Not only light quality but also mechanical stimuli are involved in height convergence in crowded Chenopodium album stands

Authors

  • Hisae Nagashima,

    1. Nikko Botanical Garden, Graduate School of Science, The University of Tokyo, 321-1435 Tochigi, Japan
    2. Graduate School of Life Sciences, Tohoku University, Aoba, Sendai, 980-8578 Miyagi, Japan
    Search for more papers by this author
  • Kouki Hikosaka

    1. Graduate School of Life Sciences, Tohoku University, Aoba, Sendai, 980-8578 Miyagi, Japan
    2. CREST, Japan Science and Technology Agency (JST), 102-0076 Tokyo, Japan
    Search for more papers by this author

Author for correspondence:
Hisae Nagashima
Tel: +81 022 795 7732
Email: hisae@ceres.ocn.ne.jp

Summary

  • In crowded stands, height is often similar among dominant plants, as plants adjust their height to that of their neighbours (height convergence). We investigated which of the factors, light quality, light quantity and mechanical stimuli, is primarily responsible for stem elongation and height convergence in crowded stands.
  • We established stands of potted Chenopodium album plants. In one stand, target plants were surrounded by artificial plants that were painted black to ensure that the light quality was not modified by their neighbours. In a second stand, target plants were surrounded by real plants. In both stands, one-half of the target plants were anchored to stakes to prevent flexing by wind. The target plants were lifted or lowered by 10 cm to test whether height convergence was affected by the different treatments.
  • Stem length was affected by being surrounded by artificial plants, anchoring and pot elevation, indicating that light quality, light quantity and mechanical stimuli all influenced stem elongation. Height convergence did not occur in the stand with artificial plants or in anchored plants.
  • We conclude that light quality and mechanical stimuli are important factors for the regulation of stem growth and height convergence in crowded stands.

Introduction

Stem growth is an important determinant in the competition for light amongst crowded plants (Schmitt et al., 1995; Dudley & Schmitt, 1996; Huber & Stuefer, 1997; Huber et al., 1998; Weinig, 2000). Taller plants can expose a larger portion of their leaf area to sunlight and shade out competitors (Anten & Hirose, 1998; Hikosaka et al., 1999). However, with increasing plant height, the risk of lodging caused by wind increases (Casal et al., 1994) and biomass allocation to the leaves (Givnish, 1982) and roots (Maliakal et al., 1999) is reduced, which may decrease total resource acquisition. Being tall can therefore be both beneficial and detrimental to plant growth.

Plants regulate stem growth depending on their environment. In crowded stands, accelerated stem elongation has been observed (Schmitt et al., 1987; Geber, 1989; Weiner et al., 1990; Weiner & Thomas, 1992; Weiner & Fishman, 1994; Nagashima, 1999). Despite this, most plants avoid overtopping their neighbours. Dominant plants, which expose their leaves to the canopy surface, have similar heights to their neighbours despite exhibiting a great variety in stem diameter and shoot mass. This phenomenon is termed ‘height convergence’ (Weiner & Thomas, 1992; Weiner & Fishman, 1994; Nagashima & Terashima, 1995). Nagashima & Hikosaka (2011) conducted an experiment with a stand of potted Chenopodium album plants, in which some of the plants were lifted and lowered to be taller and shorter, respectively, than neighbouring plants. The lifted plants exhibited a decreased stem elongation rate, whereas the lowered plants exhibited accelerated stem elongation. These results indicate that height convergence is a consequence of the regulation of stem growth, so as to maintain a similar height to neighbouring plants.

