Plant Science 184 (2012) 20–28

Contents lists available at SciVerse ScienceDirect

Plant Science
journal homepage: www.elsevier.com/locate/plantsci

Review

Is windswept tree growth negative thigmotropism?

Frank W. Telewski∗
W.J. Beal Botanical Garden and Campus Arboretum, Department of Plant Biology, Michigan State University, East Lansing, MI 48824, USA

a r t i c l e i n f o a b s t r a c t

Article history: The scientific investigation of the influence of wind on tree growth has been conducted for over 200 years.
Received 5 September 2011 One influence of wind on trees is the formation of an asymmetric crown, usually characterized as being
Received in revised form windswept under moderate windy conditions. As wind exposure increases, the terms applied to this
30 November 2011
growth form include flag-tree, banner-tree, and krummholz. The modification in crown morphology has
Accepted 1 December 2011
been widely recognized and studied, especial in the area of wind prospecting or as a bioindicator of wind
Available online 14 December 2011
speed in environments lacking monitoring stations. However, the causes and physiology underlying this
response is little understood. The windswept morphology is consistent with the morphologies associated
Keywords:
Thigmotropism
with other tropisms (i.e. phototropism and gravitropism). Tropisms are defined as a growth response
Flag-tree towards (positive) or away (negative) from an environmental stimulus. The asymmetric growth form
Biomechanics of windswept trees appears to be a negative thigmotropic growth response to wind. In this review,
Viscoelasticity evidence will be presented to support or reject two hypotheses; H1 the windswept growth form is the
Plastic deformation result of a negative thigmotropic growth response or H2 the windswept growth form is determined
by the biophysical properties of wood. It is argued that the windswept growth form is more likely the
product of biomechanical properties (accept H2 ) than of a physiological thigmotropic growth response
(reject H1 ). However, proper testing of both hypotheses is still required before a final confirmation can
be established.
© 2011 Elsevier Ireland Ltd. All rights reserved.

Contents

1. Introduction: wind and trees . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20

2. Thigmotropism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
3. Crown asymmetry: biomechanics, wood properties, and acclimation that reduces drag . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
3.1. Wood biomechanics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
3.2. Reaction wood . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
4. Hypotheses-biophysical vs. physiological . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
4.1. Physiological hypothesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
4.2. Biophysical hypothesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
5. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27

1. Introduction: wind and trees sway were shorter and had thicker trunks compared to stayed
trees were in 1803 [2]. A mechanistic theory of tree growth was
One of the earliest, if not the earliest description of the influ- proposed in 1893 stating wind was the most significant factor to
ence of wind on tree growth was recorded by Theophrastus in influence stem form in trees [3]. Over the ensuing 118 years, a mul-
300 BC [1]. He observed that trees growing in windy regions titude of researchers documented the influence of wind and other
exhibited stunted growth whereas trees growing in windless envi- mechanical perturbations on plant growth [4–9]. In 1973, Jaffe [10]
ronments grew taller. The first experiments conducted to restrict coined the term thigmomorphogenesis to describe the response
the movement of trees to wind reported that trees allowed to of plants to touch, brushing, wind, shaking, vibration and other
dynamic mechanical inducing stimuli. Thigmomorphogenesis is a
plant response to mechanical stimulation, usually resulting in a
∗ Tel.: +1 517 884 0764; fax: +1 517 432 1090. morphologically reduction in height and/or an increase in radial
E-mail address: telewski@cpa.msu.edu growth, a reduction in branch length and a reduction in leaf size

0168-9452/$ – see front matter © 2011 Elsevier Ireland Ltd. All rights reserved.
doi:10.1016/j.plantsci.2011.12.001
 F.W. Telewski / Plant Science 184 (2012) 20–28 21

[5]. The thigmomorphogenetic response in plants has been docu-

mented physiologically and requires an initial mechanoperception
phase followed by a physiological and metabolic cascade of events
that have recently been reviewed [6–8]. The response is very sen-
sitive and requires minimal stimulation [11–13]. Plants exposed to
chronic thigmostimulation also experience a physiological condi-
tioning, hardening, or accommodation reducing the intensity of the
physiological response to subsequent mechanical stimuli [8,14].
Wind is ubiquitous in the terrestrial environment and condi-
tions absent of any wind are extremely rare, if not non-existent.
Very low wind speeds in the range of 0.0018–3.2 km/h stim-
ulate growth by decreasing the boundary layer around leaves
and increasing the availability of carbon dioxide to photosyn-
thetically active leaves [15]. Higher wind speeds in the range of
7.9–97.2 km/h decrease plant growth [15] inducing a thigmomor-
phogenetic response. The staking or staying with guy wires of a
tree stem can prevent wind induced sway and eliminate the thig-
momorphogenetic response in the trunk [2,16–19]. Conversely,
artificially increasing wind drag on a tree increases the thigmo-
morphogenetic response [19]. Trees growing as subdominants or in
the subcanopy will experience far less wind than dominant canopy
trees. The dominants provide mutual support as well as protec-
tion from wind. Subdominants provided support by dominants can
actually exceed their critical height limit, experiencing buckling if
the dominant trees are removed [20–22].
Trees exposed to moderate winds of relatively equal distri-
bution from all directions will maintain a relatively symmetrical
crown and stem morphology while experiencing a thigmomorpho-
genetic response. In regions with moderate to strong prevailing
winds, which are especially characteristic of, but not restricted to
coastal areas, mountain tops, ridges, and river gorges, trees take
on a more pronounced asymmetric, windswept growth form. This
growth form is characterized by branches and in some instances,
stems curve away from the prevailing wind direction [23–33]. Stem
asymmetry can also be observed in trees growing in urban envi-
ronments where wind is redirected by large buildings and forced
to move bilaterally in the direction of the road. Crown asymme-
try ranges from moderate windswept branches, to the flagging of
crowns, to the extreme of krummholz (Fig. 1). The term krummholz
was first coined in 1898 [23] to describe the short, gnarled, and hor- Fig. 1. Two responses to wind conditions: Windswept upright European larch (Larix
decidua) and krummholz mugo pine (Pinus mugo var. pumilio) background, growing
izontal stems and branches of trees growing in subalpine regions.
near tree line in the Tatra Mountains of Slovakia.
The growth form can be determined either by genetics (dwarf
growth form) or winter desiccation induced by high wind speed
on unprotected foliage and buds [34]. as an indirect effect augmenting the formation of asymmetrical
As early as 1803, the stout characteristics of trees growing in crowns. Recent studies have added to the list of indirect stresses
windy regions were noted [2]. However, it appears that the first that can be involved in wind sculpting of crowns. In coastal regions
description of the windswept growth form was published in 1898 near salt water, the wind can carry salt spray, resulting in desic-
stating that in regions of strong prevailing winds, “the stems and cation of plant tissues augmenting the windswept growth form
branches are bent away from the prevailing wind direction” creat- [29,35]. Sand, snow, and ice crystals can abrade foliar and stem
ing this windswept growth form [23]. However, after over 200 years tissues resulting in physical injury [36].
of observation and research on the effect of wind on trees, only a few Further complicating our understanding of the formation of the
studies have addressed the potential mechanism of the windswept windswept growth form is seasonality or timing of the deforma-
growth form [25,36–39]. The subject is complex and as stated in tion. This was clearly demonstrated in a study of the vegetation of
a 1939 monograph; “The baffling intricacy of the influences that the Columbia River Gorge in the USA [26]. Two distinctly different
take part in the formation of these remarkable tree crowns is seen windswept crown morphologies occur in Douglas fir (Pseudotsuga
to involve climatological, aerological, mechanical, anatomical, and menziesii (Mirb.) Franco); storm-pruned and wind-trained or wind-
physiological conditions” [26]. bent. The storm-pruned growth form occur in the western part of
The physiological response of trees to a strong, unidirectional the canyon, characterized by many dead and broken branches and
wind is, indeed, complex. With wind pressure as the primary or many wound-induced epicormic branches emerging from dormant
direct stress, there are also a multitude of secondary or indirect buds under the bark on the downwind side of the trunk. The direc-
stresses involved in the sculpting of tree crowns [5]. The direct tion of the branch growth is consistent with winter winds. These
action of the wind resulting from tension and pressure in the liv- asymmetric growth forms develop in response to severe winter
ing tissues induces the windswept growth form [23]. Wind also storms and ice glaze injury [26]. On the contrary, the direction of
increases transpiration, increased exposure to freezing temper- the asymmetric crown of the wind-trained form occurring in the
atures by stripping the boundary layer, winter desiccation and eastern part of the canyon correlates well with summer winds and
mechanical damage (stripping of foliage and breaking of branches) does not show signs of severe physical injury or damage. In these
22 F.W. Telewski / Plant Science 184 (2012) 20–28

