical procedures are used in order to standardise
.s
"0
100
.~

short-term variations (curve fitting). "0

o
o
>;
~ 10

Maximum density curve and a latewood width curve with age trend
(A) is brought to a uniform level (B) The annual deviations from the 1730 1790 1850
smoothing function vary around the horizontal line.

78 Fritts and Swetnam, 1986.

The long-term features are brought into relief by sup-
pressing the short-term ones (low-pass filter) . 4
A
3
<!
Calculation of the characteristic parameters of a data
~ 2
series '0
c
This is done so that groups can be established which will
provide a basis for the characterization and comparison of
these series. Comparable data can thus be aggregated 1800
using calculated sum totals and means.
The parameters that are commonly used to define the
characteristics and the similarities of the tree-ring series
are expressed as mean values , correlation coefficients , B
Gleichlaufigkeit values, sensitivity and variance . Iow· pass
ro
~ I •
'0
C
- 0 - - -1-80
- 0- - - - - - - - - - . . , .19""'0:-:0- - - . . , . - - - -
The long-term variations in the curves (8) which lie within the curves
for the annual values (A) are shown by means of smoothing func-
tions (Fritts, 1976).

mean of a data sct mean of several data sets

o o ~
~
>

mean of a part of a data set

~ 21
~ I
I
I I
1
I
1
I
1
I
I,
I
0 i ~
'0 I I I I I I -

31I II II I I I+ I 6 ~
o t t t t t t t
I I I ~

OOOOOOCY
annual mean

Comparisons are made by calculating the degree of similarity:

between several sets
Within one set

synchronously placed

2
Wllh Imc·dlsplacement

of parts of the set 4 1~~--+-~-4--+--+~

79
The tree-ring series and the meteorological series are
compared as a basis for the construction of statistical
models linking tree rings and climate
The first step in the process of climatic reconstruction is
the checking of the time-series. The meteorological data
must be checked for missing values and to ensure that
they are homogeneous (right-hand column). Tree-ring
series that relate to the recent past and those covering
historical or archaeological wood should be compared and
contrasted (left-hand column). Chronologies from living
trees may only be extended using material from historical
ones when the series are at least statistically similar.
Where both the meteorological and the tree-ring series
satisfy the requirements outlined above, statistical models
may be constructed using response and transfer func-
tions.

estimate
missing statistics
chronology
and
chronology
and
statistics
L-v_al_ues---'l
statistics

years

combined
chronology

Flow-chart for the data-checking procedures

involved in the calculation of dependency
models for climatic reconstruction (Rose et al.,
1981).

Checking the reconstructions

Calculating the model Reconstruction Verification

Calibration phase Reconstruction phase Independent phase

The tree-ring series is linked to Only the dependent series The new dependent
the meteorological one. A new can provide a reliable series is compared
dependent series is constructed. measure of past weather with the unchanged
conditions. (see page 90.) meteorological series.

80
Basic statistical methods

Characterising tree-ring chronologies

Common terms = Gleichlaufigkeit for two whole series, x and y

= standard deviation of single values from the
Xi' y, = observed values of the series X or y at moment i mean of x or y
~i = difference between two successive observed x,y = mean of series x or y
values over the interval i (X i+ 1 - Xi) = sensitivity in interval i
Si
t ix , tiy = interval trend values in the interval i for series X S = mean sensitivity
ory = correlation coefficient
t, = mean interval trend value for the interval i for (linear dependence)
several series or mean series = explained variance
xp, yp = smoothed value for the series X or y at moment i = value of index at moment i
n = number of observations
Gix , giy = Gleichlaufigkeit at moment i for x or y Further information in Fritts 1976.

Arithmetic mean, standard deviation

n

Mean = x= I x;ln
i- 1

Standard = s = +
deviation x - n - 1
."./
The mean, and standard deviation of any given data set
are characteristic for that set. The mean is the average
value of the data. The standard deviation is a measure of
the mean variation of the individual data about the mean.
Its numerical expression can therefore be positive or
negative, and it can be used for graphical presentations of
the distribution (of the values about the mean). If the data
have a normal distribution, as is the case for most tree-
ring features, two thirds of the values lie within one D)
0,5 0,6 0,8 0,9 1,0 1.1
standard deviation on either side of the mean. The mean glee
and the standard deviation can be calculated from raw,
transformed or derived data. Means of single series may
reflect ecological conditions. In some cases a mean may Normal distribution. Approximately 68% of all values lie within one
characterise a site. standard deviation and 95% within two standard deviations.

81
 . /, /-. . ,. .",.-.~ / .,-_, /.-.-.-', /\ j ' -'-'v"\
, i~ '- /'. . '-
/\/. -.-' '.-' \\//-,_.;,
•• ".'

,._.. /'V;-'-./ -.-.-\/ ..\. ./.-/',

'. /....... ... . ..... - . .; -\\ • I .. • .'.

. -.
\./ . .
. ,/\ . j" .
A mean curve exhibits general features: individual dif-
./\.,.
'/~
. /',,',
\ /'-',' .;',.,.\
'.-. /. '/' ,.
'" .-. . \
.,,:~: \ 1,\ ,.J! .<-.
". ".

. . /. '. /. "'1\'
.< "./'/' /\'
/'
/'/ \ /', '. ._.\ / /\
ferences in single trees can no longer be seen. The
similarity between the mean curves for two different sites
- such curves are usually based on two cores from each . '. ,.... ~.
.'I'".
of twelve trees - is therefore often greater than that
between any two single curves.

Comparison between two mean curves exhibiting good agree-

ment (top) and two single curves exhibiting fairly good agreement
(bottom).

Standard deviation is a measure of the homogeneity of

the mean curve. For samples from ecologically uniform
sites the variation is small, while for heterogeneous sites
or badly-selected samples it is large.
Running means are discussed in the section entitled
'Filters and Smoothing Functions'.
' ':
'li)
:::l
C
C
'"

2 3 4 5 6 7 8 910 11 12

The mean and standard deviation for one series. 68% of all the
values are within the hatched area. In certain years, e.g. 2, there is
very little variation, while in others, e.g. 6, there is a lot.

Sensitivity sensitive, e.g. maximum density

(X i + 1 - x,) ·2 mean S=
Annual Si+1 = -----'----- -'-i-_ 1'-----_ __

(Xi + 1 + Xi) n-1

The mean sensitivity is a measure of change and ex-

presses the difference between two successive values in
a series by means of percentages. It is the average of the
absolute values of the individual sensitivities in a series
and can be calculated either from the raw or from the complacent, e.g. minimum density
indexed values. .. . . . . . _. ._""",-./.,.__ .___ ......... / ... . . . _._., ./0_.---·_·
By determining the sensitivities it is possible to as- .................. ""......... .---.......... ........'
certain to what extent the growth of individual species on
particular sites is influenced by environmental factors, and
in which periods growth conditions were either very stable
or very variable. In the former case the curves are smooth
or 'complacent' and in the latter uneven or 'sensitive'.
Sensitivity and Gleichlaufigkeit are always calculated
relevant to particular intervals. In contrast to Gleichlau-
figkeit, which reflects year-to-year agreement between
series, sensitivity reflects year-to-year amplitude within a sensitive complacent
series.

82
Gleichlaufigkeit (sign test)

~i = (x, +1 - x,) when ~,> 0: Gix = +~

~, = 0: Gix = 0
~i < 0: Gix =-~

1 n-1 interva! 1 2 3 4 5 6 7 8 9 10
for two curves G(x. y) = - - - E i G,x + Giy

\V1\ I\
V
!
n- 1

i
i-1

Gleichlaufigkeit is a measure of the similarity between two

curves. The intervals between successive points in time
'">
G ..
l\ /
/~ / '.
are examined for an upward or downward trend. The total Q)
/
Gleichlaufigkeit over all the intervals is a measure of the
l.-
:J
U
/" ~,
agreement between the interval trends of two curves, and
is usually expressed as a percentage. Where the intervals
for annual ring curves run parallel for many years it can be
G,y
- -
/ "
assumed that the factors influencing growth were similar Gleic hlaufigkeit 1 1 112 ,11211 1 1 1 - 7 out 0 flO
in both cases. values G - 70%
By comparing all the curves from one site for a given
time-span it is possible to assess the homogeneity of the
site by looking at the Gleichlaufigkeit. This may then be
expressed in symbols. It is often possible to synchronize An example of a Gleichlaufigkeit calculation
samples of unknown age with dated samples by calculat-
ing the progressive Gleichlaufigkeit between the two
samples.

Interval trend

~, = (X i + 1 - x;); when ~i > 0: t,x = 1

~i = 0: tix = ~ indivi dual curves
~, < 0: tix = 0

for m curves: ti = . I
1-1
t'i/m; i = 1 ... m

The mean interval trend shows how many of the intervals

of the same data display the same tendency in several
curves. It is a measure of the homogeneity of the individ-
ual intervals in a mean curve. The values are expressed
as percentages of ascending intervals. 100% can be
average 46 %
produced only by a rise, 0% indicates that all the intervals
are falling. The number of integral samples must be taken _ '" l00%]Gleich- ® : pointer year
into account in the interpretation of the values. o ~ ~ laufig kei t
Where particular years within an ecological unit display Q) ~
C1>Q)

a very high interval percentage it can be assumed that ~ . S ~% \.

one particular factor influenced tree growth in the year in ~ g'
question. Such cases are known as pointer years. For ~ :~ 0%
further definitions see Eckstein and Bauch 1969. Interval
trend is a measure similar to standard deviation/scatter
about the mean curve, except that the latter is based on
annual values rather than on intervals.

Calculation of the interval percentages from five individual curves.

The number of rising intervals is expressed as a percentage (lower
curve). Intervals in which, in ten or more samples, more than 90% of
all the individual curves move in the same direction, are here termed
pointer years.

83
Correlation, correlation coefficient

r~ + L(Xi - x)(y, - y)
skew distribution

- JL(X, - X)2L (Yi - y)2 .

The correlation coefficient is a measure of the linear

relationship between pairs of values from two series. The
correlation coefficient is expressed as a decimal fraction . o 10 20 30 40 50
(0 means no relation, +1 means identical, -1 means I inear scale
negatively identical) .
As raw data curves normally contain long-term fluctua-
tions, e.g. age trend, as well as annual variations , these
normal distribution
must be eliminated before the calculation of the correla-
tion coefficient Baillie (1973) uses the following procedure :
smoothing the curve using 5-year running means
production of comparable series by indexing
transformation of skew distribution. Since tree-ring 10 20 30 40 50 100
series usually contain more narrow than wide rings , logarithmic scale
the distribution is skew and must be transformed to a
normal one. This is done by taking the logarithm of all
values (xt, yt) · Transformation of a skew distribution with a linear scale to a normal
calculation of the correlation coefficients (between xt , one with a logarithmic scale
yt)
determination of the (-values. These are positive
values . The (-test indicates whether two curves are
t-value : t = Ir ly ~1-2
rn=2'
related, and being stringent, generally gives the syn- - r
chronous position of two curves indisputably when the
level of statistical significance is sufficiently high (see
Baillie , 1982, p. 80ff.).

"
cross-correlation
The correlation coefficients for two or more curves of
the same date indicate the similarity between the curves
-/ -, ............._ ".-........ /- /'-
_,;e.....
/', ......- -.......- ~ - .........- - --"
.......

-'
(cross-correlation) . The correlation coefficients of a num- ~
ber of tree-ring series from one ecological unit (site) can
be put into a matrix, from which it can be seen whether
all the ring series fit into the set. The mean of all the
coefficients is a measure of homogeneity and is a charac- higher - -
......

-
.......
lower -
....... ~
"\._......-,
teristic of the site. Serial correlation is a measure of correlation correlation ,_
similarity within a sequence with a time displacement of Ir~-'r-~'-~~~IIr-.--.-.--,~~r-ol

one or more years. It provides information on the relation- 1950 1957 1950 1957
ships within the series. Where the preceding year has in-
fluenced the year under investigation significantly the
serial correlation coefficient is high . correlation matrix
In contrast to Gleichlaufigkeit and sensitivity, correla- curves 1 2 3 4 1:
tion is based on the actual values rather than changes
1 - 0.68 045 0.82 063
from one value or interval to the next. en
2: 2
Q)
0.63 - 0.38 OAl 047
:J
u 3 OA5 0.38 0.26 0.36
4 0.82 OAl 026 OA7
1: 063 OA7 036 047 OA91:
1:

serial correlation = autocorrelation

•
50 • • ./ ~o
r,;-./ \'--..-./ \. ./ \ '@
" "" , ", ./"", " ./'\
\
\ . '

~. / "
~ \~ \ '\ \

\ ", 51 . '- ' , \ .• \

', •/" ,61

. "•/
, / ./
® . . . . .'.-.......... \ ./
\
~ ®
-...........:.
I I I
1950 1955 1960

84
Standardization and indexing

Xi f . Ii = Xi - X? for densities
It = -0 or widths;
Xi
Standardization, commonly termed indexing in dendro-
chronology, enables tree ring series relating to different
samples to be compared by eliminating long-term vari-
ables which have been caused by bio-ecological factors
such as tree age and microsite differences.
The smoothed value (x?) is subtracted from the actual
value, or the actual value is divided by the smoothed
value, depending on the tree ring variable being pro-
cessed decreasing ampl itude and sensit ivity

Division:
The yearly variation in widths (earlywood, latewood, measured values
ring width) is related to the widths themselves, and de-
creases with decreasing width. The values of the
indices vary around 1. (They are often multiplied by
1000 for convenience).

Subtraction:
The maximum and minimum densities are not related
to the fluctuation in the density values. This variation constant ampl i tude and sensi tiv i IV
tends to remain constant over the total life of the tree.

Mt--f~-I-----"Y--l£..:::~-=----I--A------\r---I'r--------H-+----+ annual ring width, raw values

r-*H~H--">I--..L.=-\=---:-::~---\---I\,-----L---¥----+----+- annual ring width,

indexed values

10

-#--\-H--A-j....:..-\-Jl-- ---J--=---\f-- - --+max i mu m densi ty. raw val ues

10 .0

maximum density indexed

values

glee
)0.0
:1 / .0
JUJU 1=-+---m~----c~r___r....--A"1""C*"7""I~~.,.......,:d+~tJ!'u__'L:>I.v=:::.L..:..r~:T"..q__t.......d_- - _t mi n imu m densit y , raw va l ues

1--+,j----1~l__\__-I_\_-HrP-ttlnr_N~----Il-_H_\_r4rT_dlI-_brf4____jPrftt_--t mini mum densi ty indexed

values
3' U

o 50 100 y ears

Raw data curves (thin lines) and standardized curves (thick lines) of width: measured (raw) value)
three ring features. The standardization or indexing was done as smoothed value
follows:
density: measured (raw) value - smoothed value

85
Filters and smoothing functions

Interest focuses on short-, medium- or long-term fluctua- Running means

tions, depending on the research objectives. The forms
and periods of these biological-ecological fluctuations can
be determined and presented by means of various statis-
tical methods.
Terminology.
A given number of tree ring values within a t'ime window'
is averaged, the mean being set in the center of the
window . A window of selected length is placed over the
series year by year. Short-term fluctuations are thus
eliminated from the series of means so calculated.

Trend: The characteristic course of a curve. The trend can

be described using smoothing functions and low-
pass filters. The values falling on the trend line annual ring width
may be termed 'signal'; the short-term variations a
around the trend line are 'noise', if trend is the b
c
focus of interest. d

Smoothing
function: the function is pre-selected and fitted to the
data. Typical forms are polynomial, exponen-
max imum density
tial and straight line functions.
a
Filter: the results, i.e. the form of the curve are deter- b
mined by the length and weighting of the filter. c
d
Low-pass filters bring long-term fluctuations to the
fore, while high-pass filters emphasise the short-
term annual values. Weighted running means are min i mum density
typical low pass filters.
a
The following illustrations are based on a single curve b
from a spruce growing in the subalpine zone of the Alps, c
d
(Riederalp, Valais). The series is 90 years long. Three
tree-ring features are presented: ring width, maximum
and minimum density. For clarity the curves have been
Slightly displaced vertically.
Comparison of running means:
(All figs. after Braker, unpublished.). (a) Unweighted, 5-year. Each value has the same weighting
(b) Binomially weighted, 5-year. The central value is multiplied by a
factor of 0.6, the adjacent ones by 0.4 and the outer ones by 0.1
(c) Digital filter, 13-year, weighting
(d) Digital filter, 31-year, weighting

Filters a to c are equally good when significant, medium-term

fluctuations are under investigation. The 31-year filter, on the other
hand, only shows up long-term fluctuations.

0.6

0> 0.4 0 .6
c
~
.c 0>
0>
0.2 C
04
.~ .~

.c
0>
' Qj
a ~ 0.2 / 13 year
/31 year
0 .2 , , r 1 0
6 ~ 2 a 2 4 6 10 8 6 4 2 a 2 4 6 8 10
y ears y ears

Weighting of high pass filters. Weighting of low pass filters.

Only the central value is amplified. The outer ones are suppressed. Within a given period all values are amplified; the central ones
strongly, the outer ones less so (LaMarche and Fritts, 1972).

86 Fritts and Swetnam, 1986; LaMarche and Fritts, 1972.

Band-pass filter
Here the values are weighted to emphasize a narrow fre-
quency band for which a cyclical pattern is suspected.
The central values are strongly weighted, those adjacent
to them are suppressed and the outer values are slightly
amplified.
This type of filter should enable cyclic patterns - e.g. maximum density
sunspot cycles - which may be present but not neces-
sarily recognisable as such, to be seen.

Curves above: 5 year binomial filter

Curves below: 23 year band-pass filter

a>
.S +0.1
.§. 0
~ - 0 .1 , • 1 r f I 1 .,

10 8 6 4 202 4 6 8 10
years

Weighting using band-pass filters.

Only those values which fall within a given number of years are
Polynomial amplified, the others are suppressed.

the order (highest) of the function is pre-selected, i.e. it is

determined in advance how closely the curves of the ring
values should fit together. Low order polynomials with few
or no inflections reflect the large waves. High order ones
with several inflections reflect the short-wave fluctuations
also. This is particularly suitable for the elimination of
disturbances e.g. silvicultural measures, but care must be
taken with dendroclimatological analyses, since climatic
influences may also be suppressed when such functions
are used.

Straight line
this can be regarded as a special type of first order poly-
nomial, i.e. without inflections. This function is often
appropriate for the smoothing of density values.
(a) Linear regression: straight line with any slope.
(b) Mean: horizontal line.
(c) Quadratic function: second order polynomial with two pOints of
intersection.
(d) Cubic function: third order polynomial with three pOints of
intersection.
(e) High order polynomial: with several inflections or
spline: several cubic functions spliced together at knot pOints.

Negative exponential annual ring width

a
here the type of function is pre-selected: the fitting to the
b ~ _ _ _ _ _ _ __ _
data determines the curvature and height of the curve.
This method is mainly appropriate for the smoothing of
age trends in width values. It does not however allow for a~ty
increases in width in young phases. This is taken into
account.
b ~---------------
Hugershoff's function minimum de::.:n::si~ty_ _ _ _ _ _ _ _ _ __ _ _ __

this is a combination of a polynomial and an exponential b~

function. It brings the climbing trend of the first years of
growth and the falling trend of later years to the fore, so (a) Negative exponential function
that the width values are optimally smoothed. (b) Hugershoff's function

87
Modelling (response functions)
Response functions are used to identify the relationship
between climate and tree rings. A number of different
statistical procedures, e.g. sign tests, correlations, multi-
ple regression with or without the identification of the
principal components of the population , can be used to
identify and plot the relationships between climate and
a) with itself, but in serial correlation (time displacement!.
This shows how strongly the previous year. which deter·
mines the physiological condition of the tree, influences
growth in the current year.
b) with the monthly mean temperatu res
1900 annual ring series 1950
1899 temperature series 1949
June
July
December
1900 1950
January
February
August
September
(a) Schematic presentation of tree ring/climate comparisons
Such calculations allow response function diagrams to be
constructed as a series of monthly coefficients of weights
which are plotted sequentially in relation to the central
axis. Values lying significantly above this axis indicate a
positive response, those lying significantly below the axes
indicate a negative relatioQship between tree-ring size
and climate for particular months.
Tree-ring series can also be related to other climate-
dependent series, for example, the mass balance of
glaciers, changes in glacier tongue length, wind yield, etc.
Information obtained using these sort of calculations
will only provide a basis for valid comparisons when the
following conditions are met:
there must be sufficient correctly synchronized sam-
ples relating to homogeneous sites. Here it should be
borne in mind that the concepts 'homogeneous ' and
'site' are often used loosely or incorrectly. For exam-
ple, the term site is sometimes misleadingly used to
refer to an area which falls into distinct topographical
units but which is heterogeneous with respect to
vegetation .
verified metereological data for weather stations which
are fairly close together must be used.
the limitations of response functions must be taken
into account. Where they relate to sites which are
subject to extreme weather conditions, e.g. very low
88
tree rings.
Fritts (1976) and Fritts and Wu (1986) described these
methods, which are by no means always straightforward .
A response function diagram may contain a number of
different comparisons. The diagram below compares a
ring series covering 50 years with a number of factors.
1900 1950
1899 ~1. . . . . . . . . . . . . . . . . . . .~r~~~1~1---,
1898 1- . . . . . . . . . . . . . . . . . . . .~1~ 94~~
~.~---4
1897 "'I- ............ ---~1'!!' 94'!'!7~1.-------1
I f .
c) with the total monthly precipitation
1900 annual ring series 1950
1899 precipitation series 1949
1 1900 f 19 50
I I
precipitation at one time and very low temperatures at
another, such functions only indicate which factors are
generally dominant at particular times of the year .
They do not show which factors limit growth in partic-
ular years. It is clear, then, that response functions are
only a useful analytical tool in the cases where there is
in effect only one limiting factor, i.e. for sites at the two
timberline limits, the lower dry limit and the upper
cool-moist one.
the results must be consistent with those obtained for
other sites. No conclusions of general validity can be
drawn from a response function.
only factors which are possible should be compared.
There is no point in comparing predicted monthly
mean values with tree-ring sequences.
the results must be ecologically and physiologically
feasible .
the climatic variables used should not be directly
related to each other. Where this is unavoidable this
must be taken into account when the diagrams are
evaluated .
it must be possible to test the assumptions made in
constructing the model. This means that the curves
relating to the supposed limiting factors must be
shown to correlate with the tree-ring curves (raw or
transformed) in the verification procedure.
Fritts and Wu, 1986.
Two examples of the interpretation of response functions
A positive value indicates a favourable influence on
growth, a negative value an inhibiting one. The diagrams
show the response functions (solid lines) and 3leichlau-
figkeit vatues (broken lines). The vertical bars show the
95% confidence limits. Gleichlaufigkeit values are sig-
nificant with 95% confidence if they are higher than 63%
or lower than 37% for this example.

prev ious
year
temperature preclpi ta tlor influence
AI tschwald / Grosser Sl Bernhard
0 .2 Picea abies Norway spruce 90
80
0.1 70 2
*.S 5060
on
Q)
40 - 1
- 0 ,1 :I 30
ro
- 0.2 -> 20
'iii 10
.:.L
Cl
:;:
,'"E
:I
4
u
.iii
(3 70
60
50
40
30

June Oct. Jan. June June Oct. Jan. June 1 2 3

month month years

Response function - - - - - Gleichlaufigkeit Glelchlaufigkeit with hours of sunshine

(a) Maximum densities of spruces from timberline Sites in the Alps.
The annual precipitation is around 120 cm.
The densities are compared with the temperature and precipita-
tion series for 7 months of the preceding year and 9 months
of the current year, and also with themselves, displaced by
between 1 and 3 years.
(b) Annual ring widths of pines from the lowlands of northern
Germany. The annual precipitation is around 90 cm.
The widths are compared with temperature, hours of sunshine
and precipitation as in (a).

The interpretation of example (a) Interpretation of example (b)

Given the predominantly cool-moist climatic conditions of
the timberline it was reasonable to suppose that growth in
this area was limited by temperature. The response func-
tions indicate that a high maximum density occurs when
temperatures in April/May and July to September are
high, and precipitation in September/October of the pre-
ceding year and in the current summer, especially in
August, is low. The maximum density does not appear to
be related to that of the previous year.
The effect of temperature in the current vegetation
period can be traced in several sites, but it remains
unclear why the temperatures in the preceding year
should have influenced only a few of the sites. The
calculations showed that the prior July temperatures were
unimportant. This finding must be given due weight even
though it has not as yet been explained eco-physio-
logically. It has been reproduced for several sites using
two different methods and is also eminently reasonable.
No inferences could be drawn from the general conditions
or from the phyto-sociological analysis about which con-
ditions might be limiting cambial activity. The response
functions indicate that wide rings are formed when preci-
pitation in the current vegetation period, May to August,
is high, and the temperature is low. It may also be the
case that high precipitation in the preceding winter is
favourable to growth. There does not seem to be any link
between ring width and hours of sunshine.
Growth in the current vegetation period is considerably
influenced by events in the previous year. It is during this
year that the physiological state of the tree for the next
year is determined. This in turn is reflected in the ring
width. This explains why there is a certain similarity
between the ring widths of successive years.
The results of the calculations performed here cannot
be assumed to have any general validity since the find-
ings produced by the two methods agree only in part and
the calculations of the relationship between the tree ring
and climatic factors for a similar site have produced a very
different picture.

89
Reconstruction
Using single sequences
The response functions indicate which factors most
strongly correlate with tree ring growth. In the calibration
phase of dendroclimatic reconstruction, a model of the
relationship between annual rings (the predictors) and
climatic factors (the predictands) is constructed typically
by some form of simple or multiple linear regression
analysis. This model is also known as a transfer function
(Fritts, 1976). The transfer function coefficients linking the
significant tree ring and climatic features are then used to
transform or rescale the tree-ring sequence into a new
sequence of meteorological estimates. These estimates
extend back in time for the length of the tree-ring se-
quence. It is assumed that this procedure is based on
reliable sequences, for example, tree ring series from
trees which have not been affected by air pollution, and
reliable meteorological data.
In the verification phase the model is checked and
verified. The validity of the model is tested by comparing
the reconstructed climatic data with actual data which has
not been used in the calibration. If the tree-ring estimates
are sufficiently similar to the actual data in the verification
period, then the model and the reconstruction can be
accepted. Otherwise, the model must be discarded or
improved. See Fritts (1976) for methods of model verifi-
cation. In the simplest case only one tree ring feature
reconstruct i on verifi cation
• -
annual ring
series
• -
meteorological
series
dependent
series
I I
1940 1950
Diagram showing the relationship between tree-ring series, meteorological data and dependent series.
90
affected by climate is used in the model. The model can
often be improved by including several ring features both
in synchronous and lagged positions.
For climatologists the verification phase serves to
indicate how far the model is valid. In environmental
research on the other hand it is used as a tool in the
assessment of damage to trees. Where tree growth is
impaired from a particular time, e.g. by environmental
pollution, a tree ring/climate model is constructed using
data relating to the period before the onset of the
damage. The testing phase, which relates to a period for
which meteorological data are available, enables the
extent of the damage to be assessed. (See p. 169.)
No further comparisons can be made in the reconstruc-
tion phase as no meteorological data are available. The
results of the calibration and verification phases serve as
a measure of the reliability of the reconstruction.
These methods are described in Fritts (1976) and
Blasing et al. (1976).
The statistical procedures can only yield good results
when the raw data are of high quality: e.g. from relatively
old trees which have been ecologically well-selected.
Reconstructions become questionable, or at least less
certain, when chronologies constructed from shorter se-
quences are used to provide the raw data.
verification
1,5 mm
1.0 mm
0.5 mm
- lS oC
- 17°C
- 16°C
1970
I
Blasing et al., 1976; Fritts, 1976.
An example from climatology Reconstruction
The reconstruction covers the years before 1885 and after
Calibration
1964,
The predictors used in the regression were the mean
temperatures for July and August for the period 1915 An example from the environmental sciences
to 1960 (46 years) , for the two weather stations Zurich
(Central Plateau) and Great St. Bernhard (alpine). This method is of particular value where the condition of
In the simple model the predictors used were the trees which have been damaged by environmental factors
indexed values for the maximum density from six chronol- is to be determined, The calibration and the testing are
ogies for the subalpine zone of Switzerland for the period carried out on annual ring sequences for a period prior to
1915 to 1960. In the more complex model the indexed the onset of the suspected damage. The reconstructed
values for the minimum and maximum densities for the values are compared with the actual ones.
early and latewood widths for the same period , both syn-
chronously placed and displaced by a year, were used.
!l,O 1920 1930 1940 19&0 1960 1970 1980
Verification
Meteorological data for the period 1885 to 1914 were 2,0,
compared with the reconstructed values. Thus , the two
mean curves of the calibration and verification periods 1,0
were independent in time . 5,0
0'
The values are represented as deviations from the 4.0
mean for the appropriate weather station . For July and
August for the period 1900 to 1940 the means were
- 5,0
17.4·C for Zurich and 6TC for Great St. Bernhard.
In the simple model , 66.4 % of the variance shown
could be explained on the basis of climatological informa-
tion, and the corresponding figure for the complex model ~ 1,0 0
was 70.6%. These values , which were obtained for the ~

calibration phase , sank to 39 .1 % and 62.4 % respectively ~ 3,0'

in the verification phase. In this example, the complex
model is superior to the simple model. 5,0 10
3,0

5.0
1,0
5,0]
4,0

o
I,D
1
2,5 ,
: ,
I
1920 1930 1940 19&0 1960 1970 1980
-5,0
::i_~In~d:!ep~e~~d~e~n~1;,~+:~:==T~ca~'I,bral,on
1900 1950 A comparison between the actual (solid line) and the predicted
pattern of growth (broken line) of a number of groups of trees in
5,0 i Grosser 51 Bernhard Hamburg (Eckstein etat" 1981),
I
1
1
25 ,
1

,I

{',
"
,
. ,,,
:
2,5

Independent - I Caltbrauon:.,:,==:;::,~I- . ,_-.j Using several geographically scattered sequences

-5,0 I i i
1900 1 50
The relationship between tree-ring features and summer tempera-
tures,
Simple case: calibrating the models with one ring feature (maximum
density, 6 variables),
The maximum density provides by far the most information on
summer temperatures, When other ring features are included in the
model, then those years which were very hot can be seen more
clearly, (See, e.g. 1911), Meteorological series from alpine and low-
lying stations are appropriate for calibration ,
This involves the comparison and contrasting of networks
of meteorological and dendrochronological data, The
methods developed by Fritts (1976) and Blasing et at.
(1976) can be used to reproduce climatic conditions over
large areas, Examples of this use of the method are given
on page 169.
The accuracy of the somewhat complicated canonical
correlation analyses depends largely on having raw data
of high quality,

91
 n
Characterizing the meteorologjcal data

The careful examination of the meteorological data that

are to be used in the analyses is as important as the
characterization of the tree-ring series. It cannot be taken Monthly standard deviation for the temperature series for Santa Fe,
for granted that measurements that are made day-in and New Mexico (Rose et al., 1981).
day-out over long periods of time are always reliable.
Descril2tive statistics of Santa Fe I2recil2itation (inches), A.
Sites that are in fact not suitable for the taking of mea-
surements, the fact that a meteorological station has been 95% Confidence
moved, inappropriate measuring instruments or missing Median Mean SE Interval S2

or wrongly recorded values will lead to the inclusion of January 0.47 0.64 .05 0.53-0.72 0.28
false or misleading data in a series. This means that the February 0.65 0.72 .05 0.63-0.81 0.21
data must be checked and complemented where neces-
March 0.70 0.78 .05 0.69-0.87 0.22
sary. There are a number of different methods for doing
this. Two of the more straightforward ones are described April 0.71 0.89 .08 0.73-1.06 0.69
here. May 0.99 1.26 . 11 1.04-1.47 1.18
June 0.75 1.12 .10 0.93-1.32 0.96
Descriptive statistics July 2.18 2.43 .14 2.16-2.71 1. 94
The monthly mean values are calculated and a climate August 2.09 2.35 .12 2.12-2.59 1.47
diagram is produced. The diagram is then compared September 1.24 1.51 . 11 1.29-1.74 1.31
with that for neighboring meteorological stations. October 0.95 1.09 .09 0.92-1.27 0.79
The median values, the mean values, the 95% con- 0.53-0.76 0.36
November 0.54 0.65 .06
fidence limits, the variance and the standard deviation
are calculated and expressed in graph form. December 0.56 0.73 .06 0.62-0.84 0.31
The quality of the results is considered. Cln.UI ll Wl TUIII"'UfH\II[ DlffVI£H:[ S

"",'!1'IfJW
29021' QlftA'..u..·~.(,(tI
lit ... ~IlCI)
lIliDiLtill~ .

Interpolating missing values 15o:'l20 ~ ~fA'C . Ij(WI"O:I C"

Individual values are often missing from a series, perhaps . . 'j'

because the person responsible for the measurements "'" I
was ill or because he or she was not able to record the ) " .(1 -

I
measurements, as in war-time, for example. The missing :!l;:;' j
values can be estimated and then interpolated using data ~! m,oJ
from neighboring stations. As a preliminary, the correla- ~ .~
:::. 17' a I,
tions must be calculated in order to assess the suitability 1L6-Clj
of the series for comparison. If the series prove to be
comparable, a mean line can be calculated using linear
regression. The deviation from this mean line for the
0
:0 10 "'--",,:,±
I
.:::1
19JD
:---:-:'"c:----C~........,:':'.:":--~I..,..........,±'
Itl0 U~~"II:II:: ,e 4g. 1950
:----"-:-:!:----:-:!::,I
lr?Q
--::t'
1*
corresponding months of other years at neighboring
Cumulative differences in temperature for the months July to Sep-
stations is an indication of the size of the missing value. tember for the metereological stations Santa Fe and Albuquerque.
There is a slight anomaly for both stations for the period 1910 to
Determining the homogeneity of a series
There are a number of ways of checking for variations in
1920 (Rose et al., 1981).
.
.
.... +-
.. Olt20
the homogeneity of a seri~s. The following two simple
.
..
,
'
tests can, for example, be applied: ,
'
The cumulative temperature difference for two neigh-
boring meteorological stations is compared with either , .
the monthly, the seasonal or the annual temperature
series for these stations. The differences between the ~ .
.,,...'
two stations for the period in question are then calcu- ~ ,"

.....,
lated and plotted against time. ~
The monthly, seasonal or annual precipitation series
for two neighboring stations are compared by means !
§
,
of the O-mass. The total annual precipitation for one
il "
station is plotted against that for the other. i
Where both these tests produce more or less straight ~

lines the series are homogeneous. Where the lines have i

a kink, data from other nearby stations must be used to
determine whether or not the series is usable, and if so,
how it is to be corrected and what the reasons for the
deviations are.
i " ...
29ISG'7G p£COS . I£W I£XICD SlMR ...u... • FUl • S(]I'f
D-mass analysis of precipitation for July to September for the
meteorological stations Santa Fe and Pecos. The only deviation is a
slight one around the year 1940 (Rose et al., 1981).

92 Roseetal., 1981.
Characterising meteorological data from the pre-measuring time

Recently many authors have tried to convert the subjectively written weather records into absolute values. Using very
much material Pfister 1984, 1985 found close relationships between these values and the maximum density in
conifers from the Alps. (See page 161.)
A very large, critically reviewed data set from the 16th to the 19th century (CLlMHIST) is available from
METEOTEST, Fabrikstrasse 29A, CH-3012 8ern, Switzerland.

lluiCtn.N.
m4.~ 116ttc ker ",6 i~ ~ g&bml
em tN)lg n~/frir4l' 'iir 'iin Ia..
3ni 3c"ntr ~i" Si6, B\)41t S, 'iierl
eo4ll'ff 'iem 90m muGfi 14flin cr.. r..
nnb 'ieon binn.cb "'01 ncmmcn 4~' I
\1')4& 'iir 4U mon41 Sunifdh4fli.

Pocket calendar for 1573 converted to a weather diary by Wolfgang Haller, parson in Zurich

From the 16th century onwards calendars were published yearly which contained in form of symbols astrometeorological weather
predictions for every day besides recommendations for agricultural work and personal hygiene. In addition a line was left empty for every
day in which the owner could note down personal remarks. Wolfgang Haller (1525-1601) was a schoolmaster and a parson in the Canton
of Zurich. Later he became provost at the cathedral in Zurich. From 1545 to 1576 he kept track of the daily weather in his calendars using
a number of stereotype expressions. The graph shows his comments for January 1573 (Julian Style). In 1572 the months of November
and December were among the coldest in the last 500 years. Most of the rivers and lakes in Central Europe were covered with ice.
Between Jan 6 (16) and 25 (Feb 4) 1573 a warmer period followed. A translation of some of his observations is provided below (the
Gregorian Style is given in parenthesis)
Jan 22 (Feb 1) snow and dull all day long.
23 (Feb 2) dull.
24 (Feb 3) dull and snow.
25 (Feb 4) thin cover of fog in the morning. Pretty sunny in the afternoon. No wind. Fog. Rain.
26 (Feb 5) dull and cold.
27 (Feb 6) dull and cold.
28 (Feb 7) dull and cold. Fog.
29 (Feb 8) dull very cold. Much fog on the mountains.
30 (Feb 9) dull very cold.
31 (Feb 10) very cold. Bright in the evening.
By courtesy of Ch. Pfister.

93
 III Tree-ring growth and the site

Cross-section through the twig of a clematis (Clematis campanuliflora)

F. H. Schweingruber, Tree Rings

© Kluwer Academic Publishers, Dordrecht, Holland 1988
Growth increments and tree rings in tree trunks

Growth increment layers occur in the wood and bark (rhytidome and cork.). Both the vascular and cork cambia form
cells in irregul ar sequences determined by internal or external factors.

Growth fluctuations in wood

The ability of woody plants to form growth increments or
tree rings depends on their position in the plant kingdom .
None of the Pteridophytes, neither Paleozoic forms
e.g. giant horsetails, club mosses, nor living representa-
tives, form tree rings. Within the seed-bearing plants tree
rings are absent in monocotyledons and cycads . Gymno-
sperms, on the other hand, display growth fluctuations
dictated by climatic cycles. They first appeared around
395- 345 million years ago , during the Upper Devonian
period in Europe and North America, in the Cordaites,
forerunners of the conifers . The ability of a plant to react
to environmental factors varies according to family, genus
and species; in certain taxa a change in the environment
strongly influences the form of the tree ring, in others only
a slight fluctuation results . Many living angiosperms seem
to follow endogenous growth rhythms which are inde-
pendent of the climate.

No tree rings in the wood of an arboresecent No tree rings in the wood of an arborescent No tree rings in a living woody monocotyle-
fern of the Triassic period (Hirmer, 1927). Equisetum of the Carboniferous period (Zim- don. (Smilax aspera)
mermann,1959).

Fluctuations in cell size in a Tree rings and density fluctua- Tangential banding caused by Tree rings in an oak.
gymnosperm of the Keuper tions in a gymnosperm of the internal factors, and annual ring
period (Dadoxylon sp.) present day (juniper). boundary (arrow) in a member of
the Chenopodiaceae.

96
Growth fluctuations in bark

The earliest types of bark in gymnosperms display an
endogenous rhythm; layers of sieve cells, parenchyma
cells and bark fibres are interspersed. Climatic cycles
affect this rhythm in different ways; sometimes broad
bands of particular types of cells of uniform size are
formed at the end of the vegetation period or bands of
larger or smaller cells are formed in the early or late bark.
Even in areas with a seasonal climate some species do
not form annual rings in the bark. The structures of the
bark are obscured by secondary changes such as col-
lapse of sieve elements, formation of sclereids or dilation
of rays. The fluctuations in tree-ring width in the bark are
less than those in the wood according to Holdheide and
Huber (1952)

p
S
B
S
P
S
B
p

Endogenous rhythm in bark of juniper. Sieve
cells (S) and parenchyma cells (P) and alter-
nate with bark fibres (B)
Top: 60:1 Bottom 300:1
(Holdheide and Huber, 1952).
Tree rings in bark of alder. Large sieve
tubes are formed in spring, small paren-
chyma cells in autumn.
Top: 40:1 Bottom: 60:1
(Holdheide and Huber, 1952).
Analogous tree rings in wood and bark of
larch. Only those of the wood vary in width
from year to year.
Top: 16:1 Bottom: 8:1
(Holdheide and Huber, 1952).

Growth fluctuations in the rhytidome

Only in a few species does the cork cambium, which

develops later than the vascular cambium, form a new
layer each year. The cork cambium is often only locally
active, so that the bark is formed in patches or scales e.g.
poplar. Holdheide, 1951, reports a certain periodicity in
some central European species. The cambium, which in
some cases is active for a number of years, produces
layers of cork, but these rarely correspond to annual
rings. In conifers the growth rhythms, presumably endo-
genous, are to be seen in thin walled spongy cork and
thick-walled stone cork cells. In broadleaves the patterns
are less clear. Growth rings in the bark of an aspen. 20:1
(Holdheide and Huber, 1952).

Hirmer, 1927; Holdheide and Huber, 1952; Zimmermann, 1959. 97

Sensitive tissues and growth processes

Organic compounds produced in the leaves are incor- production of new cells by cell division.
porated into the structures through two basically different differentiation of the newly formed ceUs resulting in
growth processes: varying sizes. forms and functions.

The cambium

The cambium produces new cell elements. periderm

differentiation

cell formation ~ - phellogen

cell formation ~
~ camblum
dlfferenliatlOn
xylem

Aspen (Populus tremula) lOx (Holdheide and Huber, 1952).

Cell enlargement

The enlargement of cells results in individual elements

having different sizes.

Wood with large cells 250x Wood with small cells 250 x

spruce beech spruce beech

98
Cell wall growth

Cell wall growth results in thickening of the cell wall.
Only the walls of living cells can be thickened; the
presence of plasma and nucleus is infallible proof that a
cell is alive. Earlywood tracheids and vessels live only a
few days after their formation. Shortly after the completion
of cell division, the cell enlarges rapidly but soon dies and
immediately begins to function as a conducting element.
Latewood cells, on the other hand, live for 2-3 months
after their formation in temperate regions, and during this
time the walls are thickened . The cells do not die until the
end of the vegetation period .

Transverse sections through trunks of young spruces (Picea abies) seen under polarised light 25:1.

July: the tree ring is not yet complete. Practically all the cells have November: tree-ring formation is complete. The latewood cells are
been formed, but those of the latewood are not yet thickened; they fully developed and show up in polarised light.
do not show up In polarised light.

Wood with open-textured tissue, 250 X Wood with compact tissue, 250 x

earlywood latewood compressionwood latewood

Spruoe Pine pine pine

Bosshard, 1974/5; Esau, 1969; Fahn, 1974; Panshin and Zeeuw, 1971;
Wagenfuhr, 1966. 99
Distribution and number of cells

The basic pattern of cell distribution and cell size is is influenced by external factors to such an extent that a
genetically determined but is still very flexible. Cell growth large structural variation occurs within the genetically set
framework.

Number ot pores 70 X Latewood width 70x

few many narrow wide

beech beech pine pine

Earlywood width 70x

narrow with a few large wide with many large

pores pores

The determining factors.

Genetic Environmental and physiological

factors factors

Species Soil Age

Race Climate Position in
Geotropic Weather tree.
and Location Size and form
phototropic Slope of crown.
strength inc! ination Size and form
Mechanical of trunk.
forces (wind Position
snow) within stand.

together determine number, structure and arrangement of cells

Type of tissue Amount of wood (ring width)

Cell wall and th ickness
Fibre length Arrangement of cells.
Chemical composition
Physical microstructure Proportion of cell types and
Lignification forms.
Formation of heartwood

Trendelenburg and Mayer-Wegelin, 1955

oak Oak

100
Secondary changes

The activity and life span of all cells is limited. In some microscopically, tyloses or deposits in vessels and paren-
species the breakdown can be seen morphologically. chyma cells can be recognized. These deposits protect
Macroscopically it shows in the formation of heartwood; the non-living parts from decomposition by fungi.

Tyloses in earlywood vessels 100x .

Forma Ion of yloses.

Schematic diagram of the formation of tyloses. The pit membranes Oak transverse Robinia
of the parenchyma cells neighboring the vessels divide and pene- (Black Locust)
trate the vessels themselves, thus blocking the passage of water longitudinal.
(Fahn, 1977).
Deposit of coloured Deposit of coloured
substances in the substances in the
parenchyma cells. vessels.

Douglas fir (Pseudotsuga menziesii) with reddish coloured heart- Tetraclinisarticulata 250x Cistus250X
wood and yellowish sapwood.

101
Genetically determined differences in reactions

Physiological-ecological amplitude

Ecological factors bring about a natural selection of plants associations e.g. pine (Pinus sylvestris). Beech (Fagus
with similar physiological-ecological characteristics so that sylvatica), with a narrow spectrum is found in only a few
species areas and plant associations develop. Species plant associations.
with a broad physiological spectrum occur in many plant

Ecology diagram for pine (Pinus sylvestris) Ecology diagram for beech (Fagus sylvatica)

prevail ing
dry

flourisihing

~ prevailing

acid _ alkaline aCid dlkaline

Pine has a very broad physiological spectrum and is pushed to Beech has only a narrow physiological spectrum but because it
the limits of its adaptability by other strongly competitive species is very competitive it is able to form forests on optimal sites
(Ellenberg, 1965). (Ellenberg, 1965).

beech

ITIIill pine
IHHlI beech
I::ItmI and pi ne

Distribution map for pine (Pinus sylvestris) and beech (Fagus sylvatica)

Pine, with its wide range of tolerance occurs on coastal sites in Beech, with its narrow spectrum and limitation to optimal sites,
Scotland, arid sites in Spain and arctic sites in Scandinavia and forms forests only in temperate regions in Europe.
Siberia.

102
Variability in form within a plant
Within every plant there is a great variability in forms,
because the differing environmental factors affect growth
differently. The environmental conditions affecting the
growth of a peripheral root and its functions, for example,
are fundamentally different from those affecting a twig in
the tree crown. The variability is so great that it is often
impossible to identify even the species under a micro-
scope from a specimen taken from the root, for example.
Dendrochronological studies can only succeed if samples
are always taken from the same part of the tree, such as
the trunk. By taking the average of many individual data
values from rings formed in the same year, some of the
variations can be eliminated.

Deciduous: common elm (Ulmus campestris)

thin rool Ihlck root trunk long shoot short shoot

Conifer: spruce (Picea abies)

thin root thick rool Irunk branch short shoot

Structural variability within individuals. Above a example from the group of ring-porous angiosperm, below a confer. All illustrations 2Sx.

Cutler, 1976; Baas et al., 1976. 103

Variability of form within a climatic zone

Some climatic zones seem to affect the variability of form In wet areas which have cool summers the woody
in broad-leaved trees . In the course of evolution specific tissue of bushes and dwarf shrubs has numerous tissue
structural characteristics seem to favour certain climatic cells and vessels with narrow internal diameter.
conditions.

Alpine and arctic zones

The vegetation period is short and growth is limited mainly rings and very numerous small pores.
by the low temperatures. Dwarf shrubs with narrow annual

Salix retusa Empetrum hermaphroditum Rhododendron ferrugineum Phyllodoce coerulea

All illustrations 40X

Areas with dry summers

The vegetation period is short and there are long periods is limited mainly by low precipitation . Mainly shrubs with
of drought interrupted by brief intervals of rainfall. Growth fairly narrow rings and many small pores.

Helianthemum umbel/atum Cneorum tricoccum Coronil/a juncea Cytisus nigricans

All Illustrations 40X

104
Trees and bushes on dry sites in areas which have hot,
dry summers have small cells and pores. The pore
density however is usually much less.
In tropical areas trees often have large pores and it is
not unusual to find large groups of parenchyma cells in
the region of the pores. Many species have septate fibres.
The form spectrum of species growing in temperate
zones includes characteristic features from arid, boreal
and tropical regions.

Temperate zone
The vegetation period lasts several months and there is Trees and bushes have fairly wide rings and quite a lot
plentiful rainfall in all months. Growth may be limited by of rather large pores.
excessively low temperatures or lack of rain.

Acer pseudoplatanus Salix caprea Tilia cordata Fraxinus excelsior

All illustrations 40X

Tropical zones
Growth is not interrupted by climatic events to any great not usually correspond to annual rings. The tissues are
extent. Trees form large increments. The growth areas do perforated by many large pores.

Anacordium occidentale Microberlinia brazzavillensis Shorea eximia Goethalsia oneiantha

All illustrations 40x

Bormann F. H., Berlyn G. (edit.), 1981. 105

Reactions of a species to climate and site

Conifer - Pinus sy/vestris from the Mediterranean, tem-
perate and boreal climatic zones.
The physiological-ecological spectrum of different spe-
cies can be established through numerous observations.
A clear picture of the suitability of a species as a climato-
logical data storage unit can be obtained by recording the
influencing environmental factors and the manifold reac-
tions of the tree, as described below for the Scots pine
(Pinus sy/vestris) . All the features indicate the genetic
plasticity of the species i.e. its diversification into many
races and its outstanding ability to adapt to various cli-
mates and sites.

Fort William. 57 m
Scotland
oC
Climate 50 100 JoHmo". 255 m
Swooen
40 80
Climate in the area of distribution of Scots pine. This tree 30 50
forms forests in cool, damp maritime regions as well as in 20 40
the semi-arid Mediterranean area.
10 20
o-6!zm:tm::!:::::::ItlZ1:tzzW- 0 ~"":z:tzzZ.d:ztI:z:~~ ;.1IiItz:tllt-'--'-'-'-1:1Zr>W-

maritime arctic semi-arid

Characterisation of the climate in the area of distribution of Pinus

sylvestris.

External characteristics

Area of distribution Tree form

The area is extremely large, covering almost the whole of the This is very variable and the trees range in form from 40 m giants
Eurasian continent. with rounded or very slender crowns to gnarled shrub-like speci-
mens with pOinted, rounded or lop-Sided crowns (Phillips, 1978).

Sites

Site in or on the edge of a bog on the arctic Site on windy coast Site on an extremely dry, steep slope in a region
timber line with low precipitation

106
Internal characteristics

Anatomy of the annual ring Density diagram

III J

r~
u
u
'" 0.6
i .0

0.8-
cold-weI

>-
i;i 0./1-1
cQ)
lIll\i '0 0,2 ~
l';C
o
~
..r;~
temperate
10.;
;JIW

0.8 -l
IP,
~
~ 0,4..,
Q)
'0
0,2..,

o
Scots pine (Pinus sylvestris) 40x
Wide tree ring with density Narrow tree rings with weak Regular density profiles of conifers in arctic and arid zones, irregular
fluctuation in a tree on an latewood in a tree on a sub- in temperate. (See page 71.)
optimal site. alpine, extremely well-drained
site.

Tree rings from climatically different areas produce dif-
ferent density diagrams. In arid or arctic regions the
density climbs smoothly from earlywood to latewood and
the maximum lies at the boundary. In temperate zones
the density fluctuates and the maximum occurs before the
ring boundary.
The peak density values mainly reflect climatic features
over large areas; in arctic zones the maxima are low and
the minima high, while in temperate regions the maximum
values are higher. The mean width values chiefly indicate
local site conditions,

Raw data Values for density and ring width of Pinus sylvestris in various climatic zones in Europe.
Mean values for 12 trees at 2 radii,

Density (g/ cm 3) Widths in mm Age

maximum minimum earlywood latewood whole ring years
Cold and damp- Scand i navia
Scotland
sh ort vegetati on peri od
northern Norway Narvi k Q35 0.56 0.15 072 277
073
072 0.35 0.41 0.15 058 494
Lofoten 1 1

Long vegetation period

western Scotland 0.79 0.33 2.49 1 17 3.67 74
eastern Scotland 0.80 J2J..§. 0.86 0.36 1 23 271
temperate- northern Germany
030 0.95 0.78 1 74 71
Diersfordt
Xanten ~951
0.87 027 1,88 109 2.91 46
dry- Switzerland
Krauchtal (locally dry) 'cf84 0,34 0.46 0.22 0.69 200
Bratsch, Valais 085 0.35 0.90 0.24 1 14 281
Raron, Valais _0J3Q 0.30 0.93 0.24 1 18 110

107
The reaction of a flora to climate and site

Area chosen - tropics. Deciduous, in multiples of six months.

The flora in certain areas is influenced by the develop- Biennially deciduous
ment of the environment in the past, by site conditions at Irregularly deciduous.
the present time and the climate. In boreal zones, for
example, conifers are dominant whereas in the tropics
a great variety of broadleafed trees are to be found.
Although hardly any dendrochronological studies have
been carried out in the tropics, an attempt has been made
to find out how cambium reacts to a climate which is
nearly always humid. The tropiCS were chosen because
hundreds of species are to be found here and genetic
growth processes are not obscured by extreme climatic
conditions. Only in forests which are inundated annually
for 100 to 300 days, e.g. Rio Negro, Amazonas, do trees
form annual rings (Worbes 1985).

Climate

The tropical climate is very diverse; more or less con-

stant with high precipitation in the rain forests, partly
seasonal with a period of low precipitation lasting for
several months or, in the mountains, seasonal with cool
dry phases lasting several months. Further examples can
easily be found.

External characteristics

Angiosperms with large leaves are by far the most numer-

ous. The differing physiological activity during a single
year can be clearly seen in phenology. Even in Malaysia,
which does not have a particularly seasonal climate,
Medway (1972) found trees which behave very differently
with respect to leaf-fall:
Evergreen
Annual, but often different times of activity in different
years.
Irregularly biannual.
Annually deciduous.
Dipterocarpus incanus with and without foliage on Adamanen. A
rhythm, internally controlled and independent of weather changes,
causes each individual tree to shed its leaves. The time when this
takes place varies (Walter, 1962).
Regions with tropical climates in which most trees and shrubs do
not form annual rings. In extremely constant, humid climates as well
as cool, wet mountainous ones, growth increments form which can-
not be described as annual rings (Walter and Lieth, 1961-1967).
120

100 E
E
u 80 .~
o
c c
50 .g
~
40 '5,
'w
f!?
20 a.

o
Indonesia. Sumatra Africa. Cameroons Central America Central America
Costa Rica MeXICO
P/year 2328mml 2753mm 1810mm 582mm
T/year 26.6 OC 26.6 OC 20.2 o C 2.8 o C
Altldude 50tn 6m 1135m 4221m
low·lymg low·lying central eleva ion high plateau·paramo
Constant. humid Short periods with Short dry periods Short dry periods
low precIpitation Temperate ,wet Cool, wet
Humid

108 Medway. 1972; Walter, 1962; Walter and Lieth, 1961-1967.

Internal characteristics Transverse section
Cambial activity in different species varies even with the Abrupt boundary on one side: Thick fibres and terminal
same climatic conditions. Depending on sensitivity, i.e. parenchyma. No abrupt boundary: Progressive change in
genetic tendencies, individual species react to relatively amount or width of parenchyma band.
minor changes in daylight hours, temperature and precipi-
tation. Fahn (1981) explained this in an investigation Initial band of fibres mostly without vessels or paren-
carried out in Israel in which he studied the relationship chyma.
between the time when wood was formed, flowering and Different cell and cell wall dimensions.
leaf-fall as well as tree ring boundaries. In species dis- Periodicity in vessel diameter and vessel arrangement.
playing seasonal growth behavior, cambial activity cor- False rings: periodic or frequent.
responds closely to the opening and falling of the leaves. Discontinued rings.
The cambium in trees growing on wet sites is active the Traumatic tissue: resin ducts, kino veins, other secretory
whole year round and the time when the leaves open is veins.
no longer concentrated on winter and spring. In this group Reaction wood.
are to be found species with annual growth increments Anomalous wood e.g. phloem.
(Tamarix ga//ica), with biannual (Tamarix aphy//a) and
those without any growth increments (Acacia). Irregular Longitudinal section
or regular periodicity can be seen in the structure of the Interlocked grain, different length of fibres, tracheids,
growth zones, particularly in the boundary areas. Fahn parenchyma cells, vessels.
1981 referred to the following possibilities;

Re l omQ fo olom t F"OtI l j),uh

XI

L
XI

F
I
" III IV V VI VII VIII IX X

A, U:rTII'ItO monosC)e,mo O~ I

ZY9C1phyllum chJmosum SotS!. F

L
-
L
t-
RtQumuflO poloe'llno Bolli L F

Anab<l"1 0, 11(\11010 ( Forsk )MOQ L F

I--
ColltQon1.lm comosum l Hit
r-
L F
-
f--- -
Quercus tlhobu, t l'tlll IOt CMl80tU F
L
, Que-reus Iftfee10f tQ OilY F
L
I Querc lJl c:olllp'",,01 Webb
F

L
L
- t-
CrOIQegu, (uotolul Webb
F
Plslaclo atlonhco Des I r
L
F
Pis-ioelo PQlou1rnQ 80151
.:
P' lsiOCtO ItnhSCUi L F L
--
CerQlontO Sd tqUO L F L
TomOfl), Jordonts BOlli L
F
YClf neG lvl nSIS zan F
Tomarl. Qalhca L L
ItOr mor., - IT'IOflUIIGulm' Zon F

.
E ucal)oPlus can"ddi.ile1"1S1t Of-I\n F L

r L
Tamonl!. OP"yIiO I L Kar,,1
)
r L

AcotH, IQfl.h s (f"orSII, ) HOyf'lP F

Aca clo r o0(210,.0 SOVI F

Acaclo C ~.,opt\)lIa L l'Ot F

Th.ymeloeQ 1'lIr"ulO I L ) Endl F

Annual periods of secondary xylem production in woody plants in

Israel. Dotted areas indicate cambial activity. Bold vertical stripes
indicate the growth ring formation in plants with continuous cambial
activity. F = beginning of flowering: L = beginning of leaf develop-
ment (Fahn, et al., 1981).

Different reactions within a species to relatively minor climatic dif-

ferences.
(a) Individual trees on wet sites form no annual rings.
(b) The same types on dry sites form clear annual rings.
(c) On intermediate sites density fluctuations occur during short
periods of low preCipitation.
Inga oerstiana, Nicaragua 25:1

Fahn et a/.. 1981; Worbes, 1986; Car/quist, 1980. 109

Ecologically important features in wood

All tissues capable of reacting reflect environmental in-
fluences in their behavior. Every anatomical feature of
every species, or at least every genus, is limited by
genetic factors. Most reactions to climatic-ecological
factors can be seen in the microscopic anatomy of the
plant. From the beginning, dendrochronology has been
concerned with total annual ring width; the distinction
between earlywood and latewood widths has seldom
been considered. A few time-consuming investigations
have been carried out on parameters of cell elements in
earlywood and latewood, such as vessels and tracheids,
and on the proportion of various cell types such as
vessels, parenchyma and ray cells . The importance of
climatic evaluation will increase with the use of electronic
analysis.
In what follows only a few significant anatomical topics,
which can be quickly and unmistakably identified are
discussed: annual ring width, earlywood and latewood
widths cell size and thickness of cell walls.

Tree-ring widths and boundaries

The width of the annual rings in trees is to some extent
genetically determined; a poplar forms wider rings than a
bilberry growing under the same climatic conditions. The
structure of the ring boundary is also partly fixed ; it is
always more easily identified in the ring-porous ash, for
instance, than in the diffuse-porous apple. The boun-
daries are marked by the size difference between the
earlywood and latewood elements and the number of
small, thick-walled, tangentially flattened cells in the late-
wood.

Species from alpine and arctic zones All illustrations 40X

sharp boundaries diffuse boundaries

J~
....... 1'Vr'"
- 'I"I!
.
i
~- )f'
}j

~
~~
~ ~ II: :\
~ ~
~
~ It~ ~
~
~ ~
" , ~~

Juniperus communis Salix retusa Alnus viridis Salix herbacea

110
Species from temperate zones All illustrations 40x
clear rings indistinct rings no rings

Fraxinus excelsior Rosa arvensis Heelera helix Viscum album

Species from semi-arid zones

clear rings indistinct rings no rings

;
:

g:i
t:!:

ll:5: ~ ~

~~
::g
~
Juniperus thurifera Morusalba Myrtus communis Nerium oleander

111
Widths of earlywood and latewood and their demarcation

In some cases information on climatic conditions can be In broadleafed trees it is difficult to define, and con-
obtained from the earlywood and latewood widths as well sequently to identify, the boundary between earlywood
as from their proportions in the whole ring . In conifers the and latewood , since elements of differing shapes and
demarcation , i.e. the boundary between earlywood and sizes are involved. In conifers the proportion of latewood
latewood, is partly genetically controlled; it is sharper in is significant ; it is higher in areas climatically favourable
larch than in spruce for instance. The abruptness of the than in unfavourable ones. In temperate zones latewood
transition can, however, be affected by environmental seems to be influenced by precipitation.
factors. Microscopically the boundary can be affected as
follows: if the thickness of the cell wall X2 is greater than
the width of the lumen, then the cell belongs to the
latewood. This criterion is just as arbitrary as the densito-
metric one described on pages 71, 117.

All illustrations 40 X
The abruptness of the earlywood to latewood transition
is sometimes almost a species characteristic. In larch it is
sharp, in white pine more gradual.

Larch White Pine

Douglas fir - rather dry site Douglas fir - damp site Apple - damp site Apple - dry site

Features typical for the species can be altered by eco-

logical factors. Normally sharp transitions between early-
wood and latewood may become gradual and vice versa.

112
Cell size

To date, little work has been done on the relationship last-formed cells usually mark the ring boundary. Even
between cell size and environment. In dry areas, for within one tree ring, the cells vary in size. In arid areas,
example, cells in a tree-trunk are smaller than in one of smaller cells seem to appear at the time of earlywood
the same species growing on a wet site. There are also formation, whereas in temperate zones there are fre-
variations within the tree itself; cells are generally larger in quently fluctuations; nor do cell size and cell wall thick-
the trunk than in the branches. In arctic zones cell size ness always seem to be related.
decreases continuously in the course of the year. The

All illustrations 40X

Holm oak, Quercus ilex Silver fir, Abies alba

Dry site, small cells damp site, large cells twig, small cells trunk, large cells

1. ~
tranSluon

'~ . ...
-M
" ~ </>

,.\/\ • 1\-V_
V>
ti ~ W
d>

.. J • j\r.'{'-
I ~ /; C
h!l 1\
,. ..
-
\., 1- . U

.
IL~ ~. ' u
1 ., ; 1 \/\ J
~
~\
'0 -5
'\ ""
e
\.\ /J
continuous

'?U
u
'i';
>. continuous
U
~ 10 sh.up ._---- --. ~
>. '0 00 >00

cell number along (he radial lie

12. ... Il,inmJiJI ring - - -

sharp -+-----+-0 -- .~
-- -- -- )=:(
continuous
tl
Cell wall thicknesses and radial cell widths in a wide ring of a Cell sizes and cell wall thicknesses in a twig 40x. Large cells
Sitka spruce. In this case latewood formation is not yet complete with thick and thin walls occur next to small, tangentially flat-
and the cell wall thicknesses in the latewood (from the 100th cell) tened cells with thick and thin walls. (Schweingruber, 1980).
are lower than in the earlywood (Ford and Robards, 1976).

