| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Anatomy and Morphology |
2Botanique et Bioinformatique de l'Architecture des Plantes, UMR 5120 CNRS, TA40/PS2, Boulevard de la Lironde, F-34398 Montpellier cedex 5, France; 3Plant Biomechanics Group, Institut für Biologie 2 and Botanischer Garten, Universität Freiburg, Schänzlestr. 1, D-79104 Freiburg, Germany
Received for publication December 11, 2004. Accepted for publication June 3, 2005.
ABSTRACT
Mechanical properties are investigated in Desmoncus orthacanthos and D. polyacanthos from French Guiana, South America. Differences in size and axis stiffness are related to different trellis requirements and habitats. The leaf sheath surrounds the stem, increasing stiffness of young self-supporting stages and apical parts of older climbing plants. Senescence of the leaf sheath reduces stiffness of older climbing axes of both species. Its eventual loss in D. orthacanthos facilitates deformation into coils and loops when plants slip from their supports following senescence of leaves bearing attachment organs. In smaller climbing axes of D. polyacanthos, the senescent leaf sheath remains attached and axes rarely form loops and coils below attachment. An increase in stiff mechanical properties toward the base of both species is radically different from that of many dicotyledonous lianas. Besides the presence of attachment organs, stem mechanical properties of Desmoncus are similar to those of erect though not fully self-supporting stems of Bactris major, a sympatric species of the sister group genus to Desmoncus. The climbing habit in Desmoncus may have evolved via (1) heterochronic processes including early elongation of internodes relative to increase in stem diameter (reduction of the establishment phase) and (2) increased persistence of leaf sheaths.
Key Words: Bactris major • climbing palms • Desmoncus orthacanthos • Desmoncus polyacanthos • growth form • leaf sheath • lianas • mechanical architecture
Climbers are relatively common among monocotyledons (Tomlinson and Fisher, 2000
). They vary greatly in size from climbing herbs such as Scleria secans (Cyperaceae) to liana-sized rattans (Calamoideae) of southeast Asia, Australia, and West Africa, and the morphologically similar Desmoncus (Arecoideae) of the New World tropics. Climbing palms lack secondary growth and rely on primary tissues for their entire life history. Extreme mechanical stresses applied to these stems might be potentially fatal because they apparently lack the high compliance, toughness, and potential for self-repair attributed to vines and lianas having secondary vascular cambial growth. Furthermore, most climbing palms do not develop aboveground branching, so recovery from mechanical damage via aerial branching is also limited. Despite these apparent constraints, the abundance and diversity of climbing palms suggests a considerable evolutionary success of this type of growth form.
This investigation follows recent studies into the biomechanics of plant growth forms, particularly climbers. These have demonstrated how changes in cambial development can drastically modify mechanical properties of the stem during development (Rowe and Speck, 1996
, 1998
, 2004
; Speck et al., 1996
; Gallenmüller et al., 2001
; Isnard et al., 2003a
, b
). In lianas, modifications of mechanical properties between young self-supporting and older non-self-supporting stages of growth commonly show a significant drop in structural Young's modulus—a measure of the flexibility of the stem and its constituent tissues (Speck, 1994
; Rowe and Speck, 1998
; Speck and Rowe, 1999
). In dicotyledonous lianas, this is highly adapted to the mechanical constraints where growth relies on "mechanical parasitism" (Caballé, 1980
; Putz, 1984
; Fisher and Ewers, 1991
; Putz and Holbrook, 1991
; Rowe and Speck, 2004
).
Other groups besides dicotyledons have evolved climbing growth forms, though not much attention has been paid to these in terms of their biomechanics or the evolution of this specialized growth form (Rowe and Speck, 2004
; Rowe et al., 2004
). Climbers have also evolved among plants that lack secondary growth including ferns and monocotyledons. Scandent habits are known even among plants with an archaic stem organization such as Lycopodiella cernua (Lycopodiaceae). Recent biomechanical studies have shown that mechanical properties of climbers can nevertheless be modified via strategically modulating the stiffness of primary tissues along self-supporting, procumbent, or climbing stems (Rowe and Speck, 1997
, 1998
, 2004
; Rowe et al., 2004
).
Climbers have evolved in two subfamilies Arecoideae and Calamoideae (including true rattans) among the five now recognized in palms (Dransfield et al., in press
). They occur in 16 genera (7% of all palm genera) and 580 species (22% of all palm species) (Baker et al., 2000a
, b
). Most climbing palms have developed a range of adaptive novelties including modified leaves and inflorescences forming organs of attachment (Figs. 1C, 2B) and leaf sheaths, which remain attached to the stem along extensive segments of the plant (Fig. 1A). Like many climbing dicotyledons, they also have larger vessel diameters (c. 150–180 µm in Desmoncus and up to 350 µm in Calamus) than related nonclimbers (Tomlinson and Zimmermann, 2003
). In both Desmoncus and Calamus, the vascular architecture lacks direct connection between protoxylem and metaxylem and thus differs from that found in the tree-palm model Rhapis excelsa. This has been interpreted as a possible protection against cavitation and hydraulic dysfunction (Tomlinson and Zimmermann, 2003
). Tomlinson (1962)
emphasized the persistent, tube-like leaf sheaths in climbing palms compared with the shorter, more ephemeral, and partially encircling leaf bases of nonclimbers. In many climbing palms, leaf sheaths probably contribute to the mechanics of the climbing axis for at least part of the growth trajectory, as well as form a mechanical connection with host vegetation via hooks on leaves and inflorescences. A potential constraining feature in such plants is the ultimately ephemeral nature of the leaves and leaf sheath and the effects of their senescence and breakage on the mechanical architecture, growth form, and life history in general.