trees, branches on the upwind side of the stem are curved towards
the downwind side of the tree with little or no sign of visible physi-
cal damage. Similar effects of seasonal winds on crown asymmetry
were reported in other locales where the prevailing wind during the
growing season is the dominant factor in determining the direction
of the windswept growth form [41–43].
The degree of crown deformation correlates with prevailing
wind direction and speed [25]. These observations lead to the pub-
lishing of the Griggs-Putnam Index which estimated average wind
speed based on the degree of crown asymmetry [44]. This index
was proposed as a method of wind prospecting in the book Power
from the Wind [44]. This seminal work, which in its essence is a
dose response curve of crown deformation to applied force, has
given rise to numerous wind prospecting studies to document wind
speed and direction in numerous remote environments. It has also
stimulated the development and calibration of additional wind
classification schemes for specific tree species [30,32,33]. These
methods have also been applied in ecological studies for estimating
mean wind flow and airflow patterns in hilly and basin topography
lacking wind measuring instrumentation [30–33].
Although there is a role for wind-induced physical breakage of
branches and tissue death due to desiccation (due to water loss or
salt spray) or abrasion by sand and ice, not all wind swept trees
exhibit such extreme physical damage. Many trees show a gradual
to acute curvature away from the prevailing wind direction, with
Fig. 2. Permanent upwind bending of a lodgepole pine (P. controta Douglas ex
branches on the upwind side of the trunk curving into the down-
Louden) seedling subjected to high wind speed (30.6 k/h), grown in a growth room
wind side of the trunk and crown [25,42]. These branches are alive with artificial lighting from above, showing curvature of the buds into the wind.
and apparently undamaged externally. It is this type of windswept Curvature of the elongating bud is suggestive of a positive thigmotropic response to
growth form resulting in asymmetric crown development without wind.
secondary stress damage that will be the focus of the rest of this Source: Ref. [54], by permission of Oxford University Press.
paper.
To date, no one has convincingly reported how the windswept
growth form develops in the absence of secondary stresses. Is the mechanical contact by clinging and curving, as in tendrils” [46].
windswept growth form part of the thigmotropic response of trees The term has appeared in the literature since the late 19th century.
to wind? Or, is it the result of constant and persistent bending In plants, thigmotropism is a positive growth response towards
of tissues resulting in plastic deformation without a physiologi- touch or mechanical stimulus. Charles Darwin was one of the ear-
cal growth response? A multitude of recent investigations into the liest researchers to document thigmotropic responses in plants
response of trees to wind and the biomechanics of wood shed new such as that exhibited by roots, tendrils and twining plants [47].
light on the possible development of the windswept growth form. Thigmotropism can also occur in plant stems, resulting in a pos-
In this review, two hypotheses will be considered for the origin itive growth response towards the point of stimulation [48,49].
of the windswept growth form: a physiological hypothesis involv- In the thigmotropic physiological response, an asymmetric burst
ing mechanopreception and a subsequent negative thigmotropic of ethylene was reported in perturbed hypocotyls of cucumber
growth response (H1 ) and a biophysical hypothesis based in the (Cucumis sativus L.) 4 h after they are mechanically perturbed but
biomechanics of wood involving simple redirection of plant form prior to the thigmotropic curvature [49]. The application of exoge-
by a constant, unidirectional force applied by the wind (H2 ). nous indoleacetic acid (IAA) to the side of the hypocotyl receiving
mechanical perturbation did not inhibit the positive thigmotropic
growth curvature response. The role of ethylene-regulated touch
2. Thigmotropism responses was also found in the thigmotropic growth response of
roots [50]. Mechanosensing involved in thigmotropism appears to
In plants, a tropism refers to a specific directional growth in involve a mechanism similar to that described for thigmomorpho-
response to an environmental stimulus. A more detailed defini- genesis [51–53].
tion is; “Tropisms are growth curvatures of plant structures such Although studies that specifically focused on thigmotropism in
as stems, roots, coleoptiles, petioles, and tendrils in response to woody plants do not exist, one study on the effect of wind on
environmental stimuli of unequal intensity on two sides of the the growth of lodgepole pine (Pinus contorta Douglas ex Louden)
structure” [45]. Tropic movements do not include the movements reported that the expanding shoots curved into the wind (Fig. 2,
of plant organs due to differential cell turgidity, which includes the [54]). Although not acknowledged by the authors as such, this
opening and closing of the trap leaf of the Venus flytrap (Dionaea appears to be a positive thigmotropic response to a unidirec-
muscipula Ellis), sleep movements of leaves, or the rapid move- tional wind flow in the primary growth of a woody plant [54].
ment of the leaves of the sensitive plant (Mimosa pudica L.). These Another potential thigmotropic response may have been recorded
non-growth movements are nastic movements, the direction of in secondary tissues, but not reported as such: displaced stems
movement not being related to the direction of the stimulus. A of maritime pine (Pinus pinaster Aït.) exposed to wind experi-
positive tropic response involves directional growth towards the enced a greater rate of recovery in the middle and basal stem
stimulus where as a negative tropic response involves a directional sections compared to displaced stems not exposed to wind [55].
growth away from the stimulus. The increased rate of recovery under windy conditions could not
Thigmotropism describes the response of plants or animals to be explained by gravitropism and compression wood formation as
touch and is defined in plants as: “The tendency to respond to both wind and non-wind treatments contained equal amounts of
 F.W. Telewski / Plant Science 184 (2012) 20–28
compression wood in their displaced stems, but appears to be com-
plicated by a phototropic response [56,57].
3. Crown asymmetry: biomechanics, wood properties, and
acclimation that reduces drag
The ability of a tree to develop a windswept growth form
is an important acclimation response in windy environments.
Windswept crowns have a significant reduction in speed specific
drag, essentially creating a streamlined architecture that with-
stands high winds, shedding wind load and reducing the potential
for stem failure [38]. The degree of crown asymmetry is dose
specific [25,44]. Additionally, the ability of a tree to become stream-
lined is determined, at least in part, by genetics. Not all tree
