Planta (2001) 213: 192198

DOI 10.1007/s004250000503

O R I GI N A L A R T IC L E

Julia L. Scher N. Michele Holbrook Wendy K. Silk

Temporal and spatial patterns of twining force and lignication

in stems of Ipomoea purpurea

Received: 14 September 2000 / Accepted: 27 October 2000 / Published online: 8 February 2001
Springer-Verlag 2001

Abstract Using the TWIFOR, an electronic device for

continuous, in vivo measurement of the forces exerted
by twining vines, we examined the forces generated by
vines growing on cylindrical poles of slender (6.35 mm)
and thicker (19.05 mm) diameter. In stems of Ipomoea
purpurea (L.) Roth, magnitudes of twining force (axial
tensions) were, on average, less at a particular time and
location on the more slender poles; while twining loads
(normal force per unit length of vine) were much greater
on the slender poles because of the greater curvature of
the vines. Thus, the geometry of the helix formed by the
vine on the pole aects the ability of the vine to maintain
a frictional interaction with its support. In addition, the
plant-to-plant variation in twining force was twice as
great on the thicker support poles. Metaxylem and bers
developed closer to the plant apex in vines on the slender
poles. On the thicker poles, a signicant fraction of the
maximum twining force developed during the establishment of the rst gyre, before bers were lignied,
indicating that primary growth can be sucient to establish high twining forces. On the slender poles, however, twining force increased with developmental stage
until the gyre was at least 1.5 m from the apex. Thus,
twining force can increase after cessation of primary
growth. No simple relationship was found between the
site of ber dierentiation and twining force.
Keywords Biomechanics Fiber development
Ipomoea (lignication) Lignication Twining Vine

J. L. Scher W. K. Silk (&)

Department of Land, Air, and Water Resources,
University of California,
Davis, CA 95616-8627, USA
E-mail: wksilk@ucdavis.edu
Fax: +1-530-7521552
N. M. Holbrook W. K. Silk
Department of Organismic and Evolutionary Biology,
Harvard University,
Cambridge, MA 02138, USA

Introduction
Climbing vines are a prominent feature of tropical forests (Putz 1984; Putz and Chai 1987; Hegarty and
Caballe 1991). In light-limited environments climbing
habits, including stem twining, tendril twining, and
clambering with adhesive appendages, are advantageous
because they enable vines to ascend vertical supports
with a minimum expenditure of biomass (Darwin 1867;
Niklas 1992). The stem of a twining vine grows around a
cylindrical support in a corkscrew shape that is geometrically regular, so that the stem forms a helical tube
of tissue with uniform curvature and torsion (Silk 1989).
The mechanics of the twining habit involve an innate
tendency for the vine to form a coil of smaller radius and
larger torsion than is possible on the supporting pole.
Thus, the stem hugs the pole and puts itself into tension.
In terms of a force balance in the natural coordinate
system dened by the Frenet vectors, the normal distributed load (a force per unit length here referred to as a
``load'') of the vine squeezing the pole is balanced by an
axial tension (here referred to as the ``twining force'')
within the vine's stem (Silk and Hubbard 1991 using
ideas from Love 1944 and Costello 1978).
In this paper we follow the development of the
twining force in vines growing around support poles of
two diameters. It is well known that the helical geometry
of twining vines varies with the diameter of the support
pole (Bell 1958; Putz and Holbrook 1991; Silk and
Hubbard 1991). Whether the thickness of the vertical
support aects the ability of twining vines to grasp the
pole, however, has not been measured. The growth
processes responsible for development of twining forces
are also not well understood. To examine what factors
might contribute to this, we compared the development
of the twining force to spatial patterns of lignication of
protoxylem, metaxylem and bers. Measurements of
stem elongation rate allowed us to transform our temporal force recordings into a developmental framework
(i.e., twining force as a function of distance from the

193

shoot apex). This allowed us to determine whether the

twining habit is established by growth rate patterns early
in the development of the stem segment and whether
ber development can lead to longer-term increases in
the twining forces.