|
|
; Siebert, 2000
). Interest has focused recently on some species to better understand ecological requirements and impact of harvesting (Troy et al., 1997
; Siebert, 2000
; Escalante et al., 2004
). In this study, we investigated the mechanical architecture of Desmoncus orthacanthos and D. polyacanthos. Each species occupies a different habitat in French Guiana and differs in overall size, leaf size, type and size of attachment organs, height, ramet diameter, and type of trellis. We also compared their axial mechanical properties with those of Bactris major, an upright, though ultimately unstable, species of the sister group genus Bactris. We also compared the mechanical design of these climbing palms with dicotyledonous lianas. Because radically increased flexibility during ontogeny is such a prevalent feature among climbing dicotyledons, we wished to find out how plants lacking secondary growth can maintain similar growth forms in both large- and small-bodied species of the genus.
MATERIALS AND METHODS
Taxonomy and distribution of Desmoncus
The genus is placed in the tribe Bactridinae, subfamily Arecoideae (Uhl and Dransfield, 1987
; Hahn, 2002
). It is distributed widely in Central and South America from Mexico to Brazil and Bolivia and in the Caribbean except Trinidad and Tobago (Henderson et al., 1995
). The systematics and taxonomy of the genus is in need of revision with estimates of the number of species ranging from six (Henderson et al., 1995
) to 61 (Uhl and Dransfield, 1987
). The genus is distinguished from true rattans of southeast Asia by its reflexed, thickened leaflets (acanthophylls) on the modified frond apex (cirrus) acting as hook-like grapnels. Strangely, the geographically intermediate, calamoid rattans of West Africa (Eremospatha, Laccosperma, and Oncocalamus) also possess these structures (Corner, 1966
; Dransfield, 1978
). This study is based on material collected in the field in French Guiana. Desmoncus polyacanthos Mart. was collected from an understory forest habitat at the research station known as "Piste de St. Elie" (5°17' N, 53°03' W); specimens of D. orthacanthos Mart. and B. major Jacq. were collected from disturbed open and wooded sites near mangroves along the "route de l'Anse" near the village of Sinnamary (5°39' N, 52°95' W).
Habitat, growth form, and morphology
The species differ in habitat and morphology. Desmoncus orthacanthos can be very common alongside rivers and is considered a generalist species with a preference for high light and vigorous vegetative propagation (Siebert, 2000
). Desmoncus polyacanthos is sparsely distributed in forest understory and has a preference for patchy light environments with diffuse shade (Troy et al., 1997
). It tends to grow alongside paths, beside tree-fall gaps, along creeks, and in temporarily flooded areas. The species develop and climb in basically similar ways. Individual ramets are produced from underground rhizomes lying just below the soil surface. Young axes are composed of an inner stem and an outer leaf sheath bearing leaves. In both species, shoots grow upwards, become mechanically unstable, and then climb into the surrounding vegetation via effectively deployed hooks borne on leaves. In this account, the plant axis is defined as the stem and leaf sheath combined. The true stem is defined as the axial parts of the plant that have naturally shed the leaf sheath, or parts from which the leaf sheath has been removed prior to mechanical tests.
Desmoncus orthacanthos is a clustering species producing clumps of tens of upright and climbing axes from extensive rhizomes. Young axes remain self-supporting to a height of about 2.5 m. Cirri are produced only after the plant becomes unstable and these specialized grapnel-bearing leaf tips increase in length toward the apex of the climbing axis (Fig. 1A, C). Climbing axes with intact leaf sheaths are 0.6–3 cm wide. In open areas where supports are rare, axes become procumbent but can eventually produce ascending apices, which can develop into climbing axes if they grow near an appropriate host (Fig. 1D). The species can reach canopy heights of nearly 20 m and total axis lengths of about 30 m. Mature climbing axes remain attached to host trees by up to 20 leaves below the apex.
Desmoncus polyacanthos is much smaller and delicate, producing fewer (1–4) axes per rhizome at any one time. Young axes remain upright to about 1.5 m in height (Fig. 2A) and as in D. orthacanthos, produce attachment organs only when unstable. Climbing axes are slender (0.4–1 cm in diameter) and differ from D. orthacanthos by having recurved hooks on both leaf rachis and cirrus (Fig. 2B). When unstable, axes climb on the surrounding vegetation via extremely "catchy" cirri and often show "S-shaped" growth resulting from partial falls of the growing apex and later ascending growth (Fig. 2C). In the area examined, stems do not exceed 12 m in total length and rarely climb higher than 7–8 m.
Canopy openings were estimated at 20 stations of D. polyacanthos along 2.7 km of forest trail and adjacent areas. Photographs with a digital camera and fish-eye lens were taken in diffuse light conditions just before and after sunrise. The lens was pointed vertically upwards at three locations (north, west, and east) at 1 m around the rhizome and at 1 m from the ground. Percentages of canopy openings were then calculated using standard image analysis software. The measurements provided an approximate measure of the canopy gap above rhizomes bearing upright axes.