Several environmental factors are known to influence stem growth. Of primary importance is light quality, which acts as an environmental signal for the onset of growth competition (Ballaré, 1999; Smith, 2000; Pierik et al., 2004). Because of the preferential absorption of red light by chlorophyll, light scattered by neighbouring plants has a reduced red to far-red (R/FR) ratio. Phytochrome photoreceptors in plant stems detect this signal, encouraging the plants to accelerate stem elongation (Morgan & Smith, 1976, 1978, 1979). The quantity of light induces both positive and negative effects on stem elongation. As photosynthates are necessary for stem growth, high light levels may increase stem elongation. Alternatively, several studies have shown that high light levels can retard stem elongation and that moderate shade levels are better at stimulating elongation (e.g. Grime & Jeffrey, 1965; Lecharny & Jacques, 1980; Holmes et al., 1982; Corré, 1983). Exposure to strong winds may reduce elongation of overtopping plants, probably through the mechanical effects of flexing and/or through excess transpiration from the leaves (e.g. Latimer et al., 1986; Holbrook & Putz, 1989; Retuerto & Woodward, 1992; Henry & Thomas, 2002). Plants in field stands may shield each other from wind, but, when a plant grows above its neighbours, it may be more exposed to wind. It has been shown that plants also respond to mechanical stress, such as touching and rubbing, through the process of thigmomorphogenesis (Jaffe, 1973). Manual flexing of stems lowers stem elongation rates (e.g. Smith & Ennos, 2003; Anten et al., 2005, 2006, 2009). These studies indicate that mechanical stress is an important factor in the regulation of stem growth.

Previous studies have used inventive methods to prove that phytochromes are involved in height regulation in plant stands. Ballaréet al. (1990) showed that plants did not grow to enhanced heights when covered with a far-red light filter, even when they became shaded by their neighbours. Transgenic plants, in which elongation in response to neighbours was blocked, exhibited decreased relative fitness when grown in competition with wild plants (Schmitt et al., 1995). However, no published work has indicated that plants regulate growth to a certain height solely in response to light quality. It is feasible that other factors, such as light quantity or physical stimuli, or both, can influence height growth in crowded stands, possibly in combination with various other environmental factors.

In the present study, experiments were undertaken to determine whether light quality, light quantity and mechanical stimuli are involved in height growth regulation and height convergence in field stands. Two stands of potted Chenopodium album plants were established. In one of the stands, the target plants were surrounded by living plants, whereas, in the other stand, they were surrounded by artificial plants that were painted black. It has been shown previously that, in a stand containing live plants, both the R/FR ratio and light quantity decrease with depth into the stand (Holmes & Smith, 1977; Ballaréet al., 1987; Gilbert et al., 1995), whereas, in a stand of artificial black plants, only light quantity decreases with depth with no change in the R/FR ratio. If a decrease in the R/FR ratio leads to an increase in stem growth, elongation would be expected to be greater in the stand containing live plants when compared with that in the stand containing artificial black plants. Selected target plants were anchored to stakes to minimize the effect of flexing caused by wind. If mechanical stimuli reduce stem growth in the stands, anchored plants should exhibit increased stem elongation. If excess transpiration has a significant role in inhibiting elongation, the anchoring of plants should not increase stem elongation. In addition, selected pots of some target plants (including both anchored and nonanchored plants) were lifted, lowered or left unchanged to examine which environmental factor is primarily responsible for height convergence (Nagashima & Hikosaka, 2011). If height convergence is primarily caused by a change in light quality, height convergence should not be observed in the stand containing artificial black painted plants. If height convergence is primarily caused by mechanical stimuli, height convergence should not be observed in anchored plants. If height convergence is primarily caused by light quantity, height convergence should be observed in both stands containing live plants and containing artificial black plants. The experiments also measured the effects of light quality, light quantity and mechanical stimuli on plant architecture and biomass allocation. An earlier study has shown that accelerated elongation is associated with a reduction in diameter growth and biomass allocation to leaves and roots (Nagashima & Hikosaka, 2011). We address the question of which environmental factor is primarily responsible for the changes of growth.

Materials and Methods

Plant materials

Chenopodium album L. (Chenopodiaceae) is a broad-leaved summer annual that often colonizes disturbed habitats (Ohwi & Kitagawa, 1983; Grime et al., 1988) and shows great plasticity in stem growth in response to environmental conditions (Morgan & Smith, 1979).

