SEARCH

SEARCH BY CITATION

Keywords:

  • cell cycle;
  • cell division;
  • interspecific difference;
  • leaf size;
  • meristem;
  • relative growth rate (RGR);
  • specific leaf area (SLA)

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  • Investigation of responses of meristems to environmental conditions is important for understanding the mechanisms and consequences of plant phenotypic plasticity. Here, we examined how meristem plasticity to light and soil nutrients affected leaf growth and relative growth rate (RGR) in fast- and slow-growing Festuca grass species.
  • Activity in shoot apical meristems was measured by leaf appearance rate, and that in leaf meristems by the duration and rate of cell production, which was further divided into single cell cycle time and the number of dividing cells.
  • Light and soil nutrients affected activity in shoot apical meristems similarly. The high nutrient supply increased the number of dividing cells, which was responsible for enhancement of cell production rate; shaded conditions extended the duration of cell production. As a result, leaf length increased under high nutrient and shaded conditions. The RGR was correlated positively with the total meristem size of the shoot under a low nutrient supply, implying inhibition of RGR by cell production under nutrient-limited conditions. Fast-growing species were more plastic for cell production rate and specific leaf area (SLA) but less plastic for RGR than slow-growing species.
  • This study demonstrates that meristem plasticity plays key roles in characterizing environmental responses of plant species.

Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Phenotypic plasticity, or the ability of an organism to alter its phenotype in response to changes in the environment, strongly influences the ecological success of individual plants (Bradshaw, 1965). That being so, the relationships between the pattern and magnitude of plasticity and components of fitness in plants have been examined under various conditions of abiotic and biotic environmental heterogeneity (Sultan, 2000; Miner et al., 2005; Pigliucci, 2005; Valladares et al., 2007). However, little is known of the processes and mechanisms by which phenotypic plasticity occurs in response to different environmental conditions (Ackerly & Sultan, 2006). To understand the adaptive evolution of phenotypic plasticity in natural environments, it is crucial to investigate the mechanisms by which plants sense environmental signals and how those signals influence the plant phenotype (Schmitt et al., 2003).

All plants have a developmental system, which differs from that of animals: they grow through the repeated production of modular units (leaf, stem and branch) throughout their lifetime. Given this developmental system, the study of meristem plasticity can play a vital role in improving our understanding of phenotypic plasticity in plants, for the following reasons. First, the plasticity of the whole plant is a combination of the plasticity within modules plus the plasticity in production of different types of module, each of which is derived from the growth and development of a single meristem (de Kroon et al., 2005). Second, given that the fate and activity of meristems play central roles in plant developmental processes (Leyser & Day, 2003), investigation of the responses of meristems to the environment can make a contribution to the field of ‘eco-devo’, which aims to combine an understanding of the mechanism of phenotypic expression with its consequences for the success of plants in their habitats and evolutionary changes. Finally, the rate of plant growth depends on the supply of new cells produced by meristems; hence, an investigation of meristem plasticity is essential for understanding how the rate of plant growth changes in response to changes in the environment.

Grass shoots include shoot apical meristems and leaf meristems (Fig. 1a). The shoot apical meristem, which is located at the shoot apex, initiates leaf primordia and its activity is characterized by the rate at which new leaves appear. The leaf meristem is located at the leaf base and drives the growth and development of an initiated leaf by means of the production of new cells. As shown in Fig. 1b, the initial cells at the basal end of the leaf meristem enter into the cell cycle, undergo several rounds of division, and exit from the meristematic zone. It is thus evident that a number of different factors contribute to the activity of the leaf meristem: the period during which initial cells continue to enter the cell cycle; the time the cells take to complete a single cell cycle; and the number of cell cycles that the meristematic cells undergo before leaving the meristematic zone. These factors correspond to the duration of cell production, the single cell cycle time and the number of dividing cells, respectively. Quantification of these parameters is essential for the investigation of relationships between meristem plasticity and plant growth. For the developmental system of a leaf of grass, in which the width of the leaf is first determined along the circumference of the shoot apical meristem, followed by elongation along the longitudinal axis (Beemster & Masle, 1996; Sugiyama, 2005a), these cell parameters can be quantified by measuring the rate of leaf elongation and observing the length of the cells along the longitudinal axis (Fiorani et al., 2000; Sugiyama, 2005b).