Bactris major is sympatric with D. orthacanthos and typically occurs beside mangroves forming dense monospecific stands. Axes are produced from large rhizomes and young axes are self-supporting, while older, longer stems lean against each other (Fig. 3A). Axes remain relatively short (1–10 m; Henderson et al., 1995
) and mature stems, though not fully self-supporting, do not develop specialized attachment devices. Leaf sheaths senesce rapidly along self-supporting plants and are thus only intact at the apex (Fig. 3B). A fibrous sheath surrounds the stem after the decay of the leaf lamina (Fig. 4).
|
|
The total number of axes studied included: D. orthacanthos (entire axes, N = 15; leaf sheath removed, N = 15), D. polyacanthos (entire axes, N = 11; leaf sheath removed, N = 6), B. major (entire axes, N = 4; leaf sheath removed, N = 2). Older axes of D. orthacanthos and B. major were selected that had naturally lost the leaf sheath and were measured in bending. Apical segments of all species were comprised of highly flexible, immature stems and immature leaf sheath and could not be reliably tested in bending. Axes were cut into segments suitable for three-point bending tests so that the span-to-depth ratio of tested segments was always greater than 20 (mean = 43 ± 11 for D. orthacanthos, 38 ± 9 for D. polyacanthos, and 41 ± 7 for B. major). This minimum ratio was empirically established after conducting bending tests at a range of span distances; a value of 20 was found to be sufficient to minimize the influence of shear during bending to a negligible degree.
The principal aim of the bending tests was to measure the flexural stiffness, EI of the entire axis and that of the stem without the leaf sheath. Bending test protocols followed those of recent analyses on woody plants (Vincent, 1990
; Rowe and Speck, 1996
; Gallenmüller et al., 2001
; Rowe et al., in press
). Axes were placed on a steel frame apparatus at two supporting points at a precise span distance. Consecutive weights were added to a pannier suspended from the exact center of the stem segment. Deflection of the axis was measured via a dissecting microscope equipped with an eyepiece graticule and the resulting force-deflection curve used to calculate flexural stiffness:
 | (1) |
All three species have a conspicuous node/internode structure and wherever possible, the same positions of nodes and internodes were set up during each bending experiment. Axial diameters along each segment were used to calculate the mean axial second moment of area (I) of each tested segment via the formula for an ellipse, which approximates closely the shape of the axes in cross-section:
 | (2) |
After measuring flexural stiffness (EI) and axial second moment of area (I) of each axis segment, structural Young's modulus (Estr) was calculated:
 | (3) |
RESULTS
Mechanical architecture of entire axis (stem and leaf sheath)
In young self-supporting and unstable plants of both species, leaf sheaths were intact, green, and tightly surround the stem along most of the length (Fig. 5A, B). Senescence occurs at the base of young plants of D. polyacanthos and in both species occurs progressively towards the apex in older climbing axes. In D. polyacanthos, the leaf sheath remains on the stem after prolonged senescence (Fig. 5B, right). In D. orthacanthos, the leaf sheath is progressively loosened below the apex and is finally lost (Fig. 5A, right) in cane-like older stems (Fig. 1B). Among younger procumbent axes of D. orthacanthos that have fallen from their support, leaf sheaths are often already fractured, dried, and loosened toward at the base.
|
). Initial diameter and I and EI can vary widely from one axis to another depending on the size of the rhizome and the timing of the production of axes. In D. orthacanthos, young self-supporting and unstable axes have higher values for both EI and I than older climbing axes in which the leaf sheath is still present (Fig. 6A, B) because of the presence of a thick, intact leaf sheath surrounding the stem during these stages. This trend is more marked in D. orthacanthos than in the relatively diminutive D. polyacanthos. Despite differences in EI, trends along axes generally follow shifts in size as represented by axial second moment of area (Fig. 6A–D).
|
A rather different pattern is observed in D. polyacanthos; older climbing axes consistently increase in EI toward the base whatever the age and orientation. Relatively lower and constant values extend toward the apex. Both trends in EI follow the shifts in values of I (Fig. 6C, D). Senescence of the leaf sheath base appears to affect the mechanical properties less than in D. orthacanthos.
Some self-supporting to climbing individuals of D. polyacanthos increased in height up to a maximum of 1.8 m over a period of 5 months, while other axes had little or no increase. Measurements of canopy opening above stations of D. polyacanthos ranged from 5.7–9% with an overall mean value of 7.3 ± 1.1%. The range of values observed did not correlate with or explain differences in axis elongation over the period studied.
Mechanical architecture of the stem
In both species, trends in EI along young self-supporting and young unstable stems following removal of the leaf sheath, do not differ from overall trends of axes measured with their leaf sheaths (Fig. 7A, C). Stems increase in EI toward the base, following trends in I (Fig. 7B, D). In D. orthacanthos, the increase in EI of intact axes toward the apex of long procumbent axes (up to 5.5–10 m; Fig. 6A) is not observed after removal of leaf sheaths (Fig. 7A). This highlights the importance of the leaf sheath in resisting bending forces toward the apex of the plant.
