Wind and mechanical stimuli differentially affect leaf traits in Plantago major


  • Niels P. R. Anten,

    1. Ecology and Biodiversity, Institute of Environmental Biology, Utrecht University, PO Box 800.84, 3508TB, Utrecht, the Netherlands
    Search for more papers by this author
  • Rafael Alcalá-Herrera,

    1. Ecology and Biodiversity, Institute of Environmental Biology, Utrecht University, PO Box 800.84, 3508TB, Utrecht, the Netherlands
    2. Area de Ecologia, Universidad de Córdoba, Ctra. Madrid, Km. 396, 14071 Córdoba, Spain
    Search for more papers by this author
  • Feike Schieving,

    1. Ecology and Biodiversity, Institute of Environmental Biology, Utrecht University, PO Box 800.84, 3508TB, Utrecht, the Netherlands
    Search for more papers by this author
  • Yusuke Onoda

    1. Ecology and Biodiversity, Institute of Environmental Biology, Utrecht University, PO Box 800.84, 3508TB, Utrecht, the Netherlands
    2. Department of Biology, Faculty of Sciences, Kyushu University, Fukuoka 812-8581, Japan
    Search for more papers by this author

Author for correspondence:
Niels PR Anten
Tel: +31 30 2536846


  • Analysing plant phenotypic plasticity in response to wind is complicated as this factor entails not only mechanical stress but also affects leaf gas and heat exchange.
  • We exposed Plantago major plants to brushing (mechanical stress, MS) and wind (MS and air flow) and determined the effects on physiological, morphological and mechanical characteristics of leaf petioles and laminas as well as on growth and biomass allocation at the whole-plant level.
  • Both MS and wind similarly reduced growth but their effects on morphological and mechanical plant traits were different. MS induced the formation of leaves with more slender petioles, and more elliptic and thinner laminas, while wind tended to evoke the opposite response. These morphological and mechanical changes increased lamina and petiole flexibility in MS plants, thus reducing mechanical stress by reconfiguration of plant structure. Responses to wind, on the other hand, seemed to be more associated with reducing transpiration.
  • These results show that responses to mechanical stress and wind can be different and even in the opposite direction. Plant responses to wind in the field can therefore be variable depending on overall environmental conditions and plant characteristics.


Analysing phenotypic plasticity to changes in environmental factors (e.g. temperature, light, water availability or windiness) is complicated, in part because these conditions themselves may have multiple effects on plants. The balance of these multiple effects depends on the overall environmental conditions as well as on the characteristics of the plants themselves (Bradshaw, 1965). Changes in environmental factors may thus induce multiple responses in plants, which will determine the ultimate trait values of a plant, depending on their relative magnitude and potentially interactive effects. Here we analyse the effects of one such environmental factor, wind.

Wind is a particularly complicated environmental factor, having several effects on plants (Grace, 1977; Ennos, 1997). Firstly, it reduces the leaf boundary layer, which increases gas diffusion conductance, heat exchange rate and transpiration rate, depending on leaf characteristics and wind speed. Higher transpiration reduces leaf temperature and may dehydrate plants. Therefore effects of wind on photosynthesis are not simple: it can stimulate photosynthesis as it reduces diffusive resistance for CO2 (Lambers et al. 1998), but it can also reduce photosynthesis by lowering leaf temperatures below the optimum, reducing stomatal conductance to prevent excessive water loss (Retuerto & Woodward, 1992; Ennos, 1997; Lambers et al., 1998) and by causing leaves to roll up or curl inwards, which reduces their effective leaf area (Telewski, 1995).

Secondly, wind flow also exerts drag forces on plants and thus entails mechanical stress. Plant responses to mechanical stress (touching, rubbing or flexing) typically entail inhibition of stem elongation, and increases in stem diameter and root allocation (Jaffe & Forbes, 1993; Telewski & Pruyn, 1998). These responses, denoted as thigmomorphogenesis (thigmo-responses hereafter, Jaffe, 1973), increase a plant’s resistance to mechanical stress (Niklas, 1992; Anten et al., 2005, 2009). Wind can induce similar responses (Lawton, 1982) and this prompted many authors to assume that flexing or rubbing can simulate wind effects (e.g. Niklas, 1998; Anten et al., 2005). However wind can also induce responses that are different or even opposite to those induced by pure mechanical stress; for example, the production of thinner more elongated stems under wind loading (Henry & Thomas, 2002; Smith & Ennos, 2003).