image

Figure 1.  (a) The two types of meristem in a grass shoot. The shoot apical meristem (SAM) is located at the shoot base, and the leaf meristem (LM) is located at the leaf base. Leaf primordia (LP), which are initiated on the SAM, proliferate first along the transverse axis and encircle the subapical region of the SAM, and then elongate along the longitudinal axis. (b) Cell division parameters that represent leaf meristem activity can be estimated from the cell length profile. The cell division zone is defined as the region within which cell length decreases. The initial cells of the meristem enter into the cell cycle at the basal end of the meristematic zone, undergo several rounds of the cell cycle, which consists of G1, S, G2 and M stages, and exit from the cell cycle at the distal end of the zone. These processes are governed by: (1) how long the initial cells continue to enter into the cell cycle; (2) how fast the cells complete a single cell cycle; and (3) how many cell cycles the meristematic cells complete before leaving the division zone. These factors are represented by the duration of cell production, the average time of a single cell cycle and the size of the division zone or the number of dividing cells, respectively.

Download figure to PowerPoint

Soil nutrients and light can each serve either as specific signals that trigger a particular response from a plant or as a resource for the plants to use. The former represents a signal-mediated response and the latter a resource-mediated response. For example, when plants increase their height under shaded conditions by using phytochrome molecules, they are showing a signal-mediated response (Smith, 2000; Schmitt et al., 2003) and when plants intercept incident light and convert it into assimilated carbon in a photosynthetic pathway, they are showing a resource-mediated response. It is important to distinguish between the two types of response mechanism, because they give rise to contrasting views of the ecological roles of phenotypic plasticity (Sultan, 1995; Kurashige & Callahan, 2007). The signal-mediated response can be viewed as a future-oriented response that triggers adaptive plasticity to minimize future deleterious effects of the environment, such as shading by neighbouring plants (Lambers et al., 1998; Terashima et al., 2006). By contrast, the resource-mediated response is an opportunistic, present-oriented response of plants to current resources, the effect of which is inevitable growth response (Sultan, 1995). Signal-mediated responses are expected to result in changes in plasticity that are asymmetric between light and soil nutrients, because signals are specific to a particular environment. By contrast, resource-mediated responses are expected to give rise to symmetric changes in plasticity, because carbon assimilation rate is strongly influenced by both light and soil nutrients as main resources of plants.

In this study, we specifically addressed the following three issues by examining the plasticity of meristem activity in response to light and soil nutrients in fast- and slow-growing Festuca grass species. First, we investigated whether the cell parameters that represent meristem activity showed signal- or resource-mediated responses to soil nutrients and light. Second, we examined the manner in which the cell parameters influenced the plasticity of leaf size (length, width and area) and relative growth rate (RGR). Given that the RGR of a vegetative shoot depends on the supply of new cells produced by meristems (sink activity), as well as the supply of carbohydrate from mature leaves (source activity), we examined how the plasticity of RGR is influenced by the plasticity of the meristem and the specific leaf area (SLA), which is closely associated with the rate of photosynthesis per unit leaf mass (Reich et al., 1998). Third, we explored interspecific differences in the magnitude of phenotypic plasticity. Among the four species of Festuca studied, F. arundinacea Schreb. and F. pratensis Hudson are fast-growing species, whereas F. rubra L. and F. ovina L. are slow-growing species (Sugiyama, 2005c). We examined whether the fast-growing species showed a greater magnitude of plasticity with respect to RGR and leaf size than the slow-growing species. In addition, we determined which cell parameters were responsible for interspecific differences in the magnitude of plasticity and derived the implications of our findings for the evolution of phenotypic plasticity.

Materials and Methods

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Conditions of growth

We conducted the experiments in a glasshouse under natural conditions to ensure sufficient space and high light intensity. On 10 May 2002, seeds of each species were sown in a plastic box (15 × 5.5 × 10 cm) filled with sand. Each box was filled with a 1 : 500 dilution of a commercial nutrient solution (5% N, 10% P2O5 and 5% K2O; Hyponex Co. Ltd, Tokyo, Japan). The plants were grown in an unheated glasshouse under natural sunlight from the beginning of May to early July, when the climatic conditions are the most favourable for growth of these species. At 21 d after sowing, 10 seedlings of each species were harvested, and the shoots and roots were separated and dried at 80°C for 48 h. After drying, the dry weight of each individual part was determined. Also at 21 d after sowing, two seedlings of uniform size were transplanted into a 10-cm-diameter, 12-cm-deep pot that had been filled with fertilized or unfertilized soil, both of which had a similar texture but contained different nutrients. Soil nitrogen differed 10-fold (15.2 ± 1.63 and 1.63 ± 0.23 mg per 100 g soil) and soil phosphorus differed fivefold (61.8 ± 3.8 and 13.6 ± 0.7 mg per 100 g soil) between the fertilized and unfertilized soils, respectively. Plants in both the fertilized and unfertilized soils were exposed to two different light conditions. Half of the pots were grown in a shaded plot covered with cheesecloth that allowed 30% of the light to reach the plants. The other half were grown under conditions of natural daylight (control). The red : far red ratio of the light that passed through the cheesecloth (0.92) was almost the same as that of natural light; hence, we could discount changes in plasticity caused by changes in light quality. The maximum photon flux density differed threefold; it was 1694 μmol m−2 s−1 for the control plot and 538 μmol m−2 s−1 for the shaded plot on a sunny day (29 May 2002). The averaged photon flux density during the daytime was 510 μmol m−2 s−1 for the control plot and 153 μmol  m−2 s−1 for the shaded plot. The mean temperature and relative humidity during the growing period were 22.5°C and 63%, respectively, for the control plot, and 22.1°C and 74%, respectively, for the shaded plot. After the plants had become established, the seedlings were thinned to one plant per pot.