Materials and methods

Cultivation of vines
Morning glory [Ipomoea purpurea (L.) Roth cv Heavenly Blue]
vines were grown from seed in a growth chamber with 14 h light
and 10 h darkness as described previously (Silk and Hubbard
1991). After eight leaves had emerged, stems were allowed to twine
around the TWIFOR apparatus (for force measurements) or a
wooden dowel (for anatomical studies). Two pole diameters were
used: slender poles that were 6.35 mm in diameter and thicker poles
that were 19.05 mm in diameter. The vines in this study had a
growth rate of 710 mm h1. No signicant dierence in growth
rates was found between plants growing on poles of dierent
diameters.
Measurements of the forces produced by twining stems
A novel force-measuring apparatus, the TWIFOR, was used to
record the twining load exerted as the vine grows up the pole
(Matista and Silk 1997). The apparatus has a vertical support pole
that is split longitudinally to make two poles with semicircular
cross-sections. One half-pole is mounted rigidly in a xed platform,
and the other is anchored in a swinging bar suspended at each end
from the xed platform by thin plastic strips and a thin beam load
cell (Omega Engineering). The half-poles are mounted in close
proximity so that the twining vine acts to pull the halves together,
with the forces exerted by the plant on the pole being measured by
the deformation of the load cell due to the horizontal movement of
the bar. The split-pole section was 16.0 cm long for the TWIFORS
with the slender poles and 25.5 cm long for the ones with the
thicker poles. This allowed one to two complete gyres to form on
the measurement section. A wooden dowel of the same diameter
was xed above the split-pole region to enable the vine to continue
growing vertically above the TWIFOR.
Each TWIFOR was calibrated using a 1-mm-diameter nylon
string that was wrapped around the split-pole so that the helix
formed by the string had a similar wavelength to that produced by
the vines. Thus the slender poles were calibrated with smaller helical wavelengths than the thicker poles. The string was draped over
pulleys mounted to maintain a helix of desired wavelength on the
split pole, and voltage output recorded as a function of force applied to the ends of the string. The TWIFOR output (mV) is linearly proportional to the tension in the calibrating string (Matista
and Silk 1997). Each TWIFOR was calibrated at regular intervals
in the laboratory. In addition, calibration measurements were made
immediately after each measurement run, i.e., while the TWIFORS
were still in-place within the growth chamber, allowing the zerooset to be recorded in situ.
At the onset of the experiment, each shoot was gently tethered
to the TWIFOR with string. This prevented the vine from moving
away and beginning to climb on an adjacent TWIFOR. After the
vine had begun to climb on the correct TWIFOR, the string was
removed. Approximately 812 days were needed for the vine to
reach the top of the wooden dowel. Throughout this period, force
measurements were made at 60-s intervals and the average value
over a 15-min period was recorded by a datalogger (Campbell
Scientic). Curvature on the TWIFOR, needed to calculate the
twining load from the measured twining force, was calculated from
the formula a/d2, where a is the coil radius (presumed to be the sum
of the vine and the pole radii) and d is the gyre arc length per radian
of wrap. The gap in the split pole made the TWIFOR diameter