A major difficulty with understanding wind effects is thus to distinguish its mechanical effects from its effect on the microclimate. Few attempts have been made in this direction, although wind is an important factor that has determined the evolution of land plants (Niklas, 1998) and strongly regulates plant demography (Ennos, 1997). Smith & Ennos (2003) conducted an elegant experiment in which the effects of air flow associated with wind were separated from its mechanical effect on stem morphology and mechanical properties. Air flow increased stem length and reduced rigidity, while mechanical flexing induced the opposite response. These opposite effects might explain the variation in plant responses to wind observed across species and environments (Smith & Ennos, 2003).

As regards the wind effects on plants, leaves are probably most strongly influenced. First, leaves are the primary organs of photosynthesis and transpiration, and the microclimatic effects of wind affect them directly. Secondly, the leaves of most plants have large surface area to volume ratios (Niinemets & Fleck, 2002), which maximizes the light capture per mass invested but also makes them prone to mechanical failure under bending and tearing by wind forces (Wilson, 1984). Thigmo-responses are known to vary among different organs (Fluch et al., 2008),yet no study that we know of has attempted to analyse the effects of mechanical stress and wind on leaves separately.

This study was designed to analyse the separate effects of mechanical stress (MS) and wind (i.e. air flow and MS) on leaf petiole and lamina characteristics as well as on growth and biomass allocation at the whole-plant level. To address this question we subjected Plantago major plants to different wind speeds and brushing treatments, with the latter entailing MS without air flow. We analysed a wide variety of leaf traits, including mechanical, anatomical and size-related characteristics. Because of the direct effect of air flow on leaf-level gas exchange, which may aggravate, counteract or override the effect of MS, we expected that leaves would acclimate differentially, or even in the opposite direction, to wind and MS.

Materials and Methods

Plant material and growth conditions

Plantago major L., a common herbaceous species widely distributed all over the temperate world, was used for this experiment. This species can grow in a wide range of habitats, from fertile to infertile soils and from moderately shaded to open environments, but most commonly grows in open, wind-exposed areas. The physiology and ecology of this species have been extensively studied (Kuiper & Bos, 1992).

The experiment was carried out in the glasshouse of the Botanical Garden of Utrecht University. Seeds were sown into pots filled with sand in the glasshouse on 22 October 2005 and germinated within 7 d. On 9 November, 168 seedlings were transplanted into pots (9 × 9 × 9.5 cm) filled with sand, and grown at 50% of natural daylight created by the shading of the glasshouse roof, supplemented by HPI Quick 400 W lamps (Phillips, Eindhoven, the Netherlands). Daily maximum noon photon flux density reached c. 500 μmol m−2 s−1 under sunny conditions, measured with an LI-190SA quantum sensor (Li-Cor, Lincoln, NE, USA). The temperature in the glasshouse was set to 18 : 14°C day : night. Every week, each plant was given 50 ml of 250× dilution of liquid fertilizer (Easy Gro 7 : 7 : 7; Kemira Agro BV, Rozenburg, the Netherlands), 14 mg N plant−1 wk−1.

On 12 January 2006, we selected 48 plants of intermediate height (excluding the tallest and shortest ones) and these were randomly assigned to each of two mechanical stress treatments and two wind treatments in a 2 × 2 factorial design, 12 replicates per treatment combination. All pots were placed on rotating tables that rotated at a speed of 2.5 min−1 (see Supporting Information, Fig. S1 for a picture of the experimental setup). The MS treatments were imposed by a duster placed at 75% of the mean plant height (position was adjusted as plants grew taller) such that plants were brushed either 0 or 2.5 times min−1. This treatment was chosen as it simulates the mechanical effect of wind but with minimal air movement. Wind treatments were established by a fan (BC-4618; Wind Europe SA, Lausanne, Switzerland) placed at a mean distance of c. 2.5 m from the plants. As such, plants were exposed to wind speeds either < 0.2 or c. 2.3 m s−1, measured with an anemometer (D5633; R. Fuess, Berlin, Germany). Treatments were maintained 24 h d−1. Each day, pots were rotated 45° horizontally to ensure that mechanical stress and wind exposure were similar in all directions. During the experiment we took several pictures to estimate leaf deflections during brushing and wind exposure. Overall, results indicated that brushing caused somewhat larger deflections (typically c. 90°) than wind (c. 75°; Fig. S1).

