Professional Documents
Culture Documents
Tree Rings
Basics and Applications of Dendrochronology
Reprinted in 1989
Preface XII
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
C. F. von Weizsacker
Preface
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:
These two pages show a few of the many cases of Growth processes can often only be detected through
rhythmic growth increments occurring in nature. No measurements and are rarely reflected in structures.
process of development is continuous and no organism Organisms with growth structures are as a rule short-lived
grows without interruption. Sooner or later - it may be and seldom survive for more than a few years. Trees with
seconds or years - growth is limited by internal and annual growth increments are a major exception to this.
external factors. Fortunately - for without them our particular science
Fluctuations in growth are usually irregular and only would not be possible - trees are widely distributed over
seldom do they correspond to fixed astronomic cycles the earth. We must bear in mind that the term 'annual
such as days or years. ring' is not absolutely accurate. It is true that usually only
The organisms and structures illustrated here are all one ring is produced in anyone year; its formation occurs
manifestations of periodic or aperiodic growth. The topic only within the vegetation period, however, which in many
of periodic and aperiodic growth increments in trees is areas lasts for barely four months.
dealt with throughout this book.
E
~
Ol
o
o
z
8' f - - - - - '
h
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.
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.
Reaction time
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
6
Transformation into structure
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
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:
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
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
40
Elevation
Leysin (1350 m)
Martigny (478 m)
precipitation
temper,lIure
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
53° 30>
..- /'
/
I
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
The influence of topography on soils with normal drainage and on tree growth in different climatic zones.
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
10
Influence of geological substrate
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
Every site is the product of its substrate, time and the same the whole year through. In the south in Gondwana-
eternally changing climate. The duration, type and inten- land, wood from the same period shows clear growth
sity of the influencing factors are responsible for the areas - they are perhaps annual rings. Moreover, as in
character of the present-day site. the region of present-day South America, South Africa,
Australia and the Antarctic traces of glaciation have been
Site changes in geological periods discovered, one must assume that the polar regions of
that time had a seasonal climate.
Immense climatic and geological changes have influenced
vegetation and sites fundamentally. Two examples serve (2) Tertiary Period - about 70 to 5 million years ago.
to illustrate this: From the evidence provided by fossilized plant and animal
remains it has been clear for a long time that a change
(1) Palaeozoic Era - about 300 million years ago. from tropical to temperate climate occurred in Europe in
At the end of the Palaeozoic era the present-day con- the Tertiary Period. But only after the discovery of the
tinents formed a large, practically complete landmass. In outer layers of micro-organisms in ocean-boring samples
the charred, silicified tree trunks from the Carboniferous was it possible to plot the temperature course for the
swamps in the north of the continent hardly any annual whole time period.
rings are to be found. The climate may have been the 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
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).
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).
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 nature through over-utilisation and pollution.
fluctuated only slightly. In recent times, climatically caused Plant distributions and species combinations are the
site changes have obviously occurred only in extreme products of all these changes.
areas. Over the last three millennia man has altered the Different species areas have arisen as a result of
landscape by clearing forests and in the last fifty years he differing genetic predispositions of individual species and
has gone a long way towards destroying the balance of different species combinations on ecologically similar
sites.
In mountain regions
In cold periods glacier tongues extended far down into The Rhone glacier (Switzerland) in 1850 (left) and 1970
the valleys, in warmer times they retreated to greater (right). In the course of 120 years, the glacier has sur-
heights. rendered more than 2 km of valley floor.
Herd of cattle in a cliff drawing in the present-day desert. Exposed roots of a 4000-year-old Bristlecone pine. During this time
the soil has settled, the site has changed (Muench and Lambert,
1972).
14 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- If the species areas are very distant from each other
species have arisen from one basic species through and the different species clearly distinguishable, they
isolation, and become more or less widely distributed. were probably isolated at an early stage. If the differences
This explains, for instance, why alpine flora under similar in area and form between two species are only slight,
ecological conditions form plant communities with the separation probably occurred during the Ice Age.
same composition of genera, but with different species.
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
0 Plcea sitchensis
Plcea engelmannii
0 Picea rubens
Straka, 1970. 15
Selection of the sampling site
The selection of sampling sites is a key factor in dendro- from extreme outposts, where the information stored in
climatological research. In order to build up sampling the tree ring pattern does not reflect the normal weather
networks containing similar climatic information, it is conditions. It must also be borne in mind that natural
expedient to choose samples from boundary situations, boundaries can be considerably displaced by human
where growth is limited mainly by one factor, although not activities such as animal grazing, felling and fire.
. Timberlines.
normal beech
16
Annual rings from varying sites within a restricted area of The following a(ethe foremost criteria :
a uniform climatic zone reflect different climatic events.
The delimitation of a site, however, is not easy. In every variability of the site conditions . Which plant societies
case a sound knowledge of ecology and botany is neces- are typical?
sary for the selection of a sampling site on biological evaluation of the probable ecological conditions as
grounds. The selection process is as follows: shown by indicator plants.
identification of possibly suitable tree species (age,
Getting to know the site form , annual ring width , etc.)
Every objective, whether dendroclimatological, ecological
or pathological , requires that the researcher obtain a Deciding on the sampling site:
comprehensive knowledge of the site by familiarizing The following example from a temperate zone illustrates
himself with the whole spectrum of conditions in the field . the selection and delimitation of sites in relation to the
research goals.
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.
F
250
200
150
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).
Meteorological data series from the stand under investiga- measured at the same time as the meteorological factors.
tion provide, without doubt, the best values for compari- Measurements of gas-exchange processes in trees indi-
son with annual ring parameters; tree site and measuring cate physiological activity. Technical problems limit inves-
station are in one and the same place. Only in rare tigations more or less to small trees. Dendrometers can
cases is it possible to continue the measurements over be used to register fluctuations in trunk thickness, which
decades. The physiological reactions of the tree can be form one of the expressions of physiological activity.
I
0
0
u - I
E 0 2 4 6 8 10 12 14 16 18 20 n 24
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- There are two basic ways in which a site can be
ing information on two points ; climatic changes over large described :
areas and weather conditions in particular localities. In description of site integrators i.e. those elements
the end, however, the findings from all regions of the whose behaviour depends on climate.
earth possessing annual-ring-forming trees must be fitted measurement of the main site factors over a given
together like a puzzle and related to each other. For this, number of years.
site descriptions are of the greatest importance, even
if they seem unimportant to the researcher at the time The first method has the advantage that values from
of sampling. Only by comparing the site descriptions is widely differing areas can easily be compared. The
it possible to determine which tree ring-sequences fit second allows direct comparison with tree-ring growth,
together to form a unit. The formulation of regional den- but usually permits only a poor characterization of the site,
droclimatological objectives seems to me possible only by since the biologically decisive extreme years are not
means of good site descriptions. registered. Comparison of measurements from different
regions is, for chronological reasons, simply not possible .
Climatic integrators
. u 40
"\ ~"", ~K"'""~ "I" 80 E
E
~30
8.'" humid
. \ •
/ 60g
c
!!!
0.
E
~
.>:~:-;./.
!!!
20 40 a
.'.
/
• drought
10
.
/ mean mon Iy temperature .........
20
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.
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.
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
Rock type
Weathered or unweathered
Weatherability of the rock
Soil
sy mbols:
20 AI
30
40
C
50
Symbols :
Surface layers:
organic layer with little
decomposition (li tter).
b;AJVVl iron enrichment
111111111
diffuse ---
loam-clay
I -=--1
Duchaufour, 1970; FitzPatrick, 1980; Muckenhausen, 1975; Scheffer and Schachtschnabel, 1970. 23
The worldwide sampling network
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 In the south-west of North America a sampling network
permits the construction of a sampling network with an of trees from dry sites on the lower forest limit was
almost arbitrary amount of climatic information; in practice, studied (Stockton, personal communication). In these
however, this is not the case, since the means of areas the width of tree-ring growth was limited by defi-
research are limited and worldwide dendroclimatological ciency in precipitation. As the trees on the site were very
research has only just begun. Consequently, the objec- old, information was provided about precipitation condi-
tives must be constrained and research is restricted to tions in these arid regions for the last 500 to 700 years.
sites on which the trees are exposed to similar environ- In Europe a sampling network of conifers from cool,
mental factors. One of the research goals for the northern moist sites in summer on northern and subalpine timber-
hemisphere is the reconstruction of summer tempera- lines was studied (Schweingruber, 1985). Cambial activity
tures. Since the maximum densities of conifers at the is limited from several factors which change from year to
northern and subalpine timberline reflect average summer year. Thus the maps express mainly areas of high and low
temperatures, they could provide more or less homo- growth and less climatic information.
geneous climatic information for the last 300-500 years A similar network has been established from Russian
over a wide-meshed sampling network. Such an inves- dendrochronologists over large Russian and Asian regions
tigation would also help to define the areas for which data (Bitvinskas, personal communication).
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).
The stated aim of dendrochronological research is to It is a lucky fluke when all these conditions are fulfilled.
construct chronologies covering hundreds and thousands Up to now it has been possible to build an unbroken
of years. Technically, cross-dating makes this perfectly chronology only with the extremely long-lived Pinus
possible, but the success of this process depends on the /ongaeva and the European oak (Quercus roburlQ.
coincidence of various favourable factors: petraea). The Norway spruce and European larch chronol-
ogies are still floating earlier than 900 A.D.
a chronology can quickly be built up from long-lived
Each of these chronologies has a particular impor-
trees Pinus /ongaeva, Fitzroya cupressoides and Se-
tance:
quoia gigan tea , with ages of over 2000 years are
Pinus /ongaeva served for calibration of the C14
suitable. The wood must be durable by nature or given method.
favourable conditions for preservation . Fitzroya, for Quercus sp. The sequences from central Europe pro-
example, contains heartwood substances which resist vided an outline of river and landscape history in this area.
decomposition in a moist-warm climate. Pinus lon- Currently, they are being used in the dating of prehistoric
gaeva has decomposition-inhibiting resins and grows lake dwellings in the region of the Alps, and also for
in a desert climate in which decomposition scarcely calibration of the radiocarbon method.
takes place. Oak also possesses decomposition-in- Larix decidua, Picea abies. The material comes from
hibiting tannins, and fossil stems are often present in the alpine timber line and is analysed radiodensitometri-
sediments with no aeration; decomposition cannot cally. It affords an insight into the annual weather condi-
occur under these circumstances. tions of the last 8000 years.
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 .
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
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).
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.:
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.
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 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
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.
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- Wood in good condition from dry regions is suitable for
decomposing organisms. The only potential wood-de- radiodensitometric investigations. There are often prob-
stroyers are insects which produce water during the lems with the identification of the terminal ring, i.e. the
metabolic break-down of cellulose (long-horned house last ring formed by the tree. Only under a microscope is it
beetle and other wood-boring beetles). In the arid niches possible to see whether an outermost ring is incomplete
of mesophilic zones e.g. in buildings, a small amount of because biological conditions led to a stoppage of growth
chemical decomposition takes place. In windy arid and or whether the latewood has become mechanically sepa-
arctic areas, wood is eroded by sand and ice-crystal rated from the earlywood.
blasting.
Longitudinal sectibn through a decomposed jOist. ,he cell wall Insect galleries. Damage occurs mainly in the relatively soft early-
has diSintegrated through a physical-chemical process. This WOOd.
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 Hardwood with terminal tree ring (Euonymus europea 250 X)
left: interruption of growth in June. The earlywood cells are left: interruption of growth in June. Three rows of earlywood
clearly formed. vessel elements have already been formed.
right: interruption of growth after the vegetation period. Close right: interruption of growth after the vegetation period. The cells
under the bark, the last-formed tree ring has complete adjacent to the cambium have a tangentially flattened
latewood cells. shape.
Hickin, 1973; Koenig, 1957; Liese (Edit), 1975; Rypa.cek, 1966; Schweingruber, 1978. 33
Decomposition under low-oxygen conditions (anaerobic decomposition)
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 age from lakes in warmer zones, e.g. the Swiss Central
indicates the position of the decomposition zone; in this Plateau, even the heartwood of slender oak trunks is
part only the ring widths can be used for synchronization. entirely decomposed. Decomposed wood can be found in
In trunks from areas with low average temperatures, e.g. flooded coastal forest areas, fresh-water lakes, bogs,
the Arctic and the sub-alpine zones of mountains, only sediments in meander zones of rivers and in alpine
the outer parts are decomposed. In samples of the same 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
0,5
0,4 , I
I I
,
I
,
II
0.3
"'-.' \ '/7 '
dnnual rrllg wid 111
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
35
Changes in fossil wood caused by pressure
If anaerobically decomposed tree trunks in swamps or the turned into splint coal in the second stage of carboniza-
sediment of rivers in glacial areas were covered by ice- tion . In such material tree-ring widths and densities are so
masses and then by gravel during the Quaternary period, altered that there is no question of radio-densitometric
th en their form - both the anatomical and chemical analysis. The material is of great chronological value ,
properties - was irreversibly changed . The wood had however, since it is suitable for radiocarbon dating .
Heavily compressed spruce branch from a Pleistocene gravel bed in the Aare
south of Berne.
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 terial anaerobic decomposition sets in after a certain time.
when it dries, and also partly disintegrates, it must be Artificial drying with prevention of shrinkage is expedient.
preserved artificially. Any dendrochronological investiga- At present there are several preserving methods, but
tions must be made before preservation. For short periods none are problem-free. Polyethylene glycol (PEG) is often
storage in water at relatively low temperatures is suf- used. Drying techniques using alcoholic ether give good
ficient; fungal attack is not to be expected, since the results but unfortunately are slightly dangerous to use.
easily decomposed polysaccharides are no longer pre- Freeze drying in combination with PEG is another good
sent. If the wood is not further preserved, however, bac- possibility.
This 6000 year-old post from a lakeshore settlement in Switzerland A similar post impregnated with synthetic resin (Lyofix) and dried.
shrank greatly and irregularly on drying. The sketched-in outlines The impregnating fluid did not reach the inner parts, so that the
show the original form of the entire post and its cross-section. wood there shrank and crumbled. The shape of the post has not
changed.
Cell structure preserved with synthetic resin (Lyofix). The decom- Freeze-dried wood. The form of the cells remains, but the structure
posed secondary wall is preserved in its loose structure. The pre- has changed. The decomposed, loose secondary wall has collapsed
serving agent has penetrated the cell walls 1000x (Braeker et al., 1000x (Braeker et al., 1979).
1979).
Braker et al., 1979; Coles, 1981; Mathieu, 1961; Morgan et al., 1981. 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.
The anatomy of the tree ring is, among other things, an tree-ring biology, particularly in the investigation of cam-
expression of the physiological processes in the tree. The bial activity. Until recently, however, the microscope has
form and dimensions of the cell walls, the number and been little used for work in dendrochronology. Today
size of the cells and the proportions of the different cell microscopic techniques are being increasingly applied in
types all provide information about the environmental electronic tissue analysis, where the efficient production
conditions that are experienced by the tree. of long sample series is required. In radiodensitometry
Microscopic techniques have been used in the field microscopic preparations provide the basis for the under-
of wood research mainly for differential anatomy and in standing of density curves.
Microsections
The sledge microtome has proved invaluable in the pro- the blade must be optimally sharp: sharpened with a
duction of microsections. Using this equipment it is pos- special knife-sharpener, for example
sible to obtain 10 to 20 micrometer-thick sections of the sample must be positioned in the holding device
superb quality provided that the following conditions are with great care
met: skilled handling of the blade is required.
the wood should be embedded in carbowax (poly- In addition to these requirements a considerable amount
ethylene glycol) of experience is desirable.
a top-quality blade must be used
Sledge microtome with the blade in position. Knife-sharpener: the friction of the blade
against the fine-grained whetstone produces
an optimal cutting edge.
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
In dendroclimatology, the dimensions of the cell walls The procedure is different for biological work on tree
must be accurately measured. There are several ways of rings. Here the cells themselves, their contents and the
staining sections so that such measurements can be degree of lignification are of interest. Pianeze IIIB stain
carried out. can be used. The lignified cellwalls will then become
The first step is to free the contents with Eau de Javelle green, the non-lignified ones colourless and the cell
(KOCI or NaOCI). After this the cell is stained red with contents red .
safran in, which enables the cell walls to be clearly seen.
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
LJ
properly.
Diagram showing a bore-support.
Taking cores from living trees
Coring these days is done almost exclusively with Swed- trunk from a distance of between two and three meters.
ish increment borers. These are available from forestry Where the sampling has to be done by only one person a
suppliers. The 5 mm diameter borers are the most suit- holding device is used . (See photo.) This can be adjusted
able for radiodensitometric work. The double-helix screw- in the axial direction by a peg and screw, and in the radial
thread enables the tree trunk to be easily penetrated . The direction by a joint.
wood outside the borer is pushed aside and only the non- If, after some time, the borer gets blunt or notched , it
compressed core passes into the drill tube, which is can quite easily be sharpened parallel to the inner surface
slighty wider towards the rear. When the desired depth of with a whetstone. Since it is the outermost annual rings
penetration is reached , an extractor spoon is pushed from which are decisive for dating, a bore support is used to
the back of the borer between the core and the inner wall ensure that the borer penetrates the tree trunk in a
until it is tightly wedged between the core and the top of straight line.
the borer. The core is broken off within the tree with an
anti-clockwise movement and then pulled out.
orientation borer bore-support
device
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.
42
The labelling and transport of the cores
' 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
Where fossil trunks are to be examined whole discs are that in the process of clearing up the areas the wood that
generally sawn off. If the trunks are buried in peat bogs was scattered about has been burnt. This wood is often
or moraines they very often have to be freed using centuries old. In the legendary Bristlecone Forests and in
machines. Sampling must be kept to a minimum where the Petrified forests it has become fashionable for tourists
the trunks are in deposits that are in protected areas. to take a piece of wood home with them as a souvenir.
Today the remains of trunks in arid areas are increas- This practice is banned in many of the national parks.
ingly endangered. One of the reasons for this is that the In some cases permission must be obtained from the
use of 4-wheel-drive vehicles has led to the opening up authorities in the locality when it is desired to collect wood
of a number of desert regions to tourism. This has meant 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
Material in good condition The surfaces can best be examined using a stereomicro-
scope with a spotlight or a hand lens. Small details, e.g.
Cutting a plane surface: Transverse radial strips are cut single rows of cells, can be seen only in very good light.
along discs or cores using a sharp blade, e.g. a multiple On field trips the author has often used sun-light for this
use snap-off blade knife, a carpet cutter or industrial purpose.
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 G lUlillUJ • \~
cut at an angle of between 20 and 40 degrees to the
fibres. Here a satisfactory cut can be obtained only by
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.
46
Counting tree rings
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.
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\\
naked eye. Such changes can often be dated by counting the rings
or by using pOinter years.
'-'--'--"--' 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
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)
d,stlnct minimal
Intraannual vaflauons In densltv
~.~~~
, ""
'\ \\\\' "~\'\\\\\""'I'lll A'
lll lIlIllIlIJlIl
'" '"
A A" I ~
A
N·,~ .. ,re"" . . . ~'"".... ~ ' :I ~ . . ~~ ~
~ ~ ~ ~, .. ,............. ~ \ ~"~ ;' ~' ~ m
, .............. '\ 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
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
! 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
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.
. rn
stereo-microscope
sample fUl).)~
. 0 .. • O~
.pmmuawumlit!W~@~~~::;!~mllm@l!!lc14 '--_~_"" .
1----"
sledge :
.,
I .,.
~. ~ /
I simple_ve_r_sl_on_____________~
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
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 Structure analysis of tree rings before
on the age of the cambium (Eckstein et al., and after damage. Acer saccharinum
1977). from the center of Hamburg affected
by the drainage water from road-salt-
ing (Eckstein et al., 1974).
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
,
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).
55
Photometric techniques
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
0.3
0.2 ~
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).
0 .3
.• •
0.2 •
0.1
50 60 70%
I igh t reflex ion
d'
I
I ' \\
1/ \
Removal of resins and heartwood substances
I \\
I
58
Measu rement of density with the densitometer
visual check
Analysing
choice
the film
aper lure
f: ! : :
<==>
..
\
I' ,\ '
\t
'
" ,
lighl-source
, ..-...... .J../ I I \,,j"'......... ~ . . \v.. \_. ~~. .:" /,: ... " .~ ... ' .. '... '
.' ",' " \ , ' \' . : ,,;, .. , : ...\ , ...., , " , •• - .... ~ : • " .. j ~- •
. . ......... , r· ' .. ' .. '~ . . . ,' ........ /',...,/'~. ",' .. "~ . . . .".... .;.,
. , , ' ........ , ....- ........... _ .... ' .......... . ,'.
,~ ",
................................. ............ . .
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 on the
movab le p a
l te
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.
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 lath, which is in turn carefully detached from the support.
placed underneath a twin-bladed circular saw and a lath, The samples are labelled with India ink: this does not
usually 1.25 mm thick, is cut out. The saw blades, which show up on the film. The thickness of the sample is then
are of the sort normally used only for metalwork, cut with measured with a micrometer, this measurement being
great precision (1-2% variation in width). In the case of necessary for the subsequent density calculations. The
well-bored samples only a single cut is necessary while accuracy of the coring and the cutting determines the
for badly-bored ones two or three overlapping cuts may clarity of the X-ray picture: for good samples the tissue
be needed. After sawing, the unwanted wood at the side structure is clearly visible, and for poorly-prepared ones it
of the lath is broken away with a rigid plastic strip. Inci- is blurred.
sions are made with a sharp blade along the base of the
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.
~! '!
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
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 they vary enormously in the rate at which they absorb
directly belong to the tree ring structure itself. Where X-rays: resins absorb relatively little compared to
radiodensitometric work is to be done on the wood these cellulose: water absorbs slightly more; and certain
substances must be removed because: other cell substances display considerably higher
X-ray absorption.
such substances generally enter the tissue after the
ring has been formed. Resins, for instance, are pro- Resins should be extracted in a Soxhlet extractor in
duced continually by the secretion cells of the resin alcohol or acetone. This process takes several hours. The
ducts; other substances appear in the cells during the heartwood substances of a number of species, including
transformation of the sapwood to heartwood, and larix, tsuga and fitzroya, are water-soluble and so can
water is distributed more or less locally throughout the be removed by boiling in water. Extractables may be
tree-trunk. 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.
mi nimum denSI y
O.3 +-------/7,~'-'~'~~----~~----~~~~--_1
Maximum and minimum density curves for conifer wood. The water .,.~ .......
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: Kodak RP/M. This has a density scatter of only 0.007 over
Safety regulations require that X-ray proof walls and doors the whole film surface.
be installed and that there are safety switches which
enable irradiation to take place only when the room is Developing the films:
sealed. Dosimeters must be used to check the amount of Films which are designed for rapid processing take only
radiation to which the operators are exposed. The type of 90 seconds in an automatic processor. The equipment
equipment most commonly used is produced by the that large hospitals have at their disposal is always kept in
Belgian firm, Balteau. good condition and is therefore excellent for the develop-
ment of radiodensitometric films.
The films used:
The most suitable of the many types available is the
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
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
64
Irradiation: The center of the ray geometry is determined using a
The X-ray source is 2.5 m above the film. This means that Geiger counter and the film itself. The duration of irradia-
practically only parallel rays from the center of the X-ray tion depends on the distance between the radiation
tubes penetrate the samples. This ensures that the source and the film: the longer the period of irradiation
boundaries between the early and latewood are repro- the sharper the contrast obtained. For this reason wood
duced without blurring. The samples are in a cellophane samples that are only 1.25 mm thick are irradiated at 11
carrier which is placed directly on the film. 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,
The optical density of the film is measured using a densitometer table , the rolls of paper for the strip-chart
microdensitometer. The film of the radiographed samples recorder and the tape-punch move in time with one
rests on a movable table which passes over a light source another in response to signals transm itted from the con-
with a slit-shaped aperture. The light signals of varying trol mechanism to the stepper motor.
intensity are transformed by a potentiometer into electrical A new, complete radiodensitometric apparatus is pro-
impulses. These in turn activate a strip-chart recorder . A duced by Kutschenreiter, Siccaburgstrasse 64, A-1100
magnetic storage unit registers selected values . The Vienna, Austria.
The equipment
visual chee
Analysing
the 101m
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
chart-reco rder
J
density peak values
Integrator (y)
(y) \ selective
_ _ _ _ _ _ _ _ _ _ _ J. +
I -
__ '---_ _ _I---;:===~--~
reroodlc record; ng
.. .
: :: : ,.:
x·values
COlli ler
(stepPing motor)
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
- .;""'.:
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-
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
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 The percentage of different density levels in three types of wood (5
dividing all the values measured at a distance of 10 microns by the trees for each type). The average percentage of different levels can
number of steps. In this example, the mean density above a be obtained by integrating the density above different thresholds
threshold of a glcc is 0.42 g/cc. The mean density above the (Echols, 1973).
threshold 0.6 glcc is 0.82 glcc (lenz et a/., 1976).
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
70
Measurement of early and latewood widths
For densitometric purposes earlywood and latewood are ring. It tends to lie around the mid-point between the
differentiated in one of two ways: maximum and minimum density.
Delimitation according to a predetermined density Both these methods provide an answer to the problem of
level (fixed threshold). This means that a particular differentiating between early and latewood that is accept-
density threshold is set and that where values exceed able and technically feasible. The second alternative
or fall below this threshold the boundary between the outlined here, that of demarcating in line with a floating
two woods is deemed to occur. threshold, would tend to reflect the biology of the tree
Demarcation according to a density level that is related better.
to the total density amplitude within a tree ring (float- The diagrams and photographs illustrate measurement
ing threshold). Here the threshold varies from ring to along a fixed threshold.
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.
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.
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.
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.
Microfiche reader. Checking the data. Errors that have been located are noted. The information is
then fed back into the computer for correction.
1 1 .... I (
1'1.... t
11 .. '
, ..
I ' ....
~
I I
1 ~'" I
"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.
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
•
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
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 '=
/\/\
.
/
.-.
/~./\.
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
75
Dating in practice
Baillie (1982) has given a full description of dating in The density values for decomposed wood are not
the European archaeological dendrochronology. I shall reliable. This means that only the ring width can be used
attempt here to make only one or two comments on for dating. In certain species, especially the five-needle
dating in practice. Reliable dating requires the mastery of pines (haploxylon) it is practically impossible to date using
the simple methods involved and considerable experi- the maximum density measurements from just one site.
ence and patience. The following should be borne in
mind: Dating
Problems
Short ring sequences are generally less suitable for dating How exactly the synchronization work should be carried
than long ones. The shorter the sequence, the greater the out depends on a number of factors: the importance of
likelihood there is that the dating will be inaccurate. For the sample, the length of the series and the chronological
periods which give a very smooth curve and have very information available. As a first step, poorly-covered
few pointer years, it may not even be possible to date chronologies with no missing rings are cross-dated visu-
series of 120 years or more, while 30 years may suffice ally and mathematically. The resulting mean curves form
for a dating within a period with an irregular, characteristic the basis for further synchronizations. Sections from long
pattern. Missing rings falsify the statistical calculations. sequences can often be aligned mathematically to other
Their location is possible only through visual comparison curves, and the subsequent synchronization of different
with complete chronologies. They are common in conifer mean curves is often possible. As a final step, it often
wood, where the rings are very narrow. Ring-porous proves possible to aggregate single sequences sharing a
species such as oak do not display this phenomenon. common characteristic, e.g. sensitivity, into a single
Smooth curves are less easy to date than sensitive ones. chronology.
Three synchronized ring-width curves for fossilized larch from the The lower curve, which is sensitive, could only be placed because all
subalpine zone. The upper two curves are smooth but nonetheless the pOinter years in the older part are available. The alignment here
exhibit characteristic periods and pOinter years. is not sufficiently good to enable the curves to be dated.
76
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) .
. 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
Dendrochronological research without the aid of statistics curves are constructed from single curves from dif-
and electronic data processing is today unthinkable. When ferent sites. Statistics does not obviate the careful
tree ring analysis first began, relatively few ring se- selection of samples in the field.
quences were available but now there are thousands in a large quantity of data can be brought together and
every laboratory. This accumulation of data necessitated analysed according to a set of clear principles and
the development of more appropriate and more powerful recognised tendencies be quantified.
methods of comparison. The accumulation and interpreta- the results are critically appraised and interpreted from
tion of large amounts of data has also been made practical a biological viewpoint. If the results are statistically
by electronic data processing. The two main tools may be acceptable but not explicable in terms of biology or
described as follows: climatology, they remain doubtful until confirmed for
Statistics: a method for quantitative analysis and inter- other data sets.
pretation or checking of data for the testing of assump-
Details of the methods may be found in standard text-
tions, hypotheses or models. Statistical methods do not
books e.g. Sachs 1978 and Taubenheim 1969. Fritts 1976
produce new qualitative information, but serve to extract
describes the application of these methods to dendro-
from given data the answers to particular scientific ques-
tion (Sachs, 1978). chronology.