113
Cell wall thickness

Density is to a large extent the densitometrical expression variations in density within the annual ring supply valuable
of the cell wall thickness. Peak density values as well as climatological information.

Maximum density
In areas with seasonal changes the thickening of cell walls thetic products. Under good growth conditions the cell
in the latewood continues over several months, a process wall thickness is considerable, under poor conditions
which largely represents the transformation of photosyn- reduced.

Maximum density in latewood 250x

(a) Spruce (b) Spruce (c) Larch
Small latewood cells with thin cell walls Latewood cells with fairly large lumina. Large latewood cells with thick cell walls
and large lumina. Density approx. 0.8 g/cm 3 and large lumina.
Density approx. 0.6 g/cm 3 Density approx. 1.0 g/cm 3

Usually maximum density coincides with the tree-ring boundary but in temperate regions, especially with conifers , it
often occurs before the boundary.

Three tree rings from Scots pine (Pinus sylvestris) growing in low-lying areas in the maritime-type climate of northern Germany. The
maximum densities occur:

(a) on the ring boundary (b) near the ring boundary (c) in the middle of the ring.

114 Schweingruber, 1980.

Minimum density
As a rule the first earlywood ce lls are formed within a few
days and die very shortly afterwards, so there is little time
for cell wall thickening . In temperate and arctic regions the
climatological significance of the minimum density seems
slight, whereas in arid regions it is great. The thickness of
earlywood cell walls varies from species to species.
Extremely low minimum densities in larch indicate that it
has been attacked by larch bud moth.

Minimum densities in earlywood

Larch Spruce Tetraclinis articulata
Large earlywood cells with fairly thick Small earlywood cells with thin walls. Small earlywood cells with thick walls.
walls. Density approx. 0.30 g/cm 3. Density approx. 0.40 g/cm 3.
Density approx. 0.25 g/cm 3.

Difference between maximum and minimum density

The difference between earlywood and latewood is to be assumed, however, that the variation from year to year
a certain extent genetically determined, being always has a certain climatological significance.
greater in larch than Swiss stone pine , for instance. It can

Density difference between earlywood and latewood in

(a) spruce - great difference; earlywood (b) Swiss stone pine - slight difference; All illustrations 250x
cells with thin walls, latewood cells many earlywood cells with thin walls,
with thick walls. and few latewood cells with thick
walls.

115
Intra-annual density fluctations

Density fluctuations contain much climatological informa-
tion but up to the present little has been elicited. Density
fluctuation is a change in cell wall density within the early-
wood or latewood. The number, extent and position within
the individual ring vary greatly. Only the strongest fluctua-
tions can be synchronized with each other in different
trees or parts of one trunk, since their formation is partly
dependent on the ring width, i.e. the vitality of the cam-
bium at a given point.
Not all tree species have the same ability to form cell
walls of varying thickness in response to changes in
environment. For instance, the Cupressaceae seem as a
rule to have many fluctuations, Picea fewer. In arctic
climates practically no intra-annual density fluctuations
occur; in cool-temperate zones they are very common,
although there are rarely more than two per vegetational
period. In warm, temperate and tropical areas there are
several each year.

Cupressus sempervirens

Density variations in different

species on the same site.
Pseudot~uga
menziesii

Annual rings A-F each with one density fluctuation, formed at the A
same time, but in different positions in the ring. This occurs partly in
early latewood (A) partly in earlywood (F).
B
C I
o
The number, intensity and position of the density fluctuations varies, E I
I
even under the same environmental conditions. Cupressus semper-
virons has many, for example, Larix few (Schweingruber, 1980). .. .. F
.... annual rIng Wldlh ...

~---------------------------------------------------------1-3mm
dnnual rIng Width
/
.<::

'"
80
densl tv ilUClucHions
"-
ii
~
I '";:c
?
'cc"
::>

"

Relationship between tree ring

width and frequency of density
fluctuations in trees growing
on the same site. Spruce in the
SWiss Central Plateau (Madiswil)
1950 1960 1970 (Schweingruber, 1980).

116
 2~J& ~~
Cortin •• 1800 m
Sub,)IDlne Iidly
or (Hell
A I 1\ " 12

I
%
60

20
40 Horgen 570 m
tempe-raw lO Canlo" of lunch
conllnt'r"al n ~ 13 Swll7erland
0

Drersfordl 50 m
onh·Rhrne We<lphdl."
n ~ 14 Fed. Republrc
mart li me of Germany
cool

Varying number of fluctuations per year within homogeneous sites

in different climatic zones. The percentage per site is shown
(Schweingruber, 1980).

glee spruc~. Picea abres Ar(lucarliJ araucana

The density diagram 1.0

maxImum density

0.8
The density diagram demonstrates anatomically ascertain-
0.6~
able cell thicknesses and dimensions and can thus be
said to provide information characteristic of a species or o 4 .. J minimUm ....
. densllY
genus. With electronic techniques it is possible to set 0.2 early·lalewood
thresholds and register peak values for the irregular .... - -- I
density profiles automatically. annual ring

Typical density diagrams for two species of conifers. In many cases

density amplitude and form of the increase from earlywood to late-
wood are characteristic for the species.

By fixing thresholds at particular density levels it is
possible to distinguish between earlywood and latewood.
The question arises as to whether the demarcation should
be set at one constant value for all annual rings or
whether a particular fraction of the difference between
maximum and minimum in each individual ring should be
chosen. In either case the criterion is just as artificial as
that based on anatomical analysis.
Electronic techniques allow a differentiation to be made
between earlywood and latewood values. But a useful
record of intraannual density fluctuations cannot be auto-
mated since the intensity of such fluctuations varies
greatly within one site or even one tree. The operator
himself must decide which peaks are to be recorded in
each case.

~
I (J

OH

0.6

0.4

0.2

~I bl cl
0 annual ring annual nng

Demarcation of earlywood and latewood. (c) Density integration of the earlywood and latewood zones. All
(a) at a constant density level e.g. 0.50 g/cm3 • values below the threshold are ascribed to earlywood, all values
(b) after a given percentage of the total amplitude, e.g. 20% above above to latewood.
the minimum.

Schweingruber, 1980. 117

Relationship between tree ring features

Little is known so far about densitometric relationships within annual rings. It is necessary to distinguish between
technical-statistical and biological-ecological relationships.

Technical relationships to tree ring anatomy

A relationship exists between the form and size of the the darker coloured latewood and a spuriously low late-
light integration field and the anatomy of the annual ring. If wood density is recorded. Such cases can only be clar-
the measuring slit is approximately the same size as the ified on an anatomical basis.
ring, the light coloured earlywood can get merged with

Statistical relationship to tree ring anatomy

When the percentage of latewood is low and that of the
earlywood correspondingly high, statistics show a formal
relationship between total ring width and percentage of
earlywood and latewood. The selection of a process by
which relationships are determined is important. For in-
stance, calculation of the correlation between two curves,
whose progression is similar but whose absolute levels
differ, shows no statistical relationship. If sections of
curves are compared, statistical processes may indicate
spurious relationships.

All illustrations 40X

Scots pine, Pinus sylvestris with measuring
slit of the densitometer drawn in. The width
of the ring is roughly the same as that of the
slit. The density profile cannot be correctly
determined under these conditions.
Black locust, Robinia pseudoacacia. Early-
wood constitutes four fifths of the annual
ring. Statistical relationships can be shown
between widths of eariywood and the whole
ring.
Black locust, Robinia pseudoacacia. The
earlywood constitutes only a quarter, the
latewood 3/4 of the ring. There is no rela-
tionship between earlywood and total ring
width but there may be one between late-
wood and ring width.

118 Schweingruber, 1978, 1979, 1980.

Biological-ecological relationships

Genetic and climatic-ecological relationships as well as Species characteristics

correlations between tree ring features are to be ex-
pected, since various characteristics such as length of
needles, time of leaf opening , speed of shoot growth,
distribution areas, ecological requirements such as struc-
ture of wood and bark and speed of water conduction etc.
etc. are all related to each other. Apart from the fact that
they are interdependent, our knowledge of the relation-
ships between the various annual ring features is very
limited.
As has already been shown , each species has a range of
anatomical forms and a certain variability within genetically
set limits. In a species such as Pinus palustris, with a very
wide band of latewood, there is little correlation between
widths of earlywood and latewood. In Pinus strobus, with
its narrow band of latewood, there is almost always a
relationship. In conifers from subalpine zones, a relation-
ship between latewood width and maximum density has
been proved. Density diagrams show a pattern typical for
each species.

glee
I')

1.0

0.8
c
g! 0.6
<1l
'C
0.4

0.2

o~----------------------~-------------- ____________ ______________________

~

Cedrus deodara Pinus nigra Pseudotsuga menziesii

Density diagrams for different conifers. Recordings of the density ships between the various features can be estimated from the
structure reveal the variability within one species. The relation- density profile.

When dealing with density fluctuations, it must be

borne in mind that different species growing on the same
sort of site react differently. The relationships between the
characteristics of intra-annual density fluctuations there-
fore vary from site to site.

I)
80

So 60

'" ~ Rledcralp
!i Valdsl800m
- 40
1t--t-----~--~f-~r---~it-,A.--.Hr-----_4----------~~~~~------ ,pS
"t 20
"
!'.
~
0
.>

'"
c;, 60-
s
1\lers~ hw .• ld
'0 40
--r-----t--------,-t~-+t------I_---------_+_---- Valals 2000 III
!l E,1'i r-.;
~ " 12 La
20
~ -t-----t-~~--_tTir_1--t-ti~~-~---t__.._~~~~~~_i~~~~---n llfl
aqeof l'~S
200 years
0

1960 1970

Comparison of frequency of density fluctuations in spruce and shows many fluctuations, spruce hardly any. On the unfavourable
larch on a south-facing site, Riederalp, and a north-facing site, site there are fewer fluctuations in either species.
Aletschwald. On the more favourable south-facing site, larch

Schweingruber, 1980. 119

Climatological-ecological factors subalpine colline
co Id d ry
<ll<ll
Only a few studies on various conifers in the subalpine cc
zones of the Alps have been carried out; they do not Q)(l} =
UU'lo.....c..cV)(/)
££
~V')

permit conclusive findings. 22 ~ ~ ~ ~,~

Q)<ll
c c
lil-lil-~.!!!.!!! ~ ~ 5.5.
- N M "<T ,tHO r-- co Ol

••••••• ~ annual flng Width

00 .
• latewood width

percentage latewood

0 maximum density

minimum density
earlywood width

spruce I arch
swiss
stone pi ne ••••••• •• annual ring width

colline subalpine
o 0 •
• earlywood width

••••••• •• percentage latewood

..
••
Important conifers in the Alps.
0 •• maximum density
In the xerothermic zone the relationships between
• 0

0 o • minimum denSity
features seem to be closer than in the cold-damp sub-
alpine zone. There are no obvious differences between latewood width
species.
Earlywood width is closely related to total ring width,
••••••• earlywood width

since the former comprises 60-80% of the latter. Late- 0 ••••••

•• latewood width
wood width is related to a number of features, particularly 0 •• percen (age Iatewood
the proportion of latewood in the ring. If the latewood is
wide or narrow, its proportion to the total ring width is " 0
o 0 0 •• maximum density
correspondingly high or low. In many cases, especially in
minimum density
trees from dry sites, there is also a close relationship to
maximum density. Because the latewood forms a certain annual ring width
proportion of the ring, it is also correlated to this. o 0 o 0 0 •• annual ring width

Total ring width

0 earlywood width
is the sum of earlywood and latewood widths and there-
fore related to both of these. Being related to latewood
• 0 0 ••
I" latewood width

width, it also correlates with maximum density. •• •• I" percentage latewood

minimum densitY
Percentage latewood
is closely related to latewood width and, especially on dry maximum density
sites, to maximum density. annual ring w idth
earlywood width
Maximum density
often correlates with latewood width and percentage, but
. 00
0 o • latewood width
shows only a vague relationship to total ring width.
•• percentage latewood
Minimum density maximum densi y
does not display any clear correlations to other features.
minimum density

No. Origin and altitude

1.
2.
Aletsch forest, Valais 2400rn
Rigi, northern Prealps 1400rn
•
•
0
60- 64,9
65-69,9
70-74,9 .>
••
Lowest Gleichlaufigkeit with a confidence
75-79.9
80- 84,9
85

3. Lac de Bourget, western Alps 1300rn limit of 95% IS 59.4%

4. Aletsch forest, Valais 2000rn
5. Mt. Cenis, western Alps 1900rn
5. Aletsch forest, Valais 2000 rn
7. Vever, central Alps 1900rn Measure of agreement for individual annual ring features compared
to four others in the same year. The period studied was 1900-1975;
8. Hohtenn, Valais 1200rn the species, five conifers growing in cold, damp, subalpine and
9. Raron, Valais 500rn warm, dry cOlline zones in the Alps.

120
The ageing process
Continual changes in tree ring width and density are 2
usually a sign of ageing . In some trees, however, the mms
ageing process is affected by external ecological factors.
Only after taking average values of many samples of the
same age does it become clear that after a short-term
increase in growth increments in young trees a gradual
decline sets in (Braker, 1981).

Tree ring width curves (average of values for 20 years) for Sessile
oak (Quercus petraea) at various radii. In spite of the irregularities
in the curves, a general tendency for growth to decrease with age
can often be seen (Graybill, 1982). O~~-----
18~O-O----5L
O----l~9~
OO----5~O--

percentage of latewood

g/cm3
1,0
0,9 1 - - - - - - -~~
--~.~~~~""'---v~-~-.:..-;- - - -
0,8 maximum density
0,7
0.30
0.25 j
0,20
minimum density

0 100 200 300 400 500

years

Average age trends in different tree ring features of subfossil, old larches from
Grachen, Valais, Sitzerland. The curves can be described as follows: Percentage of
latewood: straight line. Width: negative exponential function. Maximum density:
Hugershoff function. Schar and Schweingruber 1987.

Density
Minimum density decreases slightly in the early years and
then remains practically constant. Maximum density in-
creases on average during the first two decades and
decreases slightly thereafter; often hardly any trend is
noticeable.
Width
Tree-ring width sometimes increases rapidly in the first
years of a tree's life and then decreases more or less
rapidly over subsequent decades, the main change oc-
curring in earlywood rather than latewood. The age trend
curves for annual ring width and earlywood width are
similar, since these two features are closely related.

121
Relationships between tree ring features in trees on the same site
A. Within one species

The site
The extent of the agreement that is found in tree-ring
features between trees of the same species depends very
largely on which trees are selected for investigation and
on the characteristics of the site. It must be borne in
mind, however, that the characteristics of the site are based
on the conditions that prevail at the time that the samples
are taken and that in fact conditions in a wood or forest,
as for example the light conditions, are constantly chang-
ing. Even the conditions under which individual trees are
growing change in the course of time. For example, in
stands where trees of different ages and heights are
growing together, light conditions in particular vary con-
siderably.
D
o

o
o

TIlls Ire<' has been dommam for only a

~ort lime (release) liS roots are now
growing deep InlO Ihe ground where the
sotl is mOISt and rich In nulnerHS

ThIS ree is su~pressed. There IS Insufficient fighl. The

roots arP In the inorganic lon which IS poor In nutrient!':

This 1r.'C has been domlnam for a conSiderable length

of 11I"f" I!s roots are n Ihe deep, nUlrient-tlch lone.

Different growing conditions at one site due to forest management.

Variations in absolute values

Within one tree

An individual tree displays substantial structural variation.
The anatomical features (see page 103) vary consider-
ably. These features include the width of the annual ring,
early and latewood widths and densities, and the den-
sities within the early and latewood zones.

Tree-ring sequences from a Douglas fir (Pseudotsuga menziesii) from an optimal site in the insublrlc zone (southern Alps). The widths, the
maximum and minimum densities and the fluctuations .n density differ Slightly between the two cores which were both taken at chest
height from the same tree.

122
 ~e gil r
I
q c J

The density distribution within the trunk for a pine in East Prussia.
Densities are high at the base of the trunk. In the upper lighter parts
of the trunk the density patterns are extremely irregular (Trendelen-
burg and Mayer-Wegelin, 1955). (See page 239).

From tree to tree The calculation of Gleichlaufigkeit - the correlation of

single curves to the mean curve for the site - reveals the
The mean values for individual features vary from tree to homogeneity of a site over a given period. Three exam-
tree even when all the samples are taken at the same ples of this for trees from the subalpine zone of the
height from trees growing on uniform sites. Pyrenees are given below.
s:. ?
r'
"0
? t:
';;
<f)
c
'">-
<f)
c OJ OJ
OJ
0) "0 "0
OJ
"0
0 "0
C 0)"0 E E '0
.gJ
0 0 ::J
m
~
0 ::J
:;: m c E trees
E u
I I
0 0
OJ :;: .- .- u ..0
2: E ~ E ::J
c E 2 C u x u E q/cc 3 5 7 8 9 12
'E en
2 4 5
~

E c OJ
OJ
"'----
E 0) ::J
1 10
1 11
'"
r'
OJ E E co E o.-"! c 1.0

I .1
. . . ..
1a 2.38 0.47 2.85 16.6 0.21 0.85 87
. .. . . • .
.
<.>
maxImum denSIty
1b 2.08 0.47 2.55 18.4 0.22 0.86 73 ~
2a
2b
1.43
1.76
048
0.47
1 .91
2.23
254
23.6
0.24
0.25
0.88
0.90
84
93
0.8
... • • • i
N

ci
3a 1.92 0.45 2.37 21.8 0.25 0.80 109
3b 1.97 0.42 2.39 21.4 0.24 0.80 11 3 0.6

4a 109 0.35 1.45 24.9 027 0.93 142

4b 109 0.39 1.49 27.0 0.28 0.92 11 2 0,4
5a 1.30 0.33 1.63 19.3 0.26 0.85 11 3 u

.. .. ...
~
5b 147 033. 1.81 18.6 0.25 0.85 108 minimum density ....a'"
6a 1.40 0.36 1.76 20.8 0.25 0.86 11 6
0.2
• 1. ~. · 1• • II.ci
6b 1.40 0.38 1.78 21.9 0.24 0.84 11 6

Mean values of individual tree-ring features for a small number of Mean peak values for individual cores taken from spruces on sub-
sample cores. alpine sites in the south of Switzerland. (Monte Generoso, 1600 mI.
(a) core 1 (b) core 2
1-3: spruces in southern Finland
4-6: spruces on a timberline site in central Finland.

"I P,C Aubas • . : rna, den. ey to he symbols:

fir
• • nng '!ovid h
Gleichlaufigkelt lowes slgnltlcant confidence

.
blOrdes5a mdX dens values in % value in % limn in %
mourui,lrn
pine 0 •• ·• rmq \Vldlh
o 60- 649 594 95
•••. • . · •
65 699 633 99
cl R abassa
• •• •• •
mol dens
676 999
pine
•• . 0 o . 0 ' Ilq Wldlh • 70 749
• 75 799
• 80 849
Glelchliluligkeit calculations
(a) the maximum densilV and the annual ring width of the individual • 85% and above
curves agree well wit~ the mean curve
(b) the maximum density shows poor agreement, the ring width
good agreement on the whole
(c) the maximum density agrees well on the whole, the ring width
agrees well only in parts

123
Visual agreement of width and density curves

Visual synchronization reveals the degree of homogeneity of the curves for that site we can assume that the tree to
of the site: good agreement from year to year and over which it relates was on occasion subject to factors which
longer periods indicates a homogeneous site . Where a did not influence the growth of the other trees.
particular curve does not have the same pattern as most

Curves with good agreement both in single years and over longer Curves with poor agreement both in single years and over longer
periods. periods. Pointer years (I) enable the curves to be synchronized
despite the poor agreement.

Form and number of intra-annual density fluctuations

The range of forms that density profiles from different
trees in different years may take is extremely wide . Within
a range of amplitudes the density profile may exhibit no
fluctuations at all or the fluctuation may be so great as to
result in the formation of a false ring.
The tree ring structure of the terminal shoots of young
trees is particularly irregular, and where these are sub-
jected to movement (wind , transplanting) during the vege-
tation period, the density fluctuations can scarcely be
distinguished from annual rings . Synchronization is im-
possible in such a case.

Density profiles for spruces (Picea abies) from homogeneous 1961

sites.
1912: 11 trees from a subalpine site in the eastern Alps (Mariazell,
1400 mI.
1961 : 10 trees from the hilly zone of the Swiss Central Plateau
(Horgen, Lake of Zurich, 570 m) (Schweingruber, 1980).

g/ ccm

0.5 1912

0.4
~
.~ 0.3
Q)
"0

0.2

0.1 + - - - - -

xtreme variability in the terminal shoots of spruce. The young

,plants, 1W0ut 1 m high, were moved several times during the
,vegetation period.

124
B. Between different species

The genetically determined characteristics of the tree and
its ability to respond to site factors vary from species to
species. This means that trees of different species grow-
ing on the same site react quite differently to particular
factors. As a result, their rings do not provide the same
information . The differences between particular species
can be seen in the following features of the ring : the
absolute values, the pattern of the width and density
curves, and the number of fluctuations in density per year
and per site.

Differences in absolute values

g/cc
"D E
o E
c
~£
OJ "D OJ
~.~ 0)
co

Roggwil sil ver fir .88 .24 1.66 .83 2.50 34.6 104
Roggwil silver f ir .88 .24 1.37 .67 205 30 .0 117 Mean values for six tree-ring features of trees on two ecologically
Mad iswil sil ver f ir .87 .24 1.29 .46 1.76 27 .8 208 disparate forest stands in the Swiss Central Plateau. Samples were
Roggwi l spruce .89 .2 5 17.5 .45 2.20 23.4 135 taken from 12 trees on each of the sites. The density values for the
two sites are only slightly different. The differences between the
Mad iswil sp ruce .91 .27 1.40 .43 1.84 25.2 170 sites are reflected in the width values (Lenz, unpublished).

Frequency of intra-annual density fluctuations

glee
It is extremely difficult to synchronize different species
1.0
using density fluctuations since different species react
differently to environmental changes, depending on the
site.

0.6
Mean curves of tree-ring width for trees from a uniform stand in .~.'./.
• SWISS
the subalpine zone of the Alps. (Aletsch forest, Valais, Switzerland,
2000 m). 1940 50 stone pine 60 70

Characteristic curve patterns

Particular species have characteristic curve patterns.

Although the purpose of averaging single curves is to
obtain climatological information from the mean curve,
.~
5i-
1001- ~ ' .
80 .•...
60- -
J'j\j\._.IV\;" \. .
' . \.
'. . " fir

spruce
Roggwll
500m
n.lil 24 s.flr
this process also reveals the typical pattern for the - >-
~.~ 40
n=24spr
tree age
species. '"
~ ..g
!:
20 ._ _...!.....---Y \00 years
<.> !:
~;;; 0 .......
e>c:

- '"
!: 0
.;;;'" 80' Madlsw.1
~ ~ 60 · ',~•• / ' 700 m
0'):>
. n = 24 s.f,r
Comparison of percentage of firs (Abies alba) and spruces (Picea ~ 40· n = 24 spr
abies) with density fluctuations from two ecologically homogene-
ous areas in the Swiss Central Plateau at 500-700 m. The silver fir
displays more fluctuations than the spruce for both the sites
~
,f
a.
20f
o· I
I tree age
250 years

(Schweingruber, 1980). 1940 1950 1960 1970

Schweingruber, 1980. 125

The relationship between individual tree-ring features in trees on similar and
on differing sites

The grouping of similar sites together is straightforward
using floristic and phytosociological criteria. It is, on the
other hand, difficult to relate a site to particular site
factors, and even more difficult to account for the fact that
certain species are found on a particular site or that the
cambium and the other living cells react in a particular
way to a particular combination of factors. The first step
towards establishing the different causal relations that are
at work here is to determine under which sets of climatic
or ecological conditions the cambium and the other living
cells behave similarly. The reaction of these cells can be
quantified through the measurement of cell-wall width and
thickness.

The mean agreement between curves

The mean agreement between curves is used to establish
which areas or groups of sites yield historical or fossil
wood which can be grouped together for the purposes
of dendrochronological analysis. The similarities between
different sites are given by the Gleichlaufigkeit, which
expresses the correlation between mean curves over a
particular period. The similarity expressed relates more to
the types of reaction exhibited by the trees on the sites
than to the botanical compOSition of the sites. Geogra-
phic-climatic boundaries are also made clear. A number of
examples which relate to the subalpine zone of the Alps
are given below.

Same species in the subalpine zone

MarialCli
Bever 1 .. .. . .. • ~~.~_ ••o~o~
• °
:-.-_0
R~i
Chasscral 1510 m IJl~~~~~~~~~~~~~ I •.•.•.•.• , •.•0 .. . .. . .. . ."-0 .0.0 . • . • 0
° 00 . 00 . 0 .
maximum density minimum density
Mariazell ~",-o 0 (' °•
Bever I •• o~ .
_.~ o . o 0 0 0 . 0
Rigi I io. •. o•
. ••_. 10
Chasscral 1!>10 m 1-+--+::-F+-=+"t-l-=-t=77;~L&.~~-::-"7'--:--I
rr .. _ 0
o • • • _ • •

0
'\., 0 • •
•• " . I
annual ring width edrlywood w idth
Maria?cll
Bever
R igi
Chasseral 1510 m 1L-L..:...L.:-'--L.L:..I-.L..:...l.::.J...:...L::Jc.::..c=.L..:..I.~..J.:=J-L'"
latewood width percentage latewood

Key to the symbols: Gleichlaufigkeit lowest significant confidence

values value in % limit (In °0)
060 649 . 75 799 594 <)5
. 65 699 . 80 849 633 99
. 70 749 • 85% and over 676 999

Gleichliiufigkelt values for mean curves of spruces from subalpine sites in the Alps and the Carpathians. Values are given for six important
tree-ring features for the period 1899 to 1975.

The extent of agreement varies from site to site. In all
cases the similarity is greatest for the maximum density,
less for the latewood width and the percentage of late-
wood, and least for the minimum density, the earlywood
width and the total ring width. The east-west boundaries
seem to be most clearly reflected in the maximum den-
sity. It is not clear why practically all the features for the
Rigi site display good agreement with those of other sites,
while for the Chasseral site only the maximum density
does so.

126
Trees of the same species in dry regions
As in the subalpine zone there are differences in the
degree of similarity between the different tree ring fea-
tures. Maximum density displays the closest similarity
between sites. Total ring width, latewood width and per- Hohtenn
••••
l"- 0 ,"- •• • •
.1".. •
• .'" .
I"- . 0 o •

o."•
••
.~
••
Raron

• "
centage latewood agree well for sites that are close

."
Plynwald I".. • 0 o • • 0
together. Earlywood widths from different sites display
Krauchtha l .~ • .o f'..0 0 0 0

'" '"
little similarity. In contrast to trees on cool-humid sites Oierslordt •• 00 r;::
trees on dry sites show clear similarities in minimum IN·Germanyl
maximum densi ty minimum density ring width
density.

'" "• •••• ••• •

'.'"
Hohtenn 00 0
• 0.
Raron of'..0

'"
Plynwald 0 0 l"- • 0
•
o 0
Krauchthal 0 . f'.. f'.. • 0
f'..
'"
Oierslordt 0

e.rlywood width latewood width percentage latewood

Gleichlaufigkeit values for the period 1899 to 1975 for pines on dry
sites in Switzerland and Germany (Schweingruber et al., 1979).

Trees of the same species at lower elevations

Here all the tree-ring features display some agreement
with those for other sites.

Different species in different zones

Even within climatic or vegetation zones the relationship
Hi
u· ~ nn~bnn~nh ! I !
el 6 iCr.
!
.e..

E E
of individual features to each other can vary considerably.
~~
• E
~ 8 2 ~
The figure shows that in areas with high precipitation
Jti
'1'2 ~~~E~~~ ~~j~

..
maximum densities are similar at central and high eleva- ~ f .c
-'I ;~t~~~ll~~ ~~~<: ~ & S ~!
~5~ ~;£~~~~~c§C§l:~~~~
~
tions, but that they exhibit no relationship to those from iii :;
::z:cra::~

edaphically dry sites or areas with low precipitation.

Minimum density behaves quite differently: the curves for
u"",Jt"t'
~ :L~l'. ' ~
~. :1:1~ ••••·f*'M
••• •• •,
i 1n1"'_~'I.,

.
Unh'U'nlll"k"'" • 0
0 ••

ff~ '-:1t ..
\

moist sites at medium elevations agree with those for dry HOfQt'n } 11' 0 ~.+ 0 0 .0 o 0
~.~ . 0 0 0 0 • 0 eo m.I~IfIWt'I~"'h·

. . :*-+-+
Uf'lirreonlll!tdl!'fl
sites, but do not agree with those for sites at high j.... o ~ •

elevations. Features which reflect cambial growth, i.e.

widths, are not related to those in other zones. There are
HQoflJC'11
Unll'ff'fll'rlrJio"
U~~ II
+
111
~

+
.

I ,, (-• \
I •
0
+
j-+-
00
",.,. ... -.11"

~ rE-p: ~
Hnrl)l'fl
r:, o
• t t I
no similarities between total ring widths in the colline Unt~~lltekJrn • 0
.~ t ,0 0
uolvwood w.drll

- ~. ~ l .. .. i .
zone, the dry zone or the cool humid subalpine zone. Hor~" H- +--.-
0

:1 Of
.... ~w

m..
l\'"
UI1111"II!n1 Irld""
+ t

....
U/11f'rl('fl,lrldrn
I~. : f 1 : . t ~t ~+- -10 io!- Pl!rctr<I~ ta'--ood

"" ....
~~I
1II.II~'p''''' /(IoI1of ,...
~,

'"
I~ ., ....
,. tl. 9 • f~ 99
99 . '5 t'49 . 80il·~9
999 • 1.:.9 . ilrtJ.

Gleichlaufigkeit values for a number of tree-ring features in different

species from the moderately humid coUine zone, the subalpine
zone with high precipitation and the dry hilly zone of the Alps
(Schweingruber etal., 1979).

Schweingruber, 1979. 127

Spatial representation of the agreement

As can be seen in the diagram, similarities, expressed in
percentages of agreement (Gleichlaufigkeit percentages),
vary considerably from one site to another. Analysis of the
site homogenous sampling network in Europe shows that
in certain areas the curves are very similar. These groups
make up the core zones. From this point on there are
fewer and fewer similarities. When comparing maximum
density and annual ring width of various species of coni-
fers on European sites with cool , damp summers, it can
be shown that the maximum densities on average have
about a 10% higher Gleichlaufigkeit value than the tree
ring widths . The statistically significant areas of similarity
are greater for maximum density than for tree ring widths .
With both measurements the area of similarity drops as
one goes from north to south. In northern Scandinavia the
chronologies of maximum density can be synchronized at
about 1400 km , those for tree ring width within 800 km . In
southern Italy, however, the respective values are only
about 700 km and 200 km.

Lappland Southern Carpathian

?:
,..'
.;;;
c:
'"
"0
E
~

S
X
<0
~

-5
"0
.~

c"
c:
a:

Some examples of areas of similarity in curves showing maximum AA Abies alba PC Pinus cembra
density and tree ring width for conifer ring sequences from 1970 to AB Abies borisii-regis PL Pinus leucodermis
about 1980 on European sites with cool, damp summer conditions. AC Abies cephalonica PN Pinus nigra
The outer line of the lightly hatched area represents a 'Gleichlaufig- AP Abies pinsapo PP Pinus peuce
keits' value of 58.0%-64.9% (95-99.9 probability). The figure in the PA Picea abies PS Pinus sylvestris
centre of the unshaded area is the average agreement (Gleichlaufig- (Schweingruber,1985)
keits value) for the region (core zone). A dotted line indicates that
the boundary is ill-defined. The following abbreviations have been
used for the trees studied:

128
Agreement over a given year
Both the calculations and the comparisons of different - within a particular climatic zone.
curves indicate that the degree of similarity between the
curves for any given annual ring parameter decreases in This trend is not always clear in individual cases and it is
the following order: by no means always possible to decide on the basis of
the curves alone whether two cores originate from the
within an individual tree same tree or from two different trees. Two examples are
- within a given site given below.

Maximum densities

~ J\t.,...\. .\. ~ 1\
'"M,:N <~\V j\ !.v.v.....'vt(·~·\
'".v.~~ V 'oJl.: \b •.\ _.
• " • \1
•• e o ·• • •
2
...g
I~\ 7,""•.•.! ......... I\j''/ \

,.1\/\,.,...,\/.. 'J' ,.' \ . .'1\

~
,\...'!\I' .., · /\ /""'. ...
...,
.. . . . .
/\/

... + + 1'-
... .........
+ + 4x
..... 3x
•
I I

• +
...........
I I I

...
I I I t

.. ... ... ... ...

......... + +
I I I

...
..................
I

.. I I I I I I I

+ •
I

.. + ....
I I

.
I I

+ ••
I

++++++
I I I

. I agreement
I

4x
3x
I I I

~ · 1
to good poor oo good poor
<0
co agreement Cl
agreement

Agreement between single curves of maximum density for two
cores taken from the same tree. Phases of good agreement and
poor agreement alternate with each other.
The samples were taken from two spruces on a uniform timberline
Agreement between mean curves of maximum density for four
widely-scattered sites in Central Scandinavia. The Samples were
taken from spruces growing on similar sites near the timberline
between the Atlantic (1) and the Gulf of Bothnia (4). As with the
single curves there are phases of good agreement and poor agree-
ment.

The number of intra-annual density fluctuations

"0
Intra-annual density fluctuations reflect ecological factors. o
o
Their formation depends on the quality of the site and on ~

the cambial activity of the tree. In similar sites, e.g. in '"

-:;
Zol li ker .
the Swiss Central Plateau, density fluctuations occur in .'= berg
greater or smaller numbers depending on the tree, the '"c: Ho'gcn
oS?
site and the year. The more similar the sites the closer is .~
the number of fluctuations per year (e.g. Zollikerberg and '">
Horgen). Fluctuations occur in ecologically dissimilar sites 2!"
in
only in years of climatic extremes. c:
Q)
"0
"0
~
E
:::l
c:
1! f--;;:=-- --1,,£.>,<;/-----11+- -
f-

1950 1960 1970

Curves showing the number of density fluctuations per year in

spruces in the Swiss Central Plateau at an altitude of between 500
and 700 metres. Zollikerberg and Horgen are on the Lake of Zurich,
Madiswil and Roggiwil are in the centre of the plateau and Chaneaz
is in the west of Switzerland.
The Zollikerberg and Horgen sites are very similar. The other sites
reflect climatic variations in a similar fashion only in particular years,
e.g. 1959/60/61 and 1973/4 (Schweingruber, 1980).

129
Agreement over longer periods of time

Maximum densities
In order that climatic patterns which span longer periods
of time may be represented, the annual value chronol-
ogies must be smoothed. In this case 13 year weighted
running means are used. Climatic variations can be
related to each other using samples from similar sites but
different geographical regions.
From the chronologies given below it can be seen that:
all the curves display the same characteristics. They
are all relatively uneven, sometimes with high ampli-
tude, sometimes with low; in places displaying low
values and in other places high ones. The frequency
over the last 1000 years differs considerably. Short-
term fluctuations tend to last about 10 years, medium-
term ones about 50 and long-term ones about 200.
the change of pattern that is perceptible at the present
time is affecting practically the whole of the northern
hemisphere. It displays a far greater regional spread
than all previous ones since 1700.
both the temperature and the density series for neigh-
boring sites differ from one another. Since such varia-
tions often cannot be satisfactorily explained, it is pos-
sible to effect a reconstruction of climatic conditions
only where the information available is such that the
characteristic features of the region can be identified.
there are clear correlations between variations in the
summer temperatures (mean values July to Septem-
ber) and variations in the maximum density.
most of the decennial variations are, like the annual
ones, limited to particular regions.
It is reasonable to suppose that measurements for the
current period will show the occurrence of a larger de-
crease in density than would be occasioned solely by
temperature.

