ABSTRACT
A node is the point of attachment of the leaf to the stem of a plant; gaps associated with nodes have been viewed as discontinuities of the stem vascular system. We tested the hypothesis that the node/gap is a spring-like joint that impacts stem flexibility even well after the leaves have been shed, with some stems specialized for elongation and others for flexibility. Four-point bending tests were done using an Instron Mechanical Testing Device with the independent variable being the number of nodes in the stem segment and dependent variables being Modulus of Elasticity (MOE), Modulus of Rupture (MOR), and xylem density. Node anatomy was examined microscopically to assess structure and function. The stiffness of the stem was inversely proportional to the frequency of leaf nodes. Surprisingly, xylem density was inversely proportional to the frequency of leaf nodes in stems of adult trees. The tissue around nodes/gaps consisted of twisted and contorted cells that may be effective at absorbing compressive and tensile stresses. Because nodes behave as spring-like joints, the frequency of nodes relates to function, with some stems specialized for vertical expansion and others for light capture and damping of wind stress. The ultimate stems on a tree are the most bendable, which may allow the trees to avoid breakage.
INTRODUCTION
Leaves of plants must be positioned to adequately capture sunlight; for this they rely on mechanical support from the stem (Rowe & Speck 2005; Niklas et al. 2006). The point of attachment of a leaf to the stem is the node, and branch buds commonly occur just above the leaf node in the axil (Fig. 1).
Niklas (1997) reported that the nodes of the grass Arundinaria tecta act as spring-like joints. Similarly, it was found in woody branches of Cercis occidentalis (Fabaceae) that the arrangement of leaves on a stem can influence the stiffness of the stem in different bending planes, with greater flexibility in the plane parallel to the node (Caringella et al. 2014).
Diagrams showing a node on a 10-mm-wide stem of a Southern California black walnut tree. – A: with intact tissues. – B: the bark has been removed to reveal the wood grain. – C: a radial longitudinal section view; the cross-hatched area in the middle of the stem represents the chambered pith.
Citation: IAWA Journal 39, 4 (2018) ; 10.1163/22941932-20170205
In the present study we tested the hypothesis that the node/gap in a woody stem is a spring-like joint that impacts stem flexibility, even well after the leaves have been shed. We sampled stems that had exhibited a wide range of internode elongation to determine the relationship between the nodal frequency and various mechanical properties. We also tried to mimic mechanical effects of nodes and branch gaps with the use of a drill and the anatomy of leaf nodes was examined to determine structure-function relationships.
Figure 1 illustrates the leaf and bud traces and associated leaf and bud gaps in a stem of Juglans californica S. Watson, the Southern California black walnut, more than a year after the leaf was shed. The gaps are composed of parenchyma cells instead of fiber-rich xylem. The pith consists of air chambers separated by thin diaphragms of tissue. Because the parenchyma cells in the pith and bud gap are thin-walled, we expected that their artificial removal with a drill would not impact mechanical properties of the stem.
We sampled stems from irrigated and non-irrigated native stands on the Cal Poly Pomona campus, with a wide range of variation in the distance between nodes on the stem. In addition, earlier, two extreme stem types were sampled for walnuts (Utsumi et al. 2010). Following a wildfire in San Dimas, California, stems that had sprouted (resprouts) from the root crown had three-fold higher photosynthetic rates than lateral adult stems of unburned trees. Here we reexamined and reanalyzed the stems from that study, counting the number of nodes on each stem segment. We reasoned that distal lateral branches, with nodes very close together on the stem, should have greater flexibility than resprout stems, with nodes far apart.
MATERIALS AND METHODS
Stems were sampled from a native stand on the campus of the California State Polytechnic University, Pomona (34° 03' 24.24" N, 117° 49' 23.98" W). Lateral branches of irrigated (n = 145) and adjacent non-irrigated trees (n = 149) were sampled, with the irrigation treatment as described earlier (Duarte 2013). These stems were sampled in autumn (n = 95), winter (n = 128), and spring (n = 71), with stems ranging from one to six years old. The sampled trees at this site were all adults with unirrigated trees averaging 9.77 m (± SE of 0.91) in height and unirrigated trees averaging 9.38 m (± SE of 0.54).