|
Mechanical roles of the stem and leaf sheath
Leaf sheath thickness differs between species (Fig. 8); in D. polyacanthos it rarely exceeds 1 mm but in D. orthacanthos it can reach up to 6 mm in apical parts of old plants. In both species, it varies little along young self-supporting to unstable axes (Fig. 8A) but decreases significantly towards the base in older plants (D. orthacanthos: R2 = 0.43, R = 0.656, N = 27, P < 0.01; D. polyacanthos: R2 = 0.5, R = 0.705, N = 45, P < 0.01; Fig. 8B). Features of the leaf sheath and stem were explored via bivariate plots where data above the isometric relationship x (contribution of stem) = y (contribution of leaf sheath) indicate greater contribution by the leaf sheath. The parameters investigated included I (Fig. 9A, B) and EI (Fig. 9C, D). Similar bivariate plots were used to compare the relative calculated values of structural Young's modulus (Estr) of leaf sheath and stem (Fig. 9E, F). Since leaf sheaths progressively senesce and are lost from bases of old axes of D. orthacanthos, we include only plants with entire leaf sheaths, tested before and after removal of the leaf sheath.
|
|
In both species, the contribution of the leaf sheath to EI is often lower in basal segments but cannot be distinguished between self-supporting, young unstable, and climbing axes (Fig. 9C, D). In D. orthacanthos, the contribution of the leaf sheath to EI is generally greater than 50% apart from basal segments of older climbing stages (Fig. 9C). In D. polyacanthos, the leaf sheath generally contributes to about 50% of total flexural stiffness. Basal segments of both young self-supporting and older climbing axes show contributions less than that of the stem and noticeably less than that of D. orthacanthos (Fig.9D).
In both species, values of Estr for the leaf sheath are generally lower than those of the inner stem with the vast majority of points falling beneath the isometric line (Fig. 9E, F). In D. orthacanthos, values range from just below 2000 to 7800 MN/ m2 (Fig. 9E), with one value close to zero representing the dried and senesced basal portion of an old axis. In D. polyacanthos, the upper limit is somewhat higher, 10–11 000 MN/ m2 (Fig. 9F). Outlying values (Fig. 9F) represent basal and senescent parts of young unstable and older climbing axes as well as median parts of young-self-supporting axes for which leaf sheaths have senesced and Estr is consequently low.
In young axes of D. orthacanthos, Estr values of the leaf sheath are similar in basal and median parts of the plant. In young plants of D. polyacanthos, however, Estr tends to be lower in basal segments. Overall values of Estr for the stems of both species range between c. 3000 and 17 000 MN/m2 for D. orthacanthos, with rather higher upper limits of between 1300 and 19 000 MN/m2 for D. polyacanthos. For both species, a few points are distributed above the isometric line and correspond to apical segments for which the immature and mainly parenchymatous stem has a lower Estr, thus increasing the mechanical role of the leaf sheath near the apex (Fig. 9E, F). In both species, structural Young's modulus of the stem increases significantly toward the base in both young self-supporting (D. orthacanthos: R2 = 0.65, N = 6; P < 0.05; D. polyacanthos: R2 = 0.87, N = 6, P < 0.01), to unstable stems (D. orthacanthos: R2 = 0.97, N = 14; P < 0.01; D. polyacanthos: R2 = 0.79, N = 19, P < 0.01; Fig. 10A). This increase, though still significant (D. orthacanthos: R2 = 0.21, N = 62; P < 0.01; D. polyacanthos: R2 = 0.36, N = 74, P < 0.01), is less marked in older climbing stems of both species (Fig. 10B). For both basal and middle segments of D. polyacanthos, mean values of Estr are noticeably higher than in D. orthacanthos (Fig. 11).
|
|
). Structural Young's moduli of young self-supporting stems without leaf sheaths varies from 5000 to 10 000 MN/m2 towards the base (Fig. 12B) and are similar to values in Desmoncus. Two older unstable stems, which have naturally lost their leaf sheaths, have relatively stable values for Estr (16 013 ± 1810 MN/m2) and are thus somewhat stiffer than the two species of Desmoncus.
|
Variation of mechanical architecture in Desmoncus
Mechanical roles of stem and leaf sheath differ in each species during ontogeny. In the larger D. orthacanthos, the leaf sheath contributes more to resisting bending forces than in the smaller D. polyacanthos. This is clear in younger self-supporting plants of D. polyacanthos, in which leaf sheaths contribute less and senescence occurs earlier. In this species, mechanical support is mainly provided by the stem. In both species, the leaf sheath has a lower calculated Estr than the stem. The fact that the leaf sheath is comprised of a more compliant material than the stem, yet nevertheless contributes significantly to bending resistance, highlights its geometrical role (high value of I) in contributing to axis stiffness. In both species, the leaf sheath contributes significantly to axis stiffness near the apex of older climbing plants.
In D. orthacanthos, loss of the leaf sheath confers an obconical geometry and relatively low values of EI toward the base of procumbent and climbing axes. In the oldest stage of development, canes are more compliant than the entire stem/ leaf sheath of younger parts and readily form loops on or near the ground. These appear to result from slippage of the plant from host supports when apically attached leaves either lose their connection or senesce and break away from the supports. The effect of senescence of the leaf sheath on the ability of the axis to resist bending forces is less in D. polyacanthos. While this smaller-bodied climbing species had been observed to fall from its supports, it did not form the wide loops and arcs linked to the possibly continual slipping of the larger species.
Body size, growth habit, and attachment
Differences in body size and mechanical architecture are characteristic of different habitats and trellis requirements. Large basal diameter and relatively high axis stiffness enable young plants of D. orthacanthos to remain upright for up to 2–3 m in clonal groups, even in relatively open environments. Narrow basal diameter and relatively low values of axis stiffness in D. polyacanthos are more consistent with strategies of rapidly locating a support in more closed conditions or "pausing" apical growth and entering a waiting phase as observed in many dicotyledonous lianas (Putz, 1984
; Caballé, 1986
; Lee and Richards, 1991
).