~~-L~--L-4--L~L-~-L-+--L-~~~--r-~~~~--~--~~~>
Coullns. Skye. W.Scotland ~
Inverey. E.Scotland ~
Braemar. E.Scotland Co
Kew. S England §
~ ?:
'iii lauenen. N AI ps 'on
.Co
I ~Wr-J~--~~~~--~~~~~~~~~~~=f~~----~cr,o~r~tl~na~.~S~E'.AITI~PS~---~
0.
~
14
13
~~~~~~~~~~~~~~~~~~~~~JL-f~~~~~M~u~n~lc~
h __~_______
Hohenpelssenberg ~
E
'"c:c: ?:
on
0.
:>'"
c:
~
19
'"
'0
Co
~
;;; E
f-
17 ~
0
x ?:
.;;;
~
:; ~ ~~~~~~~~~rJ~~~~~~~~~~~rF~~~~~~~~~~~~~
'"c:
""c
'0
15
---+~~~~®.-,~~~~~~~-4~~~~~~~~~~~-- ~
14
'"
u
Ul
2!

1800 1850 1900 1950

Relationship between neighbouring chronologies of maximum density for conifers on boreal and subalpine sites (heavily shaded band) and
average summer temperatures (July to September) in Europe. The average values were smoothed with 13 year running means.

1-------------+-------------+---=---------t-------------L---\r7"'l;:'7'------"'o"'-''4--''-.L..>o,'''l'':r Cent ral SI bena

1700 1150 1800 1850 1900 1950

Chronologies showing the maximum density for conifers on boreal and alpine sites in the northern hemisphere. Statistical analysis as above.

130
 subalpine coli lne
The agreement between the absolute values ;:; c:
~
~0
~
l':' ii'"
"0

.
:>
The mean values for tree-ring features do not by them-
0
..,:::l '" Q. 0
'~"
D
~
.., a 'c "§ -::; C
..
c c: ~
CD
c:
selves enable different sites within a vegetation zone to '"
~
~ .: ':>"
c .:; ~
'"
J::
~ U
III Q; "0
'u" '0~" ~ Q;
''""
(;J
be distinguished from one another. Nor do they suffice 0;is ~ g,
~ D
:>
.~

;;; '"'" ~ u C
"0 :)
2'" '"
J::
,~ <i u CD a: ..J Vi Ii" ::; u a: ..J Z J: N
for the differentiation in individual cases betwe8n woods
from different vegetation zones. It can, however, be said g/cm' maximum denSity

..
that in general tree-ring width and maximum density

• . ••
decrease with increasing elevation while minimum density
1.0
• • u

increases. • u
ci.
•
•
M
• N .

•
0.8 -;.---__<
>-----------------'---
o

"0
o -0
o 0
!: 0
~E
iB E
h
.§ E
0.6-

Chasseral, 1510 m .76 .27 1.02 38 1.41 238 137 12

Chasserat, 1400 m .78 26 1.34 27 1.62 18.5 131 12

0,4 :::::=---:;===E===:
• • • :;:=•===;:====,
Rlgi, 1600 m .73 .24 1.67 30 1.98 16.3 136 12 minimum density u
Rlgl, 1400 m 79 .26 1.41 30 1.71 18.6 171 12
Aletschwald, 1900 m 72 27 85 17 103 17.9 211 10 ~
Rlederalp, 1800 m
Arosa South exposure
.83
.72
.28
28
1.13
67
29
16
1.43
84
23.0
20.8
95
262
12
6
• • • • • • •
Arosa, North exposure 65 25 99 14 1.13 13.0 288 10
~~==~-=====~~==============~~o

Mean values for tree-ring parameters in spruces (Picea abies) from Mean values for maximum and minimum density in spruces from 14
the Jura and the subalpine zone of the Alps (Schweingrubers et al., subalpine and 3 colline sites in the Alps.
1979).

The agreement between the curves in relation to distance

According to theoretical models the similarity between
curves of the same type (width and density curves)
decreases with increasing distance. And indeed Fritts
(1976) and Hollstein have provided a considerable
amount of data which supports this. The former examined
conifer wood in the south-west of North America and the
latter oaks from the west and south-west of Central
Europe. Where regions are composed of distinct topo-
graphical units it appears that very clear boundaries exist.
The Gleichlaufigkeit for maximum density curves of coni-
fers in the subalpine zones would appear to decrease
sharply. Baillie (1982) found that in Western Europe with
its temperate climate, large trees displayed significant
Gleichlaufigkeit. Correlations occur which span surpris-
ingly large distances: they have been found between the
ring widths for woods on dry sites in the Near East and
the maximum densities for conifers from the cold-moist
timberline regions of Scotland and the Alps.

Glelchl<iufig eit values

OOr------------------------------------ .
Distance and Similarity for mean curves of oak

80

. . . . '.
.' t'

'.. ', '"

70, :--~ • . ••. ~.: •• •

eo; ' • ", i':~~y:, · 76

72 . 62 HIl'll Ou I1l
lOWfl Si);Of1IV

50
-
40L---~-~~--~~~~~--~~-~~~~--~
-
o 100 200 300 400 500 600 700 km

The relationship between geographical distance between sites and
the Gleichliiufigkeit for ring width curves for oaks from a number of
sites in Central Europe over the last 2,700 years. The mean similarity
between the curves (Gleichlaufigkeit 50%) decreases exponentially
with increasing distance (Hollstein, 1980).
Gleichliiufigkeit values for oak chronologies for individual sites and
for the aggregated West European chronology (31 chronologies
covering 1850 to 1959). In 62 out of 100 cases tile climate appears
to influence growth similarly in the tree as a whole. There is no
perceptible relationship to distance (Baillie, 1982).

Hollstein, 1980; Fritts, 1976; Baillie, 1982. 131

The effect of climate on the tree ring

Tree-ring features

The following tree-ring parameters are straightforward

to measure and analyse using radiodensitometric tech-
niques.

jrnaxlmum density

I density variation

Important characteristics in tree rings of conifers.

Climatic variables

For practical reasons the variables most suitable for den- course of ecological investigations. Since it is usual to use
drochronological analysis are air temperature and precipi- mean values covering one or more months it has become
tation . Both are obtainable as monthly averages from common practice among dendrochronologists to refer to
weather stations and are also easy to measure in the these 'weather periods' as 'climate'.

Methods for the comparison of tree-ring features with meteorological measurements

Comparison Advantages Disadvantages

ecological/biological measurements Permits determination of dynamics Cell wall growth cannot be

with dendrometers of radial stem growth. Causes of determined . Practical considerations
changes in growth can to a large limit measurements to a few trees
extent be explained. and places and short periods.

meteorological/biological Comparisons covering several Dynamics of growth cannot be

data from microsections and decades are possible. Causes of determined. The statistical process
official X-ray films of annual changes in growth behaviour can is a "statistical comparison" . All
stations ring sequences. often be assessed with some vegetation periods are assumed to
certainty. The method is easy to use. begin and end at the same time.

biological with biological meas-
urements, microsections and X-ray
films of annual ring sequences.
Climatic-ecological regions of validity Causes of changes in growth
for sets of similar ring sequences can behaviour can only be conjectured.
be established. Results are based on
comparison of sequences hundreds
of years long, and can be presented
cartographically. The method is easy
to apply.

132
The pattern of radial growth in relation to weather conditions

In the course of large-scale ecological-physiological

_ _ _ _~pe_"_od_of....:g:.-.ro_w_t_h_ __ Stalde
studies it has proved possible to relate the dynamics of
- - -- - - - - - -- - Barechumma
radial growth to the weather and to eco-physiological Engiloch
processes. The evaluation of densitometric curves has Bodme
revealed the following: 100 .,..---.........~----~~......,,---~
'!.
the pattern of radial growth is determined by both
internal and external factors. In early summer the
earlywood grows rapidly and the life of the cells is
short (5-20 days). In late summer the rate of growth 80
is greatly reduced and the formation of the latewood
takes several months. The last-formed latewood cells
have a relatively long life-span (2-3 months).
the length of the vegetation period and the intensity of
the growth are determined by external factors. Kern
(1960) found that the period of ring growth for spruces
at 300 metres above sea-level began on 6th May and
spruce Barechumma
ended about the end of September. Growth at 1300 m larch Bdrechumma 11800 S)
on the other hand began a month later, on 5th June, larch Stalde 11400 TI
and finished in the middle of August. larch Bodme 11800 N I
it is not merely the amount of radial growth and in- larch Eng,'och
11800TI
10 scots prne Eng,loch
crease in density that vary with the climatic conditions,
but also the amount of time that this growth takes.

Radial growth plotted against phenological data. (Percentage in-

10 15 n Z'J 5 '2 E 15 2 '0'7 21 31 B 15 21 ltl I 10 IJ 25 1 9 ,
crease related to total annual growth). May June July August September OCI. 1976
Spruces, larches and Swiss mountain pine from the subalpine zone
of the Central Alps. (The southern end of the Simplon Pass, Valais, Stalde
- - -- - - -- - -- - - Barechumma
SWitzerland). - - - - - - - - - -- _ ._- Engib:f,
In 1976, a year favourable to growth, radial growth in trees at high - - -- - -- - - - - - - - Boorne
altitudes took place over a period of about three months. Ring width larch. phenological condilion
growth, i.e. the development of the earlywood occurred in the seven
or eight weeks before the middle of July. The buds opened directly .. buds opening - - - needles fully developed
before and the needles directly after the period of radial growth - new needles VISible - - leaves yellowrng
(Mueller, 1981).

The tree ring in relation to climate

preVIous
temperature precipitation
The relationships between the individual climatic para-
meters, expressed as mean values, can be calculated
using response functions (see page 88). Factors limiting
growth appear as values above the middle lines. Here we
can see, for example, that tree growth was limited by
precipitation in Indiana in June.
On dry sites in the Mediterranean growth is fairly
clearly determined by precipitation in nearly all months. At
the northern timberline on the other hand there appears x
to be adequate moisture at all times of the year, and the Q)
-g
rate of growth is determined by temperatures in June and
sometimes in May.
It seems likely that this method could be applied to the
analysis of arid regions with a rugged topography to
obtain climatic information over several months of the
year from quite different sites, i.e. from sites at low or
high altitudes and with humid or arid conditions. However
the most recent work done in this area indicates that the
range of applicability of these methods still needs to be -0.4 CLLLJ....!-LL.LL-Ll-Ll-1...-LL.1...L.Ll.-L.J,....L.,L,..LJ,..LJ,..LJLLL.J.......J
determined. (Hughes et a/., 1980, p. 12). J A 0 Dn F A J . A;.,..;;O;...,;D;;.._F...;,.:.,;A;....;;.J__ 1 2 3
.;;:,
J...;",
previOUS current prevIous current prevIous
year year year year year
month month
Response functions for ring widths in trees from fundamentally
different climatic zones. The main factors influencing growth are
indicated (Fritts, 1976).

133
Different tree-ring features in relation to climate

Every tree-ring feature can be related to climate. As is

shown in the following example the different features Index · C
contain different types of climatic information . It should 1.19-18
therefore be possible to obtain information about climatic
1 .09- 16 .4
conditions over several months from a relatively small
number of tree-ring sequences covering well-chosen
0.89-15
sites.
1940 1950 1960
Comparison between an optimally smoothed curve of maximum - - - maximum denSity
densities for Xanten and the mean temperature curves for July to ------- temperature
September for Bocholt, West Germany.

Density and width values

The example chosen relates to two low-lying sites in Index ·c
1.099- 10
northern Germany: Xanten and Diersfordt. With average
annual precipitation of 90 cm and an average temperature, 1.000- 7
including summer maxima and relatively mild winters, 0.900- 4
these sites lie in the transition zone between the central
European and west European climatic regions. Pine
stands were planted in Diersfordt in 1906 and in Xanten in - - - - rmnlmumdenSl ty
1930, in both cases on acid sand . Phytosociological - - - - - - - temperature
analyses indicate that the ground is moister at Xanten .
It was clear from the auto-correlations, which were
Comparison between the mean curve of minimum densities for
calculated for several years , that the conditions in the Diersfordt and the mean curve of March to April temperatures for
years which precede the formation of a ring have a Kleve, West Germany.
greater effect on widths than on density and percel'ltage
latewood . The response functions - a measure of the
influence of climate on the tree ring - indicated that the
annual ring and the meteorological curves for the two
sites were similar.
Index ~ ~
:..J t: ,
11 ~ ~ \
10 200·\'
150
0.9 100
50
0.8

Comparison between the mean curve of latewood width as a per- 1940 1950 1960 1970
centage of the total ring width for Diersfordt and the curve of total
precipitation in June and July for Kleve, West Germany.

"}i
Xanlen
minimum density N NN N N N N N N
maximum density ~.
M M M M M M M ~

percentage latewood A A A
~O
A '">-
Q)

latewood width 0 0 <D 0 0 0 42 Cil

earlywood width + no
x x x
0
x
0
C

. 1O}
annual ring width 00 0
annual ring w idth I
- •• •• •• a.

• !
'0
• •• • •
earlywood width I( I(
1(1( I( 6
0 • 8 Cil
latewood width .D
percentage latewood ~~ ~~ ~ ~ ~ ~~ 46 E
maximum density m m m ,~ In :J
c
minimum density n n n n n n

1940 1950 1960 1970

Positive and negative trend-intervals for six tree-ring features (Xanten). When more than 80% of the 24 single curves in an interval exhibit
the same trend, the year they relate to is said to be a pointer year. (The limit of 80% is arbitrary)

134
Intra-annual density fluctuations
In the Swiss Central Plateau intra-annual density fluctua-
tions seem to be a response to temperature changes.
Fluctuations in density occur most commonly in the
wide annual rings of trees on humid, fairly warm sites
which have average July/August temperatures of 14°C to
21°C. They are not found in trees at high elevations where
the average temperature for these months is less than
12°C.
On the basis of observations made I think that it is
reasonable to say that the development of density fluctua-
tions for trees in temperate zones may be traced as
follows:
low
16
·4 - -Y..-\-·#- - --/-;+--1t-:.+--H4-'
40 50
where there is a good supply of water available cam-
bial activity is not limited. As a result there are un-
differentiated cells present in the trunk at all stages of
the vegetation period. This in turn means that short-
term climatic events can exercise a considerable
influence on the density of the wide annual rings.
If we look at the fact that trees on warm sites exhibit quite
different reactions to those on cold sites it seems clear
that:
trees on warm sites can at any time mobilize available
carbohydrates and build them into the cell walls. Trees
on cold sites on the other hand do not have these
spare carbohydrates available. Almost all the photo-
synthetic product is deposited in the storage organs.
)emperalUre
~ __ ~~ _____________ 40
~
~
--~~r---~~~~~~~~---,~--30~
o
-.:...i~=l.1...'----+~I--~r<---"'--- 20 ! ::l
c:
60 1970
The relationship between the average temperature for July and
August for Berne, the number of days with below-average tempera-
tures (mean for the period 1901 to 1960: July 17.4°C, August 16.7°C)
and the percentage of annual rings with density fluctuations. (Sam-
ples taken from Roggwil, Swiss Central Plateau) (Schweingruber,
1980).
Pines from Northern Germany displaying one or more density fluc-
tuations per year. The youngest complete ring was formed in 1976.
135
The effect of climate and site on tree rings

As we made clear in the previous chapter, practically
every tree-ring series can be used to obtain climatic
information . It is nonetheless clear that, particularly in the
case of alpine zones and their different topological areas,
chronologies can display less and less similarity with each
other as differences in site altitude increases. Fritts et at.
(1965) compared relatively long chronologies from vari-
ous dry sites in the north of Arizona. Chronologies from
higher, moister sites included tree rings of average width
and low mean sensitivity and showed poor correlation
with each other. Chronologies from trees in the lower
forest border areas (1900 m) showed more narrow tree
rings and higher mean sensitivity and correlated well with
each other.
LaMarche (1974) developed on this idea and recon-
structed the decennial temperature and precipitation of
the last 1000 years on the basis of chronologies from
Bristlecone pines (Pinus longaeva).

Datable series
---+
, .. If J

"complacent" series "senslllve" series ~: m.Jl ..tnQm;jIy

I', nQft!IC d

Smoothed Bristlecone pine chronologies from lower timberline

(dotted line) compared with chronologies of the same species but
from upper timberline (solid line). The climatic characteristics for
various times are indicated by vertical lines (LaMarche 1974).

low

o~--~------~--~--
forest Interior
decreasing effeclIve preClpllation
increaSing variability In annual precIpItation The relationship between statistics of time series of tree rings
along changing altitude profiles, beginning at the lower, arid tim-
more days where moisture is limiting to processes In tree berline (Fritts et a/., 1965).
)

The climatological analysis of individual years

Kienast (1985) radiodensitometrically analysed samples

taken from trees grouped together according to sites
along different altitude profiles in Europe and North
America. The mean curves for maximum density and
annual ring width which had been aggregated from
indexed single curves were compared with each other
year-by-year with respect to homogeneous geographical
sites using dendroecograms.

moisture
content

Development of ecograms. Chronologies from both dry and moist dry .. moist
sites are compared with those for high and low altitudes (Kienast,
1985). moisture content

136 Fritts et al., 1965; LaMarche, 1974; Kienast, 1985.

 Cold·mOlst years Warm-drv years
maxlmu.~ cJ~n.II'r 'urd rrd'W Jm 1,'0 IV Color do fnf)X unum ch ns ty Jura maXim lIP dunsllV Colorado
e 9 1912 n 1 Ir). ell V 1941 P g 1921 e 9 1945
n:rl j

one year ~ummary one year summary

Cold·dry years
latewood width Colorado
e 9 1971 .n 18 from 81 V Dendrochronological diagrams for extreme years:
(a) In cold-moist years temperature is a limiting factor at
told the upper timberline. At the lower timberline conditions are
optimal.
(b) In warm-dry years precipitation is a limiting factor at the
lower timberline. Conditions are optimal at the upper timber-
line.
(c) In cold-dry years temperature is a limiting factor at the upper
and precipitation at the lower timberline. Conditions are
optimal at central altitudes (Kienast, 1985).
one year summary one year summary

The assignment of diagrams for extreme years to one of max density ring width
the three main types is unproblematic. However, the
effect of climatic factors on the formation of the average high large
annual ring is by no means straightforward. That can be
seen in the figure that follows. Of the several areas
investigated, only about half of the resulting diag'rams
could be allocated to one of the main types. ~ medium medium

D low small

Jura Colorado
1920 1921 1922 1923 1966 1967 1968 1969

warm·moist

Ecogram sequences for consecutive years for the Swiss Jura
and for the Front Range in Colorado. Climate conditions shown
are based on meteorological data.
Different climatic conditions can have quite different effects on
maximum density and ring width. This is in turn reflected in the
diagrams relating to these features. A good example of this is
the year 1923. In this year there was a large increase in the
maximum density for mOist sites at high altitudes, but only a
small increase in the annual ring width. Dendrochronological
diagrams for maximum density and ring width are similar only for
extreme years. (Kienast, 1985).

137
The different tree-ring features (maximum densities and
tree-ring width) of a chronology call display quite different
climatic reactions for the same year. It is only in years in
which extreme climatic conditions are experienced that
each of the ring features is influenced in a similar fashion.
Maximum density may for instance be limited by low
temperatures while the width of the same ring is limited
by lack of rain. It is also often difficult to obtain a clear
picture of the factors which influence growth in the case
of those climatic zones where the summer dry season
lasts several month because growth may start and stop
several times in influencing the results of maximum
density analysis. Only the relatively moist timberline sites
are thought to display consistent response to tempera-
ture. Most of the low elevation - sites in the temperate
zones may be affected at one time by temperature and
another by precipitation. Samples from lower timberline
are mostly sensitive to dryness.
Thus it is seen that comparison of results that are
obtained from sites that are uniformly high or low should
be attempted.

The climatological analysis of tree-ring series Evaluating the response functions

Climatic information is in many instances obtained from

tree-ring series through the use of response functions.
sy m·
It is, however, generally accepted that the amount of in- 0.4

••
bois
formation that can be obtained from any single response III
function is limited. Kienast (1985) has categorized the 0.2
II
positive and negative influences which occur in the re- '"
sponse functions and are related to site analysis.
III
u
:.0 0
I •
c: I 0

II 0

••
0.2
III 0
- 04
0
• • 0

SON 0 J F M A M J J A SON 0 month

Only those monthly values for wicht their respective confidence

interval (indicated by vertical bars) does not include the zero line are
assessed. Those wich are far above the line are given bold symbols,
while those close to the mean line are marked with smaller symbols.
For month prior to the year of growth, characteristically bold
monthly patterns are suppressed in that the significant values are
lowered by one step as for example in November of the previous
year (Kienast, 1985).

Colorado. Front Range/ Meteo station Denver

moist 13320 m 0 /
•
• ••f'". 1\ •
• •
~

."
0 o -- .
., .....
-1-
. "'
0 01 0

•
47
39

•• •
E dry 13400 m 0 0 0 o / 0 1
:::l >
• • • •
"
I 57
.• "•
E .!:: mo ist 12500 m 0 0
~ 0 0

.
'"
.-
x c dry 12650 m I l>- I -1<', \ • • • • 01 • 29
E "O moist 1900 m
•0 •
'" III
\ 0 0 0 0
• • r'\. . 1-- r--... • 57
d ry 1950 m 0 0 r-- 0 0 _v I'-. • • • • • I 50

moist 3320m 0 • • 51

--- \ - -
0 0

~ dry 13400 m I'- 0

o "
0 v. • 0
• 36

1\ " - - . 1\.
• •
"0
.~ mOist 12500 m 0 0 0 0 37
" 0 " V;
o
dry ~650 m
,-
\I .• - n I\. •- -
Ol
c II / 0
\ • 42
·c moist 1900 m 0 0 \ .1'-- • - I'-... 49
"
-
\0 0 0 0

dry 1950 m 0

month S 0 N
0

o J
- 0 0

F M A M J J A S 0 N 0 S 0 N
0 0

o
• • •
J F M A M J J A S 0 N 0 p l p2
49

temper-atu re preci pitat ion

The relationship between response function values altitude and the influence of the previous two years (p 1 and p 2) in the two
moisture for individual site chronologies of maximum density and colu~ns on the ri~ht. T.he va~iance explained for each chronology
tree-ring width. The relationship to temperature is shown on the left and Its corresponding climate IS stated on the extreme right (Kienast
side of the figure, to precipitation on the right side of the figure, and et al., 1985).

138 Lingg, 1986; Kienast, 1985.

The suitability of different types of sites for dendrochronological analysis

From the results of the evaluation analysis (weighting
tables) for response functions the following conclusions
can be drawn : Maximum amounts of climatic information
obtainable from maximum density series comes from
samples obtained from cold-moist zones, i.e. the sub-
alpine/ alpine and subarctic/ boreal zones. In zones with
adequate summer rainfall it is August and September
temperature that is reflected in the grown patterns, while
for those which experience a summer dry period (e.g.
Mediterranean areas) it is the temperature in early sum-
mer which is decisive. In most cases it is possible to draw
some conclusions - valid only for particular sites - about
precipitation in winter, spring and early summer months.
In May and June (immediately before the period of
greatest climatic sensitivity) climate seems to have little or
no effect on cell growth . The influence of the previous
years tends to be minimal.
Annual ring width series generally contains limited
Valais Cyprus Aggregate diagram
information about the summer of the current year, it
reflects the precipitation levels in winter and spring,
although the influence of this factor varies from site to
site. The influence of the previous year is somewhat
higher than with maximum density series.
The minimum density usually provides information only
on temperatures in late winter and in spring for dry sites.
Using information from 33 radiodensitometrically ana-
lysed chronologies for various conifer species in different
climatic zones, Kienast (1985) assessed the suitability of
various sites for the reconstruction of temperature and
precipitation data. He concluded that: "It is generally true
that annual ring width is not a useful tool for the recon-
struction of temperature. This parameter does, however,
enable a wide variety of sites to be covered on the
analysis of precipitation . Maximum density values are
well-suited to the reconstruction of temperature and also
cover a wide spectrum of sites. These data also lend
themselves to the study of precipitation on dry sites."

a
~ F M A M J J A S 0 N mfluence

•• •• -- •• •• •• ••
in
c
Cl>
-0
..c
.~ Cl>
-0
Valals • high
£
;:
-
E Jur3 medium

• • • • -• •
::J t
E 3: c;;
• small

-
x 52 Colorado

• •
<0
E
Cyprus • • • • absent

b Response of cellwall growth (maximum density) in subalpine trees to

temperature. Maximum density (cellwall growth) seems to be limited
to spring (March to May) temperature and then again in the summer

~
::: £
of growth. There is no apparent influence of temperature in June
::::. '" Cl>
':::::::::.
.........•• £- - 0
::J immediately prior to the growth period.
.......... ...-
':H1EE ::::::
3: ~
0
'"c

c
~
v;
C
<Il
-0
E
::J
E
x
<0
E

Diagram showing dendroclimatological suitability of two sites in

<Il Europe for low and high elevations as well as dry and moist sites
-0 (Kienast, 1985).
3
Based on temperature reconstructions, it is concluded that:
(a) Maximum density analysis is very good for high altitude and
moderately moist site locations.
(b) Ring width analysis is only good for very high elevation on moist
dry moist dry moist sites.
(c) Maximum density analysis is good for low, dry locations.
(d) Ring width analysis shows good results at low altitudes.

suitability

high D moderate 0 low

139
A model of the growth of a tree

A tree registers both short-term and long-term influences

These influences are converted into physiological pro-

cesses
the sum of these processes is manifested as growth
and differentiation
growth and differentiation are a response to both
internal and external influences.
;~:

,
I

;
I
(

,
I

:::
I ~

::: I
I
:;.
annual rings 4
, ,
,
/

The cell-wall acts as the information store for these internal and external stimuli

Growth and differentiation manifest themselves at the

level of cell-activity in a number of ways, e.g. as a
specific cell-wall synthesis.
it controls the type and amount of the cell-wall material
that is synthesized in accordance with physiological
conditions.
the cell-wall registers the internal and external stimuli.

The surface of the cell has a key role

It receives information from both inside and outside

the cell
it filters this information
it passes this information on - either into the cell itself
or from one cell to another
it transforms this information into anatomical structures
- the cell-wall can be said to be the end-product of
this flow of information.

140
Structure is the manifestation of all metabolic processes

The structure of individual cells is the key to their func-

span of the tracheids: short life-span in summer, long

life-span in autumn .
The stru~ture is modified by the rate of growth: rapid a
growth in early summer and relatively slow growth in
autumn (example taken from the northern timberline).

Tree-ring morphology (a) life-span of the tracheids (b) and the

position of the tracheids in the tree ring in relation to the time of cell
cl July II j !il &181
formation (c). L._ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _--'-_ _-'---'---'

A tree is a good housekeeper and follows basic rules

1. It lays in reserves: convertible cq.rbohydrates

2. It secures the transport network: tracheids, vessels
3. It builds supporting tissue: cells with thick walls
.'
"

,
Weather affects Ihe melabol,c "\
The weather," early processes dUring the vegetation
summer determines pellod
the rate at which the
metabolic processes
occur.

transformatIon

transformation
IntO

form reserves
for nex I year

Early summer Summer Late summer

and autumn

141
 IV Applied Dendrochronology

Communion cup in ash from the parish of Beggingen, Schaffhausen, Switzerland. Photograph - W. Schoch

F. H. Schweingruber, Tree Rings

© Kluwer Academic Publishers, Dordrecht, Holland 1988
Tree-ring research in the historical sciences

The fact that the irregular sequence of climatic events is
reflected in the annual rings of trees is of great impor-
tance for the historical sciences, since it allows wooden
objects to be dated to within a year. With the help of
relative chronologies constructed on a typological basis,
wood finds can be used to establish an exact chrono-
logical sequence. This in turn allows the course of human
activities, e.g. building, population growth and migrations
to be traced. When the climatic conditions of the past
have been established, it will be possible to relate human
life styles to environmental conditions over long periods.
The various types of chronologies allow relationships
covering millennia to be established for Central Europe,
as shown on the chart below.

Absolute time scale 5000 4000 3000 2000 1000 o 1000

Oak chronology
(/)
.92 Dating of floods of the main Central European rivers .
0>
o Central Europe Record of long-term climatic phases.
(5
c
2 Oak chronology Dating of prehistoric and historic buildings, e .. g record of Neolithic c(/)
..c 00>
(J
Switzerland settlement periods on Swiss lakes. .';:;.~
2~
cO'"
N-
.- cO
(5 c~
(/) Conifer chronology Dating of glacier advances. Oc
"'c
.0
<C Alps Determination of climatic fluctuations in the sub-alpine levels of -5C..c
cO
the Alps. >, .....
(/) .§:

Semi-absolute chronology Determination of the 14C content in annual rings.

Dating of annual rings.

Radiocarbon content Determination of solar activity and intensity of the geomagnetic

worldwide field.

Relative chronologies Grouping of similar artefacts.

Central Europe Assignment of the groups to annual ring chronologies.

Relative chronologies Determination of stratigraphy.

Classification according to stratigraphic boundaries
(underlying and over-lying).
Assignment to particular archaeological and historical groups.

The importance of the absolute annual ring chronologies for the calibration of relative chronologies in the history, archaeology and
geology of the last 10000 years in Central Europe.

Not all the possibilities given here have been explored
yet. Completing the chronologies and the decision on the
main correlations between the time scale and the relative
chronologies will require the work of a whole generation
of research scientists.
Tree ring analysis began in Central Europe in the
1940s and only now, forty years later, is it beginning to
bear fruit.
Out of the hundreds of investigations carried out on
prehistoric and historic objects from Europe, America and
the Near East, I have singled out only a few informative
examples which illustrate the problems and opportunities
this kind of analysis gives. Many other examples are given
by Hollstein 1980. All the issues of 'Tree Ring Bulletin'
give information on studies in North America.

144 Hollstein, 1980; Tree Ring Bulletin.

Important prerequisites for dating

As well as mastery of dendrochronological techniques it is also necessary to take certain biological and architectural
conditions into account if dating is to be carried out successfully.

Terminal tree ring

As soon as a living tree is felled, it begins to dry out, the
cells of the cambium die and wood production stops, so
that the last-formed or outermost tree ring gives the year
of felling. The time of year when the tree is felled can be
only roughly determined, since the formation of the whole
annual ring is completed within a few months, tropical
trees excepted. In hilly regions 95% of all cells are
formed in th ree months, at higher altitudes in two. The
structure of the last-formed ring shows only whether the
tree was felled in summer or autumn-winter.

latewood: August to Se~t

I earlywood : July I

Softwood (Larix decidua) 100x, from the

alpine timberline; the greater part of the
annual ring is formed within one month.

Diffuse porous hardwood (Quercus cerrls)

40X, from sea level, southern England; the
greater part of the ring Is formed in five
months.

I
earlywood: April latewood : May to September

In historic and prehistoric times trees were felled

Felling in early summer; the
last ring has no latewood.
Felling after the vegetation
period; the latewood is
complete.
Mechanically broken
terminal ring; the form is
irregular.
Forms of outermost rings found in conifer wood in1buildings
throughout the year as far as weather and topography
permitted. This can be seen in trunks found in central and
northern Europe and in the arid zones of the south-west
of North America where the outermost ring is incomplete.
It was only in mountain regions where the usual practice
seems to have been for the trees to be felled in winter, as
is standard practice today because trees are less vulner-
able to damage at this time of year than in spring or
summer.
There are both biological and practical reasons for the
frequent absence of the outermost tree ring in wood
used in building. As the sapwood of all timber is readily
attacked and destroyed by insects, it was usually hewn
off. Frequently the outer layers of trunks used in building
rotted and so the outermost ring could no longer be
found.
The outermost ring can be identified with certainty in
wood where the bark is intact. If this is missing, traces of
insects living in the cambium may indicate the outermost
ring. If the last-formed ring runs all round the trunk, this
is probably the boundary or outermost ring . If, however,
only short tangential sections are present, then the fact
that annual rings often break up in the earlywood should
not be overlooked. It is therefore necessary to look
microscopicaliy for the remnants of prematurely broken-
off earlywood cells.

Bail/ie, 1982. 145

The sapwood-heartwood boundary

As the outermost ring is often missing in wood of geo-
logical interest and that used in buildings, other features
are used to determine an approximate date of felling of
the tree. One feature which has proved useful is the
sapwood-heartwood boundary. The relationship between
the age and diameter of a tree in terms of width of
sapwood and number of annual rings has only been
thoroughly investigated for oaks.
When the sapwood boundary can be identified , a
certain number of years are added to the figure obtained
from the last heartwood ring to get the approximate felling
date. This is 20 ± 5 in Germany and Switzerland, 30 ± 5 in
England and Ireland (Hughes et aI. , 1981).

~
sapwood rings
o 50
0

0 +

..
0
'"C>
C +o~
0

.... +
"0 0 0 +
+
0
0
o 0. t 0.
+
~ ~ 0 0 0 +
a. 0
lJl 0
~
+
-0- +
0
0
15 0+

60 160 260
, D'S '0 10 JO so ~ ,a, ~2! 1$0 11S 200 2SO 100 lSO (00
sample age
heartwood rmgs

Relationship between the number of heartwood and sapwood Relationship between the age of oaks and the number of sap-
rings in 493 oak samples from the last 27 centuries. In 98% of all wood rings. There is a tendency for the number of rings to
trees analysed, the number of sapwood rings lies between 10 increase with the age of the tree but this is not always the case
and 38, the average being 20 (Hollstein, 1980). (Baillie, 1982).

glee
~--------------------------------.
1 - - - - -- - - heart/sapwood
heartwood rad,ograph ically _ sapwood radiograph, IIY ...,
1.2 / heart·/sapwood
~ 1.0 ' boundary. optically
ce
~ 0.8
u
t. 0.6
~
8' 0,4
~ 0,2 nOt tixtracted minimum density
extracted
o 1910 20 30 40 50 60 1970

Heartwood and sapwood of a beam which had been used in a
building. The sapwood zone has been destroyed to a large extent
by insects (Baillie, 1982).
Density curve of larch from the sapwood heartwood transition
zone. The components of the heartwood absorb X-rays and
therefore give a false picture of the wood density. These sub-
stances are therefore removed prior to radiography. See also
page 35 (Schweingruber et al., 1978).

146
The relationship between the felling date and the utilization of the wood
It is generally the case that the more exact the dating, the
more valuable is the historical information to be obtained.
It is important to know, therefore, how long a piece of
timber was stored before being used for building. This
varies considerably from case to case. In those areas of
Europe with large oak forests, wood seems to have been
used immediately after felling. This is indicated in the
following ways:
Signs of the wood having been worked. Prehistoric
stone and metal axes, and hatchets and mortise axes
of bygone days could be used effectively only on fresh
wood. The clearly visible wood structure on cut sur-
faces and marks left by tools indicate that the wood
was used in a fresh, unseasoned state.
Rhomboid shrinkage of trunks split into quarters,
absence of radial cracks, gaps between mortar and
wood, all indicate that the wood was used fresh.
In the few studies carried out in the subalpine belt, the
Hatchet and mortise axe, two of the most important carpenter's
tools of the past (Hollstein, 1980).
Transverse section of a beam with rhomboid distortion; in drying out
the wood has changed shape in a characteristic way. As this beam
was discovered built into a wall, its shape suggests that it was used
while the wood was still fresh (Hollstein, 1980).
dates when the wood was felled and used for building
are often identical. This is confirmed by written
records and construction dates on the buildings them-
selves. Sometimes there were several years between
felling and use. Trunks with different felling dates have
been found within the same building e.g. in log cabins
and timbered houses. The most recent felling date
shows that some of the wood had been stored for up
to six years.
Between the 15th and 17th centuries European oak
panelling seems to have been stored for 3-10 years
on average, according to Bauch et al., 1978.
In arid zones in the south-west of North America the
Tsegi Indians stored timber for years before using it for
building.
These dissimilar conclusions show the need for dating
a great many objects on sites, if an accurate estimate of
the date of building is to be obtained.
Transverse plane of an oak floorboard with clear marks of notches
and blows from an adze (Hollstein, 1980).
147
Re-use of old timber

Until the present day wood was universally regarded as a

building material to be carefully husbanded and it is
obvious that timber was used over and over again . In
areas where the climatic conditions were favorable for
preservation, e.g. dry zones, and in areas with little
forest, a high proportion of re-used timber is to be
expected. The dating of single timbers in relation to their
position within the building and marks left by tools give an
indication of the varying ages of wooden construction
elements used in a building.

Well preserved, slightly weathered beams from derelict houses are

used again in modern buildings.

Beams with sections cut away for no apparent reason which in- Marks left on a horizontal beam by a circular saw. The new beam
dicates that they were used in a previous building. was used in a recent renovation.

Long-distance transport of wood

From various dendrochronological investigations it is

known that valuable wood was carried or transported by
river or ship over great distances. Painters from northern
Germany obtained oak panelling from areas with fine oak
trees and instrument makers got resonant wood from
spruces in the Alps and Scandinavia. In the deserts of
Egypt builders fetched cedarwood and junipers from
Lebanon to build the royal burial chambers. In the south-
west of N. America Indians carried trunks of spruces
nearly 100 km to their settlements in the desert. Millions
of trunks were transported by water and it was not
unusual for such logs to get stranded in arctic, treeless
areas after a long journey by sea.

Driftwood deposited in arctic, treeless areas. As decomposition in

permafrost areas proceeds very slowly, such collections form a
dendroclimatological source of information going back over many
hundreds of years which even today has not been fully exhausted.

148
The history of house-building in the Neolithic period in Switzerland

Interpreting excavational finds from prehistoric lakeshore

and marsh settlements is often very difficult as all the
constructional elements which previously stood above
ground have long since disappeared. As often posts from
previous settlements, built one on top of the other, now
stand next to each other, the only way of construct-
ing outline plans of the buildings is by using botanical
methods and by dating the posts.
If every post possessing the outermost annual ring is
dendrochronologically dated, the constructional history of
individual houses or villages can be reconstructed on the
basis of felling dates. In a study on the Horgen period
Francuz (1981) achieved some very detailed results for a
section of the lakeshore settlement Twann, Lake Biel. The
mean curve , with 476 annual rings, synchronizes with the
Danube chronology constructed by B. Becker, and spans
the period 3348-2972 B.C. (Becker et aI., 1985) Con-
structional history of the upper Horgen period:
Field of posts at the lakeshore settlement Morigen, Lake Biel,
Felling date Switzerland, exposed by sinking of the water level in 1871172. Only
by dendrochronological methods can the ages of the various posts
3090 B.C.: two houses built with central fire- be determined (Strahm, 1972).
places.
3089 B.C.: extension to the house on the
lakeside. 3090BC 3074BC
3088-3074 B.C.: various repairs. \
3074 B.C. : total renovation of the house on the
lakeside.
3072 B.C.: renovation of two neighbouring 3075BC
houses (not shown on plan).
3071 B.C.: total renovation of house further away
from the lake. ..
These findings indicate that there was practically continu- ){!if.
ous building over 20 years in this settlement. J. Francuz
found that several settlements had been built one on top
of the other on the same site in the Horgen period. He
discovered three main building phases each lasting 10, 3
and 17 years respectively. The periods without building
activity (settlement interruption) lasted for between 12
and 187 years .
Establishing a time-scale has made it possible to
assess the influence of human activity on mostly undis-
turbed natural conditions at that time.
Reconstructed ground plans of two houses in the lakeshore settle- I
ment Twann. POints represent posts, lines join posts the outermost J.. _,. ____ --~:
rings of which were formed in the same year (Simplified: after 3090BC
Francuz, 1981). 30B9BC
Lake _

building activity

3400 3300 3200 3100 3000 BC

low r middle upper

Horgen Horgen Horgen

Main settlement phases of the Horgen period in Twann, Lake Biel (Simplified: after Francuz, 1981).

149
History of settlement in the New Stone Age and Bronze Age in Switzerland

The chronological relationship between the various settle, N

ments is of particular interest to archaeologists and to this t
end absolute and relative chronologies have to be con,
• neolithic
structed.
o bronze age

CL_5km
----~--~--------~~----~~~
Prehistoric lakeshore settlements on the lake of Zurich, Pfiiffikon
and Greifensee (eastern Switzerland) (Ruoff, 1981).

Relative chronologies

Useful archaeological relative chronologies can be con-

structed with the help of artefacts found in positions which
can be identified stratigraphically. It is also important to
have large complexes of finds concentrated in a particular

I (}
area. It is sometimes even possible to find a relationship
between stylistic features of large numbers of different
objects according to their position on a large site. (Com-
rbl~~;
bination statistics in relation to stratigraphy.) If it is ·pos-
"t
sible to compare inventories and the position of finds on I
sites which have been thoroughly investigated over wide
areas, then cultural groups - or at least groups with
similar traditions of craftsmanship - can be distinguished.
fish floats
I 1
Stratigraphy only tells us which group is older, or which is
younger. It does not fix the exact date or period of
occupation of the settlement. Egolzwll
j
Pfyn
1 Horgen SIring· early lale·
Botanical and zoological relative chronologies (pollen, ceramics bronze age bronze age
seeds, bones etc.) are only useful in connection with
Changes in artefacts over the years. Pottery is subject to fashions
stratigraphy. Their main contribution is to give information and therefore allows short periods of time to be distinguished. The
on environmental conditions at the time when they were form of axe blades did not change notably until the introduction of
deposited. metal. The shape of fishing floats remained the same for 6000 years;
only after the introduction of cork from the cork oak and polystyrene
Archaeological-biological chronology studies have did it change considerably.
been particularly successful on central European lake
shores because organic remains have been preserved in
excellent condition there and even today many settle-
ments are to be found within the same area.

Statistical distribution of forms of pottery in individual strata in the

lakeshore settlement at Yverdon, Lake Neuchatel, Switzerland. Some
types are present only in the top or bottom layer, others throughout
the whole profile (Strahm, 1977).

150
Absolute chronologies
Great progress has been made in archaeologically based
dendrochronological research in central Europe in the
past ten years. Laboratories in Hemmenhofen (F.R.G.),
ZUrich, NeuchiHel and Moudon (Switzerland) have dated
wood from the pile dwellings of central European peat
bogs and lakes. The tree-ring sequences of thousands
of oak posts from Upper Swabia, Lake Constance, lakes
in the eastern part of Switzerland e.g. Lake ZUrich, lakes
in the west of Switzerland e.g. Lake Neuchatel, Lake
Geneva and Lake Annecy could be fitted together to form
local chronologies. The absolute dating of archaeological
samples was only made possible however by geologically
dating oak finds from the gravel of rivers and peat bogs.
The synchronization of the series and the absolute dating
which occurred simultaneously became possible after B.
Becker (1983) in southern Germany and J. R. Pilcher
(1984) in Ireland, working quite independently, produced
chronologies spanning more than thousands of years.
There was now no difficulty in linking the central European
relative archaeological chronologies with the absolute
geological series from southern Germany. 130 years after
the discovery of the first pile dwellings in Meilen, Rohren-
hab on Lake ZUrich, posts from nearly 200 prehistoric
central European villages were dated accurately to the
year. A time comparison between all villages studied up to
present shows clearly that many gaps exist in our knowl-
edge of the prehistoric period. For four hundred years
1 r
:"'
'. ~
Lacs in western
Swi tzerland ? ~ ~ ~; ~
?
'"
Lacs in eastern
Switzerland
- ~I~.
9~
1
~
-
Lac of Constance
~ '"
upper Swabia ~ ~
neolithic period
Inhabitation of lake-shores in Switzerland and southern Germany by pile-dwellers. Shortly before 4000 B.C. the first non-nomadic people
were to be found in Switzerland (Egolzwiler culture). The pile-dwelling culture lasted about 3000 years. Wood for the last pile-dwelling was
felled in 848 B.C. in Cortaillod on Lake Neuchatel (Suter and Schifferdecker, 1986, Ruoff and Rychner, 1986).
The climate in relation to absolute chronologies
Little is known about this. The only key to such questions
as the interruption of settlement occupation lies in the
chronologies of high-lying regions in the Alps, which can
be interpreted climatically. As far as the lakeshore settle-
ments in eastern Switzerland are concerned, the climatic
chronology shows that the 500-year period of non-
Becker et aI., 1985; Suter and Schifferdecker, 1986; Ruoff and Rychner,
1986.
between the early and late Bronze Age and eight hundred
years between the early Bronze Age and the Neolithic
Age the land near lakes and bogs was uninhabited. In
some cases the change from populated to unpopulated
occurred much more suddenly than had been suspected.
If one considers that at the present time, a hundred
years after a settlement in an area of broad-leaved trees
is abandoned, all traces of human activity have been
destroyed by the advancing forest, then it seems likely
that the prehistoric folk had to cut down trees when they
re-built a settlement on a lakeshore near a forest. The first
non-nomadic people, Neolithic man, were hardly in a
position to destroy large areas of forest, since the number
of these people, living extremely simple lives, was very
small. As U. Ruoff in Becker et a/., 1985, showed, even at
the peak of the Neolithic Age there were scarcely more
than 20 settlements on Lake ZUrich. In total there may
have been about 100 people living on the shores of Lake
ZUrich in 2700 B.C., the period of coil pots, which is a
thousandth part of today's figure. It is therefore not
surprising that in the pollen profiles of neighboring
regions, the cultural influence is no longer recognizable.
Some questions remain unanswered. Why were the
periods of settlement interrupted? Were the comings and
goings caused by rises in water level in the lake, by
fighting or by outbreaks of disease?
lj
~ B~
~ ~ b
~
848
~ '§
,
~ ~
~
~
; 911 BC
850 B
J
~
bronze age
occupation coincided with the so-called Gbschener cold
phase around 1300 B.C. Early Bronze Age settlements,
however, were also deserted during normal or warm
periods. The question of the relationship between climate
and settlement requires much work by archaeologists and
dendrochronologists.
151
Prehistoric settlements in the south-west of North America
No area has been more thoroughly investigated den-
drochronologically than the south-west of North America.
The 'Tree Ring Bulletin' has many articles on this subject.
Bannister and Robinson, 1975, summarized the state of
research and Ceram , 1972, described in an enthralling
manner some episodes in the history of dendrochro-
nology. See also page 258. In spite of lack of written
records , parts of the history of the indigenous, mainly pre-
Columbian , population can be reconstructed in detail. In
the settlements - very often cliff dwellings - much plant
material is to be found , preserved by the desert climate.
Seeds, wood and charcoal all provide information on the
diet and environment of the Indian folk while dendro-
chronological investigations on beams with annual rings
Cliff dwellings in Betatakin, Arizona U.S.A. Overhanging cliffs pro-
tected the dwellings which were constructed near valleys with plenty
of water. Fields above the dwellings were cultivated.
Roof construction of a mud house in Betatakin, Arizona (Dean,
1969).
152
provides a chronological framework for archaeological and
biological finds. In the past 60 years workers from the
Tree Ring Laboratory in Tucson, Arizona have dated some
20000 beams from about 1000 settlements. This work is
enabling archaeologists to fit together extensive historical
connections.
It has been shown that, rather than investigating many
areas in detail, inter-disciplinary collaboration between
archaeologists, palaeo-botanists and dendrochronologists
studying settlements such as Betakin and Kiet Siel in
northern Arizona and Chetro Ketl in New Mexico (Dean
1969) can add greatly to our understanding . Douglas,
1935, Bannister 1965, Robinson 1976 and Dean 1969 and
1983 in particular, have been responsible for opening up
the 'Secrets of the Southwest'.
1270
PRE
29 [2'c;:i::P=R=E : : : 0:~1
c
Changes in the arrangement of rooms in cliff dwellings In Kiet Siel,
Arizona between 1270 and 1283 (Dean 1983).
piaster
juniper bark
~r:::::;;'-~iaI""f====--- piaster
:llii5ii_~~~~~:§ii;:-- juniper bark
IiO """'____ willow rods
primaries
Roof construction of a house in the settlement in Chetro Ketl, New
Mexico, U.S.A. As in other areas of the world, environment and tradi-
tion influenced the type of house built (Dean, 1983).
Bannister and Robinson 1975, Ceram 1972, Dean 1969, 1983, Douglass
1935, Bannister 1965, Robinson 1976.
The development of settlements in Betatakin and Kiet Siel
On the basis of 292 dated pieces of wood from Betatakin It is particularly interesting to compare the history of
and 540 architecturally related samples from Kiet Siel, settlement in one region . The Tsegi region was probably
Dean (1969) succeeded in reconstructing the settlement both populated and later abandoned in single waves of
history of these cliff dwellings which today have become a migration . Presumably the group of about 600 people was
popular tourist attraction. The first step was to analyze, driven out of the flat region of Monument Valley by
archaeologically and dendrochronologically, every room drought and had to retreat to valleys with more water. The
and group of rooms in a settlement. Just one example 'Great Drought' , a catastrophically dry period between
from Kiet Siel is given here. See figure page 152. 1276 and 1299, probably forced the Tsegi southwards
Around 1270 the first dwelling was laid out consisting into the territory of the Hopi Indians.
of a living area (16) a granary (15) surrounded by a
passage next to the back wall of the cliff (23) and a court-
yard (29). In 1273 another granary was added (not
shown). In 1275 the inhabitants divided up the large living
area (16) making it into a smaller living room, a larder and
a grinding chamber. The passage (23) became a court-
yard. In 1283 the dividing wall in room 20 was taken away.
One door in room 20 was virtually closed up leaving only
a small hole to ventilate the grinding chamber (16).
The age structure of the whole settlement could be
determined by dating all the available wooden beams
found there. All the houses still standing today were built
between 1267 and 1280. This can be verified by looking
at the most recent felling dates of wood in each house .
Many of the posts were found to have been felled in
bUilt in 1267 - 1268
1267 - an indication that the comparatively few inhabi-
tants of the dwellings were expecting a wave of immigra- D bUllton 1275
bUilt in 1276
tion at that time, which did not actually take place until
rill;ID buil ton 1277
1275. Before 1267 the refuge in the rocks was only lightly
Ul!IIII bUill In 1278
populated . Dean compared the population development in
~ builtIn after 1280
Betatakin with that in Kiet Siel, which is practically the
same age, and came to the conclusion that probably
o date of constructIon unknown

between 125 and 145 people may have been living in the 010 20 40 60 80 feet
two settlements between 1275 and 1277, according to
archaeological estimates. The end of the period of occu-
o 3 6 12 18 24 meters

pation is uncertain because the beams show only the

Plan of the settlement at Betatakin showing dates of construction of
felling date of the trees. Archaeologists believe, however, individual chambers (Dean, 1969).
that the settlement was abandoned around 1300.

~
40 1250 60 70 80 90 1300

I II I I
~~t),~a~io I site 90 1200 10 20 30 401250 60 70 80 90

J. .
Betatakon

"-
~
I
\1
fellong year -

I I ,I Kie Siel

....
i"
Betata on
I Scaffold House
~ I
I
I
Swal ow's Nest I- I--
I
I

o l\A _lit. !t~ (-

\A
-'I
,I Na-8435

Lolomaki
I
I
1/' Na-2606 I-- ~
\
\
fellong year - \ Batwoman House ,

K,et Siel II I ,
J
\
"
,
Twin Caves Pueblo
I I
I I.
jW'

1I1hab\~a!:?~ \ Nagashl B,k,n I +-

,'- ~' . \
Calamity Cave I
I
j.
,
M.. k W
Iv~ I
I Long House I I
o RUin 8 I
I I

Felling dates (solid line) compared to estimated population (broken Range of dendrochronologically determined felling dates in 13 Tsegi
line) in Betatakin and Kiet Siel (Dean, 1969). settlements. Horizontal lines: range of felling dates. Vertical bars:
accumulation of felling dates (Dean, 1969).

153
Hand in hand with dating of tree trunks goes the identifi-
cation of wood types, measurement of trunk diameter and
the search for signs of the wood having been cut to shape
and size. Together with archaeological findings, such
investigations can throw light on the life-style of the
people. In the following example Dean, 1983, reports on
the settlement of Chetro Ketl.

Supply of wood

It needed all the combined efforts of a community to

collect together the wood required to build a settlement,
as not much was available in the way of mechanical aids.
The stone axe served as both felling tool and for dressing
the timber, as axe marks show.
In very short but intensive periods of activity the wood
was felled, as can be verified by fixing the exact felling
dates. If the outermost ring is incomplete, then the tree
was felled during the vegetation period; in the case of
Chetro Ketl between March and September. The exten-
sive fellings of 1039 to 1046 and 1052 took place in
spring, that of 1063 in early summer. In Betatakin, how-
ever, the people collected the wood in late summer,
usually selecting only living trees for felling but occa- The settlement of Chetro Ketl in Chaco Canyon, New Mexico, U.S.A.
sionally pieces have been found, usually split sections, By courtesy of J. S. Dean.
having an outermost ring which was formed 100 to 300
years before the settlement was built. Dead wood was
only used when very special pieces were required. In Chetro Ketl alone as many as 26000 beams could have
The quality of the wood was regarded as very impor- been used. In the years between 1030 and 1060 A.D.
tant. These people went looking for fine, 6-30 cm thick 200000 trees were felled near the 10 settlements in the
trunks of Ponderosa pine, spruce and fir over wide areas. surroundings of Chetro Ketl. What had once been dense
Dean suspects that spruce trunks were brought from as forests became fields, heaths and parkland.
far as 75 km away. One can imagine the effort involved In a few years - for example between 1037 and 1039
in carrying logs, 30 cm thick, 5 m long and weighing and again between 1051 and 1052 - about 5100 and
up to 200 kg, a day's journey. Species like Douglas fir, 4100 trees, respectively, were felled within a few months.
pinyon, juniper and poplar growing nearby tended to be One can imagine large groups of tree fellers, two to
disregarded. Spruce and firs from higher altitudes were three days journey from base, cutting down 16000
only collected after 1030 - probably because the forests Ponderosa pines in the lower-lying mountain areas and
of ponderosa pines in the locality had already been ex- 6000 spruces and firs in the higher ones. Branches and
hausted. bark were removed and the trunks were chopped into
The felling activity had an enormous effect on vegeta- shorter pieces on the spot. Usually the wood was used for
tion over a wide area. Dean came to the conclusion, on building purposes immediately on arrival at the site. Only
the basis of his analysis of wood types, felling dates and in a few cases was it stored; some beams exist, for
the use to which the wood was put, that the 575 beams example, which were felled between about 1033 and
analyzed represent only a fraction of what was once there. 1034 but which were first used between 1037 and 1040.

",50
nincomplete terminal rings
~ ~ complete terminal rings
-@l 40
Ol
ntotal number of cutting dates
c
:: 30
::J
u
"0 20
.810
E
E o h.~~~=d~~~~~~~~~~~~~~~~~~~==~~~~~~u=~b=dh~~
1000 10 20 30 40 1050 60 70 80 9') 1100 10 20
years

Distribution of 1000 felling dates with time from Chetro Ketl, New Mexico, U.S.A. Feverish building activity went on between 1051 and 1052
and then again between 1237 and 1240 A.D. (Dean, 1983).

154
Wood in the settlement
40
Wood was sorted on the building site according to its Ql
C5.
suitability for various purposes. E 30
c. secondary beams
The straight trunks of the Ponderosa pine, spruce and lJl
fir were used chiefly as primary and secondary beams. '0 20
~

The thinner pieces of Douglas fir, spruce and fir were .D 10

designated for aperture elements. Misshapen pieces of E
E 0 + - -"",
pine were set aside for use as firewood. Pieces of wood o 50 75 100 125 150 175
have been found in roof constructions which cannot be
dated dendrochronologically. The builders used only B. pnmary beams
wood without bark. In Chetro Kiel the tree fellers in the 10
forest were responsible for removing the bark from green
trees immediately after felling. In Betatakin, however, 0+-----,,----,--- -.....-
o 25 50 75 100 125 150 175
wood was left lying around until the beetles had eaten
radius in mm
away the bark, as is shown by the traces left by the
insects. Marks left by axes indicate that roof beams were Tree thickness distribution of primary (load-bearing) and secondary
often cut to length after having been built in. Floor-boards beams (Dean, 1983).
were cut with stone axes and finally polished with sand-
stone. Wood was always a valuable material and it was
therefore not surprising that it was re-used in new and
renovated buildings. This can be proved by the number of
older beams found in more recent complexes. Older
beams were built into a cellar in 1050, for example, to
stabilize it before extensions were added. If beams were
of no further use they were burnt, as is shown by the
many charred pieces of potentially good building material
found.

Indians in the south-west of North America oHen used dead wood,

the bark of which had been destroyed by bark beetles.

Beaver teeth marks on a poplar trunk. Blows from a stone axe from the pre-Spanish Blows from a metal axe, Arizona.
period, Arizona

155
Historical buildings
Those concerned with the history of architecture are
interested in drawing up an accurate time sequence for
the construction and development of the buildings being
studied. It is very seldom that accurate dates can be given
solely on the basis of the features of the building itself.
Nor can the dates on the building or historical material
provide a complete picture, even where important build-
ings are concerned. It is really only dendrochronological
Trier Cathedral (Federal Republic of Germany)
Through the dating of embedded supporting timbers,
floor-boards , chests, etc., Hollstein established that this
building was begun in the fourth century, and that building
continued more or less continuously into the eighteenth
century. Hollstein's reconstruction of the building of the
west front is extremely impressive. It was possible to date
the oak supporting timbers, which were embedded in the
156
analysis which can furnish a full and precise account of
the history of individual buildings.
Two examples of dendrochronological investigations of
historic buildings are discussed below. Hollstein (1980)
has studied one of the most important ecclesiastical build-
ings in Germany, Trier Cathedral , and Stahle (1979) was
able to date one of the oldest buildings in Arkansas , the
Wolf House.
walls , even though the annual ring sequences were rather
short. The available written records were checked against
the dates produced dendrochronologically, and were
shown to be accurate. It was clear that the west tower was
begun by Archbishop Poppo and completed by his suc-
cessor, Bishop Eberhard, and that the west front was
completed under Bishop Udo, twenty years later.
I - ..
'.~ .......
It-,It (~ ... ' ., .,~ .... , ....
·"f'()("I.t , - c,,,
Tree-ring width curves for oak timbers in the west front of Trier
Cathedral. The first building phase lasted until the death of Arch-
bishop Poppo in 1047. In the second phase, which was from 1053 to
1056, work was carried out exclusively on the north-west tower
(Hollstein, 1980).
The west front of Trier Cathedral. The dates shown relate to the
embedded supporting timbers. This monumental building was begun
under Archbishop Poppo, and by the time of his death had reached
the height of the second gallery over the portals. With the resump-
tion of the building work the north-west tower was completed in
the years 1054 to 1056 and the south-west tower in 1074 and 1075
(Hollstein, 1980).
HoI/stein, 1980.
The Wolf House, Arkansas
In the United States the houses of the first European
immigrants are now protected buildings. Although only a
hundred years have elapsed since some of these houses
were built, it is not always possible to establish the exact
date that they were put up, even where contemporary
documents are available. In the case of the Wolf House
in Arkansas, historians discovered that Jacob Wolf emi-
grated to America from Germany in 1820 and acquired
the land on which the house now stands on July 20th,
1825. Dendrochronological analysis of the beams of
Weymouth pine (Pinus strobus) used to build the house
showed that it had been built in 1828.

anginal imber ~ repairs

1828·'-

1826'

182" -
1827
rH?4

{:\ .... -'t ___ .__ . _.J:c__

The Wolf House: front elevation. The last tree used for timber was felled in 1828, which means The Wolf House: side elevation with
that the house cannot have been built before this (Stahle, 1979). felling dates of the individual timbers
(Stahle, 1979).

Wolf·House southern ye ll ow pine

tree ring widths

mm

~
~ 0'
~~ ~.= ·~. ~J\,
l :y=v=::vv. ..
V
Lo., : v-:v: ,
~::~::: 4 A_ .-
Synchronized annual ring width curves
for the Weymouth pine beams used in
167016801690 170V 1710 1720 1730 1740 175017601770178017901800 1810 1820 1830 the building of the Wolf House.
Top: Single curves;
bottom: indexed mean curves (Stahle,
1979).

157
Sacred and secular buildings in Northern Germany
Since the early 1970s the architecture of Northern Ger-
many has been the subject of intensive dendrochrono-
logical examination . In conjunction with architectural his-
torians Eckstein has succeeded in reconstructing the
development of building methods and of building styles in
this area of Europe. The Holy Ghost Hospital in Lubeck
has probably been the subject of the most intensive study
of this type. On the basis of the analysis of a considerable
number of samples of oak - 769 in total - it was
established that all the wood used for the ceilings and in
the construction of the roof came from trees felled in the
four years 1284, 1286, 1287 and 1289, and that the wood
was used as timber in each case the year after the tree
was felled . An examination of the annual ring patterns
showed that the wood came from forests in East Holstein.
The fact that in such a short space of time hundreds of
oaks were felled, chopped up for timber and then used in
the construction of a building in accordance with detailed
plans is a measure of the rigid social organization that
must have existed at that time. The dendrochronological
examination revealed that the building was extended
SCHLESW'" -
HOLSTEIN
• villa, twohouse
• country house
.. farm house N'EDERSACHSE
The sites of buildings that have been examined and dated using
dendrochronological methods in the environs of Hamburg.
Villas and two-houses. Country houses. Farm-houses.
Farm-buildings. (Eckstein et al., 1977).
I'~ I~I"
A farmhouse with wood dating from the years 1532, 1542, 1565 and
1663 (Eckstein et al., 1977).
158
twice in the fourteenth century and that from the seven-
teenth century onwards it was rebuilt or altered about
twenty times. Conifer wood was used for most of these
alterations : this indicates that oak had become scarce in
East Holstein by this time.
An analysis, using dendrochronological methods, of
thirty-two farmhouses and farm outbuildings has enabled
researchers to form a detailed picture of many aspects
of the history and development of rural life. Changes in
the layout and organization of farms can be traced by
establishi ng when different types of build ing were first
built and when they were altered or renovated . Such
changes mirror in turn the economic development of a
whole area. In addition to providing this sort of general
information dendrochronological dates, whether exact to
the year or correct within a certain tolerance, are useful
tools in the reconstruction of the history and development
of individual houses , and provide invaluable material with
which historical documents can be checked and comple-
mented.
The Holy Ghost Hospital in Lubeck; from a lithograph about 1850
(Eckstein etal., 1982).
A country house built in 1598; from a lithograph by W. Heuer, 1885
(Eckstein, 1977).
Sacred buildings in Greece and Turkey

Both Eastern Europe and the Near East had been ne-
glected by dendrochronologists until relatively recently.
Gassner and Christiansen did the first dendroclimatological
work in this area.
It was Bannister who, in 1961, recognised the potential
for dendrochronological analysis in Greece and Turkey,
through his work in Gordion. Using juniper wood found in
a royal tomb he developed a chronology covering 806
annual rings. And twenty years later Kuniholm established
that ancient buildings in many areas of the Near East and
of south-east Europe could be dated using annual ring
series for conifer and oak. It seems that for more than fifty
of the hundred years studied the dry summer conditions
had a similar effect on the growth of trees over an area Recent conifer and oak chronologies which can be synchronized
extending 2600 km east to west. with those based on data from the forest of C;atacik in central
It proved relatively easy to take samples from churches Turkey. The 700 year conifer chronologies (Pinus nigra, Pinus
leucodermis, Abies cephalonica and Abies borisii-regis) for Greece
and other sacred buildings since many of them had been and Turkey display synchronous curve patterns (Kuniholm and
damaged in earthquakes and were being rebuilt or re- Striker, 1983).
stored. However, only samples from about a third of the
buildings that were examined were suitable for dendro-
chronological analysis, and only about half of the samples
themselves yielded absolute dates. Despite this Kuniholm
succeeded in constructing a chronology for oak covering
the period 1073 to 1984. This chronology was based on ,.._t __ . "'· ___
~/' 9

samples taken from 80 buildings in more than 40 different

places. The results of the analyses, which have not as yet .. ~' .. "-~.....
:
_~J~~'

been published in full, will provide the key to a number of

important archaeological questions.

1100 1200 "00 '400 .SOC>

Ol.allr.ilo. H. P.r •• It.

11,,011 11 5039
Th4!.U .. lonl ~U, Ioohh.. T e,.
''''1-1----------------II' ~I~.
Th. . /nh,i. H. Soph1 •• It .
IJS"'" 1 1~21
Arla. ' .. ,lI; , ..aha " .
llU I Il )oJ
5.1' 1' .... linc:lr.
U9.81 1.491
The.liaJanlki, }t, Apo.ll •• ~ Iill.
11411--1- - - - - ' ---'------11' .. ""
Sar1'"II:Ii. KeNlel Boey
112' 1-1- ---'-----'--jI1....
CannllJr,ale, Kilit Iohit
119~1 111063 I
" illll.l., H. o.etno.
alII I 11""'8 I
O.t.dyaotai,.;;hQn. Hetwclt I IquG v . roof
1186 1 IUt"
luuk, Ih luh,. I.ant
1l10' 1-1---..:..------tl ll~4
Sotlrn$ , ;tOdroeos, early
IUlI l I UI,,:i
n.../niki, VlaLAISOn
1198 1 11))9
GREEK OAK ot:NOMOCHRONOLOG't 197)- 1'986

, . ---.
T1'"i 1)'11, i!:ea.rl' 1(1 lin wEDIEVAl TO .. ODCAH SITES
1191 ' I l1l1:i
Pythion, Ca.5 t I. ""- "".
_.
II ...... UlUJoI ,GO .. 1ikArloI .. .
UQ81 lIB} :ft,
Utanbul, H. Sophia, !fill Bun . c_"cMlt II 0

. . ~
11~'9 I . 11119 ~
Grac:aniu
10"1-1- - - -- -- - - . . . ,1'291
Art .. ,
11711
PUIl,tn'lll •••
Ina6
,QIIIIII'I('"

IIU"; • QI"lfll .
" "
1I 0
· ...... 0

The main constituents of the oak chronology for northern Greece in Sampling sites in Greece for historical wood (Kuniholm and Striker,
its older part (Extract from Kuniholm and Striker, 1987). 1983).

Eckstein et al., 1977, 1982; Kuniholm and Striker 1983; Bannister 1963;
G. Gassner and E. Christiansen - Weniger, 1942; Liphschitz, 1985. 159
Village in an alpine area
Material and dating
There is no difficulty in principle in reconstructing the
architectural history of whole towns and villages using
dendrochronological methods. However, where it is nec-
essary to date a number of samples from each building
the amount of time required becomes prohibitive. There
are only a few studies in which several buildings in one
town or village have been dated. One such study has as
its subject the village of Lauenen in the Bernese Ober-
land. This village was built entirely of wood. As part of the
dendroclimatological study radiodensitometric methods
were used to construct a 700 year sequence, which was
The village of Lauenen (1250 m) in the Bernese Oberland. All the
houses are built of spruce, which grows in this area up to the
timber-line.
The interior of Saanen Church in Central Switzerland. A fire in 1940
destroyed everything but the walls and the massive spruce columns.
Dendrochronological analysis confirmed the date on the church of
1444.
160
based on material from buildings whose construction date
was already known. Samples were taken from churches,
farmhouses, a school and a barn, and it proved possible
to fit the annual ring series together at the first attempt.
The chronologies were used to correct some erroneous
dates and to provide dates for two houses of special
local interest whose construction dates were not known.
Built in 1456 and 1505 respectively, they are both from a
later period than the impressive church in the village of
Saanen.
The old mill in Lauenen has a date on it of 1756. The style of the
building, with its gently sloping roof, is characteristic of this region,
while the decorative carving and the paintings on the outside of the
building are typical of this period.
Examples of bands of carving decorating the outside of buildings.
These are all taken from the Bernese Oberland and are all typical
of particular periods. They can be used in the construction of a
chronology by providing clues to the relative dates of the buildings
(Rubi, 1980).
Rubi, 1980; Schweingruber et al., 1979, 1984; Siebenlist-Kerner, 1984.
 indices

:']
t:J. g/cm 3

- 0.1
1000- 1100

20 40 60 80 100
Mean curve for maximum densities for spruce (Picea abies) and fir (Abeis alba) derived from samples taken from old buildings and from
recent wood in Lauenen and other regions in the northern prealps. The curve is covered up to 50 times. Variations from the central axis
are shown here. The individual curves were smoothed using a Hugershoff function.

161
Dating old masters

Since the mid-1960s researchers in Great Britain, Ger- disciplines, has as its goal the attainment of objective
many and Switzerland have applied dendrochronological findings which enjoy absolute certainty. The example
techniques to the dating of the wooden panels on which discussed below illustrates how effective dendrochronol-
old masters are painted. Art history, like all other historical ogy has been in helping to achieve this goal.

Aims and objectives in art history

The art historian seeks to date a work and to establish its de:libN,UC fa e

authenticity on the basis of the techniques used.

~~~·~~~,~O !I DO
0 O! LF' 0
Pdlnw"19 In the " copy palnung In the

copy from the n /. style of the

mas,., i.:~~)'.
I f k ,; style of the master

O .i /
artl5l'SStudio copy painted I
by. puPil ,
I ~/
panlalfake
'
I lsogned)
second vefslon

copy made by Ihf!'

,n,,,h,mseif
0 t;/'
i '

/'
0'/.
)~ ~
I ,;

~ ~
later copy

copy based on

rF=l(~ L·~~:~=~-·-·-·~o-·-·-
Different types of copy which may be made of a masterpiece: those
linked by the broken lines are of particular interest (Bauch et al.,
U- 1"9ned )

1972). present-day
I,Uf!'r periods

The aims and objective of the dendrochronologist

The dendrochronologist is concerned with the wooden tion about when the work was painted. By determining
panel on which the work of art is painted: by dating the the felling date of the wood he can indicate the earliest
wood hecan provide the art historian with further informa- possible date when the picture could have been painted.

Dendrochronological methods

The measuring procedure here differs somewhat from the

techniques used in other areas of dendrochronology. The
tree rings are measured directly on the intact wooden
panel itself, the panel first having been taken out of its
frame. The edge of the panel is cut using a very sharp
razor-blade, so that the individual pores can be distin-
guished. The tree rings are then measured with a hand-
lens fitted with a measuring scale. They can be measured
to an accuracy of 0.1 mm.

In the gallery: the annual ring widths are measured on the edges of
the paintings (Bauch et al., 1972).

162 Bauch et al., 1972, 1978; Fletcher, 1978; Klein, 1980; Eckstein, 1983.
The material
Up to the eighteenth century most painting was done on
wood. The type of wood used depended on where the
painter was active.
The Atlantic coast from Denmark to Portugal: mainly oak.
Central Europe: mainly lime, oak, fir and spruce.
Southern France : mainly poplar and walnut.
Italy: mainly poplar.
Spain: mainly poplar and Scots pine.
Inl rn 00 ,
250 o anYd$
• Iroplcal wood
tV other wood~
PW p.tpt:!f on wood
200 0 opper
S '5ilate
I!>O
100
!>O
0
14!>O
The use of different materials which were used in 2066 works by 366
Dutch painters dating from 1400 to 1900. The paintings have been
dated either by art historians or were dated by the artist (Bauch et al.,
1978).
The limitations of the method
The type and condition of the wood can severely restrict
its value to the dendrochronologist. Where the last-
formed rings are missing or the length of time that the
wood was stored before being used is not known, it is
very rarely possible to date the wood to the year. Oak
panels are very often difficult to date.
Klein (1980) described the sorts of problems that may
be encountered as follows: "Where the bark still exists
(see case a) the year in which the tree was felled can be
determined exactly. Where the bark is missing but the
sapwood can be seen (as in case b), it is possible to
ascertain in which year the tree was felled by allowing for
the missing sapwood rings, i.e. 20 years, depending on
the age of the tree, will be added on." (See page 146).
"Where the samples under investigation consist entirely
of heartwood, as is often the case with painted panels and
wood carvings, the earliest possible felling date of the
wood can be given, provided that at least fifteen sapwood
rings are still in existence. Pieces of wood which display
The panels supplied for paintings generally comprised
radial sections of trunks, as these were more stable and
less liable to shrinkage than other types of panels. This
means that there is an optimum number of rings available
for the dating procedure. As a rule the sapwood was
partially or wholly removed by those making the panels,
as this wood was particularly liable to attack by insects.
Panels of poplar cannot be used for dating; those from
lime have only a limited value.
Determining the felling date of oak used as the basis for paintings.
Part of the tree from which the wood originated. Case a: bark still
existing; case b: part of the splintwood still existing; case c: the
wood partly comprises heartwood, the end of the annual rings can
be seen on the sapwood/heartwood boundary; cases d and e: the
piece of wood consists of heartwood (Klein, 1980).
only a few annual rings could have come from anyone of
a number of areas within the trunk. In such cases (d, e)
there may well be disparities between the earliest felling
date provided by the dendrochronologist and the date
suggested by art historians."
Since wood was stored for varying lengths of time
before being used, the dates given for sixteenth and
seventeenth century works are only accurate to within
between three and five years.
It is possible to obtain more information about the
probable source of the wood used in works of art by
examining the wood for growth patterns that may be
characteristic of a particular region. Eckstein (1983) has
provided regional chronologies covering the Atlantic coast
areas.
Chronologies for beech are available only for the
northern part of Central Europe (Germany). It has none-
theless proved possible to date the works of very many of
the great painters of Europe.
163
Dating individual works

Peter Paul Rubens: Child playing with a bird

This picture had been dated using conventional methods,
and the date given was 1616. It is painted on three panels
of Dutch oak. The annual ring sequences for this oak are
very similar to those of the standard curve for the Dutch
coastal area.
In two of the panels (I and II) the annual rings run
horizontally, while in the third (III) they are vertical. Den-
drochronological analysis showed that the wood for I and
II must have been obtained in about 1594, and that for
III after 1621. Even if we assume that the wood was
seasoned for the minimum possible time, the analysis still
shows that the original date given must be brought
forward by at least ten years.

'Child Playing with a Bird', B 763, Three-piece panel ~ fibre direction in the wood. (Photograph: Art Gallery Berlin-Dahlem)
(Bauch et al., 1978).

164 Bauch et aI., 1978.

Dating a painted panel from a house in the east of
Switzerland

An impressive wall-panel, painted on pieces of spruce
wood , was found recently during the renovation of a
house in the village of Gais, in the north-east of Switzer-
land . Art historians were able to identify the 3 m by 5 m
panel as sixteenth century. Dating of a layer of dust found
under the paint showed that the house predated the
painting by some time . Radiodensitometric analysis of six
samples from the spruce wood enabled a 62 year annual
ring mean curve to be constructed. This curve matched
that which had been constructed for spruces in Lauenen
and indicated that the trees had been felled in 1497.

Wall panel in Gais, Appenzell, Switzerland. The cow-herd encourages the cows into the stall with salt (Schweingruber et al., 1979).

Dating a series of paintings

The dendrochronological analysis of all the paintings
produced by an individual painter or by a school can
prove of enormous value in checking or complementing
the findings of art historians. The following studies of this
type have been carried out: Rubens (Bauch et aI., 1978),
Tudor portraits (Fletcher, 1978), Rogier van der Weyden
(Klein , 1980) and Rembrandt (Bauch and Eckstein , 1981).
Such studies aim to shed light on the following ques-
tion :
Can dendrochronological methods succeed in putting
works of art into the correct time sequence where
conventional methods of dating have failed?
Is it possible to attribute unsigned works to particular
painters on the basis of the growth patterns in the
wood on which the pictures are painted?
Did certain works that are now in different places
previously belong together: were they, for example,
originally part of a triptych?
Were the wooden panels on which the works of art are
painted produced near to where the artist was active,
or did the artist have them brought from further afield?
This question can often be answered by examining the
existing regional chronologies for oak.
How far is it possible to date sculptures, or at least to
say in which order they were produced?

Klein, 1983. 165

Criminal investigation and art dealing
Forensic scientists today have at their disposal the most
up-to-date methods and equipment. It goes practically
without saying that forensic units of police forces will
include specialists who are aware of the role that the
Determining the species or type of wood
The most important aspect of forensic work on wood is
the determination, by means of an analysis of the micro-
scopic structure of the wood, of the species or type
involved. Such information may lead to a suspects being
cleared or incriminated.
What can be identified?
Pieces of non-carbonised wood or charcoal, even
those no larger than wood shavings.
Fungal hyphae in wood: such traces say something
about the condition of the wood.
The way in which the wood has decayed or disinte-
grated and damage by insects: these features provide
information about the conditions under which the
wood was stored.
Examples:
Establishing a time sequence
In some cases the only information that is needed is
whether or not something was done before or after a tree
was planted or began to germinate. Here a straightforward
count of the number of annual rings at the base of the
trunk will provide the information.
Examples
during building work on a site which had been covered
with trees a skeleton with a damaged skull was dis-
covered. It was found under the root zone of the trees,
which meant that by ascertaining the age of the trees it
Dendrochronology
There are many ways in which dendrochronological tech-
niques can be applied to crimes, including possible
frauds. These techniques, which may involve the mea-
surement of tree-ring widths, or tissue and density analy-
sis, can be applied to cases as diverse as murder and art
frauds. Using dendrochronological techniques it is usually
possible to produce dates that are exact to the year, and
where a tree was felled during the vegetation period the
dating can be even more precise. Should the place of
origin of a tree be known it is possible to decide at which
time of the year the tree was felled - early summer, mid-
summer, (sometimes) early autumn, and autumn or winter
- on the basis of information gleaned from the annual
rings.
166
analysis of wood can play in criminal investigation. A
number of examples of the successful application of
dendrochronological analysis in police work are given
below.
A man was accused for cutting down some valuable
osiers which belonged to someone else. Shortly after
the alleged theft all the wood shavings found on the
suspect's work-bench were submitted to microscopic
examination. As this analysis showed that he had been
working exclusively with conifer and fruit-tree wood,
he was cleared of suspicion.
A large museum was about to purchase an early
Gothic crucifix from southern Germany. Previous work
done on such figures had shown that they were made
of poplar. However, an examination of this crucifix
revealed that the arms and fingers were of non-con-
temporaneous wood, in this case box and pear, and
that only the body itself was of poplar. When an X-ray
of the figure showed that one of the nails was modern,
the museum decided not to buy the crucifix.
would be possible to determine whether the death had
occurred recently enough for it to be appropriate for
the police to investigate.
a piece of land which was no longer cultivated and was
covered with bushes was put up for sale. The date that
the land had last been worked was important for the
negotiations. A count of the annual rings in dwarf
shrubs (calluna) which were growing on the land
showed that none were more than eight years old. It
was clear that the land had not been cultivated for
about ten years.
Wood carvings and other objects made of wood often
fetch high prices, and the possibility that the piece is not
genuine can never be ruled out. Generally, the main thing
that has to be established is the age of the work of art.
Ring width and density are used to determine this. There
are two ways in which this may be done:
in relative dating a mean curve is constructed using
material from a number of works of art whose authen-
ticity is undisputed. The piece under investigation is
then compared to this.
where standard curves are available for the same sort
of wood from the same region, absolute dating is
possible.
Uese and Eckstein, 1971; Schaefer, 1954.
The dating of stringed instruments
In the past justifiable doubts have been cast on the
authenticity of various stringed instruments that were
being sold as valuable or historic pieces. It has as a result
become customary to use dendrochronological methods
to verify information given about an instrument as to who
made it and when it was made.
The following example makes clear how these methods
are applied in actual cases:
Two purported Stradivarius violins formed part of an
inheritance. In view of the high value of such instruments
and justifiable doubts about the authenticity of these two
violins, expert opinion was sought. No standard curves
for annual ring widths were available for spruce in the
southern Alps, this being the most likely source of the
wood for Stradivarius. Nor was it feasible to construct
curves from instruments known to be genuine. And so the
sounding-boards - which are made of radial sections of
wood - were X-rayed . Recent work had shown that
standard curves for maximum density were valid for
practically the whole of the Alpine region . As Stradivarius
worked in Cremona, in the north of Italy, it was clear that it
would be possible to date the violins by a comparison
with the curves for the central Alps.
Dating illegal fellings of tree
A considerable number of cases of illegal felling and of
theft of timber come before the courts in times of wood
shortage. Many such cases occurred in Europe in the
period after the Second World War. Today it is more often
a case of finding out when a tree at the edge of a wood
was felled by someone because it was too near his
property or otherwise in the way.
An example:
A fine larch was felled without the knowledge of its owner.
Chips of larch wood were found at a neighbor's house
and their annual ring width pattern was compared with
that of the stump of the felled tree. The tree had been
released a few years before and had immediately begun
to grow more quickly. The resulting characteristic annual
ring pattern was also evident in the wood found at the
suspect's house (Schaefer 1954).
stump
(((({(((((((((((((((((ct ( .f -e «((~

«({«m! an
The results of the analysis were unequivocal. The latest release
measurable ring from the first violin had been formed in
1902, that from the second instrument in 1894. Since corpus delicti
these were not even the outermost rings of the original
tree, and since a period of seasoning of the wood had to Comparison of a tree-ring sequence of a tree stump with those of a
be allowed for, the violins clearly could not have been corpus delicti. The similarity of the patterns is an indication of the
provenance of the questionable object.
made before about 1910. Given that Stradivarius lived at
the turn of the seventeenth century, the instruments were
undoubtedly fakes.
The kidnapping of the Lindbergh baby
In this kidnapping case a simple homemade ladder was
used. In 1935, Arthur Koehler, who had worked on the
case for four years, was able to show that the ladder had
been made out of wood from the floor-boards in the
house of Richard Hauptmann. This finding was based
primarily on the analysis of the tree-ring patterns of wood Part of the evidence used in the Lindbergh baby case. Right: the
annual ring pattern of a piece of the ladder found at the scene of the
shavings found in Hauptmann's work-shop, and provided kidnap. Left: the pattern of a piece taken from the attic floor in the
the burden of proof which sent him to the death-chamber house of the suspect, Richard Hauptmann (Courtesy of the Forest
(Christensen 1977, Palenik 1983). Product Laboratory, Madison, U.S.A.).

111\11 ' ... .--.TTr

Xray picture of the bridge of the second of the suspec t St-;;;cii~;a-ril~s violins .
.;cl"

..\ I·,· . I'

.\j\. . I -. . .'
. . .
. l ;_.. ... ...,. .
.-~
j \ , ./\
1·'./· !\ I · t'..1 . . -..
... "...,..
/\.... "'\ ..'\J\ V • i ' 1 -, •A/\ ."\.../'1 .. ".J·V·
o' \ ••

, \.:.1\ ......
'".
\
••
'.\
viol In
1894
violin 2

I·'·,r\j'\. •,'",.,...'. standard curve

1902

~
o
...\ •.\. ,f\...
~
./.~.,...
••
......\.1\ ..... .
t' \ /..... ~..,
./ • • 0,..1 \• I ..
0'. , • •..,
•
•
,/ ~..~.t\t\
•.
~\I .

. .
j\
• ., ,.. e·. \ .
... .
..~
~. •

40 1850 60 70 80 90 1900
Mean curves for the maximum densities of the two suspect Stradivarius violins compared with the standard curve for spruces in the Tyrol.

Klein, 1984; Christensen, 1977; Palenik, 1983. 167

Tree-ring research in climatology
Climatologists have for some time been aware that an
understanding of the climatic conditions over the last few
decades -. or even over the last 200 years - by no
means provides a complete picture of the wide spectrum
of variations occurring over the last 10000 years. Nor can
weather data for these relatively short periods of time
provide an adequate basis for an understanding of the
global weather machine (Hecht et aI., 1979).
It was as recently as twenty years ago that economists
and politicians became aware how closely the fate of
mankind was linked to the global climate. Neither the
'green revolution', i.e. the breeding of more productive
crop strains, nor the manufacture of fertilizers in massive
quantities, nor improvements in transportation enable us
to become independent of the climate (Schneider, 1978).
Catastrophically cold wioters with accompanying heavy
snowfalls In U.S.A., devastating droughts and floods in
Africa and Asia, and resulting lost harvests, all serve to
call this fact to our attention. The occurrence of this sort of
natural disaster has led to a considerable amount of
palaeo-climatological research by dendrochronologists
all over the world who are working on the problem of
trying to provide a basis for the understanding of large-
scale climatic changes.
No other living organism can provide as much informa-
tion of interest to palaeoclimatologists as the tree, with its
annually del.ineated growth increments, the tree rings. The
three most Important areas in which data gained from the
analysis of tree rings can be applied to palaeoclimatology
are described below.
1. The extension of meteorological records:
Using tree ring series from trees of different ages and
OUler space
Atmosphere
TeU8sIfIst
fad.al4on
HJO. Nh OJ. CO" 0) etC
Aerosols
Interaction atmosphere earth surface
Land
Ice,
Changes In Changes In form of
81mospnerrc landmass, orography.
composJllon vegetation, albedo. etc
c::::::::::> Internal processes External processes
168
species and from a range of sampling sites, shorter
meteorological records from various parts of the world
(including the less populated areas) can be extended.
2. Lending perspective to recent meteorological data:
Dendrociimatological findings provide a background
against which recent data can be evaluated by com-
parison with historic and pre-historic times. This in turn
encourages us to question the notion that events in
our own times provide any sort of absolute norms.
Chronologies which relate to the distant past can, for
example, illustrate how quickly changes in climatic
conditions may occur and how long different climatic
phases may last. The analysis of tree ring data enables
us to put events of our own times in perspective and
to develop a picture of what is 'normal ' .
This is particularly important if we consider the
implications of our current attitude toward our environ-
ment; that is our unthinking exploitation of reserves
of fossil fuels and resulting rise in both the carbon
dioxide levels in the atmosphere and the acid content
of rain . Such events are most probably inducing long-
term changes in the earth's climate.
3. Determining the causes of long-term climatic fluctua-
tions:
The analysis of long tree ring series can contribute
considerably to our understanding of the causes of
long-term climatic fluctuations. It may be that our
understanding of such changes will be improved by
the examinatIOn and analysis of wood that has been
preserved in ice and on the sea-bed.
Change of
sola' radiation
CloudS preclpnalron
Diagram illustrating the climatic interactions
that affect the earth; the internal factors
Ice Heat el(change Evaporation (shown by the white arrows); and external
processes (black arrows) (Mitchell, 1980).
Ocean modificatiOns E.rlh
In lorm, salinity etc
Hecht et al., 1979; Schneider, 1978.
Ways of obtaining climatic information
Some tree-ring material already exists from which a
picture can be built of wide ranging climatic patterns. This
includes data from sampling networks in the west of the
United States and the Soviet Union and in Europe. One or
two very long tree-ring chronologies for the same geo-
graphical areas may exist and further add to our knowl-
edge. From such material climatological information
covering wide areas and long periods of time can be built
up step-by-step.
The first stage: constructing growth maps
In 1965 H. C. Fritts began to build a sampling network of
trees on arid sites in the west of the United States. He
calculated the mean standard deviation of ring width for
10 year periods over approximately 400 years for each
chronology and then mapped the deviation from the long
term norm for each ten year period. This method allows
the varying patterns of tree growth to be clearly seen. As
the trees under investigation were growing in dry areas, it
is supposed that these maps show fluctuations in precipi-
tation over time and space.
The present author is constructing a sampling network
for trees growing on boreal and subalpine sites in Europe
and North America. From this network, he intends to
ascertain those areas for which, for any given tree-ring
feature, a trend is discernible. Maximum density in these
tree-rings exhibits a considerable uniformity over wide
areas. Because maximum density is related to growing
season temperature, maps may be constructed to show
annual spatial variations in temperature. Because site
conditions are so important to a tree's response to
climate, little reliable climatological information can be
obtained from maps for which the site characteristics are
not known. Only those maps derived from networks of
homogeneous sites are useable for climatological inter-
pretation.

1.... ~ : ~I_'_-

Growth maps for trees on and sites. The numbers show the 10 year mean standard deviations from the average annual ring widths for the
period 1651 to 1920 (Fritts, 1965).

Interval trend maps for trees on boreal sites. The vertical lines indicate those regions in which the density increased on an annual basis, and
the horizontal lines show those regions in which it decreased. The three maps reproduced here relate to the extremely cold years 1816 and
1912, and to 1911, which was warmer than average.

169
The second stage: obtaining climatic information from
individual series
tends to be limited by only one factor. Precipitation is the
Every tree ring chronology contains certain climatic infor- main limiting factor on trees at the lower semiarid timber-
mation. As a first step towards finding out what type of line, while at the cold moist mountain or Arctic timberline
information this is , the chronology must be measured cell wall growth in the latewood is limited almost exclu-
against reliable meteorological data. Where a considerable sively by the summer temperatures. The chronologies of
number of individual chronologies are to be combined to the maximum densities reflecting cell-wall growth in late-
form a single series, the quality of the climatological wood for conifers in Scotland have been shown to be
information provided by each of the series must be known integrators of temperature by Hughes et a/. 1983 and
in order that the value of the network as a whole can be utilized in correcting the historic temperature series for
determined. Homogeneous networks generally make use Edinburgh.
of samples from extreme sites where cambial growth

A boreal conifer forest. Characteristic for this type of stand are the Trees at the lower timberline in Arizona. The wide spacing between
very slender crowns. the trees is typical for this type of site.
prior
tempera ture precipi tation growth JJASONDJFMAMJJ
0.4

~ S W S
~
3
'"
&
0.2
E
2! 0.0

c 0.2 £ 0.2
.2
,.
§ e 0.0
'0.
'0
0.0 '"
(;
.,.-
ot
Q)

0.
J J A 5 0 N D J F MAM
,~ 1 2 3
-0.2
year prior year of lag
to growth growth - 0.4 (yrs.!
I! ,!, ..L...o....-I....LJ. ....
month
JJa$O"'~J' troII "''' JJ '' S
Response function diagram for tree-ring widths for trees at or near
year prior yea r of year prior year of lag
the lower dry timberline. Here growth is limited primarily by precipi-
to growth growth to growth growth (year)
tation in the previous winter and in the current growth period (Fritts,
month month 1976).
Response function diagram for maximum latewood densities for 1.0
spruces in the Swiss Alps. Here growth is limited primarily by the G
summer temperatures (Schweingruber et al., 1978). ~ 0 .5~

E
'"
c. 0.0 ,l ,., • j

a 5~
Q) •
fet:(.In~HuctIOf1 reconStruCtiOn "0 _

,
., ·3
., .,
·3

., E 100 '

..
E
8.
~

mean me.n
E
~ rJ
~
a "
E
..
~,
2i

\ '"c.
~ - 100 '

r-;;, ---,- 1600 1650 1700 1750 1800 1850 1900 1950
,-..:;0 XI 00 00
year

Temperature reconstruction for latewood density and ring width Temperature and precipitation series for the Central Plains recon-
series for conifers at the upper timberline of the Swiss Alps (thin structed from ring widths (curves) and the actual meteorological
line), compared with actual temperatures for July and August data. Here the values have been smoothed using a low-pass filter
(Schweingruber et al., 1978). (Fritts, 1983).