It is clear that samples taken by different people with
Electronic data processing (EDP): a method for the
rapid execution of mathematical operations on large quan- differing research interests will be much more variable
than those taken by only one person. It is also widely
tities of data and for the graphical representation of the
accepted that the quality of the samples - and hence of
results.
the results obtained - can be greatly affected by such
Statistical methods can be successfully applied only if:
factors as researcher fatigue, or bad weather.
the statistical process used is suitable for the given These factors must be given due weight when a
data set and properly applied. decision is being made as to whether to commit funds to
a clearly-formulated hypothesis, based on biological carrying out costly calculations. It is also of considerable
considerations, is proposed. importance that the dendrochronologist today be on his
the data are good. The best statistical processes will guard against losing his feeling for the sampling site after
fail if the basic material is heterogeneous, i.e. if mean having taken his samples.
Methods
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.
o o ~
~
>
31I II II I I I+ I 6 ~
o t t t t t t t
I I I ~
OOOOOOCY
annual mean
synchronously placed
2
Wllh Imc·dlsplacement
79
The tree-ring series and the meteorological series are
compared as a basis for the construction of statistical historical or archaeological wood should be compared and
contrasted (left-hand column). Chronologies from living
models linking tree rings and climate
trees may only be extended using material from historical
The first step in the process of climatic reconstruction is ones when the series are at least statistically similar.
the checking of the time-series. The meteorological data Where both the meteorological and the tree-ring series
must be checked for missing values and to ensure that satisfy the requirements outlined above, statistical models
they are homogeneous (right-hand column). Tree-ring may be constructed using response and transfer func-
series that relate to the recent past and those covering tions.
estimate
missing statistics
chronology
and
chronology
and
statistics
L-v_al_ues---'l
statistics
years
combined
chronology
80
Basic statistical methods
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~ '- /'. . '-
/\/. -.-' '.-' \\//-,_.;,
•• ".'
. -.
\./ . .
. ,/\ . 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.
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.
(X i + 1 - x,) ·2 mean S=
Annual Si+1 = -----'----- -'-i-_ 1'-----_ __
82
Gleichlaufigkeit (sign test)
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
Interval trend
for m curves: ti = . I
1-1
t'i/m; i = 1 ... m
83
Correlation, correlation coefficient
r~ + L(Xi - x)(y, - y)
skew distribution
"
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:
~. / "
~ \~ \ '\ \
. "•/
, / ./
® . . . . .'.-.......... \ ./
\
~ ®
-...........:.
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.
10
10 .0
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
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
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
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
a>
.S +0.1
.§. 0
~ - 0 .1 , • 1 r f I 1 .,
10 8 6 4 202 4 6 8 10
years
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.
87
Modelling (response functions)
1900 1950
a) with itself, but in serial correlation (time displacement!. 1899 ~1. . . . . . . . . . . . . . . . . . . .~r~~~1~1---,
This shows how strongly the previous year. which deter· 1898 1- . . . . . . . . . . . . . . . . . . . .~1~ 94~~
~.~---4
mines the physiological condition of the tree, influences
1897 "'I- ............ ---~1'!!' 94'!'!7~1.-------1
growth in the current year. I f .
b) with the monthly mean temperatu res c) with the total monthly precipitation
1900 annual ring series 1950 1900 annual ring series 1950
December 1 1900 f 19 50
1900 1950
January I I
February
August
September
Such calculations allow response function diagrams to be precipitation at one time and very low temperatures at
constructed as a series of monthly coefficients of weights another, such functions only indicate which factors are
which are plotted sequentially in relation to the central generally dominant at particular times of the year .
axis. Values lying significantly above this axis indicate a They do not show which factors limit growth in partic-
positive response, those lying significantly below the axes ular years. It is clear, then, that response functions are
indicate a negative relatioQship between tree-ring size only a useful analytical tool in the cases where there is
and climate for particular months. in effect only one limiting factor, i.e. for sites at the two
Tree-ring series can also be related to other climate- timberline limits, the lower dry limit and the upper
dependent series, for example, the mass balance of cool-moist one.
glaciers, changes in glacier tongue length, wind yield, etc. the results must be consistent with those obtained for
Information obtained using these sort of calculations other sites. No conclusions of general validity can be
will only provide a basis for valid comparisons when the drawn from a response function.
following conditions are met: only factors which are possible should be compared.
There is no point in comparing predicted monthly
there must be sufficient correctly synchronized sam-
mean values with tree-ring sequences.
ples relating to homogeneous sites. Here it should be
the results must be ecologically and physiologically
borne in mind that the concepts 'homogeneous ' and
feasible .
'site' are often used loosely or incorrectly. For exam-
the climatic variables used should not be directly
ple, the term site is sometimes misleadingly used to
related to each other. Where this is unavoidable this
refer to an area which falls into distinct topographical
must be taken into account when the diagrams are
units but which is heterogeneous with respect to
evaluated .
vegetation .
it must be possible to test the assumptions made in
verified metereological data for weather stations which
constructing the model. This means that the curves
are fairly close together must be used.
relating to the supposed limiting factors must be
the limitations of response functions must be taken
shown to correlate with the tree-ring curves (raw or
into account. Where they relate to sites which are
transformed) in the verification procedure.
subject to extreme weather conditions, e.g. very low
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
(a) Maximum densities of spruces from timberline Sites in the Alps. (b) Annual ring widths of pines from the lowlands of northern
The annual precipitation is around 120 cm. Germany. The annual precipitation is around 90 cm.
The densities are compared with the temperature and precipita- The widths are compared with temperature, hours of sunshine
tion series for 7 months of the preceding year and 9 months and precipitation as in (a).
of the current year, and also with themselves, displaced by
between 1 and 3 years.
89
Reconstruction
• - 1,5 mm
• - 0.5 mm
meteorological
- lS oC
series
- 17°C
- 16°C
dependent
series
I I
1940 1950 1970
I
Diagram showing the relationship between tree-ring series, meteorological data and dependent series.
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
91
n
Characterizing the meteorologjcal data
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~ .
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 ~
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
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.
The ability of woody plants to form growth increments or period in Europe and North America, in the Cordaites,
tree rings depends on their position in the plant kingdom . forerunners of the conifers . The ability of a plant to react
None of the Pteridophytes, neither Paleozoic forms to environmental factors varies according to family, genus
e.g. giant horsetails, club mosses, nor living representa- and species; in certain taxa a change in the environment
tives, form tree rings. Within the seed-bearing plants tree strongly influences the form of the tree ring, in others only
rings are absent in monocotyledons and cycads . Gymno- a slight fluctuation results . Many living angiosperms seem
sperms, on the other hand, display growth fluctuations to follow endogenous growth rhythms which are inde-
dictated by climatic cycles. They first appeared around pendent of the climate.
395- 345 million years ago , during the Upper Devonian
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 Even in areas with a seasonal climate some species do
endogenous rhythm; layers of sieve cells, parenchyma not form annual rings in the bark. The structures of the
cells and bark fibres are interspersed. Climatic cycles bark are obscured by secondary changes such as col-
affect this rhythm in different ways; sometimes broad lapse of sieve elements, formation of sclereids or dilation
bands of particular types of cells of uniform size are of rays. The fluctuations in tree-ring width in the bark are
formed at the end of the vegetation period or bands of less than those in the wood according to Holdheide and
larger or smaller cells are formed in the early or late bark. Huber (1952)
p
S
B
S
P
S
B
p
Endogenous rhythm in bark of juniper. Sieve Tree rings in bark of alder. Large sieve Analogous tree rings in wood and bark of
cells (S) and parenchyma cells (P) and alter- tubes are formed in spring, small paren- larch. Only those of the wood vary in width
nate with bark fibres (B) chyma cells in autumn. from year to year.
Top: 60:1 Bottom 300:1 Top: 40:1 Bottom: 60:1 Top: 16:1 Bottom: 8:1
(Holdheide and Huber, 1952). (Holdheide and Huber, 1952). (Holdheide and Huber, 1952).
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
cell formation ~
~ camblum
dlfferenliatlOn
xylem
Cell enlargement
Wood with large cells 250x Wood with small cells 250 x
98
Cell wall growth
Cell wall growth results in thickening of the cell wall. immediately begins to function as a conducting element.
Only the walls of living cells can be thickened; the Latewood cells, on the other hand, live for 2-3 months
presence of plasma and nucleus is infallible proof that a after their formation in temperate regions, and during this
cell is alive. Earlywood tracheids and vessels live only a time the walls are thickened . The cells do not die until the
few days after their formation. Shortly after the completion end of the vegetation period .
of cell division, the cell enlarges rapidly but soon dies and
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
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.
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.
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
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 the tree crown. The variability is so great that it is often
impossible to identify even the species under a micro-
Within every plant there is a great variability in forms, scope from a specimen taken from the root, for example.
because the differing environmental factors affect growth Dendrochronological studies can only succeed if samples
differently. The environmental conditions affecting the are always taken from the same part of the tree, such as
growth of a peripheral root and its functions, for example, the trunk. By taking the average of many individual data
are fundamentally different from those affecting a twig in values from rings formed in the same year, some of the
variations can be eliminated.
Structural variability within individuals. Above a example from the group of ring-porous angiosperm, below a confer. All illustrations 2Sx.
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.
104
Trees and bushes on dry sites in areas which have hot, the region of the pores. Many species have septate fibres.
dry summers have small cells and pores. The pore The form spectrum of species growing in temperate
density however is usually much less. zones includes characteristic features from arid, boreal
In tropical areas trees often have large pores and it is and tropical regions.
not unusual to find large groups of parenchyma cells in
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.
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.
Conifer - Pinus sy/vestris from the Mediterranean, tem- influencing environmental factors and the manifold reac-
perate and boreal climatic zones. tions of the tree, as described below for the Scots pine
The physiological-ecological spectrum of different spe- (Pinus sy/vestris) . All the features indicate the genetic
cies can be established through numerous observations. plasticity of the species i.e. its diversification into many
A clear picture of the suitability of a species as a climato- races and its outstanding ability to adapt to various cli-
logical data storage unit can be obtained by recording the 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-
External characteristics
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
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- The peak density values mainly reflect climatic features
ferent density diagrams. In arid or arctic regions the over large areas; in arctic zones the maxima are low and
density climbs smoothly from earlywood to latewood and the minima high, while in temperate regions the maximum
the maximum lies at the boundary. In temperate zones values are higher. The mean width values chiefly indicate
the density fluctuates and the maximum occurs before the local site conditions,
ring boundary.
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,
107
The reaction of a flora to climate and site
Climate
External characteristics
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
L
XI
F
I
" III IV V VI VII VIII IX X
A, U:rTII'ItO monosC)e,mo O~ I
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
All tissues capable of reacting reflect environmental in- earlywood and latewood, such as vessels and tracheids,
fluences in their behavior. Every anatomical feature of and on the proportion of various cell types such as
every species, or at least every genus, is limited by vessels, parenchyma and ray cells . The importance of
genetic factors. Most reactions to climatic-ecological climatic evaluation will increase with the use of electronic
factors can be seen in the microscopic anatomy of the analysis.
plant. From the beginning, dendrochronology has been In what follows only a few significant anatomical topics,
concerned with total annual ring width; the distinction which can be quickly and unmistakably identified are
between earlywood and latewood widths has seldom discussed: annual ring width, earlywood and latewood
been considered. A few time-consuming investigations widths cell size and thickness of cell walls.
have been carried out on parameters of cell elements in
The width of the annual rings in trees is to some extent instance, than in the diffuse-porous apple. The boun-
genetically determined; a poplar forms wider rings than a daries are marked by the size difference between the
bilberry growing under the same climatic conditions. The earlywood and latewood elements and the number of
structure of the ring boundary is also partly fixed ; it is small, thick-walled, tangentially flattened cells in the late-
always more easily identified in the ring-porous ash, for wood.
J~
....... 1'Vr'"
- 'I"I!
.
i
~- )f'
}j
~
~~
~ ~ II: :\
~ ~
~
~ It~ ~
~
~ ~
" , ~~
110
Species from temperate zones All illustrations 40x
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.
Douglas fir - rather dry site Douglas fir - damp site Apple - damp site Apple - dry site
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
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
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.
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.
115
Intra-annual density fluctations
Density fluctuations contain much climatological informa- Not all tree species have the same ability to form cell
tion but up to the present little has been elicited. Density walls of varying thickness in response to changes in
fluctuation is a change in cell wall density within the early- environment. For instance, the Cupressaceae seem as a
wood or latewood. The number, extent and position within rule to have many fluctuations, Picea fewer. In arctic
the individual ring vary greatly. Only the strongest fluctua- climates practically no intra-annual density fluctuations
tions can be synchronized with each other in different occur; in cool-temperate zones they are very common,
trees or parts of one trunk, since their formation is partly although there are rarely more than two per vegetational
dependent on the ring width, i.e. the vitality of the cam- period. In warm, temperate and tropical areas there are
bium at a given point. several each year.
Cupressus sempervirens
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"
::>
"
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
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
By fixing thresholds at particular density levels it is Electronic techniques allow a differentiation to be made
possible to distinguish between earlywood and latewood. between earlywood and latewood values. But a useful
The question arises as to whether the demarcation should record of intraannual density fluctuations cannot be auto-
be set at one constant value for all annual rings or mated since the intensity of such fluctuations varies
whether a particular fraction of the difference between greatly within one site or even one tree. The operator
maximum and minimum in each individual ring should be himself must decide which peaks are to be recorded in
chosen. In either case the criterion is just as artificial as each case.
that based on anatomical analysis.
~
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.
Little is known so far about densitometric relationships within annual rings. It is necessary to distinguish between
technical-statistical and biological-ecological relationships.
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
When the percentage of latewood is low and that of the stance, calculation of the correlation between two curves,
earlywood correspondingly high, statistics show a formal whose progression is similar but whose absolute levels
relationship between total ring width and percentage of differ, shows no statistical relationship. If sections of
earlywood and latewood. The selection of a process by curves are compared, statistical processes may indicate
which relationships are determined is important. For in- spurious relationships.
Scots pine, Pinus sylvestris with measuring Black locust, Robinia pseudoacacia. Early- Black locust, Robinia pseudoacacia. The
slit of the densitometer drawn in. The width wood constitutes four fifths of the annual earlywood constitutes only a quarter, the
of the ring is roughly the same as that of the ring. Statistical relationships can be shown latewood 3/4 of the ring. There is no rela-
slit. The density profile cannot be correctly between widths of eariywood and the whole tionship between earlywood and total ring
determined under these conditions. ring. width but there may be one between late-
wood and ring width.
glee
I')
1.0
0.8
c
g! 0.6
<1l
'C
0.4
0.2
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.
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
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
..
••
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
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
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
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 Width
Minimum density decreases slightly in the early years and Tree-ring width sometimes increases rapidly in the first
then remains practically constant. Maximum density in- years of a tree's life and then decreases more or less
creases on average during the first two decades and rapidly over subsequent decades, the main change oc-
decreases slightly thereafter; often hardly any trend is curring in earlywood rather than latewood. The age trend
noticeable. 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
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).
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
.. .. ...
~
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.
.
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.
The range of forms that density profiles from different trees is particularly irregular, and where these are sub-
trees in different years may take is extremely wide . Within jected to movement (wind , transplanting) during the vege-
a range of amplitudes the density profile may exhibit no tation period, the density fluctuations can scarcely be
fluctuations at all or the fluctuation may be so great as to distinguished from annual rings . Synchronization is im-
result in the formation of a false ring. possible in such a case.
The tree ring structure of the terminal shoots of young
g/ ccm
0.5 1912
0.4
~
.~ 0.3
Q)
"0
0.2
0.1 + - - - - -
124
B. Between different species
The genetically determined characteristics of the tree and information . The differences between particular species
its ability to respond to site factors vary from species to can be seen in the following features of the ring : the
species. This means that trees of different species grow- absolute values, the pattern of the width and density
ing on the same site react quite differently to particular curves, and the number of fluctuations in density per year
factors. As a result, their rings do not provide the same and per site.
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).
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
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
The grouping of similar sites together is straightforward towards establishing the different causal relations that are
using floristic and phytosociological criteria. It is, on the at work here is to determine under which sets of climatic
other hand, difficult to relate a site to particular site or ecological conditions the cambium and the other living
factors, and even more difficult to account for the fact that cells behave similarly. The reaction of these cells can be
certain species are found on a particular site or that the quantified through the measurement of cell-wall width and
cambium and the other living cells react in a particular thickness.
way to a particular combination of factors. The first step
The mean agreement between curves is used to establish particular period. The similarity expressed relates more to
which areas or groups of sites yield historical or fossil the types of reaction exhibited by the trees on the sites
wood which can be grouped together for the purposes than to the botanical compOSition of the sites. Geogra-
of dendrochronological analysis. The similarities between phic-climatic boundaries are also made clear. A number of
different sites are given by the Gleichlaufigkeit, which examples which relate to the subalpine zone of the Alps
expresses the correlation between mean curves over a are given below.
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
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 seem to be most clearly reflected in the maximum den-
cases the similarity is greatest for the maximum density, sity. It is not clear why practically all the features for the
less for the latewood width and the percentage of late- Rigi site display good agreement with those of other sites,
wood, and least for the minimum density, the earlywood while for the Chasseral site only the maximum density
width and the total ring width. The east-west boundaries 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.
'"
Plynwald 0 0 l"- • 0
•
o 0
Krauchthal 0 . f'.. f'.. • 0
f'..
'"
Oierslordt 0
Gleichlaufigkeit values for the period 1899 to 1975 for pines on dry
sites in Switzerland and Germany (Schweingruber et al., 1979).
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::~
.
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 ~ •
+
.
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.
As can be seen in the diagram, similarities, expressed in about a 10% higher Gleichlaufigkeit value than the tree
percentages of agreement (Gleichlaufigkeit percentages), ring widths . The statistically significant areas of similarity
vary considerably from one site to another. Analysis of the are greater for maximum density than for tree ring widths .
site homogenous sampling network in Europe shows that With both measurements the area of similarity drops as
in certain areas the curves are very similar. These groups one goes from north to south. In northern Scandinavia the
make up the core zones. From this point on there are chronologies of maximum density can be synchronized at
fewer and fewer similarities. When comparing maximum about 1400 km , those for tree ring width within 800 km . In
density and annual ring width of various species of coni- southern Italy, however, the respective values are only
fers on European sites with cool , damp summers, it can about 700 km and 200 km.
be shown that the maximum densities on average have
?:
,..'
.;;;
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'-
... .........
+ + 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 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 Agreement between mean curves of maximum density for four
cores taken from the same tree. Phases of good agreement and widely-scattered sites in Central Scandinavia. The Samples were
poor agreement alternate with each other. taken from spruces growing on similar sites near the timberline
between the Atlantic (1) and the Gulf of Bothnia (4). As with the
The samples were taken from two spruces on a uniform timberline single curves there are phases of good agreement and poor agree-
ment.
129
Agreement over longer periods of time
Maximum densities
In order that climatic patterns which span longer periods hemisphere. It displays a far greater regional spread
of time may be represented, the annual value chronol- than all previous ones since 1700.
ogies must be smoothed. In this case 13 year weighted both the temperature and the density series for neigh-
running means are used. Climatic variations can be boring sites differ from one another. Since such varia-
related to each other using samples from similar sites but tions often cannot be satisfactorily explained, it is pos-
different geographical regions. sible to effect a reconstruction of climatic conditions
only where the information available is such that the
From the chronologies given below it can be seen that: characteristic features of the region can be identified.
all the curves display the same characteristics. They there are clear correlations between variations in the
are all relatively uneven, sometimes with high ampli- summer temperatures (mean values July to Septem-
tude, sometimes with low; in places displaying low ber) and variations in the maximum density.
values and in other places high ones. The frequency most of the decennial variations are, like the annual
over the last 1000 years differs considerably. Short- ones, limited to particular regions.
term fluctuations tend to last about 10 years, medium- It is reasonable to suppose that measurements for the
term ones about 50 and long-term ones about 200. current period will show the occurrence of a larger de-
the change of pattern that is perceptible at the present crease in density than would be occasioned solely by
time is affecting practically the whole of the northern 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!
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-
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).
80
. . . . '.
.' t'
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 Gleichliiufigkeit values for oak chronologies for individual sites and
the Gleichliiufigkeit for ring width curves for oaks from a number of for the aggregated West European chronology (31 chronologies
sites in Central Europe over the last 2,700 years. The mean similarity covering 1850 to 1959). In 62 out of 100 cases tile climate appears
between the curves (Gleichlaufigkeit 50%) decreases exponentially to influence growth similarly in the tree as a whole. There is no
with increasing distance (Hollstein, 1980). perceptible relationship to distance (Baillie, 1982).
Tree-ring features
jrnaxlmum density
I density variation
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'.
biological with biological meas- Climatic-ecological regions of validity Causes of changes in growth
urements, microsections and X-ray for sets of similar ring sequences can behaviour can only be conjectured.
films of annual ring sequences. 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
133
Different tree-ring features in relation to climate
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)
. 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
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- where there is a good supply of water available cam-
tions seem to be a response to temperature changes. bial activity is not limited. As a result there are un-
Fluctuations in density occur most commonly in the differentiated cells present in the trunk at all stages of
wide annual rings of trees on humid, fairly warm sites the vegetation period. This in turn means that short-
which have average July/August temperatures of 14°C to term climatic events can exercise a considerable
21°C. They are not found in trees at high elevations where influence on the density of the wide annual rings.
the average temperature for these months is less than
If we look at the fact that trees on warm sites exhibit quite
12°C.
different reactions to those on cold sites it seems clear
On the basis of observations made I think that it is
reasonable to say that the development of density fluctua- that:
tions for trees in temperate zones may be traced as trees on warm sites can at any time mobilize available
follows: 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.
low )emperalUre
~ __ ~~ _____________ 40
16 ~
~
--~~r---~~~~~~~~---,~--30~
o
40 50 60 1970
135
The effect of climate and site on tree rings
As we made clear in the previous chapter, practically and low mean sensitivity and showed poor correlation
every tree-ring series can be used to obtain climatic with each other. Chronologies from trees in the lower
information . It is nonetheless clear that, particularly in the forest border areas (1900 m) showed more narrow tree
case of alpine zones and their different topological areas, rings and higher mean sensitivity and correlated well with
chronologies can display less and less similarity with each each other.
other as differences in site altitude increases. Fritts et at. LaMarche (1974) developed on this idea and recon-
(1965) compared relatively long chronologies from vari- structed the decennial temperature and precipitation of
ous dry sites in the north of Arizona. Chronologies from the last 1000 years on the basis of chronologies from
higher, moister sites included tree rings of average width Bristlecone pines (Pinus longaeva).
Datable series
---+
, .. If J
I', nQft!IC d
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).
)
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
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 the year 1923. In this year there was a large increase in the
and for the Front Range in Colorado. Climate conditions shown maximum density for mOist sites at high altitudes, but only a
are based on meteorological data. small increase in the annual ring width. Dendrochronological
diagrams for maximum density and ring width are similar only for
Different climatic conditions can have quite different effects on extreme years. (Kienast, 1985).
maximum density and ring width. This is in turn reflected in the
diagrams relating to these features. A good example of this is
137
The different tree-ring features (maximum densities and several times in influencing the results of maximum
tree-ring width) of a chronology call display quite different density analysis. Only the relatively moist timberline sites
climatic reactions for the same year. It is only in years in are thought to display consistent response to tempera-
which extreme climatic conditions are experienced that ture. Most of the low elevation - sites in the temperate
each of the ring features is influenced in a similar fashion. zones may be affected at one time by temperature and
Maximum density may for instance be limited by low another by precipitation. Samples from lower timberline
temperatures while the width of the same ring is limited are mostly sensitive to dryness.
by lack of rain. It is also often difficult to obtain a clear Thus it is seen that comparison of results that are
picture of the factors which influence growth in the case obtained from sites that are uniformly high or low should
of those climatic zones where the summer dry season be attempted.
lasts several month because growth may start and stop
••
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
."
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
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
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).
From the results of the evaluation analysis (weighting information about the summer of the current year, it
tables) for response functions the following conclusions reflects the precipitation levels in winter and spring,
can be drawn : Maximum amounts of climatic information although the influence of this factor varies from site to
obtainable from maximum density series comes from site. The influence of the previous year is somewhat
samples obtained from cold-moist zones, i.e. the sub- higher than with maximum density series.
alpine/ alpine and subarctic/ boreal zones. In zones with The minimum density usually provides information only
adequate summer rainfall it is August and September on temperatures in late winter and in spring for dry sites.
temperature that is reflected in the grown patterns, while Using information from 33 radiodensitometrically ana-
for those which experience a summer dry period (e.g. lysed chronologies for various conifer species in different
Mediterranean areas) it is the temperature in early sum- climatic zones, Kienast (1985) assessed the suitability of
mer which is decisive. In most cases it is possible to draw various sites for the reconstruction of temperature and
some conclusions - valid only for particular sites - about precipitation data. He concluded that: "It is generally true
precipitation in winter, spring and early summer months. that annual ring width is not a useful tool for the recon-
In May and June (immediately before the period of struction of temperature. This parameter does, however,
greatest climatic sensitivity) climate seems to have little or enable a wide variety of sites to be covered on the
no effect on cell growth . The influence of the previous analysis of precipitation . Maximum density values are
years tends to be minimal. well-suited to the reconstruction of temperature and also
Annual ring width series generally contains limited cover a wide spectrum of sites. These data also lend
Valais Cyprus Aggregate diagram
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
~
::: £
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
suitability
139
A model of the growth of a tree
,
I
;
I
(
,
I
:::
I ~
::: I
I
:;.
annual rings 4
, ,
,
/
The cell-wall acts as the information store for these internal and external stimuli
140
Structure is the manifestation of all metabolic processes
,
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
141
IV Applied Dendrochronology
Communion cup in ash from the parish of Beggingen, Schaffhausen, Switzerland. Photograph - W. Schoch
The fact that the irregular sequence of climatic events is activities, e.g. building, population growth and migrations
reflected in the annual rings of trees is of great impor- to be traced. When the climatic conditions of the past
tance for the historical sciences, since it allows wooden have been established, it will be possible to relate human
objects to be dated to within a year. With the help of life styles to environmental conditions over long periods.
relative chronologies constructed on a typological basis, The various types of chronologies allow relationships
wood finds can be used to establish an exact chrono- covering millennia to be established for Central Europe,
logical sequence. This in turn allows the course of human as shown on the chart below.
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. >, .....
(/) .§:
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 Out of the hundreds of investigations carried out on
yet. Completing the chronologies and the decision on the prehistoric and historic objects from Europe, America and
main correlations between the time scale and the relative the Near East, I have singled out only a few informative
chronologies will require the work of a whole generation examples which illustrate the problems and opportunities
of research scientists. this kind of analysis gives. Many other examples are given
Tree ring analysis began in Central Europe in the by Hollstein 1980. All the issues of 'Tree Ring Bulletin'
1940s and only now, forty years later, is it beginning to give information on studies in North America.
bear fruit.
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.
As soon as a living tree is felled, it begins to dry out, the annual ring is completed within a few months, tropical
cells of the cambium die and wood production stops, so trees excepted. In hilly regions 95% of all cells are
that the last-formed or outermost tree ring gives the year formed in th ree months, at higher altitudes in two. The
of felling. The time of year when the tree is felled can be structure of the last-formed ring shows only whether the
only roughly determined, since the formation of the whole tree was felled in summer or autumn-winter.
I
earlywood: April latewood : May to September
As the outermost ring is often missing in wood of geo- thoroughly investigated for oaks.
logical interest and that used in buildings, other features When the sapwood boundary can be identified , a
are used to determine an approximate date of felling of certain number of years are added to the figure obtained
the tree. One feature which has proved useful is the from the last heartwood ring to get the approximate felling
sapwood-heartwood boundary. The relationship between date. This is 20 ± 5 in Germany and Switzerland, 30 ± 5 in
the age and diameter of a tree in terms of width of England and Ireland (Hughes et aI. , 1981).
sapwood and number of annual rings has only been
~
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 Density curve of larch from the sapwood heartwood transition
building. The sapwood zone has been destroyed to a large extent zone. The components of the heartwood absorb X-rays and
by insects (Baillie, 1982). 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 dates when the wood was felled and used for building
more valuable is the historical information to be obtained. are often identical. This is confirmed by written
It is important to know, therefore, how long a piece of records and construction dates on the buildings them-
timber was stored before being used for building. This selves. Sometimes there were several years between
varies considerably from case to case. In those areas of felling and use. Trunks with different felling dates have
Europe with large oak forests, wood seems to have been been found within the same building e.g. in log cabins
used immediately after felling. This is indicated in the and timbered houses. The most recent felling date
following ways: shows that some of the wood had been stored for up
to six years.
Signs of the wood having been worked. Prehistoric Between the 15th and 17th centuries European oak
stone and metal axes, and hatchets and mortise axes panelling seems to have been stored for 3-10 years
of bygone days could be used effectively only on fresh on average, according to Bauch et al., 1978.
wood. The clearly visible wood structure on cut sur-
faces and marks left by tools indicate that the wood In arid zones in the south-west of North America the
was used in a fresh, unseasoned state. Tsegi Indians stored timber for years before using it for
Rhomboid shrinkage of trunks split into quarters, building.
absence of radial cracks, gaps between mortar and These dissimilar conclusions show the need for dating
wood, all indicate that the wood was used fresh. a great many objects on sites, if an accurate estimate of
In the few studies carried out in the subalpine belt, the the date of building is to be obtained.