Experiment

The experiment was conducted in an experimental plot within Nikko Botanical Gardens, Nikko, Japan (36°75′N, 139°59′E) in 2003. Mean monthly air temperatures during the experiment were 16.7, 20.0 and 21.4°C for the experimental months of June, July and August, respectively. Vinyl pots (= 515), each 10.5 cm in diameter and 22.5 cm in height, were filled with sand and tightly arranged on a bench (width, 1.1 m; length, 5 m; height, 0.25 m), which was then placed outdoors. Seeds were obtained from plants used in previous experiments (Nagashima & Hikosaka, 2011). Following cold stratification, c. 10 seeds were sown per pot on 19 May. The pots were watered once or twice per day as required. Seedling emergence was observed on 23 May. Seedlings were thinned to leave two or three per pot by the end of May and one per pot by 10 June, in order to reduce size differences among seedlings. The final plant density was 115.5 plants m−2. Plants were fertilized with 20 ml of fivefold strength Hoagland’s complete nutrient solution (21 mg N, 3.1 mg P, 23 mg K, 16 mg Ca, 4.9 mg Mg, 6.4 mg S, 0.16 mg Fe per pot per week; Hoagland & Arnon, 1950) every week commencing on 2 June.

On 18 July, when the measured leaf area index (LAI) was 1.9, the heights of all potted plants were measured, and 84 plants whose heights were around the mean were selected as target plants (stem length, 38.8 ± 0.6 cm; stem basal diameter, 4.1 ± 0.2 mm). They were randomly divided into two groups (Fig. 1). One group of plants was grown surrounded continuously by other live plants. The other group was grown surrounded by artificial plastic plants that had a similar shape and size to real plants, but were painted black (Supporting Information Fig. S1). The artificial plants were individually placed into sand-filled pots and packed to establish a stand. Half of the target plants in each stand had their stems anchored to a stake with wire at intervals of 10 cm to eliminate the effect of sway by wind. The nonanchored target plants were touched with fingers so that any mechanical stimulus they received was similar to that received by the anchored plants.

Figure 1.

Experimental design: (a) treatments and (b) pot arrangements. Half of the target plants were surrounded by Chenopodium album plants and the other half by artificial plants that were painted black (light quality treatment). Half of the target plants were anchored to stakes and the other half were not anchored (stem anchoring treatment). Target plants were lifted (Li), lowered (Lo) or remained at the same level as surrounding plants (U) (pot elevation treatment).

In a further treatment, selected target plant pots were lifted by 10 cm (Li, lifted), lowered by 10 cm (Lo, lowered) or left at the same level (U, unchanged) as neighbouring pots. Seven pots were assigned to each combination of treatments. The height of the plastic plants was adjusted to the height of the nonanchored, unchanged target plants every 3 d during the course of the experiment. The portions of anchored plant stems that became elongated were also anchored to the stakes at intervals of 5 cm as necessary. Fingers were used to provide similar mechanical stimuli to the other target plants. The pots of target plants were randomized within each combination of treatments 1 wk after the start of the treatments. Two weeks after the start of the treatments, when LAI was 3.2, target plants were harvested and separated into various organs (leaf, stem and root). Stem length was measured (to the nearest 1 mm) from the base to the terminal shoot apex. Stem basal diameter was measured with callipers (to the nearest 0.1 mm) at the middle of the first internode in directions orthogonal to each other and averaged. Images of leaves and a scale for measurement were recorded with a digital camera (Coolpix 995, Nikon, Tokyo, Japan). Leaf area was measured using image analysis software (NIH Image v. 1.63, National Institutes of Health, Bethesda, MD, USA). Dry mass was determined following oven drying at 80°C for 3 d.