There was one rotating table for each treatment combination. To minimize possible spatial effects, we randomly rearranged the position of plants on rotating tables and the positions of the rotating tables in the glasshouse twice every week.

Gas exchange measurements

Between 27 February and 1 March, photosynthesis was measured on each plant on the fourth leaf counting down from the youngest fully expanded leaf using an open gas exchange system LI6400 (Li-Cor) equipped with an LED blue/red light source. Light response curves were constructed by measuring at different light intensities in steps from 500 to 2000 and down to 0 μmol m−2 s−1. Leaf temperature in the chamber was maintained at 20°C, CO2 concentration at 380 μmol mol−1 and vapour pressure deficit at < 1 kPa. Light-saturated gross photosynthesis (Pmax, μmol m−2 s−1), quantum yield (Φ, μmol μmol−1) and dark respiration (Rd, μmol m−2 s−1) were calculated from these measurements assuming a nonrectangular hyperbola for the relationship between photosynthesis and light (see Marshal & Biscoe, 1980). The temperature of fully expanded leaves for all individuals was measured by a far-red thermometer (IR-AH8T1; Chino, Tokyo, Japan).

Between 27 February and 1 March, three plants per treatment combination were used to measure whole-plant transpiration rates. Pots were packed in polyethylene bags keeping all leaves outside the bags. The change in weight of the pots was monitored every day over 3 d as a measure of transpiration, while plants remained in their treatment position.

Mechanical measurements

After photosynthesis measurements, leaf dimensions (lamina length and width, leaf thickness, petiole and diameter) were measured. Subsequently leaves were cut at the petiole base, wrapped in wet tissue paper and placed in polythene bags to prevent loss of turgor. Three tests (punch-and-die, tensile and bending tests) were performed using a universal mechanical testing machine (Instron model 5542, Instron, Canton, MA, USA) to determine the mechanical properties of leaves. The machine simultaneously records force (N) applied to a sample and displacement (mm) (every 50 ms).

Punch-and-die test  This test was done following Onoda et al. (2008) and we only describe it briefly here. The punch and die were installed into the universal testing machine and the punch (diameter = 1.345 mm) was placed to go through the middle of the hole (0.2 mm clearance) of the die without any friction at the speed of 0.42 mm s−1. Tests were applied to sections of intercostal lamina (between secondary veins). Several mechanical traits were calculated from the force–displacement curve (see Fig. 1 in Onoda et al. (2008)). Maximum force per unit circumference of the punch was defined as force to punch (Fp), and Fp divided by lamina thickness was defined as specific force to punch (Fps). The area under the force–displacement curve was equivalent to work required to puncture the leaf. Work is expressed per unit circumference of the punch (work to punch) and per unit fracture-area that was calculated as circumference of the punch × lamina thickness (toughness).

Figure 1.

 Leaf lamina length (a), width (b) and length-to-width ratio of Plantago major plants exposed to different brushing and wind treatments. Bars indicate standard errors (n = 12). Open squares, no wind; closed squares, wind. MS, mechanical stress.

Tensile test  Intercostal lamina stripes (excluding primary veins, c. 0.4 × 4 cm in size) were cramped by pneumatically controlled grips and tensed at a constant speed of 0.42 mm s−1. We discarded measurements where the lamina ruptured close to either of the grips. Maximum force to tear per unit width of stripe (maximum force recorded before a drop in resistance indicating leaf rupture) was defined as force to tear (Ft), and Ft divided by lamina thickness was defined as specific force to tear (Fts) (Fts is the same as tensile strength in engineering; Read & Sanson, 2003). The modulus of elasticity (Elam), indicating the tissue stiffness, was also calculated, from the initial slope of the force–displacement curve.

Bending test  The bending test was performed on the petioles, applying the three-point bending test following Liu et al. (2007) with some modifications. Petioles were placed horizontally over two supports positioned 2.5 cm apart and a vertical force was applied midway between the two supports. For samples thicker than 2.5 mm, supports were placed further apart to keep the aspect ratio larger than 10. We determined the Young’s modulus (Epet) from the force–displacement (Fδ) curve as:

image(Eqn 1)

with L the span length between the supports and I the second moment of area. As the cross-sectional shape of petioles cannot be approximated by a simple geometrical shape (e.g. a circle or square), we examined the anatomy of petioles (see ‘Anatomical measurements’ below) and calculated I based on image analysis (see Notes S1).