Two experiments were conducted as described later: one to measure the RGR and to obtain morphological measurements of mature leaves, and the other to measure meristem activity. Each experiment consisted of six replications for each species for each treatment. For each experiment, all individual plants were harvested on a single day to ensure identical climatic conditions.

RGR and leaf anatomical measurements

All plants were harvested at 24 d after transplantation (8 July 2002). The youngest mature leaf from each plant was cut and its length and width were measured. The SLA was determined for each of six plants. After the FW of the leaves had been measured, a digital image of the leaf area was recorded using an optical scanner (D660U; Canon Inc., Tokyo, Japan). The leaf area was analysed using the National Institutes of Health (NIH) Image Program (version 1.6; Bethesda, MD, USA). The leaves were then dried at 80°C for 48 h and their dry weight was recorded.

After soil had been washed from the roots, the plants were separated into leaves, stems and roots. The dry weights of these parts were measured after drying at 80°C for 48 h. The RGR value was calculated on the basis of the log-transformed difference between total dry weight of roots and shoots at transplantation and at 24 d thereafter.

Meristem activity and cell parameters

The activity of shoot apical meristems was quantified by a single variable: the leaf appearance rate (i.e. the number of new leaves that appeared per day for a single shoot). The dates of emergence of new leaves on the main stem were recorded successively during the month after transplantation, and the leaf appearance rate was obtained from the slope of the linear regression of the number of visible leaves against the number of days. For the RGR experiment, the number of meristems per shoot was estimated from the number of tillers at the second harvest.

Leaf meristem activity was quantified by five variables: (1) cell production duration by the leaf meristem (number of d for which the leaf grew in length); (2) cell production rate (number of new cells produced in the meristematic region per hour), which was further subdivided into the variables (3)–(5); (3) single cell cycle time (h required for a single cell to complete its division cycle); (4) the total number of dividing cells in the leaf meristem; and (5) mature cell length. To determine these parameters, the leaf elongation rate (length of leaf elongation per h) was determined. The lengths of developing leaves that appeared above the leaf sheath were recorded over five successive days (15–20 July 2002) for all plants measured, and the leaf elongation rate was calculated from the slope of the linear regression of the leaf length against the number of d. The cell production duration was calculated by dividing the mature leaf length by the leaf elongation rate.

The leaves used for measuring the leaf elongation rate were dissected carefully and 50-mm-long basal sections were cut and stored in FAA fixative (formalin, acetic acid and 70% ethanol in a ratio of 5 : 5 : 90 (v/v/v)) in order to investigate the length profile of mesophyll cells in the leaf meristem. The basal leaf segments were placed on glass slides and lactic acid was dripped onto them to clear the tissue. The lengths of the mesophyll cells located between two adjacent vascular bundles were measured at a magnification of ×400, using a light microscope (BX40; Olympus, Tokyo, Japan) that was equipped for differential interference contrast. The microscope was attached to a charge-coupled device (CCD) camera (DP11; Olympus) and microscopic images were displayed on a 15-inch LCD screen through the CCD camera. The cell lengths on the screen were measured using a ruler. The actual sizes were then calculated by comparison with an objective micrometer. In total, we measured 25 cells in different fields from the leaf base to a point 40 mm distal along the longitudinal axis. The measurements were made at 1.0 mm increments in the division zone and at 2.0 mm increments in the elongation zone.

