Mechanical stress experiment
The experiment was carried out in a greenhouse of Utrecht University, in The Netherlands. On 3 November 2003, seeds were sown in trays in a mixture of sand and potting soil and grown at 30% of natural daylight created by neutral density shade cloth and shading by the greenhouse roof. On 3 December, seedlings were selected for equal size, both within and between genotype, and transplanted into pots (volume 6 L) on a mixture of sand and sieved potting soil (half of each) and without shade cloth at about 60% of daylight. Additional lighting was supplied with HPI Quick 400-W lamps for 16 h, and the mean daytime photon flux density during the experiment was 187 umol m−2 s−1, measured with a quantum sensor (LI190SA, LiCor, Lincoln NE) connected to a datalogger (LI1000). At this time, 3 g of slow-release fertilizer (Osmocote 10% N + 10% P + 10% K + 3% Mg + trace elements, release time 3 months) was added to each pot. Plants were watered daily throughout the experiment.
On 17 December, plant height from soil level to the highest leaf node, stem diameter just above the first true leaf, leaf number and the length of the biggest leaf were measured. For the experiment, we then chose 48 plants of each genotype, which were of intermediate height: the shortest and the tallest plants were excluded from the experiment. On the next day, plants were assigned to one of three mechanical disturbance treatments: 0, 25 or 75 flexures each day (the control, m25 and m75 treatments, hereafter) or to a separate group used for the initial harvest (see below). There were 12 replicate plants per treatment genotype combination, and these were placed in four blocks.
The flexing treatment was carried out by gently grasping the stem at about four-fifths of its height and bending it back and forth no further than 45° from the vertical. Flexing lasted about 30 s for the m25 treatment and 90 s for the m75 treatment. We chose this type of flexing because it simulates the mechanical effect of wind on plants (compressive and tensile forces on stems and roots) without affecting their microclimate.
The first harvest was conducted on 18 December, before initiation of treatments, to determine the baseline biomass distribution at the onset of the experiment. The 12 selected plants from each genotype were cut at ground level and divided into stems and leaves, and the fresh weights of both parts were determined. Leaf area was determined with a leaf area meter (LI 3100). Root systems were carefully washed. Dry weights of all plant parts were determined after oven-drying for at least 72 h at 70°C to constant weight.
The final harvest was conducted on 13 January 2004, 26 days after the first harvest, to determine growth rates and patterns of biomass allocation. Plants were cut at ground level and their height was measured from the base to the top meristem. Basal diameter and the diameters at one-third and two-thirds of stem length were measured with a digital caliper to the closest 0.1 mm. To determine the vertical distribution of fresh leaf and stem mass, plants were divided into three equal-height segments. Leaves of each segment were removed with a razor blade and weighed, and their area was determined as described above. After the mechanical measurements on stems were completed (see below), they were clipped into the three height segments and also weighed. Dry masses of each clipping segment, stems and leaves, and root mass were measured as described above.
Growth per unit mass (relative growth rate, RGR), growth per unit leaf area (net assimilation rate, NAR) and the mean leaf area per unit plant mass (mean leaf area ratio, LAR) were determined following classic growth analysis, following Equations 3.3, 3.6 and 3.8 from Beadle (1993). Leaf losses during the experimental period were minimal and were not considered (see Anten and Ackerly 2001). Plants from the two harvests were randomly paired.
The mechanical stress experiment was independently repeated between October and December 2005. The observed responses were very similar to those found for the first experiment (supplementary material), although at the final harvest plants were smaller than in the first experiment, as they had been harvested at an earlier stage (16 instead of 26 days after the initial harvest, which was 61 instead of 71 days after sowing). The data presented in Results, therefore, refer to the experiment described in the above paragraphs.
Two material properties were measured on the stems: the elastic Young’s modulus (E, N m−2) and the breaking stress (σb, N m−2), following Anten et al. (2005). E indicates the rigidity and σb the resistance to breaking of a given tissue. Both measurements were completed no longer than 15 min after a stem had been cut from its pot and its leaves removed. Stems were fixed at both ends between small clamps that were coated with a layer of rubber foam to prevent the tissue from being crushed. To measure E, we placed an empty water bottle exactly halfway along the stem. Small loads of water (P) were weighed and then added to the bottle, and the vertical displacement (δ) was measured to the nearest millimetre. The advantage of this arrangement is that the force is held perpendicular to the stem even as the stem bends. The Young’s modulus was calculated from the linear regression of δ against P from the equation of a fixed end beam (Equations 10–19 in Gere and Timoshenko 1999):
where L is the length of the stem segment and I the second moment of area (I = 1/4πR4, where R is the stem radius). The breaking stress (σb) was measured by increasing the loads in small steps to the point where the stem failed and was calculated as:
where M is the bending moment and P, in this case, is the load that was supported just before failure occurred (Gere and Timoshenko 1999).
WT and Tetr plants were grown in a growth chamber in pots with 16 h of light at 220 umol m−2 s−1(Phillips HDS 600W, Eindhoven, The Netherlands), a temperature of 20°C, and 70% relative humidity. After 5 weeks, 48 plants were transplanted to closed chambers (volume 35 dm3), which were flushed with either of two ethylene concentrations (0 or 1 ul l−1, Praxair; Oevel Belgium) in air (see Pierik et al. 2003 for details). There were 12 replicate plants for each treatment–genotype combination. Plants were harvested after 7 days, and their shoot dry mass was determined as described for the mechanical stress experiment.