species streamline or deform their crowns in response to the same
wind exposure. Trees that did not develop asymmetrical crowns
were defined as wind-resistant species and those that developed
windswept morphology, wind-sensitive [31]. A more appropriate
classification of these two groups should be wind tolerant and wind
avoidant species. The presence of tolerant and avoidant species sug-
gests that there are two strategies for trees to cope with wind;
wind-tolerant species appear to be more able to withstand wind
loading whereas the avoidant species avoid the mechanical loading
of the crown and stems by shedding the wind load via streamlining.
A hierarchy of species producing asymmetric crowns in response to
prevailing wind was noted: larch (Larix) was most sensitive, Dou-
glas fir (Pseudotsuga) and pine (Pinus) were intermediate, and of the
group, fir (Abies) was the least sensitive to wind-induced deforma-
tion [30].
3.1. Wood biomechanics
The biomechanical terms used in this review are listed in Table 1.
A quick comparison of data from two previously published sum-
maries [58,60] of green wood biomechanical properties for the
four genera (Larix, Pseudotsuga, Pinus, and Abies) discussed above,
are presented in Tables 2 and 3. All of the properties presented in
Tables 2 and 3 are measures of physical properties of wood. Wood
density and specific gravity (sp. gr.) are both measures of mass
per unit volume, with sp. gr. a decimal ratio of oven dry weight to
green volume. As a functional trait, wood density impacts the basic
xylem function of hydraulic conductivity and mechanical support.
The modulus of elasticity (MOE) quantifies the ability of a mate-
rial to deform under bending induced by loading without suffering
permanent deformation (elastic deformation). The modulus of rup-
ture (MOR) defines the point of failure under a load. Wood density
highly correlates with the mechanical properties of MOE, MOR, and
maximum strength in compression and shear [60]. Xylem with a
higher wood density tends to be more resistant to bending and
failure under mechanical loading. Based on the data presented in
Table 2, there appears to be a relationship between wood biome-
chanical properties and the degree to which a given species will
develop crown asymmetry in response to wind. Species with lower
strength properties appear to be less likely to be streamlined than
species with greater strength properties. A similar pattern, but not
as pronounced can be observed for trees presented in Table 3. Is it
possible that the lower strength properties of wood confer greater
flexibility to a tree, allowing it to bend more freely under wind load
and return to a more symmetrical orientation once it is released
from the pressure of the wind? A tree with stiffer wood will not
be as flexible and thus it may be more advantageous for trees with
stiffer wood to maintain windswept crown morphology in order to
maintain maximum drag reduction under windy conditions. This
avoidance mechanism may come with a trade-off. By reducing
drag in a permanently asymmetrical crown, the tree may sacrifice
23
photosynthetic capacity compared to a more symmetrical crown
tree growing under the same conditions.
In thigmomorphogenetic studies on trees, wood formed under
windy conditions or flexing is termed flexure wood [61]. In flexure
wood, the cellulose microfibril angle (MFA) increases and the MOE
of flexure wood is lower than in stems of the same species that
do not undergo dynamic mechanical loading [38–40,61]. There is
a direct relationship between microfibril angle, MOE and strain to
fracture in wood. An increase in MFA decreases MOE and increases
strain to fracture [62,63]. The change in MFA in flexure wood pro-
vides a tissue which is less stiff and more capable of absorbing
bending energy. However, an increase in wood production by a
windblown tree increases radial growth parallel to the direction of
applied bending load. The increase in radial growth increases the
second moment of inertia parallel to the loading vector, resulting
in an overall increase in flexural stiffness of the stem, increasing
resistance to bending in the direction of loading [38,39]. Due to
thigmomorphogenesis, the branch wood of trees growing in windy
environments should have a lower MOE imparting greater flexibil-
ity [38,39] but should be of greater cross-sectional area, imparting
greater stiffness to the branch. Comparative data from branches of
windblown trees needs to be collected and analyzed.
Another biomechanical factor important to the development of
windswept crowns is the viscoelastic property of the wood of each
species and how wind influences the viscoelastic property of newly
formed wood. Although wood formed under conditions of wind
sway has a lower MOE, there are no data with respect to how wind
alters the viscoelastic property of newly formed wood and how any
changes will alter the elastic and viscoelastic deformations of wind-
blown branches. Within the elastic range of branch displacement,
increased viscoelasticity should allow a deflected stem or branch to
remain displaced for a period of time before returning to its original
orientation.
As previously stated, the branches of tree crowns in continu-
ous exposure to a prevailing wind under conditions favorable for
growth (no salt exposure, desiccation, abrasion or physical break-
age of branches) are alive and apparently undamaged externally.
However, there is the potential for internal damage to a branch
that will facilitate the development of the windswept growth form.
If the force applied by wind to a branch exceeds the elastic limit of
the wood, then plastic deformation will occur. Successive loading
and unloading cycles on the mechanical properties of young Aris-
tolochia macrophylla tissues and the resulting elastic, viscoelastic
and plastic deformations were studied [64]. The application of a
tensional load altered the state of the stem tissue compared to its
original state, resulting in only a partial or slow return to the origi-
nal state. It was concluded that internal microstructural prestresses
were responsible for the observed mechanical behavior in response
to tension [64]. This permanent deformation was later investigated
and described as the stick-slip or Velcro mechanism in wood, a pro-
cess by which the cellulose microfibrils slip past each other and
adhere, reforming the amorphous matrix between the microfib-
rils within the cell wall [65]. Recently, the role of the viscoelastic
nature of wood in tropic movements and reaction wood formation
and function in response to light and gravity were investigated and
modeled indicating that viscoelastic properties contribute signifi-
cantly to the developmental morphology of trees [66].
3.2. Reaction wood
Woody plants alter the anatomy and mechanical properties
of wood in response or in reaction to environmental stimuli
including gravity, light and wind [5,55,56,67]. This wood is com-
monly referred to as reaction wood. Reaction wood in conifers and
angiosperms lacking vessels (non-porous wood) is characterized
by tracheids with a rounded cross-section, a thickened S2 layer of
24 F.W. Telewski / Plant Science 184 (2012) 20–28