The original data on cell length were smoothed using a moving average of three successive measurements where the previous and next measurements were weighted by a weighting factor of 0.5. The division zone was defined as the zone in which cell length decreased (Fig. 1b). The mature cell length was calculated by averaging all data points distal to the position at which the increase in cell length between two successive points became ≤ 0 μm. Under steady-state conditions of cell production, the number of cells that enter the division zone per unit time is equal to the number of cells that leave the zone. For that reason, the cell production rate (P) was defined as

  • image( Eqn 1)

where LER (leaf elongation rate) is transformed into μm h−1, and l (mature cell length) is expressed as μm. The cell production rate is determined by the total number of dividing cells and their average single cell cycle time. The division zone was subdivided into 1.0-mm-length zones from the basal end and the cell number of each zone was calculated from the reciprocal of the mean cell length of each zone. The total number of dividing cells in the entire meristem zone (N) was estimated by adding the number of cells in each segment. Under steady-state growth conditions, the average single cell cycle time (T ) can be calculated as follows (Ivanov & Dubrovsky, 1997):

  • image( Eqn 2)

The leaf meristem size was estimated as the product of the length of the division zone and leaf width, whereas the total meristem size per plant was estimated as the product of leaf meristem size and the number of meristems (tillers) per plant.

Statistical analysis

Multivariate analysis of variance (MANOVA) was performed using five cell parameters (cell production duration, cell production rate, single cell cycle time, number of dividing cells and mature cell length) as dependent variables. Subsequently, three-way ANOVA was applied to each trait, including leaf traits and RGR. Species were nested within species groups (fast-growing and slow-growing species) to test the effects of species groups. The Tukey–Kramer honestly significant difference (HSD) test was used to compare mean values between treatments or between species. All statistical analyses were performed using JMP ver. 4 software (SAS Inc., Cary, NC, USA). The relative distance plasticity index (RDPI) was used to compare the magnitude of plasticity between different traits or between different environments (Valladares, 2006). The relative distances (RDs) between trait values for high and low intensities of light or high and low concentrations of soil nutrients were calculated as

  • image( Eqn 3)

where xij2 and xij1 signify trait values of given individuals (= 1 … 6) at a high (j2) or low (j1) intensity of light or concentration of soil nutrients. A positive RD represented an increase at the higher value of the resource; a negative RD represented the decrease at the higher value of the resource. The RD for nutrients or light was calculated for all combinations of six plants between the high and low values of a treatment in both the high (= 36) and low values (= 36) of the other condition. The RDPI for each environment was obtained from the mean of the relative distances for all combinations (= 72).

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Pattern of plasticity in response to soil nutrients and light

MANOVA using the five cell parameters as dependent variables showed highly significant effects for species, nutrients and light, and the interactions between them (< 0.001, Table 1). However, ANOVA (Table 1) revealed differential effects of nutrients and light on the cell parameters. Figure 2 shows the RDPI averaged over the four species for each trait in response to soil nutrients and light. The leaf appearance rate, which is a measure of shoot apical meristem activity, showed a positive RDPI for soil nutrients and light. The cell production rate and the number of dividing cells in the leaf meristem showed a positive and high RDPI only for soil nutrients and a negligible response to light. The cell production duration showed a negative RDPI for light and a negligible one for nutrients. The single cell cycle time and mature cell length showed a negligible RDPI for nutrients and light. Leaf width showed a positive and large RDPI for nutrients but a negligible one for light, whereas the RDPI of leaf length in response to nutrients was positive, although that to light was strongly negative. The RGR showed a similar magnitude of response to nutrients and light, and SLA showed a strongly negative RDPI for light and a negligible response to nutrients (Fig. 2).

Table 1.   MANOVA of five cell parameters of the leaf meristem (duration of cell production, cell production rate, single cell cycle time, number of dividing cells and mature cell length) and ANOVA of cell parameters, leaf growth and relative growth rate (RGR)
 F-ratio or MS
Species groupSpecies (Sp)Nutrients (N)Light (L)N × LSp × NSp × LSp × N × LError
  1. The F-ratio was shown for MANOVA and the mean square (MS) value was shown for ANOVA. Species (Sp) were nested within the species group (fast-growing vs slow-growing species) in ANOVA.

  2. *, ** and ***, significance at 5%, 1% and 0.1%, respectively.