slightly larger than the corresponding dowel diameter; curvature

calculations used the true diameters of the split pole.
The distance to the shoot apex from the middle of the rst gyre
for plants growing on the thicker poles or the top of the rst gyre
for plants growing on the slender poles was measured with a exible tape and recorded once a day or once every other day. Measurements of stem elongation allowed us to relate the twining force
generated by the plant at any moment to the developmental age
(i.e., distance from the shoot apex) of the portion of the stem on the
force-sensing region of the TWIFOR. To compare developmental
age vs. force among the dierent plants and between plants
growing on dierent diameter poles, developmental age was divided into discrete (15 cm) intervals for all plants and the force
data averaged for each plant within those intervals. These spatial
patterns were then averaged across plants for comparison with the
tabulated average values of anatomical measurements.
Anatomy and morphology
Plants for anatomical measurements were allowed to twine around
122-cm-long wooden dowels mounted on ring stands. When a stem
reached the top of the dowel, it was cut at 10, 15, 20, 25, 30 and
50 cm from the apex. Each stem segment was immediately sectioned on a Vibratome Series 1000 Sectioning System so that the
resulting 80-lm sections corresponded to a specic distance (as
above) from the shoot apex. These sections were mounted on microscope slides, with coverslips secured with nail polish to preserve
them for photomicrography and diameter measurements. A total of
10 stems was sectioned, of which ve twined around slender
(6.35 mm diameter) and ve around thicker (19.05 mm diameter)
dowels.
Cross-sections were observed and photographed with Nikon
Labophot and Olympus Vanox-T photomicroscopes using Kodak
Elite 160T color slide lm for bright-eld microscopy and Kodak
400ASA slide color lm for uorescence microscopy with a UV
lter for excitation of lignin and lignin precursors. The uorescent
photomicrographs of sections taken with 10 and 20 objectives
were observed with a projector and rated for presence and degree of
development of uorescing metaxylem, protoxylem and bers, using a subjective scale of )1 to 4. Our number )1 signies no lignication, 0 indicates that only protoxylem is lignied, 1 indicates
faint lignication at corners of metaxylem elements, 2 indicates
thicker lignication at corners, 3 indicates lignication around the
metaxylem element, and 4 indicates a thick, bright metaxylem wall.
A scale from 0 to 4 was used to rank ber development. Where no
bers are visible, a rank of 0 is assigned; 1 denotes barely visible
ber cells, 2 denotes a faint ring of ber cells 3 denotes a clear ring
of ber cells, and 4 denotes a bright ring of ber cells. Stem diameters were determined at each distance from the apex, by measuring actual sections (on microscope slides) with an ocular
micrometer on an Olympus BH microscope. Each diameter is the
average of three measurements per section.

Results
Time course of development of twining force
Vines growing on the slender (6.35 mm diameter) poles
produced twining forces that tended to increase rapidly
for 24 or 36 h after establishment on the TWIFOR,
followed by a slower rate of increase in twining force
throughout the measurement period (Fig. 1). For comparative purposes we present the twining force as the
equivalent axial tension of a string wrapped to the same
helical geometry. The force records of individual plants
showed short-term oscillations (on a time scale of hours)
superimposed on an increasing trend for the rst day,

194

followed by longer-term, daily, oscillations during the

subsequent week of growth. The entrainment of the
daily oscillations in many of the individual force records
may be caused by changes in plant water status. The
dark periods generally had smoother force trends than
the light periods. On the slender poles most vines increased in twining force for at least 100 h of growth.
A wide variation in force and in the time course of
force generation was found in plants growing on the
thicker (19.05 mm diameter) poles (Fig. 2). The time
courses for the three vines shown in Fig. 2 were collected
over the same time period, but exhibit signicant differences in their behavior. One vine developed nearly all
of its twining force during the establishment of the rst
gyre (and thus prior to onset of the force measurements);
during the succeeding 6 days, twining force increased by
only approx. 12%. A second vine showed a similar
pattern of generating most of the twining force during

the establishment of the rst gyre, but the actual forces

were nearly three times smaller. The third vine behaved
more like the ones on the slender poles, with the twining
force increasing steadily over a period of several days.
Averaging the force data for the 10 plants grown on
each pole diameter shows that vines on thicker poles had
somewhat larger twining forces, especially when the
helix was rst established on the pole (Fig. 3A). However, these dierences were small relative to the substantial plant-to-plant variation. For example, at 40 h
after the initial vine attachment to the support, the
average twining force (SD) in the vines on the slender
poles was 16746 mN, while on the thicker poles the
force was 18590 mN.
Despite the small dierences in average twining force,
dierences in their helical geometry produced large differences in the ability of the vines to grasp the two different pole diameters (Fig. 3B). The geometry of the
growth habit aects the twining load, here dened as the
normal force per unit stem length, because greater curvatures give greater twining load for the same axial
tension. The twining load is important in assessing the
stability of a twining vine because it results in a frictional
interaction between vine and pole that opposes gravity.
The helical geometry diers predictably among pole