Determination of microenvironmental variables

The vertical distribution of photosynthetically active radiation (PAR) in the stands was measured with a quantum sensor (LI-190, LI-COR Inc., Lincoln, NE, USA) on 26 July, when weather conditions were overcast. Measurements were made at height intervals of 10 cm near each of seven nonanchored, unchanged target plants (seven replicates). Three measurements were taken at each height interval and were averaged to represent PAR at the height position of the plant. PAR was expressed as a relative value against a reference PAR measured above the canopy with another LI-190 sensor. The red (655–665 nm) to far-red (725–735 nm) (R/FR) ratio of light coming from a horizontal direction was measured with a portable spectroradiometer (LI-1800, LI-COR Inc.) at height intervals of 10 cm near each of five randomly chosen target plants after carefully removing the target plants (five replicates). Measurements were made between 10:00 and 14:00 h on 27 July under clear sky conditions. Measurements were taken for each height with the sensor facing the horizon in the direction of the four cardinal azimuths, and these four measurements per height were averaged. Wind speed in the stands was measured with a portable wind meter (Kestrel 4000, Nielsen–Kellerman, Boothwyn, PA, USA) on 25 July, when the ambient average and maximum wind speeds were 0.4 and 2.3 m s−1, respectively. Speeds were measured at three points at 10-cm intervals near nonanchored, unchanged target plants in each stand (seven replicates). A reference wind speed was measured above the canopy with another portable wind meter (Kestrel 4000, Nielsen–Kellerman). The daily average and maximum wind speeds during the experiment were 1.2 ± 0.5 and 3.1 ± 1.4 m s−1 (mean ± standard deviation), respectively, at the nearest weather station (Imaichi, 9 km from the garden).

Data analyses

The effects of the black artificial plants on the vertical distribution of PAR, R/FR ratio and wind speed within the canopy were analysed using a two-way repeated measures analysis of variance (ANOVA) with plants as replicates after confirmation of parametric assumptions (JMP Statistical Software, SAS Institute Inc., Cary, NC, USA). The effects of painting artificial plants black, stem anchoring, pot elevation and their interactions on stem elongation, stem diameter growth and dry mass partitioning were analysed by a three-way ANOVA after confirmation of parametric assumptions. The Tukey–Kramer honestly significant difference test was used for post hoc pairwise comparisons. The significance of height convergence in target plants was tested as follows: the mean value of the stem length of nonanchored, unchanged plants in each stand (which was regarded as a neighbour’s height in each stand) was subtracted from the apparent heights of target plants. The apparent height is the stem length + 10 and − 10 cm for lifted and lowered plants, respectively. A normal distribution was assumed for the apparent height difference and the probability of stem elongation being ≥ 10 and ≤ − 10 cm was calculated for lifted and lowered plants, respectively. If the probability was < 5%, the target plants were regarded as height converged.

Results

Microenvironments in the stands

PAR decreased with depth in both the stand containing live plants and that containing artificial black plants. The reduction in PAR was slightly smaller in the stand containing artificial black plants (< 0.001, < 0.001 and 0.253 for the effects of height, black coloration and their interaction, respectively; Fig. 2a). The stand containing live plants showed a significant reduction in the R/FR ratio with increasing depth in the canopy, whereas the stand containing artificial black plants showed only a slight reduction (< 0.001 for all three effects; Fig. 2b). Wind speed decreased with increasing depth in both stands, and the profile was not significantly different between the stands (< 0.001, 0.558 and 0.508 for the effects of height, black coloration and their interaction, respectively; Fig. 2c).

Figure 2.

Microenvironments in the real plant stand (open circles) and the stand including artificial plants painted black (closed circles): (a) photosynthetically active radiation (PAR) relative to that c. 30 cm above the canopy; (b) red to far-red (R/FR) photon ratio in horizontally incident light; (c) wind speed in the stand relative to that c. 30 cm above the canopy. Error bars represent ± SE.

Effects of environmental factors on stem elongation

Stem elongation was affected significantly by all treatments: a large part of the variation was explained by black coloration, followed by pot elevation and stem anchoring (Tables 1, S1). The interaction effect was only significant for the interaction of black coloration and pot elevation. Stem length was greater in the stand containing live plants than in the stand containing artificial black plants for each combination of anchoring and pot elevation, indicating that reduced R/FR ratio positively influenced stem elongation (Fig. 3a). Anchored plants exhibited increased stem length irrespective of the type of surrounding plants and pot elevation (Fig. 3a), indicating that reducing the mechanical stimuli also had a significant positive effect on stem elongation. The effect of anchoring had no interactive effects with other treatments, suggesting that it influences stem elongation independently and additively to other factors (Table 1). Lifted plants exhibited reduced elongation and lowered plants exhibited increased elongation in the stand containing live plants, as observed in a previous study (Nagashima & Hikosaka, 2011). The response of stem length to pot elevation was insignificant in the stand containing artificial black plants (Fig. 3a). These results suggest that the effects of light quantity on elongation are small compared with those of light quality.