MANOVA (F ratio)
 Degree of freedom (df) 15, 2025, 735, 735, 7315, 20215, 20215, 202 
 Five cell parameters of leaf meristem 35.9***367.6***63.7***2.4*17.7***4.8***7.8*** 
ANOVA (MS)
 df1211133380
Meristem activities and cell parameters
 Cell production rate (P, d−1)13 172***2329***14 346***94*112*2011***33300***17
 Number of dividing cells (N)991 922***385 539***1 654 450***26 202119 779172 183**22 45959 89037 431
 Cell cycle time (T, h)304***86*131**3140768*49*17
 Duration of cell production (D, d)122 967***321 915***87 477***482 872***183 451***599823 994**18 495*4999
 Mature cell length (l, μm)74.5***55.2***38.5***27.2***0.94.430.51.8
 Leaf appearance rate (LAP, d−1)0.0075***0.0043***0.0432***0.0494***0.007***0.00030.001*0.000 2240.0003
 Total meristem size (cm2 per plant)125.8***4.3404.2***212.1***173.6***39.2***23.2***23.6***3.3
Leaf size and RGR
 Leaf length (LL, cm)333***2389***1540***3845***188***94***197***50*14
 Leaf width (LW, cm)7.16***0.21***1.2***00.06**0.25***0.13***0.06***0.01
 Leaf area (LA, cm2)8394***54*2392***437***140**461***160***69**14
 RGR (g g−1 d−1)0.0035***0.0003***0.0045***0.006***0.001***0.000 01 0.000 10*0.000 01 0.000 01
 SLA (cm2 g−1)102 102***14 196***3640796 250***182382227 846***9101820
image

Figure 2.  Relative distances of plasticity indices (RDPIs) of respective traits in response to light (y-axis) and soil nutrients (x-axis). Each RDPI was averaged over the four Festuca species. D, cell production duration; l, mature cell length; LA, leaf area; LAP, leaf appearance rate; LL, leaf length; LW, leaf width; N, number of dividing cells; P, cell production rate; RGR, relative growth rate; SLA, specific leaf area; T, single cell cycle time.

Download figure to PowerPoint

Given the foregoing, we classified the patterns of plasticity of the traits in response to soil nutrients and light into four types: (i) high RDPI for nutrients and light (leaf appearance rate, leaf length and RGR); (ii) high RDPI only for nutrients (cell production rate, number of dividing cells, leaf width); (iii) high RDPI only for light (cell production duration and SLA); and (iv) negligible RDPI for nutrients and light (single cell cycle time and mature cell length).

Plasticity of cell parameters

Figure 3 depicts the profiles of cell lengths in the leaf meristem for a fast-growing (F. arundinacea) and a slow-growing species (F. ovina) subjected to each type of treatment. When the plants had a good supply of nutrients, the division zone was lengthened in both species. However, the cell length profiles of either species differed little in response to the different light treatments. As expected from the cell length profiles, the cell production rate was higher when the supply of nutrients was good than when it was poor (Fig. 4). When the two components of the cell production rate (i.e. the number of dividing cells and single cell cycle time) were compared, it was apparent that the increased cell production rate when the supply of nutrients was good resulted from an increase in the number of dividing cells, rather than an increase in average single cell cycle time. The cell production duration increased significantly under conditions of shade but was affected little by nutrient concentrations.

image

Figure 3.  Profiles of cell lengths in the leaf growth zone in fast-growing (Festuca arundinacea) and slow-growing (F. ovina) species. Circles, low nutrient concentrations; triangles, high nutrient concentrations; open symbols, high light conditions; closed symbols, low light conditions.

Download figure to PowerPoint

image

Figure 4.  Mean values of cell parameters in each species under conditions of high nutrients and strong light (NL), high nutrients and low light (Nl), low nutrients and strong light (nL), and low nutrients and low light (nl). Different letters show significant differences (Tukey–Kramer HSD test) in each environment. The error bars represent ±1 SE. Fa, Festuca arundinacea; Fp, F. pratensis; Fr, F. rubra; Fo, F. ovina.

Download figure to PowerPoint

The leaf appearance rate, which is a measure of shoot apical meristem activity, showed an additive response to nutrients and light, as evidenced by the fact that conditions of a good supply of nutrients and low light, and poor supply of nutrients and strong light showed similar values and that the highest values were obtained with the combination of strong light and a good supply of nutrients.

Significant interspecific differences were observed for the cell production rate (Table 1). The fast-growing species (F. arundinacea and F. pratensis) produced cells at a significantly higher rate than the slow-growing species (F. rubura and F. ovina), as shown in Fig. 4. The number of dividing cells tended to be greater in the fast-growing species than in the slow-growing species, whereas the single cell cycle time only showed a slight difference between the species. Therefore, the difference in the cell production rate between the fast- and slow-growing species was caused by the higher number of dividing cells. The apical meristem activity (leaf appearance rate) and cell production duration did not differ significantly between the fast- and slow-growing species.