Table 1
Wood and biomechanical terms used in this review.

Biomechanical term Definition

Green wood Wood at 50% moisture content, representative of wood in the living tree
Microfibrilar angle (MFA) The angle of cellulose microfibrils in the secondary cell wall with regard to the axis of the cell. An angle of 0◦ refers to a vertical
orientation, parallel to the vertical axis of the cell, 90◦ refers to a vertical orientation perpendicular to the vertical axis of a cell
Reaction wood Most commonly associated with wood formed in reaction to a gravitropic stimulus, but now accepted as a type of wood that
maintains the equilibrium position of a tree with regard to its physical environment including gravity, light, wind and other
mechanical loads
Opposite wood Type of wood formed on the side opposite to reaction wood in a branch or stem
Compression wood Type of reaction wood commonly found in conifers. Characterized by tracheids with a rounded cross-section, a thickened S2 layer
of the secondary cell wall with an increase in cellulose microfibrillar angle, and lignin content occurring on the lower or convex
side of stems and branches
Tension wood Type of reaction wood commonly found in the wood of angiosperms containing vessels, occurring on the side of the stem which
will become concave and is under tension. Characterized by fibers with a thickened secondary cell wall structure termed the
gelatinous layer or G layer with a decrease in cellulose microfibrillar angle
Flexure wood Type of wood formed in response to dynamic flexing such as that induced by wind sway, but without gravipreception.
Characterized by xylem cells with an increase in microfibrillar angle, but without compression or tension wood type secondary
cell walls
Earlywood Low density wood formed early in an annual growth ring
Latewood The more dense wood formed in an annual growth ring late during the growing season
Wood density Measure of mass per unit volume
Wood specific gravity Measure of mass per unit volume as a decimal ratio of oven dry weight to green volume
Elastic/plastic/viscoelastic An elastic material relaxes to its original form immediately after a deforming load is removed; a plastic material remains
deformed after the load has been removed; a viscoelastic material relaxes to its original form slowly after removal of a load. Many
materials have an elastic range, with plastic deformation occurring once the elastic limit is reached. Materials usually have a
combination of these properties, partially relaxing but retaining partial deformation
Modulus of elasticity (MOE) The modulus of elasticity E is a measure of material stiffness and quantifies the ability of a material to deform under bending
induced by loading without suffering permanent deformation (elastic deformation)
Second moment of area Also the second moment of inertia, I, measured in units of length to the fourth power parallel to the loading vector
Flexural stiffness The product of E * I, and expressed as EI, it is the overall stiffness of a structure, such as a stem or branch, this value is usually the
more ecologically significant to a plant
Modulus of rupture (MOR) The maximum fiber stress at failure, or yield point under load, having reached the elastic limit resulting in plastic deformation
Toughness The resistance of a material to fracture

Table 2
Species comparison to greenwood biomechanical properties. Species are listed from most sensitive to least sensitive to crown deformation due to wind.

Species Specific Modulus of Modulus of Work to maximum Impact bending Compression Shear parallel to
gravitya rupture (kPa) elasticity (MPa)b load (kJ/m3 ) (mm) perpendicular to grain (kPa)
grain (kPa)

Western Larch 0.55–0.48 60,000–53,000 11,400–10,100 71 740 30,500–25,900 6300–6000

(Larix occidentalis)c
Tamarack 0.48 47,000 8600 – – 21,600 6300
(Larix laricina)c
Coastal Douglas fir 0.45 53,000 10,800 52 660 26,100 6200
(Pseudotsugamenziesii)
Ponderosa pine 0.38 51,000 6900 36 530 16,900 4800
(Pinus ponderosa)
Balsam Fir 0.34 36,000 7800 32 410 16,800 4700
(Abies balsamea)

Wood property data from Kretschmann [58].

a
Specific gravity is based on weight when oven dry and volume when green.
b
Modulus of elasticity measured from a simply supported, center-loaded beam, on a span depth ratio of 14/1. To correct for shear deflection, the modulus can be increased
by 10%.
c
Wade and Hewson [30] when discussing sensitivity to wind only reference Larix sp. but Owada [59] specifically studied Larix leptolepis (L. kaempferi). Unfortunately,
greenwood data for this species was not presented in Kretschmann [58].

Table 3
Species comparison to wood biomechanical properties at 50% moisture content (greenwood). Species are listed from most sensitive to least sensitive to crown deformation
due to wind.

Species Density,  Modulus of rupture, Mg Modulus of elasticity, E Maximum compressive Maximum shear
(kg/m3 ) (MN/m2 ) (MN/m2 ) strength Sc (MN/m2 ) strength Sg (MN/m2 )

European Larch 673 53 7,900 24.3 6.9

(Larix decidua)
Japanese larch 609 48 6,800 21.9 6.1
(Larix kaempferi, syn L.
leptolepis)
Coastal Douglas fir 625 53 8,300 24.6 6.8
(Pseudotsugamenziesii)
Ponderosa pine 561 38 6,000 17.7 4.8
(Pinus ponderosa)
Balsam Fir (Abies balsamea) 529 43 8,100 20.8 5.7

Wood property data from Niklas and Spatz [60].