170 Fritts, 1974, 1976, 1983; $chweingruber et al. , 1978, 1980.

The type of climatic information that can be obtained from
tree ring width and density chronologies from trees in
temperate areas differs considerably from site to site.
Depending on the prevailing climatic conditions, the
growth of the trunk may at one time be limited by precipi-
tation and at another by temperature. Response functions
are used in order that the main factors limiting growth may
be isolated. It is however characteristic of response func-
tions that the climatic patterns indicated by them are valid
only for particular years. Fritts (1984) analyzed 124 tree
ring chronologies from the American west, and found that
even in that predominantly semi-arid climate an enormous
variety of growth reactions could be observed. In almost
every single month for which growth was examined the
climate had influenced growth in some way. Since the
growth on sites like these is influenced by more than one
factor, the climatic reconstructions done on the basis of A fir-tree forest in the temperate zone (Emmental, Switzerland),
information from these sites may be less accurate than displaying characteristic wide crowns and dense canopies.
those based on extreme sites.

5 D.' 6 D.' 22
~ c.?
3
::A:::d ["
2!. 0.0

1 \ J E
~ -0.2

-.'
s w s
-0, .
w S w S
1
0 .• 0•• 0. 4 ., ".' 0 .•
c
~
--j g
~ "~\
C .<:
D.'
..
L
~ .2 ~
D.

e
Cl D.O -
§
'6. c.c
e '.c
Cl
~
'0. .;.
• .< 0, . 1 .
0 .,
'0
-0.'2'
0 -o.?
.~
'0
t':'
~,
.Q .:-1
'~
0.
[).~
C. C.

c. · j ·D. ~ -( ., Q. (,. c 1

Average response functions for tree ring widths of conifers in the
south-western U.S. The following factors are most favorable to
growth:
(a) Above average temperatures in April associated with below
average temperatures in May and June. High summer and winter
precipitation. (b) High temperatures in August of the year prior to
growth and low temperatures in autumn, winter and spring. High
precipitation during most of the previous year and low precipitation
during the growth period. (c) High temperatures in summer and low
temperatures in winter and spring. Low precipitation in June of prior
year and high precipitation during late winter and early spring of the
year of growth (Fritts, 1974).

Independent alibration

A comparison of the July/August temperature series for England in Scotland. It is clear that the meteorological data obtained before
(i = Edinburgh, ii = Central England) and a reconstructed tempera- 1800, (e.g. 1730 and 1n9) was not always accurate. (Hughes et a/.,
ture series (iii) derived from the maximum latewood widths for pines 1984).

Fritts, 1974; Hughes et ai., 1984; Serre, 1973; Bednarz, 1986; Till, 1984. 171
The third stage: the reconstruction of climatic conditions
over large areas
Climatic conditions over large areas can be reconstructed
when information is available from networks of tree ring
series and meteorological series. As Fritts et al. (1979)
have demonstrated, the use and general validity of the
network may extend far beyond the area covered by
the tree-ring series. This is so because the existence of
large-scale atmospheric circulation features and the close
relationship between climatic conditions in areas that may
be some distance apart. Where such networks are avialable
the tree ring series can be calibrated and models for
reconstruction developed. The atmospheric pressure dis-
tribution over the Western Pacific can for example be
reconstructed from tree-ring networks from the South-
western United States (Fritts et a/. , 1979).
Fritts et a/. (1979) used 65 tree-ring chronologies from
eight different species of conifer from the west of U.S.A.
and from Canada for 1660 to 1970. Schweingruber (1985)
used 123 maximum density chronologies for eight dif-
ferent conifers from Europe.

WINTER
1899-1961

PERCENT VARIANCE
CALIBRATED

PRESSURE 53.3%
TEMPERATURE 54.1 %
PRECIPITATION 50.1 %
~o .-:,- ..
TEMPERATURE

The percentage agreement between reconstructed tree ring-climatic series and measured data for atmospheric surface pressure, tem-
perature and precipitation in winter (December to February) for the period 1899 to 1961. Such maps provide a good indication of the
reliability of climatic reconstructions using tree ring-climatic series. (Fritts et al., 1979).

~
' Vi
c:
Q)
"0
E
::J
E
x
"'
~

Comparison between temperature deviations in the months July to September - average values of the years 1881-1980 - (Jones et al.,
1983) and deviations in maximum densities in conifers during the same period. Patterns from the years 1928-1931 are illustrated. Black:
below average temperatures and density values respectively. White: above average values. By showing the similarity between actual exam-
ples and Climatological theories, both parameters are justified.

172 Fritts et al., 1976, 1979, 1984.

For the period 1899 to 1963 the relationship between A comparison of temperature reconstructions for the
actual meteorological measurement series and the in- whole of the U.S. shows that correlations can be drawn
dexed tree ring values was calculated using regression for certain periods with temperature developments in the
analysis and models were then constructed. The rela- Northern Hemisphere, but tree ring chronologies covering
tionship between the actual (or measured val.ues) and a limited area are inadequate for this purpose.
the dependent tree -ring-derived climatic values was The reconstructions of temperature for the northeast-
expressed in terms of associated variance. The calcula- ern United States and of precipitation for the northwestern
tion of the variance in common indicates the reliability of part have proved to be generally reliable. They may,
the climatic reconstruction for any given measurement however, be less accurate in particular cases. The recon-
point. The variance diagram shown here suggests that the struction of climatic conditions in the winter 1899/1900
pressure distribution over the western Pacific has large shows, for example, that such reconstructions must be
correlation with climate (and the tree rings) over the somewhat generously interpreted. Here the reconstructed
western United States. Today it is no longer a question of temperatures tally with the actual ones only insofar as
whether, but rather of how we are affecting the earth's they show that the north-west had a colder, and the
atmosphere. In order to determine this , it is crucial to south-east a warmer winter than was normal for the sixty
have a picture of climatic conditions over the last several years used in the calibration. Similarly, reconstruction of
centuries, so that any changes caused by man may be precipitation must be viewed with the same discretion.
identified. Fritts and Lough (1984) applied a tree ring
network to reconstruct the past average annual tempera- I.,

ture and the pressure distribution for the whole of the ' .0
United States. Using multivariate transfer functions they
succeeded in estimating the variation in past temperature
and surface pressure for North America from the mass of
data available. They then compared the estimated data
with independent measured data sets for a number of
areas in the Northern Hemisphere. The best agreement
seems to exist between tree ring widths from the semiarid
western area of the United States and pressure distribu-
tion over the whole of the North American continent.
,.,
,.
,.
,.0
a) °c c) °c
0.'
0.0
-o .S
*0 .5
- 1 .0
-O"i-_ _ __ _ _ _ _ _ _ _ _ _ _ _ _--:--[
.
-1.0

,., ,
I..
,.0
b) Index I.'
0.'

..•.•.j - - - - ---'-'--- - - - - - - - - - - - :--r d ) mb •.•

-0 .5

e) Index
,10

,to
- 1. 0
, ... "SO '''''' nso 1100 'ISO ... ,...
,

..
, ~~~~~~~~~~~~~~~~~~~~ YEARS
Uni ts eo

.....,
• ~ j-----''-'---
Reconstruction of deviations from the average values for the period
1600 to 1960 (T): (a) annual temperature in the United States and
d) °c ~~+-~~~~~~~~~~~~~~~~~~ south-west Canada (b) annual temperature in the west of the United

........t-_ _ _ ___'"""'_'_.....
-0 . 4 States (c) annual temperature in the east of the United States, and
(d) of the atmospheric pressure at sea-level (mb). The warmest and
the coldest 30 year periods in this span of 360 years occurred as
,.
_H _....
__ . _lAHO
. -".:... _ S_If_""_ T_ _ _-;-i
follows:
warmest coldest
c) °c periods periods
. -.:.: the United States as a whole
the west of the United States
1918-1947
1637-1666
1877-1906
1898-1937
CUfTiJIIA,l CNGI.- ... ND T
the east of the United States 1925-1935 1666-1695
•.1x,' •. ·l.A;o' • • '.,Lo' • "I ., IIOCI 'I~' •• ', . ' •• ".so (Fritts and Lough, 1984)
YEARS Fritts and Lough have demonstrated that correlations exist between
Comparison between (a) reconstructd average annual temperature the U.S. and the Yukon temperature curves from the mid 17th to the
for the U.S. and (b) indexed values from annual ring series from the early 18th centuries, and generally for the early 19th to the 20th
Yukon (Canada) (c) from eastern California (Bristlecone pines at century. There seem to be no correlations, however, for the mid 18th
elevation 4000 m) (d) reconstructed temperature curves for the century. Each of the series appears to show good correlation with
whole of the Northern Hemisphere (temperature measurements and certain parts of the reconstructed temperature and pressure curves
annual ring series) and (d) temperature measurements for Central for the U.S., and poor correlation in other parts (Fritts and Lough,
England (Fritts and Lough, 1984). 1984).

173
The search for cyclical patterns
It has long been the desire of man to be able to predict
long-term climatic conditions and changes. For centuries
astronomers and climatologists have attempted to identify
cyclical climatic patterns. It was this possibility which led
A. E. Douglass to lay the foundations of dendrochronol-
ogy (Webb, 1983). He devoted much of a lifetime's work
to the attempt to discover a connection between tree
ring widths and sunspot cycles, using long tree ring
sequences, e.g . for Californian Sequoiadendron gigan-
teum . It was only with the development of computer
techniques and modern statistical methods that the weak
indications of the double sunspot cycle of 22 years could
be discerned.
In recent times a number of people have examined
various time-series for periodicity, and have achieved
results similar to those obtained in earlier analyses, effec-
tively identifying cycles of differing lengths. Stockton
(1981) found evidence of the 22 year Hale magnetic cycle
in the drought area index (DAI) for the Great plains,
U.S.A. The DAI had been calculated from a considerable
number of tree-ring width chronologies (Mitchell et a/. ,
1979). Bell 1981 has found indications that the same
chronologies partially reflect an 18.6 year lunar nodal
signal.
In evaluating the findings described here, it must be
borne in mind that many researchers over the last forty
years have found indications of cycles of quite differing
lengths, and that even the more recently discovered
indications of the 22 year cycles are extremely weak and
are by no means discernible in all the tree ring chronol-
ogies that have been examined. It has not so far proved
possible to trace with consistency the long term sunspot
cycles in individual tree-ring width chronologies. How-
ever, the carbon-14 content of the cellulose provides an
indicator of this periodicity (Stuiver and Grootes, 1980).
The results of this aspect of dendrochronological research
are still very controversial.
The fourth stage: climatic conditions in earlier times
Almost every dendroclimatological study aims to shed
light on climatic conditions in the past. Large-scale clima-
tological reconstructions extending as far back as the
fifteenth century have been successfully carried out using
data from old trees growing in semi-arid areas, for exam-
ple, the south-west of the United States (Blasing , Fritts,
1976). Trees in the boreal zones live less long, on
average, than those in semi-arid areas. Thus it is unlikely
that large-scale reconstructions for these zones can be
30 (e) posi < - I
20
1700 1750 1800
Drought Index chronologies (vertical bars) calculated from tree-ring width and showing the area in drought in Western U.S. with a Palmer
Drought Severity Index. The lower wavy line on each chronology shows smoothing with a band-pass filter, the upper line with a low-pass
filter (Mitchell et al., 1979).
174
carried out extending any further back than the sixteenth
century. It should prove possible to reconstruct the dis-
tribution of precipitation in the south-west of the United
States over the last 2000 years using series which have
been built up from archaeological data (Rose et al., 1981).
Using Bristlecone pine chronologies for trees at the upper
and lower timberlines, dendrochronologists have suc-
ceeded in reconstructing both temperature and precipita-
tion on a decennial basis (La Marche, 1974).
The reconstruction of climate on the basis of tree-ring
chronologies is fraught with problems, as is clearly shown
through the comparison of sequences:
The density of coverage is decisive for the interpreta-
tion of the curves. The lower it is, the greater the
amplitude of the curves . Variations in thinly covered
curves may be due to either differences between
individu als or between sites.
Differences between species are reflected in sensi-
tivity. Maximum density in spruce and fir is less
sensitive than that in pines or larches (Laueneni
Tornetrask). However, the less sensitive species
contain just as much climatological information as the
others.
Differences between sites are also reflected in sensi-
tivity. For instance, the sensitivity of larches on sites at
the upper tree limit (Hbhenbiel) is greater than that of
larches in the lower subalpine zone (Grachen).
Comparative studies on given calibration periods have
shown that the curves under discussion reflect tempera-
tures in the summer months; in the Alps, those of
July through September, in Lappland those of July and
August.
Despite all these reservations, the following climato-
logical conclusions may be drawn:
The amplitude of fluctuations in summer temperatures
over the past 1500 years lies within the same range as
that of the past 150 years. The meteorological mea-
surements of the 20th century do not reflect the entire
range of variation in summer temperatures .
Warm and cold phases alternate fairly rapidly, though
the phases are of differing durations . .
The transition from one phase to another is as a rule
abrupt, taking only from 2 to 10 years. Gradual
changes are more seldom, e.g. , Hbhenbiel 6, 3200-
2800 BP.
The amplitude of climatic fluctuations since 500 A.D. is
relatively small.
1850 1900 1970
Mitchell et al., 1979; Webb, 1983; Douglass, 1971; Stuiver, 1980;
Stockton et Michell, 1979; Currie, 1982; Stuiver et Grootes, 1980.
 Lauenen Alps, Picea abies/Ables alba
.1g,' cm' 982-1976AD

,
OU2
;1j _______ I~ ______•. ____________ mean \
•
An....
• ...
1\
••
f\ A
~~ .
"f'\ 1'\ 1\ n ~ " • C\
~.... ...
~
.,.- - 1 c.. clcr.loJUfIQ ......
Tornetrask. Scand.navia. Pinus sylvestri. .1"',;1 t w"I.IC
500 1980AD n,f c;mc,_'
o Q.l~
u2 t n 1)9:-) V'.!,Jrs
o
12
n·J
.1 ,

I
=:; :: I
ooverdge Torneuask

J : :=: I I::' :
of chrono· " 12 '0 9 •• e tr 11 11 t1 Ib h >

IOg'~'me ~ ,(I(
;
600
:;
11 0 900 100 11{ • 1200 1400 l' I)() 1600 1700 I! 0 ":}'lO
=
......... AD

,nd,(f>S,
~ gle;.m)

0Q.l1
g02
002 G1~'Cpldufgl.~'1
1;17%
,nUlces
.l glcm'
t·vah",
004 116%
~O?i " • 1198 years
002j

} Grachen 61902
coverage I 1 Hohenb,el 6 2 3 5 3 2
time I ! I I ! I 1 I " t ! !

]I:--MlO J800 31rn 3600 J500 3400 3300 3200 JIOC .JOI ~I 2&M '1~ 10 16 75.... 24 BP

1M"
.l fer,)

) )41
1 )2

~o;- Gle,chlduhgke,t.
582
,11·1 I-value
.l t~'~ r'll
03%

o
002
00"1 n ~ 1220 years

coverage
.1 Hohenbiel 3 6

Fluctuations in the maximum density of conifers from the subalpine zone of Switzerland (Lauenen, Grachen, HOhenbiel) and the boreal zone
of Swedish Lappland. Longest chronologies for the period 5900 BP to the present. The dating of those from Lauenen and Tornetrask is
absolute, on a dendrochronological basis; that of those from Grachen and Hohenbiel semi-absolute, i.e., based on the radiocarbon method.
The indexed annual values have been smoothed by means of a binomial filter with 31 weighted components. The mean values are based on
the partial chronologies. The chronologies stem from the following sources:
Lauenen, Bernese Oberland, Switzerland. Some samples from other regions. Material from buildings. Schweingruber et al. 1979.
Tornetrask, Lappland, Sweden. Stems from bogs. The material was kindly placed at our disposal by T. Bartholin, Lund.
Grachen, the Valais, Switzerland. Stems from a small alpine lake. Schar and Schweingruber 1987.
Hohenbiel, Urserental, Switzerland. Some stems from the Valais. Most of these stems were found in a bog above the present timberline.
-...j
(J1 Renner 1982, Bircher 1982, Rothlisberger 1976.
Tree-ring research and geomorphology
Dendrochronology can provide important information
about the changes which have taken place on the surface
of the earth. Whenever an event has an effect on trees or
forests this is reflected in the structure of the shoot and
Reaction mechanisms in trees
The tree provides a good example of an organism which
has adapted well to its environment. Since this environ-
ment has been subject to considerable change in the
course of evolution, the tree itself has developed into a
flexible organism: its living tissues are capable of adapting
to or reacting against changed conditions.
A number of parts of the tree are capable of such adapta-
tion:
The shoots
Adventitious shoots are formed where it is necessary
to change the direction and rate of growth. The shoot-
system may also form roots.
The roots
Can react in the same way as the shoots and are capable
of taking over the functions of the shoots.
Those tissues which are capable of division
In particular the cambium: these tissues can accelerate or
Adventitious roots and shoots formed by the cambium of the root
and shoot systems.
176
the tree ring. As trees put on an annual growth increment,
those changes can be dated. In many cases it is possible
to determine both the intensity and the duration - be it
days or thousands of years - of such occurrences.
retard the production of new cells; the offspring are able
to form other types of cells, such as those comprising
callous tissue or resin ducts.
Living cells
These can react to changes in the environment through
slower or faster cell wall growth and through the formation
of special substances, such as those which comprise
heartwood and compression wood.
Dead cells
Are the source of information about how and why an
organism has died. Depending on the amount of stress to
which a tree is exposed, either individual cells, or the
whole organ concerned or even the tree itself will die.
Each of these reactions can be isolated and dated.
The way in which wood has decomposed
Indicates something about the conditions to which the
tree was exposed after it died. It is often possible to
determine when decomposition occurred.
The formation of large and small cells showing the different thick-
ness of cell walls.
 1913
~
192~
-=-
--:--~.
•
!

..-

'r

1966

1975

The reaction of an individual tree to different environmental factors (Wald and Umwelt).

o
Callousing over a part of the trunk that was damaged and had begun A trunk which is half buried. As the part which is actually in the soil
to rot. The year in which the tree was damaged and the infection has been well preserved, the year in which the tree was felled can be
occurred can be determined. determined.

177
Types of damage

Although there are a considerable number of possible _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _,

causes of damage to trees, it is possible to group them
together as follows:

Damage to tissues which are capable

of division, such as the cambium,
and the tips of the shoots and ~
roots.
Change in the original position. ~~

Changes in the hydrological

conditions. ~

Change in the supply of nutrients ~ ~

and oxygen.

The location of wood finds Dead

The location of a wood find is decisive for dendrochrono-
logical studies. The more that is known about the environ-
mental conditions in the area and the time scale involved,
the more significant are the results of the analysis. Wood
is found in the following places and conditions:
In situ, living
Trees growing on the sampling site.
Stumps in peat-bogs and lakes, on sea-shores, in silted-
up river beds, on the edge of the paths of avalanches,
and under volcanic deposits.
In secondary locations
Driftwood on lakes, sea-shores and sea-shore terraces.
Sunken driftwood in lakes and peat-bogs; trunks em-
bedded in gravel in alluvial deposits and moraines; pieces
of wood in avalanche deposits.

I n situ: st ump embedded in

sediment.
Secondary position: trunk
embedded in sediment.

Secondary position: driftwood.

178
Geomorphological processes which are reflected in the growth of a tree

There are a considerable number of environmental Earth movement:

changes which may affect the growth of a tree . The the rise and fall of land masses, landslides, slope
change may help or hinder growth in some way, or even creep , rock falls , mudflows, erosion of the earth at root
cause the death of the tree . The following geomorpho- level
logical processes and occurrences can be traced in trees
which have been exposed to them (Shroder, 1987; Hydrological changes:
Timell, 1986): waves, currents, tidal amplitude, ice movements,
changes in the water-level
Volcanic activity:
ash rain, fires, lava flows, the damming up of lakes, Aeolian processes:
heavy rainfall, climatic changes loess deposition , sand dune shift

Snow movement: Human activity:

avalanches, snow creep, snow break
Ice movement:
the advance of glacier tongues into wooded areas, ice
pressure on the shores of wooded river valleys
the clearing of wooded areas may result not only in
direct damage to trees, but may also cause changes in
the run-off of rivers.

...... Volcanic activity: lava, ash rain .

... Snow: snow-creep, avaLttnches.

Erosion: washing away of the

surface soil .
... Rise and fall of land masses,
trunks on sea-shore terraces.

Erosion: undermining of river

... banks .
... Accumulation: covering of
existing surfaces.

179
Variability in the run-off of rivers

Large areas of the earth are subject to a basic shortage of
water. It is therefore particularly important, especially for
farmers, to know how often periods of extreme dryness
are likely to occur. In various parts of the world the
existing short river-flow series have been extended -
often with considerable success - using tree-ring data.
It was Jacob Keuchler, a German forester who had
emigrated to Texas, who was among the first to recognize
the link between tree ring width and rain-fall. In 1859 the
Texas State newspaper published an article by him on the
climate in Texas. Keuchler had ascertained that of the 134
years that he had examined, 19 had been dry, 11 very dry
and 12 extremely dry. Fifty-nine had been very wet, 11
wet and 22 moderately wet.
Recent work by Charles Stockton (1971) primarily on
trees in the south-west of the United States, has shown
that dendrochronology can play a useful role in hydro-
logical analysis. Jones et a/. (1983) have illustrated that
changes in river run-off can be reconstructed using data
for English oaks.
Rio Negro, Argentina
The Andes form a pronounced water-shed in Patagonia in
the south of South America. Around the main ridge, at
40 o S, the annual precipitation is 2000 mm, while 80 km
further east it is only 200 mm. Holmes et a/. (1979)
succeeded in relating sensitive Araucaria and Austro-
cedrus chronologies from the semi-arid eastern portion of
the distribution area for these species to the outflow of
the Rio Limay. The graph illustrating the close match
between the actual and the tree-ring reconstructed hydro-
graphs is impressive given how difficult it can be to
synchronize even a single tree with a chronology. A
number of different tree-ring features, for example width,
variations in the boundary between early and latewood,
variations in density and in the density of the tree-ring
tissue, must be taken into account in attempting such a
synchronization. The statistical calculations here show
that the annual ring widths reflect the annual run-off
very closely.

MAt OF .... ESTERN ARGE-NTt.~A SHOwl~C

LOCATION' or t~EUQL'1::::O A.'H) LIMY
RIVERS. STR..EA."fGA.liE SiTATIO~S, MiD
TREE·RltiC CHRONOl.OGY SITES
LEGEND

,",Ll
CH ILE .~o ...

L AR GE NTI NA

Location of the sampling sites in Argentina (Holmes et al., 1979).

Sampling sites for the Araucaria araucana at 1650 m in Nequen,

Argentina. The trees are up to 700 years old (Holmes et al., 1979).

I,f I ~.~
. .J I. ..
"1: ~" ~.~; ~\)
I. I

'"E 20000
"'"
~
~
::>
c
c:
'"
~
Rio Neuquen
§ 0
1880 1900 1920 1940 1960

Suitable tree-ring series from Araucaria araucana for reconstructing Calibration and reconstruction of annual run-off for two Argentinian
climate. rivers, using ring width chronologies for Araucaria and Astro-cedrus
(Holmes etal., 1979).

180 Stockton et al., 1980; Holmes et al., 1979; Kuechler, 1859; Jones et al., 1983.
Colorado River, U.S.A.

The 2317 km long Colorado River occupies a basin of annual amount of water which had been assumed to be
632000 km 2. The precipitation occurs mainly in the up- available was in error. Only in very few of the over 400
per basin and this water is of vital importance to the years analysed by tree rings did the Colorado River
people living in the arid southwestern area of North achieve the anticipated run-off of 16.5 maf (19800 hm 2).
America. In its travel from the upper to the lower basin The average annual run-off over the period of 450 years
water is used and re-cycled a number of times - for studied was only 13.5 maf (16200 hm2)
municipal and agricultural purposes and to generate elec-
tricity. The Colorado River compact of 1922 requires that
the states in the upper and lower basins will each get 7.5
million acre feet (mat) 9000 hm 2 and Mexico 1.5 maf
(1800 hm 2 of water each year from the Colorado River.
The analyses of the tree-ring data carried out by Stockton
and Jacoby (1976) showed, however, that the average

Sampling sites in the south-west of North America. 85% of the total A Douglas fir used in the reconstruction of precipitation patterns
precipitation fell in the area indicated by the shading. (Stockton and (Arizona). (Douglass, unpublished).
Jacoby, 1976).
30 uni f iltered
~
'" 20
E
c
.... 10
'0c
::l 1
'- O~ 1 r I

ro 1600 1700 1800 1900

::l filtered
c 20
c
'" t- • ~ ~
t' ~ + -,7'
~ 10
'"
::l
'-0
'" 0 .. I ~
T
1700
1
1800
r r 1 ~ r T
1900
1

1600
Colorado River at Lee Ferry, Arizona
The reconstructed run-off for the Colorado River. Seventeen tree ring chronologies for the period 1896 to 1966 were related to the run-off
values. A model was then constructed using statistical methods and the long term flow reconstructed this indicated that over the last 450
years the average flow of water past Lee Ferry was 13.5 mat (16200 hm2) (Stockton and Jacoby 1976).
Upper curve: annual values.
Lower curve: as above, smoothed with a 10 year running mean.

181
Recent changes in the course and flow of rivers

An understanding of the geological effects of water can

be gained by studying the morphology of tree rings, tree
age and the position of living trees growing near rivers
and streams which have an irregular run-off. It is primarily
the peak events and occurrences which can be identified
using dendrochronological methods. A number of exam-
ples of such occurrences are discussed below.

Changes in the course of a river

Natural alluvial channels of rivers are subject to a constant

interplay between erosion and deposition. Nearby older
stands of trees may be destroyed by floods whereas on
new deposits, broadleaves - in temperate and arid zones
these are primarily willow, alder, poplar and sycamore -
may germinate very quickly. Everitt (1986) took samples
from hundreds of poplars on the Little Missouri in North
Dakota and determined the age of the trees. He related
tree ring information to the channel morphology and
vegetation of the river-bed as it is today. Thus, changes
occurring in a river can be mapped from the age structure
of tree stands, and the development of floodplain vegeta-
tion can traced using data derived from tree rings. , Not well loenhfled
Isoch{ones

Construction of reservoirs The age structure of stands of poplars in the Little Missouri. This
illustrates how areas of level ground have become slowly silted up
over periods of more than a hundred years, and also that large areas
During the 20th century, many dams and reservoirs have of woodland were destroyed in a single flood. (Everitt, 1968).
been constructed on major rivers allover the world. This
has resulted in whole stands of trees being submerged.
The dead or dying trees may tower over the surface of the
water for many years. Trees which are growing on the
edge of areas where reservoirs have been built often
manifest sudden decreases in growth as a result of
the decrease in oxygen levels in the soil. Poplars alone
on the edge of a reservoir near Gunnison in central
Colorado, have recorded the first rise in the reservoir
water-level (1964/5) in the reduced growth pattern of
their annual rings. The rings became wider again in 1970,
indicating that the reservoir level fell in that year. There
was another reduction in the water level in 1980 which
was also reflected in the tree ring series.

Dead and dying poplars, Gunnison Reservoir, COlorado.

Tree ring series for poplars exposed to changes in the reservoir

water-level near Gunnison, Colorado.

182 Everitt, 1968; Helley and La Marche, 1973.

Woods that were once buried and have since been exposed
There are a number of submerged forests in the Four
Corners area of semi-arid northern Arizona (Dean, Univ. of
Arizona). In many cases only the tops or the trunkless
crowns of the dead trees can be seen sticking out of
these sandy plains. Working with geologists, archaeolo-
gists and pollen analysts, Dean has succeeded in dating
a considerable number of these exposed trunks. He has
charted the hydrological changes in a diagram bringing
together a large number of findings based on a wide-
ranging study of the river course. It seems that the
character of this dry landscape has been formed above all
by the occurrence of floods of catastrophic proportions.
The main valley of this Tertiary red sandstone is old .
The bottom sediment (w-2) was formed before 200 A. D.
Archaeological finds and carbon-14 dating indicate that
this level was colonized by five successive civilizations
(x and y), which were subsequently destroyed by flooding
or in some other way. Following 1150 there was a period
of erosion which caused the existing sediment to be
removed. In about 1400 A.D. junipers began to grow on
the original valley bottom (3). In 1700 very heavy flooding
caused them to become embedded several metres deep
(2). The trees which started to populate this new surface
were destroyed in the flood of 1880 (4). This level has
trees growing on it again today. The sediments z-1 and
z-2 must have been completely washed away sometime
before 1900, since trees were having difficulty surviving
on the inadequate deposits as early as 1910 (6) . It was
not until some pOint after 1940 that those sediments
which had been deposited between 1700 and 1880 on
the essentially steep and unstable slopes provided con-
ditions favorable to tree growth (7). Only the trees which
are on the upper slopes (1) are very old: they were able
to takp. root on the existing gravel deposits. These have
remained undisturbed for at least a thousand years.

Diagram illustrating sedimentation in northern Arizona In relation

to existing wooded areas and woods that are buried (Dean, un-
published).

Trunks which have been buried in sediment and then later exposed. It often happens in arid areas that heavy storms cause woods to
become covered up by sediment. These are later exposed by erosion. While the trees are embedded the trunks grow roots. Left: Pinus
pinea, southern Spain. Right: juniper, Arizona.

183
Rivers in prehistoric times
Oaks in the post-glacial time in Central Europe
There are thousands of fossilized tree stumps in the
Quarternary deposits of the large Central European rivers .
Such stumps are often exposed during the working of
gravel beds. As they fossilize they form a black substance
which is used industrially to produce a valuable veneer .
This substance also provides the dendrochronologist with
useful material for the construction of tree ring sequences.
A considerable amount of work has been done in this area
by Bernd Becker. In conjunction with geologists, archaeol-
ogists, and pollen analysts he has been able to build up a
picture of the making of the Central European landscape
(Becker, 1982 and Schirmer, 1979). It has also proved
possible to check and complement findings produced
The places where oak trunks have been found in the
Rhein, Main and Danube (= Donau) (Southern Central
Europe) (Simplified, after Becker, 1982).
~ argillic hOnlon I1I\IJ cambic horizon
I:;:".::'d fluVlsol {.=:"':::-:3 cleysol I --- I fossil logs
184
using radiocarbon dating by taking into account material
provided by the long annual ring series for Central Europe
and Ireland (Baillie et al. , 1983) .
Vegetation in river valleys
About 14 000 years ago the glaciers receded from the
low-lying areas. And in about 12500 B.P. the first pines
and birches began to grow in these regions (Kaiser,
1979). This type of woodland began to give way to the
mixed oakwood , comprising oak, elm, ash and sycamore,
in about 9500 B.P. It was not until the Middle Ages that
this type of woodland was in turn displaced in the low-
lying areas by poplars.
Oak trunks found in the gravel beds of rivers in Central Europe. The
branches are only very rarely present; stumps often remain. No oaks
have been found in situ. They have all been moved to another
location (From Griinhilder, 1972).
~ cambic horizon -s trongly / averagely/ weakly developed ~ fossil mollic horizon
Diagram illustrating river-bed deposits along the upper Main and the
Regnitz. 1 = Wiirm Ice-Age gravel ; 2-7 = Post-glacial and more
recent sedimentation (Schirmer, 1979).
Becker, 1982; Schirmer, 1979; Baillie, 1983.
Changes in rivers
The Rhein , the Danube and the Main once had the sort of
extremely varied flow pattern that unchannelized rivers
today still display. It is clear that climatic events set
geomorphic processes in train . These in turn are reflected
in various features of the oaks found in gravel deposits -
the state of preservation, the age and the tree-ring char-
acteristics.
The periodic destruction of these stands of oaks was
caused almost exclusively by flooding. The strong fluvial
activity eroded the wooded areas near the banks of rivers
and caused the trees to be washed away over consid-
erable distances. These trees, all of which had been
damaged to some extent, sank to ground level again as
soon as the water-level sank. It is clear from the very
good cond ition in which most of these trees have been
found , that they must have become buried in sediment
again very quickly.
The picture that has emerged of considerable numbers
of trees in different places being destroyed at various
times, indicates that there were factors at work whose
influence extended beyond regional boundaries. An un-
usually large number of trees died between 2000 and
1800 B.C. The Danube and Main regions then experi -
enced a period of recovery which lasted until about 400
B.C. The erosion phase of 1200 to 800 B.C. affected only
the Rhein . The most recent period of increased river
activity in the Main and the Danube occurred in about
800 A.D.

o so 100 ISO 200 tree ring

I
'50 ... I
~ '5O
I I
100

Donau- Blindheim
~.>JOAD

ft'l~,.nlon

3~ lMr)AO! 41~O:
L ----.: .......
dying years to-- rools Pf'esef"t!d -----t 5.lP\'oQOd ptll10trved
appro 6800 ± 160BP ::- ~tcrn ..... ,th erOSion d.J~ wl t houls.t~
- t( _I not ptt _r\-d
regeneration period
Growth diagram for oaks in the Danube valley. Finds at Blindenheim.
Trees died at different times (Becker, 1982).
Rulz&ndor'f bel W ltn

15
Growth diagram for oaks in the Danube valley. Finds at Rutzendorl.
The trees all died at practically the same time. It is probable that the
Rhein
oak woods in Rutzendorf were destroyed in a single flood, those in 10
Blindheim by floods occurring at different times (Becker, 1982). IC" "1'

IJ lj j d . j
H'O ep IBC I 2000 P' 4000
, ;,one
.,
&000 )CP
'" BC

.. ~
~ sP

Main- Regntt%
0
60810("

0 l
1000
l
BP lBC 'I
Aj
1fXXJ
.-.1.6 .....
13': ..
~jIi. ()I hOOI
-A
, B~

r--.---r--.---.---r--.----------~~ ""
Donau
'" SP

The deposition of oak trunks. Post-glacial deposition of oak trunks

in river basins in southern central Europe. The period in which the
heaviest accumulation of trunks occurred began in about 2000 B.C.
This accumulation may have been caused by the clearing of wooded
areas. The individual values were obtained as follows: the number of ____~____~OOOOBC
oaks accumulated per century was multiplied by the number of 00 8.
sites. This value was then expressed as a percentage for the last <II lJ.f ',r1 Unc;.,)I,t:,.fdl:1t'd
9000 years (Becker, 1982). : C J,,' IJ C Jd:~

Becker, 1982, 1983. 185

Site changes
From the age structure of trunk finds on various sites, the and sensitivity decreased, which was no doubt due to
course of river flow can be ascertained. As progressively more generally favourable environmental conditions.
more material was deposited in valleys, i.e. with a steadily The significant changes in the subboreal period seem
rising water table, erosion claimed older stands further not to have been caused by climatic changes, according
away from the normal course of the river. In the early to pollen analytical findings. It is much more likely that
post-glacial period trees were on average 150 to 250 erosion increased as a result of the clearing of woodland
years old whereas in Roman times oaks of up to 400 in Neolithic and Bronze Age times, causing a deep layer
years old were to be found. of clay to be formed in the valleys. This would explain the
In the course of time the character of the annual rings high growth rate occurring at the same time as a period
has changed. In the early post-glacial period tree rings of low sensitivity. This phase coincided with a period of
were narrow and showed a high degree of sensitivity. This floods which caused the destruction of older, higher-Iytng
could have been a sign of changing circumstances and tree sites. This hypothesis would explain the increasing
generally unfavourable conditions. Later, in the subboreal number of trees dying around 2000 B.C.
period from about 2000 B.C., the width of rings increased

<) DANUBE
400 - 0 0
0

:/#
350
'"~ 300
OJ
-5 250 u
o 0 ('
'0!Ij, 200 0 0 0
Cl 000'iI o 0
'" 150
100 .... ; date when trees died

400 300 200 100.BC.lA.D. , 00 200 300 400 500 600 700 800 900 1000

Changes in the age structure in single find complexes on the Danube. The diagram shows the age of oaks (ordinate) at
the time of their destruction (abSCissa) in the period 400 B.C. to 900 A.D. During the periods when material was
depOSited, the rivers have eroded older wooded valleys (Becker, 1982).

1000 AD Be 1000 2000 dOOO 5000 7000

JI ' ~
2.5

E
§ 2.0
L
i5
~ J
~ 1.5

1.0

l:
~
25 I Tree-ring widths of oaks growing in gravel of the Danube used as

~ ...
~ r)Qf'UII,I
paleoecological indicators. The height of the columns indicates the
~20-­ deviation from post-glacial average. Between 8000 and 4000 B.C.
:i!
c the annual ring width was narrow and the annual change (mean
~ sensitivity) high. In 2500 B.C. a change took place; between then and
!lenSIUIo"ty
15 I I
1000 AD B 1000 2000 3000 dOOO 5000 6000 7000 500 A.D. annual rings were wider and less variable (Becker, 1982).

186
Pines in the Late Ice Age in Switzerland

In the base of channels left by melting glaciers in the little

valley at DiiUnau, near Winterthur, Switzerland, stumps of
pine trees were exposed when clay for brick-making was
dug out. These were investigated by Kaiser in 1979.
During the whole period from Bolling to Allerod this area
was covered by forest. It needed only relatively minor
hydrological changes, e.g. blockage of water flow by
slope movement or a change to a wetter climatic period,
to weaken the resistance of the trees to such a degree
that they died. These short term changes can be seen in
a difference in the proportion of terrestrial and limnic
pOinter species in pollen and mollusc cross-sections as
well as in growth disturbances in tree-ring sequences.
Two chronologies have been constructed. The lower
one plots the earliest evidence for re-foresting in the late
ice age. Within 20 years trees were growing on the site
once more. The upper chronology in the Allerod period is
about 1000 years younger and probably represents the Cross section through a 12000 year old pine from Diittnau near
end of a longer period when the area was covered by Winterthur, Switzerland 20x (Kaiser, 1979).
forest. In this sequence the eruption of the Laachen
volcano in the Eifel region shows up in a series of nar-
rower rings. Could the missing link in the existing oak
chronology, which at present ends at about 9000 B.P., be
about to be bridged? (See also page 213.)

Molluscs embedded in clay at Diittnau (Kaiser, 1979).

" ~ > =-
~ .
• ---- \ /.<"""- -=- ~

II o~ I": ~tA \'>,----.. . =-:.----~---

r---.. - - - - - - - - - -
I

.;~~ , ~
-.l.......l-Ll" .. " .. , "
Simplified diagram of findings of pollen and mollusc analysis and
dendrochronology (Kaiser, 1979).

Kaiser, 1979. 187

Tree finds in peat bogs

Fossil trunks of oak and pine have been found in the peat
bogs and marshy areas of the maritime regions of north-
ern and western Europe. A good deal of dendrochrono-
logical work has been done on the oak remains, notably
by Baillie and Pilcher (Baillie, 1982) in Ireland and by
Delorme and Leuschner (Delorme, 1984) in the north of
Germany. The fossil pines have received relatively little
attention to date. Since they are in a remarkable state of
preservation it is to be expected that their radiodensito-
metric analysis will yield important palaeoclimatological
information.
The main applications of the analysis of the oaks that
have been found in these areas have been, first, the
construction of an oak chronology for Ireland for the
Holocene period; secondly, the calibration of radiocarbon
methods (Pearson et al., 1983); and thirdly, the derivation
of information pertaining to climatic conditions in the past
(Delorme et aI., 1983).
Where the trunks are found
This varies considerably. In those marshy areas of north-
ern Germany which have not yet been drained trunks
have been found beneath as much as 20 to 200 cm of
peat. Here the results of dendrochronological analysis
done on wood which has been well-preserved yields
information about the stratification of these areas. Where,
however, the level has sunk as a result of the peat-bog
being drained, solid trunks emerge on its surface. In such
cases it is not possible to relate dendrochronological
findings to the stratification of the area. Despite the large-
scale digging for peat that has taken place in northern
Germany, Leuschner and Delorme have succeeded in
measuring and analysing a considerable number of
stumps and trunks in situ.

The finds
The state of preservation of the wood varies enormously.
Those trunks which are in good condition still have their
sapwood intact, while in other cases all that remains of
the trunk are pieces of blackened heartwood. The tree
ring widths also vary considerably: some trunks have very
narrow rings - little more than the width of a pore -
while others have rings which are between 2 mm and 3
mm wide. Nor are the trees by any means all of the same
age: a good many of the trunks are 200 years old and one
or two of them are as much as 400 years old . The oldest
specimen found so far was a 450 year old tree. In some
cases the periods at which the different trees germinated
A huge sub-fossil pine stump recovered
and died overlap, while in others the two phases are from a peat-bog in northern Germany. Parts
clearly separate. of the trunk have been well-preserved
(Leuschner et al., 1982).

A tree-trunk which has emerged on the

surface of a peat-bog in Ireland (Baillie,
1982).

An oak stump in a peat-bog in Londonderry,

Ireland (Baillie, 1982).

188
Interpreting these findings

A picture is gradually being built up of the changes to
which peat-bogs have been subject. The information
necessary for such a survey is being gleaned from a
study of the vegetation and the topography of the bogs,
together with an analysis of dendrochronological findings.
The main factors affecting the growth of trees, especially
those limiting growth, are hydrological. In the case of
oaks, germination is not possible where the surface of
the ground has either completely dried out or is flooded.
Growth is also limited by the lack of nutrients in peat. This
can be most clearly seen in the narrow rings which trees
growing in such habitats tend to display. Frequent peri-
odic changes in annual ring width reflect changes in
drainage conditions: whenever the water level changes
suddenly this is expressed in an abrupt narrowing of the
rings. This can be the result of dryness, i.e. where the
water level has fallen, or of a lack of oxygen in the soil,
i.e. where the level has risen. It is generally possible to
relate the death of a tree to growth conditions in the peat-
B.C./AD.
100 0 100 700 300 400 !)()O 600
J ,
Kehdinger Moor
near Hammah
bog. Trees growing in the wettest areas die first, while
those on any slightly higher pieces of ground tend to
survive longer. This is the reason for stands of trees in
an area like this dying off gradually over a number of
decades. When the trees have died as a result of being
damaged in a storm or have rotted at the base of the
trunk, they quickly become subject to anaerobic condi-
tions provided that the peat-bog is growing at a suf-
ficiently high rate for the trunks to become submerged
fairly rapidly. Such trunks are extremely well preserved.
Where, however, a trunk lies on the surface of the bog for
some time it begins to decompose. In oak trunks which
have been exposed to the weather in this way the sap-
wood may begin to decay after as little as five years.
We can only start to speak of a general change in
climatic conditions - the onset of a period of unusually
high precipitation for example - when most of the trees
in a number of different such habitats die and very few
new trees in the new generation survive. The fact that the
chronologies of peat-oaks can be synchronized with
700 those of oaks beside rivers indicates that the tree rings of
trees growing on peat are a source of climatological
information. Today dendrochronological findings yielding
material which can be of help in building up a picture of
the history of this type of area can be dated with absolute
certainty, since this sort of dating is now based on the
7000 year absolute chronology for Western and Central
Europe.
175 - 350 AD

:::
="'=~::~~!~~~ near
Lengener Moor
Halsbek

St.idgeorgsfehner Moor

350 - (600) AD

• • after 675 AD

- T-
Sudgeorgslehn

,
~
~
• • . ,J,
6 ~

~
~
r r ,oil ,,,,
sphagnum peat
100 a 100 200 300 400 500 600 700
~.
"
B.C./AD.
date of germination 1 i (1 unit) ~ fen peat alder
date when
the trees died
! L:J mineral soil (sand)

A reconstruction-based on dendrochronological findings-of the

Above: the life-span of sub-fossil oaks from three peat-bogs in the
north-west of Germany. Below: bar-charts showing the frequency
distribution for the years when the oaks germinated and died
(Leuschner and Delorme, 1985).
development of trees and peat in Hammah and Kehdinger peat-bogs
in the north-west of Germany. 175-350 A.D.: the trees growing on
the damp alder peat are mainly alders, with a few oaks; on the well-
drained parts there are only oaks. After 350 A.D.: with the beginning
of the growth of the upland moor the trees on the alder peat start to
die off, followed by those on the knolls (Leuscher and Delorme,
1985).

Baillie, 1982; Leuschner and Delorme, 1984; Leuschner et al., 1985. 189
Changes in water table

Human activity has changed the level of ground water,
particularly in the 20th century, as a result of recovery of
water for drinking purposes, the building of dams, regula-
tion of river and lake levels and drainage of moorland
(Markus 1936). Such changes are reflected in the pattern
of the tree rings. The drainage of an area in the northern
Pre-Alps of Switzerland, endangered by erosion, will
serve as an example. In this area of high precipitation
(about 2000 mm/year), which is also subject to violent
storms, subalpine spruce and mountain pine woods were
over-forested in the 19th century. As a result slope move-
ment, erosion gullies and debris-covered areas were to
be found in the Flysch region, an area with a natural
tendency to heavy, sodden soil. When the level of the
ground water was lowered - it had previously often been
level with the surface - growth conditions for existing
trees improved and it was possible to give newly planted
ones a good start. Drainage channels at a distance of
5-10 m and a depth of 50-80 cm were dug accordingly.
GrUnig (1955) studied the effect of drainage on growth
rate and found that in this area of high rainfall , the results
were wholly positive. Young spruces, growing on the soil
dug out frDm drainage channels, grew well from the start
and this trend fell off only slightly, in line with normal
expectations, when the trees grew older. If the drainage
channels were neglected, water built up again and growth
dropped abruptly.

Flysch area in the canton of Obwalden. Switzerland. Because the Erosion gullies in Liitzenbach, Flysch, canton of Obwalden, Switzer-
sol/ is often waterlogged, this ateais comparatively unproductive land (Griinig, 1953).
and endangered by erosion (Griinig, 1953).

Core samples. from young spruces growing near drainage channels

in the Flysch area in the northern Pre-Alps, Switzerland.

(a) Spruces growing on soil dug out from

drainage channels. The growth rate was
good and regular.
(b) Spruces in the catchment area of the
drainage channels. When the channels
fell into disrepair, water built up again
and the growth rate of the trees slowed
down (Material from Griinig).

190
Trees react suddenly to drainage by immediately showing
growth improvement. The change from poor to strong
growth occurs from one year to the next in nearly all
trees. The reaction time of a stand, however, varies from
between one year and six. Presumably the physiological
performance is suppressed as a result of shortage of
oxygen in the soil before drainage. When the drainage
channels start to function and improvement sets in, this
can last for many years as can be seen in spruces and
mountain pine.
The growth improvements are clearly a result of drain-
age since trees in undrained areas show no similar im-
provement, as is shown by core samples. Tree ring
sequences provide information on the effects of technical
measures. In the immediate vicinity of the drainage chan-
nels, at a distance of 1-3 metres, the rate of growth
improvement was higher than at a distance of over 3
metres.

spruce
Neuenal p. Canton Obwalden, 1500m

spruce
cQ) 15 Sellbuhl, Canton Berne, 1700m 15

- n - 51
"0
"0 n = 30
~
'" C
.t= Q)
.'::: E 10 10
~ ~
~
(J
ea.
2 E
a. -
'" .t=
'0 'i 5 5
}J ~
E

o
oo 1910 1920
~ drainage drainage
1921
Growth improvement in spruces after drainage of an bog. Drainage took place in 1900 in Selibiihl, in 1921 in Neuenalp. The trees reacted
after about 1 and 6 years to the aeration of the soil. In Selibiihl all the trees reacted positively to drainage, in Neuenalp, 90% (Based on
Griinig's data).

drainage 1937

drainage 1920

drainage
appr.1900

Core samples from spruces in the catchment area of drainage

channels. The reaction occurs suddenly and lasts for a long time
(Material from Griinig).

Griinig, 1955. 191

The history of sea floods
Along the coasts of north western Europe, Heyworth
(1978) working in the British Isles, and Munaut (1967) in
Belgium, have found several large areas where stumps
and prostrate tree trunks of pines and oaks (Pinus sy/ves-
tris and Quercus sp.) had been covered in sediment.
These finds add greatly to our knowledge of the Holocene
period for the following reasons:
Flooded coastal pine forest in the West of England (Heyworth, 1978).
Remains of a forest of pines and oaks on the Belgian coast (Munaut, 1967).
192
from stumps in situ the relationship between sea level
at the time of their growth and the present day can be
found,
their age can be calculated by radiocarbon methods,
the course of their growth can be calculated by annual
ring analysis,
pollen in the surrounding sediments gives information
on the development of vegetation before, during and
after the time when these areas were covered by
forest.
Heyworth, 1978; Munaut, 1967.
The location and age of trees The growth of trees

Many sites with Holocene finds have been discovered on
the coasts of England and Ireland . Today the stumps and
trunks are lying either below the tidal level or in the inter-
tidal zone. Chronologically they span over 8000 years.
The trees died in a continuous sequence as a result of
rising sea level. What causes these fluctuations in sea
level?
a global rise in sea level?
rising of the land mass after being relieved of the
weight of the ice of the Ice Age?
tectonic vertical movements?
Such mechanisms are to a certain extent compensatory ,
for example rising of the land and rise in sea level, and
can be detected only if large areas of coast are in vesti-
gated. The findings of Hepworth suggest that the coast-
line of Great Britain has changed , mainly as a result of
vertical movements involving sinking in the southeast and
rising in the northwest.
Munaut (1967) studied the developments in a subboreal
coastal forest in Belgium. At 2.5 m under the surface of a
polder he found 56 pine trunks embedded in marl and
peat, which provided a ring sequence of 242 years . Pollen
analysis showed that sphagnum moss grew on this site
before the pines appeared , so the area must have been
wet. Then the ground-water level sank and over a period
of approximately 100 years a p n i e forest developed . It
grew normally for another century and died off gradually
during the next fifty years. High pollen counts for Erio-
phorum indicate a rise in water level.
An investigation of trunks and the sediments sur-
rounding them from various coastal areas from Portugal
to Scandinavia would certainly be worthwhile, as it would
provide much new information on vertical movement of
the land masses, dynamics of vegetation and climatic
development.
+

on
Radiocarbon dating of +
+
submerged forests.
Rising or sinking of
Y land mass in the last
8000 years in meters. +
+
+

+
+

+
+

+
+
+

+
+
+
pine +

Terneuzen . +
1962
+
+
+

+
+
+
+

+
+

o 120km o 50 100 150 200 250

Dated submerged coastal forests in the British Isles from the period
8000-450 B.P. The arrows indicate whether the land mass has
moved upwards or downwards since 8000 B.P. The figures in Circles
give the displacement in metres (Heyworth, 1978).
+ with bark
Bar chart of trunks in the submerged coastal forest in Belgium
showing number of rings per trunk. The forest grew up over 100 years
and then died off over 50 years (Munaut, 1967).

193
The history of glaciers

Age of living trees in the approaches to glaciers

Research into the history of glaciers in mountain ranges
and at the Poles has added greatly to our knowledge of
climatic conditions in the past. Valuable information about
glaciers and climate is gleaned from scientific data pro-
vided by glacier core samples, sediments from bogs near
glaciers, earth covered by debris and documentary
sources such as old pictures, photographs, maps and
written accounts. Recently wood finds from areas in the
approaches to glaciers have provided valuable informa-
tion. Fluctuations of glaciers in the Alps in the last 2000
years can be traced very accurately as a result of the
dating of living and fossil trunks and the evaluation of tree
ring curves.
Information about the recent history of glaciers has been
obtained by dating living trees in the approaches to
glaciers. When a glacier retreats, land is exposed in which
plants will grow after a certain period of time has elapsed.
When the age of these trees and bushes has been
calculated, the relationship between this and the state of
the moraine can be established. Heikkinen (1984) devel-
oped methods for studying the recent history of the
Coleman and Roosevelt glaciers on Mt. Baker in Washing-
ton State, U.S.A. These methods have not been used
very much in the Alps as many documentary sources are
available for the last 150 years in this area.
Determination of age is limited by various factors:
It is rare for trees in glacier approaches to be more
than 300-400 years old. Reconstructions are only
really possible for the last 200 years.
Often, older trees are rotten in the centre, making it
impossible to determine their exact age.
When core samples are taken at chest height it is not
possible to determine exactly how old the tree was
when it reached this height.
One has to be very lucky to hit the pith when taking
core samples.
The oldest tree in the moraine, which is always the
one to give the most accurate information, can only be
identified after 20 or more trees of the same thickness
have been analysed .
There is always an interval between land 's being ex-
posed after the retreat of a glacier and trees growing
on it again. Heikkinen , (1984) found that on Mount
Baker, U.S.A., this can be anything from between 6
and 30 years. Different conditions prevailing on the
sites are responsible for this disparity. On the Aletsch
glacier in Switzerland larches and spruces were grow-
ing again after 20 years on average, Swiss stone pine
after between 45 and 85 years.

40
CoIft"OfI _Mt Bah"
pr.,..fI' ,'udyl
......'
IBvrbo",
lilolA,.,
1981 J
S<Ol'ld,no. .... o (Korle" 1982'
4 e
_ _ 1018·70

"'"

_. .,,.
_ _ 19H _ _ 19 11
- - '9'"
- - a'9])
_ 911 " 1910

"
~
- o'~ -- '~11
'I«) -- 1~201
_ _ 1816_11
., 1180 fa _ _ .. 00'"

'I!I~- -- D e")-SO
_ 18~1 ., _

_
1161\13

'1!:5055
_ _ IISO 00

13'

-- 0lS23 _ 'II})lO

_ '8011' _ _ '"00-10
'000

Reconstruction of the advance of 2 glaciers on Mt. Baker U.S.A.
based on tree age. In the early 16th century until about 1740 the tip
of the glacier was at about 850 m above sea level. In the 19th
century, by which time the glacier had retreated, the tip was at 1250
m above sea level.
Key: 1 Glacier: 2 River: 3 Contours: 4 Date of previous position
of glacier: 5 Tip 1940: 6 Age of the oldest tree in the moraine
(Heikkinen, 1984).
Comparison of the age of the moraine, based on tree ring analysis
on Mount Baker with similar moraines in different areas. Mt. Rainier:
usually dated by lichenometric methods. Scandinavia: dated using
information from historical documents (A). Swedish Lapland: dated
by lichenometric methods (B). Obviously certain parallels exist e.g.
1850, 1910 and 1925. Inexact dating and differing reaction times of
the glaciers could be responsible for the displacements (Heikkinen,
1984).

194 Rothlisberger 1976, 1986; Heikkinen 1984; Holzhauser 1984.

The effect of ice on living trees

Advancing glaciers have a direct effect on living trees. In

many cases the advancing ice destroys whole forests. If,
however, as occasionally happens, trees are only pushed
at an angle or damaged, this can be seen in the pattern of
the tree rings, particularly in a sudden formation of com-
pression wood. If the roots of a tree were affected by
glacier movement, then an abrupt growth reduction can
also be noticed. The time between the beginning of this
growth reduction and the death of a tree or its recovery,
indicates how long the glacier stayed at a certain level.
When timber is covered by ice, the soft earlywood. is
compressed and therefore only the density values of the
latewood cells can be calculated.

Larchwood, compressed when it was covered by the Zmut glacier

in Switzerland. Only the latewood zones can be evaluated dendro-
chronologically (Rothlisberger, 1976).

The age of subfossil tree stumps

Determining the age of subfossil trees is crucial when

reconstructing events related to the history of glaciers and 1----------i10 G7
climate. Wood finds spanning over 8000 years have been o 1 G18
found in the Alps, but it is often impossible to put them in o 1 _ _ - - -...., G12
the correct dendrochronological order. Particularly when a o 1__------11 G 13
long series is being built up, it is essential to make use of o 1--------11 G 14
radiocarbon methods of dating. When the approximate
position in the time scale is known, then it is usually i I I
2000 BP 1500 1000
possible to make a dendrochronological dating. Radio-
carbon dating is an important source of information in
dendroclimatology. Comparison between the time scales fixed on the basis of den-
To establish short term glacier changes, radiocarbon drochronological and l·C dating tests, on wood samples carried out
in the catchment area of the Aletsch glacier in Switzerland. The
methods are often unsatisfactory as the 14C ± error factor horizontal lines represent the 2 sigma regions of the 14C datings,
is too great. Holzhauser 1984 found, for example, that on the circles with dots the final date of trunks based on dendro-
the basis of 14C dating a glacier had reached a high point chronological datings. According to the 14e datings 2 phases
existed, whereas dendrochronological methods indicate only one.
twice, while on the basis of dendrochronological dating, (Holzhauser, 1984).
only once.

134 year old Larch trunk with bark, from the approaches to the great
Aletsch glacier in Switzerland. In 1504 the tree came under the
immediate climatic influence of the glacier. Cold winds blowing over
the glacier and the presence of snow for long periods reduced the
vitality of the tree. Around 1523 the glacier receded slightly and
growth conditions improved but in 1588 the glacier advanced once
more and covered the tree. The resin canals show very clearly how
severe the damage was. The ice pressure later compressed the
earlywood cells in the tree rings (Holzhauser, 1984).

195
Places where tree trunks are found
The history of a glacier can be reconstructed only by
studying the circumstances of the finds and the state of
preservation of the subfossil pieces of wood.
Basically there are two types of finds:
Trunks lying free in moraines in the approaches to
glaciers. They have been brought down as a result of
movement in the moraine, without the actual glacier
being directly involved, although it may have trans-
ported them over short distances. It is usually dif-
Stump of Swiss stone pine, in situ and uncovered, in the great
Aletsch glacier, SWitzerland. The first ring was formed in 973 A.D.
and the last one about 1100 A.D. (Holzhauser, 1984).
a b
(Q:'£J moriline
El ice ~ rock H hOllo.warea
.,.,.... fossil,zed roo IS
Possible position of subfossil root stumps, in situ. After being
embedded in moraine debris, often an erosion phase sets in
(Holzhauser, 1984).
196
licult to find a direct connection between these finds
and the history of the glacier. The tree rings provide
important information about the Holocene period,
however.
Tree stumps in situ, covered at one time by the glacier
and later re-exposed. By determining the date when a
tree died and how long it had lived, information can be
obtained about the level of the glacier in relation to its
present state of retreat and also on the speed of its
advance in earlier times.
Fossil trunk lying in a moraine (Ri)thllsberger, 1976).
a
1850 1890 1920 1973
······· .... ·· .. 1850
o • 0
(] o
o
Position of subfossil tree trunks in the approaches to glaciers and
moraines in relation to the ice level in the regressive phase of the
present day (Ri)thlisberger, 1976). (a) Moraine and glacier tongue
- top view. (b) Glacier and glacier bed with characteristic levels
sketched in. This representation follows Ri)thlisberger's model,
which applies only to European glaciers.
The history of the great Aletsch glacier in Switzerland
Holzhauser (1984) was able to reconstruct fluctuations in
the Great Aletsch glacier very accurately by studying sub-
fossil pieces of wood from the periphery of the glacier and
also from old water channels, and by dating them using
14C and densitometric methods. Stumps in situ have
yielded information about changes in the glacier's level
through their age, the date of their destruction and their
position in relation to the glacier's present state of retreat.
The figures explain how this was done. Six stumps were
found at different heights in areas where melting took
place most recently. Every stump found and dated by
densitometric means could be related to a phase occur-
ring since 1850. For example, tree 1 was covered by the
glacier in 1640, tree 2 in 1588. About 1640 the glacier
was at the same level as it was in 1920. Stumps dated
by 14C methods indicate that the glacier reached its
maximum in 1350. As all the trees from series 1 to 4
germinated after 1412, it seems likely that the glacier
melted rapidly after 1350 and thus conditions for trees to
flourish prevailed in the glacier approaches.
Taking into consideration all the available data, from
absolute to relative, from the very precise to the rather
less exact, most of the significant fluctuations of the last
2000 years can be registered. It is important to realise that
Photograph of the Great Aletsch Glacier in 1980 with the position of
the low point reached at the present day has occurred no wood finds indicated (arrows) and the maximum area covered by ice
fewer than 5 times before, and the high point of 1850 has in 1850 (Holzhauser, 1984).
occurred 4 times previously. These curves put the rela-
tively short-term observation periods of our scientific age 1500 1600 AD
into perspective. ->- - 1.....-
H H
Reconstruction of fluctuations in the Aletsch glacier which took
place in the Middle Ages, based on radiodensitometric dating of
in situ stumps. Further explanations to be found in the text (Holz-
hauser, 1984).
-----.,..--1---:\- - - :.~t---~~~;;~~===== 1926
1~21

Reconstruction of fluctuations in the Great Aletsch glacier using all

available sources (Holzhauser, 1984).

~ -r-~-~-r--
~ t--+-+--;.

_ 21

1961 -k'- + --+-il1+-

1910 -t~~~I;2~:-
197'
1ge1
-
. . ... ......
v "
I " "tH --+--+---+--4--+-+---+--+--+---jf--+-+---jf--+----jf--+-+--r----- - t --+--i---- 1 ,

- ,. . . .- -. . . .,
&..o At> 2500 2000 1500 1000 ~~ ~D
,.-----.,..----,..---~=-
LEGEND t5S0151O 8X) 1700 1750 .aoo .a.so A 0

,,'''ng §~§~~j""~-~~~~
1 ::~~ D
o

•
~

wtlrmef
2

. ..~ I • colrl
_ ....'00(.\ d.Jled by ndlocarbon methods D Ir" o W.lrm

c:::::==':l hood d,ued bV <k'ndrochlonolog~cal methods 2 rna)'lmum dpnsitlC'$ f t lat '.\load II"' If r W''J$;:)I ,J lar.;;:h
IAretsd'l loreslS)

197
Slope movements
Earth and slope movements take place allover the world,
whether they be as landslides, slope creep or as rise and
fall of subsoil. In densely populated areas records usually
exist about such occurrences. In unpopulated areas, how-
ever, only trees can provide information about the extent
and frequency of such events. Surprisingly enough, only
very few scientists have used the tree as a source of
information, whether it be merely through determining the
age of trees or observing the changing growth rates in
annual ring sequences. In 1971 Alestalo tried to compile
data about the intensity of cryogenic earth movement by
observing the eccentricity of inclined trees growing on
slopes bordering moors and lakes. But the most impor-
tant work was carried out by Shroder (1978 and 1980)
Trees with bent trunks growing on an unstable, creeping slope
(Lenz, 1967).
o o
o N
0> 0>
W'/ / / / / / / / / )C'
• x ~ appearance and duration of small
"~//o/'///////
,,'t .
D ~~/_/ __/_/__/____________________ -2~~ - "-
Information chart of four trees with several different radii. In the years 1922, 1938 and 1958 in particular, earth movement changed the
position of trees and influenced their rate of growth over a long period. Shroder 1978.
198
who studied the slow creep of a wooded, volcanic debris
slope in Utah, U.S.A., in great detail. He took about 1000
core samples and took cross-sections from the trunks of
220 trees and recorded all the variations in tree ring
sequences which could possibly be attributed to slope
movement.
Variations along several radii were determined, dated
and presented in graphical form for every tree observed
and because the position of each was noted, every dated
occurrence could be charted. For instance, one single
rockfall, which could be dated by scars on trees, began a
period of slope movement which lasted from 1905 till
1910.
Cross-section through the base of the trunk of a pine tree growing
on an unstable slope above the forest limit in the Alps.
abrupt appearance and duration of
• ~ reactionwood formation
; ring formation
'" + 1111111111 1 release
/;~(
• corrosion
:II
-
<
~~
Shroder, 1978; Alestalo, 1971; Timel/, 1986; Higashi et al., 1971.
If the number of growth reactions are plotted on a graph
in a so-called event - response curve, a relationship to
climatic events such as extreme precipitation, both long
and short term, can be established. The curve takes into
account the number of events based on the number of
trees studied.
Total number of events
Event-index curve (I) = X 100
Number of trees studied

:'
--" "1'
. , ,~ .J1
c
7

In 1958 only one part of the debris slope moved. This is borne out
by the concentrated number of events in annual ring sequences
c (Shroder, 1978).