147
Re-use of old timber
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.
148
The history of house-building in the Neolithic period in Switzerland
building activity
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
CL_5km
----~--~--------~~----~~~
Prehistoric lakeshore settlements on the lake of Zurich, Pfiiffikon
and Greifensee (eastern Switzerland) (Ruoff, 1981).
Relative chronologies
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.
150
Absolute chronologies
Great progress has been made in archaeologically based between the early and late Bronze Age and eight hundred
dendrochronological research in central Europe in the years between the early Bronze Age and the Neolithic
past ten years. Laboratories in Hemmenhofen (F.R.G.), Age the land near lakes and bogs was uninhabited. In
ZUrich, NeuchiHel and Moudon (Switzerland) have dated some cases the change from populated to unpopulated
wood from the pile dwellings of central European peat occurred much more suddenly than had been suspected.
bogs and lakes. The tree-ring sequences of thousands If one considers that at the present time, a hundred
of oak posts from Upper Swabia, Lake Constance, lakes years after a settlement in an area of broad-leaved trees
in the eastern part of Switzerland e.g. Lake ZUrich, lakes is abandoned, all traces of human activity have been
in the west of Switzerland e.g. Lake Neuchatel, Lake destroyed by the advancing forest, then it seems likely
Geneva and Lake Annecy could be fitted together to form that the prehistoric folk had to cut down trees when they
local chronologies. The absolute dating of archaeological re-built a settlement on a lakeshore near a forest. The first
samples was only made possible however by geologically non-nomadic people, Neolithic man, were hardly in a
dating oak finds from the gravel of rivers and peat bogs. position to destroy large areas of forest, since the number
The synchronization of the series and the absolute dating of these people, living extremely simple lives, was very
which occurred simultaneously became possible after B. small. As U. Ruoff in Becker et a/., 1985, showed, even at
Becker (1983) in southern Germany and J. R. Pilcher the peak of the Neolithic Age there were scarcely more
(1984) in Ireland, working quite independently, produced than 20 settlements on Lake ZUrich. In total there may
chronologies spanning more than thousands of years. have been about 100 people living on the shores of Lake
There was now no difficulty in linking the central European ZUrich in 2700 B.C., the period of coil pots, which is a
relative archaeological chronologies with the absolute thousandth part of today's figure. It is therefore not
geological series from southern Germany. 130 years after surprising that in the pollen profiles of neighboring
the discovery of the first pile dwellings in Meilen, Rohren- regions, the cultural influence is no longer recognizable.
hab on Lake ZUrich, posts from nearly 200 prehistoric Some questions remain unanswered. Why were the
central European villages were dated accurately to the periods of settlement interrupted? Were the comings and
year. A time comparison between all villages studied up to goings caused by rises in water level in the lake, by
present shows clearly that many gaps exist in our knowl- fighting or by outbreaks of disease?
edge of the prehistoric period. For four hundred years
1 r
lj
:"'
~ B~
'. ~
Lacs in western ~ ~ b
Swi tzerland ? ~ ~ ~; ~ ~
848
~ '§
?
'" ,
Lacs in eastern
Switzerland
- ~I~. ~ ~
~
~
; 911 BC
9~
1
~
-
Lac of Constance
850 B
J
~ '"
~
upper Swabia ~ ~
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).
Little is known about this. The only key to such questions occupation coincided with the so-called Gbschener cold
as the interruption of settlement occupation lies in the phase around 1300 B.C. Early Bronze Age settlements,
chronologies of high-lying regions in the Alps, which can however, were also deserted during normal or warm
be interpreted climatically. As far as the lakeshore settle- periods. The question of the relationship between climate
ments in eastern Switzerland are concerned, the climatic and settlement requires much work by archaeologists and
chronology shows that the 500-year period of non- dendrochronologists.
Becker et aI., 1985; Suter and Schifferdecker, 1986; Ruoff and Rychner,
1986. 151
Prehistoric settlements in the south-west of North America
No area has been more thoroughly investigated den- provides a chronological framework for archaeological and
drochronologically than the south-west of North America. biological finds. In the past 60 years workers from the
The 'Tree Ring Bulletin' has many articles on this subject. Tree Ring Laboratory in Tucson, Arizona have dated some
Bannister and Robinson, 1975, summarized the state of 20000 beams from about 1000 settlements. This work is
research and Ceram , 1972, described in an enthralling enabling archaeologists to fit together extensive historical
manner some episodes in the history of dendrochro- connections.
nology. See also page 258. In spite of lack of written It has been shown that, rather than investigating many
records , parts of the history of the indigenous, mainly pre- areas in detail, inter-disciplinary collaboration between
Columbian , population can be reconstructed in detail. In archaeologists, palaeo-botanists and dendrochronologists
the settlements - very often cliff dwellings - much plant studying settlements such as Betakin and Kiet Siel in
material is to be found , preserved by the desert climate. northern Arizona and Chetro Ketl in New Mexico (Dean
Seeds, wood and charcoal all provide information on the 1969) can add greatly to our understanding . Douglas,
diet and environment of the Indian folk while dendro- 1935, Bannister 1965, Robinson 1976 and Dean 1969 and
chronological investigations on beams with annual rings 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
piaster
juniper bark
~r:::::;;'-~iaI""f====--- piaster
:llii5ii_~~~~~:§ii;:-- juniper bark
IiO """'____ willow rods
primaries
Roof construction of a mud house in Betatakin, Arizona (Dean, Roof construction of a house in the settlement in Chetro Ketl, New
1969). 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
152 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
~
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
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'
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
",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
~
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 analysis which can furnish a full and precise account of
interested in drawing up an accurate time sequence for the history of individual buildings.
the construction and development of the buildings being Two examples of dendrochronological investigations of
studied. It is very seldom that accurate dates can be given historic buildings are discussed below. Hollstein (1980)
solely on the basis of the features of the building itself. has studied one of the most important ecclesiastical build-
Nor can the dates on the building or historical material ings in Germany, Trier Cathedral , and Stahle (1979) was
provide a complete picture, even where important build- able to date one of the oldest buildings in Arkansas , the
ings are concerned. It is really only dendrochronological Wolf House.
Through the dating of embedded supporting timbers, walls , even though the annual ring sequences were rather
floor-boards , chests, etc., Hollstein established that this short. The available written records were checked against
building was begun in the fourth century, and that building the dates produced dendrochronologically, and were
continued more or less continuously into the eighteenth shown to be accurate. It was clear that the west tower was
century. Hollstein's reconstruction of the building of the begun by Archbishop Poppo and completed by his suc-
west front is extremely impressive. It was possible to date cessor, Bishop Eberhard, and that the west front was
the oak supporting timbers, which were embedded in the 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).
1828·'-
1826'
182" -
1827
rH?4
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- twice in the fourteenth century and that from the seven-
many has been the subject of intensive dendrochrono- teenth century onwards it was rebuilt or altered about
logical examination . In conjunction with architectural his- twenty times. Conifer wood was used for most of these
torians Eckstein has succeeded in reconstructing the alterations : this indicates that oak had become scarce in
development of building methods and of building styles in East Holstein by this time.
this area of Europe. The Holy Ghost Hospital in Lubeck An analysis, using dendrochronological methods, of
has probably been the subject of the most intensive study thirty-two farmhouses and farm outbuildings has enabled
of this type. On the basis of the analysis of a considerable researchers to form a detailed picture of many aspects
number of samples of oak - 769 in total - it was of the history and development of rural life. Changes in
established that all the wood used for the ceilings and in the layout and organization of farms can be traced by
the construction of the roof came from trees felled in the establishi ng when different types of build ing were first
four years 1284, 1286, 1287 and 1289, and that the wood built and when they were altered or renovated . Such
was used as timber in each case the year after the tree changes mirror in turn the economic development of a
was felled . An examination of the annual ring patterns whole area. In addition to providing this sort of general
showed that the wood came from forests in East Holstein. information dendrochronological dates, whether exact to
The fact that in such a short space of time hundreds of the year or correct within a certain tolerance, are useful
oaks were felled, chopped up for timber and then used in tools in the reconstruction of the history and development
the construction of a building in accordance with detailed of individual houses , and provide invaluable material with
plans is a measure of the rigid social organization that which historical documents can be checked and comple-
must have existed at that time. The dendrochronological mented.
examination revealed that the building was extended
SCHLESW'" -
HOLSTEIN
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 A country house built in 1598; from a lithograph by W. Heuer, 1885
1663 (Eckstein et al., 1977). (Eckstein, 1977).
158
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
, . ---.
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
The old mill in Lauenen has a date on it of 1756. The style of the
The village of Lauenen (1250 m) in the Bernese Oberland. All the building, with its gently sloping roof, is characteristic of this region,
houses are built of spruce, which grows in this area up to the while the decorative carving and the paintings on the outside of the
timber-line. building are typical of this period.
:']
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.
The art historian seeks to date a work and to establish its de:libN,UC fa e
O .i /
artl5l'SStudio copy painted I
by. puPil ,
I ~/
panlalfake
'
I lsogned)
second vefslon
/'
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 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
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 The panels supplied for paintings generally comprised
wood. The type of wood used depended on where the radial sections of trunks, as these were more stable and
painter was active. less liable to shrinkage than other types of panels. This
means that there is an optimum number of rings available
The Atlantic coast from Denmark to Portugal: mainly oak. for the dating procedure. As a rule the sapwood was
Central Europe: mainly lime, oak, fir and spruce. partially or wholly removed by those making the panels,
Southern France : mainly poplar and walnut. as this wood was particularly liable to attack by insects.
Italy: mainly poplar. Panels of poplar cannot be used for dating; those from
Spain: mainly poplar and Scots pine. lime have only a limited value.
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 Determining the felling date of oak used as the basis for paintings.
Dutch painters dating from 1400 to 1900. The paintings have been Part of the tree from which the wood originated. Case a: bark still
dated either by art historians or were dated by the artist (Bauch et al., existing; case b: part of the splintwood still existing; case c: the
1978). 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).
The type and condition of the wood can severely restrict only a few annual rings could have come from anyone of
its value to the dendrochronologist. Where the last- a number of areas within the trunk. In such cases (d, e)
formed rings are missing or the length of time that the there may well be disparities between the earliest felling
wood was stored before being used is not known, it is date provided by the dendrochronologist and the date
very rarely possible to date the wood to the year. Oak suggested by art historians."
panels are very often difficult to date. Since wood was stored for varying lengths of time
Klein (1980) described the sorts of problems that may before being used, the dates given for sixteenth and
be encountered as follows: "Where the bark still exists seventeenth century works are only accurate to within
(see case a) the year in which the tree was felled can be between three and five years.
determined exactly. Where the bark is missing but the It is possible to obtain more information about the
sapwood can be seen (as in case b), it is possible to probable source of the wood used in works of art by
ascertain in which year the tree was felled by allowing for examining the wood for growth patterns that may be
the missing sapwood rings, i.e. 20 years, depending on characteristic of a particular region. Eckstein (1983) has
the age of the tree, will be added on." (See page 146). provided regional chronologies covering the Atlantic coast
"Where the samples under investigation consist entirely areas.
of heartwood, as is often the case with painted panels and Chronologies for beech are available only for the
wood carvings, the earliest possible felling date of the northern part of Central Europe (Germany). It has none-
wood can be given, provided that at least fifteen sapwood theless proved possible to date the works of very many of
rings are still in existence. Pieces of wood which display the great painters of Europe.
163
Dating individual works
'Child Playing with a Bird', B 763, Three-piece panel ~ fibre direction in the wood. (Photograph: Art Gallery Berlin-Dahlem)
(Bauch et al., 1978).
An impressive wall-panel, painted on pieces of spruce painting by some time . Radiodensitometric analysis of six
wood , was found recently during the renovation of a samples from the spruce wood enabled a 62 year annual
house in the village of Gais, in the north-east of Switzer- ring mean curve to be constructed. This curve matched
land . Art historians were able to identify the 3 m by 5 m that which had been constructed for spruces in Lauenen
panel as sixteenth century. Dating of a layer of dust found and indicated that the trees had been felled in 1497.
under the paint showed that the house predated the
Wall panel in Gais, Appenzell, Switzerland. The cow-herd encourages the cows into the stall with salt (Schweingruber et al., 1979).
The dendrochronological analysis of all the paintings Did certain works that are now in different places
produced by an individual painter or by a school can previously belong together: were they, for example,
prove of enormous value in checking or complementing originally part of a triptych?
the findings of art historians. The following studies of this Were the wooden panels on which the works of art are
type have been carried out: Rubens (Bauch et aI., 1978), painted produced near to where the artist was active,
Tudor portraits (Fletcher, 1978), Rogier van der Weyden or did the artist have them brought from further afield?
(Klein , 1980) and Rembrandt (Bauch and Eckstein , 1981). This question can often be answered by examining the
Such studies aim to shed light on the following ques- existing regional chronologies for oak.
tion : How far is it possible to date sculptures, or at least to
say in which order they were produced?
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?
Forensic scientists today have at their disposal the most analysis of wood can play in criminal investigation. A
up-to-date methods and equipment. It goes practically number of examples of the successful application of
without saying that forensic units of police forces will dendrochronological analysis in police work are given
include specialists who are aware of the role that the below.
The most important aspect of forensic work on wood is A man was accused for cutting down some valuable
the determination, by means of an analysis of the micro- osiers which belonged to someone else. Shortly after
scopic structure of the wood, of the species or type the alleged theft all the wood shavings found on the
involved. Such information may lead to a suspects being suspect's work-bench were submitted to microscopic
cleared or incriminated. examination. As this analysis showed that he had been
What can be identified? working exclusively with conifer and fruit-tree wood,
he was cleared of suspicion.
Pieces of non-carbonised wood or charcoal, even A large museum was about to purchase an early
those no larger than wood shavings. Gothic crucifix from southern Germany. Previous work
Fungal hyphae in wood: such traces say something done on such figures had shown that they were made
about the condition of the wood. of poplar. However, an examination of this crucifix
The way in which the wood has decayed or disinte- revealed that the arms and fingers were of non-con-
grated and damage by insects: these features provide temporaneous wood, in this case box and pear, and
information about the conditions under which the that only the body itself was of poplar. When an X-ray
wood was stored. of the figure showed that one of the nails was modern,
the museum decided not to buy the crucifix.
Examples:
In some cases the only information that is needed is would be possible to determine whether the death had
whether or not something was done before or after a tree occurred recently enough for it to be appropriate for
was planted or began to germinate. Here a straightforward the police to investigate.
count of the number of annual rings at the base of the a piece of land which was no longer cultivated and was
trunk will provide the information. covered with bushes was put up for sale. The date that
Examples the land had last been worked was important for the
negotiations. A count of the annual rings in dwarf
during building work on a site which had been covered shrubs (calluna) which were growing on the land
with trees a skeleton with a damaged skull was dis- showed that none were more than eight years old. It
covered. It was found under the root zone of the trees, was clear that the land had not been cultivated for
which meant that by ascertaining the age of the trees it about ten years.
Dendrochronology
There are many ways in which dendrochronological tech- Wood carvings and other objects made of wood often
niques can be applied to crimes, including possible fetch high prices, and the possibility that the piece is not
frauds. These techniques, which may involve the mea- genuine can never be ruled out. Generally, the main thing
surement of tree-ring widths, or tissue and density analy- that has to be established is the age of the work of art.
sis, can be applied to cases as diverse as murder and art Ring width and density are used to determine this. There
frauds. Using dendrochronological techniques it is usually are two ways in which this may be done:
possible to produce dates that are exact to the year, and
where a tree was felled during the vegetation period the in relative dating a mean curve is constructed using
dating can be even more precise. Should the place of material from a number of works of art whose authen-
origin of a tree be known it is possible to decide at which ticity is undisputed. The piece under investigation is
time of the year the tree was felled - early summer, mid- then compared to this.
summer, (sometimes) early autumn, and autumn or winter where standard curves are available for the same sort
- on the basis of information gleaned from the annual of wood from the same region, absolute dating is
rings. possible.
«({«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.).
Xray picture of the bridge of the second of the suspec t St-;;;cii~;a-ril~s violins .
.;cl"
, \.:.1\ ......
'".
\
••
'.\
viol In
1894
violin 2
~
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.
Climatologists have for some time been aware that an species and from a range of sampling sites, shorter
understanding of the climatic conditions over the last few meteorological records from various parts of the world
decades -. or even over the last 200 years - by no (including the less populated areas) can be extended.
means provides a complete picture of the wide spectrum
of variations occurring over the last 10000 years. Nor can 2. Lending perspective to recent meteorological data:
weather data for these relatively short periods of time
provide an adequate basis for an understanding of the Dendrociimatological findings provide a background
global weather machine (Hecht et aI., 1979). against which recent data can be evaluated by com-
It was as recently as twenty years ago that economists parison with historic and pre-historic times. This in turn
and politicians became aware how closely the fate of encourages us to question the notion that events in
mankind was linked to the global climate. Neither the our own times provide any sort of absolute norms.
'green revolution', i.e. the breeding of more productive Chronologies which relate to the distant past can, for
crop strains, nor the manufacture of fertilizers in massive example, illustrate how quickly changes in climatic
quantities, nor improvements in transportation enable us conditions may occur and how long different climatic
to become independent of the climate (Schneider, 1978). phases may last. The analysis of tree ring data enables
Catastrophically cold wioters with accompanying heavy us to put events of our own times in perspective and
snowfalls In U.S.A., devastating droughts and floods in to develop a picture of what is 'normal ' .
Africa and Asia, and resulting lost harvests, all serve to This is particularly important if we consider the
call this fact to our attention. The occurrence of this sort of implications of our current attitude toward our environ-
natural disaster has led to a considerable amount of ment; that is our unthinking exploitation of reserves
palaeo-climatological research by dendrochronologists of fossil fuels and resulting rise in both the carbon
all over the world who are working on the problem of dioxide levels in the atmosphere and the acid content
trying to provide a basis for the understanding of large- of rain . Such events are most probably inducing long-
scale climatic changes. term changes in the earth's climate.
No other living organism can provide as much informa-
tion of interest to palaeoclimatologists as the tree, with its 3. Determining the causes of long-term climatic fluctua-
tions:
annually del.ineated growth increments, the tree rings. The
three most Important areas in which data gained from the The analysis of long tree ring series can contribute
analysis of tree rings can be applied to palaeoclimatology considerably to our understanding of the causes of
are described below. long-term climatic fluctuations. It may be that our
understanding of such changes will be improved by
1. The extension of meteorological records: the examinatIOn and analysis of wood that has been
Using tree ring series from trees of different ages and preserved in ice and on the sea-bed.
Change of
sola' radiation
OUler space
Atmosphere
TeU8sIfIst
fad.al4on
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).
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 growth and low temperatures in autumn, winter and spring. High
south-western U.S. The following factors are most favorable to precipitation during most of the previous year and low precipitation
growth: during the growth period. (c) High temperatures in summer and low
(a) Above average temperatures in April associated with below temperatures in winter and spring. Low precipitation in June of prior
average temperatures in May and June. High summer and winter year and high precipitation during late winter and early spring of the
precipitation. (b) High temperatures in August of the year prior to 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 reconstruction developed. The atmospheric pressure dis-
when information is available from networks of tree ring tribution over the Western Pacific can for example be
series and meteorological series. As Fritts et al. (1979) reconstructed from tree-ring networks from the South-
have demonstrated, the use and general validity of the western United States (Fritts et a/. , 1979).
network may extend far beyond the area covered by Fritts et a/. (1979) used 65 tree-ring chronologies from
the tree-ring series. This is so because the existence of eight different species of conifer from the west of U.S.A.
large-scale atmospheric circulation features and the close and from Canada for 1660 to 1970. Schweingruber (1985)
relationship between climatic conditions in areas that may used 123 maximum density chronologies for eight dif-
be some distance apart. Where such networks are avialable ferent conifers from Europe.
the tree ring series can be calibrated and models for
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.
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.'
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 carried out extending any further back than the sixteenth
long-term climatic conditions and changes. For centuries century. It should prove possible to reconstruct the dis-
astronomers and climatologists have attempted to identify tribution of precipitation in the south-west of the United
cyclical climatic patterns. It was this possibility which led States over the last 2000 years using series which have
A. E. Douglass to lay the foundations of dendrochronol- been built up from archaeological data (Rose et al., 1981).
ogy (Webb, 1983). He devoted much of a lifetime's work Using Bristlecone pine chronologies for trees at the upper
to the attempt to discover a connection between tree and lower timberlines, dendrochronologists have suc-
ring widths and sunspot cycles, using long tree ring ceeded in reconstructing both temperature and precipita-
sequences, e.g . for Californian Sequoiadendron gigan- tion on a decennial basis (La Marche, 1974).
teum . It was only with the development of computer The reconstruction of climate on the basis of tree-ring
techniques and modern statistical methods that the weak chronologies is fraught with problems, as is clearly shown
indications of the double sunspot cycle of 22 years could through the comparison of sequences:
be discerned.
In recent times a number of people have examined The density of coverage is decisive for the interpreta-
various time-series for periodicity, and have achieved tion of the curves. The lower it is, the greater the
results similar to those obtained in earlier analyses, effec- amplitude of the curves . Variations in thinly covered
tively identifying cycles of differing lengths. Stockton curves may be due to either differences between
(1981) found evidence of the 22 year Hale magnetic cycle individu als or between sites.
in the drought area index (DAI) for the Great plains, Differences between species are reflected in sensi-
U.S.A. The DAI had been calculated from a considerable tivity. Maximum density in spruce and fir is less
number of tree-ring width chronologies (Mitchell et a/. , sensitive than that in pines or larches (Laueneni
1979). Bell 1981 has found indications that the same Tornetrask). However, the less sensitive species
chronologies partially reflect an 18.6 year lunar nodal contain just as much climatological information as the
signal. others.
In evaluating the findings described here, it must be Differences between sites are also reflected in sensi-
borne in mind that many researchers over the last forty tivity. For instance, the sensitivity of larches on sites at
years have found indications of cycles of quite differing the upper tree limit (Hbhenbiel) is greater than that of
lengths, and that even the more recently discovered larches in the lower subalpine zone (Grachen).
indications of the 22 year cycles are extremely weak and Comparative studies on given calibration periods have
are by no means discernible in all the tree ring chronol- shown that the curves under discussion reflect tempera-
ogies that have been examined. It has not so far proved tures in the summer months; in the Alps, those of
possible to trace with consistency the long term sunspot July through September, in Lappland those of July and
cycles in individual tree-ring width chronologies. How- August.
ever, the carbon-14 content of the cellulose provides an Despite all these reservations, the following climato-
indicator of this periodicity (Stuiver and Grootes, 1980). logical conclusions may be drawn:
The results of this aspect of dendrochronological research
are still very controversial. The amplitude of fluctuations in summer temperatures
over the past 1500 years lies within the same range as
The fourth stage: climatic conditions in earlier times that of the past 150 years. The meteorological mea-
surements of the 20th century do not reflect the entire
Almost every dendroclimatological study aims to shed range of variation in summer temperatures .
light on climatic conditions in the past. Large-scale clima- Warm and cold phases alternate fairly rapidly, though
tological reconstructions extending as far back as the the phases are of differing durations . .
fifteenth century have been successfully carried out using The transition from one phase to another is as a rule
data from old trees growing in semi-arid areas, for exam- abrupt, taking only from 2 to 10 years. Gradual
ple, the south-west of the United States (Blasing , Fritts, changes are more seldom, e.g. , Hbhenbiel 6, 3200-
1976). Trees in the boreal zones live less long, on 2800 BP.
average, than those in semi-arid areas. Thus it is unlikely The amplitude of climatic fluctuations since 500 A.D. is
that large-scale reconstructions for these zones can be relatively small.
20
,
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 the tree ring. As trees put on an annual growth increment,
about the changes which have taken place on the surface those changes can be dated. In many cases it is possible
of the earth. Whenever an event has an effect on trees or to determine both the intensity and the duration - be it
forests this is reflected in the structure of the shoot and days or thousands of years - of such occurrences.
Reaction mechanisms in trees retard the production of new cells; the offspring are able
The tree provides a good example of an organism which to form other types of cells, such as those comprising
has adapted well to its environment. Since this environ- callous tissue or resin ducts.
ment has been subject to considerable change in the
course of evolution, the tree itself has developed into a Living cells
flexible organism: its living tissues are capable of adapting
These can react to changes in the environment through
to or reacting against changed conditions.
slower or faster cell wall growth and through the formation
A number of parts of the tree are capable of such adapta- of special substances, such as those which comprise
tion: heartwood and compression wood.
The shoots Dead cells
Adventitious shoots are formed where it is necessary Are the source of information about how and why an
to change the direction and rate of growth. The shoot- organism has died. Depending on the amount of stress to
system may also form roots. which a tree is exposed, either individual cells, or the
The roots whole organ concerned or even the tree itself will die.
Each of these reactions can be isolated and dated.
Can react in the same way as the shoots and are capable
of taking over the functions of the shoots. The way in which wood has decomposed
Those tissues which are capable of division Indicates something about the conditions to which the
tree was exposed after it died. It is often possible to
In particular the cambium: these tissues can accelerate or determine when decomposition occurred.
Adventitious roots and shoots formed by the cambium of the root The formation of large and small cells showing the different thick-
and shoot systems. ness of cell walls.
176
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
178
Geomorphological processes which are reflected in the growth of a tree
179
Variability in the run-off of rivers
Large areas of the earth are subject to a basic shortage of Rio Negro, Argentina
water. It is therefore particularly important, especially for
farmers, to know how often periods of extreme dryness The Andes form a pronounced water-shed in Patagonia in
are likely to occur. In various parts of the world the the south of South America. Around the main ridge, at
existing short river-flow series have been extended - 40 o S, the annual precipitation is 2000 mm, while 80 km
often with considerable success - using tree-ring data. further east it is only 200 mm. Holmes et a/. (1979)
It was Jacob Keuchler, a German forester who had succeeded in relating sensitive Araucaria and Austro-
emigrated to Texas, who was among the first to recognize cedrus chronologies from the semi-arid eastern portion of
the link between tree ring width and rain-fall. In 1859 the the distribution area for these species to the outflow of
Texas State newspaper published an article by him on the the Rio Limay. The graph illustrating the close match
climate in Texas. Keuchler had ascertained that of the 134 between the actual and the tree-ring reconstructed hydro-
years that he had examined, 19 had been dry, 11 very dry graphs is impressive given how difficult it can be to
and 12 extremely dry. Fifty-nine had been very wet, 11 synchronize even a single tree with a chronology. A
wet and 22 moderately wet. number of different tree-ring features, for example width,
Recent work by Charles Stockton (1971) primarily on variations in the boundary between early and latewood,
trees in the south-west of the United States, has shown variations in density and in the density of the tree-ring
that dendrochronology can play a useful role in hydro- tissue, must be taken into account in attempting such a
logical analysis. Jones et a/. (1983) have illustrated that synchronization. The statistical calculations here show
changes in river run-off can be reconstructed using data that the annual ring widths reflect the annual run-off
for English oaks. very closely.
,",Ll
CH ILE .~o ...
L AR GE NTI NA
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
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
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.
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
There are thousands of fossilized tree stumps in the using radiocarbon dating by taking into account material
Quarternary deposits of the large Central European rivers . provided by the long annual ring series for Central Europe
Such stumps are often exposed during the working of and Ireland (Baillie et al. , 1983) .
gravel beds. As they fossilize they form a black substance
which is used industrially to produce a valuable veneer . Vegetation in river valleys
This substance also provides the dendrochronologist with
useful material for the construction of tree ring sequences. About 14 000 years ago the glaciers receded from the
A considerable amount of work has been done in this area low-lying areas. And in about 12500 B.P. the first pines
by Bernd Becker. In conjunction with geologists, archaeol- and birches began to grow in these regions (Kaiser,
ogists, and pollen analysts he has been able to build up a 1979). This type of woodland began to give way to the
picture of the making of the Central European landscape mixed oakwood , comprising oak, elm, ash and sycamore,
(Becker, 1982 and Schirmer, 1979). It has also proved in about 9500 B.P. It was not until the Middle Ages that
possible to check and complement findings produced this type of woodland was in turn displaced in the low-
lying areas by poplars.
The places where oak trunks have been found in the Oak trunks found in the gravel beds of rivers in Central Europe. The
Rhein, Main and Danube (= Donau) (Southern Central branches are only very rarely present; stumps often remain. No oaks
Europe) (Simplified, after Becker, 1982). have been found in situ. They have all been moved to another
location (From Griinhilder, 1972).
~ argillic hOnlon I1I\IJ cambic horizon ~ 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).
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
<) 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).
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
" ~ > =-
~ .
• ---- \ /.<"""- -=- ~
.;~~ , ~
-.l.......l-Ll" .. " .. , "
Simplified diagram of findings of pollen and mollusc analysis and
dendrochronology (Kaiser, 1979).