After the mechanical measurements, from each leaf three lamina (1 × 2 mm) and two petiole sections were cut out for anatomical measurements and six leaf disks (0.78 cm2) were punched for fresh and dry mass and nitrogen determination.

Anatomical measurements

Petiole and lamina sections were immersed in a fixation buffer (2.5% glutaraldehyde, 100 mM phosphate buffer, pH = 7.0). They were dehydrated through a graded ethanol series (30, 50, 70, 80, 90, 100%) and infiltrated and embedded in LR white resin (London Resin Company, Reading, UK). The embedded samples were sliced with a microtome (OMU-3; Leica, Rijswijk, the Netherlands) and stained with 0.5% toluidine blue (lamina) and 0.25% safranine (petioles). Photographs were subsequently taken with a light microscope (AX-LH 100, Olympus Optical, Tokyo, Japan). On the images the thickness of the lamina, as well as of the palisade- and spongy parenchyma layers and the upper and lower epidermis, were measured using ImageJ vs 1.34s (National Institutes of Health, Bethesda, MD, USA). The same program was used to measure all dimensions of the petiole cross-section (see Notes S1).

Whole-plant biomass

On 5–6 March, all plants were harvested destructively. Plants were cut at ground level and separated into leaf laminas, petioles, flowering stems and roots. Lamina area was measured with a leaf area meter (LI3100; Li-Cor). Roots were carefully washed. The dry mass of all plant parts was determined after oven-drying for at least 72 h at 70°C.

Statistical analyses

The effects of wind (df = 1) and brushing (df = 1) on plant traits were tested with two-way ANOVA (total df for the corrected model is 47). Most variables were log-transformed to improve homogeneity of variance. ANCOVA was conducted with either petiole flexural stiffness or diameter as the dependent variable, petiole length as covariate and brushing and wind as fixed factors to test whether treatments affected the petiole allometry. Data analyses were done using SPSS 15 (SAS Institute, Cary, NC, USA).


Whole-plant responses

Both wind and MS (i.e. by brushing) strongly reduced the mass of all organs: leaf laminas, petioles, flowering stems and roots, and thus whole-plant biomass. MS resulted in a reduction in the fraction of mass in roots (RMF) and flowering stems (FlowerMF) and an increase in both the petiole (PMF) and lamina mass fractions (LMF, Tables 1, 2). Wind had similar effects on PMF, LMF and FlowerMF, but no effect on RMF. The total number of leaves produced was slightly smaller in the wind treatment than in the nonwind treatment, and not influenced by MS (Tables 1 and 2).

Table 1.   Mean values and standard errors of mean (SEM, n = 12) of whole-plant characteristics, leaf lamina size and dimensions and leaf physiological characteristics of Plantago major plants exposed to different brushing and wind treatments
ParameterNo windWind
No brushBrushNo brushBrush
  1. LMF, RMF, PMF and FlowerMF denote the leaf lamina, root, petiole and flower mass fractions, respectively; LMA and frLMA denote leaf lamina dry and fresh mass per area, respectively; water v/v, the lamina volume fraction containing water; Pmax, light-saturated photosynthesis; Rd, dark respiration; Narea and Nmass, the lamina nitrogen contents per unit area and mass, respectively.

Whole-plant characteristics
 Total mass (g)4.700.132.490.162.570.191.280.14
 Leaf (g)1.550.
 Roots (g)
 Petioles (g)0.370.
 Flowering stems (g)1.780.050.730.050.760.100.280.05
 Leaf area (cm2)325.16.5248.915.7242.114.3144.413.2
 Leaf number produces11.40.811.41.610.
Lamina size and dimensions
 Lamina area (cm2)28.50.621.40.924.41.316.41.0
 Leaf thickness (mm)
 Lamina DW (g)19.10.415.30.416.70.615.30.3
 Lamina FW (g)
 LMA (g m−2)40.60.933.00.535.
 fLMA (g m−2)206.44.6196.32.9220.53.5213.55.4
 Water v/v0.630.020.700.020.680.010.720.02
 Air fraction0.370.020.300.020.320.010.300.01
 Lamina length (cm)
 Lamina width (cm)5.560.114.680.
 Lamina length : width ratio2.290.042.670.
 Lamina toughness (kJ m−2)0.480.020.460.020.470.030.440.02
Leaf physiology
 Pmax (μmol m−2 s−1)14.40.517.20.518.20.515.60.7
 Rd (μmol m−2 s−1)
 Narea (g m−2)0.990.
 Nmass (mg g−1)2.450.073.800.063.590.114.200.08
 Transpiration (ml cm−2)
 Leaf temperature (°C)
Table 2.   Results of analysis of variance (ANOVA) with mechanical stress (MS) and wind as fixed factors
 BrushWindBrush × windTransformation
  1. LMF, RMF and PMF denote the leaf lamina, root and petiole mass fractions, respectively; LMA and frLMA denote leaf lamina dry and fresh mass per area, respectively; water v/v, the lamina volume fraction containing water; Pmax, light-saturated photosynthesis; Rd, dark respiration; Narea and Nmass, the lamina nitrogen contents per unit area and mass, respectively.