Table 1.  Three-way ANOVAs for measures with artificial black plants, stem anchoring and pot elevation as factors (see Fig. 1); F values are shown
 Artificial black plantsStem anchoringPot elevationArtificial black plants × stem anchoringArtificial black plants × pot elevationStem anchoring × pot elevationArtificial black plants × stem anchoring × pot elevationModel
  1. LAR, leaf area ratio (lamina area/total mass); SLA, specific leaf area (lamina area/lamina mass). †, P < 0.1; *, P < 0.05; **, P < 0.01; ***, P < 0.001.

df112122211
Stem length173.7***28.9***47.3***2.16.6**0.10.528.5***
Diameter14.4***0.27.3**1.10.00.30.62.9**
Total mass8.9**5.6*29.9***0.181.20.60.47.3***
Root mass33.3***2.128.3***0.53.4*1.90.99.5***
Stem mass2.54.0*18.5***2.60.20.30.34.3***
Leaf mass13.2***6.4*21.1***1.11.00.20.26.0***
Root/total mass34.4***0.44.9**0.33.4*1.90.75.2***
Stem/total mass83.7***0.612.1***0.33.1†0.20.310.5***
Leaf/total mass9.5**2.31.40.01.51.40.11.9
Lamina area7.8**6.8*0.71.60.30.10.01.7
SLA6.6*0.198.9***0.16.9**1.02.020.4***
LAR0.00.869.3***0.04.3*0.31.513.8***
Figure 3.

The length (a) and basal diameter (b) of stems of target Chenopodium album plants 2 wk after the start of the treatments (see Fig. 1). Plants were lifted (grey bars), lowered (black bars) or left unchanged (white bars). Different letters indicate significant differences (< 0.05) between treatments according to the Tukey–Kramer test (n = 7 in each group). Error bars represent + SE.

The height difference between target plants and neighbouring plants was also measured (Fig. 4). In the stand containing live plants, unchanged plants without anchoring had heights similar to those of their neighbours. In the stand containing live plants, for both unanchored lifted and lowered plants, the height difference recorded between the target plant and its neighbours was reduced by 5 cm compared with the initial difference of 10 cm. This indicates that height convergence was occurring, as observed in a previous study (Nagashima & Hikosaka, 2011). In the stand containing artificial black plants, where the height of the artificial plants was maintained at a similar height to that of unchanged plants, the unanchored lifted and lowered plants maintained a constant height difference with their neighbours throughout the experiment. These result suggests that differences in not light quantity but light quality are crucial for height convergence to occur.

Figure 4.

The difference in apparent height from that of neighbouring Chenopodium album plants 2 wk after the start of the treatments (see Fig. 1). Asterisks indicate the significance level of differences in apparent height before (0, 0.1 and − 0.1 m for unchanged, lifted and lowered plants, respectively) and after treatments (*, < 0.05; ***, < 0.001; see the Materials and Methods section). Error bars represent ± SE. Unchanged, white bars; lifted, grey bars; lowered, black bars.

Anchored, unchanged plants in the stand containing live plants grew significantly taller than their neighbours (Fig. 4). Plants that had been lowered exhibited a reduction in the initial height difference with their neighbours, whereas plants that had been lifted maintained the initial height difference throughout the experiment. Similar trends were also found in anchored plants in the stand containing artificial black plants. These results suggest that height convergence does not occur when mechanical stimuli are eliminated.

Stem diameter and biomass partitioning

Stem diameter was affected by the black coloration and pot elevation treatments, but not by stem anchoring (Table 1). No interaction was found in any combination of the treatments. Plants in the stand containing artificial black plants tended to have a larger diameter than those in the stand containing live plants (Fig. 3b). Stem diameter tended to be larger and smaller in lifted and lowered plants, respectively (Fig. 3b). The ratio of length to diameter was influenced by all treatments (Table 1). The ratio tended to be higher in the stand containing live plants than in the stand containing artificial black plants, in anchored plants relative to nonanchored plants, and in lowered plants relative to lifted plants (Table S2). Interaction effects were detected, with the effects of stem anchoring and pot elevation tending to be greater in the stand containing live plants than in the stand containing artificial black plants (Tables 1, S2).