Fig. 1 Time course of twining force, here expressed as the

equivalent axial tension in a string (mN), of three individual
morning glory (Ipomoea purpurea) vines growing on slender
(6.35 mm diameter) poles. Shaded bars indicate night periods.
Time zero indicates when the supporting string was removed, 24 h
after the vine was rst placed in contact with the pole

Fig. 2 Time course of twining force, here expressed as the

equivalent axial tension in a string (mN), of three individual
morning glory vines growing on thicker (19.05 mm diameter) poles

Fig. 3A, B Time course of average twining forces and loads. A

Average twining force, expressed as equivalent string tension, for
morning glory 10 vines growing on slender and thicker poles. Bars
indicate one SD for the forces at 40 h. B Twining load, dened as
the normal force per unit length, for vines growing on slender and
thicker poles. The twining loads, like the forces, are averages for
10 vines on each diameter. At most times, vines have smaller
twining force but considerably greater twining load on the slender
poles

195

force (Fig. 4) is similar to the temporal patterns. For

early developmental stages vines growing on slender
poles had less twining force than those twining on the
thicker poles. However, twining force on the slender
poles increased with distance from the apex such that
when the region of the stem on the TWIFOR was more
than 130 cm from the apex, twining force was greater on
the slender pole.
Morphology

Fig. 4 Twining force as a function of developmental age of the

region of the stem on the TWIFOR for morning glory vines
growing on slender and thicker support poles. The force data of
Fig. 3 were correlated to distance from the apex (recorded
separately) to establish the developmental trends. Arrows show
the location of development of metaxylem and bers. The letter M
indicates metaxylem had reached stage 2 (signicant lignication),
and the letter F indicates bers had reached stage 2 (signicant
lignication)

diameters: in our experiments the average (SD) gyre

wavelength was 71.81.47 mm on the slender poles and
144.02.53 mm on the thicker poles. Because vines have
greater curvature on the slender poles, their inward
(normal) force per unit length of stem is correspondingly
greater than occurs with vines growing on thicker poles
(Fig. 3B). This was true even early in development, despite the signicantly smaller average twining force of
vines on slender poles (Fig. 3B).
Spatial pattern of the twining force
As the vine elongates on the dowel above the TWIFOR,
the portion of the stem on the TWIFOR becomes progressively more mature. Thus, distance from the plant
apex provides a developmental index for the portion of
the stem in contact with the TWIFOR. The relation
between twining force and distance from the shoot apex
also provides a basis for examining the anatomical basis
for the twining force. Because elongation rates were
approximately equal between vines growing on the different diameter poles, the spatial pattern of the twining

Consistent with published growth analyses (Silk and

Abou Haidar 1986), vine diameter increased between 10
and 20 cm (including the primary growth zone) but did
not change signicantly between 20 and 50 cm (Table 1).
Thus radial growth did not contribute to the increase in
twining force as the region of the stem on the TWIFOR
was displaced from 20 cm to 50 cm from the apex.
Anatomy
The spatial patterns of development of the lignin-containing elements in vines growing on the slender and
thicker poles are shown in Fig. 5 and Table 2. Only
protoxylem was lignied at 10 cm from the apex. At
15 cm, metaxylem lignication was faintly evident on
thicker poles and strongly evident on the slender poles.
Lignication of the entire metaxylem wall was evident
on both pole diameters at 25 cm and progressed to the
same visual brightness on both pole diameters at 50 cm
from the apex.
Fiber development was evident at 15 cm from the apex
in vines on the slender poles, and the continuous ber ring
was present 20 cm from the apex. Lignication of bers
was evident only beyond 25 cm. On the thicker poles the
ber development was delayed so that the ring of bers
was not evident until 30 cm, and little ber lignication
had occurred even at 50 cm from the shoot tip.