  2. There were 12 replicates per treatment combination (n = 12) for a total 48 plants in the experiment.

  3. ns, not significant (P > 0.05); $, marginally significant (0.01 < P < 0.05); *0.001 < P < 0.01; **P < 0.001. (−) and (+) indicate positive and negative effects, respectively.

Whole-plant characteristics
 Total mass (g)(−)**(−)**nsLog
 Leaf (g)(−)**(−)**nsNormal
 Roots (g)(−)**(−)**nsLog
 Petioles (g)(−)**(−)**nsLog
 Flower stems (g)(−)**(−)**nsLog
 Leaf area (cm2)(−)**(−)**nsLog
 Leaf number producedns(−)*nsNormal
Lamina size area and dimensions
 Lamina area (cm2)(−)**(−)**nsLog
 Leaf thickness (mm)(−)**(+)*nsNormal
 Lamina DW (g)(−)**(−)*$Log
 Lamina FW (g)(−)*(+)*nsLog
 LMA (g cm−2)(−)**(−)**Log
 frLMA(g cm−2) (−)$(+)**nsLog
 Water v/v(+)*$(+)nsNormal
 Air fraction(−)*nsnsNormal
 Lamina length (cm)ns(−)**nsLog
 Lamina width (cm)(−)**nsnsLog
 Lamina length : width ratio(+)**(−)*$Log
Leaf lamina mechanical traits
 Punch test
  Force-to-punch (kN m−1)(−)*nsnsLog
  Specific force-to-punch (MPa)nsnsnsLog
  Work-to-punch (J m−1)(−)*nsnsLog
  Toughness (KJ m−2)nsnsnsNormal
 Tensile test
  Force-to-tear (kN m−1)(−)*nsnsLog
  Specific force-to-tear (MPa)nsnsnsLog
  Lamina Young’s modulus (MPa)ns(−)*nsNormal
Lamina anatomy
 Upper epidermis (mm)nsnsnsNormal
 Palisade (mm)ns(+)$nsNormal
 Spongy (mm) (−)$nsnsLog
 Lower epidermis (mm)(−)*(−)$nsLog
Petiole characteristics
 Length (cm)(+)$(−)*nsLog
 Diameter (mm)(−)**(−)**nsLog
 Slenderness (length/diameter)(+)**(−)$**Normal
 Thickness (mm)nsnsnsNormal
 Second moment of area (mm4)(−)$(−)$nsNormal
 Petiole Young’s modulus (MPa)nsnsnsNormal
 Flexural stiffness (N mm2)(−)**(−)$nsLog
Leaf physiology
 Pmax (μmol m−2 s−1)nsns**Normal
 Rd (μmol m−2 s−1)nsnsnsNormal
 Narea (g m−2)(+)**(+)****Log
 Nmass (mg g−1)(+)**(+)****Log
 Leaf temperature (°C)ns(−)**nsNormal

Lamina characteristics

Both MS and wind exposure resulted in a reduction in mean lamina dry mass and lamina area as well as the ratio between the two (leaf mass per area, LMA) of fully expanded young leaves (fourth youngest leaves; Tables 1, 2). Fresh mass and the ratio of fresh mass to area (frLMA) increased under wind exposure but were reduced by MS. The opposite effects of wind on LMA and frLMA were reflected in the water volume fraction, which tended to be greater in wind-exposed plants (Tables 1, 2). Lamina shape was also differentially affected by wind and MS. The laminas of mechanically stressed plants tended to be narrower but equally as long as those of control plants and thus had a more elliptic shape, while wind-exposed leaf laminas tended to be shorter but equally wide, resulting in a more rounded shape (Fig. 1).