 F.W. Telewski / Plant Science 184 (2012) 20–28
the secondary cell wall with an increase in cellulose microfibril-
lar angle, and lignin content [62,67]. This type of reaction wood
develops a compressive growth strain on what will develop as the
convex side of a curving stem and is termed compression wood.
In angiosperms containing vessels (porous wood), reaction wood
occurs on the side of the stem which will become concave and is
under tension. The type of reaction wood is termed tension wood
and is characterized by fibers with a thickened secondary cell wall
structure termed the gelatinous layer or G layer with a decrease in
cellulose microfibrillar angle. Upon maturation, tension wood gen-
erates a tensional force on the side of the stem which will become
concave [67].
The formation of reaction wood is most commonly associ-
ated with gravitropism in displaced stems and in maintaining
branch position with regard to gravity and the trunk support-
ing the branch. The long-term stability of branches and displaced
stems depends on a balance of the gravitropic response between
increasing self-weight and resultant wood properties and gener-
ated growth strains [68]. However, reaction wood can form in
stems and branches as a response of the tree that returns the dis-
placed organ(s) to “its equilibrium position in space” [67]. When a
branch of a white pine (Pinus strobes L.) is removed from a whorl
of branches, the remaining branches form compression wood on
their flanks away from the gap [69]. The compression wood induces
movement of the branches to fill the gap. This study supports the
role of reaction wood in maintaining an equilibrium position of
plant organs and also implicates a role for photomorphogenesis.
More recent studies support the role of reaction wood formation
and function in maintaining an equilibrium position with regard
to a plants physical environment including gravity, light and wind
[57,58,68]. Trees growing in windy environments tend to contain
greater amounts of reaction wood [67]. Strong, prevailing wind,
usually associated with cyclonic or frontal weather systems is well
known to displace tree trunks, creating a static lean in trees. This
displacement is then stabilized or corrected by a gravitropic induc-
tion of reaction wood formation which returns the stem to its
vertical orientation.
Trees exposed to a prevailing wind produce a pronounced
eccentric growth of xylem on the downwind side [16]. Similar
results were obtained in young tamarack trees (Larix laricina (Du
Roi) K. Koch) grown in a greenhouse and exposed to a unilateral
wind created by an electric fan [37]. The stems of the windblown
trees contained compression wood over most of their length. Trees
exposed to a moderate, multidirectional wind, did not form com-
pression wood. A minimum presentation time was required to
induce compression wood formation under windy conditions [37].
Wind can induce compression wood formation when stem dis-
placement is sufficient to meet a required presentation time, but
wind sway alone does not induce compression wood formation
[67]. This was later supported in a study on Fraser fir (Abies fraseri
(Pursh) Poir.) in which dynamic sway without prolonged or static
displacement of the stem from the vertical position (gravi-neutral)
sufficient to meet the requirement of a presentation time did not
induce reaction wood formation [61]. The wood that is formed,
termed flexure wood, is intermediate in anatomical and biome-
chanical properties between normal and reaction wood.
The biomechanical properties of compression and opposite
wood in horizontal conifer branches have been the subject of recent
research [70,71]. These studies documented for the first time the
important role of opposite wood in the biomechanical function of
wood in branches in maintaining their orientation with regard to
the main trunk and gravity. Within the same cross-section of a
branch, opposite wood has a lower MFA and thus a greater stiff-
ness compared to compression wood. Changes in MFA and stiffness
were reported with increased distance from the trunk and with
age of the wood formed [70]. Branch wood (both compression and
25
Fig. 3. Cross-section of spruce branches exposed to prevailing winds. Arrows indi-
cate wind direction. (A) Cross-section of a spruce branch exposed to a westerly wind
for eight years. Note the bands of compression wood on the left hand side which cor-
responds to the downwind flank of the branch. (B) Cross-section of a spruce branch
exposed to wind for four years from below. Note the bands of compression wood
which formed on the upper portion (downwind) of the branch.
Reproduced from Ref. [74].
opposite) exhibited the greatest MFA and flexibility at the branch
junction closest to the trunk resulting in wood “with a relatively low
stiffness and strength, but a higher flexibility and toughness” [70].
This increased flexibility and toughness is critical to the function
of a branch in response to mechanical self-loading (cone crop and
continuous growth) and loading due to snow and wind (or a large
flock of birds), especially in the function of branches to damping
wind-induced sway of the entire tree [72] and by protecting the
stem from load transmission from the branch [73]. Branch junc-
tions were reported to be more swollen in trees growing in windy
environments than in more calm environments [38]. The branch
junction of a young pine tree exposed to wind was analyzed using
computer tomography to determine density, and for microfibrilar
angle and compared to a branch junction from a more protected
tree [73]. The study revealed the junction exposed to wind exhib-
ited increased radial growth, with an increase in latewood density
and MFA in the latewood of the outermost growth rings on the
upwind side when compared to junctions from a less exposed site.
Wind can also induce compression wood on the downwind side
of displaced branches, or in the case of branches blown upwards
by the wind, where the compression wood developed on the upper
side of the displaced branch (Fig. 3, [74]). This is counter to what
would be anticipated if compression wood formation was strictly
a response to gravity.
In conifers, the formation of compression wood on the down-
wind side of stems and branches displaced by wind confounds
the interpretation of the development of the windswept form. If
compression wood creates a compressive force via internal growth
strains which returns a displaced branch or stem back to “its equi-
librium position in space” [67], then the windswept growth form
should not be maintained. Stems should be returned to the verti-
cal position and branches should be returned to their symmetrical
distribution around the stem. This conundrum was recognized as
the tendency for a bent branch to straighten and to resist a unilat-
eral wind induced pressure through reaction wood formation [26].
A 72-year-old statement is still valid: “Much experimental study
of pressure effects on trees, as well as much observational study
of the natural development of one-sided crowns, will be required
before the phenomena here considered superficially may be satis-
factorily understood [26].” We know even less about the formation
of asymmetric crowns in angiosperms.
4. Hypotheses-biophysical vs. physiological
With the observations in the above text box in mind, two
hypotheses are proposed to explain the formation of the asymmet-
ric, windswept growth form in trees.
26 F.W. Telewski / Plant Science 184 (2012) 20–28
Summary of wind effects
1. Trees (conifers and angiosperms) can form asymmetric
crowns when exposed to moderate to strong prevailing
winds during favorable growth conditions without apparent
physical injury.
2. Asymmetric crowns are effectively streamlined reducing
wind loading by reducing drag thereby reducing the poten-
tial for failure.
3. The degree to which a tree will form an asymmetric crown
is dose specific (determined by wind speed and duration).
4. The degree to which a species will form an asymmetric
crown appears to be genetically determined and related to
the biomechanical properties of its wood.
5. Wood formed in response to wind sway has a lower elastic
modulus (greater flexibility) and higher microfibrilar angle.
6. Wood can undergo elastic, viscoelastic and plastic deforma-
tions in response to loading resulting in internal changes in
the biomechanical properties of the wood via the stick-slip
mechanism.
7. The expanding shoots of pines are negatively gravitropic,