In D. orthacanthos, acanthophylls on the cirrus and robust spines of the leaf sheath and petioles assure attachment between ramets as well as other neighboring plants. Large leaves (mean length of 1 m with cirrus) provide large "search and attachment" ranges for relatively distant supports. Large size and wide angle of acanthophylls in D. orthacanthos can furthermore ensure attachment to host branches up to 1.5 cm in diameter. Initial pulling/fracture tests have shown that a single acanthophyll can withstand forces of up to 200–250 N before breaking. This suggests that a single grapnel can suspend a significant portion of otherwise unattached axis. In summary, large and stiffer axes are heavier but can span larger distances in self-supporting and searcher phases of growth. Such larger axes require a more robust attachment, which can interlock with sturdier and larger diameter host branches during the climbing phase.
In D. polyacanthos, small acanthophylls on the cirrus and recurved spines on both cirrus and leaf rachis only interlock with supports one to several millimeters in diameter. Climbing individuals were mainly attached to palm leaves or woody branches under 10 mm in diameter. Initial pulling/fracture tests indicate that the largest hooks and grapnels of D. polyacanthos can sustain loads of 3–9 and 20 N, respectively. The height of the self-supporting phase is more limited before becoming unstable, and the climbing strategy relies on attaching smaller lighter axes to relatively small supports across smaller distances.
In D. orthacanthos, senescence of primary tissues probably plays a decisive role in the climbing strategy and life history processes. Leaves and leaf sheath senesce from the base of the plant upwards so that both the attachment of the climber by any given leaf and the stiffness imparted by the leaf sheath are limited over time. To remain attached to supports during and after the loss of leaves beneath the apex, the plant must continue to grow and produce new leaves and new points of attachment. This explains why the overall length of plants (20–30 m) is greater than the overall height. This type of growth form is similar to that observed among rattans from southeast Asia (Corner, 1966
; Putz, 1990
). Observations suggest that senescence of leaves and leaf sheath and slipping or ratcheting of the mature climbing stem in the surrounding vegetation might have important implications on other aspects of the plants life history. The fact that the plant crown might be under almost constant tensile strain from the weight of the suspended cane below might limit its lateral growth to new supports of the canopy. Furthermore, the production of new leaves for assuring mechanical attachment will also be linked to photosynthetic requirements. Further studies should, if possible, investigate these dynamics around crowns of climbing palms such as D. orthacanthos.
The developmental constraints linked to climbing in D. polyacanthos are quite different. The species does not reach the heights and lengths observed in D. orthacanthos. Although older leaves below the apex of mature climbing plants senesce and fragment, this smaller species does not have the same tendency to slip or ratchet downwards from its point of attachment. Furthermore, older axes do not entirely lose the outer leaf sheath, and old stages of axis do not form extensive loops and coils of cane-like stems seen in the larger, heavier D. orthacanthos.
Leaf sheaths of D. orthacanthos may fracture as result of excessive bending stress or falling from host plants (Fig. 13). Local fractures of the leaf sheath strongly reduce EI thereby providing local flexibility. The compliancy of the surviving inner stem allows repositioning of axes but because the fractured leaf sheath is only broken locally, the rest of the axis remains intact and sufficiently rigid to retain an overall climbing position. This kind of mechanical response to traumatic mechanical stress has also been observed among calamoid rattans (Rowe et al., 2004
). The mechanism was not observed in D. polyacanthos in which lighter axes often remain firmly attached to the surrounding vegetation by highly catchy cirri. In cases of severe disturbance, slender axes can bend relatively easily and reorientation of the growing apex produces a characteristic "S"-shaped axis if a growing apex detaches or emerges from an understory support (Fig. 2C).
|
, 1998
, 2004
; Speck and Rowe, 1999
; Isnard et al., 2003a
, b
) are not observed. Lacking secondary growth, the potential for modifying mechanical properties of stems might be more constrained than in dicotyledonous lianas. Older climbing axes of D. orthacanthos had an increase in Estr at the very base in the part of the axis where the leaf sheath had been shed. A localized increase towards the base was also observed in older stages of D. polyacanthos. These patterns differ from most tested woody lianas, which have a more general decrease in Estr from the youngest apical parts of mature plants toward the base. They also differ from most self-supporting plants tested that show a progressive and continuous increase in structural Young's modulus toward the base (Speck, 1994
; Speck and Rowe, 1999
). A comparison of Desmoncus with old stages of a wide range of dicotyledonous lianas (Table 1) shows higher values of Estr in the climbing palms by a factor of 10–20. High values of Estr and relatively little overall change in material properties during ontogeny are, however, consistent with some kinds of woody dicotyledonous climbers referred to as semi-self-supporting growth forms (Rowe and Speck, 1998
; Speck and Rowe, 1999
). These often have relatively stiff and straight stems compared with twining or winding lianas and among tropical climbers include species of Strychnos and Croton (Rowe and Speck, 1998
; Gallenmüller et al., 2000
, 2001
; Rowe et al., in press
), which can show relatively high values of Estr and stiff material properties in older ontogenetic stages (Table 1). In Desmoncus, the outer leaf sheath provides stiffness during maturation of stem tissues in youngest stages of growth and ensures a self-supporting orientation before the stem contributes much to overall stiffness. Furthermore, as suggested by Putz (1990)
for Asian rattans, young stems and searchers of Desmoncus with thick outer leaf sheaths are much stiffer in terms of EI than most young woody vine leaders. This difference relying on geometry-enhanced stiffness allows climbing palms to span large gaps between host supports.