To compare adults and resprouts, trees were also sampled from a site 3 km north-northeast of the Pomona site at Frank G. Bonelli Regional Park in San Dimas, California (34° 05' 23" N, 117° 48' 59" W). At the San Dimas site some of the trees had been burned to the ground by wildfire. Those plants had resprout stems that were sampled. These were compared to lateral branches of adjacent mature trees that had not burned. Twenty stems were sampled at both the burned and unburned sites; all of those sam-ples were taken in November of 2007. The average height of adult trees was 8.45 m (± SE of 0.73) with an average age of 44.5 years (± SE of 4.1 years). The sampled lateral branches ranged from two to six years old with a mean of 3.44 years (SE = 0.32). The average height of resprouts was 2.53 m (± 0.13) and their stems were all one year old at the time of sampling (Utsumi et al. 2010).
For both sites and all treatments 10-mm-diameter stems (including bark) were targeted and were cut to 160 mm in length for four-point bending tests. Mechanical tests were done using an Instron Mechanical Testing Device (model 3342; Instron, Canton, Ma., USA) with a compression load cell of 500 N; the crosshead speed was 20 mm min-1, as described earlier (Caringella et al. 2014).
Flexural stiffness (EI) was calculated according to the equation EI = P/V (a 2/12) (3L-4a) with P equal to load or force, V the vertical displacement, a the distance (45 mm) between support and loading point, and L (135 mm) the span between the two supported ends. The second moment of area for the xylem (I) was calculated for each stem, using the formula for a hollow cylinder, I = π (R 4 -r 4 )/4 where R is the radius from stem center to the outside of the xylem, and r is the radius of the pith (Gere & Timoshenko 1984). This was based upon previous studies wherein we found that for 10-mm-diameter stems the xylem was the dominant mechanical tissue and the pith and bark should be eliminated in calculations of I (Utsumi et al. 2010; Caringella et al. 2014). The relationship EI/I was used to calculate E, the modulus of elasticity (MOE), also known as Young’s modulus.
For most stems (n = 294 at the Pomona site and n = 40 at the San Dimas site), the bending experiments were allowed to continue to the point of stem failure (F max , load at stem failure). MOR was then calculated from the equation MOR = (F max × a × R)/I, modified from Ugural (1991).
Xylem density was calculated as the dry mass per saturated volume of xylem, measured by the displacement method and corrected to temperature. Xylem samples of 50 mm long were taken from above or below the failure region of each test segment, including a representative number of nodes. The pith and bark (including the phloem and vascular cambium) were removed from the xylem, but gaps were left intact. Specific MOR and Specific MOE were calculated by dividing the parameters by the xylem density for that stem.
In certain experiments the stems (n = 48, with an equal number of untreated controls) were not bent to failure, and were instead retested within the elastic range following manipulations. In these cases, tests were run for only 2 mm extension by the load cell. In one experiment (n = 12), measurements were made before and after removal of the pith with a wire that was fed through the stem segment and pulled completely through the pith at least three times to remove all the pith tissues. In another experiment (n = 20), the bud gaps were removed with the use of a 3 mm drill-bit, to see if the gap tissue was providing mechanical support. In a third experiment (n = 7) both pith and bud gaps were removed. In a fourth experiment (n = 7), a series of artificial nodes were created with a 3 mm drill-bit, in an approximation of the spiral phyllotaxy. The drill went completely through the xylem into the pith. Stems were retested, then new holes were drilled and retesting was repeated until 8 artificial nodes were present in each stem.
A sliding microtome was used to cut 25-μm-thick transverse and tangential sections through leaf and bud gaps, and sections were generally stained with safranin and fast green for microscopic observation. Other sections were stained with a solution of 1% phloroglucinol in 20% HCl to detect lignin (Ruzin 1999).
For stems from the Pomona site, ANCOVA analysis was done with a best fit model using corrected R2 and using least square means. Models were generated for MOR, MOE, xylem density, and leaf nodes per meter. The parameters were set with the initial variables of season, irrigation type, MOR, MOE, xylem density, and leaf nodes per meter. In examining possible seasonal effects on mechanical properties, Kruskal-Wallis nonparametric analysis of variance was performed, followed by Dunn’s pairwise comparisons because the data were not normally distributed.
For stems from the San Dimas site, comparisons of the adult and resprout stems were done with one-way ANOVA followed by Fisher’s Least Significant Difference Comparison. Statistical analyses were performed using SAS version 9.2 (SAS Institute Inc., North Carolina, USA).