Discussion
The climbing habit provides plants with an economical
growth habit due to the ability to forego the biomass
costs associated with being self-supporting (Niklas 1992,
1994; Rowe and Speck 1996). Twining vines are able to
ascend smooth vertical poles and make use of the me-

Table 1 Stem diameter (mm) as a function of distance from the shoot apex of morning glory (Ipomoea purpurea) vines growing on slender
(6.35 mm diameter) and thicker (19.05 mm diameter) support poles. Each value is the average SD of ve vines per pole diameter
Poles

Slender
Thicker

Distance from shoot apex (cm)

10

15

20

25

30

50

1.210.18
1.160.11

1.270.27
1.190.13

1.320.22
1.280.02

1.300.25
1.310.07

1.220.21
1.240.10

1.280.21
1.270.04

196

197
b

chanical strength of these vertical supports via their

ability to grasp on and create a frictional force. Analysis
of the mechanics of the twining habit suggests that an
innate tendency to form a coil of smaller radius and
larger torsion than is possible on the supporting pole
provides the basis for this interaction (Silk and Hubbard
1991). The internal forces within the stems of twining
vines that allow the vine to grip the pole are dominated
by an axial component (tension). We provide here the
rst in situ measurements of these contact forces in developing vines. We demonstrate that the twining forces,
here dened as the equivalent axial tension in an ideal
vine having no mechanical rigidity, develop early during
development, especially in plants growing on thicker
poles. We also found greater plant-to-plant variation in
total twining force with the larger diameter poles. The
reason for this between-plant variation is not clear, but
may be due to individual dierences in stem mechanical
properties.
The economy of form achieved by twining vines is
obtained at the cost of having to locate a series of suitable supports on which to grow (Putz and Holbrook
1991). The ability to utilize a wide range of support sizes
greatly enhances the probability that a vine will be able
to ascend to the forest canopy (Putz 1984; Putz and
Holbrook 1991). The twining forces measured in this
study indicate that the magnitudes of the tensions within
the stems are similar for vines growing on dierent-diameter poles (Fig. 3A). However, the contact forces
between vine and pole that result from these axial forces
dier considerably as a result of the dierences in curvature on the two pole diameters. Thus, vines growing
on more-slender poles have greater normal force per unit
stem length than do plants on thicker poles (Fig. 3B).
These data help explain why vines growing on larger
poles can be unstable and sometimes even slip down the

pole, while conspecics growing on more-slender supports are very stable.

One of the principal objectives of this study was to
test the hypotheses that the twining habit is established
by growth gradients during primary growth, and that
twining forces (both axial tensions and twining loads)
increase in magnitude due to ber development. These
hypotheses were inspired by the observation that the
twining stem puts itself into tension as it hugs the pole
and the theory that the tension would be augmented by
the shortening that occurs during the development of
bers. We were also inuenced by the literature on the
formation of reaction wood. Okuyama et al. (1994) have
shown that growth stresses and strains associated with
the formation of reaction wood have a strong spatial
correlation to alpha cellulose content of developing bers (and a weak inverse correlation to lignin content).
Furthermore, Koehler et al. (2000) have recently shown
that lignication occurs in the parts of vines that would
be expected to have strong twining forces.
Our rst hypothesis, that primary growth is required
for establishment of the twining habit, is supported by
the data of this study. We observed a substantial fraction of the total twining force was generated during the
formation of the apical one to two gyres. This was
particularly true on the thicker poles. The growth zone
of Ipomoea purpurea is up to 13.5-cm-long in some varieties (Silk and Abou Haidar 1986), indicating that
twining forces can be generated by regions undergoing
elongation.
A simple relationship between ber dierentiation
and magnitude of the twining force, however, did not
emerge from our data. For the plants growing on the
thicker poles, ber dierentiation occurred after the
high twining force is established. On the thinner poles
the initial twining force is low, yet this low stem tension
is associated spatially with ber dierentiation (Table 2,
Fig. 4). A case can still be made that the presence of
dierentiating bers is necessary for continuing increase of the twining force after primary growth has
ceased. However, it may also be that a general tissue
contraction or a decrease in stem elasticity occurs well
below the zone of primary growth to increase the
twining force. Further studies are needed to understand
the generation of the axial tensions produced by the
twining habit.