1905 1906 1907

-, I

j
I
r-
•
r

-
~
!""" . I

!""'!"': ....,j I

!l"""
.-i
~ :;....j

Characteristic in tree-ringsequences growing on instable slopes.

cw/c: compression wood, gr: growth suppression, i: injury.
I
"r'_--",;<'_--"'T''''O_~,o'---->;"'---""i'---='i'-0-~'ro---=r---'T''---,o '0 •

,00
190 a 1909 1910

Mean Ann ual

0
P'IIC.IPllotU)n

.,

'"'
a,yu Canyon.
," P'rIC'pllaJIOn
S ... , roqall

"

Tree-ring curve of trees growing on a debris slope in Utah. Frequent

earth movements and variations in annual rings in 1905 seem to In 1905 a rockfall triggered off slope movement which continued for
bear a relationship to the extremely high annual precipitation six years. During extremely wet periods tree-ring events are dis-
(Shroder, 1978). tributed quite regularly over the whole area (Shroder, 1978).

199
Tree-ring research and wind
Strong winds may affect both the shape and appearance
of individual trees and the pattern of vegetation over
whole regions. In valleys with strong prevailing winds the
trees grow to one side, on t; Ie coast the trees develop
characteristic lop-sided crowns, and on the edges of
woods or forests with katabatic winds the trees tend to
be stunted. During bad storms trees may be damaged ,
uprooted or tilted to one side. Such phenomena have
been extremely well documented . (Wendel, Hewson et
al., 1977) .
These events usually show up in the trunk and the
annual rings. Where a tree is affected by prevailing winds
anomalies occur in the formation of the trunk; the amount
and type of compression wood that a tree develops reflect
differing wind velocities, and a change in the angle of the
tree will cause it to start forming compression wood
A characteristically deformed tree growing on an extremely windy
site on the coast of southern Spain. The crown is lop-sided, and the
bottom of the trunk is oval.
Picea engelmannii on the upper timberline in Colorado. The crown
has been deformed by the wind.
200
zones, a process which will continue over a long period of
time.
This sort of biological signal assumes particular impor-
tance in times when the possibility of harnessing wind
power is being considered, because it is rare for meteor-
ologists to record wind speeds in the same place over a
period of years. Today the main application of annual ring
research in this area lies in attempting to trace the rela-
tionship between the effect of wind on trees and its
climatological aspects. As is shown in more detail below,
the wind may have a direct effect on a tree or it may affect
its growth in some indirect way, through the movement of
slopes or snow for example, or through changes in light
conditions or in the length of the vegetation period or
through the effect of salt-spray in coastal areas.
Poplars in a river valley, growing at an angle.
Wind-stressed trees commonly exhibit sudden changes In the tree
ring width, or the formation of compression wood.
Wende/-Hewson et a/., 1977; Wade and Hewson, 1980; Timel/, 1986.
The relationship to wind velocity

There is a direct relationship between the annual wind
velocity on the one hand and the reaction and shape of
the tree on the other. Changes in the wind speed can be
seen in the variation, from year to year, in the proportion
of compression wood: such changes can be expressed in
terms of the compression ratio.
This area has received relatively little attention from
dendrochronologists: neither the reaction of particular
species to wind-related phenomena nor the response
of the tree to an increase in wind speed have been
thoroughly investigated.

PREVAILING WIND DIRECTION--

10
Ui 0
E
8

~"
III

c
'"
- ,- ,- !l.
" 3
~ 2
IV V - 1
C ~-~-~-~--~--~--~--~--~--~~
1.1 1.2 1. 3 1.4 1. 0 1.6 1.7 1.8 1.9
con pres"ol1 r,nio C

The classification of crown shapes in conifers. Trees with regular
crowns have not been affected by wind (0). With increasing wind
speed the crowns become correspondingly lopSided, the extremes
being flag (IV) and dwarf forms. This classification is generally
referred to as the Griggs-Putnam index (Wade et sl., 1979).
By plotting the crown shapes of a number of different trees against
annual mean wind speed it is possible to illustrate the close relation-
ship between meteorological factors and biological reactions (Wade
et sl., 1979).

25 compression ratio

1.67
l.25
2.0
l.5
l.0
1.5
1.0
0.5 lee side growth

0
1960 1966 1970 1976
year

Anomalous growth in trees affected by wind can be expressed as an
annual compression ratio. This is obtained by dividing the tree ring
widths for the leeward side by those for the windward side. The
annual wind stress can thus be clearly seen (Wade et sl., 1977).
A section through a trunk showing strong asymmetry, taken from a
tree in an exposed stand of dwarf trees. Conifers form larger and
thicker-walled cambial cells on the leeward than they do on the
windward side. The cambium on the windward side of the trunk may
die off completely, with the result that the tree grows only on one
side.

201
The effect of wind direction on the tree

In very exposed areas the direction of the prevailing slope 10°

winds can be clearly seen in the shape that the crowns of valley - -
the trees take on. There are a number of crown shapes
which are characteristically caused by particular types of
wind: the flag-shape crown, with branches on only one
side, or those of the knee-high dwarf trees. Not all
species react in the same way to the same influences,
however.
All stunted or dwarf trees have eccentrically-formed
trunks. A study of a closed 90-year-old Lodgepole pine
stand (Pinus contorta) revealed that trunk eccentricity at
chest height is a reliable indicator of the direction of the
prevailing winds.
The influence of prevailing winds is very seldom sig-
nificant where the growth of young trees is concerned.
Here growth, and hence the annual ring pattern, is more
strongly influenced by local factors, such as competition
from other trees or the burden of snow on the tree's
branches in winter.

basal stem eccenticity

stem eccentricity at breast heigh

basal stem eccentricity

• eccentricity at breast heigh
o crown eccentricity

Picea abies trunks: growth is eccentric and compression wood is
formed all round the trunks and on one side of them as a result of
the trees being subjected to strong winds when they were young.
Abnormalities in crowns schape, stem eccentricity at chest-height
and in ground level in a 90 year old Lodgepole pine stand (Pinus
contorta) in the very exposed Front Range area, Colorado. The
results from the stem eccentricity at chest-height are the most
reliable.

The effect of extreme wind conditions on the tree

branch lOP. adapted on
Heavy storms or other extreme conditions may upset the
new conditions
equilibrium of an individual tree for a short time. Here the
tree reacts by forming reaction wood over a period of one
or two years. Trees which have been pushed sideways or
bent over in some way in a storm react to the new
conditions in various ways: there is a change in direction
of growth in the crown, branches produce vertical shoots
and tension (deciduous) compression wood (coniferous)
are formed in the trunk. In woods or forests which are not
subject to forestry management the frequency with which
storms occur can be ascertained by dating these reac-
tions. The rate of decomposition of trees lying on the
forest floor can be determined in the same way.
Heavy storms often affect ecological conditions in a
whole stand of trees. Where the crown or the roots are
damaged the tree reacts by forming very narrow rings.
Trees which have been tilted sideways put on a spurt of
very rapid growth and form pressure wood. Space and
consequently better light conditions are created for the
trees which survive storms. As a result their growth rate
increases sharply. Quite often the age structure and the
distribution of species in a stand also undergo a change. A damaged fir showing cross-sections through individual shoots.
This sort of effect on woodlands has received relatively When sections are taken through the main shoots it is possible to
little attention from dendrochronologists. determine when the tree started to grow lop-sidedly by looking at
the formation of the compression wood. From the age of the upright
shoots the length of time that the tree needed to regenerate can be
determined.

202 Hoffmeier, 1971.

The effect of wind on environmental conditions

The extent to which wind influences the tree's habitat can
be seen especially clearly in subalpine areas with stands
of dwarf or stunted trees (Holtmeier, 1981). Here the
growth of the tree is affected both by changes in wind
speed and direction and by different amounts of snow.
Shoots which are protected by a layer of snow will survive
the winter but will nonetheless have a shorter vegetation
period, depending on the time at which the buds appear.
Shoots which are above the snow cover can benefit from
warm winter days but are at the same time exposed to the
various dangers associated with frost, such as frost-
drought. Here the needles continue to transpire, and
because the necessary replacement cannot be drawn
from the frozen soil, there is a danger of their drying out.
Even relatively minor changes in the tree's growing
conditions , such as the loss of a branch which had pre-
viously afforded protection, slight changes in the topo-
graphy resulting from avalanches, landslides or road
building, or minor climatic changes, will affect the growth
of the tree. Most of these types of changes can be seen
in the age structure, the growth rate and the proportion of
reaction wood in different parts of the trunk. The exam-
ples given below illustrate how sensitively trees react to
changes. Every single change has a direct effect on the
tree, either changing the growth pattern of the tree or
causing the shoots or even the trees themselves to die
off. These stunted trees enjoy an extremely precarious
equilibrium. Unlike trees in wooded areas in temperate
zones, they have no sort of buffer to protect them from
extreme climatic conditions. The fact that individual stands
are able to survive over hundreds and even thousands of
years is attributable solely to their capacity for vegetative
regeneration through the formation of roots by the
branches.

During storms older trees quite often fall onto younger ones, push-
ing them sideways or causing them to fall to the ground. It is
possible to say when such storms occur by examining the formation
of reaction wood in trees which have been affected in this way or in
the branches which start to grow vertically after such damage has
ocurred.

16 45 Engelmann spruce
engelmannii

west wind

subalpine fir
Abies lasiocarpa

f'f cross-section 1: 2
o
o

n
.
o
.
2M

Age structure in stunted trees on exposed timberline sites with heavy snowfall. The subalpine fir (Abies lasiocarpa); and Engelmann spruce
(Picea engelmannii). The growth pattern is reflected in the age of the individual shoots. Damage caused by frost-drought, avalanches or
rockfalls results in abrupt changes in the annual ring width, while that caused by strong prevailing winds results in the formation of
compression wood. The study of such phenomena enables a picture to be built up of local micro-climatic conditions. (Kienast and
Schweingruber, 1986.)

203
Forest fires

Since time immemorial forest fires have constituted an become dominant: Araucaria araucana over large areas of
important natural aspect of the ecosystem in many parts South America and certain pine species in the west of
of the world (White, 1972). It has sometimes happened North America.
that the development of flora and fauna, and even of The effect of forest fires and their history has been
human cultures, has been largely determined by the documented (White, 1972; Alexander 1979; Stokes et aI.,
occurrence of forest fires. Fires were caused naturally by 1980).
lightning; later human activity led deliberately or acciden-
tally to fires being started. In the course of time trees
which are to some extent resistant to fire became domi-
nant.
Thick bark and scales on the twigs and buds provide
direct protection against the effects of fire. Wood that is
rich in resin helps to prevent the biological decomposition
that follows damage caused by fire. A considerable num-
ber of species developed the capacity to produce a
second set of leaves and shoots in years when the first
set had been destroyed by fire. The species most resis-
tant to fire include larches and many pines, e.g. Pinus
ponderosa. Spruces, firs and the majority of broadleaves
are less resistant.
Forest fires have both short and long-term effects on
the vegetation in any given area. In the short-term, stands
become less dense, and species of trees and shrubs that
are resistant to fire are encouraged. Fires may also have a
profound long-term effect on the development of the
pattern of vegetation. The earth's vegetation bears un-
mistakable signs of the effects of forest fires. In Mediter-
ranean regions, such as the Maccie, the Garigue and the
Chapparall, a characteristic pattern of vegetation devel-
oped with species which have the capacity to produce a
second set of leaves and shoots becoming dominant. The
frequent occurrences of forest fires has to a certain extent
determined where a number of different species have Pinus /eucodermis showing the characteristically thick bark. Barks
which are thick, and which are therefore poor conductors of heat,
protect the cambium from the effects of fire.

An olive tree which has been damaged by fire and has as a result A sparse Pinus ponderosa stand in the dry zone of the south-west
produced shoots. It is primarily broadleaves which have the capacity of the U.S.A. It is very often the young trees which are destroyed in
for vegetative regeneration. forest fires.

204 Stokes et al., 1980; Alexander, 1979.

The effect of fire on the tree-ring pattern

Fire affects the tree-ring pattern in a number of different Only those species with cells containing substances with
ways. According to the degree and extent of the damage a fungicidal effect such as pines (resin) and larches and
to the crown - the assimilating organ - the cambium Tsuga species (heartwood substances) remain free of the
reacts in a particular way: fungoid infection and subsequent damage that may result
from fire.
the action of heat on the trunk causes callus tissue to
be formed locally or false rings to develop;
in conifers tangential traumatic resin canals may be
formed;
the most obvious effects of damage by fire are the
scars it leaves.

Section through the trunk of a Pinus ponderosa, showing three

fire scars. Since 1904 the Forest Service has been successful A Pinus ponderosa which has sustained a lot of damage by fire.
in containing the number and extent of forest fires in the U.S.A. This characteristic scarring is commonly referred to as a 'catface'.
(Dietrich and Swetnam, 1982).

, 1977

I II 1\, Here the latcwood for 1977 is missing.

) The trees died very soon after the fi reo

; II III
, This tree was badly damaged in the fire.

! llll,! I t managed to survive for a futher three

years, but only formed very narrow
annual rings.

Ilillll) These trees were hardly damaged at

all. They died three years later as a

11111111 result of pine- beetle attack.

Photograph (taken with a macro-lens) of a cambium which has been

Diagrams illustrating different cambial reactions to fire. This fire badly damaged by fire. The fire interrupted the formation of the
occurred in the summer of 19n in an area of Ponderosa pine in earlywood, which means that it must have occurred in June or
Boulder, Colorado. thereabouts (Swetnam, unpublished).

Dietrich et aI., 1983. 205

Dating forest fires
Using information obtained from pollen analysis it is pos-
sible to build up a picture of the relative frequency with
which forest fires have occurred in particular centuries in
the Holocene era.
In those areas where there are relatively few fires, e.g.
the temperate and boreal zones, the geographical extent
of fires can still be determined after hundreds of years.
The growth release which occurs in those trees which are
on the edge of an area affected by a fire enables the date
of the last fire in that area to be ascertained. A fire which
broke out in 1905 in the Front Range near Boulder,
Colorado, for example, was provisionally dated from
growth release in Engelmann spruce. This dating was
then confirmed by determining the age of the vigorous
young growth which had occurred in these trees. All the
firs (Abies lasciocarpa) now growing on the area affected
by the fire germinated after 1905.
The most accurate picture of forest fire history can be
derived by dating the fire scars on the trees. Species
which are resistant to fire and to the decomposition that
follows fire damage tend to be affected only at the base of
the trunk. Where this occurs the cambium immediately
begins to protect the injured area with callus tissue, which
chronologically, either with the help of skeleton plots, or
by identifying a number of climatically-dependent pointer
years. (Dietrich et al., 1983) .
Since some trees remain undamaged in almost any
fire, an accurate picture of the incidence of fires in a given
area can really be obtained only by examining both sides
of a fire scar, and by looking at a good number of trees.
Dates obtained for fires in the recent past should be
checked against fire records kept by forestry agencies
or against other contemporary documents wherever pos-
sible. The geographical spread of a fire can, where neces-
sary, be ascertained using aerial photographs. For the
purposes of forest management it is often only the aver-
age incidence of fire in a given area that is required. In
such cases it is therefore generally sufficient to divide the
number of years in question by the number of known
fires. In the U.S.A. this method is only used for the period
before 1880, i.e. for the time before there was active fire
prevention .
.
00
t."
_ 00

.
OJ). ~ ... rtv A
_ 00 • •'Iqr 1 'f'DIJI"I;I "
_ 00
forms a scar. In order to determine precisely the number
.--
- 00
__ 0

of fires that have occurred and when they occurred, it is o

0 •••

necessary to examine discs from the trunks of trees,

~ O ·t· o

which of course involves felling the trees. It is not pos-
sible to say in which years fires broke out merely by
examining and counting the rings, since - especially in
trees in dry areas - some annual rings may be missing.
It is therefore necessary to date the fire scars dendro-
'!'lO 10 30
The numbers of trees at the edge of an area affected by fire showing
sudden growth release. The figures relate to a timberline site in
Niwot Ridge, Boulder, Colorado. The fire occurred in 1905 and the
instances of growth release concentrate on the period 1906 to 1908.
(Kienast and Schweingruber, 1986.)

MFI

1700 1750 1800 1850

a AI
Counti ng date 96 yrs
-':/ ~~ -~ Z; ."'1 lY ."'1 D: .!>I: 'll ~ ~ -'lY
::L. . ~ ~
a .~.!:i. *,G;\<.~ ~
b C{' E.- N li « Adiusted counting date 9.4 yrs
c A A A ... A ... A ... A A ... A A A A ... Dendrochronological Date
• • • • Coincidence
10.5 yrs

A comparison between dates obtained by fire-scars using different

methods: (a) dates obtained by counting the tree rings; (b) as above,
but corrected, using existing scar chronologies; (c) dendrochrono- Di sc 1 Disc 2
logically produced dates using cross-dating. Only three of the
Side A Side B Side A SideB
sixteen dates agree in all cases. The mean fire intervals, on the other
hand, show good agreement (Madany et al., 1982).
1900 1900 1900 1900
1898 1898 1898 1898
1889 1889 1889 1889
1885
Samp Ie tree nu mber Compo site
fire scar 1875 1875 1875 1875
13 16 17 18 chronology 1872 1872 1872 1872
1865 1865
1871 1871 1863 1863
1842 1842 1861 1861
1785 1785 1785 1785 1785 1859 1859 1859 1859
1757 1757 1757 1757 1857 1857 1857
1752 1752 1855
1632 1632 1853 1853 1853
1595 1595 1851 1851

A comparison between the fire scars for a number of different trees A comparison between the dates obtained for fire scars by dating
from a narrowly-defined site. It is possible to determine the exact the left and right-hand sides of catfaces. The exact number of fires
number of fires that have occurred in a particular area only by that have occurred can be ascertained only by dating whole catfaces
examining a large number of trees (Arno and Sneck, 1983). dendrochronologically (Dietrich and Swetnam, 1983).

206 Arno and Sneck, 1977; Madanyet al., 1982; Dietrich and Swetnam, 1982; Swetnam and Dietrich, 1983.
 1800 1850 1900 1950 1980
SITES "'"-i

A '~
2 •
3 •
.
rl
---;
2
3
4 • : snog 4

:i
H 6
.....
Q)
---;
.D
E 8 •
• ---;
7
8
B Z
j

II • H II
c 12 • rl 12
QJ

..
E 13 • snog 13
U
a. 9 •
QJ
i '/ rl 9
C (/)
10~•
.
H 10

snog 14

::~
"1
snag 15
D 16'" I II ' downed log 16

.. .. .
h)
17 : i i i snag 17
0; .; ;;;
2 0 :; ~ ~
~
2

1800 1850 1900 1930 1960 1980

SITES

A '~
2 •
12 :
13 :
1-------1
,
2
12
13

.....
14 :
.. , snag
14
4

.~•
Q)

.
I.,snag 5
.D I"i
5 •
B E , Isnag 7
:l 7 •
Z
I'
t, ) ,.
• I 6

l
6 •

C
c
QJ

E
u
(I'

downed log
.
, ., 8
9
10 : 10
,
Q)
a. II
CJ) II :
~snog

"~
15
16 :
~~I , 16

L
, ,
D 17 • 17
18 :
38 :
. .."'.". - .. .. .. .
{
I snag
I snag
18
38
..'"
0
0;
~ ~- ! '"
~ ~
; iii;
~
0;
~ ~

1800 1850 1900 1950 1980

... I • • ~

2 • 2
'"
.D
E
:3 •
• :3
(II
Z
:l 4 • • 4
........
c
5 • • (,1
5

'" 7 ••• 7
E 8
•• ii i ....... 8
u 9
'"a. • n It! ........
9
CJ) 10 • I li , I I 10

. ....
....
( I]
II : II
0;
: :
N ~
~ ~i
.oN ~o
~
.... :
~ ~
i :..

Dendrochronological dating of fire scars on Pinus ponderosa from three separate sites in the Gila Wilderness in southwestern New Mexico.
In that many of the fires ranged over wide areas it is likely that this dating will yield information about precipitation in summer. The fire
scar free period before 1837 corresponds to a very wet period. Since 1904 the U.S. Forest Servive has had considerable success in
keeping the number and extent of fires under control (Adapted from Swetnam, Dietrich, 1983).

Swetnam and Dietrich, 1983. 207

The effects of forest management on the incidence of fires

Since the end of the last century fire prevention mea-
sures, particularly in the American West, have ensured
that forest fires have been kept under control. This in turn
has caused dramatic changes in the pattern of vegetation
that had been dominant in these areas for thousands of
years. Today the effect of this sort of intervention is pretty
well understood. The example of the Ponderosa pine
forest near Flagstaff, Arizona, is certainly typical of many
other cases. (Kennedy-Sutherland, 1983). In the last
hundred years there have been no fires at all in the stand:
before this time fires occurred once every two or three
years. The stand was described as park-like and open by
early explorers, but now it is extremely dense with many
thickets of young trees. Since 1920, which marked the
end of a relatively wet climatic phase, the growth rate of
both young and old trees has decreased .
There are a number of reasons for this marked de-
crease in growth
The older trees are in competition with the younger
ones particularly in the root area. The young trees take
a lot of rain-water from directly under the surface.
The crowns of the trees, especially of the young ones,
have to compete for light and space
The fact that the ground is relatively densely covered
inhibits the flow of water to the root layer: a lot of
water is lost through interception.
In dry areas the organic materials (needles, twigs and
cones) decompose slowly. This means that most of
3
the nutrients remain on the surface and so are not
available to the trees. The occurrence of fires used
to ensure that these potential nutrients were continu-
ously being recycled in the soils.
The structure of the present stand has considerable dis-
advantages. The dense forests are much more suscepti-
ble to large scale crown fires that destroy entire stands,
because the young dense trees and accumulated fuels
provide a 'ladder' for fire to the crown level. Insect popu-
lations, e.g . the spruce bud worm, flourish in dense
stands and mistletoe, particularly Arceuthobia, have an
adverse effect on the vitality of the stand.
It is mainly for the sorts of reasons outlined here that
foresters and biologists have started to call into question
the concept of total fire prevention. The work done by
dendrochronologists in this area have helped to provide a
background against which today's essentially short-term
measures can be evaluated.
Conclusion
The effect of forest management in semi-arid and boreal
regions has been the subject of fairly intensive study.
Relatively little work has been done on the effect of fire in
broadleaf woodlands in the time before forestry manage-
ment, however. Studies in this area would enable an
important gap in research to be filled.

2
><
Q)
"0
c:

40 60 80 1700 20 40 60 80 1800 20 40 60 80 1900 20 40 60 80

years

Tree-ring chronologies for Ponderosa pine in northern Arizona. This area, wl1ich is not subject to forestry management such as harvest or
thinning, was last damaged by fire about a hundred years ago. The trees reacted to the high precipitation at the beginning of this century
by forming wide rings. The extremely narrow rings that are characteristic of the last few years are a result of lack of oxygen and nutrients
caused by the fact that there has been insufficient mineralization. This can be traced to the absence of fires and to the fact that young
growth has absorbed a lot of precipitation mOisture (Kennedy-Sutherland, 1983).

'';. ! II 'II d 1/1 ill I ill I { I{Ji I~ If I / ~ l 1 J., 1111111 U~ III {fl lJ UJlU HrmITII 911HYnmm
;illl.IPlllJ
I / .
.J /I J II/J i U ,"1/ r ,.·,'·W 'I -""'\""'j; --.--.r.,~1 1/ ~
J' ~ !..'=.Y£:: .• .'i'. 'l:

,-,-' i H·I/j/ II 11/111 J li~JJ1]} ll ) ?' 11nlllll!1 t~ '.L. . .

Single cores relating to the above chronologies. There has been a continual decrease In ring width since 1910 (Courtesy of Kennedy-
Sutherland).

208 Kennedy-Sutherland. 1983; Tande, 1979

The effect of weather conditions and vegetation on the incidence of forest fires

Fires which occur when the weather is dry and there is a
lot of wind - in high summer, for example - cause more
damage, in terms both of the intensity of the damage
to individual trees and the geographical extent of the
damage, than those which occur at wetter times of the
year, when there is less wind . This difference is most
clearly reflected in the extent of the area affected by a fire.
The extent of past fires can be mapped fairly accurately
provided that a good number of trees are examined. Fires
occur more often in dry than in wet periods. This can be
clearly seen by comparing the average intervals between
fires for different vegetation zones.
The incidence of fires in semi-arid zones is very high.
In some areas fires occur as often as every two years. In
boreal and sub-artic regions, on the other hand , forest
fires may occur on average every hundred to two hundred
years. Thus the relationship of fire occurrence to weather
is probably more strongly linked in the low frequency
types than in the high frequency types. On the other hand,
the absence of fire may be strongly linked to weather in
the higher fire frequency types.
The frequency and intensity of forest fires is largely
dependent on the type of vegetation in the stand . Where,
for example, there is a lot of grass the fire spreads quickly
and leaves little trace . By contrast, dense stands of young
trees may be totally devastated over wide areas. In woods
composed of species which are resistant to fire - in
Pinus ponderosa stands, for example - it often happens
that only the grass and shrubs are destroyed and the
large trees sustain damage in the form of a fire scar.
Woodland comprising trees which are not resistant to fire,
e.g. spruce, are often destroyed over wide areas.

1889 1906-1907 1908

The extent of forest fires in the Jasper National Park in Alberta, Canada, mapped using Information gained from dendrochronological
analysis. Some of the fires covered only a limited area (1808), while others affected a whole valley (1889) (Tande, 1979).

Principal Underground alti- Forest fire interval (years)

tree tude
species m.a.s.1. region A region B region C

Pinus ponderosa grass and dwarf shrubs 1500 2- 20 (6) 2-18(11) 2-18(10)
Pseudotsuga dwarf shrubs 2000 3-31 (10) 4-29 (16) 2-48 (19)
Pseudotsuga
grass and dwarf shrubs 2300 3- 33 (17) 5- 52 (27) 5-67 (28)
Abies /asiocarpa
Abies /asiocarpa
grass 2700 8-50 (41) 4-78 (30) 2-68 (33)
Pinus a/bicau/is

The incidence of forest fires related to altitude and woodland type. This information is based on dendrochronological dating of fires in the
west of North America. Fires occur far more often in forests of Douglas fir than in stands of Abies lasiocarpa which are at higher altitudes,
even where both types of stand have a considerable amount of grass as ground cover (Arno and Sneck, 1977).

Tande, 1979; Arno and Sneck, 1977. 209

Volcanic activity
The eruption of a volcano has always caused a consider-
able amount of destruction and change. Living creatures
in the immediate vicinity are destroyed; particular sites are
laid waste while others are improved in some way; clouds
of volcanic dust may cause climatic changes which affect
the whole of the earth. The different effects of volcanic
activity on trees, how the damage caused is reflected in
the annual ring picture, and how this information can con-
tribute to an understanding of the incidence of volcanic
activity are discussed below.
The direct effects of the eruption of volcanoes
Practically everything in the areas immediately surround-
ing a volcano is destroyed in an eruption. Pressure waves
cause trees to be knocked the ground, snapping them
like matchsticks. Trees in more sheltered positions are
tilted.
After an eruption the countryside lies under a thick
blanket of volcanic ash. Following the eruption of the
Laacher volcano in the Eifel region of West Germany, for
example, the scorched trees were left under 20 metres of
Destroyed forests in the west of North America. When Mount St.
Helens erupted on May 18th, 1980, the conifer forests at the
timberline were completely destroyed by heat and pressure waves.
Many trunks are now buried under the ash (Anonymous, 1980).
The waves of air pressure which followed the eruption of Mount
Katmai flattened whole forests. The position of the trees and the
date of their last ring provide information about the nature of these
occurrences (Griggs, 1928).
210
ash. This very fact made it possible to give a precise date
for the eruption of the volcano on the basis of radiocarbon
dating. An analysis of leaves found buried in the sediment
indicated that the volcano erupted in early summer. Only
a few km away from the site of the eruption the layer of
dust may be as little as twenty to thirty centimetres thick.
Clouds of ash may be blown over distances of hundreds
of km by strong winds.
Trees which are directly in the path of a lava flow are
charred by the tremendous heat which the lava emits.
Those that are just on the edge of a flow will only be
damaged to the extent that their trunks may sustain some
burn marks or, where their crowns have been singed,
they will produce narrow rings for a number of years. By
analysing the last-formed rings of trees destroyed by a
rise in the water-level caused by lakes being silted up
with lava, it is possible to determine the year in which the
volcano erupted. A volcanic eruption often causes ice
melting. The resultant flooding may cause the trees which
were 'felled' by the effect of the volcano to be carried
over long distances.
In 1912 whole forests were covered with ash when Mount Katmal
erupted. Years later river activity exposed the roots of conifers
whose branches had in the meanwhile produced adventitious
branches (Griggs, 1928).
Trees exposed in a mudflow in a stream draining Mount St. Helens.
Hot debris from the eruption induced by rapid melting snow and ice
in the upper slopes of the volcano, causing mudflow that carried
trees many km downsteam from the volcano.
Further away from the volcano, at the edge of the area
most affected by an eruption, the heat has a selective
effect. Young trees protected by a layer of snow may
hardly be damaged, while larger mature trees are killed.
The nature of the tree's vegetative organs are crucial to
whether or not it survives. Trees like Araucaria may be
protected from the searing heat by their thick needles and
bark, while those species whose needles are rich in resin
or which have thin leaves will be set alight immediately
and die as a result. Species with thick bark, e.g. Araucaria
or Ponderosa pine, are far more resistant to the heat that
those with thin bark such as beech, southern beech or
spruce.
Tiny particles of volcanic ash often adhere to the
surface of the leaves, inhibiting photosynthesis. This
effect is particularly significant in semi-arid regions where
the layers of dust may remain on the leaves for months.

Changes in the pattern of vegetation in a stand of Araucaria and
southern beech in Chile (Araucaria araucana and Nothofagus
antarctica), caused by the effects of hot volcanic ash. The
Araucaria survived almost unharmed thanks to their thick leaves
while the beeches were badly burnt. After a relatively short time,
however, they were able to produce new shoots.
Young conifers which were protected by a layer of snow survived
the waves of pressure and heat sent out when Mount St. Helens
erupted. By contrast, the dominant trees, Douglas firs up to 1.2
meters thick, were flattened and seared.

Forests over wide areas were covered with volcanic lapilli as large
as hail-stones. Two years later the needles and branches of the
affected trees still had a layer of volcanic dust and stones on
them. All the assimilating organs were covered with tiny particles
of volcanic ash. The resulting impairment of the trees' capacity to
photosynthesize was reflected in the formation of narrower annual
rings.

The effect of more light becoming available to trees in a forest stand as a result of other trees being .destroyed by hot volcanic ash.
The Nothofagus dombeyi were badly scorched while the Araucaria survived. The latter began to grow more quickly as a result of the
increase in light.

211
 Volcanic ash changes the physical conditions in the
soil. This effect can be seen very clearly in the tree rings
of conifers which are growing in volcanic ash. Thus
volcanic events can be dated. An example of this was the
eruption of Sunset Crater near Flagstaff, northern Arizona.
Here the trunks of those trees which were damaged in the
blast itself were so badly charred that it proved impossible
to date them dendrochronologically. However Smiley
(1958) identified a number of trunks which had been used
to build the Indian settlement of Wupatki. This wood was
characterized by narrow rings with low sensitivity for the
Section through the trunk of a fir near Mount St. Helens. Photo-
synthesis was impeded through the effect of a layer of dust on the
needles. The release can be seen as early as 1981.
1050 1064
Four tree-ring curves of pieces of timber from the Wupatki settle-
ment. The onset of the period of poorer growth and lower sensitivity
is a expression of the defoliation by the volcano Sunset Crater
erupted in 1064 (Smiley, 1958).
212
years after 1064-1067. It is probable that the wood was
used to rebuild the settlement when the Indians returned
to the site which had been severely damaged in the
eruption of the volcano. It is likely that the trees originate
from an area which was affected by the eruption of the
volcano. The analysis of these pieces of pine shows
clearly that the volcano erupted approx. 1064 AD and that it
was a good 50 years before the area was resettled. It proba-
bly took this long for the layers of volcanic ash to weather
down into fertile soil.
Section through a fir growing in the area devastated by Mt. St.
Helens. The growth of this tree had been limited for 50 years
because it had been growing close to other trees. When these trees
were destroyed by the volcano much more light became available to
the tree, and so from 1982 its growth is appreciably better.
A disc from a beam from the Wupatki settlement. In 1064 the tree
rings started to narrow abruptly. The wood was dated using the
skeleton-plot method (Courtesy of the University of Arizona).
Smiley, 1958; Kaiser, 1979; Yamaguchi, 1983.
Examples

Mount St. Helens, Washington. U.S.A.

It can be of enormous value to know the frequency with
which a volcano erupts in an area where people are living .
In the case of Mt. 81. Helens the existence of two near
surface soil horizons which had been buried under
volcanic ash provided clear evidence of the fact that the
volcano had erupted before. Yamaguchi (1983) dated
these eruptions by identifying abrupt reductions in growth
in Douglas firs growing in the ash sediments. No such
reductions occurred in trees which were growing on sites
a good distance away from this sedimentation. Trees
within a radius of between 5 and 20 kilometres on the
other hand display clear reductions in growth for the
periods following the two years 1480 and 1820. These
reductions can be traced to damage caused by volcanic
ash. Fossil soil covered by volcanic ash, Mt. St. Helens, U.S.A.

Il lfH\Htll Straight Creek

beyond the area covered with

'1/11/JIllJ111111 , II , ,II, \"

ash by the eruption of 1800
Jefferson Creek

1800

IJ11I1H/IJIIIJ1111111111KII///I///II////IIII Commonwealth Mine

I U\\1 \\-\ \ \\\ \\1 l)B\l\\~\ti\llilfu\ \\\ H I ndependence Pass

within the area covered with
ash by the eruption of 1800

I JI111 ~r(II/IIit/,}OJ»1H1"'~
1800
Spirit Lake

I III· I I 1111 U/JHHilI 1 III

J
1480
Bear Meadon
within the area covered with
ash in 1480.

Tree-ring sequences for Douglas firs near Mt. St. Helens. Trees close to the volcano display abrupt growth reductions which can be traced to
the eruption of the volcano. (Yamaguchi, 1983).

Laacher Volcano, Germany

It is probable that the eruption of the Laacher volcano in the Eifel

region of West Germany in the Allerod (approx. 12000 yrs. BP) era
was responsible for these tree-ring patterns in pines near Zurich,
Switzerland (Kaiser, 1979). Ash depOSits from volcanic activity in the 'V\ .'
Eifel region have been found in the western Alps. The action of
volcanic dust in reducing light and temperature must have been so
great that the physiological functions of the trees were impaired for
some time afterwards. This hypothesis still remains to be checked
using precise radio-carbon dating (Kaiser, 1979).

-
t

213
Frost rings

Katmai
1912
V

Unusually low latewood density in the annual ring for 1912 in timber-
linetrees in the Alps. Using a hand-lens tree-ring sequences can be
dated on the basis of such characteristics.
On June 6th, 1912, the volcano Katmai in Alaska erupted sending
ash more than 20 km 3 up into the air. Griggs (1928) reported that the
dust particles reduced solar radiation in Algiers by 18% in July and
August. Summers everywhere were colder than usual on account of
this dust barrier. Cell wall growth in trees in the Alps and the
Carpathians was drastically reduced. Over practically the whole of
Europe latewood density in conifers for the year 1912 is lower than
that for 1911 .

a 0 0 0 N Lf) r--.<x>~r--. <D

~
NN Lf)

!! ,I
... 0 <D <X> M <D o ~NMM <D o~ <;t <D
~ co

, l gl
IV <D <D <D r--. r--. <X> <X> <X> <X> <X> <X> (J)(J) (J) (J)
:0 '" ..- ...-,,---
'" C
'"
(5
C
~ IVIV> Ig gg l 0
C
0
C
I gl 0
C
I }gllJ
<l>C
~ .Q
'" Lf)
co
r--.
J
r--.
c'" Lf) M
::J ::J 5,000 co
0)0.
b 0 <;t
Lf)
co
~

0''-
~~
4,000 (J)
Lf)
r--. ~N M
<DM 0 <D N M <D
> C '"
3,000 tg(J)o~
~ <;t
co co (J)
'0 ::J
0'0 2,000
0
0 L!) <D <D ~ co
.~.> L!)
0..-
::J'O
1,000
'- C
0) . - 0
~ ~

i ~
700 j ~j
h.:!! ::
. •
0
c ; ii ~~ e t
; :
,!

..ii: 1
~

600 :i :\ . r; ;~~
:i :!!'i
1 0-

JJ h
.;; 0)
j l~ .; 1'~
Q;
'0 .r 500 ~ r ~ ~ cu i~ ? u c
joi
a.
. ,.~
o -
~!! M "
~
'0 f;;::;
'" 'E
<I> ;~~ j 1~ ~:"8 ;~
_~l

vn
.A~ -

.Ir ","::j
IV 400 g Jr "t - •• - - .!
~
S~!
l.J 2: il
~-

- !.;;~ !!i
::J 0)
~-
"iO .r ~
300 j !f
• .: CI "...
>
E jjj ~
>
'i: '"
-5 0
l"i
::0
!
c
'"
> 0
IV 200
C

100
0
AD 1500 1600 1700 1800 1900 1970

Frost rings in bristlecone pines in the western USA in relation to major volcanic eruptions on global scale. (a) Dates of notable frost events.
Arrow indicates associ2ted eruption and NO indicates absence of notable frost event at time of major eruption. (b) Dust Veil I
(DVI/Emax ) and dates of eruptions and eruption sequences of 1000 or greater. NO indicates apparent absence of major eruption correspond
to frost-ring date shown in (a). (c) Integrated yearly DVI with names of volcanoes and dates of major eruptions. Some other large erupt are
indicated by names of volcanoes in parentheses (La Marche, Hirschboeck, 1984).

214 La Marche et al., 1984; Kaiser, 1987.

The effect of ash rain on climatic conditions and annual ring growth

Where large amounts of volcanic dust (silicate particles
and sulfur aerosol) are released into the earth's atmo-
sphere this has a direct effect on irradiation from the sun.
Should this dust veil remain in the stratosphere over a
period of some months the weather will become colder
and there will be more periods of extreme cold than
normal. Such periods when the weather is colder than
usual are clearly reflected in the annual ring patterns of
conifers in the subalpine and arctic zones and in the arctic
glaciers: summer temperatures which are lower than
average are reflected in low latewood densities. When on
the other hand there are two or more nights in the two
months August and September with temperatures below
freezing (-5°C for Bristlecone pines) the latewood sus-
tains frost damage. Temperatures of this order cause the
cells to freeze and be damaged. Non-lignified cells - and
in some cases callus tissue - within the latewood are
typical reactions to this sort of attack by the frost.
It is possible to relate frost damage to a fall in tempera-
ture. The frost ring formed by Bristlecone pines in 1884,
for example, is known to be a reaction to a frost in the
night of 9-10th September; the ring formed in 1965
resulted from the cold snap that occurred from 17th to
19th September. The extreme acidity of the ice that is to
be found in Greenland - where the ice is split into layers
every year by the action of the cold - can be attributed to
volcanic sulfur vapor, while the layers of dust in the ice
itself are composed of silicate aerosol.
LaMarche et al. (1984) plotted the incidence of frost
rings in the long annual ring sequences for American
Bristlecone pines against volcanic activity. In many cases
the occurrence of an unusually high number of frost rings
could be directly related to large-scale volcanic activity. It
must of course be borne in mind that the formation of
frost rings may take place independently of the occur-
rence of volcanic activity, and that this activity in turn does
not necessarily result in a higher incidence of frost rings
(Timell 1986). Nonetheless a causal connection has been
established between the eruption of volcanoes and the
occurrence of frost rings: in some cases, for example,
more than six times as many such rings as normal are
formed in years when a volcano erupts.
The discovery of volcanic events in the ice and the tree
rings open is up the possibility of determining both how
often and when volcanic activity occurred in times for
which no records exist. The following table illustrates how
far the attempt made by LaMarche et al. to carry out this
sort of dating has been successful:

r ' Frost rings Volcano Date Source

~ ~
B.C. 2035 St. Helens, 1900 14C
U.S.A.
B.C. 1626 Santorini, 1390 ±50 14C and
Greece? archaeological
~~~ ~ evidence
]; 1'1
'Y.! B.C. 42 Etna, Italy 44 B.C. written records
.c Joj'" ~
~ 687
~ ~
~
...
~I-I
~
9 1003
1099 no parallels
~ ~~~ JA H 1171

~
~
g
~
1200
r ~ )" ~ ~ K H
1453 Kelud, Java 1453-1454 Greenland ice
tl~ ~
~ Ij\ j.
1500 St. Helens, 1480 tree rings
Frost damage in the earlywood of a larch. Cambia that are U.S.A.
damaged by frost produce traumatic tissue for a short time (Lenz 1601 not known 1600-1601 Greenland ice
1967). '
1884 Krakatau 1883 written records

The· probable dating of the serious eruption of Mount

Santorini on the Aegean coast of Greece is of particular
interest. Since this eruption was almost certainly con-
siderably more severe than that of Krakatau in 1883 it
probably had a global effect on the weather, causing the
frost ring in the Bristlecone pines in the White Mountains.
14C dating of pieces of wood and charred grain found in
settlements destroyed by Santorini indicates that the
volcano erupted around the year 1680 B.C.

Here a narrow annual ring with low maximum density can be seen
between two narrow rings of normal density. In cold-moist climates
low maximum density is an indicator of cold summers (RothliS-
berger, 1976).

LaMarche et al., 1984; Rothlisberger, 1976; Timel!, 1986; Yamaguchi, 1983. 215
Earthq uakes

Earthquakes can affect the growth of trees in a variety of (LaMarche et aI., 1972, Meisling and Sieh, 1980) have
ways, e.g. changes in the hydrological conditions caused been confirmed in this way. Earthquakes which are not
by changes in the level of a site, trees on slopes or in documented have not proved possible to date with any
fault zones being tilted sideways or their root-system or degree of certainty, the exception being a number which
trunk being torn or damaged in some other way. have occurred in California (LaMarche et aI., 1972).
Since other sorts of occurrences such as storms, fires
or lightning may have the same effect on the growth of
trees as earthquakes, it is often difficult to show conclu-
sively from the annual ring picture alone that an earth-
quake has occurred. Changes caused by earthquakes
may affect growth in the following ways: the abrupt onset
of a period of increased growth on the side under pres-
sure; the formation of compression wood and the onset of
reduced growth on the released side of trees tilted side-
ways; a slower or faster rate of growth over a number of
years; changes in the axial growth of the trunk.
When the year in which an earthquake occurred is
known its effects can generally be identified in the annual
ring pattern. The movements along the Fairwater fault
in Alaska, which occurred in 1958 (Page, 1970) and
those along the S1. Andreas fault in California in 1905

1958
Topped spruce, abrupt growth reduction

Annual ring patterns of trees in areas of Alaska affected by earth-

quakes. The abrupt changes in the growth patterns enable the
earthquake to be dated (Page, 1970).

original center
(ca. 1400)

,'....r..-<--- fracture
(1906? )

- apprOXimate
ring pattern

scale,
5 ft.
A Sequoia gigantea which has been tilted sideways. This tree lies in Section through a Sequoia which was tilted sideways in the earth-
the St. Andreas fault, Fort Ross, California. V. C. La Marche is in the quake of 1650 (La Marche and Wallace, 1972).
foreground of the picture (La Marche and Wallace, 1972).
1960

Tree-ring sequence of a Sitka spruce which was severely disturbed in the earthquake in Alaska in 1964. A period of considerably reduced
growth followed, probably as a result of severe damage to the roots. The tree was growing normally again 12 years later (Jacoby,
unpublished).

216 Page, 1970; LaMarche and Wallace, 1972; Meisling and Sieh, 1980.
Giant waves
During the Holocene epoch wooded areas in the moun-
tains were quite frequently destroyed by rock or ice falls.
Such falls may have devastating effects when the masses
of rock or ice plummet into fjords or seas and cause giant
waves to form . Such waves occur fairly frequently in
Japan, Norway and the U.S.A., and have resulted in
woodland being destroyed over wide areas. Miller (1960)
documented the occurrence of such waves in Lituya Bay,
Alaska. In the last hundred years alone earthquakes have
caused a number of glacial ice falls (1864, 1899, 1936
Aerial photograph of Lituya Bay, Alaska, in 1954. (g) and (k) trim-
line of the giant wave 1853/44. (m) lateral moraline (Miller, 1960).
Impact slope and the SOO m high wave (Miller, 1960).
Impact slope of the 500 m high wave (Miller, 1960).
and 1958). In 1958 thirty-five million cubic metres of rock
hurtled from a height of 100 m into the bay. The resulting
giant wave swept along the 6.5 km long bay with a
velocity of 200 kph. , leaving large areas of devastated
woodlands in its wake. The pattern of vegetation in areas
affected by giant waves indicates the exact extent of their
effect, while the callusing over of injuries on trees grow-
ing at the upper limit of the devastated area provides the
key to the dating of their occurrence.
4#
Aerial photograph of Lituya Bay in 1950, immediately after the area
had been hit by the giant wave of 9 July, 1958. (r) rocks which
were broken up, (d) impact slope of the highest wave, (b) a fishing
boat was swept over the sea-wall, (e) a fishing boat sank here
(Miller, 1960).
Cross-section through a tree trunk. This was taken in 1953 above
the trimline of 1936. The callusing enables the year in which the
giant wave occurred to be established (Miller, 1960).
Cross-section through a tree trunk. This was taken in 1953 above
the trimline of 1953/4. The number of rings formed since the tree
was damaged indicate that another giant wave occurred in 1899
(Miller, 1960).
217
Tectonic movements

Earthquakes often result in land-masses rising or falling. and crown systems had not been able to develop prop-
Trees growing on sites which fall are very frequently erly. Growth conditions improved appreciably with the rise
affected by the flooding which this brings in its wake, and in the level of the land. One tree was tilted over by the
they die fairly rapidly. Where the level of river banks and force of the earthquake and as a result developed com-
the surrounding areas is raised this causes a significant pression wood from 1900.
change in the growth conditions on these sites. The effects of the earthquake can be clearly seen both
Jacoby (1983) studied this sort of chain of events in in the reaction of individual trees and in the mean curves
Alaska. It is known that there were severe earthquakes on for the area.
3rd and 10th September, 1899, and 9th October, 1900,
which caused the land-masses in Icy Cape in the Gulf of
Alaska to rise by between two and three meters. This can
be clearly seen in the trees growing near the coast. In
1980 the trees on the sea-shore terrace IV had a maxi-
mum age of 77 years, which indicates that germination ~
first became possible after 1900; i.e. when the ground
was relatively dry. The trees on the edge of terrace III
grew much more quickly with the rise in the level of the
'~~=:::::-;:::--------
.-
~
0

land, commencing this period of improved growth some-

time between 1900 and 1903. Before this time they had
been exposed to strong winds from the sea as well as the
effects of salt-spray. In addition, the fact that the water-
table had often been very high and that they were subject
to periods of flooding or erosion had meant that their root

Drawing illustrating some of the effects of tectonic movements.

Landslides cause trees from higher-lying areas (conifer forests) to
be pushed Into areas lower down (broadleaf woodlands) and be
buried. Some trees are tilted sideways. Tectonic falls cause the
water-level to rise, with the result that a number of trees die.
Tectonic uplifting brings with it an improvement in the hydrological
conditions on lake and sea-shores, which in turn promotes the
growth of the trees on these Sites (Drawing by Sebek).

before uplifting after uplifllng

III

III
I
I
I
IV uplift'ng I
IV
-1- I
-- 1---I
~pa Influpn(~ : normal a 'nfluen(e I normal conditions
conditions I '
I I

Morphological and ecological conditions on the coast of Alaska

before and after tectonic uplifting (Jacoby and Ulan, 1984).

1850 60 70 80 90 1900 10 20 30 40 1950

1900
Cross-section through a spruce in ICY Cape. Growth conditions
improved following tectonic uplifting In 1900 or thereabouts. The
resulting improvement in growth can be clearly seen in the annual
ring picture (Jacoby and Ulan, 1984).
years
Comparison of annual ring chronologies for trees which were at one
time growing near a river bank (thick line) and for trees permanently
on sites away from rivers (thin line). The growth of the former
improved following the uplifting that occurred around 1900 (Jacoby
and Ulan, 1984).

218 Jacoby, 1983; Jacoby et al., 1984.

Ice jams and icings

In the winter, rivers in the northern parts of the northern
hemisphere freeze over for a number of months. They
then thaw very quickly, usually with dramatic effects, with
the ice-cover cracking at various points. This causes the
ice floes to be pushed one on top of the other. Some of
these enormous masses of ice are forced onto the river
banks, parts of which are wooded, while others form
dams behind which the water builds up.
Parker and Jones (1981) have described the varied
morphological reactions which this chain of events brings
in its wake, the most obvious of these being the ice-
ridges which are formed on the river banks. Here the
action of the ice causes the vegetation to be prized away
from the banks, while loose stones build up into moraine-
like ridges. The different types of vegetation growing on
different parts of the banks are a clear indication of the
fact that earlier accumulations of ice have laid waste the
area at some time. Remains of sediments and driftwood
which are found well below the average level of the river
are evidence of past flooding. The effects of the build-up
of the ice can be seen in the damage sustained by the
trees and in the fact that some of the trees have been
tilted sideways. Ice floes which have been forced onto the
banks cause damage to the bark and cambium on one
side of the tree. The incidence of such occurrences can
be determined by examining the callusing on damaged
trees, and the level reached by the ice can be judged
by looking at where exactly on the trees the injuries
occurred.
Parker and Jones (1971) made an intensive study of
the build-up of ice in the MacKenzie region of Canada.
Their aim was to discover how far above the summer
water level ice scars on trees could be identified. It is
clear that it is inadvisable to build pipelines, roads, build-
ings or electric lines in areas where such damage is likely
to occur.
In stretches of the river that are wide and straight the
ice tends not to reach a height much above the water
level. Where, on the other hand, the river narrows, or at
the confluence of two rivers, the ice may accumulate, with
the result that trees up to 17 metres above the level of the
river will be damaged. Water may build up behind these
natural dams over an area of up to 100 km. The ice floes
in turn cause the trees in this area to be damaged. The
fact that the trees on level ground near river banks are
relatively young probably indicates that whole stands of
trees have been destroyed in the severe build-ups of ice
that tend to occur per century in these areas.

14m.a.r.1. edge
of terrace

low bushes 9.5m.a.r.1.

grass sedges

ice shove ridges 5.5m.a.r.1.

Ice jamming in the Bay of Bothnia (Alestalo and Hiiikio, 1976).

Mackenzie river

Schematic view of the river bank of the MacKenzie River, Green

Island. Of particular interest are the mounds of stones, the areas
with no vegetation at all and the recent phases of vegetation com-
prising willows and poplars (Parker et al., 1971).

Section through an elm which had been growing by the Tenile River
in New York State. Ice jamming caused damage to this tree in the
winters of 1975/6, 19n18 and 1980/81 (Jacoby, unpublished).

Parker et al., 1971; Alestala et al., 1976. 219

Tree ring research on fossilized woods
In 1847, a good half a century before dendrochronology
began to develop as a subject in its own right, the
palaeo botanist D. F. Unger investigated the rings of fos-
silized woods. And in 1911 W. Gothan sought an explana-
tion for the fact that woods from the Palaeozoic era do not
have tree rings. Since then biologists and geologists have
attempted to obtain information about climatic and eco-
logical conditions in particular geological periods. Today it
is widely accepted that dendrochronological analysis in
the strict sense cannot be carried out on fossilized woods,
since the scope and validity of the method is limited here
by the paucity of the material available and the huge time-
spans that come into question. It took 30 years, for
instance, to build up a 7000 year absolute chronology for
Europe based on a wealth of material from quaternary
gravel and peat-bogs. Is it reasonable to expect that a
chronology which would be more than a thousand times
longer than this, reaching back into the early Tertiary
period, could be constructed using very little material?
~~~
220
Despite the considerable amount of work that has been
done in this area it is really still only possible to tackle the
question of whether or not fossilized woods display
annual rings. And this question is worth asking only where
studies of the flora and fauna and the geology of the area
enable the following points to be cleared up:
What sort of organisms does the sediment contain?
By determining the species of wood involved and analys-
ing the pollen and diatom profiles and the zoological
macro- and micro-organisms on the same site it is pos-
sible to ascertain the ecology and sometimes even the
geological age of the deposits containing the wood.
Where the fossilized wood displays rings this provides
evidence of seasonality. Only where several species are
available for investigation, however, can the question of
seasonality be answered definitively, since even within
one climatic zone by no means all species possess the
same capacity to form rings.
Sections through tertiary (Eocene) silicified woods from the Amethyst
Mountain in the Yellowstone National Park, U.S.A. There are horizon-
tal trunks displaying tree-rings with boundaries of varying degrees
of definition in the strata between 248 and 253 metres. The trunks
originate from different sites. The fact that there is a variety of
species is a result of slope movements. (a-c) Sequoia magnifica,
(d) Podocarpus? (e, f) Pinus, (g-i) Aralia? (j) Artocarpus? (Fritz,
1982). (See page 218.)
Radial section through a Araucarioxylon stem from the petrified
forest in Arizona (Ash and May 1981).
Antevs, 1954; Creber and Chaloner, 1984; Chaloner et al., 1972; Fritz, 1982.
In the case of sub-tropical woods from the Palaeozoic era
it is often difficult to determine whether the growth pat-
terns display annual rings or not. Nor can we be confident
that certain fern species which are now extinct were even
capable of forming tree rings.
Where the fossilized woods are found
In the case of tree-trunks it can be said with any degree of
certainty that they are in situ only when they have
roots and are upright. Such finds are, however, made only
rarely. Stumps in carboniferous deposits are generally in
situ. Where the specimen is found with a large quantity of
other wood it may be part of an aCGumulation of driftwood
or of trunks that have been pushed together a~ a result of
a landslide. In such cases trunks from different climatic
zones - e.g. dry mountain areas and moist-warm low
elevations - and from different types of sites - dry
south-facing slopes and damp marshy sites, for instance
- may be found together. This means that pieces of
fossilized wood with and without rings can be present in
the same deposits.
In the case of fluviatile sediments fossilized wood from
different geological time periods may be found together.
In the Tertiary conglomerates in the north of the Swiss
Jura, for instance, fossilized palm and Angiospermae
occur with Dadoxylon. The latter originates from the
Liassic deposits of the Black Forest and must have been
eroded out of its primary site and then moved about 100
million years after it had become fossilized.
Seasonality
Chaloner and Creber's analysis of fossilized woods
(1973) has yielded the following information about sea-
sonality:
The first species with properly-developed xylem occur
in the Upper Devonian period. Only in very few cases
were growth increment zones formed (Cal/ioxoJyn
newberry, Indiana, U.S.A.). There are no tree rings
present, even in woods found in mountainous areas.
Conditions in the Carboniferous period were similar
for the whole of the Northern Hemisphere.
In the Permian system, on the other hand, conifers in
the Northern Hemisphere display no tree rings, while
those in the Southern Hemisphere formed clear rings
(AntarcticoxyJon sp.). (See map on page 12.)
From the Triassic period to the present day there are
woods in each of the geological periods which formed
tree rings.
These findings indicate that a seasonal climatic system
first developed on the earth with the splitting off of
Gondwanaland, i.e. in the Upper Carboniferous period.

What proportion of the original vegetation do the finds

represent?
Only a fraction of the original plant life has been pre-
served. This has occurred in the main through silification.
Since each of the stages of decomposition, even at the
level of microorganisms, has been preserved through the
process of fossilization, pieces of fossil wood can provide
a clue to conditions prior to sedimentation.

Standing tree trunks which have been fossilized in situ, Yellowstone Horizontal fossilized trunks in the Petrified Forest in Arizona, U.S.A.,
National Park, U.S.A. In the Tertiary Period, i.e. about 60 million have been exposed through the action of erosion. In the Triassic
years ago, a number of forested areas were buried in volcanic ash period, i.e. about 200 million years ago, vast numbers of trunks from
flows and destroyed (Fritz, 1982). fallen trees accumulated over millions of years (Ash and May, 1981).

221
Tree-ring analysis in entomology

Damage to tree-ring patterns

In forestry research the problem of bad infestations by
insects was previously regarded primarily from the eco-
nomic point of view because dying trees, reduced growth,
increased risk of disease, delays in producing new
shoots, all have an adverse effect on profits (Timell 1986).
At the present time, thanks in some measure to annual
ring research, it has been realized that 'harmful' insects
play a part in the ecological cycle and therefore their
activities are of considerable importance. The most valu-
able aspect of tree-ring research for the entomologist is
that it can throw light on those insect behavior patterns
which have long term effects. Insects are selected for
investigation which profoundly influence the physiology of
the tree, especially those which damage the light-assimi-
lating foliage of the crown, thus impairing the process of
photosynthesis. The resulting poor rate of growth can be
seen clearly in the annual rings.

,J.. Abrupt halt in growth. There is no gradual slowing down in growth

just before the tree dies. Insects which attack the cambium, e.g.
bark beetle, can cause the death of a tree from one year to the next.
Lodgepole pine, Bend, Oregon, 1984.
t

t Short-term growth reduction before the tree dies. Spruce budworms

caused severe damage to the crown of the tree in two to three con-
secutive years. The tree died in the course of the third year. Colorada
blue spruce (Picea pungens), Nederland, Colorado. Douglas fir
t iPseudotsuga menziesii) Nederland, Colorado, 1983.

t Recurring growth reduction over a period of several years before a

tree died. Spruce budworms attacked the tree during the whole of
this period. Red spruce (Picea rubens), Quebec, Canada. First attack
- 9 years before the tree died in 1983. Douglas fir (Pseudotsuga
menziesii), Colorado Springs. First phase of attack from about 1950
t to 1955. Second phase - from about 1963.

Tree-ring series of an oak in Denmark showing periodic damage to

tree rings, caused by cockchafers (Christensen, unpublished).

year of infestation 2/3 11

T T
year I
o
III I
10
III

Tree-ring sequence of a larch in the Valais, Switzerland, showing periodically recurring growth disturbances as a result of regular and
repeated, severe infestations of grey larch bud moth (Schweingruber, 1979).

222 Swetnam et al., 1985.

Insect damage recurring at regular intervals

Cockchafers occur in large numbers every 3 to 5 years; Larch bud moth does not seem to have existed in
the larch bud moth has a cycle of between 6 and 10 the Alps in pre-Roman times. There is no evidence of
years. Many tree trunks from buildings in the Valais show damage attributable to this insect in fossil trunks from pre-
that since the 15th century there have been attacks on Christian times. During the last 2000 years varieties of
approximately 12 occasions each century. Observations larch seem to have evolved which are resistant to attack
of tree-ring series from the 10th century and Roman by larch bud moth.
times confirm this.
E 6,-------~---------r----------._--------~--------_,._--------~--------~~
E5-
I\i'· .,. j \ ..
' . i'
· /\. \iV·\ '.
/! \ \./ \./ ·\ .e'· / · 1 V.
-£ 4-
~ 3- / i \ / \ / ..\ .1
§ 2-
j. i 'o,., / 'e. / •.•\
\ /\ • /\ ' \ \ \ .'\
~ .,. f / .e\o . ,••/ .,
.~. 01 - '.
t
.,
_
f
.e
f
.f . 'j •.•I \
e..
•
I'·
/
I I I J t I I
1840 50 60 70 80 90 1900
Tree-ring curve of 2 oaks from Denmark for the period 1830-1900, showing recurring damage at regular 4-year intervals (Christensen,
unpublished).

"':J E ?:
;;;
~ E fii
'" '0
1.0 -

8 8 8 8 9 length of
cycle
- -- - - - -. density

Curves showing tree-ring width (top) and maximum density (bottom) of a larch in the Valais, Switzerland. Damage occurs at irregular
intervals (Schweingruber, 1979).
• Munich

4 years •
Warsaw

rv larch -g rOWing areas m t heal ps

•Paris ~ 1970 )
3 years
P 11972 mfested areas

IllllIIl 1974
• Ice Baltflnsweiier

Distribution map for cockchafers (Melolontha melolontha and M.
hippocastani) with 3, 4 and 5 year development cycles. By examin-
ing tree-ring patterns of trees in these areas, it can be established if
environmental conditions on the sites had varied, either as a result
of climatic changes or human activity (Simplified, Schwenke 1974-
1978).
Map showing the extent of damage caused by larch bud moth in
various years. In 1970 it was concentrated in the western Alps, in
1974 in the east. A study of tree rings would show whether this
recently discovered shift from west to east actually began hundreds
of years ago (Baltensweiler 1971-1975, simplified).

Baltensweiler,·1976; Kulman, 1971; Schweingruber, 1979; Varley, 1978. 223

Insect damage which does not recur at regular intervals

Most of the insects which are responsible for defoliating British Columbia
trees do not develop in short, regular cycles - a conclu-
sion based on studies spanning a long period of time and Washington
annual ring sequences. It is well known that individual Idaho
trees and whole stands have suffered from insect damage 'U'
for centuries. An important addition to our knowledge has
u
c:: Montana
.s:
been that, even in a bad infestation, not all the trees of
one species in a given area are attacked simultaneously.
.
0
e=.
Oregon
I
I
California
Taking the example of the Tussock moth it can be shown
that within the distribution area of the host tree, Douglas
~
!! Nevada I
rIl
fir, there are very great differerces in the severity and Colorado
duration of the attacks as well as the extent of the areas
affected. These range from Arizona
New Mexico
(a) something approaching regular cycles and severe
infestations as in Idaho and Washington. 1900 10 20 30 40 SO 60 70 80
(b) irregular, sporadic infestations as in Colorado and Year
Arizona.
(c) areas free from infestations such as northern British Years with bad infestations of Tussock moth in the west of North
Columbia. America (Mason, 1978).