Fossil trunks of oak and pine have been found in the peat Where the trunks are found
bogs and marshy areas of the maritime regions of north-
ern and western Europe. A good deal of dendrochrono- This varies considerably. In those marshy areas of north-
logical work has been done on the oak remains, notably ern Germany which have not yet been drained trunks
by Baillie and Pilcher (Baillie, 1982) in Ireland and by have been found beneath as much as 20 to 200 cm of
Delorme and Leuschner (Delorme, 1984) in the north of peat. Here the results of dendrochronological analysis
Germany. The fossil pines have received relatively little done on wood which has been well-preserved yields
attention to date. Since they are in a remarkable state of information about the stratification of these areas. Where,
preservation it is to be expected that their radiodensito- however, the level has sunk as a result of the peat-bog
metric analysis will yield important palaeoclimatological being drained, solid trunks emerge on its surface. In such
information. cases it is not possible to relate dendrochronological
The main applications of the analysis of the oaks that findings to the stratification of the area. Despite the large-
have been found in these areas have been, first, the scale digging for peat that has taken place in northern
construction of an oak chronology for Ireland for the Germany, Leuschner and Delorme have succeeded in
Holocene period; secondly, the calibration of radiocarbon measuring and analysing a considerable number of
methods (Pearson et al., 1983); and thirdly, the derivation stumps and trunks in situ.
of information pertaining to climatic conditions in the past
(Delorme et aI., 1983).
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).
188
Interpreting these findings
A picture is gradually being built up of the changes to bog. Trees growing in the wettest areas die first, while
which peat-bogs have been subject. The information those on any slightly higher pieces of ground tend to
necessary for such a survey is being gleaned from a survive longer. This is the reason for stands of trees in
study of the vegetation and the topography of the bogs, an area like this dying off gradually over a number of
together with an analysis of dendrochronological findings. decades. When the trees have died as a result of being
The main factors affecting the growth of trees, especially damaged in a storm or have rotted at the base of the
those limiting growth, are hydrological. In the case of trunk, they quickly become subject to anaerobic condi-
oaks, germination is not possible where the surface of tions provided that the peat-bog is growing at a suf-
the ground has either completely dried out or is flooded. ficiently high rate for the trunks to become submerged
Growth is also limited by the lack of nutrients in peat. This fairly rapidly. Such trunks are extremely well preserved.
can be most clearly seen in the narrow rings which trees Where, however, a trunk lies on the surface of the bog for
growing in such habitats tend to display. Frequent peri- some time it begins to decompose. In oak trunks which
odic changes in annual ring width reflect changes in have been exposed to the weather in this way the sap-
drainage conditions: whenever the water level changes wood may begin to decay after as little as five years.
suddenly this is expressed in an abrupt narrowing of the We can only start to speak of a general change in
rings. This can be the result of dryness, i.e. where the climatic conditions - the onset of a period of unusually
water level has fallen, or of a lack of oxygen in the soil, high precipitation for example - when most of the trees
i.e. where the level has risen. It is generally possible to in a number of different such habitats die and very few
relate the death of a tree to growth conditions in the peat- new trees in the new generation survive. The fact that the
chronologies of peat-oaks can be synchronized with
B.C./AD.
100 0 100 700 300 400 !)()O 600 700 those of oaks beside rivers indicates that the tree rings of
J ,
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
Kehdinger Moor the history of this type of area can be dated with absolute
near Hammah
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)
Baillie, 1982; Leuschner and Delorme, 1984; Leuschner et al., 1985. 189
Changes in water table
Human activity has changed the level of ground water, ground water was lowered - it had previously often been
particularly in the 20th century, as a result of recovery of level with the surface - growth conditions for existing
water for drinking purposes, the building of dams, regula- trees improved and it was possible to give newly planted
tion of river and lake levels and drainage of moorland ones a good start. Drainage channels at a distance of
(Markus 1936). Such changes are reflected in the pattern 5-10 m and a depth of 50-80 cm were dug accordingly.
of the tree rings. The drainage of an area in the northern GrUnig (1955) studied the effect of drainage on growth
Pre-Alps of Switzerland, endangered by erosion, will rate and found that in this area of high rainfall , the results
serve as an example. In this area of high precipitation were wholly positive. Young spruces, growing on the soil
(about 2000 mm/year), which is also subject to violent dug out frDm drainage channels, grew well from the start
storms, subalpine spruce and mountain pine woods were and this trend fell off only slightly, in line with normal
over-forested in the 19th century. As a result slope move- expectations, when the trees grew older. If the drainage
ment, erosion gullies and debris-covered areas were to channels were neglected, water built up again and growth
be found in the Flysch region, an area with a natural dropped abruptly.
tendency to heavy, sodden soil. When the level of the
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).
190
Trees react suddenly to drainage by immediately showing mountain pine.
growth improvement. The change from poor to strong The growth improvements are clearly a result of drain-
growth occurs from one year to the next in nearly all age since trees in undrained areas show no similar im-
trees. The reaction time of a stand, however, varies from provement, as is shown by core samples. Tree ring
between one year and six. Presumably the physiological sequences provide information on the effects of technical
performance is suppressed as a result of shortage of measures. In the immediate vicinity of the drainage chan-
oxygen in the soil before drainage. When the drainage nels, at a distance of 1-3 metres, the rate of growth
channels start to function and improvement sets in, this improvement was higher than at a distance of over 3
can last for many years as can be seen in spruces and 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
Along the coasts of north western Europe, Heyworth from stumps in situ the relationship between sea level
(1978) working in the British Isles, and Munaut (1967) in at the time of their growth and the present day can be
Belgium, have found several large areas where stumps found,
and prostrate tree trunks of pines and oaks (Pinus sy/ves- their age can be calculated by radiocarbon methods,
tris and Quercus sp.) had been covered in sediment. the course of their growth can be calculated by annual
These finds add greatly to our knowledge of the Holocene ring analysis,
period for the following reasons: 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.
Remains of a forest of pines and oaks on the Belgian coast (Munaut, 1967).
Many sites with Holocene finds have been discovered on Munaut (1967) studied the developments in a subboreal
the coasts of England and Ireland . Today the stumps and coastal forest in Belgium. At 2.5 m under the surface of a
trunks are lying either below the tidal level or in the inter- polder he found 56 pine trunks embedded in marl and
tidal zone. Chronologically they span over 8000 years. peat, which provided a ring sequence of 242 years . Pollen
The trees died in a continuous sequence as a result of analysis showed that sphagnum moss grew on this site
rising sea level. What causes these fluctuations in sea before the pines appeared , so the area must have been
level? wet. Then the ground-water level sank and over a period
of approximately 100 years a p n i e forest developed . It
a global rise in sea level? grew normally for another century and died off gradually
rising of the land mass after being relieved of the
during the next fifty years. High pollen counts for Erio-
weight of the ice of the Ice Age?
phorum indicate a rise in water level.
tectonic vertical movements?
An investigation of trunks and the sediments sur-
Such mechanisms are to a certain extent compensatory , rounding them from various coastal areas from Portugal
for example rising of the land and rise in sea level, and to Scandinavia would certainly be worthwhile, as it would
can be detected only if large areas of coast are in vesti- provide much new information on vertical movement of
gated. The findings of Hepworth suggest that the coast- the land masses, dynamics of vegetation and climatic
line of Great Britain has changed , mainly as a result of development.
vertical movements involving sinking in the southeast and
rising in the northwest.
+
on
Radiocarbon dating of +
+
submerged forests.
Rising or sinking of
Y land mass in the last
8000 years in meters. +
+
+
+
+
+
+
+
+
+
+
+
+
pine +
Terneuzen . +
1962
+
+
+
+
+
+
+
+
+
193
The history of glaciers
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. Comparison of the age of the moraine, based on tree ring analysis
based on tree age. In the early 16th century until about 1740 the tip on Mount Baker with similar moraines in different areas. Mt. Rainier:
of the glacier was at about 850 m above sea level. In the 19th usually dated by lichenometric methods. Scandinavia: dated using
century, by which time the glacier had retreated, the tip was at 1250 information from historical documents (A). Swedish Lapland: dated
m above sea level. by lichenometric methods (B). Obviously certain parallels exist e.g.
Key: 1 Glacier: 2 River: 3 Contours: 4 Date of previous position 1850, 1910 and 1925. Inexact dating and differing reaction times of
of glacier: 5 Tip 1940: 6 Age of the oldest tree in the moraine the glaciers could be responsible for the displacements (Heikkinen,
(Heikkinen, 1984). 1984).
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 licult to find a direct connection between these finds
studying the circumstances of the finds and the state of and the history of the glacier. The tree rings provide
preservation of the subfossil pieces of wood. important information about the Holocene period,
Basically there are two types of finds: however.
Tree stumps in situ, covered at one time by the glacier
Trunks lying free in moraines in the approaches to and later re-exposed. By determining the date when a
glaciers. They have been brought down as a result of tree died and how long it had lived, information can be
movement in the moraine, without the actual glacier obtained about the level of the glacier in relation to its
being directly involved, although it may have trans- present state of retreat and also on the speed of its
ported them over short distances. It is usually dif- advance in earlier times.
Stump of Swiss stone pine, in situ and uncovered, in the great Fossil trunk lying in a moraine (Ri)thllsberger, 1976).
Aletsch glacier, SWitzerland. The first ring was formed in 973 A.D.
and the last one about 1100 A.D. (Holzhauser, 1984).
a b
a
o • 0
(] o
o
(Q:'£J moriline Position of subfossil tree trunks in the approaches to glaciers and
El ice ~ rock H hOllo.warea
.,.,.... fossil,zed roo IS moraines in relation to the ice level in the regressive phase of the
present day (Ri)thlisberger, 1976). (a) Moraine and glacier tongue
Possible position of subfossil root stumps, in situ. After being - top view. (b) Glacier and glacier bed with characteristic levels
embedded in moraine debris, often an erosion phase sets in sketched in. This representation follows Ri)thlisberger's model,
(Holzhauser, 1984). which applies only to European glaciers.
196
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
~ -r-~-~-r--
~ t--+-+--;.
_ 21
- ,. . . .- -. . . .,
&..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, who studied the slow creep of a wooded, volcanic debris
whether they be as landslides, slope creep or as rise and slope in Utah, U.S.A., in great detail. He took about 1000
fall of subsoil. In densely populated areas records usually core samples and took cross-sections from the trunks of
exist about such occurrences. In unpopulated areas, how- 220 trees and recorded all the variations in tree ring
ever, only trees can provide information about the extent sequences which could possibly be attributed to slope
and frequency of such events. Surprisingly enough, only movement.
very few scientists have used the tree as a source of Variations along several radii were determined, dated
information, whether it be merely through determining the and presented in graphical form for every tree observed
age of trees or observing the changing growth rates in and because the position of each was noted, every dated
annual ring sequences. In 1971 Alestalo tried to compile occurrence could be charted. For instance, one single
data about the intensity of cryogenic earth movement by rockfall, which could be dated by scars on trees, began a
observing the eccentricity of inclined trees growing on period of slope movement which lasted from 1905 till
slopes bordering moors and lakes. But the most impor- 1910.
tant work was carried out by Shroder (1978 and 1980)
Trees with bent trunks growing on an unstable, creeping slope Cross-section through the base of the trunk of a pine tree growing
(Lenz, 1967). on an unstable slope above the forest limit in the Alps.
o o
o N
0> 0>
:II
-
<
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 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).
j
I
r-
•
r
-
~
!""" . I
!""'!"': ....,j I
!l"""
.-i
~ :;....j
,00
190 a 1909 1910
.,
'"'
a,yu Canyon.
," P'rIC'pllaJIOn
S ... , roqall
"
199
Tree-ring research and wind
Strong winds may affect both the shape and appearance zones, a process which will continue over a long period of
of individual trees and the pattern of vegetation over time.
whole regions. In valleys with strong prevailing winds the This sort of biological signal assumes particular impor-
trees grow to one side, on t; Ie coast the trees develop tance in times when the possibility of harnessing wind
characteristic lop-sided crowns, and on the edges of power is being considered, because it is rare for meteor-
woods or forests with katabatic winds the trees tend to ologists to record wind speeds in the same place over a
be stunted. During bad storms trees may be damaged , period of years. Today the main application of annual ring
uprooted or tilted to one side. Such phenomena have research in this area lies in attempting to trace the rela-
been extremely well documented . (Wendel, Hewson et tionship between the effect of wind on trees and its
al., 1977) . climatological aspects. As is shown in more detail below,
These events usually show up in the trunk and the the wind may have a direct effect on a tree or it may affect
annual rings. Where a tree is affected by prevailing winds its growth in some indirect way, through the movement of
anomalies occur in the formation of the trunk; the amount slopes or snow for example, or through changes in light
and type of compression wood that a tree develops reflect conditions or in the length of the vegetation period or
differing wind velocities, and a change in the angle of the through the effect of salt-spray in coastal areas.
tree will cause it to start forming compression wood
A characteristically deformed tree growing on an extremely windy Poplars in a river valley, growing at an angle.
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 Wind-stressed trees commonly exhibit sudden changes In the tree
has been deformed by the wind. ring width, or the formation of compression wood.
200 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 This area has received relatively little attention from
velocity on the one hand and the reaction and shape of dendrochronologists: neither the reaction of particular
the tree on the other. Changes in the wind speed can be species to wind-related phenomena nor the response
seen in the variation, from year to year, in the proportion of the tree to an increase in wind speed have been
of compression wood: such changes can be expressed in thoroughly investigated.
terms of the compression ratio.
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 By plotting the crown shapes of a number of different trees against
crowns have not been affected by wind (0). With increasing wind annual mean wind speed it is possible to illustrate the close relation-
speed the crowns become correspondingly lopSided, the extremes ship between meteorological factors and biological reactions (Wade
being flag (IV) and dwarf forms. This classification is generally et sl., 1979).
referred to as the Griggs-Putnam index (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 A section through a trunk showing strong asymmetry, taken from a
annual compression ratio. This is obtained by dividing the tree ring tree in an exposed stand of dwarf trees. Conifers form larger and
widths for the leeward side by those for the windward side. The thicker-walled cambial cells on the leeward than they do on the
annual wind stress can thus be clearly seen (Wade et sl., 1977). 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
Picea abies trunks: growth is eccentric and compression wood is Abnormalities in crowns schape, stem eccentricity at chest-height
formed all round the trunks and on one side of them as a result of and in ground level in a 90 year old Lodgepole pine stand (Pinus
the trees being subjected to strong winds when they were young. contorta) in the very exposed Front Range area, Colorado. The
results from the stem eccentricity at chest-height are the most
reliable.
The extent to which wind influences the tree's habitat can in the age structure, the growth rate and the proportion of
be seen especially clearly in subalpine areas with stands reaction wood in different parts of the trunk. The exam-
of dwarf or stunted trees (Holtmeier, 1981). Here the ples given below illustrate how sensitively trees react to
growth of the tree is affected both by changes in wind changes. Every single change has a direct effect on the
speed and direction and by different amounts of snow. tree, either changing the growth pattern of the tree or
Shoots which are protected by a layer of snow will survive causing the shoots or even the trees themselves to die
the winter but will nonetheless have a shorter vegetation off. These stunted trees enjoy an extremely precarious
period, depending on the time at which the buds appear. equilibrium. Unlike trees in wooded areas in temperate
Shoots which are above the snow cover can benefit from zones, they have no sort of buffer to protect them from
warm winter days but are at the same time exposed to the extreme climatic conditions. The fact that individual stands
various dangers associated with frost, such as frost- are able to survive over hundreds and even thousands of
drought. Here the needles continue to transpire, and years is attributable solely to their capacity for vegetative
because the necessary replacement cannot be drawn regeneration through the formation of roots by the
from the frozen soil, there is a danger of their drying out. branches.
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
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.
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.
, 1977
; II III
, This tree was badly damaged in the fire.
.
OJ). ~ ... rtv A
_ 00 • •'Iqr 1 'f'DIJI"I;I "
_ 00
forms a scar. In order to determine precisely the number
.--
- 00
__ 0
'!'lO 10 30
which of course involves felling the trees. It is not pos-
sible to say in which years fires broke out merely by The numbers of trees at the edge of an area affected by fire showing
examining and counting the rings, since - especially in sudden growth release. The figures relate to a timberline site in
Niwot Ridge, Boulder, Colorado. The fire occurred in 1905 and the
trees in dry areas - some annual rings may be missing. instances of growth release concentrate on the period 1906 to 1908.
It is therefore necessary to date the fire scars dendro- (Kienast and Schweingruber, 1986.)
MFI
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
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;
~ ~
... 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).
Since the end of the last century fire prevention mea- the nutrients remain on the surface and so are not
sures, particularly in the American West, have ensured available to the trees. The occurrence of fires used
that forest fires have been kept under control. This in turn to ensure that these potential nutrients were continu-
has caused dramatic changes in the pattern of vegetation ously being recycled in the soils.
that had been dominant in these areas for thousands of
years. Today the effect of this sort of intervention is pretty The structure of the present stand has considerable dis-
well understood. The example of the Ponderosa pine advantages. The dense forests are much more suscepti-
forest near Flagstaff, Arizona, is certainly typical of many ble to large scale crown fires that destroy entire stands,
other cases. (Kennedy-Sutherland, 1983). In the last because the young dense trees and accumulated fuels
hundred years there have been no fires at all in the stand: provide a 'ladder' for fire to the crown level. Insect popu-
before this time fires occurred once every two or three lations, e.g . the spruce bud worm, flourish in dense
years. The stand was described as park-like and open by stands and mistletoe, particularly Arceuthobia, have an
early explorers, but now it is extremely dense with many adverse effect on the vitality of the stand.
thickets of young trees. Since 1920, which marked the It is mainly for the sorts of reasons outlined here that
end of a relatively wet climatic phase, the growth rate of foresters and biologists have started to call into question
both young and old trees has decreased . the concept of total fire prevention. The work done by
There are a number of reasons for this marked de- dendrochronologists in this area have helped to provide a
crease in growth background against which today's essentially short-term
measures can be evaluated.
The older trees are in competition with the younger
ones particularly in the root area. The young trees take Conclusion
a lot of rain-water from directly under the surface.
The crowns of the trees, especially of the young ones, The effect of forest management in semi-arid and boreal
have to compete for light and space regions has been the subject of fairly intensive study.
The fact that the ground is relatively densely covered Relatively little work has been done on the effect of fire in
inhibits the flow of water to the root layer: a lot of broadleaf woodlands in the time before forestry manage-
water is lost through interception. ment, however. Studies in this area would enable an
In dry areas the organic materials (needles, twigs and important gap in research to be filled.
cones) decompose slowly. This means that most of
3
2
><
Q)
"0
c:
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:
Fires which occur when the weather is dry and there is a years. Thus the relationship of fire occurrence to weather
lot of wind - in high summer, for example - cause more is probably more strongly linked in the low frequency
damage, in terms both of the intensity of the damage types than in the high frequency types. On the other hand,
to individual trees and the geographical extent of the the absence of fire may be strongly linked to weather in
damage, than those which occur at wetter times of the the higher fire frequency types.
year, when there is less wind . This difference is most The frequency and intensity of forest fires is largely
clearly reflected in the extent of the area affected by a fire. dependent on the type of vegetation in the stand . Where,
The extent of past fires can be mapped fairly accurately for example, there is a lot of grass the fire spreads quickly
provided that a good number of trees are examined. Fires and leaves little trace . By contrast, dense stands of young
occur more often in dry than in wet periods. This can be trees may be totally devastated over wide areas. In woods
clearly seen by comparing the average intervals between composed of species which are resistant to fire - in
fires for different vegetation zones. Pinus ponderosa stands, for example - it often happens
The incidence of fires in semi-arid zones is very high. that only the grass and shrubs are destroyed and the
In some areas fires occur as often as every two years. In large trees sustain damage in the form of a fire scar.
boreal and sub-artic regions, on the other hand , forest Woodland comprising trees which are not resistant to fire,
fires may occur on average every hundred to two hundred e.g. spruce, are often destroyed over wide areas.
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).
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).
The eruption of a volcano has always caused a consider- ash. This very fact made it possible to give a precise date
able amount of destruction and change. Living creatures for the eruption of the volcano on the basis of radiocarbon
in the immediate vicinity are destroyed; particular sites are dating. An analysis of leaves found buried in the sediment
laid waste while others are improved in some way; clouds indicated that the volcano erupted in early summer. Only
of volcanic dust may cause climatic changes which affect a few km away from the site of the eruption the layer of
the whole of the earth. The different effects of volcanic dust may be as little as twenty to thirty centimetres thick.
activity on trees, how the damage caused is reflected in Clouds of ash may be blown over distances of hundreds
the annual ring picture, and how this information can con- of km by strong winds.
tribute to an understanding of the incidence of volcanic Trees which are directly in the path of a lava flow are
activity are discussed below. charred by the tremendous heat which the lava emits.
Those that are just on the edge of a flow will only be
The direct effects of the eruption of volcanoes damaged to the extent that their trunks may sustain some
burn marks or, where their crowns have been singed,
Practically everything in the areas immediately surround- they will produce narrow rings for a number of years. By
ing a volcano is destroyed in an eruption. Pressure waves analysing the last-formed rings of trees destroyed by a
cause trees to be knocked the ground, snapping them rise in the water-level caused by lakes being silted up
like matchsticks. Trees in more sheltered positions are with lava, it is possible to determine the year in which the
tilted. volcano erupted. A volcanic eruption often causes ice
After an eruption the countryside lies under a thick melting. The resultant flooding may cause the trees which
blanket of volcanic ash. Following the eruption of the were 'felled' by the effect of the volcano to be carried
Laacher volcano in the Eifel region of West Germany, for over long distances.
example, the scorched trees were left under 20 metres of
Destroyed forests in the west of North America. When Mount St. In 1912 whole forests were covered with ash when Mount Katmal
Helens erupted on May 18th, 1980, the conifer forests at the erupted. Years later river activity exposed the roots of conifers
timberline were completely destroyed by heat and pressure waves. whose branches had in the meanwhile produced adventitious
Many trunks are now buried under the ash (Anonymous, 1980). branches (Griggs, 1928).
The waves of air pressure which followed the eruption of Mount Trees exposed in a mudflow in a stream draining Mount St. Helens.
Katmai flattened whole forests. The position of the trees and the Hot debris from the eruption induced by rapid melting snow and ice
date of their last ring provide information about the nature of these in the upper slopes of the volcano, causing mudflow that carried
occurrences (Griggs, 1928). trees many km downsteam from the volcano.
210
Further away from the volcano, at the edge of the area and die as a result. Species with thick bark, e.g. Araucaria
most affected by an eruption, the heat has a selective or Ponderosa pine, are far more resistant to the heat that
effect. Young trees protected by a layer of snow may those with thin bark such as beech, southern beech or
hardly be damaged, while larger mature trees are killed. spruce.
The nature of the tree's vegetative organs are crucial to Tiny particles of volcanic ash often adhere to the
whether or not it survives. Trees like Araucaria may be surface of the leaves, inhibiting photosynthesis. This
protected from the searing heat by their thick needles and effect is particularly significant in semi-arid regions where
bark, while those species whose needles are rich in resin the layers of dust may remain on the leaves for months.
or which have thin leaves will be set alight immediately
Changes in the pattern of vegetation in a stand of Araucaria and Young conifers which were protected by a layer of snow survived
southern beech in Chile (Araucaria araucana and Nothofagus the waves of pressure and heat sent out when Mount St. Helens
antarctica), caused by the effects of hot volcanic ash. The erupted. By contrast, the dominant trees, Douglas firs up to 1.2
Araucaria survived almost unharmed thanks to their thick leaves meters thick, were flattened and seared.
while the beeches were badly burnt. After a relatively short time,
however, they were able to produce new shoots.
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 years after 1064-1067. It is probable that the wood was
soil. This effect can be seen very clearly in the tree rings used to rebuild the settlement when the Indians returned
of conifers which are growing in volcanic ash. Thus to the site which had been severely damaged in the
volcanic events can be dated. An example of this was the eruption of the volcano. It is likely that the trees originate
eruption of Sunset Crater near Flagstaff, northern Arizona. from an area which was affected by the eruption of the
Here the trunks of those trees which were damaged in the volcano. The analysis of these pieces of pine shows
blast itself were so badly charred that it proved impossible clearly that the volcano erupted approx. 1064 AD and that it
to date them dendrochronologically. However Smiley was a good 50 years before the area was resettled. It proba-
(1958) identified a number of trunks which had been used bly took this long for the layers of volcanic ash to weather
to build the Indian settlement of Wupatki. This wood was down into fertile soil.
characterized by narrow rings with low sensitivity for the
Section through the trunk of a fir near Mount St. Helens. Photo- Section through a fir growing in the area devastated by Mt. St.
synthesis was impeded through the effect of a layer of dust on the Helens. The growth of this tree had been limited for 50 years
needles. The release can be seen as early as 1981. 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.
1050 1064
Four tree-ring curves of pieces of timber from the Wupatki settle- A disc from a beam from the Wupatki settlement. In 1064 the tree
ment. The onset of the period of poorer growth and lower sensitivity rings started to narrow abruptly. The wood was dated using the
is a expression of the defoliation by the volcano Sunset Crater skeleton-plot method (Courtesy of the University of Arizona).
erupted in 1064 (Smiley, 1958).
1800
I JI111 ~r(II/IIit/,}OJ»1H1"'~
1800
Spirit Lake
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).
-
t
213
Frost rings
Katmai
1912
V
Unusually low latewood density in the annual ring for 1912 in timber- On June 6th, 1912, the volcano Katmai in Alaska erupted sending
linetrees in the Alps. Using a hand-lens tree-ring sequences can be ash more than 20 km 3 up into the air. Griggs (1928) reported that the
dated on the basis of such characteristics. 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 .
!! ,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).
Where large amounts of volcanic dust (silicate particles be found in Greenland - where the ice is split into layers
and sulfur aerosol) are released into the earth's atmo- every year by the action of the cold - can be attributed to
sphere this has a direct effect on irradiation from the sun. volcanic sulfur vapor, while the layers of dust in the ice
Should this dust veil remain in the stratosphere over a itself are composed of silicate aerosol.
period of some months the weather will become colder LaMarche et al. (1984) plotted the incidence of frost
and there will be more periods of extreme cold than rings in the long annual ring sequences for American
normal. Such periods when the weather is colder than Bristlecone pines against volcanic activity. In many cases
usual are clearly reflected in the annual ring patterns of the occurrence of an unusually high number of frost rings
conifers in the subalpine and arctic zones and in the arctic could be directly related to large-scale volcanic activity. It
glaciers: summer temperatures which are lower than must of course be borne in mind that the formation of
average are reflected in low latewood densities. When on frost rings may take place independently of the occur-
the other hand there are two or more nights in the two rence of volcanic activity, and that this activity in turn does
months August and September with temperatures below not necessarily result in a higher incidence of frost rings
freezing (-5°C for Bristlecone pines) the latewood sus- (Timell 1986). Nonetheless a causal connection has been
tains frost damage. Temperatures of this order cause the established between the eruption of volcanoes and the
cells to freeze and be damaged. Non-lignified cells - and occurrence of frost rings: in some cases, for example,
in some cases callus tissue - within the latewood are more than six times as many such rings as normal are
typical reactions to this sort of attack by the frost. formed in years when a volcano erupts.
It is possible to relate frost damage to a fall in tempera- The discovery of volcanic events in the ice and the tree
ture. The frost ring formed by Bristlecone pines in 1884, rings open is up the possibility of determining both how
for example, is known to be a reaction to a frost in the often and when volcanic activity occurred in times for
night of 9-10th September; the ring formed in 1965 which no records exist. The following table illustrates how
resulted from the cold snap that occurred from 17th to far the attempt made by LaMarche et al. to carry out this
19th September. The extreme acidity of the ice that is to sort of dating has been successful:
~
~
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
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
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- and 1958). In 1958 thirty-five million cubic metres of rock
tains were quite frequently destroyed by rock or ice falls. hurtled from a height of 100 m into the bay. The resulting
Such falls may have devastating effects when the masses giant wave swept along the 6.5 km long bay with a
of rock or ice plummet into fjords or seas and cause giant velocity of 200 kph. , leaving large areas of devastated
waves to form . Such waves occur fairly frequently in woodlands in its wake. The pattern of vegetation in areas
Japan, Norway and the U.S.A., and have resulted in affected by giant waves indicates the exact extent of their
woodland being destroyed over wide areas. Miller (1960) effect, while the callusing over of injuries on trees grow-
documented the occurrence of such waves in Lituya Bay, ing at the upper limit of the devastated area provides the
Alaska. In the last hundred years alone earthquakes have key to the dating of their occurrence.
caused a number of glacial ice falls (1864, 1899, 1936
4#
Aerial photograph of Lituya Bay, Alaska, in 1954. (g) and (k) trim- Aerial photograph of Lituya Bay in 1950, immediately after the area
line of the giant wave 1853/44. (m) lateral moraline (Miller, 1960). 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).
Impact slope and the SOO m high wave (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).