Leaf lamina thickness was slightly but significantly larger in wind-exposed plants and smaller in MS plants than in control plants (Tables 1, 2; Fig. S2). The effect of wind tended to be the result of a marginally significant increase in thicknesses of the palisade layers (P = 0.02). The effect of MS was mostly through a marginally significantly thinner spongy layer and a thinner lower epidermis (Fig. S2).

Force to punch (Fp), which reflects the resistance of the leaf lamina to puncture, was smaller for laminas of MS plants but was not affected by wind (Fig. 2a; Table 2). The specific punch strength (Fps), which indicates the material strength of the leaf, was unaffected by either treatment (Fig. 2b). Work-to-punch (the energy required to puncture a leaf per unit circumference) was reduced by MS but unaffected by wind (Fig. 2c), while punch toughness (work-to-punch divided by lamina thickness) was unaffected by either treatment (Tables 1, 2). Tensile tests revealed somewhat similar results. The force to tear (Ft), reflecting the overall resistance of the lamina to tear per unit width, was smaller in MS laminas than in unstressed ones while wind had no effect (Fig. 2d). The specific force to tear (Fts), indicating material strength, was not significantly affected by either treatment (Fig. 2e). Overall these results indicate that brushed laminas were mechanically less resistant to fracture because they were thinner but not because their tissues were weaker. Finally the tensile Young’s modulus of leaf laminas was not significantly affected by either treatment (Fig. 2f).

Figure 2.

 Mechanical properties of leaf laminas of Plantago major plants exposed to different brushing and wind treatments: force-to-punch (a), specific force-to-punch (b), work-to-punch (c), force to tear (d), specific force to tear (e) and apparent Young’s modulus measured in tension (f). Bars indicate standard errors (n = 12). Open squares, no wind; closed squares, wind. MS, mechanical stress.

We roughly estimated the second moment of area (I) of leaf laminas, by assuming the lamina cross-section to be a rectangle (see Fig. 3.3 in Niklas, 1992), and found it to be 40–50% smaller in MS leaves than in unstressed laminas (data not shown). The flexural stiffness (EI), which measures the degree of resistance to bending, is the product of I and the Young’s modulus, E. As E was not affected by MS, the leaf laminas of the MS plants had a considerably lower EI than those of nonMS plants and were thus more flexible, enabling them to bend more easily under mechanical stress. Wind, on the other hand, had no effect on lamina EI.

Both wind and MS resulted in leaves having greater N contents per unit mass and per unit leaf area. However, light-saturated photosynthesis and respiration were not significantly affected by either treatment (Tables 1, 2). Transpiration rates were 50–100% higher in wind-exposed plants than in nonexposed ones, while MS had no effect. Leaf temperatures were c. 0.7°C lower in wind-exposed plants than in nonexposed ones (Tables 1, 2).

Petiole characteristics

Petiole length tended to be longer in MS plants and was shorter in wind-exposed plants (Fig. 3a, Table 2). Petiole diameter was reduced by both treatments, and these reductions resulted in smaller second moments of area (I), which is the geometrical contribution to flexural stiffness (Notes S1, Fig. 3b). MS thus resulted in considerably more slender petioles (i.e. with a greater length-to-diameter ratio). Among MS plants, slenderness was lower in wind-exposed than in non-exposed plants, but among the nonMS plants there was no difference (Fig. 3d; Table 2). Since petiole length and diameter are generally positively correlated, we conducted an ANCOVA, with petiole length as a covariate and diameter as an independent variable (both parameters log-transformed). This analysis indicated that this relationship was not affected by wind (P = 0.356) but that it shifted downwards, that is, thinner petioles for a given petiole length, with the MS treatment (< 0.0001). The Young’s modulus (Epet) was not affected by either treatment and as a result the flexural stiffness (the product of Epet and I) was reduced by both treatments (Fig. 3c,e). As with petiole diameter, ANCOVA showed that the allometric relationship between petiole length and EpetI was shifted downwards by MS (< 0.0001) but not by wind (P = 0.223).

Figure 3.