Attacks or outbreaks often show up in the pattern of the

tree rings. If a comparison can be made with ring pattern
in trees which have not been attacked or with resistant CANADA
varieties or species, it is reasonable to attribute any
reduction in growth in the host tree to insect damage. A
correlation exists between the extent of the damage to
the crown and growth reduction; the more severe the USA
defoliation, the greater the growth reduction. It must be
stressed that such statements apply only to whole stands c
and do not necessarily hold good for every single radius ~
o
of every single tree. u
;;::
Obviously trees suffer damage when attacked by in- '0
sects but taking a broader view because more space and "'
0..

light is created for the surviving trees, growth in some

stands is actually stimulated.

Distribution area of Douglas fir (Pseudotsuga menziesii) - dotted

area. Areas of severe damage (large dots) and light damage (small
dots) as a result of a bad infestation by Tussock moth in the west of
North America (Beckwith, 1978).

apo

r -":.~~:: ....::::~.-:.: I::.:.:::::::~::.:.:~

NATIONAL STAND 50 1 50 1 ! 0 50 80
FOREST ID

rLaEM
II)
W
Roo .... " DGU
II)
ij
L ...
II>
u u •• u.· 0
...I ...0
0
o X

~. ~: . u·.· .;z.:.. :.:::"~":~: .....:: ::"l::::::: ~:: ::~~~.:::::::···1 ;.:::::.:. . ~:::"::: ~

t- Z
o 0
-<
II:
~
0
0::
X
o II: ()
...I 0
o() IL
0
Z
0
z ¥
s.n ...b.{OCK
OCR .--- ~- -
:
-- -. ~- - - -. - - - - - -. -- .-- --T ----- __.--x... _-- ______
'Lf------------. ~------- -- J.
.1
0 w
;:: II:
-<
o::
W
t-
::> ::IE
0 0
...
o
"'-.-J~~:----~~~--. ~~---~~r:~--~~~~.~----~
0 0::
Z
___ _ ....T. ___.Y.. _.---- --
()
>C C.rlon [::: -< 0w
w 0 >
::E
~
GAR _. _ _ • ______ • __ _ - - .- . ~-.- - -- . ....l...:r.__ .___..L. _______ ...----:!. ._--1.L. .. _. _______ . ~. __~ ___ ~ z iii:
i w
w ;:: 0
z OSH •• ___ • .:L..-.. _________ .L -_______ ~- L...-n--muo ! ------.i....--- ------ ~ --- ....:!._u y
i ' I ; , , , I
170010 20 30 40 50 60 10 80 -VEARS
PERIODS OF INCREASE D
SPRUCE 8UOWORM
ACTIVrrV

Spruce Budworm growth loss (reduction in radial increment) and duration of growth loss periods, estimated from growth differences of
Douglas fir (host species) and Ponderosa pine (not host species) in some National Forests in Colorado and New Mexico. The infestations do
not occur in regular intervals (Swetnam, unpublished).

224 Wickman, 1978, 1980; Beckwith,1978; Mason, 1978.

The extent of damage
This has been well documented in publications on fores - the latewood within 1 to 2 years.
try-related subjects (Kulman , 1971 ). Damage to the crown The best way of assessing the extent of damage is to
takes many shapes and forms. These range from the study a great many trees on one uniform site.
shedding of just a few leaves, to whole sections of the
tree being enmeshed in a web, woven by insects, to the tree 8 tree 9
spring •...•..
tree being stripped completely bare of leaves. Every increment summer ___
reduction in the light-assimilating fol iage means a reduc- Q;

tion in the amount of tissue formed by photosynthesis.

0;
E 10
caterpillars _ _ .......:
This shows up in the annual rings, except when the loss lt 8 1\
7 I \ E
is made good by drawing on reserves of stored nutrients. Q;
6 E
"
c. 'I \
Losses in the crown and trunk are highly variable, even in a0'" 5 \ c
2
the same tree. If only a few branches are defoliated, then 4
.<:
i
e
~

the growth increments are only affected in these .,;

i§ 3 CJI
branches. Where a whole section of the crown is de- 3
foliated , then the annual rings , particularly at the base of "e
~ 4
the trunk are narrow whereas those in the crown and the !:l 2 5
root flange are formed normally. Only when the whole of 6
P <0.01 P <0.01 7
crown is severely damaged , particularl y the young leaves
which are very important to the physiology of the tree , 1951 52 53 54 55 56 57 58 51 52 53 54 55 56 57 58
does radial growth and the production of cell tissue for the Relationship between the extent of damage caused by an infestation
of cockchafers (number of caterpillars per mO) and the widths of
whole tree fall to a fraction of what it was previously. If earlywood and latewood in the annual rings. The latewood width
there is only one isolated attack the width of the annual correlates closely with the total area of the leaves. (Varley and
rings is back to normal within 3 to 4 years, the density of Gradwell, 1962).
Pseudotsuga menziesii Picea pungens
Tree 1981 1982 1983 Tree 1978 197 9 1980 1981 1982 1983 Needle-
No. Ew Lw Ew Lw Ew Lw No. cw Lw Ew Lw Ew Lw Ew Lw Ew Lw Ew Lw lost, %
1
• • 0 0 • • dead 1 •0 • • •0 • • 0• 0• 0 0 0 0 50
2
• • 0 0 • • dead 2 0 0 0 0 0 0 0 0 50
3 • • • • • • dead 3 • • • • • • 00 00 • • • • 60
4
• • • • • - dead 4
• • • • 0 0 • • • • 60
5
• • • • • - dead 5 • • • • • • • • • • • • 90
6 • • • • • - dead 6
• • • • • • 0• • • • • • 90
7
• • • • • - dead 7 • • • • • 0 0 • • • - 60
8 • • • • • - dead 8 • • • • •0 • 0 0•
0 • • • - 95
9
• • • • - - dead 9 • • • 0 0 • • • - 50
10
• • • - - - dead 10 • • • 0 • • • • • • • - 50
Symbo ls for ri ng w id th 11 • • 0 0 • • • • • • • - 90

• no rmal •
heavi ly red uced
-
12
• • • • • • • • • • • - dead
o reduced missing 13
• • 0 0 0 0 0 0 • • • • 75
Nederland, mo ist site Nederland, moist site
Semi-quantitative representation of earlywood and latewood widths Spruces, were first attacked in 1978 but survived several infesta-
in Douglas fir (Pseudotsuga menzies;;) and Blue Colorado spruce tions. By 1983 the crowns were certainly thin but the trees were still
(Picea pungens) on a uniform Site in Nederland, Colorado, infested alive. The Douglas firs were first attacked in 1981, after the third
by spruce bud worm. infestation they were either dead or dying (Kienast and Schwein-
year 1780 90 1800 10 20 a5 ber. 1986). 40 50 60 70 80
~ 20 Munster. Goms. 1400 m
- 0>
~ g> 16
_"0
~ ~ 12
0.<: 8
g .~ 4
n~~ __~~~-U~__~~~~~~~~~~~~~~-L~~~~______~____~"~-~'~-L~~-U~~__~

I
year 1880 90

I ~ ~~ Munst.r. G;;m; i
I4OO
m
~ ~ 12
'0 ~ 8 .
g.~ : A ,A., I
Extent of an infestation of larch bud moth on a uniform stand in an area favorable for the development of the insect in the Upper Valais,
Switzerland. The degree of damage is measured according to the number of trees affected per year, the annual ring width (upper line) and
the maximum density (lower line). Even in an area where conditions are optimum for the insect, the degree of severity of damage varies in a
period of time. The periods 1780 to 1820 and 1920 to 1960 stand out as being very bad, while the intervening hundred years were practically
free from damage (Schweingruber, 1979).

Schweingruber, 1979; Kulman, 1971; Varley and Gradwell, 1962. 225

Relationship between a site and the climate
Climatic factors and site conditions influence the com-
plicated mechanism governing the development of the
insect population and the reaction of trees to harmful
attacks. This whole area of interaction can be seen in the
way that the area affected changes from year to year and
also in the annual growth rates. Particularly in the case of
damage which has long term effects, such as that caused
by spruce bud worm, fluctuations within the same site are
very noticeable. When an insect such as the larch bud
I~
226
moth, which recurs in regular cycles, is studied it can be
seen that within a large area some climatic regions are
affected more often than others. Presumably unfavourable
climates, e.g. cool ones or those with long dry summer
periods, inhibit the growth of this particular insect. Annual
ring ana1ysis investigations, taking both time and geo-
graphical area into account, can throw light on this com-
plicated relationship.
Regions with average and heavy defoliation in firs (Abies balsamea)
as a result of attack by spruce bud worm (Choristoneura fumiferana)
in the Laurentide Park region, Quebec, Canada, between 1946 and
1954. The severity of the attack and the shape and size of the area
affected changed from year to year (Blais, 1964).
Semi-schematic diagram of annual ring width of 13 Douglas firs
(Pseudotsuga menziesii) in Colorado Springs, U.S.A. The abrupt
reduction in annual ring width indicates an attack, the subsequent
phases of narrow ring width formation show the duration of the
infestation. The trees were felled in different years. The years 1960
to 1962 are noticeable for being prac1ically free from damage. A
second wave of damage occurred in 1963.
Blais 1964
Relationship between defoliation, radial growth and mortality rate
.~
~
o
E
I
/
".
I I;s
I
2 0 2 4 6 8 10 years
Ild1111M11 I i
Effect of short-lived attacks e.g. by larch bud moth (Zeiraphera
diniana) recurring at regular intervals. Larches may be attacked
twice in the early summer. As new needles are formed each time,
wood growth comes practically to a halt and no reserves are laid
down in such a year. In alpine regions two attacks in one year mean
there is a 5 to 10% chance of a tree dying. When only one attack
occurs, the risk is reduced to around 1 %. Radial growth returns to
normal after 3 to 4 years; the density of the latewood is back to
normal levels after 2 to 3 years.
Conclusions
The following ideas originate primarily from Wickmann,
1979. Trees and tree-damaging insects have existed in
every geological period. Insect infestations have affected
woodland in epidemic proportions for millions of years. It
is known that larches in the Swiss alps are infested by
larch bud moths every 7 or 8 years. The areas where
damage occurs are clearly defined for each type of insect.
Growth reduction occurs in trees over the entire infested
area, but actual cases of mortality are limited to a few
individual trees in a stand and to a few places within an
affected area. Looked at from the economic point of view
what has been described must be regarded as a loss
(growth reduction, dead trees, under-stocked stands), but
from the biological point of view the death of some trees
can be regarded as an opening up of the forests which,
once the insect population has departed, will lead to an
increase in the number of germinating seedlings and a
much faster growth rate in the surviving trees. Particularly
in the rather unstable situation which exists near ava-
lanche gullies is this an ecological advantage, for insect
plagues mean a return to the situation which existed in an
Wickman, 1979; MacLean, 1981.
--
I
I --
1/
I"li;r.:i,
/i
~------------~~~--~~
;1mill1111 IIII
2 2 4 6 8 10 years
Effect of a longer-lasting attack, e.g. by spruce budworm (Choriston-
eura fumifera). After the first defoliation, severe but not complete,
radial growth rate drops dramatically. If the trees continue to lose
needles in the succeeding years, growth rate is reduced appreciably.
If the density of the crown becomes so low that it is no longer able
to supply nutrients for the formation of new needles and earlywood
cells, the tree dies as soon as all the reserves laid down in the
parenchyma cells have been used up. The longer the duration of the
attack, the greater the mortality rate; it can reach 90%. What usually
happens is that the attacks are just severe enough to keep the tree
on the border-line between living and dying for years. (Adapted from
MacLean 1981.)
earlier stage of forest development when stands were
very well stocked and they performed a valuable protec-
tive role .
If today's areas of damage are larger than before, and
this hypothesiS has still to be proved, it could be as a
result of the age structure and species composition of the
forests, in other words, infestations are a reaction to the
'man-made' forest. Fighting the attacks with chemical or
biological measures is wrong and can be justified only in
terms of short-sighted, economic interests. The aim of
forestry planners should be to create a forest which corre-
sponds as closely as possible to natural conditions.
Taking a broader view, insect activities can be regarded
as regulating rather than destroying. A new topic for
discussion arises. Is there a connection between the
relatively large areas of insect infestations and changes in
the environment? When we find out where and for how
long large areas of insect infestation have occurred, it will
be possible to get an idea as to whether such a connec-
tion could exist.
227
Tree ring research in phytopathology

Fungal attacks
The defence mechanism in the tree trunk
Damage in the region of the crown, trunk and roots offer
points of entry for fungal infections. As life on earth has
evolved trees have developed a biological system which
holds in check or even prevents the spread of mycelia. In
the center of the tree, particularly in the heartwood zone,
substances which inhibit fungal development are present
and in the sapwood zone living cells are able to produce
the appropriate substances as and when necessary. If an
infection sets in, the tree begins to form a barrier to stop
it spreading , aprocess known as compartmentalization.
1. Fungi spread quickly along the trunk axis.
2. The infection spreads more slowly from the outside
towards the center because thick cell walls in the
latewood and parenchyma cells on the annual ring
b(1)undary, producing fungal inhibitors, check growth .
3. Spread of infection tangentially is held very much in
check by the ray cells which secrete fungal inhibitors.
4. New tissue formed after a wound which has been
covered over with callus tissue will not be re-infected
from the inside.
9 years ago an oak sapling lost most of its bark. Cambium began
In the present-day situation of intensive forestry man- to form over the damaged places. Fungii aHacked the trunk (light
agement an effective system of defence against fungal areas) and caused discolouration (dark areas) but they could not
disease becomes very important. Mechanical damage, infect the new tissue (Shigo, 1983).
caused by heavy machinery or grazing animals, creates
vulnerable places but the natural defence system of the
tree keeps the damage within acceptable limits.

The biological defence system in the trunk (Shlgo, 1977).

radial section cross section
The branch of a bush was broken off. Fungal decomposition of the
branch and the tissue below the wound set in. The wound then
began to heal over. Fungal filaments spread along the longitudinal
axis in the centre of the trunk but came to a halt at the annual ring
boundary (Shigo, 1983).

Tangential barriers in an American elm after being infected by

Dutch elm disease. First of all a broad band of parenchyma cells
was laid down. The vessels then formed were soon blocked by
tyloses and filled with dark-coloured substances (Shigo et al.,
1980).

228 Time1/1986; Shiga 1977, 1980, 1986.

Damage due to peeling

In a meadow in the Swiss alps Bazzigher (1973) dis-

covered extensive peeling damage caused by goats
which led to a bad fungal infection in the open wounds.
74% of all spruces were damaged. By studying the callus
tissue, the time when the damage occurred could be
fixed as being between 1935 and 1945. Of the 441 trees
examined 43% showed signs of fungal attack in the
wounds, 10% had areas of discoloration and only 17%
were healthy.

1900 1910 1920 1930 1940 1950 1960 tree

no.
1 1 .1.
* = first ring • = Injury
156
176
181
169
40
189
131
242
166
14
22
250
75 1958
36
17
126 Cross-section through a spruce suffering from peeling damage.
105 Wood formed before 1935 has been decomposed by fungii. Wood
formed at a later date has been attacked by a second fungal infec-
tion. Along the annual ring boundary signs of the activity of the
The relationship between various spruces and the time when peeling natural defence system can be seen, either in the stages of decom-
damage occurred. Callous tissue shows that the damage occurred position (dark areas) or in stages of penetration (discoloured zones)
mainly between 1935 and 1945 (Bazzigher, 1973). (Bazzigher, 1973).

Condition of spruce forests

Niederer and Nippel (1984) working in the Rhone Valley,

Switzerland, examined the condition of trees in mountain
areas. About a quarter of all apparently healthy spruces
were found to be rotten in the center when core samples
were taken at chest height. The stability of these trees is
maintained only by the tangential fungal defence system,
formed annually, parallel to the tree rings. This supports
the healthy tubular-shaped woouy section.
The specific, fungus inhibiting qualities of the outer-
most, water-conducting rings is the only explanation for
Proportion of spruces with rotten cores in the lower Rhone val·
the fact that the mountain forests continue to function as ley, Switzerland 16 trees were bored at chest height on each site
protective entities despite being badly diseased. (Niederer and Nippel, 1984).

Growth-inhibiting effect of Armillaria Me/lea (Honey

Fungus)
On the basis of one investigation it emerged that fungi
can inhibit the growth of spruces. (Hrib et al., 1983). The
honey fungus causes growth reductions, particularly in
artificial stands. Losses are incurred as a result of growth
reduction and the death of trees.

Healthy spruces have broad annual rings; those attacked by the

honey fungus have narrow ones (Hrib, 1983).

8azzigher, 1973; Hfib et al., 1983. 229

Mistletoe Growth damage

Forestry research workers, particularly in North America,
have investigated the problem of the growth of the para-
site mistletoe. Many different varieties of the species
Arceufhobium, Viscum and Phorodendron attack trees,
usually conifers. The problem has not yet been studied by
dendrochronoiogists.
Damage to trees from this source is considerable.
Every year 3.2 million board feet are lost in the American
west as a result of attacks of mistletoe, which cause either
a reduction in growth in the host tree or its death.
Once a seedling has attached itself to a branch the vitality
of the tree is reduced. Suckers penetrate the vascular
system of the host tree, tapping water, nutrients and
assimilated substances. At the point where the mistletoe
penetrates the tree the cambium reacts by producing
callus tissue and an above average number of cells. The
formation of new cells, which is normally distributed
throughout the tree, becomes concentrated at this point
of infection, which means the vigor in this area is in-
creased.
EASTERN
WASKINGTON

r-"""T""- - - .
C]) /
~

[ASTERN
OREGON

Arceuthobium americanum (Hawksworth and Arceuthobium sp. on oak. Percent of comercial forest area infested with
Wiens, 1972). dwarf mistletoe in California (BolSinger, 1978).

1m

in reduction
~ period
no in relase period
reduction

-
() -~
'00
Suckers of Vlscum slbum on Ables slba.

"""
Radial growth of pines (Pinus sylvestris) with and without Viscum attak in forests near Chur, Switzerland. On each site 16 trees have
been cored. The black part of the circle shows the percentage of trees reduced in growth. Growth is in this case not negatively influ-
enced by mistletoes (Schweingruber et al., 1986).

230 Hawksworth et a/., 1972; Bo/singer, 1978; Scharpf et al. (edit.), 1978.
Damage

The condition of a tree or a whole stand can be changed,
both in the short and long-term by a mistletoe attack.
The extent of damage ranges from the loss of a few
twigs to the destruction of whole forests. Generally weaker
trees rather than stronger are attacked.
The longer the attack lasts, the more serious is its
effect on stand vigor. This is evident both in reduced
vertical growth and wood production (annual ring widths,
volume). Seed production declines, another sign of re-
duced vigor, and so propagation of the trees is also
adversely affected. Dead mistletoe shoots make a con-
venient point of entry for fungi. The resulting increase in
rotten wood leads to greater structural instability, which
makes the trees more vulnerable to storm damage. As
a result the mortality rate again increases. The ultimate
effect of mistletoe depends on the host-parasite relation-
ship, the severity of the attack, the climatic and ecological
conditions and the presence of harmful insects. However,
generally speaking, the mortality rates are influenced by a
combination of three primary factors; presence of mistle-
toe, insects and fungi.

The devastating effect of mistletoe (Arceufhobium vagina fum) on vigorous Ponderosa pines within the last 20 years In the Grand Canyon,
Arizona (Lightle and Hawksworth, 1973).

Ways of keeping the damage under control

Various measures are undertaken by foresters to reduce reduction of shrubs in the underbrush. The effectiveness
mistletoe damage and improve stand vigor. These include of these measures is clearly evident as crown size is
selective pruning, removal of severely affected trees, correlated with the process of regeneration; the larger the
which allows more light to penetrate the stands, and crown, the more successful is seed production.
+ heigh
• d.b.h .
o volume
20 E 81-100% live crown
80 .E
~ .l:
~
4

>
-5 60 e 61-80% live crown
'iii '"
~
Q)
.l: o 'i5 41-60% live crown
0 ~
40 ~
E :>
21-40% live crown
c:
'"
u c
8. 20
'"c
.,'"
2
0
E
20
0 2~

0 10 20 30 40 50 60 70 none medium heavy

time onfected ye~rs infection rating

Relationship between relative height, d.b.h., and cubic foot VOlume Relationship between radial growth, live crown ratio and dwarf
of dominant and codominant Lodgepole pines and the time Since mistletoe rating of red fir, 10 years after release (Scharpf, 1978).
the infection on the basis of 133 plot (Hawksworth and Hinds, 1964).

231
Tree-ring research in forestry

The aim of forestry is to cultivate the forests in such a way
as to achieve optimum production without disturbing eco-
logical balances. Measures are taken to stimulate organic
production while at the same time preserving or improv-
ing many other forest functions, such as providing recrea-
tional use and acting as a natural barrier to erosion etc.
Specific operations such as thinning, pruning and
fertilization lead to increases in yield in the short term,
and improved timber quality, in the long run. Although
several studies have been carried out on the effects of
such measures on individual trees and on whole stand
growth, few generally applicable conclusions can be
drawn, as there are wide discrepancies between the
findings. As a result, little is known about the connection
between forestry practices and tree ring structure.

Thinning

The value of a commercial forest depends to a great volume increases by measuring the diameter of the trunks
extent on the effective use of thinning, but production can at chest height and the height of the tree than by measur-
only be increased as far as site conditions permit. For ing annual rings in core samples. Variations from year to
example, on stony soil with poor nutrient supply only a year are less significant than differences between sites
slight improvement is possible, whereas large increases and forestry methods. Only in a few European countries is
can be expected in stands on rich deep soils. Foresters growth increment measured by means of core samples.
look for more than enhanced growth, for this factor alone
does not satisfy the demand for high quality wood in
today's market. They therefore try to influence growth
along certain lines by practices such as thinning, taking
the following factors into account:
production , i.e. establishing optimum conditions for
growth.
trunk quality, i.e. long straight trunks with only a few
branches in the lower part and with regularly shaped
crowns are preferred.
stability of the stand, i.e. a tree species that can with-
stand heavy snowfalls, freezing rain and storms.
Forest yield science uses a modified form of annual ring
analysis; volume increments in individual trees or whole
stands are measured periodically. It is easier to measure
height
m

Neglected beech stand. Inadequately thinned beech stand

(Leibundgut et al., 1971).
20
.')

15 0 -hr-....;;:::;r;:bF"""-----T'-----""'T- - . , ma ture
.., beech

10 .
+ 10
')
0 +'Ir""--.--t::....,---,-----.----, spruce
'i
10
5 I'>

0 1 I '1 t ~ t l) I I years
before thinning after thinning

Distribution of wood density in the trunk of a 68 year old fir of
approx. 40 cm diameter. 40 years after it germinated, when its
diameter was about 30 cm, thinning was carried out and space was
created for the tree to develop. For a short time density decreased
but returned to normal within a few years (von Pechmann, 1974).
Mean radial increment in years before and after thinning as a per-
centage of mean tree-ring width. The favorable effect of moderate
thinning lasts only a few years (Mitscherlich, 1970-1975).

232
Pruning

Pruning should improve the quality of the wood. Well- Eight of the trees had never been pruned, 8 had had their
formed branches and long sections of trunk without branches reduced by 50% in 1966. The effects of this
branches increase the value of the timber. Many reports heavy pruning could be clearly seen in a reduction in ring
have been made on the practical and economic effects width and a more gradual transition from earlywood to
of pruning, e.g. Lepetre, 1957, Polge et a/., 1973, who latewood. The maximum and minimum densities were
studied the effects of pruning on tree ring structure. In slightly higher than normal. Only very heavy pruning has
1971 core samples were taken from 16 trees with full any lasting influence on growth; as long as the tree has
crowns, growing in fairly open situations. The species a relatively large crown it can survive unfavorable con-
used were Pseudotsuga menziesii and Abies grandis. ditions.

-_~~~~==~~~~== - - 0em

......~~:;;::::;;~.~::---- 4.5
r----.. . ~.It~~~ -- 12
.........- -' --

- -21
6

- -24

Improvement in the quality of the timber as a result of pruning. If Features of the density curve used to determine the effects of
the place where a branch is removed is quickly covered by new pruning.
tissue, the wood fibres run more or less straight within about 2 cm (a) tree ring width
of the scar. In the case illustrated here all signs of a branch having (b) maximum density
been removed had disappeared within 6 years (Lenz, unpublished). (c) minimum density
(d) abruptness of transition from earlywood to latewood (Polge
et al., 1973).

.
:i Ables grandls
prunong "80
,, \ ~
,, \
\
:!!
.!!! Abies grandi s
pruning

,, g 6
"tl5 " \

E \ 0
0 4
\
\
E \
3 \
~ 3
.s:: , ~,
-
\ \
' '
,~

2
i5
~ 2
\
\
.. --- ~ ..."' 1
\
\
\- - -- , '
.,.---- - ~'"
E 0

:~
en .,.----,
c
.;:: ~
Pseudolsuga menziesii pruning c
<6
:0 ,g
C .;;;
C
,, c 5
"' ,---- , ~ 4 Pseudotsuga menzles ii
4 ,, a pruning
3
3
\
, ,, - /-------- 1lc'! 2
---- ,

-' - 0. 1
,
..... _----- -- - -- - -----
.0 01962
:0
2 t r I
1962 63 64 65 66 67 68 69 70 71 '" 63 64 65 66 67 68 69 70 71

Tree ring width decreases immediately after pruning but gradually Increase in density from earlywood to latewood occurs more
returns to normal. rapidly after pruning. As ring width increases the transition
_ average of 8 control samples becomes more gradual once more. (Polge et al., 1973).
- - - average of 8 pruned trees
xx significant differences.

Mitscherlich. 1970-1975; Paige et al.• 1973. 233

Fertilization and soil improvement
As with all forestry measures, soil improvement is in-
tended to raise the vigor of weak stands and increase
organic production . In cases where the soil is exhausted
or poor in nutrients and the exact conditions are known ,
appropriate measures can be taken to improve it. These
include:
application of fertilizers such as lime, potash, phos-
phates and nitrogen.
cultivation of legumes, e.g . clover, lupins, broom.
tilling of the soil; the effectiveness of fertilizers on the
growth of legumes and trees is increased by aeration
of the soil and improved soil structure.
Trees growing on sites with normal water supply react
quickly to the application of appropriate fertilizers. The
effects vary but are as a rule fairly short-lived, but a
considerable improvement in the value of stands or in-
Effect of application of fertilizer on needle and shoot growth in
young spruces. No fertilizer was applied to the left-hand shoot
(Baule and Fricker, 1967).
Effect of the application of fertilizer on the cambial activity of Pinus radiata on a poor site on New Zealand. The reaction began 3 years
after the fertilizer was applied (Kozlowski, 1971).
234
dividual trees can be achieved by means of fertilizers in
conjunction with soil cultivation and general forestry mea-
sures as described above. Whether such measures are
economically worthwhile , however, depends primarily on
the cost and effectiveness of the fertilizer and the produc-
tion costs and selling price of the timber.
Much has been written about the physiological, pedo-
logical, technical and forest yield aspects of fertilizers and
soil improvement (8aule and Fricker, 1967). Little work
has been done, as yet, on the effect of fertilizers on the
anatomical structure of tree rings.
Effect of an application of nitrogen to pines in the Black Forest (left
- no fertilizer). After the application of the fertilizer to the 190-year-
old trees, the needles became denser and the tree ring widths
doubled (Assmann, 1961).
 The following example describes the effects of applica-
tion of fertilizer (von Pechmann, 1960). A 63-year-old
spruce growing on variegated sandstone was treated with
nitrogen and phosphate fertilizer (between 1953 and
1955). As early as one year after the first application
changes could be seen in the annual ring structure:
the width, especially of the earlywood increased
greatly.
the latewood density was lower than in the previous
year.
the density sank from 0.475 glcc to 0.412 glcc and
the compression strength showed a parallel reduction
from 513 kg/cm 2 to 410 kg/cm 2
It seems that cambial activity had been stimulated by the
nitrogen. It follows that the yield of forests on acid soils
with poor nutrient supply can be increased by the cUltiva-
tion of nitrogen-assimilating plants, e.g. lupin, on the site.

mm
3.5
fertil ized
3.0 surface area
fertilization
2.5
~

U
.~
2.0
Ol
c:
.-=
m
:::l 1.5
c:
c:
<t>

1.0

0.5
t

0
1930 1957

Effect of fertilization with nitrogen on annual ring width in spruce on Earlywood-Iatewood transition in spruce before (left) and after (right)
a site with poor nutrient supply in Germany; the ring width increases application of fertilizer. The transition becomes more gradual (von
(von Pechmann, 1960). Pechmann,1960).

m '"
'"
<1l 10
_;:~u =t 5
w~
u ~
0
2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 mm

less than 15).1 III 16 -30)..1 . 31 - 45,u UIIID 46 - 60).1 radial cell diameter

Tree-ring structure of a 63-year-old spruce on a site with poor nutrient supply before and after application of fertilizer. Widths of
earlywood, latewood and the whole ring increase. The density difference between earlywood and latewood, and the average latewood
density decrease (von Pechmann, 1960).

Assmann, 1961; von Pechmann, 1960. 235

Study of yield in tropical forests

As has already been shown (p. 109), traditional methods tively small area. It is not uncommon to find 200 types of
of dendrochronology are completely unsatisfactory when tree species growing on one hectare of land. It can be
determining yield characteristics such as tree age and regarded as a challenge, therefore, to work on growth
radial and volumetric growth in tropical forests. It is not rings in certain tropical regions . Eckstein et a/. (1981) and
surprising, therefore, that relatively few dendrochronol- Worbes (1985) showed that such work could be success-
ogists have studied tropical trees. Considering the enor- ful by a study carried out in a seasonal tropical climate.
mous variety of species, it is even difficult to investigate Using 70 cross-sections taken from the trunks of 20
the basic underlying botanical characteristics of a rela- Pisonia grandis trees Eckstein et al. tried to establish the
relationship between ring width and meteorological fac-
150 tors. Although the definition, counting and measurement
humid of tree rings was not always possible, a clear relationship
between precipitation in the rainy periods and growth
emerged. In hot, humid, rainy periods the carbon balance
seemed to be unfavorable, presumably because the high
preci pi tation
temperatures which stimulate respiration and photosyn-
100 E thesis are reduced as a result of dense cloud cover.
E
mean max. c 14
temperature .2
..-
30
--.- ......... e 1.2 800

-......, ""- ...... a.

'uQl 51.0
I ·,...... .-. 50 a. -s 0.8
"0
600 ~
mean min. • _._. .~ 0.6
temperature '"
c
c 0.4
0.2
preClpllallon
arid O ~~~-r~~--r-~~~~~----~~-+ o
a 1980 1975
dale
1970 1965
J F M A MJ J A SON 0
Climatic diagram for Heron Island. The yearly change from humid to Mean ring width along radii of maximum growth of Pisonia grandis
arid is typical (Eckstein et al., in Bormann et al., 1981). on Heron Island, and wet season rainfall, defined as the total for
January, February and March in each year. Ring width chronology
based on 8 trees and 14 radii up to 1971, 6 trees and 10 radii up to
1965 (Eckstein et al., in Bormann et al., 1981).

Methods of determining growth

Studying cross-sections of the trunk

By examining well-polished sections of tree-trunk, it is In addition to the traditional methods of microscopy and
possible to judge whether a tree is forming definable, mechanical-physical examination, the suitability of the
incremental growth. What had appeared to be definable following newer methods should be considered as to
annual growth when core samples were taken, was often their suitability: X-ray densitometry, image-analysis, or
found to be localized, one-sided trunk growth when optical photometry.
cross-sections were studied. Often tangential bands of
parenchyma cells or ill-defined tangential pigmented
zones can be distinguished from rhythmic incremental Tropical Aride Zone Florida Java
limits only on closer examination. Only rarely can distinct rain forest Savannah
rings be seen all around the trunk. Even then, it is Africa Africa
possible that several rings have been formed in the same distinct rings
year. Various authors have tried to build up a picture of not necessarily
the dendrochronological applicability of some species in annual 16 20 15 10
certain areas. The following table merely indicates the
morphological spectrum that can be represented within a indistinct rings 1 5 21 32
tropical area. without rings 7 3 51 21
A variety of methods are available for investigating
growth structure. Only occasionally, however, are these
methods more accurate in defining areas of growth than Number of trees investigated with specific growth forms (Fahn et al.,
1981). Trees with and without rings are to be found in both tropical
what can be achieved when examining trunk cross-sec- zones with a consistently wet climate as well as those with a
tions with the naked eye. seasonal climate (Eckstein et al., in Bormann et al., 1981).

236 Eckstein et al., 1981; Fahn et al., 1981.

Age determination of trees in plantations of known age
Deciding whether growth rings are formed annually or not
is easiest if the exact date of planting is known as thus
tree age can be compared with growth structures in
cross-sections taken from the trunk. Of the many tropical
species - and there may be as many as 7000 - there
Cross-sections of tropical woods.
Top: Da/bergia retusa with distinct areas of growth and possibly
even annual rings.
Middle: Mllletia /aurentli with continuous, tangential growth areas but
no annual ring formation.
Bottom: Ochroma /agopus without any rhythmiC growth structures
(Gottwald, 1958).
Fahn et al., 1981; Gottwald, 1958.
are only about 100 being cultivated intensively which are
suitable for study by dendrochronologists. It will never be
possible to reproduce accurately the natural conditions of
the tropical forest in plantations, because only a few
varieties can be grown in the test areas and these are
often conifers growing outside their natural habitat. More-
over, cultivation of trees in these test areas is very dif-
ferent from the growth conditions which prevail in the
dense forest. But in spite of all this, investigations con-
cerning growth structure carried out in such areas have
great advantages: the relationship between the actual age
of the tree, annual climate and site and the morphology of
growth can be examined. In one such investigation in
Brazil it was found that the same variety of tree under the
same climatic conditions, but on different sites, formed
rings on one site, but not the other.
Determination of growth by dendrometers and markings
Dendrometers give information about growth. A simple
tape dendrometer records the extent of radial growth
whereas dendrometers with the facility to make daily
recordings are capable of providing information about
chronological growth rhythm. The relationship to the
structure can be established only if chemical substances
are injected into the layers of cambium in the trunk with
needles. When the relevant tree sections are examined at
a later stage, it is possible to ascertain the stage of
development that the xylem had reached at a given point
in time. (Fahn et al. 1981 .)
alt itude
20
I
annual rings Ipresent' absent present
drainage good I bad good
climate sub- I sub.-trop. tropical
trop. I
Pinus eliotti plantation. On well-drained sites trees only form annual
rings when a water shortage exists in dry periods of short duration.
Density diagram of a tropical deciduous tree. Neither In the cross-
sections nor in the density diagrams can seasonal growth fluctua-
tions be distinguished from annual rings.
237
Tree-ring research and wood technology

The density of wood is one of its most important techno-

logical features. It is influenced by biological, physical and
chemical characteristics and a relationship exists between
such things as cell wall thickness, cell size, compression
strength, breaking load, modulus of elasticity and lignin
content. It is often the case that density determines the
suitability or otherwise of a particular wood for a partic-
ular purpose, which is the reason why radiodensitometric
analysis is so important in research into wood technol-
ogy. It is hardly surprising that even in the early days
such studies (Green, 1965, Hughes 1968, Keller 1968,
Phillips 1966, Sanderson et aI., 1960) raised technological
questions, an even more frequent occurrence in later
research.
Little mention has been made in this book of tropical
woods, an omission which will now be rectified by refer-
ring to some relevant material from the work of such
authors as Plumtree (1984). Pines native to the Caribbean
and Central America (Pinus caribaea, Pinus patu/a) are
cultivated extensively in tropical regions today and are
very important to forestry.
Radiographic analysis is made more difficult by the
presence of resins which can make up to 60% of the total
weight. After they have been extracted the irregularities in
density can be seen clearly in the density curves. On sub-
tropical sites annual rings are present with maximum
densities often in the middle of the rings. The proportion
of latewood is very high as a rule. The average density of
the wood is therefore higher than that of pine-wood in
temperate zones (0.5-0.7 g/cm 3). The width of the
annual rings in trees grown on plantations is striking. In
the center it is not unusual for them to measure 2-3 cm,
decreasing rapidly towards the outside. The ageing pro-
cess is completed significantly faster under these condi-
tions than in cooler or more arid zones.
Ring width, average ring density and cross-sectional
areas vary in a characteristic way. These reactions, similar Pinus caribaea var. hondurensis. 7 year-old trees in a plantation on
from species to species, seem to be partly controlled by Melville Island, Northern Territory, Australia (Plumtree, 1984).
climate. For example, Plumtree (1978) discovered that the
annual volumetric growth and dry matter content in Pinus
patu/a increased with heavy rainfall.
To ensure the optimal use of wood from conifers
growing on tropical plantations it is important to know
something about annual rings and density structures
within the tree. So Plumtree (1978, 1984) drew up density
profiles for Pinus Caribaea along several radii and at
several different heights. By extrapolation he succeeded
in determining the density variations using suitable
mathematical methods.

Cross-section through a 18 year-old Pinus caribaea growing in an

area of afforestion in Cuba. The annual ring width decreases rapidly
towards the outside as a result of crowding on all sides from other
trees. Characteristic of all tropical trees are the large number of
intra-annual density fluctuations, the high proportion of latewood
and the high average wood densities. In the example shown they are
about 0.7 g/cm 3 (Specimen J. F. Hughes).

238 Plumtree, 1984; Sauvala, 1982.

 KWAMOONAY81 RING wlOfH £ 010 Ot.NSIYv X secYJONAl ARo.
l O'~

.>0 C C7

." o C8 ear ... ban

I'" I'" INt '9W 1"0 '." .'1. ..,.

YEAR

Annual variation in width of tree rings, mean density and cross- of what it was in the first. Density increases with decreasing ring
sectional area of pines in an afforestation in Uganda, Africa. The ring width. Annually changing ecological conditions are reflected prin-
width decreases very rapidly; by the fourth year it is only one third cipally in cross-sectional area (Plumtree, 1974).
m~ volume em
5 tree average
kg dry matter em 0,14 140
35 350
0,12 120
30 300
0,10 100
25 250 c
~
c
2 <l> 0,08 80
g
200 E
~ 20 § :::l
a.
E 0. "0
> 0,06 G
60 ~
>- 15 150 ;)
-0 ~
a.
a.

10 100 0,04 40
preclpl allan
volume
5 50 0,02 20

056 0 0,00 0
60 65 69 56 60 65 69
Years Years
Relationship between annual rainfall and wood characteristics with production of wood is lightly stimulated. The relationships between
importance to forestry in Pinus patula in an afforestation in Uganda cross-sectional growth and rainfall are partly positive (1963), partly
(Plumtree, 1978). There seems to be a positive correlation between negative (1967).
annual rainfall and production of dry matter. With high rainfall the
tree·ring width density g/cm 3
> 18mm 27 >0,5
27
12 18mm 0,4-0.5
6-12mm 0.3-0,4
24 24
<6mm <0,3

21 21

18 18

~ ~
-
.C)
r:
'Qi 12
15
....
.<:
Ol
'Qi 12
15

.r: .<:

9 9

6 6

3 3

o
180 o 30 60 90 120 150 180
- bar~ plth- diameter (mm)
diameter (mm)
Contour diagram of Pinus patula trunks from Uganda. With the aid periphery of the trunk. Densities along the pith in the centre of the
of densitometric photographs of trunk sections, taken at different trunk are less than those at the outside. The highest values are to be
tree-heights, the relationship between tree ring width and density for found in the lower, outer sections of the trunk (Plumtree, 1978).
the entire trunk was extrapolated. Ring widths are usually greater in
the centre of the tree (right side of diagram) than those on the

239
Tree-ring research in environmental sciences

One of the main tasks of dendrochronologists in the
future will be to work on the problems of environmental
pollution. Finding the answers to the following questions
will be of paramount importance:
How do different species of trees on different sites
behave when subjected to various kinds of pollution?
When and where did growth disturbances in trees first
become evident and in which areas?
Dendrochronology has on the one hand to supply the
information required for political decisions to be made
and, on the other, to carry out biological follow-up studies
of a political-technical nature. In this respect dendro-
chronologists are in the hot seat in the current proble-
matic situation.
Methods
Most dendrochronological studies to date have been con-
cerned with solving historical, ecological, and climatologi-
cal problems by means of statistically random sampling.
For this purpose efficient but time-consuming measure-
ment and calculation techniques have been developed. In
the current situation, where large areas of forest are
dying, it is often necessary to define the areas where
(Johnson and Siccama, 1983; Schweingruber et al.,
1983). Sudden reductions in growth should be investi-
gated when localized and regional instances of forest
damage occur because they are an indication of some
kind of disturbance. Because recognition, dating, and
quantifying of disturbance are simple, such an investiga-
tion gives important results very quickly.
Every site can provide the following information based
on 20-30 trunk cross sections or core samples.:
Percentage of trees suffering from growth reduction.
Age distribution of trees with growth reduction; be-
ginning and duration of phases of inhibited growth.
Connection between growth reduction and Ule out-
ward appearance of the tree, e.g. loss of needles,
outward signs of damage, its performance compared
with other trees in the same forest.
Relationship between damage and site, e.g. hydro-
logical conditions, elevation, proximity to buildings
emitting toxic substances. All available dendrochrono-
logical methods are being used in research into pollu-
tion, including methods of identifying chemical trace
elements in tree rings (Baes III and McLaughlin, 1984).

{...
regions
damage is occurring by means of large scale sampling. reductions

The large number of samples requires the use of very

simple methods to get an indication of the cause. Often pme II c 1 . I 1~
measurements are not necessary, as all that might be
I IL~1
(flul
• I h . Itil

r"
needed is a date for interpretation.
..1 ~. ..1.1 I
Dating of sudden changes in growth behavior spruce
11 .. /1. I ..1
A.I'~<lLJ
dill
Previous studies have shown that growth irregularities SOil)

occur in trees weakened by the effects of pollution. Ih I'll

.. 11111 11.1 III
I. 1.1111/ I.
Wal
. -........1!I.. 1980
1974
fir
tlur
. .1 I II lIul _.J....
I

diLl 1/ ludl/1I11I JJll

,I'\,JrlJt.lU
I.
'ioo
Thurn

{,-,.
relEases
II

pine . 1 U
Lhw
f. .1 .. II JU

r"
JLL
spruc(~ A:,'g 'u
.. ilii.
5010
'tIUII1
IJ LU/ .1.111

~ pointer year with small latewood

or small ring width
WillS
Llll
~ duration of change fir .~

Modern firs with abrupt growth changes (See page 48.)

~Q
I L._I _I .~ J lID
Percentages of different conifers from various parts of Switzerland Thurn
with abrupt reductions (top) or recovery (below). In this century, 1930 1940
reduction phases occur every 11-16 years. Supraregional trigger
years are 1934, 1962, 1970, 1972, 1973, 1974, and 1976. The year percentoqe of trees wl\h
1956 is typical only for fir (Schweingruber et al., 1986). beglnnong growth changes . 1 1 1 I 2 <)3 101 ~.'l

240 Johnson et aI., 1983; Schweingruber. et aI., 1983, 1986; Baes III, 1983; Baas and Bauch, 1986.
Regional growth reductions: United States and Central Europe

Both in the eastern part of the United States and in
Europe areas of forest are suffering from damage, the
cause of which is not clear (Johnson and Siccama, 1983;
McLaughlin et aI., 1983; Schweingruber et aI., 1983;
Hornbeck and Smith, 1985). In both areas an abrupt
growth reduction and a high mortality rate have been
detected in conifers. The trees on both continents began
to show signs of damage around the 1950s but the full
extent of the damage was not recognized until the 1970s,
when the high mortality rate became obvious. In the
forests not cultivated for commercial reasons in the
U.S.A. over half the trees have died in some localities in
the 30 years since the sudden growth reduction set in. In
the European forests, which are important commercial
concerns, conditions may be similar but are not so obvi-
ous as the dead trees are felled, albeit prematurely, by
foresters. In both cases structural changes in the forests
must take place. Johnson et al. (1983) discovered that in
certain stands in Vermont there were about 40% fewer
mature red spruces but 300% more silver birches.

28
o WhIteface MI.. N.Y.
24 Mounl MarCIeId. V1.
• Mounl WuhIng1on. N.H.
g 20 % spruce. Swiss Plaleau and Jura (Aargaul
~ 16 n = 657
14
~
u:12

8
*'
c:
"'
12
10
4 ii '"
.r; c
~ .2 8
o ~~~~~--~--"
1950 55 75 o U
Year ~~
c
6 ~

'cc
28 i 2
-5
24 .~
0
~ 20
.,c: ~ 2
g 16 '0
I
u: 12
i!J, N4
~ ~
~ ~ 6
! B I
Year
10
1900
I
10
I
20 30 40 50 60
I
70 1980

Frequency of the occurrence of tree-ring reduction in red spruce

(above) and short-leaf pine (below). Damage to spruces began in
1960, to firs in 1956. It appears not to increase, but this is only fir. Swiss Plateau and Jura (Aargaul
because the majority of the trees suffered continuous damage in the %
n · 620
1960s (Johnson and Siccama, 1983). 14
I
l00 r---~~------------------, 12
Key
• Soulhern 'oresl
l> Northern boreal foresl
80
a Northern Ir8n51110n 'oresl
o Northern hardwood foresl

60
0
l>
40
0
808
20

• • @8~

10 20 30 40 50 50 70
I
1980

Percentage of dead conifers in forests in the south and north of the Percentages of spruce and fir in the Swiss Mittelland and the Swiss
U.S.A. Damage occurs much more frequently in the north than in the Jura with incipient growth changes (reductions above the central
south (Johnson and Siccama, 1983). line, recovery below). Spruce is less sensitive to environmental in-
fluences. (Schweingruber et al., 1986.)

241
Regional growth reductions

In the inner alpine dry valley of Valais, Switzerland, trees The fact that damage to the crown was greater than to
both in orchards and forests have been showing signs the tree-rings in the period 1982/83, when the investiga-
of damage for many years. Most probably waste gases tion took place, remains for the time being an unsolved
containing fluorine, emitted by aluminum factories in the problem. The following data have been compiled on the
area, are at the root of the trouble. An abrupt reduction in basis of tree ring analysis:
the width of annual rings in conifers provides the clearest
indication of the damage. An ecologically oriented sam- 1975 - the percentage of trees with an abrupt reduction
in tree ring width was 66% for pines, 69% for firs, and
pling network was set up and 600 firs, 600 pines and
1300 spruces were investigated dendrochronologically. 37% for spruces. After 1976 many trees recovered so
The tree species reacted very differently to adverse that by 1982 only 46% of the pines, 37% of the firs, and
external conditions. The pines reacted earliest, the first 22% of the spruces were in a phase of reduced growth.
time in 1921 and at intervals ever since, while firs showed No differences in damage ratios could be detected be-
the first signs of damage in 1962. Spruces came some- tween dry and wet sites or those with basic or acid soils.
where in the middle. In the fifties some trees, especially
pines and spruces, recovered. At the present time, pre-
sumably as a result of tighter controls on the level of
fluorine in factory emissions, all three species of conifers
are recovering. Significant differences came to light con-
cerning time and place. In 1921, a bad year for damage,
spruces in central Valais were worst affected. In 1962 the
damage was worst in the lower Valais and in 1974 in the
neighbourhood of Martigny. All in all, the lower Valais has
suffered most. Above 1600 m, the level of the inversion
layer, hardly any damage has been recorded.
spruce, Wall is
0.
'" %
pine, Wallis
% ~
70 n = 1327 100 "&E n = 621
:J!
70 100 ~
:J!
. 80 l"
"0
"0
'"C 60 80 ~
'"c 60 ~
.Q '"~
.,
0
u::J '13 C>

50 60 .S ::J
.S
e
"0

..<:
'0 e
"0 50 60
'0
..<:
~ 40 40 ~ ~ 40 ~
e
Ol e
• 40

£ 30 '"
£ 30
.~
.~

~
~ 20
'0
~ 20
'0
0 0
Z 10 Z 10

0 ~=E~~~~--~~~""4 o ~. .-~. .
1900 10 20 30 40 50 60 70 1980
1900 10 20 30 40 50 60 70 1980
fir, Wallis
~
n = 634
~__~~------~----; loo ~ Percentages of modern conifers undergoing growth reduction in the
present century.
:J!
"0 black: heavy reduction (> 70%)
80 ~ shaded: reduction 56-70% in comparison with previous
l" periods.
fir white: reduction 40-55%
60.S
Each species reacts individually to environmental conditions: while
'0 pine shows strong reactions as from the 1920s, fir exhibits severe
40 ~ responses only from the '70s onwards. Today, all three species show
more or less vigorous recovery. It is noticeable that growth reduc-
tions increase spasmodically (Kontic et al., 1986). (see page 48.)

Z
o 10
Percentage of trees in the Valais with abrupt growth reduction in
tree rings (dark segments) and percentage of those recovering
(shaded segments). Core samples were taken from 16 trees on each
o-=~~±::::Cl:::::r:...~ site. Damage occurs in the whole of the Valais but is concentrated in
the area between Martigny and the Lake of Geneva for firs and
1900 10 20 30 40 50 60 70 1980 spruces (Kontic et al., 1986).

242 Kantie et al., 1986.

 Symbols

1000m 75% (12 trees)

\"h ,,_'" ,,~I

altitude with growth red uctlons
of the site of these 25% (4
Fir

- S,erre

Pine

'liSP

Spruce

Brig

CJ
~)]~

'.00
~"9"Y

243
Damage due to local environmental factors
There can hardly be a more unsuitable location for trees
to grow in than the present-day town, inundated with
traffic. The vitality of the tree is sapped in various ways:
by damage to its roots, trunks, and crown, when the soil
around it is compacted, covered over, or contaminated by
hydrocarbon gas, oil, or salt, and by air pollution (Hoester,
1977). It is thought that about half the trees in the cities of
Europe are threatened with winter salting of roads being
the main cause. Petersen (1983) and Joos (1985) indicate
Cu Hlng green branches
.J: _10
~~~8
eg"'6
i~~4
ci ~.& 2 1 ~~
c 0 l:I:=l..LJ:-~Jt=t=lJ:l~tr.~JC:~::3...L_--L 0 ~
1920 30 40 50 1960
bl ROOl damages and soil compressIon
Pollution effects in city trees, Basel (Joos, 1985).
Key:
Shaded areas: Number of trees (maximum 10) with noticeably nar-
rower rings than in normal periods. (Left hand scale)
Columns: Average width of tree rings in mm in periods of damage
and regeneration. (Right hand scale)
During the regeneration periods the rings formed are more than
twice as wide as in the periods of damage (After Joos, 1985).
244
that the trouble began with the introduction of motorized
vehicles at the beginning of the century. Eckstein et al.
(1974) found that the composition of woody tissue and
also the resulting physiological conditions were changed
with decreasing cambial activity.
Horticulturalists, traffic planners, and biologists are
cooperating in an attempt to prevent further air pollution in
towns.
(a) Effect of cutting off branches form the crown of plane
5 E trees (Platanus X hybrida) (arrows). The crowns of the
4E:~ trees were cut at irregular intervals of between 4 and 10
3-::E years. As a result radial growth dropped, in the short
2~~ term, from 4 to 2 mm but one or two years later, the trees
had already recovered.
(b) Effect of building operations. During the building of an
ice-rink in 1933 and subsequent extensions in 1942 and
1960, the growth of plane trees (Platanus X hybridus)
was impaired for between 4 to 8 years as a result of root
damage and soil compaction. The trees recovered, how-
ever, on each occasion.
(c) Effect of soil compaction and paving on horse-chest-
nut trees (Aesculus hippocastanum) during road con-
struction. Trees have been suffering from the effects of
soil compaction and insufficient water supply since the
beginning of the century. Particularly since the introduc-
tion of the steam roller in road construction (Macadam
process) the width of new growth in annual rings has
gone down from 2 mm in 1902 to 0.3 mm today. Re-
generation is almost impossible.
(d) Effect of winter salting on horse-chestnut trees (Aes-
culus hippocastanum). Since the practice of sprinkling
salt on roads was introduced about 1960, conditions for
the growth of trees have deteriorated rapidly. Contamina-
tion and the resulting physiological disturbances have
caused a decrease in cambial activity. Even on only lightly
salted roads, the width of the tree rings has gone down
from 4 mm to 0.5 mm within a few years. Regeneration
occurs only to a modest extent, and then only if less salt
is used for several consecutive winters, if less salt perco-
lates through to the roots, or if a lot of salt is washed out
of the soil by heavy rainfall.
(e) Effect of hot, dry summers on sycamore trees (Acer
pseudo-platanus) The summers of 1943, 1947, 1949 and
1952 showed clearly the effects of periods of dry weather
on trees, in this case sycamores, already under stress.
Hoester, 1977; Joos, 1985; Eckstein et al., 1974.
Growth reductions related to natural environmental factors

Certain environmental factors are responsible for the weather conditions or damage to the crown after two or
variability in the width of annual rings. Many authors have three years. It is much more likely that the effects of a
tried to explain the sudden reduction in width by correlat- long-term problem such as environmental pollution are
ing it with natural environmental factors. In the Valais intensified by additional, natural factors.
region of Switzerland (Kienast et a/., 1981) and also in Information about the effect of elevation is contradic-
New Jersey in the eastern part of the U.S.A. (Johnson et tory. Johnson et al. (1983) and Schuett (1981) found that
al., 1983) certain connections between damage and low damage increased with elevation in the northeastern
precipitation were found. In both cases low precipitation U.S.A. and in the Black Forest of Federal Germany.
seemed to trigger off a reduction in annual ring growth in According to Johnson, many factors might be respon-
certain years as in the Valais between 1941 and 1946 and sible: "Wind speed, exposure to cloud moisture, sulfur
in New Jersey between 1954 and 1957. However, a direct dioxide deposition and heavy-metal content of the soil
connection does not exist and it is certainly not possible increase with elevation whereas soil mOisture-holding
to explain the length of time that the damage persists by capacity, soil pH, base status and soil and air temperature
means of climatic conditions. Under normal circum- decrease with elevation." Kontic (1983) suggests that
stances trees recover from the effects of unfavorable trees growing above frequently occurring inversion layers
mm in the Valais may be exposed to fewer harmful agents and
c for this reason suffer less damage.
0 40
'';::; >-
"'-
.~ . ~ 30
.9- I
1il ~ 20
es..E
10
c 0
0
'';::;
<.>
:J 10
".c ~
20
~
e
0)
30
40
'0
ci
c
1950 1960 1970 1980
Relationship between precipitation in the months of May through
July and the frequency of new cases of sudden reduction to growth Relationship between precipitation in the months of April to July and
rings in shortleaf pine in New Jersey, U.S.A. The surge of damage the frequency of new cases of sudden reduction to growth rings in
could be connected to the dry summers around 1955 (Johnson and Scots Pine in the Valais, Switzerland. The relationship between
Siccama, 1983). damage and low prepitation was particularly noticeable in the Forties
(Kienast et al., 1981).
100rr--~r---'----.----r---.----.----"

•
90 Key
Southern forest
80 6 Northern boreal forest
0 Northern transition forest
70 0 Northern hardwood fares
~
~ 60
a
E
0
50
C l::.
Q)
<.>
a; 40 /"
a. 0 6 ... 6t::'
30
0 //~
20 " 6
/"'''~ ceJ
10
0
0 400
/6
600 800
-e
e
1000 1200 1400 1600 1800
• .-e- e-
elevation (m)
Relationship between mortality and increased elevation in red Inversion layers: damage is more frequent below these layers than
spruce in the north- and south-east of the U.S.A. Damage to trees above. Spruce forests on the forest limit at about 2100 m demon-
increased with elevation but not in Virginia and North Carolina strate no comparable reduction in tree rings.
(Johnson and Siccama, 1983).

Johnson and Siccama, 1983; Kienast et al., 1981; Kontic et al., 1986;
Schuett, 1982. 245
Relationship between growth reduction and recovery and the biological characteristics of trees

Damage to the metabolism of individual trees, entire effects of various polluting agents on individual trees and
stands, and whole species are reflected in a number of species but it is enough to pOint to the complexity of the
different ways. This is why it was necessary to investigate situation. The external picture, i.e. the total leaf area,
34 different species of trees in an extensive study carried corresponds more or less with the cambial activity. The
out in the eastern part of North America. From the den- more severe the needle loss, the less growth in the
drochronological angle, very little is known about the annual rings, and the higher the mortality rate is. Com-
parisons between trees from different sites e.g. dry and
moist, are unreliable because of the dissimilarity in growth
Percenl spruce showIng dleback
conditions. There are also inconsistencies in the defini-
tions of crown.
The time lag between an abrupt drop in growth and the
death of a tree varies. In Scots pine the 'sick' phase may
last for 25 years on average but there are wide deviations
even within a species. Some survive for 10 years, others
for 60. The time lag is shorter in firs. In Switzerland it is
between 15 and 20 years on average. Norway spruce
usually die within 3 to 6 years whilst beeches survive for
a maximum of one year from the time when the first
[>0 symptoms of disease appear.
00 [> o
o 00 t:I>
00 [> 0 [>

Comparison between the extent of needle cover and the mortality

rate in red spruce stands in the southern (black dots) and the
northern Appalachians. With thick needle cover the mortality rate is
low; with poor needle cover, it is high (Johnson and Siccama, 1983).

c: pine 1940 1950 1960 1970

.2 15 I I I
U
"
"C
~
10
-5.<:
~ ~ 5
'" 0
~~
-en 0
-c:
o .~
ci- 5
z ~
c:
c: 10
"'
~ 15
c: for 1950 1960 1970 1980
Q
tJ 15
I I I

"
"C
~ 10
-5-6
.~
i: 5
.,'" 0
~

~~
-en
-c:
0 -..::::
ci- 5
z c:~
c: Comparison between the vitality claSSification with crown and tree
'" ring characteristics on spruces in many stands in northern Switzer-
~ 15 land. The black part in each circle represents:
- the percentage of trees with poor needle cover (dark: loss
Time lag between physiological damage, seen by an abrupt reduc- > 30%, point 15-25%)
tion in growth (light columns) and the formation of the last annual - the percentage of trees with abrupt growth reductions since
ring and the death of trees in different stands in Switzerland. Pine: 1965.
reduction occurring in the thirties and forties led to the trees dying On each site 16 trees were investigated. There are severe disagree-
in the sixties and seventies. Fir: trees damaged in 1956 and 1963 ments between the crown aspect and the expected growth behavior.
died at the end of the seventies (Schweingruber et al., 1983). (Schweingruber etal., 1986.)

Johnson et aI. , 1983; Schweingruber et al., 1983, 1986; Kontic et al.,

246 1986; Fritts and Swetnam, 1986.
The question of causality

The most frequently asked question when forest damage
is being discussed is 'What causes it?' Dendrochronol-
ogists are trying to correlate statistically the connection
between the timing and degree of severity of damage,
and the spread of emissions, by means of tree-ring
analysis. The diagrams below represent an attempt to
correlate available knowledge but should not be taken as
proof positive. An attempt to find a direct connection
between acid rain and annual ring patterns failed (John-
son and 8iccama, 1983). Results are beginning to come
in on the subject of the CO 2 effect. Correlations are
coming to light regarding the connection between 80 2
emissions within small areas, tree-ring width suppression,
and the chemical constituents of the rings.
As many more results are expected in the near future,
only a few examples - partially controversial - are given
here to indicate the direction of current research.

Cades Cove
400 sulphur in the ice of a gl acier in the Alps
(4500 m.a.s.I.)

'°1
C>
~ 300 :§ 2.0
EE E
co

"' 200 '"c:

~u 0.0
Q

~c:
'f 100
B 6.0
c:
8
II)
c ~
n a 3.0
~.g % growth reductions in firs in the Alps
~u 60 (approx . 1000 m.a.s.U 199
_::J
o~
~~ 40
E~0
Q) Shortleaf pines in the Smokey mountains, Tennessee, formed nar-
~ S, 20 rower rings with higher iron and titanium content as a result of the
~..c high SO. content in emissions from a copper-smelting plant between
.~
~
1863 and 1912 (Baes III and McLaughlin, 1984).
O~~,--.~~--~~.---.---,--,,---
1900 10 20 30 40 50 60 70 80

Comparison of the sulfur content in approximately dated ice layers

in the glacier of the Colle Gniffetti near Zermatt, 4500 m.a.s.1.
/
/\\
(Schotterer 1985) and the percentage of firs with abrupt growth 70 / \
reductions in three alpine valleys. Firs grow in altitudes from
approximately 900-1400 m.a.s.l. After a first slight increase in sulfur
--,/ \
emissions in the early '60s followed a second wave in the early '70s. \
The curves of the sulfur content in the ice and the percentage of firs 'x---x---K..... . .
with growth reductions look very similar. But the question is still ........... .•• •••• ....x Fir SOlothurn
open whether this obvious agreement of events is only a statistical 60

/x~:~\ Pine Wail is

correlation. We don't have measurements from other glaciers and
the growth behaviour of firs in other regions is different. Dendro-
chronological date from Schweingruber et at. 1986 and L. Bronzini, *.S (+ F)
Firenze, unpublished. '"c:
.S'

/
U 50

7
ppm "e
"0
Fluoride ...\ ........ .
~...
'"c: 100 r.
8
~ •• ••• Pone Chur
c:
g 5
~
.
~80
1.2
Cl
-5 40
§ Fir Wall is I+F I
3 (;
260 ~
u. ,:: .x ... ·... F" Chur
1 Fluoride in leafs
40 30

§
'0
0
Spruce Wallis
Spruce Solothurn
(+ F)

.=
'"
c I
20 +---'---.---r-~~-'---,-
.,., I
I 74 75 76 77 78 79 80 years
t= I
1910 1930 1940
Correlation between the fluoride content in and on leaves of apricot
Years trees in the Valais, Switzerland (Zuber et a/., 1981) and the number
of trees suffering from growth reduction in various areas of Switzer-
In Nevada U.S.A. the phases of growth suppression in Pinus edulis land. As also in areas not affected by fluoride the proportion of trees
seem to bear a direct relationship to SO. emissions from a nearby with growth reduction is decreasing, this is purely a statistical and
copper-smelting plant (Thompson, 1981). not a biological relationship (Schweingruber et a/., 1986).

Johnson and Siccama, 1983; 1980; 8aes II/ and McLaughlin, 1984;
Stuiver and Quai. 1981; Schweingruber et al., 1986. 247
Radiocarbon method
At the end of the 'forties Willard Frank Libby developed
and proved the theory that cosmic radiation produced
unstable radioactive carbon C4 C) which was assimilated
by plants in the process of photosynthesis and incor-
porated in organic compounds. Since all living things are
directly or indirectly nourished on plant substances, radio-
active carbon must also be present in the bodies of all
animals, including humans.
Libby calculated the half-life of 14C to be 5568 years.
On the basis of the state of knowledge at that time, it
was assumed that a sample with a 14C content of only
50% was 5568 years old and one with a 25% 14C con-
tent was 11140 years old. As the concentration of radio-
active carbon compared with nonradioactive is very small,
(1 : 1000000000000) measuring instruments for testing
the method had to be developed. In 1949, Arnold and
Libby, using wood and annual rings of known age,
showed that these methods were suitable for determining
the age of organic substances. Libby was awarded a
Nobel prize for his work in this field in 1960. In the 'fifties
counting techniques were improved and thousands of
measurements were taken. By the early 'sixties it was
clear that 14C datings for the period before 1000 B.C. did
not agree with the dendrochronological absolute ones.
Libby suspected at first that mistakes in dendrochrono-
logical datings caused by false annual rings were the
reason. However investigations on dated tree-ring series
verified the deviation from the physical assumption of the
constant half-life.
13r---------------------------------,
samples of known age
12
1.1 ?tree ring (580 ± 50 A.D.)
......
C)

E 10
Q:
i!"
'> 9 Tayinad675 ± 50 B.C.)
"8 Redwood(979 ± 52 B.C.)
'"
u
;;:
'0
Q) 8 ,
;

a.
'" curve calculated
from present day point
and known half life of Zoser; Sneferu
7 radio-carbon, (2650 ± 75 B.C .)
5720± 47 years

6
0
historical age (years)
The first graphical representation of the relationship between the
rate of decay of the radioactive carbon Isotope and dated wood from
various periods of time: Tree ring 580 A.D. - beam from the Indian
settlement Broken Flute in Arizona. Dated by A. E. Douglas, 1931 .
Ptolemy: sarcophagus from the period of Ptolemy, Egypt. Dated on
the basis of its style. Zoser: Cypress and acacia wood from an
Egyptian grave. Dated by archaeological methods (Arnold and Libby,
1949). ''''' -
Willard Frank Libby (From Libby, 1980).

Long-term deviation ro , ..,

<0
.
LO
After first Schulmann (1958) and later Ferguson had ~
"'"
Q)
extended the Bristlecone pine chronologies by thousands C)

of years, Suss, 1967, and later many other authors '"c:

0
discovered a long-term deviation. Annual ring sequences -e .""
of living and fossil Bristlecone pines and sequoia from '"
u
.Q
the south-west of North America and oaks from Ireland '0
and southern Germany confirmed this trend. Recently ~ 6000

four laboratories have defined this deviation as being

sinusoidal (Klein et al., 1982). At the time of the greatest
deviation, about 4000 B.C. to 6000 B.C. the 14C dates are
about 800 to 900 years too young. At the present time, as
a result of the combustion of fossil fuels, they are about
300 years too old (De Vries effect).
Presumably the oak and Scots pine chronologies of
Europe with that of the Bristlecone pine of North America
can be constructed for periods of up to 10000 years ago.
(Ferguson and Graybill, 1983; Kaiser, 1979.)
' roo
calibrated date (AD- Be)
Relationship between radiocarbon ages (ordinate) and dendrochron-
ological ones. As a result of radiocarbon datings and absolute dated
wood, (Bristle cone pine, oak) a long-term deviation from the ex-
pected half-life was discovered (!<Iein et al. , 1982):

Libbyet al.• 1944, 1949; Klein et al., 1982; Schulmann, 1958; Ferguson
248 and Graybill, 1983; Siiss, 1967; Kaiser, 1979; Douglass, 1931.
 '~[
I

.., 1

b'
Q) 600
Cl

v'"
I
U ' 00
III
:::l
.S
E
u 200
·c
-0
C
Q)

·200

"

· . 00
I I I I I I I ..L -.l I -,-_',--,--,_
6OOO 11C SSOCIIC SOOO IlC ' 5OO11C ' ~IIC 35C01IC JOOOoc 1SOOoc ?OOOIIC 15OO8C I~ec 5000c .~ec RC 500 AO 1000 AO ,500 RC Z~

calendric date (AD-Be)

Discrepancies between the ,.c dates, in years, and the dendrochronological age. Taking the year 5000 B.C., the dendrochronological age is
about 800 years older than the ,.c age. The curve showing the deviation was compiled from data from five laboratories working on ,.c
dating (Klein et al., 1982).

Short-term deviations
Repeated measurements on wood samples of the same
age comprising only 1-10 annual rings, using the most
modern technical equipment, has led to the knowledge
that a long-term deviation is superimposed by a more or
less irregular short-term one. These so-called 'wiggles'
have a world-wide application for they can be detected in
a wide variety of trees , growing under very different
conditions in various parts of the world . As absolute
chronologies, extending back over thousands of years,
have been constructed in California, Ireland and southern
Germany, work on the calibration of 14C curves is going
on in several laboratories.
Stui ver (1982) calibrated the last 2000 years, and
Pearson et al. (1983) the period from 2000 to 4000 B.C.

600

500 <* , f

a..
H
co 400
¢ ~~
~.

Q) .~ .
* ~~~*
Cl

'"c0 t~ " , 00
E

-e
300
~ ~
u
.. ." "

'"
u
.2
$Qt
0 "
-<l
.....
91j!9~
-0
200
VanCOuver Islimd
e 09
*
0
P,eH... Co
o Seattl e douglas f ir t o9 00
· O'egon

* ,: I
100 01 {mplC Peri
Ml Ralnlcr
X Belfast irish oak

0
000 100 800 AD

1400 1500 1600 1700 1800

dendroyear AD

,.c content of tree rings in Douglas firs (Pseudotsuga menziesii)
(circles) and Irish oaks (crosses). The level in both types is practi-
cally identical. The relatively short-term, synchronous ,.c variations
In the northern hemisphere are shown. The length of the vertical
lines denotes standard deviations (Stuiver, 1982, in press).
,.c content in tree rings of Douglas firs on various sites on the
Pacific coast of the U.S.A. from 1000 to 1880 in relation to the oxalic
acid standard measurement (O-line). The deviations (called wiggles)
can be clearly seen (Stuiver et al., 1980).