Impact slope of the 500 m high wave (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
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
In the winter, rivers in the northern parts of the northern trees, and the level reached by the ice can be judged
hemisphere freeze over for a number of months. They by looking at where exactly on the trees the injuries
then thaw very quickly, usually with dramatic effects, with occurred.
the ice-cover cracking at various points. This causes the Parker and Jones (1971) made an intensive study of
ice floes to be pushed one on top of the other. Some of the build-up of ice in the MacKenzie region of Canada.
these enormous masses of ice are forced onto the river Their aim was to discover how far above the summer
banks, parts of which are wooded, while others form water level ice scars on trees could be identified. It is
dams behind which the water builds up. clear that it is inadvisable to build pipelines, roads, build-
Parker and Jones (1981) have described the varied ings or electric lines in areas where such damage is likely
morphological reactions which this chain of events brings to occur.
in its wake, the most obvious of these being the ice- In stretches of the river that are wide and straight the
ridges which are formed on the river banks. Here the ice tends not to reach a height much above the water
action of the ice causes the vegetation to be prized away level. Where, on the other hand, the river narrows, or at
from the banks, while loose stones build up into moraine- the confluence of two rivers, the ice may accumulate, with
like ridges. The different types of vegetation growing on the result that trees up to 17 metres above the level of the
different parts of the banks are a clear indication of the river will be damaged. Water may build up behind these
fact that earlier accumulations of ice have laid waste the natural dams over an area of up to 100 km. The ice floes
area at some time. Remains of sediments and driftwood in turn cause the trees in this area to be damaged. The
which are found well below the average level of the river fact that the trees on level ground near river banks are
are evidence of past flooding. The effects of the build-up relatively young probably indicates that whole stands of
of the ice can be seen in the damage sustained by the trees have been destroyed in the severe build-ups of ice
trees and in the fact that some of the trees have been that tend to occur per century in these areas.
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
14m.a.r.1. edge
of terrace
grass sedges
Mackenzie river
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).
In 1847, a good half a century before dendrochronology Despite the considerable amount of work that has been
began to develop as a subject in its own right, the done in this area it is really still only possible to tackle the
palaeo botanist D. F. Unger investigated the rings of fos- question of whether or not fossilized woods display
silized woods. And in 1911 W. Gothan sought an explana- annual rings. And this question is worth asking only where
tion for the fact that woods from the Palaeozoic era do not studies of the flora and fauna and the geology of the area
have tree rings. Since then biologists and geologists have enable the following points to be cleared up:
attempted to obtain information about climatic and eco-
logical conditions in particular geological periods. Today it What sort of organisms does the sediment contain?
is widely accepted that dendrochronological analysis in
By determining the species of wood involved and analys-
the strict sense cannot be carried out on fossilized woods,
ing the pollen and diatom profiles and the zoological
since the scope and validity of the method is limited here
macro- and micro-organisms on the same site it is pos-
by the paucity of the material available and the huge time-
sible to ascertain the ecology and sometimes even the
spans that come into question. It took 30 years, for geological age of the deposits containing the wood.
instance, to build up a 7000 year absolute chronology for
Where the fossilized wood displays rings this provides
Europe based on a wealth of material from quaternary
evidence of seasonality. Only where several species are
gravel and peat-bogs. Is it reasonable to expect that a
available for investigation, however, can the question of
chronology which would be more than a thousand times seasonality be answered definitively, since even within
longer than this, reaching back into the early Tertiary
one climatic zone by no means all species possess the
period, could be constructed using very little material?
same capacity to form rings.
~~~
220 Antevs, 1954; Creber and Chaloner, 1984; Chaloner et al., 1972; Fritz, 1982.
In the case of sub-tropical woods from the Palaeozoic era Seasonality
it is often difficult to determine whether the growth pat-
terns display annual rings or not. Nor can we be confident Chaloner and Creber's analysis of fossilized woods
(1973) has yielded the following information about sea-
that certain fern species which are now extinct were even
capable of forming tree rings. sonality:
The first species with properly-developed xylem occur
Where the fossilized woods are found in the Upper Devonian period. Only in very few cases
In the case of tree-trunks it can be said with any degree of were growth increment zones formed (Cal/ioxoJyn
newberry, Indiana, U.S.A.). There are no tree rings
certainty that they are in situ only when they have
roots and are upright. Such finds are, however, made only present, even in woods found in mountainous areas.
rarely. Stumps in carboniferous deposits are generally in Conditions in the Carboniferous period were similar
situ. Where the specimen is found with a large quantity of
for the whole of the Northern Hemisphere.
In the Permian system, on the other hand, conifers in
other wood it may be part of an aCGumulation of driftwood
or of trunks that have been pushed together a~ a result of the Northern Hemisphere display no tree rings, while
a landslide. In such cases trunks from different climatic those in the Southern Hemisphere formed clear rings
zones - e.g. dry mountain areas and moist-warm low (AntarcticoxyJon sp.). (See map on page 12.)
From the Triassic period to the present day there are
elevations - and from different types of sites - dry
woods in each of the geological periods which formed
south-facing slopes and damp marshy sites, for instance
- may be found together. This means that pieces of tree rings.
fossilized wood with and without rings can be present in These findings indicate that a seasonal climatic system
the same deposits. first developed on the earth with the splitting off of
In the case of fluviatile sediments fossilized wood from Gondwanaland, i.e. in the Upper Carboniferous period.
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.
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
In forestry research the problem of bad infestations by able aspect of tree-ring research for the entomologist is
insects was previously regarded primarily from the eco- that it can throw light on those insect behavior patterns
nomic point of view because dying trees, reduced growth, which have long term effects. Insects are selected for
increased risk of disease, delays in producing new investigation which profoundly influence the physiology of
shoots, all have an adverse effect on profits (Timell 1986). the tree, especially those which damage the light-assimi-
At the present time, thanks in some measure to annual lating foliage of the crown, thus impairing the process of
ring research, it has been realized that 'harmful' insects photosynthesis. The resulting poor rate of growth can be
play a part in the ecological cycle and therefore their seen clearly in the annual rings.
activities are of considerable importance. The most valu-
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).
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
IllllIIl 1974
• Ice Baltflnsweiier
Distribution map for cockchafers (Melolontha melolontha and M. Map showing the extent of damage caused by larch bud moth in
hippocastani) with 3, 4 and 5 year development cycles. By examin- various years. In 1970 it was concentrated in the western Alps, in
ing tree-ring patterns of trees in these areas, it can be established if 1974 in the east. A study of tree rings would show whether this
environmental conditions on the sites had varied, either as a result recently discovered shift from west to east actually began hundreds
of climatic changes or human activity (Simplified, Schwenke 1974- of years ago (Baltensweiler 1971-1975, simplified).
1978).
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).
apo
rLaEM
II)
W
Roo .... " DGU
II)
ij
L ...
II>
u u •• u.· 0
...I ...0
0
o X
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).
• 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).
Climatic factors and site conditions influence the com- moth, which recurs in regular cycles, is studied it can be
plicated mechanism governing the development of the seen that within a large area some climatic regions are
insect population and the reaction of trees to harmful affected more often than others. Presumably unfavourable
attacks. This whole area of interaction can be seen in the climates, e.g. cool ones or those with long dry summer
way that the area affected changes from year to year and periods, inhibit the growth of this particular insect. Annual
also in the annual growth rates. Particularly in the case of ring ana1ysis investigations, taking both time and geo-
damage which has long term effects, such as that caused graphical area into account, can throw light on this com-
by spruce bud worm, fluctuations within the same site are plicated relationship.
very noticeable. When an insect such as the larch bud
I~
.~
~
o
E
I
/
". --
I
I --
I I;s
I
1/
I"li;r.:i,
/i
~------------~~~--~~
;1mill1111 IIII
2 0 2 4 6 8 10 years 2 2 4 6 8 10 years
Ild1111M11 I i
Effect of short-lived attacks e.g. by larch bud moth (Zeiraphera Effect of a longer-lasting attack, e.g. by spruce budworm (Choriston-
diniana) recurring at regular intervals. Larches may be attacked eura fumifera). After the first defoliation, severe but not complete,
twice in the early summer. As new needles are formed each time, radial growth rate drops dramatically. If the trees continue to lose
wood growth comes practically to a halt and no reserves are laid needles in the succeeding years, growth rate is reduced appreciably.
down in such a year. In alpine regions two attacks in one year mean If the density of the crown becomes so low that it is no longer able
there is a 5 to 10% chance of a tree dying. When only one attack to supply nutrients for the formation of new needles and earlywood
occurs, the risk is reduced to around 1 %. Radial growth returns to cells, the tree dies as soon as all the reserves laid down in the
normal after 3 to 4 years; the density of the latewood is back to parenchyma cells have been used up. The longer the duration of the
normal levels after 2 to 3 years. 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.)
Conclusions
The following ideas originate primarily from Wickmann, earlier stage of forest development when stands were
1979. Trees and tree-damaging insects have existed in very well stocked and they performed a valuable protec-
every geological period. Insect infestations have affected tive role .
woodland in epidemic proportions for millions of years. It If today's areas of damage are larger than before, and
is known that larches in the Swiss alps are infested by this hypothesiS has still to be proved, it could be as a
larch bud moths every 7 or 8 years. The areas where result of the age structure and species composition of the
damage occurs are clearly defined for each type of insect. forests, in other words, infestations are a reaction to the
Growth reduction occurs in trees over the entire infested 'man-made' forest. Fighting the attacks with chemical or
area, but actual cases of mortality are limited to a few biological measures is wrong and can be justified only in
individual trees in a stand and to a few places within an terms of short-sighted, economic interests. The aim of
affected area. Looked at from the economic point of view forestry planners should be to create a forest which corre-
what has been described must be regarded as a loss sponds as closely as possible to natural conditions.
(growth reduction, dead trees, under-stocked stands), but Taking a broader view, insect activities can be regarded
from the biological point of view the death of some trees as regulating rather than destroying. A new topic for
can be regarded as an opening up of the forests which, discussion arises. Is there a connection between the
once the insect population has departed, will lead to an relatively large areas of insect infestations and changes in
increase in the number of germinating seedlings and a the environment? When we find out where and for how
much faster growth rate in the surviving trees. Particularly long large areas of insect infestation have occurred, it will
in the rather unstable situation which exists near ava- be possible to get an idea as to whether such a connec-
lanche gullies is this an ecological advantage, for insect tion could exist.
plagues mean a return to the situation which existed in an
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.
Forestry research workers, particularly in North America, Once a seedling has attached itself to a branch the vitality
have investigated the problem of the growth of the para- of the tree is reduced. Suckers penetrate the vascular
site mistletoe. Many different varieties of the species system of the host tree, tapping water, nutrients and
Arceufhobium, Viscum and Phorodendron attack trees, assimilated substances. At the point where the mistletoe
usually conifers. The problem has not yet been studied by penetrates the tree the cambium reacts by producing
dendrochronoiogists. callus tissue and an above average number of cells. The
Damage to trees from this source is considerable. formation of new cells, which is normally distributed
Every year 3.2 million board feet are lost in the American throughout the tree, becomes concentrated at this point
west as a result of attacks of mistletoe, which cause either of infection, which means the vigor in this area is in-
a reduction in growth in the host tree or its death. 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, venient point of entry for fungi. The resulting increase in
both in the short and long-term by a mistletoe attack. rotten wood leads to greater structural instability, which
The extent of damage ranges from the loss of a few makes the trees more vulnerable to storm damage. As
twigs to the destruction of whole forests. Generally weaker a result the mortality rate again increases. The ultimate
trees rather than stronger are attacked. effect of mistletoe depends on the host-parasite relation-
The longer the attack lasts, the more serious is its ship, the severity of the attack, the climatic and ecological
effect on stand vigor. This is evident both in reduced conditions and the presence of harmful insects. However,
vertical growth and wood production (annual ring widths, generally speaking, the mortality rates are influenced by a
volume). Seed production declines, another sign of re- combination of three primary factors; presence of mistle-
duced vigor, and so propagation of the trees is also toe, insects and fungi.
adversely affected. Dead mistletoe shoots make a con-
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).
>
-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~
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 and improved timber quality, in the long run. Although
as to achieve optimum production without disturbing eco- several studies have been carried out on the effects of
logical balances. Measures are taken to stimulate organic such measures on individual trees and on whole stand
production while at the same time preserving or improv- growth, few generally applicable conclusions can be
ing many other forest functions, such as providing recrea- drawn, as there are wide discrepancies between the
tional use and acting as a natural barrier to erosion etc. findings. As a result, little is known about the connection
Specific operations such as thinning, pruning and between forestry practices and tree ring structure.
fertilization lead to increases in yield in the short term,
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
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 Mean radial increment in years before and after thinning as a per-
approx. 40 cm diameter. 40 years after it germinated, when its centage of mean tree-ring width. The favorable effect of moderate
diameter was about 30 cm, thinning was carried out and space was thinning lasts only a few years (Mitscherlich, 1970-1975).
created for the tree to develop. For a short time density decreased
but returned to normal within a few years (von Pechmann, 1974).
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.
As with all forestry measures, soil improvement is in- dividual trees can be achieved by means of fertilizers in
tended to raise the vigor of weak stands and increase conjunction with soil cultivation and general forestry mea-
organic production . In cases where the soil is exhausted sures as described above. Whether such measures are
or poor in nutrients and the exact conditions are known , economically worthwhile , however, depends primarily on
appropriate measures can be taken to improve it. These the cost and effectiveness of the fertilizer and the produc-
include: tion costs and selling price of the timber.
Much has been written about the physiological, pedo-
application of fertilizers such as lime, potash, phos- logical, technical and forest yield aspects of fertilizers and
phates and nitrogen. soil improvement (8aule and Fricker, 1967). Little work
cultivation of legumes, e.g . clover, lupins, broom. has been done, as yet, on the effect of fertilizers on the
tilling of the soil; the effectiveness of fertilizers on the anatomical structure of tree rings.
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 Effect of an application of nitrogen to pines in the Black Forest (left
young spruces. No fertilizer was applied to the left-hand shoot - no fertilizer). After the application of the fertilizer to the 190-year-
(Baule and Fricker, 1967). old trees, the needles became denser and the tree ring widths
doubled (Assmann, 1961).
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
The following example describes the effects of applica- the density sank from 0.475 glcc to 0.412 glcc and
tion of fertilizer (von Pechmann, 1960). A 63-year-old the compression strength showed a parallel reduction
spruce growing on variegated sandstone was treated with from 513 kg/cm 2 to 410 kg/cm 2
nitrogen and phosphate fertilizer (between 1953 and
1955). As early as one year after the first application It seems that cambial activity had been stimulated by the
changes could be seen in the annual ring structure: nitrogen. It follows that the yield of forests on acid soils
with poor nutrient supply can be increased by the cUltiva-
the width, especially of the earlywood increased tion of nitrogen-assimilating plants, e.g. lupin, on the site.
greatly.
the latewood density was lower than in the previous
year.
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).
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
alt itude
20
I
annual rings Ipresent' absent present
drainage good I bad good
climate sub- I sub.-trop. tropical
trop. I
.>0 C C7
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 (Johnson and Siccama, 1983; Schweingruber et al.,
future will be to work on the problems of environmental 1983). Sudden reductions in growth should be investi-
pollution. Finding the answers to the following questions gated when localized and regional instances of forest
will be of paramount importance: damage occur because they are an indication of some
kind of disturbance. Because recognition, dating, and
How do different species of trees on different sites quantifying of disturbance are simple, such an investiga-
behave when subjected to various kinds of pollution? tion gives important results very quickly.
When and where did growth disturbances in trees first Every site can provide the following information based
become evident and in which areas? on 20-30 trunk cross sections or core samples.:
Dendrochronology has on the one hand to supply the Percentage of trees suffering from growth reduction.
information required for political decisions to be made Age distribution of trees with growth reduction; be-
and, on the other, to carry out biological follow-up studies ginning and duration of phases of inhibited growth.
of a political-technical nature. In this respect dendro- Connection between growth reduction and Ule out-
chronologists are in the hot seat in the current proble- ward appearance of the tree, e.g. loss of needles,
matic situation. outward signs of damage, its performance compared
with other trees in the same forest.
Methods Relationship between damage and site, e.g. hydro-
Most dendrochronological studies to date have been con- logical conditions, elevation, proximity to buildings
cerned with solving historical, ecological, and climatologi- emitting toxic substances. All available dendrochrono-
cal problems by means of statistically random sampling. logical methods are being used in research into pollu-
For this purpose efficient but time-consuming measure- tion, including methods of identifying chemical trace
ment and calculation techniques have been developed. In elements in tree rings (Baes III and McLaughlin, 1984).
the current situation, where large areas of forest are
dying, it is often necessary to define the areas where
{...
regions
damage is occurring by means of large scale sampling. reductions
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)
{,-,.
relEases
II
pine . 1 U
Lhw
f. .1 .. II JU
r"
JLL
spruc(~ A:,'g 'u
.. ilii.
5010
'tIUII1
IJ LU/ .1.111
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 forests not cultivated for commercial reasons in the
Europe areas of forest are suffering from damage, the U.S.A. over half the trees have died in some localities in
cause of which is not clear (Johnson and Siccama, 1983; the 30 years since the sudden growth reduction set in. In
McLaughlin et aI., 1983; Schweingruber et aI., 1983; the European forests, which are important commercial
Hornbeck and Smith, 1985). In both areas an abrupt concerns, conditions may be similar but are not so obvi-
growth reduction and a high mortality rate have been ous as the dead trees are felled, albeit prematurely, by
detected in conifers. The trees on both continents began foresters. In both cases structural changes in the forests
to show signs of damage around the 1950s but the full must take place. Johnson et al. (1983) discovered that in
extent of the damage was not recognized until the 1970s, certain stands in Vermont there were about 40% fewer
when the high mortality rate became obvious. In the 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
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).
- 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 that the trouble began with the introduction of motorized
to grow in than the present-day town, inundated with vehicles at the beginning of the century. Eckstein et al.
traffic. The vitality of the tree is sapped in various ways: (1974) found that the composition of woody tissue and
by damage to its roots, trunks, and crown, when the soil also the resulting physiological conditions were changed
around it is compacted, covered over, or contaminated by with decreasing cambial activity.
hydrocarbon gas, oil, or salt, and by air pollution (Hoester, Horticulturalists, traffic planners, and biologists are
1977). It is thought that about half the trees in the cities of cooperating in an attempt to prevent further air pollution in
Europe are threatened with winter salting of roads being towns.
the main cause. Petersen (1983) and Joos (1985) indicate
Cu Hlng green branches (a) Effect of cutting off branches form the crown of plane
.J: _10 5 E trees (Platanus X hybrida) (arrows). The crowns of the
~~~8 4E:~ trees were cut at irregular intervals of between 4 and 10
eg"'6 3-::E years. As a result radial growth dropped, in the short
i~~4 2~~ term, from 4 to 2 mm but one or two years later, the trees
ci ~.& 2 1 ~~ had already recovered.
c 0 l:I:=l..LJ:-~Jt=t=lJ:l~tr.~JC:~::3...L_--L 0 ~
1920 30 40 50 1960 (b) Effect of building operations. During the building of an
bl ROOl damages and soil compressIon 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.
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).
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 [>
"
"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.)
The most frequently asked question when forest damage son and 8iccama, 1983). Results are beginning to come
is being discussed is 'What causes it?' Dendrochronol- in on the subject of the CO 2 effect. Correlations are
ogists are trying to correlate statistically the connection coming to light regarding the connection between 80 2
between the timing and degree of severity of damage, emissions within small areas, tree-ring width suppression,
and the spread of emissions, by means of tree-ring and the chemical constituents of the rings.
analysis. The diagrams below represent an attempt to As many more results are expected in the near future,
correlate available knowledge but should not be taken as only a few examples - partially controversial - are given
proof positive. An attempt to find a direct connection here to indicate the direction of current research.
between acid rain and annual ring patterns failed (John-
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
~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
/
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 the age of organic substances. Libby was awarded a
and proved the theory that cosmic radiation produced Nobel prize for his work in this field in 1960. In the 'fifties
unstable radioactive carbon C4 C) which was assimilated counting techniques were improved and thousands of
by plants in the process of photosynthesis and incor- measurements were taken. By the early 'sixties it was
porated in organic compounds. Since all living things are clear that 14C datings for the period before 1000 B.C. did
directly or indirectly nourished on plant substances, radio- not agree with the dendrochronological absolute ones.
active carbon must also be present in the bodies of all Libby suspected at first that mistakes in dendrochrono-
animals, including humans. logical datings caused by false annual rings were the
Libby calculated the half-life of 14C to be 5568 years. reason. However investigations on dated tree-ring series
On the basis of the state of knowledge at that time, it verified the deviation from the physical assumption of the
was assumed that a sample with a 14C content of only constant half-life.
50% was 5568 years old and one with a 25% 14C con- 13r---------------------------------,
tent was 11140 years old. As the concentration of radio-
active carbon compared with nonradioactive is very small,
(1 : 1000000000000) measuring instruments for testing samples of known age
12
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
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).
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~
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 conditions in various parts of the world . As absolute
age comprising only 1-10 annual rings, using the most chronologies, extending back over thousands of years,
modern technical equipment, has led to the knowledge have been constructed in California, Ireland and southern
that a long-term deviation is superimposed by a more or Germany, work on the calibration of 14C curves is going
less irregular short-term one. These so-called 'wiggles' on in several laboratories.
have a world-wide application for they can be detected in Stui ver (1982) calibrated the last 2000 years, and
a wide variety of trees , growing under very different 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
,.c content of tree rings in Douglas firs (Pseudotsuga menziesii) ,.c content in tree rings of Douglas firs on various sites on the
(circles) and Irish oaks (crosses). The level in both types is practi- Pacific coast of the U.S.A. from 1000 to 1880 in relation to the oxalic
cally identical. The relatively short-term, synchronous ,.c variations acid standard measurement (O-line). The deviations (called wiggles)
In the northern hemisphere are shown. The length of the vertical can be clearly seen (Stuiver et al., 1980).
lines denotes standard deviations (Stuiver, 1982, in press).
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 It is difficult to furnish proof of an 11 year cycle of sunspot
reasons for the short and long-term variations: "a solar or activity as the amplitude of the relevant 14C variations
geomagnetic modulation of the cosmic rays or a fluctua- is very small. A relationship between solar activity and
tion in the carbon cycle and the associated dynamic climate is also difficult to prove (Stuiver and Quay, 1980)
exchange, which bring about a change in the CO 2 content but according to Suss (1979) such a possibility cannot be
of the atmosphere." ruled out.
The short-term 14C fluctuations in the tree rings cor- Long-term variations are to be found only in the 14C
respond to the beryllium-10 COBe) in ice in Greenland. content in the tree rings and not in the 10Be content of
Both fluctuations bear a close correlation to the medium- Greenland ice. Oescher (1985) explained this as follows:
term sunspot cycle of the last 800 years. The short-term "Ocean currents have become stronger since the end of
fluctuations in the 14C content are certainly due to the last Ice Age which led to an equalization in the 14C
changes in the rate at which it was formed. By measuring content in shallow and deep sea water. This has been
the 14C content in the tree ring accurately and by taking responsible for a drop in the 14C content of the atmo-
the different lengths of time that the radioactive carbon sphere."
remains in the atmosphere, biosphere and in the oceans,
medium-term changes in solar activity can be determined.
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
~
250
Hydrogen bomb explosions The consequences for historical research
There has been an incease in the level of 14C in tree rings Present hypotheses on the sequence of events in pre -
which is clearly due to the hydrogen bomb tests of the historic times are based to a large extent on uncorrecte,d
sixties. When heartwood is formed, organic substances dates, determined by radiocarbon methods. Taking long-
are transported to the inner part of the trunk, leading to term variations into account, a start has been made on
additional biologically determined variations in the 14C correcting the errors, but this has led to confusion among
content. 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
{l"e jOioo) existing chronological systems until the variations in time
800 r--------------------------------------,
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-
700 /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 Relationship between dates determined by tree ring measurements
means of 14C measurements. In fast-growing tropical trees the bomb and those by means of radiocarbon dating. (dark band). As 14C
peak of between 1962 and 1964 can be determined, using 14C mea- levels in the atmosphere vary, the relation between annual ring
surements. The degree of 14C contamination in the trunk provides datings and 14C datings is not linear. This has various consequences
Information on the rate of growth (Stuiver et al., 1981). 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).
lO
7 ! _
,..... •...........
.. J
.... ...
«>
-4
,.,.
~ \ .. ~ 6 , 8 ..,
Rthft SDj"'-l JuI"IQ'(I
Relationship between the 0 ' .0 in precipitation and cellulose in tree Relationship between the deuterium content of the cell sap and
rings in various conifers on the Pacific coast of the U.S.A. according cellulose nitrate in Pinus Strobus in New York at various pOints
to Burk and Stuiver, 1981 (diagram on the left), and in conifers and in time during the vegetation period. A close connection exists
peat in the Rocky Mountains (diagram on the right) (Gray et al., between both parameters (White and Lawrence, 1980).
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
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 at a later stage during the development of heartwood.
to disguise the fact that often isotopic measurements on Together with these biological problems, various unsolved
annual rings do not reveal any clear connection to en- technical ones exist too.
vironmental factors. Long (1981) points out that many The positive results lead us to suppose that in spite of
problems remain unsolved. When CO 2 and water undergo obvious gaps in knowledge, it is possible to collect in-
biological changes certain processes take place, some of formation on the environment. Our present state of
which are temperature-dependent, some not. Physical knowledge suggests that local site factors and differences
signals are masked either in the cellulose, the lignin or between the various species of trees influence the separa-
throughout the wood during the formation of new wood or 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.
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-
256
By the beginning of the First World War the main ring widths with the corresponding weather data and
foundations of tree-ring research had been laid. Antevs found a strong measure of agreement.
(1917), lists over 300 papers on this topic in the period In 1892 the Russian, F. N. Shvedov, working in Odessa,
1840-1917. Until about 1920, however, actual dendro- found a clear correlation between annual ring widths of
chronological studies, in which tree-ring sequences were Robina pseudoacacia and precipitation values. (Shvedov
compared with long-term meteorological records, were in Fletcher and Linnard , 1977).
rare. According to Holmsgard (1956), the first annual ring
In 1859 the German immigrant, Jacob Keuchler, living analytical studies in Denmark were carried out by Revent-
in Texas, was interested to know whether the drought low, 1879, and Lutken, 1891. According to Schulman
occurring at the time was an isolated event, or whether he (1937), Kaptein (1914) recognized the relationship be-
had to be prepared for periodic water shortages in his tween the weather and annual ring width. In the rest of
new homeland (Stallings, 1937). He tried to find the Scandinavia the first dendroclimatological studies were
answer to this question by studying the tree rings of trees conducted by Laitakari (1920), and von Eide (1926).
on exposed sites in the neighborhood. (Hoeg 1956.)
In 1869 Pokorny compared the 10-year mean of tree-
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.
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).
259
Dendrochronology in Europe
It is remarkable that the new science was not immediately tists, or did archaeologists and biologists working in tem-
adopted by others, in spite of the spectacular results perate and boreal zones fail to recognize the potential of
achieved. Was it because ring sequences of trees in tree ring analysis? Not until the end of the thirties did the
temperate and boreal zones are less irregular? Was there principles gain a foothold in Europe. Bruno Huber was the
too little contact between American and European sci en- man who helped to make the breakthrough.
(
Bruno Huber, about 1965.
Andrew E. Douglass and Bruno Huber laid the founda- In 1963, Harold C. Fritts of Tucson , Arizona, intro-
tions of modern dendrochronology. Spurred on by their duced computers as an aid to modern statistical
ideas, new laboratories were opened in various parts of methods in dendroclimatology. His findings are now
Europe from 1950 onwards. By 1970 Eckstein listed 20; used, in one form or another, by every tree ring
today there could be more than 30. In both the east and laboratory.
west of Russia several laboratories are working on tree In the same year Hubert Polge of Nancy, France,
ring analysis (Fletcher and Linnard , 1977). invented a method of measuring wood density using
In North America development was slow. Was no X-rays. It soon became apparent that particularly the
laboratory other than Tucson interested? Or were too few latewood densities of trees in temporate and boreal
willing to compete with the large, successful Tree Ring zones provide much more climatological information
Laboratory? The importance of tree ring research for the than the annual ring widths. Only a few dendrochrono-
life and historical sciences is now recognized in every logical laboratories are equipped with the necessary
continent. In recent times two developments have given apparatus for the application of this method although
dendrochronological research a great impetus: 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:
261
Bibliography 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.
Alestalo, J., 1971: Dendrochronological interpretation of Balteau, O. J.: Radiographie par rayons X mous. Balteau
geomorphic processes. Societas geographica Fenniae company catalog.
105, 1-140. Baltensweiler, W., 1976: Die Massenvermehrung des
Alestalo, J., and Haikio, J., 1976: Ice features and ice- grauen Larchenwicklers in den Alpen. Int. Ber. Entom-
thrust shore forms at Luodonselka, Gulf of Bothnia in ologisches Institut, ETH Zurich, 15 pp.