Biomass and its partitioning among organs were affected by the treatments (Table 1). Total mass was greater in the stand containing artificial black plants than in the stand containing live plants, in anchored plants than in nonanchored plants, and in lifted plants than in lowered plants (Table S2). Biomass partitioning was influenced by the black coloration and pot elevation treatments, but not by anchoring (Table 1). There was an interaction effect between the black coloration and pot elevation treatments on root mass fraction (root/total mass, RMF) and stem mass fraction (stem/total mass, SMF). In the stand containing live plants, RMF tended to be greater in lifted plants than in both unchanged and lowered plants, whereas SMF exhibited the opposite response (Fig. 5). This response to pot elevation in RMF and SMF was not observed in the stand containing artificial black plants. Leaf mass fraction (leaf/total mass, LMF) was not affected by pot elevation. These results suggest that biomass partitioning was affected primarily by light quality rather than by light quantity or mechanical stimuli.

Figure 5.

The proportion of biomass of target Chenopodium album plants partitioned to the roots (a), stems (b) and leaves (c) 2 wk after the start of the treatments (see Fig. 1). Anchoring and nonanchoring treatments were pooled. Different letters indicate significant differences (< 0.05) between each combination of pot elevation and light quality treatments according to the Tukey–Kramer test (n = 14 in each group). Error bars represent + SE. Elevation treatments: unchanged, white bars; lifted, grey bars; lowered, black bars.

The specific leaf area (SLA) and leaf area ratio (LAR) were affected by pot elevation, and there was an interaction effect observed between the black coloration and pot elevation treatments (Table 1). SLA and LAR were higher in plants that had been lowered than in plants that had been lifted in the stand containing live plants (Table S2). There was an interaction effect observed between the black coloration and pot elevation treatments with responses to pot elevation being smaller in the stand containing artificial black plants than in the stand containing live plants.

Discussion

The results of this study clearly show that stem elongation is influenced by light quality, light quantity and mechanical stimuli. They also indicate that two of the factors, light quality and mechanical stimuli, are important environmental cues that induce height convergence, given that height convergence did not occur when the effects of light quality and mechanical stimuli were eliminated. It is well known that light quality has a significant effect on stem growth in field stands (Ballaréet al., 1990; Ballaré, 1994), and it has been shown in an elegant experiment that, in a light quality gradient, plants keep up with their neighbours (Vermeulen et al., 2008). The involvement of the other factors, however, has not been demonstrated previously. The results of this study suggest that stem elongation of plants in crowded stands is governed by a complex interaction of different environmental factors.

Anchored plants showed greater stem elongation, which is consistent with previous studies demonstrating that wind stimuli reduce elongation (e.g. Latimer et al., 1986; Holbrook & Putz, 1989; Henry & Thomas, 2002) and that manual swaying affects stem length (e.g. Smith & Ennos, 2003; Anten et al., 2005, 2006, 2009). These results suggest that mechanical stimuli caused by wind action can be a significant constraint to stem elongation. Meanwhile, it has been argued that enhanced evapotranspiration resulting from wind action may constrain overtopping (Nagashima & Hikosaka, 2011). In the present study, however, anchored unchanged plants overtopped neighbours and produced greater biomass, suggesting that enhanced evapotranspiration was not a constraint. It should be noted that the wind speed during our experiment was relatively low. For example, we observed a maximum wind speed of 25 m s−1 in our previous study (Nagashima & Hikosaka, 2011), whereas it was only 5 m s−1 in the present study. Therefore, it is still uncertain whether or not evapotranspiration is a constraint under strong winds.

