Intraspecific variation in temperature dependence of gas exchange characteristics among Plantago asiatica ecotypes from different temperature regimes

Authors


Author for correspondence: Kouki Hikosaka
Tel:+81 22 7957735
Fax:+81 22 7956699
Email: hikosaka@mail.tains.tohoku.ac.jp

Summary

  • • There are large inter- and intraspecific differences in the temperature dependence of photosynthesis, but the physiological cause of the variation is poorly understood. Here, the temperature dependence of photosynthesis was examined in three ecotypes of Plantago asiatica transplanted from different latitudes, where the mean annual temperature varies between 7.5 and 16.8°C.
  • • Plants were raised at 15 or 30°C, and the CO2 response of photosynthetic rates was determined at various temperatures.
  • • When plants were grown at 30°C, no difference was found in the temperature dependence of photosynthesis among ecotypes. When plants were grown at 15°C, ecotypes from a higher latitude maintained a relatively higher photosynthetic rate at low measurement temperatures. This difference was caused by a difference in the balance between the capacities of two processes, ribulose-1,5-bisphosphate regeneration (Jmax) and carboxylation (Vcmax), which altered the limiting step of photosynthesis at low temperatures. The organization of photosynthetic proteins also varied among ecotypes. The ecotype from the highest latitude increased the Jmax : Vcmax ratio with decreasing growth temperature, while that from the lowest latitude did not.
  • • It is concluded that nitrogen partitioning in the photosynthetic apparatus and its response to growth temperature were different among ecotypes, which caused an intraspecific variation in temperature dependence of photosynthesis.

Introduction

Temperature dependence of the light-saturated rate of photosynthesis shows a large interspecific variation. In general, plants inhabiting lower temperature regions have a lower optimal temperature that maximizes the photosynthetic rate (Mooney & Billings, 1961; Berry & Björkman, 1980). Cunningham & Read (2002) showed that not only the optimal temperature but also the shape of the curve differed between temperate and tropical species. Such differences have also been observed among populations of a single species. Several studies have reported that plants originated from high altitude or latitude had lower optimal temperatures than those from low altitude or latitude even when they were grown at the same temperature (Treharne & Eagles, 1970, Fryer & Ledig, 1972, Slatyer, 1977). These differences may contribute to adaptation to respective habitat temperature, and may be related to species differentiation.

Temperature dependence of photosynthesis is potentially affected by various biochemical factors (Hikosaka et al., 2006). It has been suggested that photosynthetic rates at high temperatures are related with thermal tolerance of photosynthetic proteins (Badger et al., 1982, Salvucci & Crafts-Brandner, 2004b; Wise et al., 2004; Yamori et al., 2006). Salvucci & Crafts-Brandner (2004a) showed that an Antarctic species had a lower optimal temperature for photosynthesis than temperate species as a result of inactivation of Rubisco (ribulose-1,5-bisphosphate carboxylase/oxygenase) at high temperatures. However, the Rubisco activation state is not necessarily responsible for the temperature dependence of photosynthesis in some other species (Cen & Sage, 2005). Several recent studies have indicated that the electron transport capacity is an important determinant for temperature dependence of photosynthesis, especially at higher temperatures (Wise et al., 2004), but its relation to inter- and intraspecific variation in temperature dependence of photosynthesis is not known.

Temperature dependence of photosynthesis potentially changes without alteration in the thermal tolerance of photosynthetic proteins. Photosynthetic rates at a normal CO2 concentration are limited either by RuBP (ribulose-1,5-bisphosphate) carboxylation or by RuBP regeneration (Farquhar et al., 1980, Kirschbaum & Farquhar, 1984). Since the temperature dependence of the two processes are different from each other, the temperature dependence of photosynthesis changes depending on which of the two steps limits photosynthesis (Hikosaka, 1997, Hikosaka et al., 1999, 2006). It is known that in some species the balance between carboxylation and regeneration of RuBP changes with growth temperature, leading to an alteration in temperature dependence of photosynthesis (Hikosaka et al., 1999, 2006; Onoda et al., 2005a,b; Yamori et al., 2005). Onoda et al. (2005b) found that the response of this balance to growth temperature is different among species: Polygonum cuspidatum, a perennial herb, altered the balance, whereas Fagus crenata, a deciduous tree, did not. Thus the balance between carboxylation and regeneration of RuBP is one of the causes of interspecific differences in the temperature dependence of photosynthesis. However, it is still unclear what kind of species characteristics are related to these differences, such as growth form, climate conditions of the habitat, or phylogeny.

In the present study we questioned what biochemical or physiological factors are related to intraspecific differences in the temperature dependence of photosynthesis. Hypothesizing that such differences are a result of adaptation to the climate conditions of the habitat, we studied ecotypes of Plantago asiatica from different latitudes, where mean annual temperature differs from 7.5 to 16.8°C. Gas exchange characteristics were analysed according to the biochemical model of Farquhar et al. (1980). Changes in temperature dependence of photosynthesis are attributable to changes in the following four traits: (i) internal CO2 concentration; (ii) temperature dependence of the maximum rate of RuBP carboxylation (Vcmax); (iii) temperature dependence of the maximum rate of RuBP regeneration, expressed as the rate of electron transport (Jmax); and (iv) the ratio of Jmax to Vcmax (Hikosaka et al., 2006). If the ecotypic difference in the temperature dependence of photosynthesis is caused by the thermal tolerance of photosynthetic proteins, a difference will be found in (ii) or (iii). If it is related with the balance between carboxylation and regeneration of RuBP, a difference in the Jmax : Vcmax ratio will be found.