Winter 1972/73, Fenniae 144, 5-24. Bannister, B., 1963: Dendrochronology in the Near East:
Alexander, M. E., 1973: Bibliography and resume of Current research and future potentialities. Abstract and
current studies on fire history. Supplements 1980 and report. VII. Congres international des sciences anthro-
1983. Great Lakes Forest Research Center. Sault St. pologiques et ethnologiques. V, 336-340.
Marie, Ontario, Canada. 52 pp. Bannister, B., 1965: Tree-ring dating of archaeological
Anonymous, 1962: Andrew Ellicott Douglass. Obituary sites in the Chaco Canyon region. New Mexico. South-
and Bibliography. Tree-Ring Bull. 24, 2-10. west Parks and Monuments Association. Technical
Anonymous, 1979: Die Schweiz und ihre Gletscher; von Series 6(2).
der Eiszeit bis zur Gegenwart. Hrgs. Schweiz. Ver- Bannister, B., and Robinson W. J., 1975: Tree-ring dating
kehrszentrale. 191 pp., Bern, Kummerly und Frey. in archaeology. World Archaeology 7,210-225.
Anonymous, 1980: Mount St. Helens Holocaust. The Baer, 0., 1977: Geographie Europas. Zurich, Lehrmittel-
Columbian, Inc., P. O. Box 180, Vancouver, Wash. 64 verlag des Kantons Zurich. 315 pp.
pp. Bauch, J., Eckstein, D., and Meier-Siem, M., 1972: Dating
Anonymous, 1981: Saurer Regen uber Deutschland. Der the wood panels by a dendrochronological analysis of
Wald stirbt. Der Spiegel 35(45), 96-110. tree-rings. Nederl. Kunsthist. Jaarboek 23, 485-496.
Antevs, E., 1917: Die Jahresringe der Holzgewachse und Bauch, J., Eckstein, D., and Brauner, G., 1978: Den-
die Bedeutung derselben als klimatischer Indikator. drochronologische Untersuchungen an Eichenholzta-
Progr. Reih. Bot. 5, 285-386. feln von Rubens-Gemalden, Jahrbuch Berliner Museen
Arno, S. F., and Sneck, K. M., 1977: A method for 20, 209-221.
determining fire history in coniferous forests of the Bauch, J., and Eckstein, D., 1981: Woodbiological inves-
mountain west. USDA Forest Service. Gen. Techn. tigations on panels of Rembrandt Paintings. Wood Sci.
Rep.INT-12. Technol. 15, 251-263.
Arnold, J. R., and Libby, W. F., 1949: Age determinations Baule, H., and Fricker C., 1967: Die Dungung von Wald-
by radiocarbon content: Checks with samples of known biiumen. Munchen, Basel, Wien, BLW-Verlag. 259 pp.
age. Science 110, 678-680. Baumann, F., 1949: Die freilebenden Saugetiere in der
Artmann, A., 1949: Jahrringchronologische und -klimatolo- Schweiz. Bern, Huber. 492 pp.
gische Untersuchungen an der Zirbe und andern Bazzigher, G., 1973: Wundfaule in Fichtenwaldungen mit
Baumen des Hochgebirges. Dissertation, Univ. Munich. alten Schalschaden. Eur. J. For. Phat. 3, 71-82.
Ash, S. R., and May, 0. D., 1981.' Petrified forest. The Becker, B., and Schirmer, W., 1977: Paleoecological
story behind the scenery. Petrified Forest. Nat. Park, study on the Holocene valleys development of the river
Arizona, Holbrook. 32 pp. Main, Southern Germany. Boreas 6,303-321.
Assmann, E., 1961: Waldertragskunde. Organische Pro- Becker, B., 1983: Dendrochronologie und Palaookologie
duktion, Struktur, Zuwachs und Ertrag von Waldbe- subfossiler Baumstamme aus Flussablagerungen. Ein
standen. 490 pp., Munchen, Bayer. Beitrag zur nacheiszeitlichen Auenentwicklung im sud-
lichen Mitteleuropa. Mitt. Komm. Ouartarforschung der
Baas, P., Bauch, J. (ed.) 1986: The effects of environ- osterr. Akad. Wiss. 5, 120 pp.
mental pollution on wood structure and quality. IAWA- Becker, B., Billamboz, A, Egger, H., Gassmann, P., Orcel
Bull. 7, n.s. 267-415. Internat. Assoc. Wood Anato- A, Orcel, Chr., and Ruoff, U., 1985: Dendrochronologie
mists. Leiden, Netherland. in der Ur- und Fruhgeschichte. Die absolute Datierung
Bass, P., Bolton, A J., and Catling, D. M., 1976: Wood von Pfahlbausiedlungen nordlich der Alpen im Jahrring-
structure in biological and technological research. In: kalender Mitteleuropas. Antiqua 11, Veroff. der schweiz.
Proc. of the Anglo-Dutch wood anatomy meeting. Ges. Ur- und Fruhgeschichte. 68 pp.
Oxford and Kew. 5th-8th April 1976, 280 pp., Leiden, Beckwith, R. C., 1978: Biology of the insect. pg. 25-36.
Univ. Press. In: The Douglas-Fir Tussock-Moth, a Synthesis. Forest
Baes III, C. V., and McLaughlin, S. B. 1984: Trace Ele- Service, Science and Education Agency, Technical
ments in tree Rings: Evidents of recent and historical air Bulletin No. 1585. U.S. Dept. of Agriculture. Washing-
pollution. Science 224,494-497. ton D.C. 391 p.
Baillie, M. G. L., 1982: Tree-Ring Dating and Archaeol- Bednarz, Z., 1984: The comparison of dendroclimato-
ogy. Croom Helm, London and Canberra. 274 pp. logical reconstructions of summer temperatures from
Baillie, M. G. L., 1984: Is there a single British Isles Oak the Alps and the Tatra Mountains from 1741-1965.
tree-ring signal? Proc. 22nd Symp. on Archaeometry Dendrochronologia 2, 63-72.
(A. Aspinall and S. E. Warren) at 73-82, Univ. of Brad- Bircher, W., 1982: Zur Gletscher- und Klimageschichte
ford. des Saastales. Glazialmorphologische und klimatolo-
Baillie, M. G. L., and Pilcher, J. R., 1973: A simple cross- gische Untersuchung. Physische Geographie, Universi-
dating program for tree-ring research. Tree-Ring Bull. tat Zurich. 9, 233 pp.
33,7-14.
262
Blais, J. R., 1964: Account of a recent Spruce Budworm Christensen, D. J., 1977: Lindbergh Kidnapping - The
outbreak in the Laurentide Park Region of Quebec and ladder link. Forests and People 27, 3-11.
measures for reducing damage in future outbreaks. Coles, J. M., 1981: Conservation of wooden artefacts
Forestry Chronicle 40, 313-323. from the Somerset Levels: 3. Somerset Levels Papers
Blasing, T. J., and Fritts, H. C., 1976: Reconstructing past 7,70-80.
climatic anomalies in North America from tree-, ing data. Cropper, J. P., 1979: Tree-ring skeleton plotting by
Quaternary Res. 6, 563-579. computer. Tree Ring Bull. 39, 47-60.
Bolsinger, Ch. L., 1978: The extent of dwarf Mistletoe in Cutler, D. F., 1976: Variation in root wood anatomy. In:
six principal softwoods in California, Oregon and Wash- Proc. Anglo-Dutch wood anatomy meeting. Oxford and
ington. Proceedings of the Symposium on Dwarf Mistle- Kew, 5th-8th April 1976. Leiden, Univ. Press. 143-
toe control through forest management. Pacific South- 176.
west Forest and Range Experimental Station. General Currie, R. G., 1982: Evidence for a 18.6 year MN term in
Technical Report PSW-31 , 45-54. air pressure in Japan and geophysical implications.
Borman, H., and Berlyn, G., eds. 1981: Age and growth Geophys. J. R. Astron. Soc. 69,321-327.
of tropical trees: New directions for research. Yale Uni-
versity: School of Forestry and environmental studies. Dean, J. S., 1969: Chronological analysis of Tsegi phase
New Haven: Yale University. Bull. No. 94. sites in northeastern Arizona. Papers Lab. Tree-Ring
Bosshard, H. H., 1974-1975: Holzkunde. 3 Vols. Basel Res. 3, 207 p., Tucson, Arizona University Press.
und Stuttgart, Birkhauser. Dean, J. S. and Warren, R. L., 1983: Dendrochronology in
Braeker, O. U., and Bill, J. (Eds.), 1979: Zum derzeitigen S. H. Lekson (edit): The Architecture and Dendrochro-
Stand der Nassholzkonservierung. Zeitschr. schweiz. ·nology of Chetro Ketl, Chaco Canyon, New Mexico.
Archaologie u. Kunstgesch. 36, 97-145. Nat. Park Service, U.S. Dept. of the Interior. Albu-
Braeker, O. U., 1981: Der Alterstrend bei Jahrringdichten querque, New Mexico 1983, 356 pp.
und Jahrringbreiten von Nadelhblzern und sein Aus- Dean, J. S., in press: Dendrochronology and paleoen-
gleich. Mitt. forst/. Bundesvers anst. Wien. 142, 75- vironmental reconstruction on the Colorado Plateaus. In
102. The Anasazi and Their Environment, edited by George
Brehme, K., 1951: Jahrringchronologische und -klimatol- J. Gumerman. School of American Research Press,
ogische Untersuchungen an Hochgebirgslarchen des Santa Fe.
Berchtesgadener Landes. Ztschr. Weltforstwirtsch. 14, Delorme, A., 1973: Uber die Reichweite von Jahrring-
64-80. chronologien unter besonderer Berucksichtigung mittel-
Brenninkmeijer, C. A. M., 1983: Deuterium Oxygen-18 europaischer Eichenchronologien. Prahist. Zeitschr. 48,
and carbon-13 in tree rings and peat deposits in relation 133-143.
to climate. Dissertation, Rijksuniversiteit Groningen, Dietrich, J. H., and Swetnam, T. W., 1982: Pyrodendro-
Holland. 146 pp. chronology of fire-scarred Ponderosa Pine. 30, 238-
Bucharat, B., 1978: Oxygen isotope palaeotemperatures 247. Forest Science. 16 p.
from Tertiary period in the North Sea area. Nature 275, Douglass, A. E., 1929: The secret of the Southwest
121-123. solved by talkative tree rings. Nat. Geogr. Mag. 54,
Burger, H., 1947: Blattmenge und Zuwachs. VIII. Mittei- 737-770.
lung. Die Eiche. Mitt. Eidg. Anstalt fur das forstliche Douglass, A. E., 1935: Dating Pueblo Bonito and other
Versuchswesen 25,207-279. ruins of the Southwest. Washington, Nat. Geogr. Soc.,
Burk, R. L., and Stuiver, M., 1981: Oxygen isotope rations Pueblo Bonito Ser. 1, 1-74.
in trees reflect mean annual temperature and humidity. Douglass, A. E., 1971: Climatic cycles and tree-growth.
Science 211,1417-1419. Cramer Lehre (New impression of the work of 1919,
Cain, W. F., 1979: 14C in modern American trees. In: R. 127 pp.; 1928, 166 pp.; 1936, 171 pp.).
Berger and H. E. Suss (Eds.). Radiocarbon dating. Proc. Duchaufour, P., 1970: Precis de pedologie. 481 pp.,
9th Int. Cont., Los Angeles and La Jolla, 1976, 495- Paris, Masson & Cie.
510.
Campbell, T. N., 1949: The pioneer tree-ring work of Echols, R. M., 1973: Uniformity of wood density, as-
Jacob Kuechler. Tree-Ring Bull. 15, 16-20. sessed from x-rays of increment cores. Wood Sci.
Carlquist, S., 1980: Further concepts in ecological wood Techn. 7, 34-44.
anatomy, with comments on recent work in wood Eckstein, D., 1972: Tree-ring research in Europe. Tree-
anatomy and evolution. Aliso 9, 499-553. Ring Bull. 32, 1-18.
Ceram, C. W., 1972: Der erste Amerikaner. 373 pp., Eckstein, D., and Bauch, J., 1969: Beitrage zur Rationa-
Zurich, Ex Libris. lisierung eines dendrochronologischen Verfahrens zur
Chaloner,.w. G., and Creber, G. T., 1967: In D. H. Tarling Analyse seiner Aussagesicherheit. Forstwiss. Central-
and S. K. Runcorn (Eds.). Implication of continental drift blatt 88, 230-250.
for the earth sciences. London, Academic Press. Eckstein, D., Frisse, E., and Liese, W., 1974: Holzanato-
Creber, G. T., and Chaloner, W. G., 1984: Influence of mische Untersuchungen an umweltgeschadigten Stras-
environmental factors on the wood structure of living senbaumen der Hamburger Innenstadt. Eur. J. For.
and fossil trees. Bot. Rev. 50,357-448. Path. 4, 232-244.
Creber, G. T., and Chaloner, W. G., 1987: The contribu- Eckstein, D., Grote, R.-J., and Mathieu, K., 1977: Den-
tion of growth ring studies to the reconstruction of past drochronologische Untersuchungen zur landlichen und
climates. In: R. G. W. Ward, Applications of tree-ring stadtischen Architektur Hamburgs im 15.-18. Jahrhun-
studies. BAR. Internat. Series 333.37-67.
263
dert. Deutsche Kunst- und Denkmalpflege. Jahrgang Fletcher, J., 1978: Tree-ring analysis of panel paintings.
1977,33-74. In: Dendrochronology in Europe. BAR Int. Series 51,
Eckstein, D., Breyne, A, Aniol, R. W., and Liese, W., 303-306.
1981: Dendroklimatologische Untersuchungen zur Ford, E. D., and Robards, A W., 1976: Short term varia-
Entwicklung von Strassenbaumen. Forstw. Cbl. 100, tion in tracheid development in early wood of Picea
381-396. sitchensis. Leiden Botanical Series 3, 212-221 .
Eckstein, D., Neugebauer, M., and Brauner, G., 1982: Fox, C. A., and Nash III, T. H., 1980: The effect of air
Die Baugeschichte der Holzkonstrukturen im Heiligen- pollution on western larch as detected by tree-ring
Geist-Hospital zu Luebeck. Lubecker Schriften zur analysis. Symposium on Effects of Air Pollutants on
Archiiologie und Kulturgeschichte. 6, 123-162. Mediterranean and Temperate Forest Ecosystems.
Eckstein, D., 1983: Overview of existing tree-ring chro- June 22-27, 1980, Riverside, California U.S.A.
nologies in Europe. In: D. Eckstein, D. Wrobel und R. Francuz, J., 1980: A brief explanation to the Twann
W. Aniol (eds.): Dendrochronology and Archaeology in Horgen dendro-work. Vortrag an der Jahressitzung der
Europe. Mitt. Bundesforschungsanstalt fUr forst- und Arbeitsgemeinschaft fur die Urgeschichtsforschung der
Holzwirtschaft, Hamburg 414, 125-136. Schweiz. Unpublished, 5 pp.
Egger, H., and Lambert, J., (unpublished): Dendrochron- Frey-Wyssling, A., 1959: Die pflanzliche Zellwand. Berlin,
ologische Bibliographie. Laboratoire de Dendrochronol- Gbttingen, Heidelberg, Springer, 367 pp.
ogie du Musee Cantonale d'Archeologie de Neuchatel. Friedman, I., and Gleason, J., 1980: Deuterium content of
Eide, E., 1926: Uber Sommertemperatur und Dicken- lignin and the methyl and hydrogen of cellulose. In: G.
wachstum im Fichtenwald. Medd. Norske Skogforsoks- C. Jacoby, (Ed.) Carbon Dioxide Effects: Res. and
vesen 2, 87-104 (Norwegian). Assessment Progr. Proc. Int. Meeting on Stable Iso-
Eidem, P., 1955: Dendrochronological dating of a cast- topes in Tree-Ring Res., New York, Lamont-Doherty
house from Istad, Slidre, Valdres. Blyttia 13, 65-70 Geol. Observ., Columbia University. 50-55.
(Norwegian). Fritts, H. C., 1965: Tree-ring evidence for climatic
Eklund, B., 1956: Variations in the width of the annual changes in western North America. Monthly Weather
rings in pine and spruce due to climatic conditions in Rev. 93, 421-443.
northern Sweden during the years 1900-1944. Tree- Fritts, H. C., Smith, D. G., Cardis, J. w., and Budelsky, C.
Ring Bull. 21, 21-25. A, 1965: Tree-ring characteristics along a vegetation
Ellenberg, H., 1965: Vegetation Mitteleuropas mit den gradient in Northern Arizona. Ecology 46, 393-401.
Alpen. 843 pp., Stuttgart, Ulmer. Fritts, H. C., 1969: Bristlecone pine in the White Mountain
Ellenberg, H., 1967: Vegetations- und bodenkundliche of California. Tucson, Arizona, Univ. Press., Pap. Lab. of
Methoden der forstlichen Standortkartierung: Ergeb- Tree-ring Res. 4, 44 p.
nisse eines into Methodenvergleichs im Schweizer Mit- Fritts, H. C., 1974: Relationships of ring width in arid-sites
telland. 296 pp., Zurich, Geobot. Inst. ETH, Stiftung conifers to variations in monthly temperature and pre-
Rubel. cipitation. Ecological Monographs 44, 411-440.
Ellenberg, H., 1978: Vegetation Mitteleuropas mit den Fritts, H. C., 1976: Tree rings and climate, London, New
Alpen in 6kologischer Sicht. 981 pp., Stuttgart, Ulmer. York, San Francisco, Academic Press, 567 pp.
Esau, K., 1969: Plant anatomy - Pflanzenanatomie. 594 Fritts, H. C., 1977: Tree rings: A record of past climate.
pp., Stuttgart, Fischer. US Dept. of Commerce, EDS, 31-42.
Everitt, B. L., 1968: Use of the cottonwood in an investi- Fritts, H. C., Lofgren, G. R., Gordon, G. A, 1979: Varia-
gation of the recent history of a flood plain. American tions in climate since 1602 as reconstructed from tree
Journal of Science, 1969,417-438. rings. Quaternary Res. 12,18-46.
Evertsen, J. A, 1982: Interlaboratory standardisation sur- Fritts, H. C., 1983: Tree-ring dating and reconstructed
vey. Wood Microdensitometry Bulletin 2, 3-24. variations in Central Plains climate. Transactions of the
Nebraska Academy of Sciences XI (special Issue), 37-
Fahn, A, 1974: Plant Anatomy, 2nd Ed., Reprint 1977. 41.
Oxford etc. Pergamon Press. Fritts, H. C., and Lough, J. M., 1984: An estimate of aver-
Fahn, A, Burley, J., Longman, K. A, Mariaux, A, 1981: age annual temperature variations for North America
Possible contributions of wood anatomy to the deter- 1602-1960.
mination of the age of tropical trees. In: F. H. Borman, Fritts, H. C., and Swetnam, T. W., 1986: Dendroecology:
G. Berlyn (eds.) Age and growth rate of tropical trees: A tool for evaluating variations in the past and present
New directions for research. Yale University: School of forest environments. Lab. of Tree-Ring Research, Univ.
forestry and environmental studies. Bull. 94, 31-54. of Arizona 85721, Tucson, Arizona, 61 pp.
Ferguson, C. W., 1968: Bristlecone pine: science and Fritts, H. C., and Wu Xiangding, 1986: A comparison
esthetics. Science 159, 839-846. between response - function analysis and other re-
Ferguson, C. W., and Graybill, D. A, 1983: Dendrochro- greSSion techniques. Tree-ring Bull. 46, 31-46.
nology of Bristlecone pine: A progress report. Radio- Furrer, G., Gamper-Schollenberger, B., and Suter, J.,
carbon 25, 287-288. 1980: Zur Geschichte unserer Gletscher in der
Firbas, F., 1949-1952: Spfit- und nacheiszeitliche Wald- Nacheiszeit - Methoden und Ergebnisse. In: H.
geschichte Mitteleuropas nordlich der Alpen. 2 Vols. Oeschger, B. Messerli, M. Svilar (Eds.) Das Klima.
Jena, Fischer. Analysen und Modelle. Geschichte und Zukunft. Berlin,
Fletcher, J. M., and Linnard, W., 1977: Russian papers on Heidelberg, New York, Springer. 296 pp.
dendrochronology and dendroclimatology, 142 pp.,
Oxford, Res. Lab., Arch. and History of Art, University
Press.
264
Gassner, G. and Christiansen-Weniger, E., 1942: Den- Hawksworth, F. G., Hinds, T. E., 1964: Effects of Dwarf
droklimatologische Untersuchungen uber die Jahresrin- Mistletoe on immature Lodgepole pine stands in Colo-
gentwicklung der Kiefern in Anatolien. Nova Acta Leo- rado. J. Forestry 62, 27-32.
poldina 12, Halle. 1-137. Haselhoff, E., Lindau, G., 1903: Die Besch8digung der
Geiger, R., 1961: Das Klima der bodennahen Luftschicht; Vegetation durch Rauch. Berlin, Leipzing, Borntrager.
ein Lehrbuch der Mikroklimatologie. Braunschweig. 4. 412 pp.
Aufl. 646 pp. Hawksworth, F. G., 1961: Dwarf Mistletoe of Ponderosa
Gothan, W., and Weyland, H., 1973: Lehrbuch der Palao- Pine in the Southwest. Tech. Bull. no. 1246, US Dept.
botanik. Munich etc. BLV Verlagsgesellschaft. 677 pp. of Agriculture 111 pp.
Gottwald, H., 1958: Handelshblzer, Hamburg. Hawksworth, F. G., and Hinds, T. E., 1964: Effects of
Graslund, M., 1984: The history of Dendrochronology in Dwarf Mistletoe on immature Lodgepole pine stands in
the Nordic Countries. Dendrochronologia 2, 31-62. Colorado. J. Forestry 62, 27-32.
Gray, J., and Thompson, P., 1980: Natural variations in Hawksworth, F. G., and Wiens, D., 1972: Biology and
the 18 0 content of cellulose. In: G. C. Jacoby, (Ed.) classification of Dwarf Mistletoe (Arceuthobium). Agri-
Carbon Dioxide Effects; Res. and Assessment Pro- culture Handbook no. 401, Forest Service, US Dept. of
gram, Proc. Int. Meeting on Stable Isotopes in Tree- Agriculture.
ring Res., New York, Lamont-Doherty Geol. Observ., Hecht, A D., (Ed.) 1979: Paleoclimatic research: Status
Columbia Univ., 84-92. and opportunities. Quaternary Res. 12,6-17.
Graybill, D. A, 1982: Chronology development and analy- Heer, 0., 1865: Die Urwelt der Schweiz. Schulthess,
sis. In: Hughes, M. K., Kelly, P. M., Pilcher J. R., and .Zurich, 713 pp.
LaMarche V. C. (eds.). Climate from tree rings. Cam- Hegi, G., 1957: ///ustrierte Flora von Mittel-Europa s.
bridge, Univ. Press London, 21-31. Band 3, Teil 1, Aufl. 2, 452 pp. Munchen, Hanser.
Green, H. V., 1965: The study of wood characteristics by Heikkinen, 0., 1984: Dendrochronological evidence of
means of a photometric technique. In: Proc. meeting of variations of Coleman Glacier, Mount Baker, Washing-
working groups on wood quality, sawing and mechin- ton, USA Arctic and Alpine Research 16, 53-64.
ing, wood and tree chemistry. Ed. Section 41, IUFRO, Helley, E. J., and LaMarche, V. C., 1973: Historic flood
Melbourne, 4th-15th Oct. 1965, vol. 2, 12 p., Mel- Information for Northern California streams from geo-
bourne: CSIRO, Div. forest products. logical and botanical evidence. Geol. Survey prof
Greve, U., 1984: Holzbiologische Untersuchungen an paper 485 - E. 15 p.
Fichten (Picea abies (L.) Karst.) unterschiedlicher Im- Hess, H. E., Landolt, E., and Hirzel, R., 1967-1972: Flora
missionstoleranz und Immissionsbelastung. Disserta- der Schweiz und angrenzender Gebiete. 3 Bande.
tion, Universitat Hamburg, Fachbereich Biologie. 137 Basel, Stuttgart, Birkhauser.
pp. Heyer, E., 1977: Witterung und Klima, 4. Aufl., Leipzig,
Griggs, F. F., 1928: Das Tal der zehntausend Dampfe. Teubner, 460 pp.
Leipzig. Brockhaus. Heyworth, A, 1978: Submerged forests around the
Grootes, M., 1977: Thermal diffusion isotopic enrichment British Isles: Their dating and relevance as indicators of
and radiocarbon dating beyond 50000 years B.P., postglacial land and sea level changes. In: J. Fletcher,
Groningen. 221 pp. Dendrochronology in Europe. BAR Int. Series 51, 279-
Grosser, D., 1977: Die HOlzer Mitteleuropas. Ein mikro- 288.
photographischer Lehratlas. Berlin, Heidelberg, New Hickin, N. E., 1973: The insect factor in wood decay; an
York, Springer. 208 pp. account of wood boring insects with particular reference
Grunhilder, I., 1972: Auf der Spur vorgeschichtlicher to timber indoors. 3 Edition. London, Robin Edwards.
Katastrophen. Griin, Das Gartenmagazin E. 5948. 24- Higashi, S., Fujiwara, K., Araya, T., and Murai, N., 1971:
30. Dendrochronological studies on the transition of the
Grunig, P. E., 1953: Die Flyschaufforstung in wirtschaft- creeping land. Res. Bull. Coli. Experiment Forest,
licher, naturwissenschaftlicher und waldbaulicher Betra- College of Agr. Hokkaido University 28, 339-420.
chtung. Schweiz. Ztschr. Forstwesen 6, 1-16. Hirmer, M., 1927: Handbuch der Palaobotanik. Munchen,
Grunig, P. E., 1955: Ueber den Einfluss der Entwasser- Berlin, R. Oldenburg. 708 pp.
ung auf die Flachmoorvegetation und auf den Zuwachs Hoeg, A 0., 1956: Growth-ring research in Norway.
der Fichte und Bergfbhre im Flyschgebiet der Voralpen. Tree-Ring BUll. 21, 2-15.
Mitt. Schweiz. Anstalt forstl. Versuchswesen. 31, 419- Holdheide, W., Huber, B., 1951: Aehnlichkeiten und
492. Unterschiede im Feinbau von Holz und Rinde. Holz als
Roh- und Werkstoff 10, 262-268.
Haesler, R., 1982: Net photosynthesis and transpiration of Hollstein, E., 1978: Wood technology and the dating of
Pinus montana on east and north facing slops at alpine oak. West German chronologies for oak and beech. In:
timberline. Oecologia (Berf) 54,14-22. J. Fletcher, Dendrochronology in Europe. BAR Int.
Harlow, W. M., Harrar, E. S., 1950: Textbook of dendro- Series 51, 33-44.
logy. London, New York, Toronto, McGraw-HilI. 555 p. Hollstein, E., 1980: Mitteleuropaische Eichenchronologie.
Harms Geographie 1979: Physische Geographie und Trierer dendrochronologische Forschungen zur Archa-
Nachbarwissenschaften, Munchen, 424 pp. ologie und Kunstgeschichte. Mainz a. Rh., Trierer Gra-
Hawksworth, F. G., Wiens, D., 1972: Biology and classifi- bungen und Forschungen 11, 273 pp.
cation of Dwarf Mistletoe (Arceuthobium). Agriculture Holmes, R. L., Stockton, D. W., and LaMarche, V. C.,
Handbook No. 401. Forest Service, US Dept. of Agri- 1979: Extension of river flow records in Argentina from
culture. living tree-ring chronologies. Water Resources Bull. 15,
1081-1085.
265
Holmsgard, E., 1956: Tree-ring analysis of Danish forest Johnson, A H., Siccama, T. G., 1983: Acid deposition
trees. Tree-Ring Bull. 21, 25-27. and forest decline. Environmental Science and Technol-
Holtmeier, F. K., 1971: Der Einfluss der orographischen ogy 17, 294-305A
Situation auf die Windverhaltnisse im Spiegel der Vege- Jones, P. D., Briffa, K. R., and Pilcher, J. R., 1984: River-
tation. Erdkunde 25, 178-195. flow reconstruction from tree rings in Southern Britain.
Holzhauser, H. P., 1984: Zur Geschichte der Aletsch- Journal of Climatology 4, 461-472.
gletscher und des Fieschergletschers. Physische Jong de, A. F. M., Mook, W. G., 1980: Medium-term
Geographie 13, Universitat Zurich, 1-488. atmospheric 14C variations. Radiocarbon 22, 267-272.
Hornbeck, J. W., and Smith, R. B., 1985: Documentation Joos, K., 1985: Year ring analysis - a new method of
of red spruce growth decline. Can. J. For. Res. 15, assessing the condition of trees in cities. Phoenix
1199-1201. International 5, 14-19.