Stuiver, 1982; Stuiveret al. , 1980; Klein et ai. , 1982; Pearson et al. , 1983. 249
Reasons for the variations

According to Oeschger (1985) there are two possible
reasons for the short and long-term variations: "a solar or
geomagnetic modulation of the cosmic rays or a fluctua-
tion in the carbon cycle and the associated dynamic
exchange, which bring about a change in the CO 2 content
of the atmosphere."
The short-term 14C fluctuations in the tree rings cor-
respond to the beryllium-10 COBe) in ice in Greenland.
Both fluctuations bear a close correlation to the medium-
term sunspot cycle of the last 800 years. The short-term
fluctuations in the 14C content are certainly due to
changes in the rate at which it was formed. By measuring
the 14C content in the tree ring accurately and by taking
the different lengths of time that the radioactive carbon
remains in the atmosphere, biosphere and in the oceans,
medium-term changes in solar activity can be determined.
It is difficult to furnish proof of an 11 year cycle of sunspot
activity as the amplitude of the relevant 14C variations
is very small. A relationship between solar activity and
climate is also difficult to prove (Stuiver and Quay, 1980)
but according to Suss (1979) such a possibility cannot be
ruled out.
Long-term variations are to be found only in the 14C
content in the tree rings and not in the 10Be content of
Greenland ice. Oescher (1985) explained this as follows:
"Ocean currents have become stronger since the end of
the last Ice Age which led to an equalization in the 14C
content in shallow and deep sea water. This has been
responsible for a drop in the 14C content of the atmo-
sphere."

30 r

1«'" 20 1
10 ·
OL
V)
100 [
8. 80 l Wolf Sporer
mini mum Number of sunspots (bottom) and intensity
~ 60 (velocity) of solar wind (top) calculated on
~

'" the basis of measurements of 14C content of

annual rings Since 1000 A.D. The number of
sunspots observed (11 year mean) agrees
well with the calculated values for the period
1640-1980. The minimum phases, which
1800 have been given names, are of interest
1200 1400 1600 (Stuiver, 1980).

Combustion of organic fossil fuels

There is a connection between the drop in the 14C content I .......,..........~--~.,....,-.----rj-.~

of annual rings in this century and the use of fossil fuels.

I
By burning organic material, millions of years old, radio
inactive carbon is released into the atmosphere which
then mixes with active carbon already present, so that the
14C level sinks. As plants absorb CO 2 from the surround,
ing air and convert it into plant tissue by photosynthesis,
the 14C level fixed by these processes corresponds to the
<J . 1
1,
level in the atmosphere at that time. By measuring the 14C
content of the alpha-cellulose in individual tree rings it can ••
be proved that the 14C level has sunk significantly since
the beginning of this century.
Radioactive carbon formed during the hydrogen bomb
explosions and fixed in the wood in the inner part of
the trunk can be removed by chemical extraction of sub-
stances in the heartwood. The radiocarbon content of the
extract reflects the artificially acquired radioactivity and
14C content of tree rings of Douglas firs on the west coast of the
that of the lignins the combustion products of the in- U.S.A. in industrialized times. Obviously the 14C content dropped
dustrial age. after about 1900 (Stuiver and Quay, 1981).

250
Hydrogen bomb explosions The consequences for historical research

There has been an incease in the level of 14C in tree rings
which is clearly due to the hydrogen bomb tests of the
sixties. When heartwood is formed, organic substances
are transported to the inner part of the trunk, leading to
additional biologically determined variations in the 14C
content.
{l"e jOioo)
800 r--------------------------------------,
700
Present hypotheses on the sequence of events in pre -
historic times are based to a large extent on uncorrecte,d
dates, determined by radiocarbon methods. Taking long-
term variations into account, a start has been made on
correcting the errors, but this has led to confusion among
historians as allowances have not been made for inac-
curacies in the original measurements or short-term varia-
tions. The time will not be opportune for a revision of the
existing chronological systems until the variations in time
have been established on the basis of long, absolute
annual ring sequences. If the results of first attempts to
revise existing chronological time scales are any indica-
/1 tion , significant changes in certain historical periods are

.\
/.
to be expected. Renfrew, (1977) for example, proved that
600 the theory that the western European Megalithic culture
developed out of the Mycenaean culture of south-east-
\ ern Europe is untenable. When it was thought that the
500 •\ Mycenaean and Megalithic cultures occurred at about the
.... same time, this hypothesis was tenable. But, as new
studies have shown the absolute date of the western
400 European culture to be centuries older than was pre-
viously thought, it is impossible that the one developed
from the other.
300

200 ..
I

100

.. tt •• .
O ~-------- -- . --- ~.~~------------------__1
.+ ~.·I ~'I' III.
/ " I ••• ..". 1'+
- t E! ' .~·lfi.<tf tt
-:
100 __ ____ __ _ __ _ ____ _ _ _ L_ _ _ _ 0 400 " c radiocarbon dates
1900 1910 1920 1930 1940 19:'0 1960 19/0 <i. from annual nngs
0
age of annual ri ngs l!)
to

14C level in the tree rings of an oak tree in Central Park, New York. 300
The roughly 700% increase in the 14C level as a result of the
hydrogen bomb explosions can be clearly seen. As substances are Cl
c:
-- . -----
transported to the inner parts of the trunk by biological processes
the 'bomb peak' does not show up in the sapwood-heartwood iti ~
transition zone as clearly as might be expected (Cain, 1979).
"0 220
c
200
-e
0
~ r----r----r---~----r---~----~---" to
u
0
A '6
"0 600 ~
c¥- 100
u 400
~ ci
<l 200
<i.
~ 0
OJ
0 100 200 300 400
O r-----------~------ (1950 A.D.) (1550 A.D.)
1910 1920 1930 1940 1950 1960 1970 1980 annual ring dales before 1950

Determination of the growth rate of trees without tree rings by
means of 14C measurements. In fast-growing tropical trees the bomb
peak of between 1962 and 1964 can be determined, using 14C mea-
surements. The degree of 14C contamination in the trunk provides
Information on the rate of growth (Stuiver et al., 1981).
Relationship between dates determined by tree ring measurements
and those by means of radiocarbon dating. (dark band). As 14C
levels in the atmosphere vary, the relation between annual ring
datings and 14C datings is not linear. This has various consequences
for dating measurements e.g. a 14C dating of 220 ± 50 years (hori-
zontal band) can mean that a sample is actually 150-210, 280-320
or 410-420 years older (Stuiver, 1978, simplified).

Cain, 1979; Stuiver, 1978 and 1981; Renfrew, 1977. 251

Stable isotopes

When water containing carbon dioxide changes from one

physical state to another, i.e. evaporates, condenses or 3? ~-------------.
freezes, the ratio of the isotopes changes; (2H/,H = (PH 30 .........., +.. +. ,-- 098
[deuterium]), the oxygen C6 0/ '8 0 = 0 '8 0) and the carbon
( •
cell u lose '+'+ ~ 28
C2 C/ '3 C = 0'3C). These changes can take place in the ~
o
20 .......... o
~ 24
•
0 .......

atmosphere, the soil or in plants, particularly in the leaves. iA ........

10
~ ....
Current research into radiometric dating of tree rings is
trying to find the connection in the chain of events linking
;. 0
.-;,
"!..
o
20
,,'7.... .... 00
drnonton

external isotopic values with those in plants, particularly in

P
- -10 '0 . . . r' = 094
. 16 ....
<l precipitation 0b-. -!
the cellulose in the tree rings. Up to the present, no ..., Coppermlnc
12 +--.--,--,--,-~
general relationship has been found but a few examples -20 0 15 304560 7590
15 10 5 0 5 10"(:
have indicated that in the future, climatic reconstructions latitude
N • S
based on measurements of isotopic rations can be
expected.
Burke and Stuiver (1981) discovered the connection 0' .0 values in precipitation at various elevations and the cellulose in
between 0 ,8 0 and elevation and latitude. Gray and annual rings of Douglas firs growing at the corresponding heights
above sea level SMOW = Standard mean ocean water (Burk and
Thompson (1979), working on the Pacific coast, estab- Stuiver, 1981).
lished that both east and west of the Rocky Mountains the
0 '8 0 values sink as the temperature drops. This trend is
repeated in the cellulose of trees growing in both south-
ern and northern latitudes, from California to Alaska and 1500
the Yukon and in the lOW-lying and higher parts of Wash-
ington, from the Pacific coast to Mount Rainier on the r' = 0.98
timber-line at 1900 m.
]: 1000 0"",
White (1983) working on white pine (Pinus strobus) in
the north east of the U.S.A., discovered various connec- ~ 0
"0
"0

tions between values of the hydrogen isotope 02H in 2 o

"'-0
""
trees and external conditions. Choosing one dry site with <Q 500
shallow soil, one average site and one damp site with 0",
deep soil, he studied the 02H values on a yearly basis.
Only the 02H in the sap of trees on the shallow site
corresponded, for a few days, to that of the precipitation.
The 02H in the sap of the trees on the average site and
0
13 -12 - 11
1:.' "0 in precipitation
%. SMOW
-10 - 9 25 26
1:.'"0 In cellulose
%.SMOW
27 "
28

the wet one reflected a mixture of the values in the

precipitation and the ground water. The 02H value of the
cellulose in the annual rings seems to have a direct Relationship between the 02H of water in the sapwood of white pines
correlation with that in the sap. It is therefore possible to on a dry site, an intermediate one, and a wet one, a single fall of rain
on 1st July, 1979 and ground water (diagram on the right). Also the
calculate the 02H value of the precipitation from that of relationship between the water in the sapwood in the tree on the dry
the annual rings on shallow sites. Thus Lawrence et a/. site and the precipitation (diagram on the left) (White, 1983).
(1982) were able to establish the area of origin of storm
trajectories in the north east of the U.S.A. as the 02H
value of each storm is characteristic. The more southerly '"
and the nearer to the sea it is and the lower the tempera-
tures are, the lower is the 02H in the precipitation and the
cellulose of the white pines.

lO
7 ! _
,..... •...........
.. J
.... ...
«>
-4

,.,.

~ \ .. ~ 6 , 8 ..,
Rthft SDj"'-l JuI"IQ'(I

Relationship between the 0 ' .0 in precipitation and cellulose in tree
rings in various conifers on the Pacific coast of the U.S.A. according
to Burk and Stuiver, 1981 (diagram on the left), and in conifers and
peat in the Rocky Mountains (diagram on the right) (Gray et al.,
1980).
Relationship between the deuterium content of the cell sap and
cellulose nitrate in Pinus Strobus in New York at various pOints
in time during the vegetation period. A close connection exists
between both parameters (White and Lawrence, 1980).

Burk and Stuiver, 1981; Grey and Thompson, 1979; White and Lawrence, 1980;
252 White, 1983; Lawrence and White, 1984; Long, 1981; Leavitt and Long, 1984.
 It is hoped that measurement of a13C isotope will give
us an insight into the nature of the carbon cycle, in
particular the carbon dioxide in the atmosphere. Many
questions in this branch of research remain unanswered.
The fact that many trees, at least those growing in dry
a
regions, contain the same amount of '3 C suggests that
not only signals specific to one individual type are pre-
sent. It remains to be seen what significance this will
have.
The variations in the values of the hydrogen, oxygen
and carbon isotopes in the tree rings are considerable
and these change from year to year. In 1983 Brenninkl
meijer observed an elm growing on a damp site with
deep soil in northern Holland and found differences in the
isotopic values which could not be explained climatically.
No single environmental factor seemed to have influenced ,... 1910 I~!O 1910 I~ t9SO
the separation of the isotopes more than another. Y all

013e variations in tree ring series (5 year mean) of pinyon (Pinus

edulis) on a semi-arid site in the south-west of the U.S.A. (Leavitt et
al.,1984).

r1
-.0
6 2H 1'/.,1 6 2 H1'/•• 1 ,......J

l L
- 60 J L rf
.r J
, fr t J ~
J '
- 80 • J 1-

+)0
5,e OI ·/.. 1 S,eOI'I.. 1 L.,
+28

+ 26
1

I J
'- f L

1
L

l (11':
- 22 5'lel'/.. 1 s"e 1'1.. 1
f r "
- 24 r l
r

- 26 1976
1975

'~
~
70 65 60 55 50 45 .0 30 20 10
dIs tance f r~ c.amblum [mm)

Variations in the hydrogen, oxygen and carbon values in the annual rings of an elm in Holland. The 02H rises at regular intervals after the
formation of the first earlywood vessels. These seem to consist of assimilated materials from the previous year. In this case neither
climatic factors nor biological proporties of the tree rings seem to have a direct relationship to the 0'·0 and o'3e curves (Brenninkmeijer,
1983).

The good results presented here should not be allowed
to disguise the fact that often isotopic measurements on
annual rings do not reveal any clear connection to en-
vironmental factors. Long (1981) points out that many
problems remain unsolved. When CO 2 and water undergo
biological changes certain processes take place, some of
which are temperature-dependent, some not. Physical
signals are masked either in the cellulose, the lignin or
throughout the wood during the formation of new wood or
at a later stage during the development of heartwood.
Together with these biological problems, various unsolved
technical ones exist too.
The positive results lead us to suppose that in spite of
obvious gaps in knowledge, it is possible to collect in-
formation on the environment. Our present state of
knowledge suggests that local site factors and differences
between the various species of trees influence the separa-
tion of isotopes more than general climatic factors. Future
research should take these facts into account.

253
 V The History of Dendrochronology

Transverse end section of a piece of historic wood. Bruno Huber dated pile dwellings with pieces of oak.

F. H. Schweingruber, Tree Rings

© Kluwer Academic Publishers, Dordrecht, Holland 1988
The history of dendrochronology (Europe and USA)

If a detailed history of dendrochronology were to be Phase Field of research Methods

published, the bibliography of publications on annual rings Pioneer Climatology Measurement of ring
would fill a small volume on its own. H. Egger has already
Botany of trees width by means of
over 2000 titles in his data bank in Neuchatel. Let us magnifying glass and
Astronomy
leave such a task to an historian . I shall restrict myself microscope.
to outlining the main points with the help of some exam-
ples and to describing some of the particularly exciting Development History Measurement of ring
episodes. Archaeology width by machines
Basically dendrochronology has passed through the fol- Expansion History Measurement of ring
lowing stages: Archaeology width by machines
Climatology computers
Geomorphology X-rays, image analysis

The first tentative steps

In the densely forested Emmental in Switzerland every logical knowledge to such an extent that by 1855 Theodor
farmer knows that the narrowest annual ring of the last 40 Hartig (1805-1880) had a clear conception of the de-
years was formed in 1976. In the extreme drought of this velopment of tree rings. Robert Hartig devoted his life
year wells ran dry, luscious meadows turned brown and fir to tree-ring research and published 34 papers on the
trees produced less wood than usual. It is common anatomy and ecology of tree rings between 1869 and
knowledge that there is a connection between tree growth 1901 (Antevs, 1917). By the end of the last century he
and the weather. Observant people in days gone by, was dating hail, frost and insect damage in trees.
among them foresters of note, became interested in the
formation of tree rings. Early research began in the follow-
ing areas:
Anatomy of wood .
- Botany of trees (ecology, physiology)
- Climatology
In Europe tree-ring research has a long history. As early
as the 15th century the relationship between tree rings
and precipitation during the vegetation period was recog-
nized by Leonardo da Vinci and the first references to this
subject were written by him (Stallings, 1937). According
to Schmucker and Linnemann (1951), Marcello Malpighi
(1628-1694) in Italy, and Nehemiah Grew (1628-1711)
in England, laid the anatomical foundations of tree-ring
research using the newly-invented microscope. Their
major works on plant anatomy appeared in 1682. The
Frenchman, H. L. Duhamel de Monceau identified the part
of the wood where new tissue was formed in 1758 but it
was not until 1828 that C. F. Mirbel recognized the signif-
icance of the cambium. John Hill, in his illustrated book
The Construction of Timber, published in 1770, distin-
guished the "circles of the seasons" , i.e. the earlywood
and the latewood .
In the first half of the 19th century, botanists such as J.
J. P. Moldenhawerof Kiel (1766-1827) , C. F. Mirbel of
Paris (1776-1827), and H. von Mohl of TLibingen (1805-
1872), had added to the state of anatomical and physio-

Theo Hartig (1805-1880)

256
 By the beginning of the First World War the main
foundations of tree-ring research had been laid. Antevs
(1917), lists over 300 papers on this topic in the period
1840-1917. Until about 1920, however, actual dendro-
chronological studies, in which tree-ring sequences were
compared with long-term meteorological records, were
rare.
In 1859 the German immigrant, Jacob Keuchler, living
in Texas, was interested to know whether the drought
occurring at the time was an isolated event, or whether he
had to be prepared for periodic water shortages in his
new homeland (Stallings, 1937). He tried to find the
answer to this question by studying the tree rings of trees
on exposed sites in the neighborhood.
In 1869 Pokorny compared the 10-year mean of tree-
ring widths with the corresponding weather data and
found a strong measure of agreement.
In 1892 the Russian, F. N. Shvedov, working in Odessa,
found a clear correlation between annual ring widths of
Robina pseudoacacia and precipitation values. (Shvedov
in Fletcher and Linnard , 1977).
According to Holmsgard (1956), the first annual ring
analytical studies in Denmark were carried out by Revent-
low, 1879, and Lutken, 1891. According to Schulman
(1937), Kaptein (1914) recognized the relationship be-
tween the weather and annual ring width. In the rest of
Scandinavia the first dendroclimatological studies were
conducted by Laitakari (1920), and von Eide (1926).
(Hoeg 1956.)

Tree-ring research develops

Although the principles of cross-dating had been dis- ogy. Only he was far-sighted enough to relate simple
covered in the last century, it is undisputed that the dendrochronological principles to historical research, cli-
American, A. E. Douglass, is the father of dendrochronol- matology and astronomy.

Andrew Ellicott Douglass (1867-1962)

Hardly any other science has been so influenced by one
single person as dendrochronology (Robinson, 1976). In
1901, the young astronomer, then first assistant and later
Director of the Lowell Astronomical Observatory in Flag-
staff, Arizona, was trying to elucidate the relationship
between solar activity and the earth's climate. At that time
there where no long-term meteorological records for the
area, and Douglass hoped to find living records in tree-
trunks. This hope was based on the following reasoning:
tree rings reflect the supply of nutrients at the time of
growth.
food supply is strongly dependent on available mois-
ture.
- therefore tree rings must reflect to some extent the
amount of precipitation.
He tested his hypothesis on 350-300-year-old Pondero-
sa pines (Pinus ponderosa) growing in the neighborhood
and soon established that in dry years narrow rings were
formed in all the trees. Such years are now known as
pointer years. Examining the stump of a tree, he recog-
nized the typical pattern of the last century. It was then
easy for him to calculate the felling date - much to the
astonishment of the owner of the forest!

Andrew E. Douglass (1867-1962)

Gras/und, 1984; Antevs, 1917 257

Douglass' association with the University of Arizona began
in 1906 and lasted for the rest of his life.
The year 1911 brought a further important discovery:
Douglass noticed that trees growing near Prescott, almost
50 miles away, had the same ring pattern as those in the
Flagstaff area. He immediately recognized the significance
of this and set about cross-dating living trees from further
and further afield. By 1914 he had constructed a 500-
year chronology for a large area. Using simple statistical
methods, he established that annual ring formation was
mainly influenced by precipitation in the months preced-
ing the vegetation period; using the long sequence he
continued his search for cyclic patterns. In broad-based
publications he described his investigations on the se-
quoias of the Sierra Nevada and the Pacific coast and
other conifers in the west of the U.S.A. and even in
Europe.
In 1914 Douglass began to date wood from various
historical and archaeological sites. He had been collecting
pieces from different regions since 1911 since he recog-
nized the importance of cross-dating long before others in
the field. In fact it is only in comparatively recent times
that its value has been widely acknowledged.
The story of the dating of the Indian settlements in the
American south-west is a dramatic one. A paper delivered
by Douglass in Washington brought annual ring research
to the attention of the archaeologist, Clark Wissler. As it
was possible to date wood from recent times, he won-
dered whether pieces from archaeological sites could also
be dated. Wissler sent Douglass some beam sections
from the Aztec ruins in New Mexico. Straight away he was The charred log, HH39, from Showlow which, in 1929, provided
able to place them in a 139-year chronology, which did the link between Douglass' floating and absolute chronologies
(Douglass, 1929). (HH is the abbreviation for Hargrave and Haury).
not match the Flagstaff sequence, however. This was the
first 'floating chronology' but by 1920 Douglass had suc-
ceeded in synchronizing it with that of the Pueblo Bonito 'Rosetta Stone' was found - in the form of a charred log
settlement. This was a tremendous breakthrough for - in the settlement of Showlow in Arizona. During that
Douglass was able to tell archaeologists that the people of night Douglass became convinced that the log, labelled
Pueblo Bonito had built their settlement 39 years before HH39, closed the gap between the absolute and the
the Aztecs. The date was exact to the year! relative chronologies, as was indeed the case. Imme-
With the first 'Beam Expedition' of 1923, sponsored by diately 40 settlements, including the famous cliff-dwell-
the National Geographic Society, began a period of inten- ings of Mesa Verde, were absolutely dated (Douglass,
sive dating. In the next five years the absolute chronology 1935). Douglass himself was not a little surprised when it
was extended back to 1260 and a floating chronology of became apparent that actually there had been no gap; the
585 years constructed. In further excavations the archae- two chronologies already overlapped by 25 years, but
ologists L. Hargrave and E. Haury sought for the link there were too few pointer years in the overlap period for
between the two sequences. On June 22nd, 1929 the it to have been recognized as such.

I
1100
I
1200
I , 1300
I I
1400
1
1500
I

700 A.D. _
floating chronology 1285
I
I
absolute chronology (585 years)
1260

, ,
,, HH39 Showlow ,
1237 1380

Bar chart showing how the chronologies overlapped in 1929. The piece from Showlow extended the absolute chronology so far back that
the floating chronology, spanning 585 years, could be absolutely dated.

258
Tree rings of the log, HH39, the "key to the secrets of the south-west". This ring sequence made possible the dating of 40 Indian
settlements from before 1260 A.D. (Douglass, 1929).

Eight years later, in December 1937, the Laboratory of

Tree Ring Research was founded in Tucson, Arizona.
Douglass was appointed director and continued in this
capacity until 1958. The laboratory was housed under the
terraces of the university sports stadium. The following
years were characterized by work on dating. Eventually
the absolute chronology was extended back to the year
322 B.C. and by 1976 the laboratory had collected 20000
beams from 1320 settlements (Robinson, 1976.)

A. E. Douglass (on the right) and E. Schulman at the entrance to the

Laboratory of Tree Ring Research under the terraces of the sports
stadium (Photo B. Huber, 1956).

In 1954 E. Schulmann discovered the Bristlecone

pines, more than 4000 years old, on the alpine timberline
in the White Mountains of California. This was the start of
a research project which is of great scientific importance
today. Using absolutely dated annual ring sequences it is
possible to calibrate methods of radiocarbon dating
(Schulman, 1958.)

E. Schulman discovered the "oldest living things" in the Bristlecone

pine stands of the White Mountains.

259
Dendrochronology in Europe
It is remarkable that the new science was not immediately
adopted by others, in spite of the spectacular results
achieved. Was it because ring sequences of trees in
temperate and boreal zones are less irregular? Was there
too little contact between American and European sci en-
tists, or did archaeologists and biologists working in tem-
perate and boreal zones fail to recognize the potential of
tree ring analysis? Not until the end of the thirties did the
principles gain a foothold in Europe. Bruno Huber was the
man who helped to make the breakthrough.

Bruno Huber (1899-1969)

About 1937 the German botanist, Bruno Huber, realized
the significance of the new discipline, dendrochronology.
A man with wide interests, he took on the task of inves-
tigating the problems identified by Douglass, working in
the temperate zone of central Europe. His life's work is
characterized by breadth of vision.
In 1941 he presented a paper to the German Academy
of Forest Sciences on the construction of a 250-year
sequence and the synchronization of Bronze Age posts.
In subsequent years he systematically extended both
absolute and floating chronologies. In 1963, having ab-
solutely dated a great many objects in southern Germany,
he succeeded in established the temporal relationship
between three lakeside settlements - known as pile
dwellings - in Switzerland. It was not long before he
became interested in tree rings as sowces of climato-
logical information and its regional application. The studies
he initiated on trees growing at high altitudes (Artmann,
1949; Brehme, 1951) and also species in hilly regions
(Muller-Stoll H., 1951) are still of major importance today.
Huber and his students published 56 articles, some of
which are of fundamental importance in European den-
drochronology. For 30 years he was a leading figure in
this field, exerting a considerable influence on research
work.

(
Bruno Huber, about 1965.

Tree-ring research as a recognized science

Andrew E. Douglass and Bruno Huber laid the founda-
tions of modern dendrochronology. Spurred on by their
ideas, new laboratories were opened in various parts of
Europe from 1950 onwards. By 1970 Eckstein listed 20;
today there could be more than 30. In both the east and
west of Russia several laboratories are working on tree
ring analysis (Fletcher and Linnard , 1977).
In North America development was slow. Was no
laboratory other than Tucson interested? Or were too few
willing to compete with the large, successful Tree Ring
Laboratory? The importance of tree ring research for the
life and historical sciences is now recognized in every
continent. In recent times two developments have given
dendrochronological research a great impetus:
In 1963, Harold C. Fritts of Tucson , Arizona, intro-
duced computers as an aid to modern statistical
methods in dendroclimatology. His findings are now
used, in one form or another, by every tree ring
laboratory.
In the same year Hubert Polge of Nancy, France,
invented a method of measuring wood density using
X-rays. It soon became apparent that particularly the
latewood densities of trees in temporate and boreal
zones provide much more climatological information
than the annual ring widths. Only a few dendrochrono-
logical laboratories are equipped with the necessary
apparatus for the application of this method although
more are being set up and others will be established
within the next few years.

260
 The time seems to be ripe for collaboration on dendro-
climatological projects with world-wide application which
are being researched in several different laboratories.
With this in mind a group of experts met in 1981 and drew
up a series of guidelines regarding the major research
objectives for the coming years. (Hughes et at. (Edit.)
1981.)
What will happen now? Will the various branches
diverge? Will those directly interested such as archaeolo-
gists and climatologists take over tree ring research? It
seems likely.
However, within the framework of dendrochronology,
two definite areas seem to be developing strongly:

Harold C. Fritts 1975 Hubert Polge 1981

the use of electronic image analysis, mainly in inves-

tigations on deciduous wood . When such instruments
are used on oaks and other broadleaves, will the kind
of climatic information be unfolded which a whole
generation has been waiting for?
the importance of ecology is growing. The key to
understanding historic and fossil wood lies in recent
tree ring sequences of trees whose ecological back-
ground is known.
In spite of all the specialization it is to be hoped that tree
ring research will remain an interdisciplinary science filling
the role that Weizsacker envisaged when he said : "One
specialized science is not able to provide a complete
global picture which, in the complexity of our modern
existence, would give us something firm to hold on to .
That is why we are looking for a synthesis; we want a
comprehensive view."