Among the four species, the shortest cell lengths were observed in F. pratensis under all treatment conditions (Fig. 4). With the exception of this species, mature cell length did not show any significant differences between nutrient or light treatments. However, the mature cell length was correlated strongly and negatively with the cell production rate among individual plants (= −0.742**; Fig. 5), which indicated that cell size was influenced strongly by cell division activity in the leaf meristem.

image

Figure 5.  Relationship between rate of cell production and length of mature mesophyll cells among individual plants. The data were combined across all species (Festuca arundinacea, F. pratensis, F. rubra, F. ovina) and treatments. ***Statistical significance at the 0.1% level.

Download figure to PowerPoint

Plasticity of leaf size and RGR

Although leaf width increased only when the plant received a good supply of nutrients, leaf length increased both when the plant received a good supply of nutrients and was shaded from light (Fig. 6). The fast-growing species, F. arundinacea and F. pratensis, had a significantly larger leaf area than the two slow-growing species, F. rubra and F. ovina, under all treatment conditions (Fig. 6). This interspecific difference in leaf area was determined by leaf width rather than leaf length, because the leaves of the fast-growing species were significantly wider than those of the slow-growing species but no clear differences were apparent in leaf length.

image

Figure 6.  Mean values of leaf area, leaf length, leaf width, relative growth rate (RGR), total meristem size and specific leaf area (SLA) in each species under conditions of high nutrients and strong light (NL), high nutrients and low light (Nl), low nutrients and strong light (nL), and low nutrients and low light (nl). Different letters show significant differences (Tukey–Kramer HSD test) in each environment. The error bars represent ±1 SE. Fa, Festuca arundinacea; Fp, F. pratensis; Fr, F. rubra; Fo, F. ovina.

Download figure to PowerPoint

The highest RGR occurred when the supply of nutrients was good and the light was strong, although similar RGRs occurred when the supply of nutrients was good and light was low, and when the supply of nutrients was poor and the light was strong (Fig. 6). The total meristem size, which was estimated as the product of the length of the division zone, the leaf width and the number of tillers, differed greatly between a good supply of nutrients and a poor supply of nutrients (Fig. 6). In particular, the fast-growing species had a larger total meristem size than the slow-growing species when the supply of nutrients was good. Under strong light, for example, the total meristem size of F. pratensis when the supply of nutrients was good (14.8 cm2 per plant) was 43 times greater than when the supply was poor (0.35 cm2 per plant). The RGR showed a linear relationship with total meristem size when the supply of nutrients was poor, but a hyperbolic relationship when the supply was high (Fig. 7). In addition, in the fast-growing species, the RGR levelled off at the higher values of meristem size when the supply of nutrients was good (Fig. 7). By contrast, the RGR showed a significantly positive correlation with SLA under shaded conditions (= 0.788**), although it showed no clear correlation with SLA under strong light (= 0.428, Fig. 7).

image

Figure 7.  Relationship of relative growth rate (RGR) to total meristem size per shoot (a) and specific leaf area (SLA) (b). Circles, low nutrient concentrations; triangles, high nutrient concentrations; open symbols, strong light conditions; closed symbols, low light conditions. * and **, statistical significance at 5% and 1% levels, respectively.

Download figure to PowerPoint

Interspecific differences in magnitude of plasticity

The significant interactions between species and soil nutrients with respect to the cell production rate and the number of dividing cells (Table 1) indicate the existence of interspecific differences in the magnitude of plasticity. The fast-growing species had significantly higher RDPIs for the cell production rate and the number of dividing cells in response to nutrients than those of the slow-growing species (Fig. 8). By contrast, the RDPI of cell production duration in response to light showed no consistent difference between the fast- and slow-growing species, although the RDPIs of the fast-growing species with respect to soil nutrients were significantly lower than those of the slow-growing species. No consistent differences were found in the RDPIs for leaf size (length, width and area) between the fast- and slow-growing species.

image

Figure 8.  Relative distance plasticity indices (RDPIs) in response to soil nutrients (closed circles) and light (open circles) in each species. Different letters show significant differences among the species in each environment. The error bars represent ±1 SE. Fa, Festuca arundinacea; Fp, F. pratensis; Fr, F. rubra; Fo, F. ovina. RGR, relative growth rate; SLA, specific leaf area.

Download figure to PowerPoint

Although the fast-growing species showed a significantly higher RGR than the slow-growing species, irrespective of environmental conditions (Fig. 6), the RDPI for RGR was significantly lower for the fast-growing species than for the slow-growing species (Fig. 8). The fast-growing species had a significantly more positive RDPI for total meristem size to soil nutrients and a significantly more negative RDPI for SLA to light than did the slow-growing species (Fig. 8).