positively phototropic and may be positively thigmotropic.
8. Conifer trees form compression wood on the downwind side
of stems and branches displaced by wind.
9. Wind appears to accelerate the corrective ability of com-
pression wood to return a displaced stem to the vertical
position
4.1. Physiological hypothesis
H1 : The windswept habit of tree canopies growing in environ-
ments dominated by a prevailing wind is the result of negative
thigmotropism. The branches and stem perceive the asymmetric
wind pressure and stimulate growth on the downwind side to curve
branches away from the prevailing wind direction.
As in the case of other tropic movements, such as phototropism
or gravitropism, differential growth is established by a gradient
in plant growth regulators, and is the driving force behind the
movement or induction of curvature of a plant organ in response
to the environmental stimulus [75–78]. This is also the true for
thigmotropism [79]. Lateral branches are plagiogravitropic, main-
taining a position in equilibrium with gravity, light and the tree
trunk. Reaction wood forms within branches to maintain this equi-
librium. At present, it is unknown if a differential distribution of
plant growth regulators develops in wind-deflected branches. Cer-
tainly, the presence of compression wood on the downwind flank
of horizontally deflected branches of conifers strongly supports the
existence of such a gradient. In wind-deflected stems, which are
negatively gravitropic, the situation is complicated as the displaced
stem is responding to a gravistimulus in addition to a mechanos-
timulus induced by wind and a photostimulus [5,55–57]. However,
based on the current understanding of the function of compression
wood, the location of compression wood on the downwind flank
of branches and stems would appear to function to return these
organs to their equilibrium position [67].
In stems, as opposed to branches, light also plays a confounding
role in terms of phototropism and heliotropism impacting mor-
phology and anatomy. A heliotropic entrained circumnutation of
the young expanding shoot of loblolly pine (Pinus taeda L.) was pre-
viously reported [80]. This circumnutational pattern was proposed
to be responsible for inducing the formation of a spiral pattern
of compression wood in the tissue below the nutating, expand-
ing shoot. Similarly, wind can interact with both gravity and light
in the corrective response of displaced stems of maritime pine
(Pinus pinaster) seedlings. Stems displaced by an artificial wind
source in a direction away from the sun angle curved southwards
towards the natural light source [56] indicating an interaction
between wind displaced stems, gravity, and light. However, neither
of these studies investigated the influence of light on the forma-
tion of windswept branches. The direction of branch orientation
of windswept trees growing in windy environments consistently
points down wind of the prevailing wind direction, regardless of
the direction of the light source.
Observations suggest that woody plants may have a positive
thigmotropic response in both primary and secondary tissues
[54,55]. The curvature of the expanding pine shoot into a prevail-
ing wind (Fig. 2, [54]) could be an over compensation reaction to
displacement by the wind with regard to gravity. It might also be
an interaction with phototropism as other studies might suggest
[56,57]. However, in the case of the study presented in Fig. 2, this
is highly unlikely since the trees in this experiment were grown in
a growth room with artificial light coming from directly above the
trees [54]. No conclusions can be drawn regarding this observation
at the present time. Additional research on the tropic movements
of expanding shoots of conifers is required.
The pressing question with regard to the maintenance of an
asymmetric, windswept growth form when the wind is not actively
exerting pressure on the crown is how is the asymmetry main-
tained? If primary growth is positively thigmotropic as previously
suggested herein, and wind is capable of accelerating the forma-
tion of reaction wood and the corrective, reorientation growth in
stems [55], and in conifers, compression wood forms on the down-
wind flank of displaced branches and stems [67,74], then what is
maintaining a branch in the windswept, downwind position?
Interpretation of a biological role in the development and main-
tenance of the windswept growth form is further complicated by
the timing of formation. Most trees growing in a windy environ-
ment are continuously exposed to this mechanical stimulus and
all tissue development, from juvenile to mature, is impacted dur-
ing ontogeny [81]. A second scenario involves a tree previously
protected from a strong prevailing wind and then exposed to that
wind as a more mature organism. Tissues previously formed under
non- or low-wind conditions are fixed and cannot alter their biome-
chanical characteristics biologically. However, the tree, specifically
the branches, deflect and form reaction wood in the newly formed
xylem [74]. Is it possible that in both cases the ontogenetic pattern
of mechanical properties of the differentiating wood formed after
deflection, including any reaction wood and opposite wood, is actu-
ally maintaining the deflected growth form, rather than returning it
to its original orientation? And if so, can this modified growth that
maintains curvature, but not necessarily induces the curvature, be
considered negatively thigmotropic?
4.2. Biophysical hypothesis
H2 : The windswept habit of tree canopies growing in envi-
ronments dominated by a prevailing wind is the result of the
biomechanical properties of the wood, including strength and elas-
tic, viscoelastic and plastic deformation. Physical pressure from the
wind, exerted on the branches and stem, continually reposition the
tree in a windswept position which cannot be returned to the orig-
inal equilibrium position by the development of reaction wood.
Subsequent plastic deformation and new wood formation hold the
branches and stem in the windswept position.
Much of the collected data and observations presented
above could support a biophysical hypothesis. As presented in
Tables 2 and 3, there appears to be an association between
biomechanical properties and susceptibility to development of a
windswept crown. At least for conifers, stronger wood trees are
more susceptible to streamlining that shed wind loading. The dif-
ferences in biomechanical wood properties also account for the
 F.W. Telewski / Plant Science 184 (2012) 20–28
observed genetic difference in species response to wind. These
properties could also account for the dose response previously
observed and used to calibrate the estimation of wind speed in
areas devoid of meteorological instrumentation [44]. Although vis-
coelastic data are not available for comparison, a compilation of
these data would be useful in further testing this hypothesis. The
trunk and branches of tree with greater viscoelasticity would tend
to remain deformed for a longer period before returning to their
original position.
Plasticity can be conferred from the MOR. If plasticity is to play
a significant role in the formation of the windswept growth form
then a lower MOR should be anticipated in the wood of trees sensi-
tive to developing asymmetric crowns. This is not supported by the
data presented in Tables 2 and 3. However, the data presented in
Tables 2 and 3 are general biomechanical data for each species, and
not specific for wood differentiating from the cambium under the
successive loading–unloading cycles associated with wind induced
flexure. The wood of hybrid poplar (Populus trichocarpa Torr. & A.
Gray. × P. deltoides Bartr. ex Marsh.) differentiated in response to
mechanical flexing has a lower MOR compared to the wood of non-
flexed control trees [82]. This is also true for the wood of conifers
[40,61]. Presently, these data do not exist for branch wood in a
multi-species comparison. Measurements need to be obtained from
branches and stems for different trees species exhibiting different
degrees of sensitivity to windswept growth formation.
Measurements of growth strains and the development of
growth strains in deflected branches must be collected to further
elucidate the role of reaction wood formation in branches under
moderate to strong prevailing wind conditions. The challenge of
this hypothesis is to understand how the biomechanical properties
of wood in a windswept tree overcome the biological responses