RESULTS
We sampled walnut stems from irrigated (n = 145) and non-irrigated (n = 149) native stands on the Cal Poly Pomona campus, with nodal frequencies ranging from less than 10 nodes m-1 to over 160 nodes m-1. The average xylem diameter of sampled stems was 9.37 mm (SE = 0.52), with stem ages ranging from one to six years old. Irrigation had no significant impact on any of the measured parameters. However, flexural stiffness, MOE, Specific MOE, MOR, and Specific MOR were all inversely related to the frequency of leaf nodes. The only biomechanical parameter that was positively correlated with the frequency of leaf nodes was xylem density (Fig. 2).
For the ANCOVA of MOE the best-fit model was predicted with two variables, leaf nodes m-1 (LN) and season (MSE = 6729361.4, R2 = 0.146). In Type III Sum of Square analysis, LN had a greater contribution to the model (F = 26.0, P < 0.0001) than that of season (F = 3.47, P = 0.032).
Various stem mechanical parameters as functions of the frequency of leaf nodes on the stem (leaf nodes per meter), including, from top to bottom, flexural stiffness (exponential curve fit), xylem density (logarithmic curve fit), Specific Modulus of Rupture (MOR – exponential curve fit) and Specific Modulus of Elasticity (MOE – logarithmic curve fit). A total of 294 stems were sampled as specified in Table 1. Data were binned at intervals of 6 leaf nodes m- 1 with frequencies from 125 to 250 nodes m-1 pooled as a single bin. Values are means ± 1 standard error.
Citation: IAWA Journal 39, 4 (2018) ; 10.1163/22941932-20170205
In the ANCOVA of MOR the best-fit model was with two variables, LN and season (MSE = 1422.8, R2 = 0.204) with the seasons of autumn (n = 95), winter (n = 128), and spring (n = 71) represented. In Type III Sum of Squares analysis, LN had a similar contribution (F = 34.42, P < 0.0001) to that of seasonal effects (F = 7.847, P < 0.0004). In both MOR and MOE, xylem density did not contribute to the best predictive model.
Lastly, in the ANCOVA analysis of xylem density, the best-fit model was with two variables, LN and season (MSE = 0.002, R2 = 0.403). In Type III Sum of Squares analysis, seasonal variation had a greater contribution (F = 75.0, P < 0.0001) to the model than LN (F = 5.63, P = 0.018).
In comparing results by season, the winter samples had about 20% lower values of Specific MOE and Specific MOR than the autumn and spring samples (P < 0.001). The autumn and spring samples were not significantly different from one another. However, the winter samples had about 50% more leaf nodes than the autumn and spring samples (P < 0.001), concomitant with the greater flexibility of the winter samples (Table 1).
Mechanical properties and leaf nodes of lateral stems of Juglans californica sampled in Pomona, California. Mean values (± 1 standard error) are shown by season for Density Specific Modulus of Rupture (Specific MOR in N mm-2), Specific Modulus of Elasticity (Specific MOE in N mm-2), and Nodal Frequency (leaf nodes per meter). N = number of stems sampled by season. P-values are for Kruskal-Wallis nonparametric analysis of variance; values followed by different letters in columns indicate significant differences between seasons after Dunn’s pairwise comparisons (P < 0.001).
A comparison of the mechanical properties of 10 mm diameter stems for one-year-old resprouts versus adult lateral branches. The differential node frequency enhances support for rapidly growing vertical shoots in resprouts and flexibility for slow growing lateral shoots in adults. Parameters included Leaf nodes per meter, Flexural stiffness, Specific MOE, and Specific MOR. For each parameter N = 20; means (± 1 SE) are shown. * differences between growth form were significant at P < 0.05.
Citation: IAWA Journal 39, 4 (2018) ; 10.1163/22941932-20170205
Stem microscopic sections showing node anatomy. A lower magnification tangential longitudinal section (A) showing a bud gap (bg) and leaf gap (lg). The box in A outlines the area shown at higher magnification in B, which shows the contorted xylem (cx) tissue adjacent the bud gap. – C: a transverse section showing a bud gap (bg) that runs from the central pith, with a pith chamber (p), to the exterior of the xylem.
Citation: IAWA Journal 39, 4 (2018) ; 10.1163/22941932-20170205
In comparing growth forms, the resprout stems had a 3.75 times greater internode elongation than lateral branches of mature trees, based upon having 32.5 ± 1.9 nodes m-1 (mean + SE) versus 121.9 ± 30.6 nodes m-1 for adult stems (Fig. 3). Concomitant with having fewer nodes, the resprout stems had 2.5 times greater flexural stiffness and 2.6 times greater Specific MOE but only a 1.5 times greater Specific MOR (Fig. 3). The number of nodes had a greater impact on the Specific MOE, reflecting greater flexibility, than on Specific MOR, reflecting resistance to breakage.