|
). Anomalous cambial development also promotes healing of injured stems (Dobbins and Fisher, 1986
; Fisher and Ewers, 1989
, 1991
). The increase in Estr towards the base and lack of secondary growth in Desmoncus are not at all consistent with such strategies. Furthermore, climbing palms cannot replace damaged vascular tissue and consequently might be viewed as being highly vulnerable to mechanical damage (Tomlinson and Zimmermann, 2003
). In Desmoncus, greater flexibility is obtained via a more passive mechanism of leaf sheath fracture, but nevertheless constitutes an effective response to potentially critical mechanical forces. Though quite different from dicotyledonous lianas, we suggest that the composite stem/leaf sheath structure characterizing Desmoncus and possibly other climbing palms might represent at least part of the adaptive mechanical control over stiffness and flexibility, which is so important for the climbing habit, particularly among larger-bodied species.
The ability to resprout from the stem after damage to the shoot allows many dicotyledonous lianas to survive mechanical damage (Caballé, 1977
, 1980
; Putz, 1984
). Many climbing palms including Desmoncus are not able to produce reiterations or adventitious roots from the aboveground stem. Desmoncus does have additional characteristics that are probably adapted to the climbing habit. First, the ability to sprout from the rhizome (Putz, 1983
) after damage and/or dieback of a ramet is an alternative way to replace rather than survive mechanical damage. Second, axes of D. polyacanthos are narrow and in spite of relatively high values of Estr, resistance to bending forces remains small by retaining a narrow diameter.
Evolutionary context
Tree palms (Welfia, Euterpe, Iriartea; Rich, 1986
, 1987
) have mechanical architectures similar to the tested species of Desmoncus with an increase of EI and Estr toward the base. The comparison of Desmoncus and Bactris revealed little difference in stem stiffness and only marginally higher values of Estr in Bactris, although trends along stems are similar. The comparison based on mechanical properties is consistent with findings on the hydraulic architecture where the stem vascular system of Desmoncus resembles that of tree palms and the palm model based on Rhapis excelsa. Interestingly, this organization differs from the highly modified vascular architecture found in Calamus (Tomlinson and Zimmermann, 2003
).
In terms of the mechanical and hydraulic organization observed in Desmoncus and possibly in general in the Arecoideae, differences between arborescent, semi-self-supporting, and climbing growth forms do not involve profound modifications of stem organization and material properties. This contrasts markedly to the situation in many dicotyledonous families in which trees, shrubs, and climbers can have drastically different patterns of developmental and mechanical properties, largely but not always mediated by variations in secondary growth.
Previous authors have noted the extraordinary morphological convergence of climbing forms in Desmoncus and the Calamoideae (Corner, 1966
; Tomlinson and Fisher, 2000
) as well as the great diversity of climbing species in Calamoideae (Uhl and Dransfield, 1987
; Baker et al., 2000a
, b
). Beyond a broad similarity of height, size, and attachment in the two groups, it would be of great interest to know whether (1) the mechanical architecture involving a central stem, leaf sheath, and fracturing of the leaf sheath after extreme mechanical stress is similar for large bodied members of the Calamoideae and (2) whether smaller-bodied species show different architectures linked to habitat preferences as observed in Desmoncus.
Previous observers including Holttum (1955)
noted the uniformity of growth pattern among monocotyledons that nevertheless gave rise to a wide variety of body sizes and growth forms. These observations led to the concept of establishment growth characterizing development in palms where the stem base thickens without internodal expansion and forms an obconical stem base (Tomlinson and Zimmermann, 1966
). Development in Desmoncus does not pass through an extended establishment phase, and internodes extend early after the second leaf opens (Putz, 1983
). As suggested by Tomlinson (1979)
, habit evolution in palms could involve heterochronic modification of this developmental trajectory and retention of juvenile characteristics. In both species of Desmoncus, especially in D. polyacanthos, reduction of the establishment phase results in a slender stem. It is possible that a decrease in rate of stem thickening (the establishment phase) combined with an increase in internode elongation occurred in Desmoncus and possibly other climbing palms.
Recent investigations of dicotyledonous lianas have demonstrated that in some groups the transition from self-supporter to climber involved a wide array of hydraulic and mechanical novelties that strongly influenced and possibly constrained future growth form change and evolution in the group. In some families of dicotyledons such as the Aristolochiaceae, there is evidence that lianoid forms rarely reversed to "truly self-supporting forms" (Rowe and Speck, 2004
). Such reversals observed in other groups can differ in their mechanical and geometrical characters compared with self-supporting species (Speck et al., 1997
; Civeyrel and Rowe, 2001
). Interestingly, in Desmoncus and possibly other climbing monocotyledons, the prevalent changes in development compared with self-supporting species appear to involve changes in stem size, geometry, and leaf sheath morphology rather than profound changes in the integral mechanical properties of the stem.