Hoester, H. R., 1977: Veranderungen der Holzstruktur als
Indikator fur Umweltbelastungen bei Baumen. Ber. Kaelin, I., 1983: Versteinerte Holzer in vulkanischen
deutsch. bot. Ges. 90, 253-260. Tuffen der Insel Lipari. Mikrokosmos 1983, 65-70,
Hrib, J., Kyncl, J., and Cerny, A., 1983: A tree-ring study 111-113, 140-148.
of Norway spruce infected with the wood-destroying Kaiser, N. F. J., 1979: Ein spateiszeitlicher Wald im
fungus Armillaria mellea. European Journal of Forest Dattnau bei Winterthur, Schweiz. Winterthur, Ziegler. 90
Pathology 13, 160-165. pp.
Huber, B., 1941: Aufbau einer mitteleuropaischen Jahr- Kaiser, F., and Kaiser, C., 1987: The Katmai eruption von
ring-Chronologie. Mitt. H. G. Akad.deutsch. Forstwiss. 1912 and the Alaska earthquake of 1964 as reflected in
3,137-142. the annual rings of sitka spruce (Picea sitchensis) on
Huber, B., 1961: Grundzuge der Pflanzenanatomie. Ver- Kodiak island. Dendrochronologia 5, in press.
such einer zeitgemassen Darstellung. Berlin, Gottingen, Kaptein, J. C., 1914: Tree growth and meteorological
Heidelberg, Springer. 243 pp. factors. Receuil des travaux botaniques neerlandais 11,
Huber, B., and Merz, W., 1963: Jahrringchronologische 70-93.
Synchronisierung der jungsteinzeitlichen Siedlungen Kasser, P., 1954: Sur Ie bilan hydrologique des bassins
Thayngen-Weier und Burgaschisee-Sud und Sudwest. glaciaires avec application au Grand Glacier d'Aletsch.
Germania 41, 1-9. Publ. no. 39 de I'assoc. intern. d'hydrologie (Assem-
Huber, B., and Jazewitsch von, W., 1965: Tree-ring blee generale de Rome, tome 4) 331-350.
studies of the Forestry-Botany Institutes of Tharandt Keller, R., 1968: Des characteristiques nouvelles pour
and Munich, Tree-Ring Bull. 21, 28-30. I'etude des proprietes mechaniques des bois: les com-
Huber, B., 1970: Lichtmikroskopische Untersuchungen posantes de la densite. Ann. Sci. For. 25,237-249.
on Holzern, besonders die Bestimmung ihrer systema- Kennedy-Sutherland, E., 1983: The effects of fire exclu-
tischen Zugehorigkeit. In: Handbuch der Mikroskopie in sion on growth in mature Ponderosa Pine in Northern
der Technik, Band V. Mikroskopie des Holzes und des Arizona. Unpublished, University of Arizona. 20 pp.
Papiers. Frankfurt AM., Umschau-Verlag 37-103 und Kern, K. G., 1960: Der jahreszeitliche Ablauf des Dicken-
171-211. wachstums von Fichten verschiedener Standorte im
Hughes, J. F., 1968: Density as an index of wood quality Trockenjahr 1959. Allgemeine Forst- und Jagdzeitung
with special reference to the development of rapid and 131, 97-116.
efficient methods of estimating density. Journ. Oxford Kienast, F., Fluhler, H., and Schweingruber, F. H., 1981:
Univ. For. Soc. (Ser. 6), 5-6. Jahrring-analysen an Fohren (Pinus silvestris L.) an
Hughes, M. K., Kelly, P. M., Pilcher, J., and LaMarche Jr., immissionsgefahrdeten Waldbestanden des unteren
V. C., 1980: Report and recommendations. Second Wallis (Saxon, Schweiz). Mitt. schweiz. Anst. forstl.
international workshop on global dendroclimatology Versuchswes. 57,415-432.
Belfast, July 1980, Norwich, Edmund Norvic Press, Kienast, F., 1985: Dendrookologische Untersuchungen an
ISBN 0-950-7344-0-3. Hohenprofilen aus verschiedenen Klimabereichen. Eine
Hughes, M. K., Milsom, S. J., and Leggett, P. A, 1981: radiodensitometrische Studie uber den Einfluss der
Sapwood estimates in the interpretation of tree-ring Witterung, der Hohenlage und den Standortseigens-
dates. Journal of Archaeological Science 8, 381-390. chaften auf das Jahrringwaschstum von Nadelbaumen.
Hughes, M. K., Schweingruber, F. H., Cartwright, D., and Diss. Univ. Zurich, 129 pp.
Kelley, O. M., 1984: July-August temperature at Edin- Kienast, F., and Schweingruber, F. H., 1986: Dendroeco-
burgh between 1721 and 1975 from tree-ring density logical studies in the Front Range, Colorado, U.S.A.
and width data. Nature 308, 341-344. Arctic and Alpine Research 18, 277-288.
Kittredge, J., 1948: Forest influences; the effects of
Jacoby, G. C., and Ulan, L. D., 1983: Tree ring indications woody vegetation on climate, water and soil, with
of uplift at Icy Cape, Alaska related to 1899 earth- applications to the conservation of water and the control
quakes. Journal of Geophysical Research 88, 9305- of floods and erosion. New York, Mc Graw Hill, 394 pp.
9313. Klein, J., Lerman, J. C., Damon, P. E. and Ralph, E. K.,
Jahrig, M., and Baumgarten, K., 1975: Dendrochronol- 1982: Calibration of radiocarbon dates. Radiocarbon 24,
ogische Untersuchungen an Bauernhausern des meck- 103-105.
lenburgischen Sudwestens. Jahrbuch der bayerischen Klein, P., 1979: Alte Gemalde auf Holz gemalt. Holz-ZbI.
Denkmalpflege 29,35-78. 105, 2287/88.
Jesse, A, 1974: Bibliography on automatic image analy- Klein, P., 1980: Dendrochronologische Untersuchungen
sis. The Microscope 22, 89-115. an Eichenholztafeln von Rogier Van der Weyden.
Jahrbuch der Berliner Museen 23, 113-123.
266
Klein, P., 1983: Dating of art historical objects. In: D. Lawrence, J. R., and White, J. W., 1984: Growing season
Eckstein, S. Wrobel, R. W. Aniol (eds.) Dendrochronol- precipitation from D/H ratio of Eastern White Pine.
ogy and archaeology in Europe. Mitt. Bundesforschung- Nature 311, 558-550.
sanstalt fur Forst- und Holzwirtschaft. Hamburg 141, Leavitt, S. W., and Long, A., 1984: Sampling strategy for
209-222. stable carbon isotope analysis for tree rings in pine.
Klein, P., 1984: Tree-Ring chronology of spruce wood Nature 311,145-147.
and its application in the dating of stringed instruments. Leibundgut, H., Auer, C., and Wieland, C., 1971: Ergeb-
ICOM Committee for Conservation 7th Triennial Meet- nisse von Durchforstungsversuchen 1930-1965 im
ing, 1984 Kopenhagen. 84.1, 69-72. Sihlwald. Mitt. Eidg. Anst. forst/. Versuchswes. 47,
Koenig, E. 1957: Tierische und pflanzliche Holzschiid- 259-389.
tinge. Stuttgart, Holz-Zentralblattverlag, 330 pp. Lenz, 0., 1957: Utilisation de la radiographie pour I'ex-
Koenigswald von. W., and Taute, W., 1979: Zwei bedeu- amen des couches d'accroissement. Mitt. Eidg. Anstalt
tende Quartar-profile in der Burghohle von Dietfurt bei fUr das forstl. Versuchswes. 33, 125-134.
Sigmaringen an der Donau. N. Jb. Geol. Pal§ont. Mh. 4, Lenz, 0.,1967: Action de la neige et du gel sur les arbres
216-236. de montagne, en particulier sur leur forme et I'anatomie
Kontic, R., Niederer, M., Nippel, C. -A., and Winkler- de la tige. Mitt. Eidg. Anstalt fUr das forstl. Versuch-
Seifert, A., 1986: Jahrringanalysen an Nadelbaumen swesen 43, 288-316.
zur Darstellung und Interpretation von Waldschaden Lenz, 0., Oswald, K., 1971: Ueber Schaden durch Bohr-
(Wallis, Schweiz), Bericht 283, Eidg. Anstalt fUr das spanentnahme an Fichte, Tanne und Buche. Mitt. Eidg.
forstliche Versuchswesen. CH-8903 Birmensdorf. 46 . Anst. forst!. Versuchswesen 47 (1) 1-29.
pp. Lenz, 0., Schar, E., and Schweingruber, F. H., 1976:
Kozlowski, T. T., 1971: Growth and development of trees, Methodische Probleme bei der radiographisch-densito-
2 Bande, New York, London, Academic Press. metrischen Bestimmung der Dichte und der Jahrring-
Kuechler, A. W., 1966: Vegetation maps of Europe. breiten von Holz. Holzforschung 30, 114-123.
Lawrence Univ. of Kansas libraries, 584 pp. Leptere, R., 1957: Influence of high pruning on the
Kulmann, H. M., 1971: Effects of insect defoliation on growth of Douglas fir. Forestry 30.
growth and mortality of trees. Ann. Rev. Entomology Leuschner, H.-H., Delorme, A., Tuxen, J., and Hofle,
16, 289-324. H.-Chr., 1985: Eichenstammlagen in Mooren und
Kuniholm, P. I., and Striker, C. L., 1983: Dendrochrono- Talauen und die Klimaverschlechterung im Subboreal.
logical investigations in the Aegean and neighboring Flora 177, 283-':"295.
regions, 1977-1982. Journal of Field Archaeology 10, Leuschner, H. H., and Delorme, A.: Dendrochronolo-
411-420. gische Befunde zu Torfeichen aus dem Kehdinger
Kuniholm, P. I., and Striker, C. L., 1987: Dendrochrono- Moor bei Hammah, Kreis Stade. Unveroffentlichtes
logical investigations in the Aegean and neighboring Manuskript.
regions, 1983-1986. J. Field Arch. 14, in press. Libby, W. F., Anderson, E. C., and Arnold, J. R., 1949:
Laitakari, E., 1920: Untersuchungen uber die Einwirkung Age determination by radiocarbon content. World-wide
der Witterungsverhaltnisse auf das Langen- und Dick- assay of natural radiocarbon. Science 109, 227-228.
enwachstum der Kiefer. Acta Forest. Fenn 17, 53 pp. Libby, W. F., 1963: Accuracy of radiocarbon dates.
(Finnisch). Science 140, 278-280.
LaMarche, V. C. Jr., and Fritts, H. C., 1972: Tree-ring and Libby, W. F., 1960: Radiokohlenstoff-Datierung. Nobel-
sunspot numbers, Tree-Ring. Bull. 32, 19-32. Vortrag vom 12. Dez. 1960. Angewandte Chemie 73,
LaMarche, V. C. Jr., and Wallace, R. E., 1972: Evaluation 225-229.
of effects of past movements on the San Andreas Fault, Liese, W., and Eckstein, D., 1971: Die Jahrringchronol-
Northern California. Geological Soc. of America, Bull. ogie in der Kriminalistik. In: H. Schafer, (Ed.) Grund-
83, 2665-2676. lagen der Kriminalistik. Hamburg, Steintor, 395-422.
LaMarche, V. C. Jr., and Stockton, C. W., 1974: Chro- Liese, W. (ed.), 1975: Biological transformation of wood
nologies from temperature-sensitive Bristlecone pines by microorganisms. Proc. of the Sessions on wood
at upper treeline in western United States. Tree-Ring products path., 2nd Int. Congr. Plant path., Sept. 10-
Bulletin 34, 21-45. 12, 1973. Minneapolis. Berlin, Springer, 203 p.
LaMarche, V. C. Jr., 1974: Paleoclimatic inferences from Liese, W., Schneider, M., and Eckstein, D., 1975: His-
long tree-ring records. Science 183, 1043-1048. tometrische Untersuchungen am Holz einer rauchges-
LaMarche, V. C. Jr., and Hirschboeck, K. K., 1984: Frost chadigten Fichte. European J. Forest Pathology 5,
rings in trees as records of major volcanic eruptions. 152-161.
Nature 307, 121-145. Liese, W., 1978: Bruno Huber: The poineer of European
Lamb, H. ·H., 1972-1977: Climate; present, past and dendrochronology. BAR Int. Series 51, 1-10.
future, 2 Vols. London: Methuen. New York: Barnes Lightle, P. C., and Hawksworth, E. C., 1973: Control of
und Noble. Dwarf Mistletoe in a heavily used Ponderosa Pine
Larcher, W., 1973: Oekologie der Pflanzen. Stuttgart, recreation forest: Grand Canyon, Arizona. USDA Forest
Ulmer. 320 pp. Res. Paper RM-1 06. Rocky Mountain Forest and Range
Larsen, J. B., 1981: Waldbauliche und ertragskundliche Experimental Station. US Dept. of Agriculture. 22 p.
Erfahrungen mit verschiedenen Provenienzen der Lingg, W., 1986: Dendrookologische Studie an Fichte
Weisstanne (Abies alba Mill.) in Danemark. Forstw. Cbl. (Picea abies) und Weisstanne (Abies alba) im subkon-
100,275-286. tinentalen Klimagebiet Wallis, Schweiz. Ber. Eidg.
Anstalt fur das forstliche Versuchswesn 287, 1-81.
267
Liphschitz, N., 1986: Overview of the dendrochroarchae- Waldwachstums. 3 Vols. Frankfurt am Main, Sauerlan-
logical research in Israel. Dendrochronologia 4,37-58. der.
Long, A, 1982: Stable isotopes in tree rings. In Hughes, Morgan, R. A, Hillam, J., Coles, J. M., and McGrail, S.,
M. K., Kelly, P. M., Pilcher, J. R., and Lamarche, V. C.: 1981: Reconciling tree-ring sampling with conservation.
Climate from tree rings. Cambridge, London, New York, Antiquity 55, 90-95.
New Rochelle, Melbourne, Sydney. 13-21. Mueckenhausen, E., 1975: Die Bodenkunde und ihre
LUtken, C., 1891: Nogletagtagelser over de meteoriske geologischen, geomorphologischen, mineralogischen
Forhold Indflydelse pa Traernes Tilvaxtgang. Tidsskr. und petrologischen Grundlagen: Frankfurt am Main,
Skovbrug 12. DLG-Verlag. 579 pp.
Mueller, H. N., 1981: Jahrringwachstum und Klimafak~
Madany, M. H., Swetman, T. W., and West, N. E., 1982: toren. Angewandte Pflanzensoziologie. Ver6ff. Forstl.
Comparison of two approaches of determining fire Bundesversuchsanstalt Wien 25, 81 pp.
dates from tree scars. Forest Science 28, 856-861 . Mueller-Stoll, H., 1951: Vergleichende Untersuchungen
Markus, R., 1936: Einfluss der Entwasserung auf den Uber die Abhangigkeit der Jahrringfolge von Holzart,
Zuwachs der Kiefer und Fichte auf MoorbOden in Lett- Standort und Klima. Bibl. Bot. 122, 1-93.
land. Mitt. forestliche Versuchsanstalt Lettlands 5 Mueller-Stoll, W., 1947: Photometrische Holzstruktur-
(Latvian with German Summary). Untersuchungen. I. Mitt.: Ueber die Ermittlung von
Mason, R. R., 1978: Population Ecology. Pg. 39-61. In: Jahrringaufbau und Spatholzanteil auf photometrischem
The Douglas-Fir Tussock-Moth, a Synthesis. Forest Wege. Planta 35, 397-426. II. Mitt.: Ueber die Bezie-
Service, Science and Education Agency, Technical hungen der Lichtdurchlassigkeit von Holzschnitten zur
Bulletin No. 1585, US Dept. of Agriculture. Washington Rohwichte und Wichtekontrast. Forstw. Cbl. 68, 21-63.
D.C.,391 p. Munaut, A, 1967: La foret Ensevelie de Ternuzen. Revue
Mathieu, H., 1961: La conservation du materiau bois. Industrie. 9 pp.
Paris, Bailliere, 318 pg. Munaut, A, and Casperie, W. A, 1971: Etude dendro-
McGregor, J. C., 1936: The effect of a volcanic cinder fall chronologique de Pinus silvestris L. subfossiles pro-
on tree growth. Tree-Ring Bull. 3, 11-13. venants de la tourbiere d'Emmen (Drenthe, Pays-Bas).
McLaughlin, S. B., West, D. C., and Blasing, T. J., 1983: Rev. Palaeobot., Palynology 11, 201-226.
Measuring effects of air pollution stress on forest pro- MUnch, D., and Lambert, D., 1972: Tiberline ancients.
ductivity: Some perspectives, problems, and ap- 2000 N.w. Wilson, Oregon, Belding. 128 pp.
proaches. Proceedings TAPPI, Annual meeting, 321-
333. Nageli, W., 1935: Aussetzende und auskeilende Jahr-
McLaughlin, S. B., Blasing, T. J., Mann, L. K., and Duvick, ringe. Schweiz. Z. Forstwesen 1935.1-7.
D. N., 1983: Effects of acid rain and gaseous pollutants Nogler, P., 1981: Auskeilende und fehlende Jahrringe in
on forest productivity: A regional scale approach. Jour- absterbenden Tannen (Abies alba Mill.). Allg. Forstz.
nal of the Air Pollution Control Association 33, 1042- 28, 709-711.
1048. Nultsch, W., 1977: Allgemeine Botanik. Kurzes Lehrbuch
MacLean, D. A, 1981: Impact of defoliation by Spruce tilr Mediziner und Naturwissenschaftler. 6. Aufl. Stutt-
budworm populations on radial and volume growth of gart, Georg Thieme. 438 pp.
Balsam fir: A review of present knowledge. Mitt. der
forstlichen Bundesversuchsanstalt, Wien 142, 293- Oeschger, H., and Rothlisberger, H., 1961: Datierung
306. eines ehemaligen Standes des Aletschgletschers durch
Medway, L., 1972: Phenology of a tropical rainforest in Radioaktivitatsmessung an Holzproben und Bemer-
Malaya. BioI. J. Linn. Soc. 4, 117-146. kungen zu Holzfunden an weiteren Gletschern. Ztschr.
Meisling, K. E., and Sieh, K., 1980: Disturbance of trees Gletscherkunde und Glaziologie 4, 191-205.
by the 1857 Fort Tejon earthquake, California. Journal Oeschger, H., 1985: Langfristige Klimastabilitat. Neue
of Geophysical Research 83, 3225-3238. Zurcher Zeitung vom 27. Februar 1985. P. 65.
Meusel, H., 1965-1978: Vergleichende Chronologie der
zentraleuropaischen Flora. 4 Vols. Jena Fischer. Page, R., 1970: Dating episodes of faulting from tree
Mikola, P., 1956: Tree-ring research in Finland. Tree-Ring rings: Effect of the 1958 rupture of Fairweather Fault on
BUll. 21, 16-20. tree growth. Geological Soc. of America Bulletin 81,
Miller, D. J., 1960: Giant waves in Lithuya Bay, Alaska. 3085-3094.
US Geol. Survey Prof. Paper 354 - C. 51-86. Palenik, S., 1983: Microscopic trace evidence. The over-
Mitchell, J. M., Stockton, C. W., and Meko, D. M., 1979: looked clue. Part IV Arthur Koehler-Wood Detective.
Evidence of a 22-year rhythm of drought in the western The Microscope 31, 1-13.
United States related to the hale solar cycle since the Panshin, A J., and Zeeuw, C., 1970: Textbook of wood
17th century. In: Solar-terrestrial influences in climate. technology, 3rd ed., 705 p., New York, McGraw-HilI.
B. M. McCormac and T. A Seliga (Eds.). Reidel Publ. Parker, M. L., 1971: Dendrochronological techniques
Co. Holland. pp. 125-143. used by the geological survey of Canada. In: Paper
Mitchell, J. M., 1980: History and mechanisms of climate. 71-25, 1-30. Ed. Geological Survey of Canada.
In: H. Oeschger, B. Messerli and M. Svilar: Das Klima. Parker, M. L., and Jozsa, L. A., 1971: Dendrochrono-
Analysen und Modelle, Geschichte und Zukunft. Berlin, logical investigation along the Mackenzie, Liard and
Heidelberg, New York, Springer. 296 pp. south Nahanni Rivers, N.w.T. Technical Report 10 to
Mitscherlich, G., 1970-1975: Wachstum und Umwelt; Glaciology Division. Water Resources Branch. Dept. of
eine Einfuhrung in die 6kologischen Grundlagen des the Environment. 187 pp.
268
Parker, M. L., and Jozsa, L. A., 1973: X-ray scanning Renfrew, C., 1977: Ancient Europe is older than we
machine for tree ring width and density analysis. Wood thought. Nat. Geogr. Mag. 152,615-62.
and Fibre 5,192-197. Renner, F., 1982: Beitrage zur Gletschergeschichte des
Patzelt, G., and Bortenschlager, S., 1973: Die post- Gotthardgebietes und dendroklimatologische Analysen
glazialen Gletscher- und Klimaschwankungen in der an fossilen Holzern. Physische Geographie. Universitat
Venedigergruppe (Hohe Tauern, Ostalpen). Ztschr. Zurich, 8, 182 pp.
Gletscherkunde und Glaziologie 9,5-57. Reventlow, C. D. F., 1879: Forslag til en forbedret skov-
Pearson, G. W., and Baillie, M. G. L., 1983: High-preci- drift grundet paa undersogelser over traeernes vegeta-
sion 14C measurement of Irish oaks to show the natural tion i Danmarks og Slesvigs skove. Kopenhagen,
atmospheric 14C variations of the AD period. Radio- Hauberg. 206 pp.
carbon 25,187-196. Richter, K., and Eckstein, D., 1986: Estudio dendrochro-
Pearson, G. W., Pilcher, J. R., and Baillie, M. G. L., 1983: nol6gico en Espana. Dendrochronologia 4, 59-74.
High precision 14C variations from 200 B.C. to 4000 Robinson, W. J., 1967: Tree-ring materials as a basis for
B. C. Radiocarbon 25, 179-186. cultural interpretations. Diss. Univ. of Arizona, Tucson.
Pechmann von, H., 1960: Haben Mineraldungung und Unpublished, 107 pp.
Lupinenanbau einen Einfluss auf die Eigenschaften von Robinson, W. H., 1976: Tree-ring dating and archaeology
Fichten- und Kiefernholz. Forstw. Cbl. 79, 91-105. in the American Southwest. Tree-Ring Bull. 36, 9-20.
Pechmann von, H., 1974: Der Einfluss der Durchforstung Rose, M. R., Dean, J. S., and Robinson, W. J., 1981: The
auf die Holzqualitat. Forstarchiv 45, 34-38. past climate of the Arroyo Hondo New Mexico recon-
Petersen, A., and Eckstein, D., 1984: Holzbiologische . structed from the tree rings. School of American re-
Untersuchungen an Strassenbaumen. Wiss. Ber. Univ. search press. Arroyo Hondo arch. Series 4, 114 pp.
Hamburg. 18,37-40. Rothlisberger, F., 1976: 8000 Jahre Walliser Gletscher-
Pfister, Ch., 1984: Klimageschichte der Schweiz, 1525- geschichte: ein Beitrag zur Erforschung des Klimaver-
1860. Academica helvetica 6. Paul Haupt, Bern und laufs in der Nacheiszeit. 2. Teil. Die Alpen 52, 59-152.
Stuttgart. Rothlisberger, F., 1986: 10000 Jahre Gletschergeschichte
Pfister, Ch., 1986: Veranderungen der Sommerwitterung der Erde. Aarau, Frankfurt a.M., Salzburg. Sauerlander,
im sudlichen Mitteleuropa als Auftakt zum Gletscher- 416 pp.
hochstand der Neuzeit. Geographica helvetica 1985, Rubi, Chr., 1980: Das Simmentaler Bauernhaus. Berner
186-195. Heimatbucher 125. Bern, Paul Haupt. 155 pp.
Philipps, R., 1978: Trees in Britain, Europe and North Ruoff, U., 1979: Neue dendrochronologische Date aus
America. London, Pan. 224 pp. der Ostschweiz. Zeitschr. Schweiz. Arch. und Kunst-
Phillips E. W. J., 1966: The use of beta-particle radiation gesch. 36, 94-96.
methods in timber research. IAWA Bull. 2, 17-28. Ruoff, U., and Rychner, V., 1986: Die Bronzezeit im
Pilcher, J. R., Baillie, M. G. L., Schmidt, B., and Becker, schweizerischen Mittelland. Antiqua 15, 73-79.
B., 1984: A 7272-year tree-ring chronology for western Rypacek, V., 1966: Biologie holzzerstorender Pilze. Jena,
Europe. Nature 312, 150-152. Fischer. 211 pp.
Plumtree R. A., 1979: Some techniques used in the study
of wood density. Unpublished report 6 pp. Sachs, L., 1978: Angewandte Statistik. Statistische
Plumtree R. A., 1984: Pinus caribaea Vol. 2, Wood prop- Methoden und ihre Anwendung. 5. Edn. Berlin, Heidel-
erties. Tropical Forestry paper 17, 1-148. berg, New York, Springer. 522 pp.
Pokorny, A., 1892: Methode um den meteorologischen Sanderman, W., Schweers, W., and Gaudert, P., 1960:
Coeffizienten des jahrlichen Holzzuwachses der Di- Messung der Holzdichte und Bestimmung der Holz-
cotyledonenstamme zu ermitteln. Tharandter forstl. Jb. Jahrringbreite mit Hilfe von Beta-Strahlen. Forstarchiv
22, 81 ff. 31,126-128.
Polge, H., 1966: Etablissement des courbes de variation Sauvala, K., 1982: Wood microdensitometry bibliography.
de la densite du bois par I'exploration densitometrique Wood Microdensitometry Bulletin 2(1), 6-54.
a
de radiographies d'echantillons preleves la tariere sur Schafer, H., 1954: Forstkriminalistik. Hamburg, 85 pp.
des arb res vivants. Ann. Sci. Forest. 23, 1-206. Schar, E., Schweingruber, F. H., 1987: Nacheiszeitliche
Polge, H., and Thiercelin, F., 1970: Degats occasionnes Stammfunde aus Grachen im Wallis. Schweiz. Z Forst-
par les sondages de tariere. Techn. et For. 6, 629- wesen, 138, 497-515.
636. Scharpf, R. F., 1978: Control of Dwarf Mistletoe on true
Polge, H., Keller, R., and Thiercelin, F., 1973: Influence Firs in the West. Pacific Southwest Forest and Range
de I'etage de branches vivantes sur la structure des Experimental Station. Gen. Tech. Rep. PSW - 31,
accroissements annuels et sur quelques caracteris- 117-123.
tiques du bois de Douglas et de Grandis. Ann. Sci. Scheffer, F., and Schachtschabel, P., 1970: Lehrbuch der
Forest. 30,127-140. Bodenkunde. 7. Aufl. Stuttgart. 448 pp.
Polge, H., Lemoine, M., and Deret, E., 1977: Etude de la Schichtl, H. M., 1973: Karte der aktuellen Vegetation
variabilite specifique et infraspecifique de la structure Tirols. Blatt 5, Silvretta und Lechtaler Alpen. Documents
juvenile du bois de chene a I'aide d'un analyseur de Cartographie Ecologique 11, Grenoble.
d'image. Ann. Si. Forest. 34(4), 285-293. Schirmer, W., 1979: Rannen im Mainschotter. Frankische
Pollanschutz, J., 1971: Die ertragskundlichen Methoden Heimat am Obermain, Heft 16. Beilage zum Jahres-
zur Erkennung und Beurteilung von forstlichen Rauch- bericht 1978179 des Merainer-Gymnasiums lichten-
schaden. Mitt. Forstl. Bundesvers'anst. Wien 92, 155- fels. 44 pp.
206. Schmidt-Vogt, H., 1977: Die Fichte; Band I: Taxonomie,
269
Verbreitung, Morphologie, Oekologie, Waldgesellschaf- Schweingruber, F. H., Kontic, R., and Winkler-Seifert, A.,
ten. Hamburg und Berlin, Parey. 647 pp. 1983: Eine jahrringanalytische Studie zum Nadelbaum-
Schmucker, Th., and Linnemann, G., 1951: Geschichte sterben in der Schweiz. Ber. der Eidg. Anstalt fUr das
der Anatomie des Holzes. In: H. Freud, Handbuch der forstliche Versuchswesen 253, 29 pp.
Mikroskopie in der Technik, Bd. 5/1, 1-78. Schweingruber, F. H., Albrecht, H., Beck, M., Hessel, J.,
Schneebeli, W., 1976: 8000 Jahre Walliser Gletscherge- Joos, K., Keller, D., Kontic, R., Lange, K., Niederer, M.,
schichte; ein Beitrag zur Erforschung des Klimaverlaufs Nippel, C., Spang, S., Spinnler, A., Steiner, B., and
in der Nacheiszeit. 1. Teil. Die Alpen 52(3/4), 1-57. Winkler-Seifert, A., 1986: Abrupte Zuwachsschwan-
Schneider, S. H., 1978: Klima in Gefahr. Strategien zur kungen in Jahrringabfolgen als oekologische Indika-
Beherrschung des Welters. Frankfurt am Main, Fischer; toren. Dendrochronologia 4, 125-183.