At the node, the bud gap was composed of small diameter parenchyma cells and tissue surrounding the gap was dense, appearing with interlocked grain and very low vessel frequency. Small circular vessels were seen adjacent the gap in some cases. A twisted web of fibers, vessels and parenchyma was especially prominent adjacent each gap (Fig. 4A, B). The bud gaps often extended all the way across the xylem as seen in transverse view (Fig. 4C). The gap parenchyma cells had thin lignified cell walls, based upon red staining with phloroglucinol/HCl.
Removal of the chambered pith with a wire had no measurable impact on flexural stiffness. Likewise, drilling out the existing gaps with a 3-mm-diameter drill-bit had no measurable effect on stiffness. Adding “artificial nodes” by drilling additional 3-mm-diameter holes through the xylem, to simulate gaps, had an effect, but less than expected. Going from 10 to 60 artificial nodes per meter resulted in a 6% reduction of flexural stiffness, whereas a similar increase in natural nodes corresponded to decreased flexural stiffness of about 20%.
DISCUSSION
Leonardo da Vinci may have been the first to model plant architecture when he described the proportionality of growth and cross-sectional area between different orders of branches and the trunk of trees (Richter 1970). Here we report the surprising result that greater frequency of leaf nodes results in greater flexibility (less stiffness) of woody stems. Woody plants may be more articulated than is generally appreciated, which has implications for tree architecture, the lumber industry, and the field of biomimetics. Mimicking the design of leaf nodes could improve the performance of various materials and structures for use by humans. However, to assess this further it will be necessary to develop methods to separately measure the flexibility of individual leaf nodes and individual internodes in woody stems.
Leaf nodes have previously been reported as areas that impact water transport properties (Zimmermann & Sperry 1983; Salleo & Lo Gullo 1986; Tyree & Zimmermann 2002). Because the xylem of plants functions both for water transport and mechanical support, it is not surprising that junctions between organs can act as both hydraulic and mechanical inflection points. Now we know that the mechanical effects of nodes can apply not just to herbaceous plants (Niklas 1997) but to woody plants such as Cercis occidentalis (Caringella et al. 2014) and Juglans californica, reported herein.
Surprisingly, xylem density was inversely proportional to flexural stiffness, contrary to the norm in the fields of material science and biomechanics (Gere & Timoshenko 1984; Ugural 1991; Niklas 1992; Vogel 2003). This was due to the overwhelming impact of leaf nodes, which increased flexibility despite higher xylem density. The results are consistent with results by Caringella et al. (2014) who reported that the nodes of Cercis occidentalis were denser and more flexible than the internodes. In both Cercis and Juglans the tissue surrounding the gap is very dense with interlocked grain and very low vessel frequency. Related to this, the knots in lumber, which are from old branches on the trunk of a tree, are of high density but decrease the flexural stiffness and MOR (Panshin & De Zeeuw 1980; Grant et al. 1984).
The irrigation treatment had no measurable effect on mechanical properties, and the apparent seasonal effect that we found in this study, with the winter stems being more flexible than autumn and spring stems, may have been spurious. We inadvertently sampled stems of higher nodal frequency in the winter season, resulting in lower MOR and MOE values for those samples. Thus nodal frequency appears to be more important than season in determining mechanical properties.
As noted first by da Vinci, young, vigorously growing branches often produce leaves that are far apart from one another (Richter 1970). Thus the frequency of nodes on a stem may be related to the primary growth rate and the photosynthetic capacity of different shoots. In walnut there appears to be a continuum between, at the extremes, rapidly growing pioneer stems with great internode elongation and low xylem density, and established stems with low internode elongation and high xylem density.