FOOTNOTES
1 The authors thank Meriem Fournier, Lilian Blanc, and Pascal Pétronelli for their kind support during fieldwork (2002–2004) in French Guiana at the scientific campus at Kourou and the field camp at Paracou. We also thank two anonymous reviewers. N. P. R. and T. S. acknowledge a PROCOPE exchange grant (2000–2003) and an Alexander von Humboldt award to NPR (2004). Part of this work was supported from a grant to S. I. from the Agence Universitaire de la Francophonie (AUF) (France). 
4 Author for correspondence (e-mail: sisnard{at}cirad.fr
) phone: +33 (0) 467617553, fax: +33 (0) 467615668
LITERATURE CITED
Baker J. W. J. Dransfield T. A. Hedderson 2000a Phylogeny, character evolution, and a new classification of the calamoid palms. Systematic Botany 25: 297-322
Baker W. J. T. A. Hedderson J. Dransfield 2000b Molecular phylogenetics of Calamus (Palmae) and related rattan genera based on 5S nrDNA spacer sequence data. Molecular Phylogenetics and Evolution 14: 218-231[CrossRef][ISI][Medline]
Caballé G. 1977 Multiplication végétative en forêt dense du Gabon de la liane Entada sclerata (Mimosoideae). Adansonia 17: 215-220[ISI]
Caballé G. 1980 Caractéristiques de croissance et multiplication végétative en forêt dense du Gabon de la "liane à eau" Tetracera alnifolia Willd. (Dilleniaceae). Adansonia 19: 309-320
Caballé G. 1986 Sur la biologie des lianes ligneuses en forêt gabonaise. Thèse de Doctorat d'Etat en Sciences, Université de Montpellier 2 Sciences et Techniques du Languedoc, Montpellier, France
Civeyrel L. N. P. Rowe 2001 Phylogenetic relationships of Secamonoideae based on the plastid gene matK, morphology and biomechanics. Annals of the Missouri Botanical Garden 88: 583-602[CrossRef][ISI]
Corner E. J. H. 1966 The natural history of palms. University of California Press, Berkeley, California, USA
Dobbins D. R. J. B. Fisher 1986 Wound responses in girdled stems of lianas. Botanical Gazette 147: 278-289[CrossRef]
Dransfield J. 1978 Growth forms of rain forest palms. In P. B. Tomlinson and M. H. Zimmermann [eds.], Tropical trees as living systems, 247– 268. Cambridge University Press, Cambridge, UK
Dransfield J. N. W. Uhl C. B. Asmussen W. J. Baker M. M. Harley C. E. Lewis In press An outline of a new phylogenetic classification of the palm family, Arecaceae. Kew Bulletin
Escalante S. C. Montaña R. Orellana 2004 Demography and potential extractive use of the liana palms, Desmoncus orthacanthos Martius (Arecaceae), in southern Quintana Roo, Mexico. Forest Ecology and Management 187: 3-18[CrossRef][ISI]
Fisher J. B. F. W. Ewers 1989 Wound healing in stems of lianas after twisting and girdling. Botanical Gazette 150: 251-265[CrossRef]
Fisher J. B. F. W. Ewers 1991 Structural responses to stem injury in vines. In F. E. Putz and H. A. Mooney [eds.], The biology of vines, 99–124. Cambridge University Press, Cambridge, UK
Gallenmüller F. U. Müller N. P. Rowe T. Speck 2000 Variability of habit and mechanical properties in the neotropical liana Croton nuntians. In H. C. Spatz and T. Speck [eds.], Plant biomechanics 2000, 63– 70. Thieme, Stuttgart, Germany
Gallenmüller F. U. Müller N. Rowe T. Speck 2001 The growth form of Croton pullei (Euphorbiaceae)—functional morphology and biomechanics of a neotropical liana. Plant Biology 3: 50-61[CrossRef][ISI]
Hahn W. J. 2002 A phylogenetic analysis of the arecoid line of palms based on plastid DNA sequence data. Molecular Phylogenetics and Evolution 23: 189-204[CrossRef][ISI][Medline]
Henderson A. F. Chavez 1993 Desmoncus as a useful palm in the western Amazon basin. Principes 37: 184-186
Henderson A. G. Galeano R. Bernal 1995 Field guide to the palms of the Americas. Princeton University Press, Princeton, New Jersey, USA
Holttum R. E. 1955 Growth-habits of monocotyledons—variation on a theme. Phytomorphology 5: 399-413
Isnard S. N. P. Rowe T. Speck 2003a Growth habit and mechanical architecture of the sand dune-adapted climber Clematis flammula var. maritima L. Annals of Botany 91: 1-11
Isnard S. T. Speck N. P. Rowe 2003b Mechanical architecture and development in different growth forms of Clematis L.: implications for canalised evolution of growth forms. New Phytologist 158: 543-559[CrossRef][ISI]
Lee D. W. J. H. Richards 1991 Heteroblastic development in vines. In F. E. Putz and H. A. Mooney [eds.], The biology of vines, 205–243. Cambridge University Press, Cambridge, UK
Niklas K. J. 1992 Plant biomechanics: an engineering approach to plant form and function. University of Chicago Press, Chicago, Illinois, USA
Putz F. E. 1983 Developmental morphology of Desmoncus isthmius, a climbing colonial, cocosoid palm. Principes 27: 38-42
Putz F. E. 1984 The natural history of lianas on Barro Colorado Island, Panama. Ecology 65: 1713-1724[CrossRef][ISI]