261
Bibliography
Alestalo, J., 1971: Dendrochronological interpretation of
geomorphic processes. Societas geographica Fenniae
105, 1-140.
Alestalo, J., and Haikio, J., 1976: Ice features and ice-
thrust shore forms at Luodonselka, Gulf of Bothnia in
Winter 1972/73, Fenniae 144, 5-24.
Alexander, M. E., 1973: Bibliography and resume of
current studies on fire history. Supplements 1980 and
1983. Great Lakes Forest Research Center. Sault St.
Marie, Ontario, Canada. 52 pp.
Anonymous, 1962: Andrew Ellicott Douglass. Obituary
and Bibliography. Tree-Ring Bull. 24, 2-10.
Anonymous, 1979: Die Schweiz und ihre Gletscher; von
der Eiszeit bis zur Gegenwart. Hrgs. Schweiz. Ver-
kehrszentrale. 191 pp., Bern, Kummerly und Frey.
Anonymous, 1980: Mount St. Helens Holocaust. The
Columbian, Inc., P. O. Box 180, Vancouver, Wash. 64
pp.
Anonymous, 1981: Saurer Regen uber Deutschland. Der
Wald stirbt. Der Spiegel 35(45), 96-110.
Antevs, E., 1917: Die Jahresringe der Holzgewachse und
die Bedeutung derselben als klimatischer Indikator.
Progr. Reih. Bot. 5, 285-386.
Arno, S. F., and Sneck, K. M., 1977: A method for
determining fire history in coniferous forests of the
mountain west. USDA Forest Service. Gen. Techn.
Rep.INT-12.
Arnold, J. R., and Libby, W. F., 1949: Age determinations
by radiocarbon content: Checks with samples of known
age. Science 110, 678-680.
Artmann, A., 1949: Jahrringchronologische und -klimatolo-
gische Untersuchungen an der Zirbe und andern
Baumen des Hochgebirges. Dissertation, Univ. Munich.
Ash, S. R., and May, 0. D., 1981.' Petrified forest. The
story behind the scenery. Petrified Forest. Nat. Park,
Arizona, Holbrook. 32 pp.
Assmann, E., 1961: Waldertragskunde. Organische Pro-
duktion, Struktur, Zuwachs und Ertrag von Waldbe-
standen. 490 pp., Munchen, Bayer.
Baas, P., Bauch, J. (ed.) 1986: The effects of environ-
mental pollution on wood structure and quality. IAWA-
Bull. 7, n.s. 267-415. Internat. Assoc. Wood Anato-
mists. Leiden, Netherland.
Bass, P., Bolton, A J., and Catling, D. M., 1976: Wood
structure in biological and technological research. In:
Proc. of the Anglo-Dutch wood anatomy meeting.
Oxford and Kew. 5th-8th April 1976, 280 pp., Leiden,
Univ. Press.
Baes III, C. V., and McLaughlin, S. B. 1984: Trace Ele-
ments in tree Rings: Evidents of recent and historical air
pollution. Science 224,494-497.
Baillie, M. G. L., 1982: Tree-Ring Dating and Archaeol-
ogy. Croom Helm, London and Canberra. 274 pp.
Baillie, M. G. L., 1984: Is there a single British Isles Oak
tree-ring signal? Proc. 22nd Symp. on Archaeometry
(A. Aspinall and S. E. Warren) at 73-82, Univ. of Brad-
ford.
Baillie, M. G. L., and Pilcher, J. R., 1973: A simple cross-
dating program for tree-ring research. Tree-Ring Bull.
33,7-14.
262
Baillie, M. G. L., Pilcher, J. R., and Pearson, G. W., 1983:
Dendrochronology at Belfast as a Background to high
precision calibration. Radiocarbon 25, 171-178.
Balteau, O. J.: Radiographie par rayons X mous. Balteau
company catalog.
Baltensweiler, W., 1976: Die Massenvermehrung des
grauen Larchenwicklers in den Alpen. Int. Ber. Entom-
ologisches Institut, ETH Zurich, 15 pp.
Bannister, B., 1963: Dendrochronology in the Near East:
Current research and future potentialities. Abstract and
report. VII. Congres international des sciences anthro-
pologiques et ethnologiques. V, 336-340.
Bannister, B., 1965: Tree-ring dating of archaeological
sites in the Chaco Canyon region. New Mexico. South-
west Parks and Monuments Association. Technical
Series 6(2).
Bannister, B., and Robinson W. J., 1975: Tree-ring dating
in archaeology. World Archaeology 7,210-225.
Baer, 0., 1977: Geographie Europas. Zurich, Lehrmittel-
verlag des Kantons Zurich. 315 pp.
Bauch, J., Eckstein, D., and Meier-Siem, M., 1972: Dating
the wood panels by a dendrochronological analysis of
tree-rings. Nederl. Kunsthist. Jaarboek 23, 485-496.
Bauch, J., Eckstein, D., and Brauner, G., 1978: Den-
drochronologische Untersuchungen an Eichenholzta-
feln von Rubens-Gemalden, Jahrbuch Berliner Museen
20, 209-221.
Bauch, J., and Eckstein, D., 1981: Woodbiological inves-
tigations on panels of Rembrandt Paintings. Wood Sci.
Technol. 15, 251-263.
Baule, H., and Fricker C., 1967: Die Dungung von Wald-
biiumen. Munchen, Basel, Wien, BLW-Verlag. 259 pp.
Baumann, F., 1949: Die freilebenden Saugetiere in der
Schweiz. Bern, Huber. 492 pp.
Bazzigher, G., 1973: Wundfaule in Fichtenwaldungen mit
alten Schalschaden. Eur. J. For. Phat. 3, 71-82.
Becker, B., and Schirmer, W., 1977: Paleoecological
study on the Holocene valleys development of the river
Main, Southern Germany. Boreas 6,303-321.
Becker, B., 1983: Dendrochronologie und Palaookologie
subfossiler Baumstamme aus Flussablagerungen. Ein
Beitrag zur nacheiszeitlichen Auenentwicklung im sud-
lichen Mitteleuropa. Mitt. Komm. Ouartarforschung der
osterr. Akad. Wiss. 5, 120 pp.
Becker, B., Billamboz, A, Egger, H., Gassmann, P., Orcel
A, Orcel, Chr., and Ruoff, U., 1985: Dendrochronologie
in der Ur- und Fruhgeschichte. Die absolute Datierung
von Pfahlbausiedlungen nordlich der Alpen im Jahrring-
kalender Mitteleuropas. Antiqua 11, Veroff. der schweiz.
Ges. Ur- und Fruhgeschichte. 68 pp.
Beckwith, R. C., 1978: Biology of the insect. pg. 25-36.
In: The Douglas-Fir Tussock-Moth, a Synthesis. Forest
Service, Science and Education Agency, Technical
Bulletin No. 1585. U.S. Dept. of Agriculture. Washing-
ton D.C. 391 p.
Bednarz, Z., 1984: The comparison of dendroclimato-
logical reconstructions of summer temperatures from
the Alps and the Tatra Mountains from 1741-1965.
Dendrochronologia 2, 63-72.
Bircher, W., 1982: Zur Gletscher- und Klimageschichte
des Saastales. Glazialmorphologische und klimatolo-
gische Untersuchung. Physische Geographie, Universi-
tat Zurich. 9, 233 pp.
Blais, J. R., 1964: Account of a recent Spruce Budworm
outbreak in the Laurentide Park Region of Quebec and
measures for reducing damage in future outbreaks.
Forestry Chronicle 40, 313-323.
Blasing, T. J., and Fritts, H. C., 1976: Reconstructing past
climatic anomalies in North America from tree-, ing data.
Quaternary Res. 6, 563-579.
Bolsinger, Ch. L., 1978: The extent of dwarf Mistletoe in
six principal softwoods in California, Oregon and Wash-
ington. Proceedings of the Symposium on Dwarf Mistle-
toe control through forest management. Pacific South-
west Forest and Range Experimental Station. General
Technical Report PSW-31 , 45-54.
Borman, H., and Berlyn, G., eds. 1981: Age and growth
of tropical trees: New directions for research. Yale Uni-
versity: School of Forestry and environmental studies.
New Haven: Yale University. Bull. No. 94.
Bosshard, H. H., 1974-1975: Holzkunde. 3 Vols. Basel
und Stuttgart, Birkhauser.
Braeker, O. U., and Bill, J. (Eds.), 1979: Zum derzeitigen
Stand der Nassholzkonservierung. Zeitschr. schweiz.
Archaologie u. Kunstgesch. 36, 97-145.
Braeker, O. U., 1981: Der Alterstrend bei Jahrringdichten
und Jahrringbreiten von Nadelhblzern und sein Aus-
gleich. Mitt. forst/. Bundesvers anst. Wien. 142, 75-
102.
Brehme, K., 1951: Jahrringchronologische und -klimatol-
ogische Untersuchungen an Hochgebirgslarchen des
Berchtesgadener Landes. Ztschr. Weltforstwirtsch. 14,
64-80.
Brenninkmeijer, C. A. M., 1983: Deuterium Oxygen-18
and carbon-13 in tree rings and peat deposits in relation
to climate. Dissertation, Rijksuniversiteit Groningen,
Holland. 146 pp.
Bucharat, B., 1978: Oxygen isotope palaeotemperatures
from Tertiary period in the North Sea area. Nature 275,
121-123.
Burger, H., 1947: Blattmenge und Zuwachs. VIII. Mittei-
lung. Die Eiche. Mitt. Eidg. Anstalt fur das forstliche
Versuchswesen 25,207-279.
Burk, R. L., and Stuiver, M., 1981: Oxygen isotope rations
in trees reflect mean annual temperature and humidity.
Science 211,1417-1419.
Cain, W. F., 1979: 14C in modern American trees. In: R.
Berger and H. E. Suss (Eds.). Radiocarbon dating. Proc.
9th Int. Cont., Los Angeles and La Jolla, 1976, 495-
510.
Campbell, T. N., 1949: The pioneer tree-ring work of
Jacob Kuechler. Tree-Ring Bull. 15, 16-20.
Carlquist, S., 1980: Further concepts in ecological wood
anatomy, with comments on recent work in wood
anatomy and evolution. Aliso 9, 499-553.
Ceram, C. W., 1972: Der erste Amerikaner. 373 pp.,
Zurich, Ex Libris.
Chaloner,.w. G., and Creber, G. T., 1967: In D. H. Tarling
and S. K. Runcorn (Eds.). Implication of continental drift
for the earth sciences. London, Academic Press.
Creber, G. T., and Chaloner, W. G., 1984: Influence of
environmental factors on the wood structure of living
and fossil trees. Bot. Rev. 50,357-448.
Creber, G. T., and Chaloner, W. G., 1987: The contribu-
tion of growth ring studies to the reconstruction of past
climates. In: R. G. W. Ward, Applications of tree-ring
studies. BAR. Internat. Series 333.37-67.
Christensen, D. J., 1977: Lindbergh Kidnapping - The
ladder link. Forests and People 27, 3-11.
Coles, J. M., 1981: Conservation of wooden artefacts
from the Somerset Levels: 3. Somerset Levels Papers
7,70-80.
Cropper, J. P., 1979: Tree-ring skeleton plotting by
computer. Tree Ring Bull. 39, 47-60.
Cutler, D. F., 1976: Variation in root wood anatomy. In:
Proc. Anglo-Dutch wood anatomy meeting. Oxford and
Kew, 5th-8th April 1976. Leiden, Univ. Press. 143-
176.
Currie, R. G., 1982: Evidence for a 18.6 year MN term in
air pressure in Japan and geophysical implications.
Geophys. J. R. Astron. Soc. 69,321-327.
Dean, J. S., 1969: Chronological analysis of Tsegi phase
sites in northeastern Arizona. Papers Lab. Tree-Ring
Res. 3, 207 p., Tucson, Arizona University Press.
Dean, J. S. and Warren, R. L., 1983: Dendrochronology in
S. H. Lekson (edit): The Architecture and Dendrochro-
·nology of Chetro Ketl, Chaco Canyon, New Mexico.
Nat. Park Service, U.S. Dept. of the Interior. Albu-
querque, New Mexico 1983, 356 pp.
Dean, J. S., in press: Dendrochronology and paleoen-
vironmental reconstruction on the Colorado Plateaus. In
The Anasazi and Their Environment, edited by George
J. Gumerman. School of American Research Press,
Santa Fe.
Delorme, A., 1973: Uber die Reichweite von Jahrring-
chronologien unter besonderer Berucksichtigung mittel-
europaischer Eichenchronologien. Prahist. Zeitschr. 48,
133-143.
Dietrich, J. H., and Swetnam, T. W., 1982: Pyrodendro-
chronology of fire-scarred Ponderosa Pine. 30, 238-
247. Forest Science. 16 p.
Douglass, A. E., 1929: The secret of the Southwest
solved by talkative tree rings. Nat. Geogr. Mag. 54,
737-770.
Douglass, A. E., 1935: Dating Pueblo Bonito and other
ruins of the Southwest. Washington, Nat. Geogr. Soc.,
Pueblo Bonito Ser. 1, 1-74.
Douglass, A. E., 1971: Climatic cycles and tree-growth.
Cramer Lehre (New impression of the work of 1919,
127 pp.; 1928, 166 pp.; 1936, 171 pp.).
Duchaufour, P., 1970: Precis de pedologie. 481 pp.,
Paris, Masson & Cie.
Echols, R. M., 1973: Uniformity of wood density, as-
sessed from x-rays of increment cores. Wood Sci.
Techn. 7, 34-44.
Eckstein, D., 1972: Tree-ring research in Europe. Tree-
Ring Bull. 32, 1-18.
Eckstein, D., and Bauch, J., 1969: Beitrage zur Rationa-
lisierung eines dendrochronologischen Verfahrens zur
Analyse seiner Aussagesicherheit. Forstwiss. Central-
blatt 88, 230-250.
Eckstein, D., Frisse, E., and Liese, W., 1974: Holzanato-
mische Untersuchungen an umweltgeschadigten Stras-
senbaumen der Hamburger Innenstadt. Eur. J. For.
Path. 4, 232-244.
Eckstein, D., Grote, R.-J., and Mathieu, K., 1977: Den-
drochronologische Untersuchungen zur landlichen und
stadtischen Architektur Hamburgs im 15.-18. Jahrhun-
263
 dert. Deutsche Kunst- und Denkmalpflege. Jahrgang
1977,33-74.
Eckstein, D., Breyne, A, Aniol, R. W., and Liese, W.,
1981: Dendroklimatologische Untersuchungen zur
Entwicklung von Strassenbaumen. Forstw. Cbl. 100,
381-396.
Eckstein, D., Neugebauer, M., and Brauner, G., 1982:
Die Baugeschichte der Holzkonstrukturen im Heiligen-
Geist-Hospital zu Luebeck. Lubecker Schriften zur
Archiiologie und Kulturgeschichte. 6, 123-162.
Eckstein, D., 1983: Overview of existing tree-ring chro-
nologies in Europe. In: D. Eckstein, D. Wrobel und R.
W. Aniol (eds.): Dendrochronology and Archaeology in
Europe. Mitt. Bundesforschungsanstalt fUr forst- und
Holzwirtschaft, Hamburg 414, 125-136.
Egger, H., and Lambert, J., (unpublished): Dendrochron-
ologische Bibliographie. Laboratoire de Dendrochronol-
ogie du Musee Cantonale d'Archeologie de Neuchatel.
Eide, E., 1926: Uber Sommertemperatur und Dicken-
wachstum im Fichtenwald. Medd. Norske Skogforsoks-
vesen 2, 87-104 (Norwegian).
Eidem, P., 1955: Dendrochronological dating of a cast-
house from Istad, Slidre, Valdres. Blyttia 13, 65-70
(Norwegian).
Eklund, B., 1956: Variations in the width of the annual
rings in pine and spruce due to climatic conditions in
northern Sweden during the years 1900-1944. Tree-
Ring Bull. 21, 21-25.
Ellenberg, H., 1965: Vegetation Mitteleuropas mit den
Alpen. 843 pp., Stuttgart, Ulmer.
Ellenberg, H., 1967: Vegetations- und bodenkundliche
Methoden der forstlichen Standortkartierung: Ergeb-
nisse eines into Methodenvergleichs im Schweizer Mit-
telland. 296 pp., Zurich, Geobot. Inst. ETH, Stiftung
Rubel.
Ellenberg, H., 1978: Vegetation Mitteleuropas mit den
Alpen in 6kologischer Sicht. 981 pp., Stuttgart, Ulmer.
Esau, K., 1969: Plant anatomy - Pflanzenanatomie. 594
pp., Stuttgart, Fischer.
Everitt, B. L., 1968: Use of the cottonwood in an investi-
gation of the recent history of a flood plain. American
Journal of Science, 1969,417-438.
Evertsen, J. A, 1982: Interlaboratory standardisation sur-
vey. Wood Microdensitometry Bulletin 2, 3-24.
Fahn, A, 1974: Plant Anatomy, 2nd Ed., Reprint 1977.
Oxford etc. Pergamon Press.
Fahn, A, Burley, J., Longman, K. A, Mariaux, A, 1981:
Possible contributions of wood anatomy to the deter-
mination of the age of tropical trees. In: F. H. Borman,
G. Berlyn (eds.) Age and growth rate of tropical trees:
New directions for research. Yale University: School of
forestry and environmental studies. Bull. 94, 31-54.
Ferguson, C. W., 1968: Bristlecone pine: science and
esthetics. Science 159, 839-846.
Ferguson, C. W., and Graybill, D. A, 1983: Dendrochro-
nology of Bristlecone pine: A progress report. Radio-
carbon 25, 287-288.
Firbas, F., 1949-1952: Spfit- und nacheiszeitliche Wald-
geschichte Mitteleuropas nordlich der Alpen. 2 Vols.
Jena, Fischer.
Fletcher, J. M., and Linnard, W., 1977: Russian papers on
dendrochronology and dendroclimatology, 142 pp.,
Oxford, Res. Lab., Arch. and History of Art, University
Press.
264
Fletcher, J., 1978: Tree-ring analysis of panel paintings.
In: Dendrochronology in Europe. BAR Int. Series 51,
303-306.
Ford, E. D., and Robards, A W., 1976: Short term varia-
tion in tracheid development in early wood of Picea
sitchensis. Leiden Botanical Series 3, 212-221 .
Fox, C. A., and Nash III, T. H., 1980: The effect of air
pollution on western larch as detected by tree-ring
analysis. Symposium on Effects of Air Pollutants on
Mediterranean and Temperate Forest Ecosystems.
June 22-27, 1980, Riverside, California U.S.A.
Francuz, J., 1980: A brief explanation to the Twann
Horgen dendro-work. Vortrag an der Jahressitzung der
Arbeitsgemeinschaft fur die Urgeschichtsforschung der
Schweiz. Unpublished, 5 pp.
Frey-Wyssling, A., 1959: Die pflanzliche Zellwand. Berlin,
Gbttingen, Heidelberg, Springer, 367 pp.
Friedman, I., and Gleason, J., 1980: Deuterium content of
lignin and the methyl and hydrogen of cellulose. In: G.
C. Jacoby, (Ed.) Carbon Dioxide Effects: Res. and
Assessment Progr. Proc. Int. Meeting on Stable Iso-
topes in Tree-Ring Res., New York, Lamont-Doherty
Geol. Observ., Columbia University. 50-55.
Fritts, H. C., 1965: Tree-ring evidence for climatic
changes in western North America. Monthly Weather
Rev. 93, 421-443.
Fritts, H. C., Smith, D. G., Cardis, J. w., and Budelsky, C.
A, 1965: Tree-ring characteristics along a vegetation
gradient in Northern Arizona. Ecology 46, 393-401.
Fritts, H. C., 1969: Bristlecone pine in the White Mountain
of California. Tucson, Arizona, Univ. Press., Pap. Lab. of
Tree-ring Res. 4, 44 p.
Fritts, H. C., 1974: Relationships of ring width in arid-sites
conifers to variations in monthly temperature and pre-
cipitation. Ecological Monographs 44, 411-440.
Fritts, H. C., 1976: Tree rings and climate, London, New
York, San Francisco, Academic Press, 567 pp.
Fritts, H. C., 1977: Tree rings: A record of past climate.
US Dept. of Commerce, EDS, 31-42.
Fritts, H. C., Lofgren, G. R., Gordon, G. A, 1979: Varia-
tions in climate since 1602 as reconstructed from tree
rings. Quaternary Res. 12,18-46.
Fritts, H. C., 1983: Tree-ring dating and reconstructed
variations in Central Plains climate. Transactions of the
Nebraska Academy of Sciences XI (special Issue), 37-
41.
Fritts, H. C., and Lough, J. M., 1984: An estimate of aver-
age annual temperature variations for North America
1602-1960.
Fritts, H. C., and Swetnam, T. W., 1986: Dendroecology:
A tool for evaluating variations in the past and present
forest environments. Lab. of Tree-Ring Research, Univ.
of Arizona 85721, Tucson, Arizona, 61 pp.
Fritts, H. C., and Wu Xiangding, 1986: A comparison
between response - function analysis and other re-
greSSion techniques. Tree-ring Bull. 46, 31-46.
Furrer, G., Gamper-Schollenberger, B., and Suter, J.,
1980: Zur Geschichte unserer Gletscher in der
Nacheiszeit - Methoden und Ergebnisse. In: H.
Oeschger, B. Messerli, M. Svilar (Eds.) Das Klima.
Analysen und Modelle. Geschichte und Zukunft. Berlin,
Heidelberg, New York, Springer. 296 pp.
Gassner, G. and Christiansen-Weniger, E., 1942: Den-
droklimatologische Untersuchungen uber die Jahresrin-
gentwicklung der Kiefern in Anatolien. Nova Acta Leo-
poldina 12, Halle. 1-137.
Geiger, R., 1961: Das Klima der bodennahen Luftschicht;
ein Lehrbuch der Mikroklimatologie. Braunschweig. 4.
Aufl. 646 pp.
Gothan, W., and Weyland, H., 1973: Lehrbuch der Palao-
botanik. Munich etc. BLV Verlagsgesellschaft. 677 pp.
Gottwald, H., 1958: Handelshblzer, Hamburg.
Graslund, M., 1984: The history of Dendrochronology in
the Nordic Countries. Dendrochronologia 2, 31-62.
Gray, J., and Thompson, P., 1980: Natural variations in
the 18 0 content of cellulose. In: G. C. Jacoby, (Ed.)
Carbon Dioxide Effects; Res. and Assessment Pro-
gram, Proc. Int. Meeting on Stable Isotopes in Tree-
ring Res., New York, Lamont-Doherty Geol. Observ.,
Columbia Univ., 84-92.
Graybill, D. A, 1982: Chronology development and analy-
sis. In: Hughes, M. K., Kelly, P. M., Pilcher J. R., and
LaMarche V. C. (eds.). Climate from tree rings. Cam-
bridge, Univ. Press London, 21-31.
Green, H. V., 1965: The study of wood characteristics by
means of a photometric technique. In: Proc. meeting of
working groups on wood quality, sawing and mechin-
ing, wood and tree chemistry. Ed. Section 41, IUFRO,
Melbourne, 4th-15th Oct. 1965, vol. 2, 12 p., Mel-
bourne: CSIRO, Div. forest products.
Greve, U., 1984: Holzbiologische Untersuchungen an
Fichten (Picea abies (L.) Karst.) unterschiedlicher Im-
missionstoleranz und Immissionsbelastung. Disserta-
tion, Universitat Hamburg, Fachbereich Biologie. 137
pp.
Griggs, F. F., 1928: Das Tal der zehntausend Dampfe.
Leipzig. Brockhaus.
Grootes, M., 1977: Thermal diffusion isotopic enrichment
and radiocarbon dating beyond 50000 years B.P.,
Groningen. 221 pp.
Grosser, D., 1977: Die HOlzer Mitteleuropas. Ein mikro-
photographischer Lehratlas. Berlin, Heidelberg, New
York, Springer. 208 pp.
Grunhilder, I., 1972: Auf der Spur vorgeschichtlicher
Katastrophen. Griin, Das Gartenmagazin E. 5948. 24-
30.
Grunig, P. E., 1953: Die Flyschaufforstung in wirtschaft-
licher, naturwissenschaftlicher und waldbaulicher Betra-
chtung. Schweiz. Ztschr. Forstwesen 6, 1-16.
Grunig, P. E., 1955: Ueber den Einfluss der Entwasser-
ung auf die Flachmoorvegetation und auf den Zuwachs
der Fichte und Bergfbhre im Flyschgebiet der Voralpen.
Mitt. Schweiz. Anstalt forstl. Versuchswesen. 31, 419-
492.
Haesler, R., 1982: Net photosynthesis and transpiration of
Pinus montana on east and north facing slops at alpine
timberline. Oecologia (Berf) 54,14-22.
Harlow, W. M., Harrar, E. S., 1950: Textbook of dendro-
logy. London, New York, Toronto, McGraw-HilI. 555 p.
Harms Geographie 1979: Physische Geographie und
Nachbarwissenschaften, Munchen, 424 pp.
Hawksworth, F. G., Wiens, D., 1972: Biology and classifi-
cation of Dwarf Mistletoe (Arceuthobium). Agriculture
Handbook No. 401. Forest Service, US Dept. of Agri-
culture.
Hawksworth, F. G., Hinds, T. E., 1964: Effects of Dwarf
Mistletoe on immature Lodgepole pine stands in Colo-
rado. J. Forestry 62, 27-32.
Haselhoff, E., Lindau, G., 1903: Die Besch8digung der
Vegetation durch Rauch. Berlin, Leipzing, Borntrager.
412 pp.
Hawksworth, F. G., 1961: Dwarf Mistletoe of Ponderosa
Pine in the Southwest. Tech. Bull. no. 1246, US Dept.
of Agriculture 111 pp.
Hawksworth, F. G., and Hinds, T. E., 1964: Effects of
Dwarf Mistletoe on immature Lodgepole pine stands in
Colorado. J. Forestry 62, 27-32.
Hawksworth, F. G., and Wiens, D., 1972: Biology and
classification of Dwarf Mistletoe (Arceuthobium). Agri-
culture Handbook no. 401, Forest Service, US Dept. of
Agriculture.
Hecht, A D., (Ed.) 1979: Paleoclimatic research: Status
and opportunities. Quaternary Res. 12,6-17.
Heer, 0., 1865: Die Urwelt der Schweiz. Schulthess,
.Zurich, 713 pp.
Hegi, G., 1957: ///ustrierte Flora von Mittel-Europa s.
Band 3, Teil 1, Aufl. 2, 452 pp. Munchen, Hanser.
Heikkinen, 0., 1984: Dendrochronological evidence of
variations of Coleman Glacier, Mount Baker, Washing-
ton, USA Arctic and Alpine Research 16, 53-64.
Helley, E. J., and LaMarche, V. C., 1973: Historic flood
Information for Northern California streams from geo-
logical and botanical evidence. Geol. Survey prof
paper 485 - E. 15 p.
Hess, H. E., Landolt, E., and Hirzel, R., 1967-1972: Flora
der Schweiz und angrenzender Gebiete. 3 Bande.
Basel, Stuttgart, Birkhauser.
Heyer, E., 1977: Witterung und Klima, 4. Aufl., Leipzig,
Teubner, 460 pp.
Heyworth, A, 1978: Submerged forests around the
British Isles: Their dating and relevance as indicators of
postglacial land and sea level changes. In: J. Fletcher,
Dendrochronology in Europe. BAR Int. Series 51, 279-
288.
Hickin, N. E., 1973: The insect factor in wood decay; an
account of wood boring insects with particular reference
to timber indoors. 3 Edition. London, Robin Edwards.
Higashi, S., Fujiwara, K., Araya, T., and Murai, N., 1971:
Dendrochronological studies on the transition of the
creeping land. Res. Bull. Coli. Experiment Forest,
College of Agr. Hokkaido University 28, 339-420.
Hirmer, M., 1927: Handbuch der Palaobotanik. Munchen,
Berlin, R. Oldenburg. 708 pp.
Hoeg, A 0., 1956: Growth-ring research in Norway.
Tree-Ring BUll. 21, 2-15.
Holdheide, W., Huber, B., 1951: Aehnlichkeiten und
Unterschiede im Feinbau von Holz und Rinde. Holz als
Roh- und Werkstoff 10, 262-268.
Hollstein, E., 1978: Wood technology and the dating of
oak. West German chronologies for oak and beech. In:
J. Fletcher, Dendrochronology in Europe. BAR Int.
Series 51, 33-44.
Hollstein, E., 1980: Mitteleuropaische Eichenchronologie.
Trierer dendrochronologische Forschungen zur Archa-
ologie und Kunstgeschichte. Mainz a. Rh., Trierer Gra-
bungen und Forschungen 11, 273 pp.
Holmes, R. L., Stockton, D. W., and LaMarche, V. C.,
1979: Extension of river flow records in Argentina from
living tree-ring chronologies. Water Resources Bull. 15,
1081-1085.
265
Holmsgard, E., 1956: Tree-ring analysis of Danish forest
trees. Tree-Ring Bull. 21, 25-27.
Holtmeier, F. K., 1971: Der Einfluss der orographischen
Situation auf die Windverhaltnisse im Spiegel der Vege-
tation. Erdkunde 25, 178-195.
Holzhauser, H. P., 1984: Zur Geschichte der Aletsch-
gletscher und des Fieschergletschers. Physische
Geographie 13, Universitat Zurich, 1-488.
Hornbeck, J. W., and Smith, R. B., 1985: Documentation
of red spruce growth decline. Can. J. For. Res. 15,
1199-1201.
Hoester, H. R., 1977: Veranderungen der Holzstruktur als
Indikator fur Umweltbelastungen bei Baumen. Ber.
deutsch. bot. Ges. 90, 253-260.
Hrib, J., Kyncl, J., and Cerny, A., 1983: A tree-ring study
of Norway spruce infected with the wood-destroying
fungus Armillaria mellea. European Journal of Forest
Pathology 13, 160-165.
Huber, B., 1941: Aufbau einer mitteleuropaischen Jahr-
ring-Chronologie. Mitt. H. G. Akad.deutsch. Forstwiss.
3,137-142.
Huber, B., 1961: Grundzuge der Pflanzenanatomie. Ver-
such einer zeitgemassen Darstellung. Berlin, Gottingen,
Heidelberg, Springer. 243 pp.
Huber, B., and Merz, W., 1963: Jahrringchronologische
Synchronisierung der jungsteinzeitlichen Siedlungen
Thayngen-Weier und Burgaschisee-Sud und Sudwest.
Germania 41, 1-9.
Huber, B., and Jazewitsch von, W., 1965: Tree-ring
studies of the Forestry-Botany Institutes of Tharandt
and Munich, Tree-Ring Bull. 21, 28-30.
Huber, B., 1970: Lichtmikroskopische Untersuchungen
on Holzern, besonders die Bestimmung ihrer systema-
tischen Zugehorigkeit. In: Handbuch der Mikroskopie in
der Technik, Band V. Mikroskopie des Holzes und des
Papiers. Frankfurt AM., Umschau-Verlag 37-103 und
171-211.
Hughes, J. F., 1968: Density as an index of wood quality
with special reference to the development of rapid and
efficient methods of estimating density. Journ. Oxford
Univ. For. Soc. (Ser. 6), 5-6.
Hughes, M. K., Kelly, P. M., Pilcher, J., and LaMarche Jr.,
V. C., 1980: Report and recommendations. Second
international workshop on global dendroclimatology
Belfast, July 1980, Norwich, Edmund Norvic Press,
ISBN 0-950-7344-0-3.
Hughes, M. K., Milsom, S. J., and Leggett, P. A, 1981:
Sapwood estimates in the interpretation of tree-ring
dates. Journal of Archaeological Science 8, 381-390.
Hughes, M. K., Schweingruber, F. H., Cartwright, D., and
Kelley, O. M., 1984: July-August temperature at Edin-
burgh between 1721 and 1975 from tree-ring density
and width data. Nature 308, 341-344.
Jacoby, G. C., and Ulan, L. D., 1983: Tree ring indications
of uplift at Icy Cape, Alaska related to 1899 earth-
quakes. Journal of Geophysical Research 88, 9305-
9313.
Jahrig, M., and Baumgarten, K., 1975: Dendrochronol-
ogische Untersuchungen an Bauernhausern des meck-
lenburgischen Sudwestens. Jahrbuch der bayerischen
Denkmalpflege 29,35-78.
Jesse, A, 1974: Bibliography on automatic image analy-
sis. The Microscope 22, 89-115.
266
Johnson, A H., Siccama, T. G., 1983: Acid deposition
and forest decline. Environmental Science and Technol-
ogy 17, 294-305A
Jones, P. D., Briffa, K. R., and Pilcher, J. R., 1984: River-
flow reconstruction from tree rings in Southern Britain.
Journal of Climatology 4, 461-472.
Jong de, A. F. M., Mook, W. G., 1980: Medium-term
atmospheric 14C variations. Radiocarbon 22, 267-272.
Joos, K., 1985: Year ring analysis - a new method of
assessing the condition of trees in cities. Phoenix
International 5, 14-19.
Kaelin, I., 1983: Versteinerte Holzer in vulkanischen
Tuffen der Insel Lipari. Mikrokosmos 1983, 65-70,
111-113, 140-148.
Kaiser, N. F. J., 1979: Ein spateiszeitlicher Wald im
Dattnau bei Winterthur, Schweiz. Winterthur, Ziegler. 90
pp.
Kaiser, F., and Kaiser, C., 1987: The Katmai eruption von
1912 and the Alaska earthquake of 1964 as reflected in
the annual rings of sitka spruce (Picea sitchensis) on
Kodiak island. Dendrochronologia 5, in press.
Kaptein, J. C., 1914: Tree growth and meteorological
factors. Receuil des travaux botaniques neerlandais 11,
70-93.
Kasser, P., 1954: Sur Ie bilan hydrologique des bassins
glaciaires avec application au Grand Glacier d'Aletsch.
Publ. no. 39 de I'assoc. intern. d'hydrologie (Assem-
blee generale de Rome, tome 4) 331-350.
Keller, R., 1968: Des characteristiques nouvelles pour
I'etude des proprietes mechaniques des bois: les com-
posantes de la densite. Ann. Sci. For. 25,237-249.
Kennedy-Sutherland, E., 1983: The effects of fire exclu-
sion on growth in mature Ponderosa Pine in Northern
Arizona. Unpublished, University of Arizona. 20 pp.
Kern, K. G., 1960: Der jahreszeitliche Ablauf des Dicken-
wachstums von Fichten verschiedener Standorte im
Trockenjahr 1959. Allgemeine Forst- und Jagdzeitung
131, 97-116.
Kienast, F., Fluhler, H., and Schweingruber, F. H., 1981:
Jahrring-analysen an Fohren (Pinus silvestris L.) an
immissionsgefahrdeten Waldbestanden des unteren
Wallis (Saxon, Schweiz). Mitt. schweiz. Anst. forstl.
Versuchswes. 57,415-432.
Kienast, F., 1985: Dendrookologische Untersuchungen an
Hohenprofilen aus verschiedenen Klimabereichen. Eine
radiodensitometrische Studie uber den Einfluss der
Witterung, der Hohenlage und den Standortseigens-
chaften auf das Jahrringwaschstum von Nadelbaumen.
Diss. Univ. Zurich, 129 pp.
Kienast, F., and Schweingruber, F. H., 1986: Dendroeco-
logical studies in the Front Range, Colorado, U.S.A.
Arctic and Alpine Research 18, 277-288.
Kittredge, J., 1948: Forest influences; the effects of
woody vegetation on climate, water and soil, with
applications to the conservation of water and the control
of floods and erosion. New York, Mc Graw Hill, 394 pp.
Klein, J., Lerman, J. C., Damon, P. E. and Ralph, E. K.,
1982: Calibration of radiocarbon dates. Radiocarbon 24,
103-105.
Klein, P., 1979: Alte Gemalde auf Holz gemalt. Holz-ZbI.
105, 2287/88.
Klein, P., 1980: Dendrochronologische Untersuchungen
an Eichenholztafeln von Rogier Van der Weyden.
Jahrbuch der Berliner Museen 23, 113-123.
Klein, P., 1983: Dating of art historical objects. In: D.
Eckstein, S. Wrobel, R. W. Aniol (eds.) Dendrochronol-
ogy and archaeology in Europe. Mitt. Bundesforschung-
sanstalt fur Forst- und Holzwirtschaft. Hamburg 141,
209-222.
Klein, P., 1984: Tree-Ring chronology of spruce wood
and its application in the dating of stringed instruments.
ICOM Committee for Conservation 7th Triennial Meet-
ing, 1984 Kopenhagen. 84.1, 69-72.
Koenig, E. 1957: Tierische und pflanzliche Holzschiid-
tinge. Stuttgart, Holz-Zentralblattverlag, 330 pp.
Koenigswald von. W., and Taute, W., 1979: Zwei bedeu-
tende Quartar-profile in der Burghohle von Dietfurt bei
Sigmaringen an der Donau. N. Jb. Geol. Pal§ont. Mh. 4,
216-236.
Kontic, R., Niederer, M., Nippel, C. -A., and Winkler-
Seifert, A., 1986: Jahrringanalysen an Nadelbaumen
zur Darstellung und Interpretation von Waldschaden
(Wallis, Schweiz), Bericht 283, Eidg. Anstalt fUr das
forstliche Versuchswesen. CH-8903 Birmensdorf. 46
pp.
Kozlowski, T. T., 1971: Growth and development of trees,
2 Bande, New York, London, Academic Press.
Kuechler, A. W., 1966: Vegetation maps of Europe.
Lawrence Univ. of Kansas libraries, 584 pp.
Kulmann, H. M., 1971: Effects of insect defoliation on
growth and mortality of trees. Ann. Rev. Entomology
16, 289-324.
Kuniholm, P. I., and Striker, C. L., 1983: Dendrochrono-
logical investigations in the Aegean and neighboring
regions, 1977-1982. Journal of Field Archaeology 10,
411-420.
Kuniholm, P. I., and Striker, C. L., 1987: Dendrochrono-
logical investigations in the Aegean and neighboring
regions, 1983-1986. J. Field Arch. 14, in press.
Laitakari, E., 1920: Untersuchungen uber die Einwirkung
der Witterungsverhaltnisse auf das Langen- und Dick-
enwachstum der Kiefer. Acta Forest. Fenn 17, 53 pp.
(Finnisch).
LaMarche, V. C. Jr., and Fritts, H. C., 1972: Tree-ring and
sunspot numbers, Tree-Ring. Bull. 32, 19-32.
LaMarche, V. C. Jr., and Wallace, R. E., 1972: Evaluation
of effects of past movements on the San Andreas Fault,
Northern California. Geological Soc. of America, Bull.
83, 2665-2676.
LaMarche, V. C. Jr., and Stockton, C. W., 1974: Chro-
nologies from temperature-sensitive Bristlecone pines
at upper treeline in western United States. Tree-Ring
Bulletin 34, 21-45.
LaMarche, V. C. Jr., 1974: Paleoclimatic inferences from
long tree-ring records. Science 183, 1043-1048.
LaMarche, V. C. Jr., and Hirschboeck, K. K., 1984: Frost
rings in trees as records of major volcanic eruptions.
Nature 307, 121-145.
Lamb, H. ·H., 1972-1977: Climate; present, past and
future, 2 Vols. London: Methuen. New York: Barnes
und Noble.
Larcher, W., 1973: Oekologie der Pflanzen. Stuttgart,
Ulmer. 320 pp.
Larsen, J. B., 1981: Waldbauliche und ertragskundliche
Erfahrungen mit verschiedenen Provenienzen der
Weisstanne (Abies alba Mill.) in Danemark. Forstw. Cbl.
100,275-286.
Lawrence, J. R., and White, J. W., 1984: Growing season
precipitation from D/H ratio of Eastern White Pine.
Nature 311, 558-550.
Leavitt, S. W., and Long, A., 1984: Sampling strategy for
stable carbon isotope analysis for tree rings in pine.
Nature 311,145-147.
Leibundgut, H., Auer, C., and Wieland, C., 1971: Ergeb-
nisse von Durchforstungsversuchen 1930-1965 im
Sihlwald. Mitt. Eidg. Anst. forst/. Versuchswes. 47,
259-389.
Lenz, 0., 1957: Utilisation de la radiographie pour I'ex-
amen des couches d'accroissement. Mitt. Eidg. Anstalt
fUr das forstl. Versuchswes. 33, 125-134.
Lenz, 0.,1967: Action de la neige et du gel sur les arbres
de montagne, en particulier sur leur forme et I'anatomie
de la tige. Mitt. Eidg. Anstalt fUr das forstl. Versuch-
swesen 43, 288-316.
Lenz, 0., Oswald, K., 1971: Ueber Schaden durch Bohr-
spanentnahme an Fichte, Tanne und Buche. Mitt. Eidg.
. Anst. forst!. Versuchswesen 47 (1) 1-29.
Lenz, 0., Schar, E., and Schweingruber, F. H., 1976:
Methodische Probleme bei der radiographisch-densito-
metrischen Bestimmung der Dichte und der Jahrring-
breiten von Holz. Holzforschung 30, 114-123.
Leptere, R., 1957: Influence of high pruning on the
growth of Douglas fir. Forestry 30.
Leuschner, H.-H., Delorme, A., Tuxen, J., and Hofle,
H.-Chr., 1985: Eichenstammlagen in Mooren und
Talauen und die Klimaverschlechterung im Subboreal.
Flora 177, 283-':"295.
Leuschner, H. H., and Delorme, A.: Dendrochronolo-
gische Befunde zu Torfeichen aus dem Kehdinger
Moor bei Hammah, Kreis Stade. Unveroffentlichtes
Manuskript.
Libby, W. F., Anderson, E. C., and Arnold, J. R., 1949:
Age determination by radiocarbon content. World-wide
assay of natural radiocarbon. Science 109, 227-228.
Libby, W. F., 1963: Accuracy of radiocarbon dates.
Science 140, 278-280.
Libby, W. F., 1960: Radiokohlenstoff-Datierung. Nobel-
Vortrag vom 12. Dez. 1960. Angewandte Chemie 73,
225-229.
Liese, W., and Eckstein, D., 1971: Die Jahrringchronol-
ogie in der Kriminalistik. In: H. Schafer, (Ed.) Grund-
lagen der Kriminalistik. Hamburg, Steintor, 395-422.
Liese, W. (ed.), 1975: Biological transformation of wood
by microorganisms. Proc. of the Sessions on wood
products path., 2nd Int. Congr. Plant path., Sept. 10-
12, 1973. Minneapolis. Berlin, Springer, 203 p.
Liese, W., Schneider, M., and Eckstein, D., 1975: His-
tometrische Untersuchungen am Holz einer rauchges-
chadigten Fichte. European J. Forest Pathology 5,
152-161.
Liese, W., 1978: Bruno Huber: The poineer of European
dendrochronology. BAR Int. Series 51, 1-10.
Lightle, P. C., and Hawksworth, E. C., 1973: Control of
Dwarf Mistletoe in a heavily used Ponderosa Pine
recreation forest: Grand Canyon, Arizona. USDA Forest
Res. Paper RM-1 06. Rocky Mountain Forest and Range
Experimental Station. US Dept. of Agriculture. 22 p.
Lingg, W., 1986: Dendrookologische Studie an Fichte
(Picea abies) und Weisstanne (Abies alba) im subkon-
tinentalen Klimagebiet Wallis, Schweiz. Ber. Eidg.
Anstalt fur das forstliche Versuchswesn 287, 1-81.
267
Liphschitz, N., 1986: Overview of the dendrochroarchae-
logical research in Israel. Dendrochronologia 4,37-58.
Long, A, 1982: Stable isotopes in tree rings. In Hughes,
M. K., Kelly, P. M., Pilcher, J. R., and Lamarche, V. C.:
Climate from tree rings. Cambridge, London, New York,
New Rochelle, Melbourne, Sydney. 13-21.
LUtken, C., 1891: Nogletagtagelser over de meteoriske
Forhold Indflydelse pa Traernes Tilvaxtgang. Tidsskr.
Skovbrug 12.
Madany, M. H., Swetman, T. W., and West, N. E., 1982:
Comparison of two approaches of determining fire
dates from tree scars. Forest Science 28, 856-861 .
Markus, R., 1936: Einfluss der Entwasserung auf den
Zuwachs der Kiefer und Fichte auf MoorbOden in Lett-
land. Mitt. forestliche Versuchsanstalt Lettlands 5
(Latvian with German Summary).
Mason, R. R., 1978: Population Ecology. Pg. 39-61. In:
The Douglas-Fir Tussock-Moth, a Synthesis. Forest
Service, Science and Education Agency, Technical
Bulletin No. 1585, US Dept. of Agriculture. Washington
D.C.,391 p.
Mathieu, H., 1961: La conservation du materiau bois.
Paris, Bailliere, 318 pg.
McGregor, J. C., 1936: The effect of a volcanic cinder fall
on tree growth. Tree-Ring Bull. 3, 11-13.
McLaughlin, S. B., West, D. C., and Blasing, T. J., 1983:
Measuring effects of air pollution stress on forest pro-
ductivity: Some perspectives, problems, and ap-
proaches. Proceedings TAPPI, Annual meeting, 321-
333.
McLaughlin, S. B., Blasing, T. J., Mann, L. K., and Duvick,
D. N., 1983: Effects of acid rain and gaseous pollutants
on forest productivity: A regional scale approach. Jour-
nal of the Air Pollution Control Association 33, 1042-
1048.
MacLean, D. A, 1981: Impact of defoliation by Spruce
budworm populations on radial and volume growth of
Balsam fir: A review of present knowledge. Mitt. der
forstlichen Bundesversuchsanstalt, Wien 142, 293-
306.
Medway, L., 1972: Phenology of a tropical rainforest in
Malaya. BioI. J. Linn. Soc. 4, 117-146.
Meisling, K. E., and Sieh, K., 1980: Disturbance of trees
by the 1857 Fort Tejon earthquake, California. Journal
of Geophysical Research 83, 3225-3238.
Meusel, H., 1965-1978: Vergleichende Chronologie der
zentraleuropaischen Flora. 4 Vols. Jena Fischer.
Mikola, P., 1956: Tree-ring research in Finland. Tree-Ring
BUll. 21, 16-20.
Miller, D. J., 1960: Giant waves in Lithuya Bay, Alaska.
US Geol. Survey Prof. Paper 354 - C. 51-86.
Mitchell, J. M., Stockton, C. W., and Meko, D. M., 1979:
Evidence of a 22-year rhythm of drought in the western
United States related to the hale solar cycle since the
17th century. In: Solar-terrestrial influences in climate.
B. M. McCormac and T. A Seliga (Eds.). Reidel Publ.
Co. Holland. pp. 125-143.
Mitchell, J. M., 1980: History and mechanisms of climate.
In: H. Oeschger, B. Messerli and M. Svilar: Das Klima.
Analysen und Modelle, Geschichte und Zukunft. Berlin,
Heidelberg, New York, Springer. 296 pp.
Mitscherlich, G., 1970-1975: Wachstum und Umwelt;
eine Einfuhrung in die 6kologischen Grundlagen des
268
Waldwachstums. 3 Vols. Frankfurt am Main, Sauerlan-
der.
Morgan, R. A, Hillam, J., Coles, J. M., and McGrail, S.,
1981: Reconciling tree-ring sampling with conservation.
Antiquity 55, 90-95.
Mueckenhausen, E., 1975: Die Bodenkunde und ihre
geologischen, geomorphologischen, mineralogischen
und petrologischen Grundlagen: Frankfurt am Main,
DLG-Verlag. 579 pp.
Mueller, H. N., 1981: Jahrringwachstum und Klimafak~
toren. Angewandte Pflanzensoziologie. Ver6ff. Forstl.
Bundesversuchsanstalt Wien 25, 81 pp.
Mueller-Stoll, H., 1951: Vergleichende Untersuchungen
Uber die Abhangigkeit der Jahrringfolge von Holzart,
Standort und Klima. Bibl. Bot. 122, 1-93.
Mueller-Stoll, W., 1947: Photometrische Holzstruktur-
Untersuchungen. I. Mitt.: Ueber die Ermittlung von
Jahrringaufbau und Spatholzanteil auf photometrischem
Wege. Planta 35, 397-426. II. Mitt.: Ueber die Bezie-
hungen der Lichtdurchlassigkeit von Holzschnitten zur
Rohwichte und Wichtekontrast. Forstw. Cbl. 68, 21-63.
Munaut, A, 1967: La foret Ensevelie de Ternuzen. Revue
Industrie. 9 pp.
Munaut, A, and Casperie, W. A, 1971: Etude dendro-
chronologique de Pinus silvestris L. subfossiles pro-
venants de la tourbiere d'Emmen (Drenthe, Pays-Bas).
Rev. Palaeobot., Palynology 11, 201-226.
MUnch, D., and Lambert, D., 1972: Tiberline ancients.
2000 N.w. Wilson, Oregon, Belding. 128 pp.
Nageli, W., 1935: Aussetzende und auskeilende Jahr-
ringe. Schweiz. Z. Forstwesen 1935.1-7.
Nogler, P., 1981: Auskeilende und fehlende Jahrringe in
absterbenden Tannen (Abies alba Mill.). Allg. Forstz.
28, 709-711.
Nultsch, W., 1977: Allgemeine Botanik. Kurzes Lehrbuch
tilr Mediziner und Naturwissenschaftler. 6. Aufl. Stutt-
gart, Georg Thieme. 438 pp.
Oeschger, H., and Rothlisberger, H., 1961: Datierung
eines ehemaligen Standes des Aletschgletschers durch
Radioaktivitatsmessung an Holzproben und Bemer-
kungen zu Holzfunden an weiteren Gletschern. Ztschr.
Gletscherkunde und Glaziologie 4, 191-205.
Oeschger, H., 1985: Langfristige Klimastabilitat. Neue
Zurcher Zeitung vom 27. Februar 1985. P. 65.
Page, R., 1970: Dating episodes of faulting from tree
rings: Effect of the 1958 rupture of Fairweather Fault on
tree growth. Geological Soc. of America Bulletin 81,
3085-3094.
Palenik, S., 1983: Microscopic trace evidence. The over-
looked clue. Part IV Arthur Koehler-Wood Detective.
The Microscope 31, 1-13.
Panshin, A J., and Zeeuw, C., 1970: Textbook of wood
technology, 3rd ed., 705 p., New York, McGraw-HilI.
Parker, M. L., 1971: Dendrochronological techniques
used by the geological survey of Canada. In: Paper
71-25, 1-30. Ed. Geological Survey of Canada.
Parker, M. L., and Jozsa, L. A., 1971: Dendrochrono-
logical investigation along the Mackenzie, Liard and
south Nahanni Rivers, N.w.T. Technical Report 10 to
Glaciology Division. Water Resources Branch. Dept. of
the Environment. 187 pp.
Parker, M. L., and Jozsa, L. A., 1973: X-ray scanning
machine for tree ring width and density analysis. Wood
and Fibre 5,192-197.
Patzelt, G., and Bortenschlager, S., 1973: Die post-
glazialen Gletscher- und Klimaschwankungen in der
Venedigergruppe (Hohe Tauern, Ostalpen). Ztschr.
Gletscherkunde und Glaziologie 9,5-57.
Pearson, G. W., and Baillie, M. G. L., 1983: High-preci-
sion 14C measurement of Irish oaks to show the natural
atmospheric 14C variations of the AD period. Radio-
carbon 25,187-196.
Pearson, G. W., Pilcher, J. R., and Baillie, M. G. L., 1983:
High precision 14C variations from 200 B.C. to 4000
B. C. Radiocarbon 25, 179-186.
Pechmann von, H., 1960: Haben Mineraldungung und
Lupinenanbau einen Einfluss auf die Eigenschaften von
Fichten- und Kiefernholz. Forstw. Cbl. 79, 91-105.
Pechmann von, H., 1974: Der Einfluss der Durchforstung
auf die Holzqualitat. Forstarchiv 45, 34-38.
Petersen, A., and Eckstein, D., 1984: Holzbiologische
Untersuchungen an Strassenbaumen. Wiss. Ber. Univ.
Hamburg. 18,37-40.
Pfister, Ch., 1984: Klimageschichte der Schweiz, 1525-
1860. Academica helvetica 6. Paul Haupt, Bern und
Stuttgart.
Pfister, Ch., 1986: Veranderungen der Sommerwitterung
im sudlichen Mitteleuropa als Auftakt zum Gletscher-
hochstand der Neuzeit. Geographica helvetica 1985,
186-195.
Philipps, R., 1978: Trees in Britain, Europe and North
America. London, Pan. 224 pp.
Phillips E. W. J., 1966: The use of beta-particle radiation
methods in timber research. IAWA Bull. 2, 17-28.
Pilcher, J. R., Baillie, M. G. L., Schmidt, B., and Becker,
B., 1984: A 7272-year tree-ring chronology for western
Europe. Nature 312, 150-152.
Plumtree R. A., 1979: Some techniques used in the study
of wood density. Unpublished report 6 pp.
Plumtree R. A., 1984: Pinus caribaea Vol. 2, Wood prop-
erties. Tropical Forestry paper 17, 1-148.
Pokorny, A., 1892: Methode um den meteorologischen
Coeffizienten des jahrlichen Holzzuwachses der Di-
cotyledonenstamme zu ermitteln. Tharandter forstl. Jb.
22, 81 ff.
Polge, H., 1966: Etablissement des courbes de variation
de la densite du bois par I'exploration densitometrique
a
de radiographies d'echantillons preleves la tariere sur
des arb res vivants. Ann. Sci. Forest. 23, 1-206.
Polge, H., and Thiercelin, F., 1970: Degats occasionnes
par les sondages de tariere. Techn. et For. 6, 629-
636.
Polge, H., Keller, R., and Thiercelin, F., 1973: Influence
de I'etage de branches vivantes sur la structure des
accroissements annuels et sur quelques caracteris-
tiques du bois de Douglas et de Grandis. Ann. Sci.
Forest. 30,127-140.
Polge, H., Lemoine, M., and Deret, E., 1977: Etude de la
variabilite specifique et infraspecifique de la structure
juvenile du bois de chene a I'aide d'un analyseur
d'image. Ann. Si. Forest. 34(4), 285-293.
Pollanschutz, J., 1971: Die ertragskundlichen Methoden
zur Erkennung und Beurteilung von forstlichen Rauch-
schaden. Mitt. Forstl. Bundesvers'anst. Wien 92, 155-
206.
Renfrew, C., 1977: Ancient Europe is older than we
thought. Nat. Geogr. Mag. 152,615-62.
Renner, F., 1982: Beitrage zur Gletschergeschichte des
Gotthardgebietes und dendroklimatologische Analysen
an fossilen Holzern. Physische Geographie. Universitat
Zurich, 8, 182 pp.
Reventlow, C. D. F., 1879: Forslag til en forbedret skov-
drift grundet paa undersogelser over traeernes vegeta-
tion i Danmarks og Slesvigs skove. Kopenhagen,
Hauberg. 206 pp.
Richter, K., and Eckstein, D., 1986: Estudio dendrochro-
nol6gico en Espana. Dendrochronologia 4, 59-74.
Robinson, W. J., 1967: Tree-ring materials as a basis for
cultural interpretations. Diss. Univ. of Arizona, Tucson.
Unpublished, 107 pp.
Robinson, W. H., 1976: Tree-ring dating and archaeology
in the American Southwest. Tree-Ring Bull. 36, 9-20.
Rose, M. R., Dean, J. S., and Robinson, W. J., 1981: The
past climate of the Arroyo Hondo New Mexico recon-
. structed from the tree rings. School of American re-
search press. Arroyo Hondo arch. Series 4, 114 pp.
Rothlisberger, F., 1976: 8000 Jahre Walliser Gletscher-
geschichte: ein Beitrag zur Erforschung des Klimaver-
laufs in der Nacheiszeit. 2. Teil. Die Alpen 52, 59-152.
Rothlisberger, F., 1986: 10000 Jahre Gletschergeschichte
der Erde. Aarau, Frankfurt a.M., Salzburg. Sauerlander,
416 pp.
Rubi, Chr., 1980: Das Simmentaler Bauernhaus. Berner
Heimatbucher 125. Bern, Paul Haupt. 155 pp.
Ruoff, U., 1979: Neue dendrochronologische Date aus
der Ostschweiz. Zeitschr. Schweiz. Arch. und Kunst-
gesch. 36, 94-96.
Ruoff, U., and Rychner, V., 1986: Die Bronzezeit im
schweizerischen Mittelland. Antiqua 15, 73-79.
Rypacek, V., 1966: Biologie holzzerstorender Pilze. Jena,
Fischer. 211 pp.
Sachs, L., 1978: Angewandte Statistik. Statistische
Methoden und ihre Anwendung. 5. Edn. Berlin, Heidel-
berg, New York, Springer. 522 pp.
Sanderman, W., Schweers, W., and Gaudert, P., 1960:
Messung der Holzdichte und Bestimmung der Holz-
Jahrringbreite mit Hilfe von Beta-Strahlen. Forstarchiv
31,126-128.
Sauvala, K., 1982: Wood microdensitometry bibliography.
Wood Microdensitometry Bulletin 2(1), 6-54.
Schafer, H., 1954: Forstkriminalistik. Hamburg, 85 pp.
Schar, E., Schweingruber, F. H., 1987: Nacheiszeitliche
Stammfunde aus Grachen im Wallis. Schweiz. Z Forst-
wesen, 138, 497-515.
Scharpf, R. F., 1978: Control of Dwarf Mistletoe on true
Firs in the West. Pacific Southwest Forest and Range
Experimental Station. Gen. Tech. Rep. PSW - 31,
117-123.
Scheffer, F., and Schachtschabel, P., 1970: Lehrbuch der
Bodenkunde. 7. Aufl. Stuttgart. 448 pp.
Schichtl, H. M., 1973: Karte der aktuellen Vegetation
Tirols. Blatt 5, Silvretta und Lechtaler Alpen. Documents
de Cartographie Ecologique 11, Grenoble.
Schirmer, W., 1979: Rannen im Mainschotter. Frankische
Heimat am Obermain, Heft 16. Beilage zum Jahres-
bericht 1978179 des Merainer-Gymnasiums lichten-
fels. 44 pp.
Schmidt-Vogt, H., 1977: Die Fichte; Band I: Taxonomie,
269
 Verbreitung, Morphologie, Oekologie, Waldgesellschaf-
ten. Hamburg und Berlin, Parey. 647 pp.
Schmucker, Th., and Linnemann, G., 1951: Geschichte
der Anatomie des Holzes. In: H. Freud, Handbuch der
Mikroskopie in der Technik, Bd. 5/1, 1-78.
Schneebeli, W., 1976: 8000 Jahre Walliser Gletscherge-
schichte; ein Beitrag zur Erforschung des Klimaverlaufs
in der Nacheiszeit. 1. Teil. Die Alpen 52(3/4), 1-57.
Schneider, S. H., 1978: Klima in Gefahr. Strategien zur
Beherrschung des Welters. Frankfurt am Main, Fischer;
Translation of the American edition: The genesis stra-
tegy - Climate and global survival (1976). New York,
Plenum Press. 396 pp.
Schoepfer, W., 1962: Beitrage zur Erfassung des Assimi-
lations-apparates der Fichte. Diss. Univ. Freiburg. 127
pp.
Schotterer, U., and Oeschger, H., 1979: 1m Eis gespei-
cherte Information. In: Kasser P.: Die Schweiz und ihre
Gletscher. Kummerli und Frey, Bern.
Schotterer, U., 1986: Der Gletscher als Reinraumarchiv
von Umweltvorgangen. Swiss Chem. 8, 49-52.
(Schweizerische Zeitschrift fUr die chemische Industrie.)
Schroeder, D., 1969: Bodenkunde in Stichworten. Kiel,
Ferdinand Hirt. 144 pp.
Schulman, E., 1937: Some early papers on tree-rings. J.
C. Kaptein. Tree-Ring Bull. 3,28-29.
Schulman, E., 1958: Bristlecone pine, oldest known living
thing. Nat. Geogr. Mag. 113,355-372.
Schuett, P., 1981: Ursache und Ablauf des Tannenster-
bens - Versuch einer Zwischenbilanz. Forstw. Cbl.
100, 286-287.
Schweingruber, F. H., 1978: Mikroskopische Holzana-
tomie; Formspektren mitteleuropaischer Stamm- und
Zweighblzer zur Bestimmung von rezentem und sub-
fossilem Material. Eidg. Anst. forst/. Versuchswes. 226
pp.
Schweingruber, F. H., Fritts, H. C., Braeker, O. U., Drew,
L. G., and Schar, E., 1978: The X-ray technique as
applied to dendroclimatology. Tree-Ring Bull. 38, 61-
91.
Schweingruber, F. H., 1979: Auswirkungen des Larchen-
wicklerbefalls auf die Jahrringstruktur der Larche;
Ergebnisse einer Jahrringanalyse mit rbntgendensito-
metrischen Methoden. Schweiz. Z. Forstwes. 130(12),
1071-1093.
Schweingruber, F. H., Braeker, O. U., and Schaer, E.,
1979: Dendroclimatic studies on conifers from central
Europe and Great Britain. Boreas 8,427-452.
Schweingruber, F. H., and Ruoff, U., 1979: Stand und
Anwendung der Dendrochronologie in der Schweiz. Z.
schweiz. Arch. Kunstgesch. 36,69-96.
Schweingruber, F. H., 1980: Dichteschwankungen in
Jahrringen von Nadelhblzern in Beziehung zu klima-
tisch-bkologischen Faktoren, oder das Problem der
falschen Jahrringe. Ber. Eidg. Anst. forstl. Versuch-
swes. 213, 35 pp.
Schweingruber, F. H., Schar, E., and Braker, O. U., 1984:
Jahrringe aus sieben Jahrhunderten. Saaner Jahrbuch
1984,1-30.
Schweingruber, F. H., 1985: Dendroecological zones in
the coniferous forests of Europe. Dendrochronologia 3,
67-75.
Schweingruber, F. H., 1987: Flachenhafte dendroklima-
tische Temperaturrekonstruktionen fur Europa. Natur-
wissenschaften 74, 205-212.
270
Schweingruber, F. H., Kontic, R., and Winkler-Seifert, A.,
1983: Eine jahrringanalytische Studie zum Nadelbaum-
sterben in der Schweiz. Ber. der Eidg. Anstalt fUr das
forstliche Versuchswesen 253, 29 pp.
Schweingruber, F. H., Albrecht, H., Beck, M., Hessel, J.,
Joos, K., Keller, D., Kontic, R., Lange, K., Niederer, M.,
Nippel, C., Spang, S., Spinnler, A., Steiner, B., and
Winkler-Seifert, A., 1986: Abrupte Zuwachsschwan-
kungen in Jahrringabfolgen als oekologische Indika-
toren. Dendrochronologia 4, 125-183.
Schwenke, W., 1972, 1974-1978: Die Forstsch3.dlinge
Europas. 3 Bande. Hamburg, Berlin, Parey.
Sell, J., 1978: Quantitative automatische Bildanalyse in
der Materialprufung. Prinzip, Systemaufbau, Anwen-
dungen. Material und Technik 6, 79-83.
Serre, F., 1973: Contribution a I'etude dendroclimato-
logique du pin d'alep (Pinus halepensis). Ph.D. these,
I'universite D'Aix-Marseille III.
Shane, D., and Harper, L. T., 1979: Influence of precipita-
tion and temperature on ring annual branch increment
and needle-growth of White Fir and Douglas Fir in
central Utah. Great Basin Naturalist 39, 219-225.
Shigo, A. L., and Marx, H. G., 1977: Compartmentaliza-
tion of decay in trees. Forest Service US Dept. of Agri-
culture. Agriculture Information Bul/' 405, 73 pp.
Shigo, A. L., 1986: A new tree biology. Shigo and Trees
Assoc. Durham, NH 03424, USA, 595 pp.
Shigo, A. L., Campana, R., Hyland, F., and Anderson, J.,
1980: Anatomy of Elms injected to control Dutch Elm
Disease. Journal of Arboriculture 6, 96-100.
Shroder, J. F., 1976: Dendrogeomorphological analysis of
mass movement on Table Cliffs Plateau, Utah. Quarter-
nary Research 9, 108-185.
Shroder, J. F., 1980: Dendrogeomorphology: review and
new techniques of tree-ring dating. Progress in Physical
Geography 4, 161-188.
Shroder, Jr. J. F., and Butler, D. R., 1987: Tree-ring
analysis in the earth sciences. In press.
Shvedov, F. N., 1892: Trees as chroniclers of droughts.
In: J. M. Fletcher and W. Linnard, 1977. Russian papers
on dendrochronology and dendroclimatology 1962/
1968, 1970, 1972. Res. Lab. Arch. and Hist. of Art,
Oxford University 39-49.
Sieben list-Kerner, V., 1984: Der Aufbau von Jahrring-
chronologien fUr Zirbel kiefer, Larche und Fichte eines
alpinen Hochgebirgsstandortes. Dendrochronologia 2,
9-29.
Smiley, T. L., 1958: The geology and dating of Sunset
Crater, Flagstaff, Arizona. In: Guidebook for the Black
Mesa Basin, Northeastern Arizona, 186-190.
Smith, L. P., and Stockton, Ch. W., 1981: Reconstructed
stream flow for the Salt and Verde Rivers from tree-ring
data. Water Resources Bul/. 17, 939-947.
Stahle, D. W., 1979: Tree-ring dating of historic buildings
in Arkansas. Tree-Ring Bull. 39, 1-28.
Stallings, W. S., 1937: Some early papers on tree-rings.
Tree-Ring Bull. 3, 27-28.
Stockton, Ch. W. 1971: The feasibility of augmenting
hydrologic records using tree-ring data. Ph.D. Diss.
University of Arizona, Tucson.
Stockton, Ch. W., and Boggess, W. R., 1980: Tree rings:
A proxy data source for hydrologic forecasting. In North
R. M. et al. (eds.) Unified Basin management. Sym-
posium proceedings, May 4-7, 1980. Gatlingburg,
Tennessee, 605-624.
Stockton, Ch. W., and Jacoby G. C. Jr., 1976: Long-term
surface water supply and streamflow trends in the
upper Colorado River basin based on tree-ring analysis.
Lake Powell Research Project BUll. 18, University of
California, Los Angeles, Institute of Geophysics and
Planetary Physics, 70 pp.
Stockton, Ch. W., Mitchell, L. M., and Meko, D. M., 1981:
Tree-ring evidence of a relationship between drought
occurence in the western United States and the Hale
Sunspot Cycle. In: Lawson, M. P., Baker, M. E., (eds.)
The Great Plains. Perspectives and Prospects. Uni-
versity of Nebraska Press, Lincoln and London, 83-
110.
Stokes, M. A., and Smiley, T. L., 1968: An introduction to
tree-ring dating. Chicago and London. University of
Chicago Press, 73 pp.
Stokes, M. A., and Dietrich, J. H., (eds.) 1980: Proceed-
ings of the fire history workshop. Tucson, Arizona.
General Tech. Report RM-81. Rocky Mountain Forest
and Range Experiment Station. Forest Service. US
Dept. of Agriculture, 142 p.
Strahler, A. N., 1975: Physical geography. 4th Ed. New
York, John Wiley and Sons, Inc.
Strahm, Ch. 1977: Kontinuitat und Kulturwandel im Neo-
lithikum der Westschweiz. Fundberichte aus Baden
Wurttemberg 3, 11 5-143.
Straka, H., 1970: Walter, H., Arealkunde; f/oristisch his-
torische Geobotanik. 2. Aufl. Stuttgart, Ulmer. 478 pp.
Stuiver, M., and Quay, D., 1980: Changes in atmospheric
carbon-14 attributed to a variable sun. Science 207,
11-19.
Stuiver, M., 1980: Solar variability and climatic change
during the current millennium. Nature 286, 868-871.
Stuiver, M., and Grootes, P. M., 1980: Trees and the
ancient record of heliomagnetic cosmic ray flux modula-
tion. In: Pepin, R. 0., Eddy, J. A. and Merrill, R. B.,
Proceedings of the Conference Ancient Sun, 165-
1973.
Stuiver, M., Rebello, A. L., White J. C., and Broecker,
w., 1981: Isotopic indicators of age/growth in tropical
trees. School of Forestry and environmental studies.
Yale University Bull. No. 94, 75-82.
Stuiver, M., and Quay, P. D., 1981: Atmospheric 14C
changes resulting from fossil fuel CO 2 release and
cosmic ray flux variability. Earth and Planetary Science
Letters 53, 349-362.
Stuiver, M., 1982: A high-precision calibration of the AD
radiocarbon timescale. Radiocarbon 24, 1-26.
Stuiver, M., 1983: Statistics of the climatic and carbon
isotopic change. Radiocarbon 25, 219-228.
Suess, H. E., 1967: Bristlecone pine calibration of the
radiocarbon time scale 5200 B.C. to 1500 B.C. In:
Radioactive dating and methods of low-level counting:
Vienna IAEA, 143-151.
Suess, H. E., 1979: 1st die Sonnenaktivitat fUr Klima-
schwankungen verantwortlich? Umschau 79, 312-316.
Suter, P. J., and Schifferdecker, F., 1986: Das Neolithi-
kum im schweizerischen Mittelland. Antiqua 15, 34-
43.
Swetnam, T. W., 1985: Radial growth losses in Douglas-
Fir and White Fir caused by Western Spruce Budworm
in northern New Mexico: 1700-1983. Forest Pest
Management Report R-3, 86-2. 62 pp. US Dept. of
Agriculture, Forest Service, Southwestern Region.
Swetnam, T. w., and Dietrich, J. H., 1983: Fire history of
Ponderosa Pine forest in the Gila Wilderness, New
Mexico. Unpublished paper, presented at the Wilder-
ness Fire Symposium, Missoula, Montana, Nov. 1983.
Swetnam, T. W., Thompson, M. A., and Kennedy Suther-
land, E., 1985: Using dendrochronology to measure
radial growth of defoliated trees. US Dept. of Agricul-
ture, Forest Service, Agriculture Handbook 639, 1-39.
Tande, G. F., 1979: Fire history and vegetation pattern of
coniferous forests in Jasper National Park, Alberta.
Canadian Journal of Botany 57, 1912-1932.
Taubenheim, J., 1969: Statistische Auswertung geophy-
sikalischer und meteorologischer Daten. Akad. Ver-
lagsges. Geest und Portig K.-G., Leipzig. 386 pp.
Terskov, I. A., Vaganov, E. A., Zyrjanov, G. I., and Phili-
monov, V. S., 1978: The use of cell size distribution-
curves for tree growth analysis. Mitt. der sibirischen
Abt. der Akad. Wiss. USSR (Russian). Serie Biolo-
gische Wissenschaften, 26-36.
Thenius, E., 1977: Meere und Lander im Wechsel der
Zeiten. Verstandliche Wissenschaft Bd. 114. Berlin,
Heidelberg, New York, Springer. 200 pp.
Thompson, M. A., 1981: Tree rings and air pollution: A
case study of Pinus monophylla growing in east Central
Nevada. Environmental Pollution (Series A) 26, 251-
265.
Till, C., 1984: A synthesis of response functions from
eight Cedar forests located in Northern Africa. Den-
drochronologia 2,73-82.
Timell, T. E., 1986: Compression wood in Gymnosperms.
3 vols. Berlin, Heidelberg, New York, Tokyo. Springer,
2150 pp.
Tranquillini, W., 1979: Physiological ecology of the alpine
timberline. Berlin etc. Springer, 137 pp.
Trendelenburg, R., and Mayer-Wegelin, H., 1955: Das
Holz als Rohstoff. Munchen, Carl Hanser. 541 pp.
Ur- und Fruhgeschichte Archaologie der Schweiz 1968:
Band I. Die Aeltere und Mittlere Steinzeit. Schweiz.
Ges. Ur- und Fruhgesch., Basel, 146 pp.
Vaganov, E. A., and Terskov, I. A., 1977: Die Analyse des
Baumwachstums anhand der Jahrringstruktur (Rus-
sian). SA der Akad. der Wiss. UdSSR Sibirische Abt.
Forst- u. Holzinst. W. N. Suskatschow. Verlag Wis-
sensch. Nouosibirsk. 93 pp.
Various authors, 1975: (SBN) Wald und Umwelt; Informa-
tion des schweizerischen Bundes fUr Naturschutz und
der liechtensteinischen Gesellschaft fUr Umweltschutz.
Various authors, 1979: Harms Geographie. Physische
Geographie und Nachbarwissenschaften. 424 pp.,
Munchen, List.
Varley, G. C., 1978: The effects of insect defoliation on
the growth of oaks in England. In: J. Fletcher, Den-
drochronology in Europe. BAR Int. Series 51, 179-
186.
Wade, J. E., and Hewson, E. W., 1980: A guide to
biological wind prospecting. Final report for US Dept. of
Energy, Div. of distributed solar technOlogy, Federal
wind energy program. 112 pp.
Wade, J.E., and Wendel-Hewson, E., 1979:Trees as a
local climatic wind indicator. Journal of Applied Meteor-
ology. 18, 1182-1187.
271
WagenfOhr, R., 1966: Anatomie des Holzes. Leipzig. VEB
Fachbuch-Verlag. 377 pp.
Walter, H., and Lieth, H., 1960/1964/1967: Klimadia-
gramm-Weltatlas. 3 Bde. Jena, Fischer.
Walter, H., 1964: Vegetation der Erde. Bad I Tropische
und subtropische Zonen. Fischer, Jena.
Walter, H., 1964, 1968: Die Vegetation der Erde in oko-
physiologischer Betrachtung. Band I: Die tropischen
und subtropischen Zonen. Band 2: Die gemassigten
und arktischen Zonen. Stuttgart, Fischer.
Webb, G. E., 1983: Tree rings and telescopes. The
scientific career of A. E. Douglass. The University Press
of Arizona, Tucson, Arizona. 242 pp.
Welten, M., 1981: Gletscher und Vegetation im Lauf der
letzten hunderttausend Jahre. Verh. Schweiz. natur-
forsch. Ges., Gletschersymposium in Brig 1978. Basel.
Weltforstatlas 1951-1971: Weltforstatlas - World fores-
try atlas - Atlas des forets du monde - Atlas forestal
del mundo. Hg. Bundesforschungsanstalt fOr Forst- und
Holzwirtschaft, Reinbeck. Hamburg, Berlin, Parey.
Wendel-Hewson, E., Wade, J. E., and Baker, R. W., 1977:
Vegetation as an indicator of high wind velocity. Final
Report for US Dept. of Energy, Div. of Solar Energy.
Federal wind Program. Erstellt von Oregon State
University, Dept. of Atmospheric Sciences, Corvallis,
Oregon. 58 pp.
White, G. F., 1972: History of fire in North America. In:
Fire in the environment, Symp. Proc., May 1-5, 1972.
Forest Service, US Dept. of Agricult., North American
Forestry Commission, FAO, FS-276, 3.
White, J. W. C., and Lawrence, J. R., 1980: The relation-
ship between the non-exchangeable hydrogens of tree-
ring cellulose and the source waters for trees. In: G. C.
Jacoby, (Ed.) Carbon Dioxide Effects; Res. and Assess-
ment Program, Proc. Int. Meeting on Stable Isotopes in
Tree-Ring Research, New York, Lamont-Doherty Geol.
Observ., Columbia Univ., 58-65.
White, J. W. C., 1983: The climatic significance of D/H
ratios in White Pine in the northeastern United States.
Diss. Columbia University New York. 331 pp.
Wickman, B. E., 1978: A case study of a Douglas-Fir
Tussock Moth Outbreak and stand conditions 10 years
later. USDA Forest Service research Paper PNW-244.
22 pp.
Wickman, B. E., Henshaw, D. L., and Gollob, S. K., 1980:
Radial Growth in Grand Fir and Douglas-Fir related to
defoliation by Douglas-Fir Tussock Moth in Blue Moun-
tains Outbreak. US Dept. of Agriculture. Forest Service.
Pacific Northwest Forest and Range Experiment Station.
Research Paper NNW-269. 23 pp.
272
Wieler, A., 1905: Untersuchungen tJber die Einwirkung
schwefliger Saure auf die Pflanzen. Berlin und Leipzig.
Borntrager. 427 pp.
Willkomm, H., 1976: Alterbestimmungen im Quartar.
Munchen, Thiemig. 276 pp.
Wininger, J., 1976: Feldmeilen-Vorderfeld. Die Ausgra-
bungen 1970-1971. Schweiz. Ges. Ur- und FrtJh-
gesch. Basel, 1-106.
Worbes, M., 1985: Structural and other adaptions to long-
term flooding by trees in Central Amazonia. Amazonia
9,459-484.
Worbes, M., 1986: Lebensbedingungen und Holzwach-
stum in zentralamazonischen Ueberschwemmungs-
waldern. Scripta Geobotanica 17, 112 pp.
Yamaguchi, D. K., 1983: New tree ring dates for recent
eruptions of Mount St. Helens. Quarternary Reserach
20, 246-250.
Yanosky, T., 1983: Evidence of floods on the Potomac
River from anatomical abnormalities in the wood of
flood-plain trees. US Geol. Survey Prof. Paper 1296, 42
pp.
Zimmermann, M. H., and Brown, C. L., 1971: Trees:
structure and function. Berlin, Springer, 336 pp.
Zimmermann, W. 1959: Die Phylogenie der Pflanzen.
Stuttgart, Fischer, 777 pp.
Zoller, H., Schindler, C., and Rothlisberger, H., 1966:
Postglaziale Gletscherstande und Klimaschwankungen
im Gotthardmassiv und Vorderrheingebiet. Verh. Natur-
forsch. Ges. Basel 77, 97-1 64.
Zuber, R., Tschannen, W., and Bovey, E., 1981: Controle
de la teneur en fluor des feuilles d'abricotiers de la
Vallee du Rhone de 1974 a 1980. Rev. Suisse Vitic.
Abric. HortiG. 13, 133-138.
Recently two volumes have been published who reflect the
modern standard in dendroecology:
Jacoby G. c., Hornbeck J. W. (editors) 1987: Proceedings of
the international symposium on ecological aspects of
tree-ring analysis. August 17-21, 1986, Marymount Col-
lege, Tarrytown, New York. 726 pp. (Available from the
National Technical Information Service, U.S. Department
of Commerce, Springfield, Virginia 22161, USA.)
Isaev et al. (editors) 1987: Dendrochronological methods in
forest science and ecological forcasting. Irkutsk USSR,
313 pp.
General index
Abies alba 171, 240, 242, 247
Abies cephalonica 128, 159
Abies grandis 233
Abrupt growth changes 48, 190,
195,211,212,227,240
Acclimatization 63
Acer pseudoplatanus 244
Acersaccharinurn 55
Acid rain 247
Actinornycetes 34
Aeolian processes 179
Aerobic 34
Afforestation 238
Age determination 96, 108, 109,
237,251
distribution 240
structure 203
trend 78, 121
Aging process 121, 238, 239
Air pressure 18, 210
Alpine zone 104, 110
Alps 126, 127
Anatomical technique 40
American buildings 152-155,
157
America, northern 152-155, 157,
172, 206-210
America, southern 180, 211
Anaerobic 34
Annual growth analysis 136-139
Aperiodic increments 3
Applied dendrochronology 143-
253
Araucaria araucana 17, 180, 211
Arceutobiurn sp. 230,231
Arctic zone 104, 110, 231
Arizona 45,152-155,183,221
Arid zone 104, 111
Arithmetic mean 81
Armillaria rnellea 229
Art history 162-165
Austrocedrus chilensis 180
Axes, stone 150-155
Bacteria 34
Balkan 159
Barcharts 51,159,185,188,198
Bark 50,97
Beaver-teeth marks 155
Beech 100, 102
Beetle 33, 222
Beetle galleries 155
Belt sander 46
Biological defence 228
Biological errors 72
Boreal climate 126, 128
Bog oak 184-186,188,189
Borer 42,43
power driven 43
Bore hole 44
repairing 45
sealing 45
Bore support 42, 43
Branch cutting 244 Complacent 49, 74, 82
Bristlecone pine 14, 27, 31, 136, Composite 50
174,215,249 Compression wood 36, 49, 72,
Bronze age 154 99,177,200
Building activity 149, 153, 154 Conifer 7, 10, 11, 55, 61, 63, 70-
Buried trunks 45,178,184-189, 72,99,103,114,115,145
196 Core cutting 58
holder 43
14C-Dating 248-253 mounting 43
14C-Age 248, 249 orientation 42, 60
Calibration 57, 68, 90 support 60
Callixylon newberry 36 transport 43
Callusing 44, 205, 215, 219 Coring 58
Cambial activity 19, 98, 109, 139 Correlation 84, 247
Carbonization 35 auto- 84
Carving 160 coefficient 84
Caterpillar 225 serial 84
Catface 205 Cosmic radiation 248
Cathedrals 156, 158 Coverage 51
Cell differentiation 7,98, 140 Criminal investigation 166, 167
enlargement 98 Cross dating 50, 51, 76, 77
form 113 Crown damage 246
formation 98 Counting tree rings 47
numbers 100 Cumulative temperature differ-
size 99,113 ence 18,92
Cellwall density 55 Curve fitting 75
growth 99 trend 75
thickness 114, 115, 176 Cutting 40, 52
Cells dead 176 Cycles, lunar 174
living 99,141,176 sunspot 174
Cellulose acetate 62 insect damage 223, 225
Cellulose nitrate 252
Charcoal 35,47,258,259 Deciduous trees 28, 33, 35, 98,
Chemical trace elements 247 99,103-105,109-112,118
Churches 159, 160 Density diagram 68,117,119,
Chronology, absolute 151 237
bristlecone pine 26, 27 difference 115
douglas fir 249 fluctuations 116, 117, 129, 135
larch 26,29,225 integrator 69
oak 26,184-186,188,189, integration 69
209 maximum 114
relative 150 minimum 115
regional 128 level 71
scar 206, 207 pattern 123
spruce 26,29 threshold 69, 71
City trees 91, 244 profile 33, 35, 41, 56, 107, 124
Cliff dwellings 152-155 variations 116, 124
Climate, boreal 126, 128 Destruction period 185
mediterranean 139 Deuterium 252, 253
temperate 127,139 Diffuse porous 100
Climatic diagrams 8, 9, 20, 106, Discoloration 44
108,236 Distortion beam 147
fluctuations 170, 175 Distribution, normal 84
information 23, 24 skew 84
integrators 20 Diurnal rings 2
reconstruction 80, 90, 170- Double-mass analysis 92
174 Douglass, A. E. 257
zones 8 Douglas fir 155, 181
Climatology 168-175 Drainage 190, 191, 237
CO 2 -exchange 6 Driftwood 33, 148, 21 0
Cockchafers 222 Daily radial growth 19
Colorado river 181 Data correcting 72
Colored deposits 57, 101 index 78,85,90
273
 meteorological 18, 92
processing 73,78-92
raw 73
Dating 76
accuracy 77
methods 76
plantation 237
problems 76
radiocarbon 76, 251
Decomposition 33,34,37,45,47
Defoliating 222-227
Dendroclimatology 168, 175
Dendroecology 54,95-141
Dendrochronology, applied 143-
253
Dendrochronological age
249
Dendrometres 19, 237
Drinking straw 43
Dry sites 25
Dwarf tree 203
Earth movement 179
Earthquakes 216
Earlywood 52, 71, 99, 100
Ecogram 102,136,137,139
Ecklund machine 53
Ecological factors 19, 102
gradient 136
limitation 16
Ecophysiology 19
Elm 253,103
Elm disease 228
Entomology 222-227
Environmental sciences 240
Erosion 179, 190
Europe 128,169,172
Event diagram 48, 191, 198,
240,241,246
Exposure time 65
Extractives 62
Fagus sy!vatica 98, 100, 102
False ring 47, 205
Farm house 158, 160
Felling date 145, 147, 152, 162-
167
Fertilisation 234
Film developing 64
Fire 35
map 209
scar 205-207
Filter 86, 87
band-pass 87
binomial 86
Fitzroya cupressoides 26, 32
Flag tree 200
Floods 183, 192, 193, 210
Fluoride 247
Flow chart 80
Fomes annosus 32
Forest fire 204-209
Fossil fuels 250
pine 36
soil 213
wood 96,220,221
274
Fossilization 36 Icings 219
Freeze drying 37 Ice jams 219
Frost ring 214, 215 Ice movement 179,194-197
Fritts, H. C. 261 Identification of wood 166
Fungal infection 32, 228 Image analysis 55
Fungicidal substance 32 Increment borer 42
Function, exponential 87 Increment layers 3
Hugershoff 87 Indexing 85
negative exponential 87 Indians 152-155, 212
polynomial 87 Insect galleries 33, 72, 145, 155
straight line 87 damage 222-235
Interval trend 83, 134
Gas exchange 6 Inversion layers 246
Geiger counter 65 Irradiation 65
Geographical distance 126, 128,
248, 131 Katmai 214
Geological substrate 11, 23 Knife sharpener 40
periods 12 Krakatau 215
Geomorphology 176-179
Giant waves 217 Laachen volcano 187,213
Glacier history 194-197 Labelling 43, 61
Gleichlaufigkeit 77, 83, 123, Lake shore settlement 149, 150
126-128 Landscape damage 45
Grafting wax 44 Larix decidua 26, 29, 71, 195-
Ground water 190 197
Growth change 7,48,239 Larix sibirica 56
abrupt 190, 191, 195, 211, Larch bud moth 223, 225
212,227,238 Larch 27, 29, 56, 70, 195-197
fluctuations 96 Latewood 52, 71, 99, 100, 112
increments 96 Lath 61
layers 96, 97, 109, 220, 221, Light influence 11
237 reflection 56, 57
model 140,141 slit 59
reduction see Growth change transmittance 54, 56
-reduction diagrams 242, 244 Lightning 204
release see Growth change Lindbergh baby 167
rhythms 96, 237 Living cells 141, 176
maps 169 Log constructions 30
Logarithmic scale 51
Half-life time 248 Long chronologies 26
Hand lens 52 Long term deviation 249
Hartig, T. 256 Lower forest limit 25, 170
Hazelwood 52 Libby, W. F. 248
Heartwood 57,101,146
History, Art- 162-165 Marks, Axe- 147
House- 156-160 Maximum density, mainly at
Settlement- 149, 152 pages 58-73, 114, 120, 123,
Historical buildings 156-160 127, 128, 175
Holding device 60 Mean values 79
Hollstein, E. 32 Measuring slit 70
Honeyfungus 229 Mechanical damage 44, 228,
Horgen period 149 229
House construction 32, 148, 149, Medieval ruin 31
152,157,158 Mediterranean climate 133, 139
ground plans 149, 152, 153, Megalithic culture 251
154 Meteorological data 92
Human activity 32, 144, 179 station 18
Huber, B. 28, 260 Microdensitometer 66, 67
Hydrogen bomb 251 Microfiche reader 73
Hydrological analysis 180, 190, Microscopic technique 40
191 Microsections 46, 57
changes 179 Microstructure 28, 29, 32, 33,
Hydrology 23, 180-193, 216 35-37, 52, 96-105, 107, 109,
111-115
Ice age 13 variability 103
Ice falls 217 Mineralisation 36
Minimum density, mainly at pages contorta 206,231 Running mean 86, 130
115,123,127,131 edulis 247,253 Runoff 180
Missing ring 47,54 leucodermis 128, 159, 204
Mistletoe 230,231 longaeva 14, 26, 27, 136, 174, Sampling site 16, 17, 21, 22
Moisture content 32, 63 215,249 Sandpaper 46
Molluscs 187 nigra 15 Sapwood 57,101,146
Monitor 54 patula 239 Santorini 215
Moraine 196 ponderosa 155, 204, 205, 207, Saw blade 61
Mortality 231, 245, 246 208,231 Sawing 60
Mount St. Helens 210-213 radiata 234 Scale arithmetric 51
strobus 62, 157 logarithmic 51
Near East 159 sylvestris 11, 21, 35, 55, 71, Scar chronology 206, 207
Needle cover 246 102,106,187,192,193 Schulman, E. 27, 259
length 7 Polge, H. 261 Sensitive 49,74,82
Neolithic house 149 Pointer years 49, 134, 240 Sensitivity 82
settlement 31 Polishing surface 46, 57 Sequoia gigantea 26, 216
Network 18, 24, 25, 169, 246 Pollution 240-247 Settlement, history 149, 152
Normal distribution 81 Polyethylene glycol 37 phases 149,151,153
Norway spruce see spruce Poplar 47,182 Shoot 7,176
Northern America see America Posts 149, 155 length 7
Nothofagus pumilio 17, 211 Pottery 140 terminal 124
Power planers 46 Short term variation 249
Oak 27, 47, 62, 100, 146, 147, Preparation methods 40 Showlow 258
183 Preservation 37 Shrinkage 34
Bog- 184-186,188,189 Principal components 88 Signatures 49
chronology 32, 184-186 Print plot 73 Sign test 83
subfossil 184-186, 188, 189 Pruning 232 Similarity 126-131
Ocean currents 9 Pseudotsuga menZleSIl 155, Site characteristics 20
Optical density 55, 66 181,224,233 conditions 17
Orientation equipment 60 description 22, 23
Outermost tree ring 33, 51, 145. Quercus petraea 28 selection 17,21,22
154 robur 28 Skeleton plot 50
Overlapping 51, 76 sp. 5, 26, 49, 62, 184-186, Sledge microtome 40
Oxygen C6 0, 180) 252,253 188,189 Slope movement 198
Smoothing functions 78, 86
Paintings 162-165 Radiocarbon 248-253 Snap off blade 46
Paleotemperatures 13 age 248,249 Snow cover 203
Peeling 229 dating 28, 29, 76 movement 179
PEG 37 Radiation 6 S02-emission 247
Periodic damage 222, 223, 225 Radiodensitometry 40, 58-73 Soil 23
Petrified forest 45, 222 Radiographic density 69 Soil compression 244
wood 36, 220, 221 Raw data 58,107,123 improvement 234, 235
Period atlantic 13 Ray geometry 65 Solar wind 250
boreal 13 Reaction time 6 Soxhlet extractor 62
late glacial 13, 187 Regional chronology 128 Spatial agreement 128
poste glacial 184-189, 192, Release see growth change Spiral growth 60
193 Renovation 148, 149 Spruce 29, 46, 98, 103, 114,
subboreal 13 Repairs 148,149 115, 120
tertiary 12 Reservoir 182 Spruce bud worm 222,225,226
Periodicity 75,174,240 Response function 88, 89, 133, Stable isotopes 252
Photomicrography 40 143,170,171 Staining 47, 52
Photometric technique 56 Red rot 32 Standardization 85
Physiological processes 19, 140 Resin 62 Stand stability 232
Phytopathology 228-231 duct 62 Standard deviation 81, 82
PHh 50,9~ 96,103,113 duct, traumatic 72, 205 Statistics 78-92
Plantation date 237 extraction 58 Stem eccentricity 198-202
Picea abies 21, 26, 29, 103 secretion 44 Stepping motor 66
engelmannii 15 Rhythmic growth 2, 3, 96, 97 Stereomicroscope 52
mariana 15 Ring-porous wood 100, 111, Stone axe 150, 155
obovata 29 118, 145 Storm 202
rubens 15, 56 Ring width measurement 52, 53 Stradivarius 167
sitchensis 15 River change 182 Street salting 244
Pine, fossil 36,187,192,193 Road salting 54, 244 Stringed instruments 167
Pinus aristata, see P. longaeva Robinia pseudoacacia 118 Structure analysis 54, 55
caribaea 238 Root, adventitous 176 Subalpine trees 126, 133
cembra 19, 29 damage 244 Subfossil oak 184-186, 188, 189

275
 pine 36, 187, 192, 193
wood 31,35,184-189,195-
197
Submerged forest 183, 192
Sulfate 247
Summer temperature 24, 25, 170,
171,175
Sunset crater 212
Sunspot cycles 174, 250
Supression 11, 212
Surface cutting 40
polishing 40
Synthetic resin 37
Technical error 72
Tectonic movement 193, 218
Temperate climate 127, 139
zone 105, 111
Temperature amplitude 8
Terminal ring 33, 145, 154
Tetrac/inis sp. 62
Thinning 232
Timber 156, 157, 160
re-use 148
Timberline 16, 24, 136, 170
upper 24
Time lag 246. See also response
function
Tissue analysis 40,54,55
276
Topography 10 Volume increment 232, 239
Transmitted light 56 Volumetric gravimetric density 62
Transportation 6, 148
Transport, cores 43 Weathered beam 38, 33, 148
Traumatic tissue 49 Weather reconstruction 24
Tree form 106, 120 Wedge cellulose 68
Tree ring boundaries 11 0-115 optical 68
characteristics 49, 107, 119, Wheat harvest 8
120,132 Wind influence 10, 106, 200-203
Tropical trees 108, 109, 236- velocity 201
239,251 Wood, compression, see
wood 47 compression wood
zone 105 identification 166
Tussock moth 224 preservation 32, 37
Twin-bladed saw 61 technology 238, 239
Tylosis 101, 228
X-ray absorption 62,69
Ulmus campestris 103 chamber 64
Uplifting 218 equipment 64
film 59,64
Varves 3 film coating 65
Verification 90, 91 picture 61
Vessel area 54 source 65
Viscum album 228 technique 58-73
Visual data checking 74 tube 65
Visual matching 49
Volcanic activity 179, 187, 210- Year, felling 145
215
ash 211 Zeiraphera diniana 222, 223, 225