of reaction wood formation which would appear to counter the
formation of an asymmetric crown. The formation of compres-
sion wood in a downwind flank of a wind displaced branch should
return the branch to its original position, countering the windswept
growth form. However, such a return of branches in windswept
trees to the original position has not been observed in nature.
Both viscoelastic and plastic deformation would provide time
for differentiating xylem to secure the branch or trunk in the
deflected position. This would also account for why windswept
growth develops without significant external damage due to other
physical stresses when formed during the growing season [26].
However, once again, the observation that compression wood
forms on the downwind side of wind deflected branches and stems
(Fig. 3B, [74]) does not support this hypothesis. Some light can
be shed on this paradox with a new understanding of the biome-
chanical properties of wood. Compression wood samples in the
wet condition, subjected to tensile deformation beyond the elastic
limit maintain their original stiffness [65]. The authors of this study
suggest within wood exists a recovery mechanism that allows the
reforming of the amorphous matrix linking cellulose microfibrils
within the cell wall, thus maintaining the mechanical properties.
The authors go on to describe this phenomenon as the slip-stick
mechanism akin to Velcro [65]. These results were confirmed for
wet compression wood from Norway spruce (Picea abies) in a cycli-
cal loading study [83]. These observations provide a mechanism
that allows for the movement of a branch or stem without physi-
ological activity, given that tracheids are dead at maturity. It also
allows for the alteration of wood in multiple annual growth rings.
Is it possible that the continuous cycling of wind load on branches
will induce the slip-stick mechanism in compression wood of wind-
blown branches overcoming the physiological growth strain which
was originally induced by reaction wood, resulting in a permanent
displaced windblown branch? The uncertainty with this concept is
that the tests in the slip-stick study were conducted in tension. The
further bending of wind deflected branches by wind would place
27
the compression wood on the downwind flank of a branch under
further compressive load, not a tensile load. It is not clear if the
slip-stick mechanism will function under a compressive load.
5. Conclusions
By definition, thigmotropism is a directional growth response
of a plant organ to a mechanical stimulus, such as touch. If wind
is accepted as a mechanical stimulus as in the case of thigmo-
morphogenesis, than wind should be appropriately included as a
mechanical stimulus in the case of thigmotropism. To date, no one
has documented a thigmotropic response to a unilateral wind pres-
sure in any plant system, although some work [54,55] suggests one
may exist in conifers. The majority of evidence that exists regard-
ing the windswept growth form is derived from studies on conifers.
This evidence suggests that the asymmetry of tree crowns exposed
to moderate to strong prevailing winds can be explained solely by
the action of wind and the physical properties of wood, especially
viscoelastic and plastic deformation and possibly the slip-stick
mechanism. However, the formation of compression wood on the
downwind flank of deflected conifer branches [74], and the accel-
erated growth movement in wind exposed displaced stems [55]
indicates the existence of a physiological growth response that
counters wind induced displacement. Clearly, the role of compres-
sion wood in the windswept growth form is not fully understood.
Even less understood is the role of tension wood in angiosperm
trees. However, the majority of evidence available strongly suggests
that the windswept growth form is more likely due to biophysical
viscoelastic and plastic deformation and slippage and subsequent
re-adherence between cellulose microfibrils in mature xylem cells,
if the slip-stick mechanism functions under a compressive load. It
appears unlikely that a physiological growth response induces cur-
vature of branches and stems in response to wind. Although further
investigation into this phenomenon is required, it is unlikely that
the windswept growth form represents a negative thigmotropic
response.
Acknowledgements
This review is dedicated to the memory of Dr. Mordecai (Mark) J.
Jaffe. Mark provided me with the knowledge, tools, and experience
to pursue my career in plant biology and to continue his work in the
field of plant response to mechanical stimuli. The author also thanks
Drs. Hans-Christof Spatz, Jonathan Gressel, Jameel Al-Haddad and
two anonymous reviewers for their helpful comments to improve
this manuscript.
References
[1] Theophrastus, De causis plantarum: In Three Volumes, Trans. B. Einarson, G.K.K.
Link, Harvard University Press, Cambridge, MA, 1976.
[2] T. Knight, Account of some experiments on the descent of sap in trees, Philos.
Trans. R. Soc. London 93 (1803) 277–289.
[3] C. Metzger, Der Wind als maßgebender Faktor für das Wachsthum der Bäume,
Mündener Forstliche Hefte 3 (1893) 35–86.
[4] C.A. Mitchell, P.N. Myers, Mechanical stress regulation of plant growth and
development, Hort. Rev. 17 (1995) 1–42.
[5] F.W. Telewski, Wind-induced physiological and developmental responses in
trees, in: M.P. Coutts, J. Grace (Eds.), Wind and Trees, Cambridge University
Press, Cambridge, 1995, pp. 237–263.
[6] F.W. Telewski, A unified hypothesis of mechanoperception in plants, Am. J. Bot.
93 (2006) 1466–1476.
[7] E. de Langre, Effects of wind on plants, Annu. Rev. Fluid Mech. 40 (2008)
141–168.
[8] C. Coutand, Mechanosensing and thigmomorphogenesis, a physiological and
biomechanical point of view, Plant Sci. 179 (2010) 168–182.
[9] W.E. Chehab, E. Eich, J. Braam, Thigmomorphogenesis: a complex plant
response to mechano-stimulation, J. Exp. Bot. 60 (2009) 43–56.
[10] M.J. Jaffe, Thigmomorphogenesis: the response of plant growth and develop-
ment to mechanical stimulation, with special reference to Bryonia dioica, Planta
144 (1973) 143–157.
28 F.W. Telewski / Plant Science 184 (2012) 20–28
[11] M.J. Jaffe, R. Biro, K. Bridle, Thigmomorphogenesis: calibration of the parame-
ters of the sensory function in beans, Physiol. Plant 49 (1980) 410–416.
[12] F.W. Telewski, M.L. Pruyn, Thigmomorphogenesis: a dose response to flexing
in Ulmus americana seedlings, Tree Physiol. 18 (1998) 65–68.
[13] C. Coutand, L. Martin, N. Leblanc-Fournier, M. Decourteix, J.-L. Julien, B. Moulia,
Strain mechanosensing quantitatively controls diameter growth and PtaZFP2
gene expression in poplar, Plant Physiol. 151 (2009) 223–232.
[14] F.W. Telewski, M.J. Jaffe, Thigmomorphogenesis: changes in the morphology
and chemical composition by mechanical perturbation in 6-month-old Pinus
taeda seedlings, Can. J. Forest Res. 11 (1980) 380–387.
[15] R.M. Wadsworth, An optimum wind speed for plant growth, Ann. Bot. 23 (1959)
195–199.
[16] M.R. Jacobs, The effect of wind sway on the form and development of Pinus
radiata D. Don, Aust. J. Bot. 2 (1954) 35–51.
[17] P.L. Neel, R.W. Harris, Motion-induced inhibition of elongation and induction
of dormancy in Liquidambar, Science (1971) 58–59.
[18] N.M. Holbrook, F.E. Putz, Influence of neighbors on tree form: effects of lat-
eral shade and prevention of sway on the allometry of Liquidambar styraciflua
(Sweet Gum), Am. J. Bot. 76 (1989) 1740–1749.
[19] E. Valinger, Effects of wind sway on stem form and crown development of Scots
pine (Pinus sylvestris L.), Aust. Forest 55 (1992) 15–21.
[20] L. Tirén, Einige Untersuchungen über die Schaftform, Medd Stat Skogsförsök-
sanst 24 (1928) 81–152.
[21] H. Härdtl, Die Gestalt der Pflanzen in ihrer Abhängigkeit von der Schwerkraft,
Forstwiss Cbl 58 (1936) 765–775.
[22] H. Härdtl, Die formbildende Wirkung der Eigenlast von Stamm und Zweigen
bei einem Nadelbaum Forstwiss Cbl 60 (1938) 343–352, 424–429.
[23] A.F.W. Schimper, Pflanzen-Geographie auf physiologischer Grundlage, Verlag
von Gustav Fischer, Jena, 1898.
[24] M.S.W. Jefferson, Wind effects, J. Geog. 3 (1904) 2–20.
[25] R.F. Griggs, The timberlines of northern America and their interpretation, Ecol-
ogy 27 (1938) 275–289.
[26] D.B. Lawrence, Some features of the vegetation of the Columbia River Gorge
with special reference to asymmetry in forest trees, Ecol. Mono. 9 (1939)
217–257.
[27] R.F. Daubenmire, Vegetational zonation in the rocky mountains, Bot. Rev. 9