Putz F. E. 1990 Growth habits and trellis requirements of climbing palms (Calamus spp.) in north-eastern Queensland. Australian Journal of Botany 38: 603-608[CrossRef]
Putz F. E. M. Holbrook 1991 Biomechanical studies of vines. In F. E. Putz and H. A. Mooney [eds.], The biology of vines, 73–95. Cambridge University Press, Cambridge, UK
Rich P. M. 1986 Mechanical architecture of arborescent rain forest palms. Principes 30: 117-131
Rich P. M. 1987 Mechanical structure of the stem of arborescent palms. Botanical Gazette 148: 42-50[CrossRef]
Rowe N. P. S. Isnard T. Speck 2004 Diversity of mechanical architectures in climbing plants: an evolutionary perspective. Journal of Plant Growth Regulation 23: 108-128[CrossRef][ISI]
Rowe N. P. S. Isnard F. Gallenmüller T. Speck In press Diversity of mechanical architectures in climbing plants: an ecological perspective. In A. Herrel, T. Speck, and N. P. Rowe [eds.], Ecology and biomechanics: a mechanical approach to the ecology of animals and plants. CRC Press, Boca Raton, Florida, USA
Rowe N. P. T. Speck 1996 Biomechanical characteristics of the ontogeny and growth habit of the tropical liana Condylocarpon guianense (Apocynaceae). International Journal of Plant Sciences 157: 406-417[CrossRef][ISI]
Rowe N. P. T. Speck 1997 Biomechanics of Lycopodiella cernua and Huperzia squarrosa: implications for inferring growth habits of fossil small-bodied lycopsids. Mededelingen Nederlands Instituut voor Toegepaste Geowetenschappen TNO 58: 293-302
Rowe N. P. T. Speck 1998 Biomechanics of plant growth forms: the trouble with fossil plants. Review of Palaeobotany and Palynology 102: 43-62
Rowe N. P. T. Speck 2004 Hydraulics and mechanics of plants: novelty, innovation and evolution. In I. Poole and A. R. Hemsley [eds.], The evolution of plant physiology, 297–325. Royal Botanic Gardens, Kew, Richmond, UK
Siebert S. F. 2000 Abundance and growth of Desmoncus orthacanthos Mart. (Palmae) in response to light and ramet harvesting in five forest sites in Belize. Forest Ecology and Management 137: 83-90[CrossRef][ISI]
Speck T. 1994 Bending stability of plant stems: ontogenetical, ecological, and phylogenetical aspects. Biomimetics 2: 109-127
Speck T. N. P. Rowe 1999 A quantitative approach for analytically defining size, growth form and habit in living and fossil plants. In M. H. Kurmann and A. R. Hemsley [eds.], The evolution of plant architecture, 447–479. Royal Botanic Gardens, Kew, Richmond, UK
Speck T. N. P. Rowe F. Brüchert W. Haberer F. Gallenmüller H. C. Spatz 1996 How plants adjust the "material properties" of their stems according to differing mechanical constraints during growth: an example of smart design in nature. In A. E. Engin [ed.], Bioengineering Proceedings of the Engineering Systems Design and Analysis Conference. PD-vol. 77 5: 233-241 American Society of Mechanical Engineers, New York, USA
Speck T. C. Neinhus F. Gallenmüller N. P. Rowe 1997 Trees and shrubs in the mainly lianescent genus Aristolochia s. l.: secondary evolution of the self-supporting growth habit?. In G. Jeronomidis and J. F. V. Vincent [eds.], Plant biomechanics 1997, I, 201–207. Centre for Biomimetics, University of Reading, Reading, UK
Tomlinson P. B. 1962 The leaf base in palms its morphology and mechanical biology. Journal of the Arnold Arboretum 43: 23-45
Tomlinson P. B. 1979 Juvénilité et néoténie chez les Monocotylédones. Bulletin de la Société Botanique de France 126: 227-232
Tomlinson P. B. J. B. Fisher 2000 Stem vasculature in climbing monocotyledons: a comparative approach. In K. L. Wilson and D. A. Morrison [eds.], Monocots: systematics and evolution, 89–96. CSIRO Publishing, Melbourne, Australia and Royal Botanic Gardens, Kew, UK
Tomlinson P. B. M. H. Zimmermann 1966 Anatomy of the palms Raphis excelsa. III. Juvenile phase. Journal of the Arnold Arboretum 47: 301-312
Tomlinson P. B. M. H. Zimmermann 2003 Stem vascular architecture in the American climbing palm Desmoncus (Arecaceae-Arecoideae-Bactridinae). Botanical Journal of the Linnean Society 142: 243-254[CrossRef]
Troy J. R. P. M. Ashton B. C. Larson 1997 A protocol for measuring abundance and size of a neotropical liana, Desmoncus polyacanthos (Palmae), in relation to forest structure. Economic Botany 51: 339-346[ISI]
Uhl N. W. J. Dransfield 1987 Genera palmarum: a classification of palms based on the work of Harold E. Moore Jr. Allen Press, Lawrence, Kansas, USA
Vincent J. F. V. 1990 Structural biomaterials. Princeton University Press, Princeton, New Jersey, USA
This article has been cited by other articles:
|
|
 |
|
  D. Barthelemy and Y. Caraglio Plant Architecture: A Dynamic, Multilevel and Comprehensive Approach to Plant Form, Structure and Ontogeny Ann. Bot., March 1, 2007; 99(3): 375 - 407. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||