Table 2 Average degree of lignication of xylem (X) elements

(meta- and protoxylem) and bers (F) as a function of distance
from the shoot apex of morning glory vines growing on slender
(6.35 mm diameter) and thicker (19.05 mm diameter) support

poles. Stem cross-sections were scored on a scale from )1 to 4

(xylem) or 0 to 4 (bers), where the lowest value indicates no lignication. For further details see Materials and methods. Each
value is the average score of ve vines per pole diameter

Fig. 5AL Cross-sections of morning glory stems showing uorescence of lignin and lignin precursors in protoxylem (p),
metaxylem (m) and bers (f). Sections are 10 cm (A, G), 15 cm
(B, H), 20 cm (C, I), 25 cm (D, J), 30 cm (E, K), and 50 cm (F, L)
from the apex on a 6.35-mm-diameter pole (AF) and a 19.05-mmdiameter pole (GL). Not all images within one pole diameter
represent the same plant. Ratings for protoxylem and metaxylem
(meta) and ber (bers) development are indicated, with 0
indicating no development and 4 mature with lignication

Pole

Distance from shoot apex

10 cm

Slender
Thicker

15 cm

20 cm

25 cm

30 cm

50 cm

0
0.1

0
0

2.2
1.1

0.6
0

2.5
2.3

2.0
0.2

3.1
3.2

2.1
1.0

3.8
3.6

3.3
1.6

4.0
4.0

3.8
2.8

198
Acknowledgements This research was supported by Grant IBN
9604230 from the National Science Foundation to W.K.S. A Bullard Fellowship from the Harvard Forest funded the visit of
W.K.S. to the laboratory of N.M.H. S. Wiederkehr provided
skillful technical assistance.

References
Bell PR (1958) Twining of the hop. Nature 181:10091010
Costello GA (1978) Analytical investigation of wire rope. Appl
Mech Rev 31:897899
Darwin C (1867) On the movements and habits of climbing plants.
J Linn Soc Lond Bot 9:1118
Hegarty H, Caballe G (1991) Distribution and abundance of vines
in forest communities. In: Putz FE, Mooney HA (eds) The biology of climbing plants. Cambridge University Press, Cambridge, pp 313335
Koehler L, Speck T, Spatz, H.-C (2000) Micromechanics and anatomical changes during early ontogeny of two lianescent
Aristolochia species. Planta 210:691700
Love AEH (1944) A treatise on the mathematical theory of elasticity. Dover, New York
Matista A, Silk W (1997) An electronic device for continuous, in
situ measurement of the forces exerted by twining vines. Am J
Bot 84:11641168

Niklas KJ (1992) Plant biomechanics: an engineering approach to

plant form and function. University of Chicago Press, Chicago
Niklas KJ (1994) Plant allometry: the scaling of form and function.
University of Chicago Press, Chicago
Okuyama T, Yamamoto H, Yoshida M, Hattori Y, Archer RR
(1994) Growth stresses in tension wood: role of microbrils and
lignication. Ann Sci For 51:291300
Putz FE (1984) The natural history of lianas on Barro Colorado
Island, Panama. Ecology 65:17131724
Putz FE, Chai P (1987) Ecological studies of lianas in Lambir
National Park, Sarawak, Malaysia. J Ecol 75:523531
Putz FE, Holbrook NM (1991) Biomechanical studies of vines. In:
Putz FE, Mooney HA (eds) The biology of climbing plants,
Cambridge University Press, Cambridge, pp 5378
Rowe NP, Speck T (1996) Biomechanical characteristics of the
ontogeny and growth habit of the tropical liana Condylocarpon
guianense (Apocynaceae). Int J Plant Sci 157:406417
Silk WK (1989) Growth rate patterns which maintain a helical
tissue tube. J Theor Biol 138:311327
Silk WK, Abou Haidar S (1986) Growth of the stem of Pharbitis
nil: Analysis of longitudinal and radial components. Physiol
Veg 24:109116
Silk WK, Hubbard M (1991). Axial forces and normal distributed loads in twining stems of morning glory. J Biomech 24:
599606