The pot elevation treatment influenced stem growth in the stand containing artificial black plants, suggesting that light quantity also affects stem elongation. This is consistent with previous work demonstrating that the photon fluence rate itself affects elongation (Ballaréet al., 1991). In the present study, stem length was longer in plants that had been lowered relative to those that had been lifted, despite having a smaller total biomass. This indicates that stem elongation is not simply related to the availability of photosynthates, but is also influenced by the reduction in light quantity. However, the effect of reduced light quantity on stem elongation was too small to cause height convergence, suggesting that it is less important than the effects of light quality and mechanical stimuli.

In the present study, PAR was 15% higher in the black-painted than in the real plant stand. One may therefore consider that it is hard to distinguish to what extent the increased length of the target plants surrounded by living plants was caused by reduced PAR or R/FR ratio and the lesser response to pot elevation in the black-painted stand might result from greater PAR. In our previous study, however, we used a lower plant density (77 vs 116 plant m−2 for previous and present studies, respectively) and a smaller height difference between lifted and lowered plants (7 vs 10 cm). The relative PAR at the top of lowered plants was greater in the previous study than in the present black-painted plants (95% vs 60%, data not shown). In the previous study, therefore, height convergence occurred even when the variation in light quantity was smaller than in the black-painted stand in the present study. These results strongly indicate that height convergence occurs in response to light quality and mechanical stimuli, rather than light quantity.

The light quality and mechanical stimuli treatments produced different effects on stem diameter growth and biomass allocation. Accelerated elongation in plants experiencing reduced R/FR ratio was associated with a decrease in diameter growth and an increase in biomass allocation to stems at the expense of that to roots, consistent with previous studies (Maliakal et al., 1999). Stem elongation in plants experiencing a reduction in mechanical stimuli was not accompanied by significant changes in diameter growth or biomass allocation, but simply by an increase in total biomass. Although both light quality and mechanical stimuli treatments affected significantly the length to diameter ratio, the effects of the former were greater (Table S1). These results suggest that stem shape and biomass allocation are more sensitive to variations in light quality than to mechanical stimuli.

These results suggest that stem growth in field stands is regulated primarily by two independent systems, with one responding to variations in light quality and the other responding to mechanical stimuli. This is supported by the observation that there was no interactive effect of light quality and mechanical stimuli on stem elongation, and that the effects on diameter growth and biomass allocation also varied between the light quality and mechanical stimuli treatments. These regulatory systems are interesting from an ecological aspect. The former system enables plants to avoid the effects of significant shading by neighbours, whereas the latter system may assist plants in avoiding the mechanical stresses that can occur when they overtop their neighbours. Indeed, in an experiment with Impatiens capensis, elongated plants at high density were more prone to fail mechanically and to show reduced reproduction (Huber et al., 2011), suggesting that the control of mechanical robustness may influence fitness in elongating plants directly. These regulatory systems may also be consistent with constraints assumed in game-theoretical models for plant height (Givnish, 1982; Iwasa et al., 1985; Sakai, 1991; Falster & Westoby, 2003). Height convergence is a result of the two regulatory systems and may not occur if either one of them is not effective.

Conclusion

This study presented three novel findings. First, variations in light quality, quantity and mechanical stimuli all influence stem elongation in a crowded stand in the field. Second, light quality and mechanical stimuli are crucial for height convergence of plants. If one of these environmental factors is not present, height convergence does not occur. Third, light quality and mechanical stimuli affect plant growth in different ways. Plants can detect the presence of neighbours and physical stresses on stems, each of which influences stem elongation independently. Regulatory systems are employed by plants to avoid both being shaded by neighbours and suffering from mechanical stresses, which could contribute to the success of individual plants under conditions of competition for light in field populations.

Acknowledgements

The authors thank anonymous reviewers, T. Hirose and M. Tateno for their valuable comments and suggestions, H. Takahashi for the experimental set-up, N. Osada, Y. Osone and H. Taneda for assistance with the experiment and for their valuable comments, and M. Aiba for statistical advice. This research was supported by fellowships from the Japan Society for the Promotion of Science (nos. 3025 and 40172) and KAKENHI (no. 12601) to HN and by KAKENHI (nos. 20677001 and 21114009) to KH.

Ancillary