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Plasticity of meristems

Increased intensities of light and increased concentrations of soil nutrients both led to a similar degree of acceleration of the leaf appearance rate. Wilhelm & McMaster (1995) have reported that the rate of leaf initiation is increased by a good supply of nutrients, high irradiance and elevated CO2. These results suggest that shoot apical meristem activity is increased by the increased availability of resources and thus represents resource-mediated plasticity, although other environmental factors, such as the thermal environment and water stress, also influence the rate of leaf initiation in the shoot apex (Granier & Tardieu, 1998, 1999).

By contrast, the differences in plasticity of the rate and duration of cell production in response to soil nutrients or to light imply that these cell parameters are controlled by specific light or nutrient signals, rather than by the resource-mediated rate of carbon assimilation. Previous studies have also shown that cell division and the rate of leaf elongation are inhibited when the availability of nitrogen is low (MacAdam et al., 1989; Fricke et al., 1997; Kavanováet al., 2008). Rahayu et al. (2005) have revealed that plants grown in soil in which there is a good supply of nitrates show an increased amount of cytokinins, which are plant hormones that are transported over long distances in their xylem. Those authors suggested that cytokinin synthesis is activated after the perception of nitrate in roots, and that this hormone acts as a signal molecule to regulate the increase in leaf meristem activity. Sakakibara et al. (2006) have reported that cytokinins are induced by inorganic nitrogen in the root and that they play important roles in cell division and the consequent growth and development of the shoot. Therefore, it is likely that the nitrate–cytokinin signalling pathway plays a key role in determining the number of dividing cells in the leaf meristem, which is the parameter that is responsible for the increase of cell production rate when the supply of nutrients is good. In addition to nitrate deficiency, phosphorus deficiency also inhibits the rate of cell production (Assuero et al., 2004; Kavanováet al., 2006), which implies the involvement of a similar cytokinin pathway that is regulated by phosphorus.

Cookson et al. (2005) and Cookson & Granier (2006) have reported that conditions of shade increase the duration of leaf expansion. Our study also showed that leaf elongation has a longer duration under shaded conditions; this might result from the extended entry of initial cells into the division zone at the basal end of the leaf meristem. On the other hand, the negligible changes in single cell cycle time indicate that soil nutrients and light have less influence on the cell cycle time than on the duration and rate of cell production. As shown in Fig. 1, high plasticity of the rate and duration of cell production is associated closely with the control of the G1–S transition during the cell cycle in the leaf meristem.

Plasticity of leaf size and RGR

Leaf width, which is determined mainly by cell division activity in the subapical region of the shoot apical meristem, showed a similar response to that of the number of dividing cells in the leaf meristem. This suggests that the same mechanisms control cell proliferation in the subapical region of the shoot apical meristem and the rate of cell production along the longitudinal axis in the leaf meristem. By contrast, leaf length increased both when the supply of nutrients was good and under shaded conditions. When the supply of nutrients was good, leaf length increased as a result of an increase in the cell production rate, whereas under shaded conditions the cell production duration increased but the cell production rate remained the same. Granier & Tardieu (1999) and Cookson et al. (2005) have reported that when intensities of light were low, the duration of leaf expansion increased slightly and the rate of cell production decreased a lot, which resulted in a big reduction in leaf area. In their experiments, the photon flux density used (60–75 μmol m−2 s−1) was lower than that in the present study (154 μmol m−2 s−1). Therefore, the plants were more likely to be limited severely by a poor supply of sugar as a result of low irradiance. The cell cycle is known to be regulated strictly by sugar supply (Menges et al., 2006). This result indicated that the plasticity of leaf length in response to light included a direct signal-mediated response of duration of cell production and a resource-mediated response of the rate of cell production to the availability of light. Whether the leaf length increased or decreased depended on the relative strength of these two types of response, which in turn depended on the intensity of irradiance.

Given that cell size did not differ significantly between the different conditions of soil nutrients and light, the plasticity of plant mass or RGR should depend on the responses of cell proliferation in the meristems. The total meristem size per plant was affected more strongly by nutrients than by light: the size of leaf meristems increased a great deal when the supply of nutrients was good. Consequently, the RGR was correlated strongly and positively with the total meristem size per plant when the supply of nutrients was poor (Fig. 7), which suggested that the RGR was limited by the supply of new cells (sink activity). However, when the supply of nutrients was good, the limitation imposed by the sink activity was alleviated. For example, the cell production rate in F. pratensis under conditions of high irradiance increased from 25 cells ha–1 for low nutrient concentrations to 92 cells ha–1 for high nutrient concentrations. When the supply of nutrients was good, the RGR of this species levelled off at high values of total meristem size (Fig. 7). Consequently, there was an apparent overproduction of cells when the supply of nutrients was good, which resulted in a reversal of the relative balance between sink and source functions. These results suggest that cell production (the sink activity) in plants is not necessarily well balanced with the carbohydrate supply (source activity). RGR might be restricted more frequently by the rate of cell production than by carbohydrate supply when the supply of nutrients is poor, although the supply of carbohydrate is more likely to be limited when the supply of nutrients is good.