The stiff, rapidly elongating resprout stems may quickly establish space for leaves in the vertical dimension. In contrast, the lateral stems of mature adults remain flexible and produce leaves close together near the tip of the stem. In response to wind, the more flexible stems of adults could avoid breakage by bending with the wind and reducing the drag of the leaves. Similarly, it was recently reported that the frequency of leaf nodes impact the mechanics of coppice sprouts (shoots that form after cutting down the tree trunk) in trees of chestnut, sycamore and ash (Ozden 2016). Low nodal frequency can result in the useful structural properties of coppice growth for the lumber industry
Why is the node area more flexible than expected? It may have to do with the structure of the tissues surrounding the bud gap. Gaps and associated tissues may allow for the absorption of compressive and tensile forces as the stem is bent, without breakage. In the case of Asimina triloba, it has been shown that the high flexibility and “twist-ability” of small stems and petioles may protect the large leaves when exposed to high winds (Goodrich et al. 2016). Similarly, at a higher level of organization, at stem-branch junctions the tortuous and interlocked wood grain of many species adds density and helps prevent breakage that might otherwise occur with wind or snow loading (Slater et al. 2014). Higher nodal frequency near the ultimate tips might help Juglans trees avoid breakage due to wind stress; in effect the flexible distal stems may bend to help prevent breakage.
ACKNOWLEDGEMENTS
The authors thank the staff of Frank G. Bonelli Regional Park for providing the opportunity to conduct this study. Karl Niklas, Duncan Slater and two anonymous reviewers provided insightful reviews of the manuscript.
REFERENCES
Caringella MA , Bergman BA , Stanfield RC , Ewers MM , Bobich EG , Ewers FW . 2014. Effects of phyllotaxy on biomechanical properties of stems of Cercis occidentalis (Fabaceae). Am. J. Bot. 101: 206–210. DOI: .
Duarte D. 2013. Plastic responses of wood development in California black walnut (Juglans californica): effects of irrigation and post-fire growth. Master’s Thesis, California State Polytechnic University, Pomona.
Gere JM , Timoshenko SP . 1984. Mechanics of materials, Ed. 2. Belmont Inc., Wadsworth, California.
Goodrich KR , Ortiz LA , Coughlin DJ . 2016. Unusual twig “twistiness” in pawpaw (Asimina triloba) provides biomechanical protection for distal foliage in high winds. Am. J. Bot. 103: 1872–1879. DOI: .
Grant DJ , Anton A , Lind P . 1984. Bending strength, stiffness, and stress-grade of structural Pinus radiata: Effect of knots and timber density. New Zeal. J. For. Sci. 14: 331–348.
Niklas KJ . 1992. Plant biomechanics: an engineering approach to plant form and function. University of Chicago Press.
Niklas KJ . 1997. Responses of hollow, septate stems to vibrations: biomechanical evidence that nodes can act mechanically as spring-like joints. Ann. Bot. 80: 437–448. DOI: org/10.1006/anbo.1997.0463.
Niklas KJ , Spatz H , Vincent J . 2006. Plant biomechanics: an overview and prospectus. Am. J. Bot. 93: 1369–1378. DOI: .
Ozden S. 2016. The relationship between the anatomy and mechanical properties of different green wood species. Doctoral Dissertation, University of Manchester.
Panshin AJ , De Zeeuw C . 1980. Textbook of wood technology, Ed. 4. McGraw-Hill, New York.
Richter JP . 1970. The notebooks of Leonardo da Vinci: Vol. 1. Dover Publications, New York.
Rowe N , Speck T . 2005. Plant growth forms: an ecological and evolutionary perspective. New Phytologist 166: 61–72. DOI: .
Ruzin SE . 1999. Plant microtechnique and microscopy. Oxford University Press.
Salleo S , Lo Gullo MA . 1986. Xylem cavitation in nodes and internodes of whole Chorisia insignis H. B. et K. plants subjected to water stress: relations between xylem conduit size and cavitation. Ann. Bot. 58: 431–441.
Slater D , Bradley RS , Withers PJ , Ennos AR . 2014. The anatomy and grain pattern in forks of hazel (Corylus avellana L.) and other tree species. Trees 28: 1437–1448.
Tyree MT , Zimmermann MH . 2002. Xylem structure and the ascent of sap. Ed. 2. Springer-Verlag, New York.
Ugural AC . 1991. Mechanics of materials. McGraw-Hill, New York.
Utsumi Y , Bobich EG , Ewers FW . 2010. Photosynthetic, hydraulic and biomechanical responses of Juglans californica shoots to wildfire. Oecologia 164: 331–338. DOI: .
Vogel S. 2003. Comparative biomechanics: life’s physical world. Princeton University Press.
Zimmermann MH , Sperry JS . 1983. Anatomy of the palm Rhapis excelsa. IX: Xylem structure of the leaf insertion. J. Arnold Arbor. 64: 599–609.
Corresponding author: e-mail: fwewers@cpp.edu