 Petiole properties of Plantago major plants exposed to different brushing and wind treatments: petiole length (a), petiole diameter at the base (b), apparent Young’s modulus measured in bending (c), petiole slenderness (length-to-diameter ratio) (d) and flexural stiffness (e). Bars indicate standard errors (n = 12). Open squares, no wind; closed squares, wind. MS, mechanical stress.


Wind and brushing may evoke different responses in Plantago major

The effects of wind on plants are complicated as they entail mechanical stimulation and changes in microclimate. Here we show that, with respect to a number of traits, effects of MS and wind were different and, in some cases, even in the opposite direction. For example, MS induced the formation of leaves with longer, more slender petioles and more elongated leaf blades with thinner laminas, whereas wind tended to have the opposite effect. In the field, responses of plants to wind will thus depend on the relative importance of air flow and mechanical stress effects (Smith & Ennos, 2003) which in turn depend on the overall environmental conditions as well as the characteristics of the plants themselves. These factors include humidity, the magnitude, frequency and duration of wind loading, leaf shape and size, and the overall shape and drag coefficient of the vegetation stand in which a plant is growing (Smith & Ennos, 2003; Speck, 2003). This could explain the variable effects of wind that have been found (Lawton, 1982; Henry & Thomas, 2002). In the field, responses of plants to wind and other forms of MS (e.g. brushing by animals or neighbouring plants) should also be expected to be different. While many studies that analyse the effects of flexing or brushing on plant growth and allocation implicitly extrapolate their results to wind effects (e.g. Niklas, 1998; Anten et al., 2005), our results indicate that this may not always be correct.

Microclimatic and mechanical effects on leaf traits

The differential effects of wind and MS could be mostly associated with the fact that wind entails both microclimatic effects in addition to mechanical disturbance. Regarding the effects of air flow, we observed a significant increase in transpiration and a concomitant small reduction in leaf temperature (−0.7°C) in wind-exposed plants. The latter effect was probably minor, but responses in a number of plant traits to wind might be attributed to desiccation stress. Leaves of wind-exposed plants had thicker laminas and tended to have higher water content than those of control plants (Table 1; Cordero 1999), which may be associated with water-saving strategies, while brushed plants showed the opposite response. Wind-exposed plants also had shorter petioles and rounder (relatively shorter and wider) leaf blades. Shorter leaves have a shorter water pathway and have smaller hydraulic resistance, and may therefore be less prone to embolisms (Sack et al., 2002), although shorter leaves are also more resistant to deformation. Wider leaf blades tend to have a larger boundary layer resistance, which reduces transpiration (Lambers et al., 1998). At the molecular level it was shown that, in poplar, wind evoked expression of genes associated with plant responses to limit water loss (Fluch et al., 2008). Together these results suggest that wind-induced phenotypic changes in at least some leaf traits were associated with preventing dehydration.

Plants can prevent mechanical failure under external forces (e.g. wind or water flow) by producing strong structures that resist large forces or by producing flexible structures that deflect and thus reduce the impact of forces (Wainwright et al., 1976; Niklas, 1996; Read & Stokes, 2006). Both wind exposure and MS reduced leaf size and flexural stiffness of petioles, which reduced mechanical loads and increased leaf flexibility (Read & Stokes, 2006). However, only MS plants had lower lamina flexural stiffness and more slender leaves and petioles. These characteristics further facilitate deformation under mechanical loads. Slender leaves are often found in plants growing along rivers, where they are periodically subjected to strong hydrodynamic forces (e.g. Lytle & Poff, 2004; Nomura et al., 2006).

Our finding that wind and MS caused a reduction in the flexural rigidity of petioles is consistent with Niklas (1996), who found similar differences between wind-exposed and sheltered Acer leaves, but runs contrary to the other findings that both wind and mechanical stimulation seem to induce the production of shorter but also thicker stems (e.g. Biro et al., 1980; Telewski, 1990; Anten et al., 2005, 2009) or tree branches (reaction wood; Jaffe 1973). These different responses to mechanical stress could be associated with the mechanical roles of these plant structures. For example, petioles or distal branches support only individual leaf laminas or a relatively small part of the crown, respectively. A reduction in EI increases flexibility and thus helps their ability to reconfigure under wind loading (Vogel, 1994). Stems or primary tree branches, on the other hand, need to maintain a whole structure, and as the plant grows, they tend to support increasingly heavy loads. A reduction in flexural rigidity could thus make them prone to global buckling (Niklas, 1992). Increases in diameter and rigidity, which are considered to be typical thigmo-responses, may be found in these cases. Our results suggest that plants can prevent mechanical failure by two contrasting strategies depending on the mechanical role of organs.