Translation of the American edition: The genesis stra- Schwenke, W., 1972, 1974-1978: Die Forstsch3.dlinge
tegy - Climate and global survival (1976). New York, Europas. 3 Bande. Hamburg, Berlin, Parey.
Plenum Press. 396 pp. Sell, J., 1978: Quantitative automatische Bildanalyse in
Schoepfer, W., 1962: Beitrage zur Erfassung des Assimi- der Materialprufung. Prinzip, Systemaufbau, Anwen-
lations-apparates der Fichte. Diss. Univ. Freiburg. 127 dungen. Material und Technik 6, 79-83.
pp. Serre, F., 1973: Contribution a I'etude dendroclimato-
Schotterer, U., and Oeschger, H., 1979: 1m Eis gespei- logique du pin d'alep (Pinus halepensis). Ph.D. these,
cherte Information. In: Kasser P.: Die Schweiz und ihre I'universite D'Aix-Marseille III.
Gletscher. Kummerli und Frey, Bern. Shane, D., and Harper, L. T., 1979: Influence of precipita-
Schotterer, U., 1986: Der Gletscher als Reinraumarchiv tion and temperature on ring annual branch increment
von Umweltvorgangen. Swiss Chem. 8, 49-52. and needle-growth of White Fir and Douglas Fir in
(Schweizerische Zeitschrift fUr die chemische Industrie.) central Utah. Great Basin Naturalist 39, 219-225.
Schroeder, D., 1969: Bodenkunde in Stichworten. Kiel, Shigo, A. L., and Marx, H. G., 1977: Compartmentaliza-
Ferdinand Hirt. 144 pp. tion of decay in trees. Forest Service US Dept. of Agri-
Schulman, E., 1937: Some early papers on tree-rings. J. culture. Agriculture Information Bul/' 405, 73 pp.
C. Kaptein. Tree-Ring Bull. 3,28-29. Shigo, A. L., 1986: A new tree biology. Shigo and Trees
Schulman, E., 1958: Bristlecone pine, oldest known living Assoc. Durham, NH 03424, USA, 595 pp.
thing. Nat. Geogr. Mag. 113,355-372. Shigo, A. L., Campana, R., Hyland, F., and Anderson, J.,
Schuett, P., 1981: Ursache und Ablauf des Tannenster- 1980: Anatomy of Elms injected to control Dutch Elm
bens - Versuch einer Zwischenbilanz. Forstw. Cbl. Disease. Journal of Arboriculture 6, 96-100.
100, 286-287. Shroder, J. F., 1976: Dendrogeomorphological analysis of
Schweingruber, F. H., 1978: Mikroskopische Holzana- mass movement on Table Cliffs Plateau, Utah. Quarter-
tomie; Formspektren mitteleuropaischer Stamm- und nary Research 9, 108-185.
Zweighblzer zur Bestimmung von rezentem und sub- Shroder, J. F., 1980: Dendrogeomorphology: review and
fossilem Material. Eidg. Anst. forst/. Versuchswes. 226 new techniques of tree-ring dating. Progress in Physical
pp. Geography 4, 161-188.
Schweingruber, F. H., Fritts, H. C., Braeker, O. U., Drew, Shroder, Jr. J. F., and Butler, D. R., 1987: Tree-ring
L. G., and Schar, E., 1978: The X-ray technique as analysis in the earth sciences. In press.
applied to dendroclimatology. Tree-Ring Bull. 38, 61- Shvedov, F. N., 1892: Trees as chroniclers of droughts.
91. In: J. M. Fletcher and W. Linnard, 1977. Russian papers
Schweingruber, F. H., 1979: Auswirkungen des Larchen- on dendrochronology and dendroclimatology 1962/
wicklerbefalls auf die Jahrringstruktur der Larche; 1968, 1970, 1972. Res. Lab. Arch. and Hist. of Art,
Ergebnisse einer Jahrringanalyse mit rbntgendensito- Oxford University 39-49.
metrischen Methoden. Schweiz. Z. Forstwes. 130(12), Sieben list-Kerner, V., 1984: Der Aufbau von Jahrring-
1071-1093. chronologien fUr Zirbel kiefer, Larche und Fichte eines
Schweingruber, F. H., Braeker, O. U., and Schaer, E., alpinen Hochgebirgsstandortes. Dendrochronologia 2,
1979: Dendroclimatic studies on conifers from central 9-29.
Europe and Great Britain. Boreas 8,427-452. Smiley, T. L., 1958: The geology and dating of Sunset
Schweingruber, F. H., and Ruoff, U., 1979: Stand und Crater, Flagstaff, Arizona. In: Guidebook for the Black
Anwendung der Dendrochronologie in der Schweiz. Z. Mesa Basin, Northeastern Arizona, 186-190.
schweiz. Arch. Kunstgesch. 36,69-96. Smith, L. P., and Stockton, Ch. W., 1981: Reconstructed
Schweingruber, F. H., 1980: Dichteschwankungen in stream flow for the Salt and Verde Rivers from tree-ring
Jahrringen von Nadelhblzern in Beziehung zu klima- data. Water Resources Bul/. 17, 939-947.
tisch-bkologischen Faktoren, oder das Problem der Stahle, D. W., 1979: Tree-ring dating of historic buildings
falschen Jahrringe. Ber. Eidg. Anst. forstl. Versuch- in Arkansas. Tree-Ring Bull. 39, 1-28.
swes. 213, 35 pp. Stallings, W. S., 1937: Some early papers on tree-rings.
Schweingruber, F. H., Schar, E., and Braker, O. U., 1984: Tree-Ring Bull. 3, 27-28.
Jahrringe aus sieben Jahrhunderten. Saaner Jahrbuch Stockton, Ch. W. 1971: The feasibility of augmenting
1984,1-30. hydrologic records using tree-ring data. Ph.D. Diss.
Schweingruber, F. H., 1985: Dendroecological zones in University of Arizona, Tucson.
the coniferous forests of Europe. Dendrochronologia 3, Stockton, Ch. W., and Boggess, W. R., 1980: Tree rings:
67-75. A proxy data source for hydrologic forecasting. In North
Schweingruber, F. H., 1987: Flachenhafte dendroklima- R. M. et al. (eds.) Unified Basin management. Sym-
tische Temperaturrekonstruktionen fur Europa. Natur- posium proceedings, May 4-7, 1980. Gatlingburg,
wissenschaften 74, 205-212. Tennessee, 605-624.
270
Stockton, Ch. W., and Jacoby G. C. Jr., 1976: Long-term Swetnam, T. w., and Dietrich, J. H., 1983: Fire history of
surface water supply and streamflow trends in the Ponderosa Pine forest in the Gila Wilderness, New
upper Colorado River basin based on tree-ring analysis. Mexico. Unpublished paper, presented at the Wilder-
Lake Powell Research Project BUll. 18, University of ness Fire Symposium, Missoula, Montana, Nov. 1983.
California, Los Angeles, Institute of Geophysics and Swetnam, T. W., Thompson, M. A., and Kennedy Suther-
Planetary Physics, 70 pp. land, E., 1985: Using dendrochronology to measure
Stockton, Ch. W., Mitchell, L. M., and Meko, D. M., 1981: radial growth of defoliated trees. US Dept. of Agricul-
Tree-ring evidence of a relationship between drought ture, Forest Service, Agriculture Handbook 639, 1-39.
occurence in the western United States and the Hale
Sunspot Cycle. In: Lawson, M. P., Baker, M. E., (eds.) Tande, G. F., 1979: Fire history and vegetation pattern of
The Great Plains. Perspectives and Prospects. Uni- coniferous forests in Jasper National Park, Alberta.
versity of Nebraska Press, Lincoln and London, 83- Canadian Journal of Botany 57, 1912-1932.
110. Taubenheim, J., 1969: Statistische Auswertung geophy-
Stokes, M. A., and Smiley, T. L., 1968: An introduction to sikalischer und meteorologischer Daten. Akad. Ver-
tree-ring dating. Chicago and London. University of lagsges. Geest und Portig K.-G., Leipzig. 386 pp.
Chicago Press, 73 pp. Terskov, I. A., Vaganov, E. A., Zyrjanov, G. I., and Phili-
Stokes, M. A., and Dietrich, J. H., (eds.) 1980: Proceed- monov, V. S., 1978: The use of cell size distribution-
ings of the fire history workshop. Tucson, Arizona. curves for tree growth analysis. Mitt. der sibirischen
General Tech. Report RM-81. Rocky Mountain Forest Abt. der Akad. Wiss. USSR (Russian). Serie Biolo-
and Range Experiment Station. Forest Service. US gische Wissenschaften, 26-36.
Dept. of Agriculture, 142 p. Thenius, E., 1977: Meere und Lander im Wechsel der
Strahler, A. N., 1975: Physical geography. 4th Ed. New Zeiten. Verstandliche Wissenschaft Bd. 114. Berlin,
York, John Wiley and Sons, Inc. Heidelberg, New York, Springer. 200 pp.
Strahm, Ch. 1977: Kontinuitat und Kulturwandel im Neo- Thompson, M. A., 1981: Tree rings and air pollution: A
lithikum der Westschweiz. Fundberichte aus Baden case study of Pinus monophylla growing in east Central
Wurttemberg 3, 11 5-143. Nevada. Environmental Pollution (Series A) 26, 251-
Straka, H., 1970: Walter, H., Arealkunde; f/oristisch his- 265.
torische Geobotanik. 2. Aufl. Stuttgart, Ulmer. 478 pp. Till, C., 1984: A synthesis of response functions from
Stuiver, M., and Quay, D., 1980: Changes in atmospheric eight Cedar forests located in Northern Africa. Den-
carbon-14 attributed to a variable sun. Science 207, drochronologia 2,73-82.
11-19. Timell, T. E., 1986: Compression wood in Gymnosperms.
Stuiver, M., 1980: Solar variability and climatic change 3 vols. Berlin, Heidelberg, New York, Tokyo. Springer,
during the current millennium. Nature 286, 868-871. 2150 pp.
Stuiver, M., and Grootes, P. M., 1980: Trees and the Tranquillini, W., 1979: Physiological ecology of the alpine
ancient record of heliomagnetic cosmic ray flux modula- timberline. Berlin etc. Springer, 137 pp.
tion. In: Pepin, R. 0., Eddy, J. A. and Merrill, R. B., Trendelenburg, R., and Mayer-Wegelin, H., 1955: Das
Proceedings of the Conference Ancient Sun, 165- Holz als Rohstoff. Munchen, Carl Hanser. 541 pp.
1973.
Stuiver, M., Rebello, A. L., White J. C., and Broecker, Ur- und Fruhgeschichte Archaologie der Schweiz 1968:
w., 1981: Isotopic indicators of age/growth in tropical Band I. Die Aeltere und Mittlere Steinzeit. Schweiz.
trees. School of Forestry and environmental studies. Ges. Ur- und Fruhgesch., Basel, 146 pp.
Yale University Bull. No. 94, 75-82.
Stuiver, M., and Quay, P. D., 1981: Atmospheric 14C Vaganov, E. A., and Terskov, I. A., 1977: Die Analyse des
changes resulting from fossil fuel CO 2 release and Baumwachstums anhand der Jahrringstruktur (Rus-
cosmic ray flux variability. Earth and Planetary Science sian). SA der Akad. der Wiss. UdSSR Sibirische Abt.
Letters 53, 349-362. Forst- u. Holzinst. W. N. Suskatschow. Verlag Wis-
Stuiver, M., 1982: A high-precision calibration of the AD sensch. Nouosibirsk. 93 pp.
radiocarbon timescale. Radiocarbon 24, 1-26. Various authors, 1975: (SBN) Wald und Umwelt; Informa-
Stuiver, M., 1983: Statistics of the climatic and carbon tion des schweizerischen Bundes fUr Naturschutz und
isotopic change. Radiocarbon 25, 219-228. der liechtensteinischen Gesellschaft fUr Umweltschutz.
Suess, H. E., 1967: Bristlecone pine calibration of the Various authors, 1979: Harms Geographie. Physische
radiocarbon time scale 5200 B.C. to 1500 B.C. In: Geographie und Nachbarwissenschaften. 424 pp.,
Radioactive dating and methods of low-level counting: Munchen, List.
Vienna IAEA, 143-151. Varley, G. C., 1978: The effects of insect defoliation on
Suess, H. E., 1979: 1st die Sonnenaktivitat fUr Klima- the growth of oaks in England. In: J. Fletcher, Den-
schwankungen verantwortlich? Umschau 79, 312-316. drochronology in Europe. BAR Int. Series 51, 179-
Suter, P. J., and Schifferdecker, F., 1986: Das Neolithi- 186.
kum im schweizerischen Mittelland. Antiqua 15, 34-
43. Wade, J. E., and Hewson, E. W., 1980: A guide to
Swetnam, T. W., 1985: Radial growth losses in Douglas- biological wind prospecting. Final report for US Dept. of
Fir and White Fir caused by Western Spruce Budworm Energy, Div. of distributed solar technOlogy, Federal
in northern New Mexico: 1700-1983. Forest Pest wind energy program. 112 pp.
Management Report R-3, 86-2. 62 pp. US Dept. of Wade, J.E., and Wendel-Hewson, E., 1979:Trees as a
Agriculture, Forest Service, Southwestern Region. local climatic wind indicator. Journal of Applied Meteor-
ology. 18, 1182-1187.
271
WagenfOhr, R., 1966: Anatomie des Holzes. Leipzig. VEB Wieler, A., 1905: Untersuchungen tJber die Einwirkung
Fachbuch-Verlag. 377 pp. schwefliger Saure auf die Pflanzen. Berlin und Leipzig.
Walter, H., and Lieth, H., 1960/1964/1967: Klimadia- Borntrager. 427 pp.
gramm-Weltatlas. 3 Bde. Jena, Fischer. Willkomm, H., 1976: Alterbestimmungen im Quartar.
Walter, H., 1964: Vegetation der Erde. Bad I Tropische Munchen, Thiemig. 276 pp.
und subtropische Zonen. Fischer, Jena. Wininger, J., 1976: Feldmeilen-Vorderfeld. Die Ausgra-
Walter, H., 1964, 1968: Die Vegetation der Erde in oko- bungen 1970-1971. Schweiz. Ges. Ur- und FrtJh-
physiologischer Betrachtung. Band I: Die tropischen gesch. Basel, 1-106.
und subtropischen Zonen. Band 2: Die gemassigten Worbes, M., 1985: Structural and other adaptions to long-
und arktischen Zonen. Stuttgart, Fischer. term flooding by trees in Central Amazonia. Amazonia
Webb, G. E., 1983: Tree rings and telescopes. The 9,459-484.
scientific career of A. E. Douglass. The University Press Worbes, M., 1986: Lebensbedingungen und Holzwach-
of Arizona, Tucson, Arizona. 242 pp. stum in zentralamazonischen Ueberschwemmungs-
Welten, M., 1981: Gletscher und Vegetation im Lauf der waldern. Scripta Geobotanica 17, 112 pp.
letzten hunderttausend Jahre. Verh. Schweiz. natur-
forsch. Ges., Gletschersymposium in Brig 1978. Basel. Yamaguchi, D. K., 1983: New tree ring dates for recent
Weltforstatlas 1951-1971: Weltforstatlas - World fores- eruptions of Mount St. Helens. Quarternary Reserach
try atlas - Atlas des forets du monde - Atlas forestal 20, 246-250.
del mundo. Hg. Bundesforschungsanstalt fOr Forst- und Yanosky, T., 1983: Evidence of floods on the Potomac
Holzwirtschaft, Reinbeck. Hamburg, Berlin, Parey. River from anatomical abnormalities in the wood of
Wendel-Hewson, E., Wade, J. E., and Baker, R. W., 1977: flood-plain trees. US Geol. Survey Prof. Paper 1296, 42
Vegetation as an indicator of high wind velocity. Final pp.
Report for US Dept. of Energy, Div. of Solar Energy.
Federal wind Program. Erstellt von Oregon State Zimmermann, M. H., and Brown, C. L., 1971: Trees:
University, Dept. of Atmospheric Sciences, Corvallis, structure and function. Berlin, Springer, 336 pp.
Oregon. 58 pp. Zimmermann, W. 1959: Die Phylogenie der Pflanzen.
White, G. F., 1972: History of fire in North America. In: Stuttgart, Fischer, 777 pp.
Fire in the environment, Symp. Proc., May 1-5, 1972. Zoller, H., Schindler, C., and Rothlisberger, H., 1966:
Forest Service, US Dept. of Agricult., North American Postglaziale Gletscherstande und Klimaschwankungen
Forestry Commission, FAO, FS-276, 3. im Gotthardmassiv und Vorderrheingebiet. Verh. Natur-
White, J. W. C., and Lawrence, J. R., 1980: The relation- forsch. Ges. Basel 77, 97-1 64.
ship between the non-exchangeable hydrogens of tree- Zuber, R., Tschannen, W., and Bovey, E., 1981: Controle
ring cellulose and the source waters for trees. In: G. C. de la teneur en fluor des feuilles d'abricotiers de la
Jacoby, (Ed.) Carbon Dioxide Effects; Res. and Assess- Vallee du Rhone de 1974 a 1980. Rev. Suisse Vitic.
ment Program, Proc. Int. Meeting on Stable Isotopes in Abric. HortiG. 13, 133-138.
Tree-Ring Research, New York, Lamont-Doherty Geol.
Observ., Columbia Univ., 58-65.
White, J. W. C., 1983: The climatic significance of D/H Recently two volumes have been published who reflect the
ratios in White Pine in the northeastern United States. modern standard in dendroecology:
Diss. Columbia University New York. 331 pp. Jacoby G. c., Hornbeck J. W. (editors) 1987: Proceedings of
Wickman, B. E., 1978: A case study of a Douglas-Fir the international symposium on ecological aspects of
Tussock Moth Outbreak and stand conditions 10 years tree-ring analysis. August 17-21, 1986, Marymount Col-
later. USDA Forest Service research Paper PNW-244. lege, Tarrytown, New York. 726 pp. (Available from the
22 pp. National Technical Information Service, U.S. Department
Wickman, B. E., Henshaw, D. L., and Gollob, S. K., 1980: of Commerce, Springfield, Virginia 22161, USA.)
Radial Growth in Grand Fir and Douglas-Fir related to Isaev et al. (editors) 1987: Dendrochronological methods in
defoliation by Douglas-Fir Tussock Moth in Blue Moun- forest science and ecological forcasting. Irkutsk USSR,
tains Outbreak. US Dept. of Agriculture. Forest Service. 313 pp.
Pacific Northwest Forest and Range Experiment Station.
Research Paper NNW-269. 23 pp.
272
General index
Abies alba 171, 240, 242, 247 Branch cutting 244 Complacent 49, 74, 82
Abies cephalonica 128, 159 Bristlecone pine 14, 27, 31, 136, Composite 50
Abies grandis 233 174,215,249 Compression wood 36, 49, 72,
Abrupt growth changes 48, 190, Bronze age 154 99,177,200
195,211,212,227,240 Building activity 149, 153, 154 Conifer 7, 10, 11, 55, 61, 63, 70-
Acclimatization 63 Buried trunks 45,178,184-189, 72,99,103,114,115,145
Acer pseudoplatanus 244 196 Core cutting 58
Acersaccharinurn 55 holder 43
Acid rain 247 14C-Dating 248-253 mounting 43
Actinornycetes 34 14C-Age 248, 249 orientation 42, 60
Aeolian processes 179 Calibration 57, 68, 90 support 60
Aerobic 34 Callixylon newberry 36 transport 43
Afforestation 238 Callusing 44, 205, 215, 219 Coring 58
Age determination 96, 108, 109, Cambial activity 19, 98, 109, 139 Correlation 84, 247
237,251 Carbonization 35 auto- 84
distribution 240 Carving 160 coefficient 84
structure 203 Caterpillar 225 serial 84
trend 78, 121 Catface 205 Cosmic radiation 248
Aging process 121, 238, 239 Cathedrals 156, 158 Coverage 51
Air pressure 18, 210 Cell differentiation 7,98, 140 Criminal investigation 166, 167
Alpine zone 104, 110 enlargement 98 Cross dating 50, 51, 76, 77
Alps 126, 127 form 113 Crown damage 246
Anatomical technique 40 formation 98 Counting tree rings 47
American buildings 152-155, numbers 100 Cumulative temperature differ-
157 size 99,113 ence 18,92
America, northern 152-155, 157, Cellwall density 55 Curve fitting 75
172, 206-210 growth 99 trend 75
America, southern 180, 211 thickness 114, 115, 176 Cutting 40, 52
Anaerobic 34 Cells dead 176 Cycles, lunar 174
Annual growth analysis 136-139 living 99,141,176 sunspot 174
Aperiodic increments 3 Cellulose acetate 62 insect damage 223, 225
Applied dendrochronology 143- Cellulose nitrate 252
253 Charcoal 35,47,258,259 Deciduous trees 28, 33, 35, 98,
Araucaria araucana 17, 180, 211 Chemical trace elements 247 99,103-105,109-112,118
Arceutobiurn sp. 230,231 Churches 159, 160 Density diagram 68,117,119,
Arctic zone 104, 110, 231 Chronology, absolute 151 237
Arizona 45,152-155,183,221 bristlecone pine 26, 27 difference 115
Arid zone 104, 111 douglas fir 249 fluctuations 116, 117, 129, 135
Arithmetic mean 81 larch 26,29,225 integrator 69
Armillaria rnellea 229 oak 26,184-186,188,189, integration 69
Art history 162-165 209 maximum 114
Austrocedrus chilensis 180 relative 150 minimum 115
Axes, stone 150-155 regional 128 level 71
scar 206, 207 pattern 123
Bacteria 34 spruce 26,29 threshold 69, 71
Balkan 159 City trees 91, 244 profile 33, 35, 41, 56, 107, 124
Barcharts 51,159,185,188,198 Cliff dwellings 152-155 variations 116, 124
Bark 50,97 Climate, boreal 126, 128 Destruction period 185
Beaver-teeth marks 155 mediterranean 139 Deuterium 252, 253
Beech 100, 102 temperate 127,139 Diffuse porous 100
Beetle 33, 222 Climatic diagrams 8, 9, 20, 106, Discoloration 44
Beetle galleries 155 108,236 Distortion beam 147
Belt sander 46 fluctuations 170, 175 Distribution, normal 84
Biological defence 228 information 23, 24 skew 84
Biological errors 72 integrators 20 Diurnal rings 2
Boreal climate 126, 128 reconstruction 80, 90, 170- Double-mass analysis 92
Bog oak 184-186,188,189 174 Douglass, A. E. 257
Borer 42,43 zones 8 Douglas fir 155, 181
power driven 43 Climatology 168-175 Drainage 190, 191, 237
Bore hole 44 CO 2 -exchange 6 Driftwood 33, 148, 21 0
repairing 45 Cockchafers 222 Daily radial growth 19
sealing 45 Colorado river 181 Data correcting 72
Bore support 42, 43 Colored deposits 57, 101 index 78,85,90
273
meteorological 18, 92 Fossilization 36 Icings 219
processing 73,78-92 Freeze drying 37 Ice jams 219
raw 73 Frost ring 214, 215 Ice movement 179,194-197
Dating 76 Fritts, H. C. 261 Identification of wood 166
accuracy 77 Fungal infection 32, 228 Image analysis 55
methods 76 Fungicidal substance 32 Increment borer 42
plantation 237 Function, exponential 87 Increment layers 3
problems 76 Hugershoff 87 Indexing 85
radiocarbon 76, 251 negative exponential 87 Indians 152-155, 212
Decomposition 33,34,37,45,47 polynomial 87 Insect galleries 33, 72, 145, 155
Defoliating 222-227 straight line 87 damage 222-235
Dendroclimatology 168, 175 Interval trend 83, 134
Dendroecology 54,95-141 Gas exchange 6 Inversion layers 246
Dendrochronology, applied 143- Geiger counter 65 Irradiation 65
253 Geographical distance 126, 128,
Dendrochronological age 248, 131 Katmai 214
249 Geological substrate 11, 23 Knife sharpener 40
Dendrometres 19, 237 periods 12 Krakatau 215
Drinking straw 43 Geomorphology 176-179
Dry sites 25 Giant waves 217 Laachen volcano 187,213
Dwarf tree 203 Glacier history 194-197 Labelling 43, 61
Gleichlaufigkeit 77, 83, 123, Lake shore settlement 149, 150
Earth movement 179 126-128 Landscape damage 45
Earthquakes 216 Grafting wax 44 Larix decidua 26, 29, 71, 195-
Earlywood 52, 71, 99, 100 Ground water 190 197
Ecogram 102,136,137,139 Growth change 7,48,239 Larix sibirica 56
Ecklund machine 53 abrupt 190, 191, 195, 211, Larch bud moth 223, 225
Ecological factors 19, 102 212,227,238 Larch 27, 29, 56, 70, 195-197
gradient 136 fluctuations 96 Latewood 52, 71, 99, 100, 112
limitation 16 increments 96 Lath 61
Ecophysiology 19 layers 96, 97, 109, 220, 221, Light influence 11
Elm 253,103 237 reflection 56, 57
Elm disease 228 model 140,141 slit 59
Entomology 222-227 reduction see Growth change transmittance 54, 56
Environmental sciences 240 -reduction diagrams 242, 244 Lightning 204
Erosion 179, 190 release see Growth change Lindbergh baby 167
Europe 128,169,172 rhythms 96, 237 Living cells 141, 176
Event diagram 48, 191, 198, maps 169 Log constructions 30
240,241,246 Logarithmic scale 51
Exposure time 65 Half-life time 248 Long chronologies 26
Extractives 62 Hand lens 52 Long term deviation 249
Hartig, T. 256 Lower forest limit 25, 170
Fagus sy!vatica 98, 100, 102 Hazelwood 52 Libby, W. F. 248
False ring 47, 205 Heartwood 57,101,146
Farm house 158, 160 History, Art- 162-165 Marks, Axe- 147
Felling date 145, 147, 152, 162- House- 156-160 Maximum density, mainly at
167 Settlement- 149, 152 pages 58-73, 114, 120, 123,
Fertilisation 234 Historical buildings 156-160 127, 128, 175
Film developing 64 Holding device 60 Mean values 79
Fire 35 Hollstein, E. 32 Measuring slit 70
map 209 Honeyfungus 229 Mechanical damage 44, 228,
scar 205-207 Horgen period 149 229
Filter 86, 87 House construction 32, 148, 149, Medieval ruin 31
band-pass 87 152,157,158 Mediterranean climate 133, 139
binomial 86 ground plans 149, 152, 153, Megalithic culture 251
Fitzroya cupressoides 26, 32 154 Meteorological data 92
Flag tree 200 Human activity 32, 144, 179 station 18
Floods 183, 192, 193, 210 Huber, B. 28, 260 Microdensitometer 66, 67
Fluoride 247 Hydrogen bomb 251 Microfiche reader 73
Flow chart 80 Hydrological analysis 180, 190, Microscopic technique 40
Fomes annosus 32 191 Microsections 46, 57
Forest fire 204-209 changes 179 Microstructure 28, 29, 32, 33,
Fossil fuels 250 Hydrology 23, 180-193, 216 35-37, 52, 96-105, 107, 109,
pine 36 111-115
soil 213 Ice age 13 variability 103
wood 96,220,221 Ice falls 217 Mineralisation 36
274
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 Topography 10 Volume increment 232, 239
wood 31,35,184-189,195- Transmitted light 56 Volumetric gravimetric density 62
197 Transportation 6, 148
Submerged forest 183, 192 Transport, cores 43 Weathered beam 38, 33, 148
Sulfate 247 Traumatic tissue 49 Weather reconstruction 24
Summer temperature 24, 25, 170, Tree form 106, 120 Wedge cellulose 68
171,175 Tree ring boundaries 11 0-115 optical 68
Sunset crater 212 characteristics 49, 107, 119, Wheat harvest 8
Sunspot cycles 174, 250 120,132 Wind influence 10, 106, 200-203
Supression 11, 212 Tropical trees 108, 109, 236- velocity 201
Surface cutting 40 239,251 Wood, compression, see
polishing 40 wood 47 compression wood
Synthetic resin 37 zone 105 identification 166
Tussock moth 224 preservation 32, 37
Technical error 72 Twin-bladed saw 61 technology 238, 239
Tectonic movement 193, 218 Tylosis 101, 228
Temperate climate 127, 139 X-ray absorption 62,69
zone 105, 111 Ulmus campestris 103 chamber 64
Temperature amplitude 8 Uplifting 218 equipment 64
Terminal ring 33, 145, 154 film 59,64
Tetrac/inis sp. 62 Varves 3 film coating 65
Thinning 232 Verification 90, 91 picture 61
Timber 156, 157, 160 Vessel area 54 source 65
re-use 148 Viscum album 228 technique 58-73
Timberline 16, 24, 136, 170 Visual data checking 74 tube 65
upper 24 Visual matching 49
Time lag 246. See also response Volcanic activity 179, 187, 210- Year, felling 145
function 215
Tissue analysis 40,54,55 ash 211 Zeiraphera diniana 222, 223, 225
276