(1943) 325–393.
[28] T.d.A. Richmond, D. Mueller-Dombois coastline ecosystems on Oahu, Hawaii,
Plant Ecol. 25 (1972) 367–400.
[29] A.P. Smith, Notes on wind-related growth patterns of paramo plants in
Venezuela, Biotropica 4 (1972) 10–16.
[30] J.E. Wade, E.W. Hewson, Trees as a local climatic wind indicator, J. Appl. Mete-
orol. 18 (1979) 1182–1187.
[31] Y. Noguchi, Deformation of trees in Hawaii and its relation to wind, J. Ecol. 67
(1979) 611–628.
[32] A. Robertson, Estimating mean windflow in hilly terrain from Tamarack (Larix
laricina (Du Roi) K. Koch) deformation, Int. J. Biometeor. 30 (1986) 333–349.
[33] G. Wooldridge, R. Musselman, B. Connely, D. Fox, Airflow patterns in a small
subalpine basin, Theor. Appl. Climatol. 45 (1992) 37–41.
[34] F.-K. Holtmeier, What does the term krummholz really mean? Observations
with special reference to the Alps and the Colorado Front Range, Mt. Res. Dev.
1 (1981) 253–260.
[35] B.W. Wells, I.V. Shunk, Seaside shrubs; wind forms vs. spray forms, Science 85
(1937) 499.
[36] L.G. Klikoff, Microenvironmental influence on vegetational pattern near tim-
berline in the Central Sierra Nevada, Ecol. Mon. 35 (1965) 187–211.
[37] P.R. Larson, Stem form of young Larix as influenced by wind and pruning, Forest
Sci. 4 (1965) 412–424.
[38] F.W. Telewski, M.J. Jaffe, Thigmomorphogenesis: field and laboratory studies
of Abies fraseri in response to wind or mechanical perturbation, Physiol. Plant
66 (1986) 211–218.
[39] M.L. Pruyn, B.J. Ewers, F.W. Telewski, Thigmomorphogenesis: changes in the
morphology and mechanical properties of two Populus hybrids in response to
mechanical perturbation, Tree Physiol. 20 (2000) 535–540.
[40] G. Koch, J. Bauch, J. Puls, E. Schwabe, Biological, chemical, and mechanical char-
acteristics of ‘Wulstholz’ as response to mechanical stress in living trees of Picea
abies [L.] Karst, Holzforschung 54 (2000) 137–143.
[41] M.M. Yoshino, Wind-shaped trees in the subalpine zone in Japan, Arct. Alpine
Res. 5 (1973), pp. A155-A126.
[42] T. Sekiguti, Studies on local climatology, III. On the prevailing wind in early sum-
mer judged by bending shapes of top-twigs of persimmon trees at the Akaho
Fan Nagano Prefecture, Japan, Pap. Meteorol. Geophys. 2 (1951) 168–179.
[43] E.W. Holroyd III, Prevailing winds on Whiteface Mountain as indicated by flag
trees, Forest Sci. 16 (1970) 222–229.
[44] P.C. Putnam, Power from the Wind, D. Van Nostrand Reinhold, New York, 1948.
[45] V.A. Greulach, Plant Function and Structure, MacMillan Publishing Co., Inc., New
York, 1973.
[46] I.F. Henderson, W.D. Henderson, A Dictionary of Scientific Terms, D. Van Nos-
trand Company, New York, 1920.
[47] C. Darwin, F. Darwin, The Power of Movement in Plants. The Power of Move-
ment in Plants, D. Appleton and Company, New York, 1881.
[48] M. Huberman, M.J. Jaffe, Thigmotropism in organs of the bean plant (Phaseolus
vulgaris L.), Ann. Bot. 57 (1986) 133–137.
[49] H. Takahashi, J. Jaffe, Thigmotropism and the modulation of tropistic curvature
by mechanical perturbation in cucumber hypocotyls, Physiol. Plant 80 (1990)
561–567.
[50] C. Yamamoto, Y. Sakata, T. Taja, T. Baba, S. Tanaka, Unique ethylene-regulated
touch responses of Arabidopsis thaliana roots to physical hardness, J. Plant Res.
121 (2008) 509–519.
[51] J.M. Fasano, G.D. Massa, S. Gilroy, Ionic signaling in plant responses to gravity
and touch, J. Plant Growth Regul. 21 (2002) 71–88.
[52] C.A. Esmon, U.V. Pedmale, E. Liscum, Plant tropisms: providing the power of
movement to a sessile organism, Int. J. Dev. Biol. 49 (2005) 665–674.
[53] A. Eshel, A. Katz, Y. Waisel, N. Ohad, Roots of tomato respond to mechan-
ical stimulation by induction of touch genes, Plant Biosyst. 139 (5) (2005)
209–213.
[54] D.J. Rees, J. Grace, The effects of wind on the extension growth of Pinus contorta
Douglas, Forestry 53 (1980) 145–153.
[55] S. Berthier, A. Stokes, Righting response of artificially inclined maritime pine
(Pinus pinaster) saplings to wind loading, Tree Physiol. 26 (2006) 73–79.
[56] S. Berthier, A. Stokes, Phototropic response induced by wind loading in Mar-
itime pine seedlings (Pinus pinaster Aït.), J. Exp. Bot. 56 (2005) 851–856.
[57] R. Herrera, C. Krier, C. Lalanne, E.M. Ba, A. Stokes, F. Salin, T. Fourcaud, S. Claverol,
C. Polmion, (Not) Keeping the stem straight: a proteomic analysis of maritime
pine seedlings undergoing phototropism and gravitropism, BMC Plant Biol. 10
(2010) 217. doi:10.1186/1471-2229-10-217.
[58] D.E. Kretschmann, Chapter 5: mechanical properties of wood, in: Wood Hand-
book, Wood as an engineering material, General Technical Report FPL-GTR-190,
US Department of Agriculture, Forest Service, Forest Products Laboratory,
Madison, WI, 2010, pp. 5-1–5-44.
[59] M. Owada, Prevailing winds in southeastern Hokhido and Donse-Genya, Geogr.
Rev. Jpn. 46 (1973) 505–515.
[60] K. Niklas, H.-C.H. Spatz, Worldwide correlations of mechanical properties and
green wood density, Am. J. Bot. 97 (2010) 1587–1594.
[61] F.W. Telewski, Structure and function of flexure wood in Abies fraseri, Tree
Physiol. 5 (1989) 113–121.
[62] A. Reiterer, H. Lichtenegger, S. Tschegg, P. Fratzl, Experimental evidence for a
mechanical function of the cellulose microfibril angle in wood cell walls, Philos.
Mag. A 79 (1999) 2173–2184.
[63] P. Fratzl, I. Burgert, H.S. Gupta, On the role of interface polymers for the
mechanics of natural polymeric composites, Phys. Chem. Chem. Phys. 6 (2004)
5575–5579.
[64] H.-C.H. Spatz, L. Köhler, K.J. Niklas, Mechanical behavior of plant tissues: com-
posite materials or structures? J. Exp. Biol. 202 (1999) 3269–3272.
[65] J. Keckes, I. Burgert, Frühmann, M. Müller, K. Kölln, M. Hamilton, M. Burgham-
mer, S.V. Roth, S. Stanzl-Tschegg, P. Fratzl, Cell-wall recovery after irreversible
deformation of wood, Nat. Mater. 2 (2003) 810–814.
[66] C. Coutand, J-D. Mathias, G. Jeronimidis, J-F. Destrebecq, TWIG: a model
to simulate the gravitropic response of a tree axis in the frame of elastic-
ity and viscoelasticity, at intra-annual time scale, J. Theor. Biol. 273 (2011)
115–129.
[67] T.E. Timell, Compression Wood in Gymnosperms, vol. 3, Springer-Verlag, Berlin,
1986.
[68] T. Alméras, M. Fournier, Biomechanical design and long-term stability of
trees: morphological and wood traits involved in the balance between weight
increase and the gravitropic reaction, J. Theor. Biol. 259 (2009) 370–381.
[69] A.H. Westing, Compression wood in the regulation of branch angle in gym-
nosperms, Bull Torrey Bot. Club 92 (1965) 62–66.
[70] J. Färber, H.C. Lichtenegger, A. Reiterer, S. Stanzl-Tschegg, P. Fratzl, Cellulose
microfibril angles in a spruce branch and mechanical implications, J. Mater.
Sci. 36 (2001) 5087–5092.
[71] I. Burgert, K. Jungnikl, Adaptive growth of gymnosperm branches-
ultrastructural and micromechanical examinations, J. Plant Growth Regul. 23
(2004) 76–82.
[72] H-C. Spatz, F. Brüchert, J. Pfisterer, Multiple resonance damping or how do trees
escape dangerously large oscillations? Am. J. Bot. 94 (2007) 1603–1611.
[73] K. Jungnikl, J. Goebbels, I. Burgert, P. Fratzl, The role of material properties for
the mechanical adaptation at branch junctions, Trees 23 (2009) 605–610.
[74] R. Hartig, Holzuntersuchungen Altes und Neues, Julius Springer, Berlin, 1901.
[75] J.W. Hart, Plant Tropisms and Other Growth Movements, Unwin Hyman, Lon-
don, 1990.
[76] S. Du, F. Yamamoto, An overview of the biology of reaction wood formation, J.
Int. Plant Biol. 49 (2007) 131–143.
[77] J.J. Holland, D. Roberts, E. Liscum, Understanding phototropism: from Darwin
to today, J. Exp. Bot. 60 (2009) 1969–1978.
[78] M.L. Molas, J.Z. Kiss, Phototropism and gravitropism in Plants, Adv. Bot. Res. 49
(2009) 1–34.
[79] L. Reinhold, Induction of coiling in tendrils by auxin and carbon dioxide, Science
158 (1967) 791–793.
[80] F.W. Telewski, Intra-annual spiral compression wood: a record of low-
frequency gravitropic circumnutational movement in trees, IAWA Bull. (N.S.)
9 (1988) 269–274.
[81] L. Köhler, T. Speck, H.-C.H. Spatz, Micromechanics and anatomical changes dur-
ing early ontogeny of two lianescent Aristolochia species, Planta 210 (2000)
691–700.
[82] K.A. Kern, F.W. Ewers, F.W. Telewski, L. Koehler, Mechanical perturbation
affects conductivity, mechanical properties and aboveground biomass of
hybrid poplars, Tree Physiol. 25 (2005) 1243–1251.
[83] I. Burgert, J. Keckes, P. Fratzl, Mechanics of the wood cell wall, in: D.D.
Stokke, L.H. Broom (Eds.), Characterization of the Cellulosic Cell Wall, Blackwell,
Oxford, UK, 2006, pp. 30–37.