The imbalance between source and sink functions during plant growth seemed to arise from inconsistent control between cell division and cell growth. Cell production that occurs in the division zone is regulated mainly by nutrient signals, whereas cell growth that occurs in the elongation zone depends on the supply of carbohydrate. Suppression of the rate of cell production in the leaf meristem by nutrient signals enables the plant to avoid the overproduction of cells when the supply of nutrients is poor, in which case the photosynthetic rate in leaves is expected to be low. Leaves that have a poor supply of nutrients or that have been exposed to prolonged high concentrations of CO2 frequently accumulate starch (Stitt & Krapp, 1999); this is likely to represent a shortage of sink activity relative to source function.

On the other hand, the increase in the rate of cell production in the leaf meristem when the supply of nutrients is good requires a proportional increase in the carbohydrate supply to maintain the cell growth rate. If the increased carbon assimilation does not meet the requirements of increased cell production, the shortage of carbohydrate must be compensated for by changes in cell growth. A negative correlation between the cell production rate and mature cell length (Fig. 5) suggests that changes in cell size compensate for the shortage of carbohydrate. Tsukaya (2002) has also demonstrated the existence of developmental coordination between cell proliferation and cell growth with respect to the determination of leaf size.

Using a comparison of grass species that received a good supply of nutrients as a basis, Sugiyama (2005b) have reported that the high rate of cell production in fast-growing species correlates positively with SLA, which in turn is related to a high rate of photosynthesis per unit leaf mass (Reich et al., 1998). These results suggest that evolutionary changes in cell division activity in leaf meristem accompany changes in traits pertaining to leaf function that lead to an increased rate of carbon assimilation.

Interspecific differences in meristem plasticity

Although the fast-growing species had a significantly larger leaf area and width than the slow-growing species, no significant differences in the magnitude of plasticity of leaf size in response to nutrients and light were found between the two groups. By contrast, significant differences between the two groups were found for the cell production rate and number of dividing cells (Fig. 8). Fast-growing species had significantly higher RDPI for the cell production rate and the number of dividing cells than those of the slow-growing species (Fig. 8). Consequently, the difference in phenotypic plasticity between the fast- and slow-growing species of Festuca could be traced to the response of the cell cycle in the leaf meristem.

In addition to the greater plasticity of the cell production rate in response to soil nutrients, the present study also revealed that plasticity of the fast-growing Festuca species is characterized by the greater plasticity of SLA in response to light, as compared with the slow-growing species. Poorter & Remkes (1990) demonstrated the importance of SLA in determining interspecific differences in RGR, especially under low-irradiance conditions (Meziane & Shipley, 1999; Shipley, 2002). In the current study, SLA showed a positive correlation with RGR only under shaded conditions. The high magnitude of plasticity of SLA in the fast-growing species can avoid a large decline in the photosynthetic rate per unit leaf mass under shaded conditions because of a large increase in SLA. Therefore, the high magnitude of plasticity of SLA could cause the low magnitude of plasticity in the RGR of the fast-growing species between low and high amounts of resources (Fig. 8).

The mechanisms by which leaf size is regulated dynamically have been explored at the cellular level in many species, including dicotyledonous plant species (Fiorani et al., 2000; Tsukaya, 2002; Cookson et al., 2005; Granier & Tardieu, 2009). However, the regulatory mechanisms of RGR plasticity have been investigated only rarely at the cellular level. The present study revealed that when the supply of nutrients in the soil was poor, the RGR was affected strongly by the signal-mediated rate of cell production, rather than by the resource-mediated rate of carbon assimilation. This implies that plants implement a forward response to control growth rate through the modulation of meristem activities. The main factors that characterize the primary strategy of plants in their response to environmental gradients are the pattern and magnitude of the plant responses (Grime, 2001). These are the main physiological attributes that contribute to the replacement of dominant species during secondary succession (Bazzaz, 1979). The results of the present study demonstrate that the plasticity of meristem activities plays a key role in determining the responses of plants to the environment.

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

This study was supported by a Grant-in Aid for Exploratory Research (no. 13876072) from the Japan Society for the Promotion of Science (JSPS).

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References