The relative mechanical and microclimatic effect of wind may depend on the growth form of plants. In rosette plants, such as P. major and Arabidopsis thaliana (a model system in the research on the physiological basis of MS and wind responses; see Braam, 2005), the mechanical effect may be relatively smaller than in erect plants. In Helianthus annuus, a tall erect annual, responses to wind seemed to be associated with increasing mechanical stability rather than securing water transport (Smith & Ennos, 2003), while our results suggest a response that is more associated with water conservation. Evidently more research on plants with different growth forms is needed.

It has been proposed that mechanical stress can increase the resistance of leaf lamina tissue to be torn or punctured and thus reduce susceptibility to herbivory of plants (Cipollini, 1997), but to the best of our knowledge the effects of mechanical stress and wind force on leaf lamina mechanical properties have not been measured. Contrary to this proposition we found that lamina force to punch and force to tear were reduced by MS. This reduction was mainly the result of MS leaves having thinner laminas, rather than being caused by changes in material strength of the lamina tissue. The thinner lamina, as mentioned earlier, was more flexible, which had the advantage of avoiding damage, but which, at the same time, resulted in reduced structural strength. In the presence of large herbivores, reduced mechanical strength of leaves might also be advantageous, as the leaves would break before the whole plant is uprooted.

Whole-plant growth, allocation and size

Both wind and MS reduced biomass increment, a result that is consistent with various other studies on the growth effects of wind (Retuerto & Woodward, 1992; McArthur et al., 2010) and pure mechanical stress (Niklas, 1998; Cipollini, 1999), although some studies show no growth reduction (Anten et al., 2005). Wind may also negatively affect whole-plant carbon gain through its effect on stomatal conductance, leaf temperature and by reducing the effective leaf area by causing leaves to roll up (Telewski, 1995; Ennos, 1997). We detected neither a reduction in stomatal conductance or leaf photosynthetic capacity nor a large decrease in leaf temperature, suggesting that the former two factors may not have played a big role. We did, however, clearly observe leaves being folded or rolled up in the wind treatment, and therefore the continuous wind condition should have had a strong negative effect.

Negative effects of mechanical stress on growth have been attributed to the existence of an internal resource allocation trade-off; investment of resources allocated to support structures (e.g. stems, branches or petioles) to maintain mechanical stability cannot simultaneously be allocated to resource (light) harvesting structures (Goodman & Ennos, 1996; Cipollini, 1999; Selaya et al., 2007). However, this trade-off may be less straightforward than it seems. As observed here, plants may respond to mechanical stress by producing thinner, more flexible support structures and this does not necessarily entail a greater biomass investment to structural tissue. In the current study, we found that MS plants had larger lamina mass fractions and smaller petiole mass fractions (Table 1).

Allocation of dry mass between organs might also be influenced by plant size; for example, small, less developed plants tend to allocate less to support structures. This could partly explain the larger leaf mass fraction observed in the MS and wind-treated plants. Plant size potentially influenced other plant traits as well. However, MS and wind effects on most of the characteristics observed in this study were very different and often in the opposite direction, even though plant mass was similarly reduced by both treatments. In addition, the number of leaves produced was unaffected by MS and only slightly reduced by wind, suggesting that plant development was not so different. The key result of this study – different and opposite effects of MS and wind – did not therefore result from differences in plant size or developmental stage.

Concluding remarks

This study suggests that plant responses to pure MS (brushing in this case) and wind (combination of air flow and mechanical stress) can be different, or even be in the opposite direction. At the leaf level, the overall responses to MS resulted in increased flexibility and could be associated with avoidance of mechanical stress. Responses to wind, on the other hand, seemed to be more associated with reducing transpiration. These results exemplify the complexity of understanding phenotypic plasticity in plants in response to changes in environmental factors, as these factors themselves may have multiple effects.


We thank Heinjo During and Thijs Pons for valuable comments on the manuscript, Yuko Yasumura for help with biomechanical measurements and Fred Siesling and Betty Verduyn for technical support. This study was partly supported by Grant-in-Aid from JSPS for Young Research Fellows (Y.O.) and an ERASMUS exchange fellowship to R.A.H.