Materials and Methods

Plantago asiatica L., a perennial herb, is widely distributed in Japan as well as in Taiwan and China. This species is known to have a large genetic variation across latitudinal gradient (Sawada et al., 1994, Ishikawa et al., 2006). In 2003, living individuals of P. asiatica that had grown at a sunny area in Tomakomai (42°37′-N, 141°32′-E), Sendai (38°15′, 140°50′), and Shimada (34°51′, 138°05′), Japan were collected. The annual average of daily, minimum and maximum temperatures was 7.5, 3.6 and 11.4 in Tomakomai, 12.5, 8.9 and 16.5 in Sendai, and 16.8, 12.7 and 21.2 in Shimada, respectively (2001–05, Japan Meteorological Agency). Individuals were transplanted into 1.5 l pots filled with washed river sand (one plant per pot). Plants were grown at the experimental garden of Tohoku University (Sendai) until the experiment started. During the experiment, plants were grown in growth cabinets, in which air temperature was 15 or 30°C. These two growth temperatures have been used in previous studies for temperature acclimation (Hikosaka et al., 1999, Hikosaka, 2005; Yamori et al., 2005, 2006). A temperature of 30°C is equivalent to the highest air temperature in Tomakomai (30.3°C in 2001–05); 15°C was selected because Plantago asiatica plants in Sendai shed their leaves in November, where the mean daily temperature becomes lower than 15°C. Light (16 h) was provided by fluorescent lamps (50 µmol m−2 s−1 above the plant) and additionally by halogen lamps (total 700 µmol m−2 s−1) for 4 h. Relative humidity was 80%. Plants received 100 ml of nutrient solution that contained 0.3 mol N, 0.23 mol P, 0.09 mol K (Hyponex, Murakami Bussan, Osaka, Japan) every week. Two growth cabinets were used and their temperature condition periodically rotated, in order to avoid any chamber effects.

Four plants per growth temperature per ecotype were used. Photosynthetic rates were determined for c. 30-d-old leaves, which newly emerged in the growth cabinets, with a portable gas exchange system (LI-6400, LiCor Inc., Lincoln, NE, USA). For each leaf, the CO2-response curve of the photosynthetic rate was obtained at 10, 15, 20, 25, 30, 35 and 40°C leaf temperature. For each leaf temperature, the photosynthetic rates were determined at 370, 220, 100, 50, 550, 750, 1000, 1500 and 2000 µmol CO2 mol−1. Photon flux density was 1750 µmol m−2 s−1 and vapour pressure deficit was < 1 kPa at 10–30°C and < 2 kPa at 35 and 40°C. After photosynthetic measurements, leaf discs of 1 cm diameter were punched out and stored at –85°C until biochemical analyses. Fully activated activity of stroma fructose-1,6-bisphosphatase (FBPase) and Rubisco content were determined after Hikosaka (2005).

The temperature dependence of the photosynthetic rate at a CO2 concentration of 370 µmol mol−1 was fitted by the following curve (Cunningham & Read, 2002):

image(Eqn 1)

(A, photosynthetic rate; Tk, leaf temperature in Kelvin; inline image and inline image, minimum and maximum temperatures at which the photosynthetic rate is zero; b and c, fitting parameters). Curve-fitting was performed with Kaleidagraph (Synergy Software, Reading, PA, UK). A data set of a leaf was not used if there was no convergence of the curve. From the curve, we numerically obtained Tmin and Tmax, the temperatures below and above the optimum that realize 80% of maximum photosynthetic rate, respectively.

The CO2 dependence curve of photosynthesis was fitted with the biochemical model of Farquhar et al. (1980). At lower CO2 concentrations (< 300 µmol mol−1), the following curve was applied assuming that the photosynthetic rate is limited by RuBP carboxylation:

Ac = (Vcmax (Ci–Γ*)/( Ci + Kc (1 + O/Ko))) –Rd(Eqn 2)

(Ac, carboxylation-limited photosynthetic rate; Vcmax, maximal velocity of RuBP carboxylation; Kc and Ko, Michaelis–Menten constants of Rubisco for CO2 and O2, respectively; Ci and O, intercellular partial pressures of CO2 and O2, respectively; Γ*, CO2 compensation point in the absence of day respiration; Rd is the rate of day respiration). At higher CO2 concentrations (> 500 µmol mol−1), the following curve was applied assuming that the photosynthetic rate is limited by RuBP regeneration:

Ar = (Jmax (Ci–Γ*)/(4Ci+ 8Γ*)) –Rd(Eqn 3)

(Ar, regeneration-limited rate of photosynthesis; Jmax, maximum rate of RuBP regeneration expressed as the rate of electron transport). The temperature dependence of the parameters (f) was fitted using the Arrhenius model if there was an exponential increase with temperature.

f = f(25) exp(Ea (Tk–298)/298RTk)   (Eqn 4)

(f(25), value of f at 25°C; Ea, activation energy of f; R, universal gas constant (8.314 J mol−1 K−1)). If there was a considerable deactivation at high temperatures, a peak model was applied (von Caemmerer, 2000; Medlyn et al., 2002a):

g = g(25) exp(Ha (Tk – 298)/298RTk) × (1 + exp((298ΔS   – Hd)/298R))/(1 + exp((T kDaS – Hd)/RTk)) (Eqn 5)

(g(25), value of g at 25°C; Ha, activation energy of g; ΔS, entropy term; Hd; deactivation energy). We assumed 200 kJ mol−1 for Hd (Medlyn et al., 2002a).

It is known that Rubisco kinetic parameters are different among species (Galmés et al., 2005) and many studies have presented various sets of Rubisco kinetics (Harley & Baldocchi, 1995, Bernacchi et al., 2001). A combination of Rubisco kinetics parameters was sought that well explained the CO2-response curve of photosynthesis in P. asiatica leaves. It was assumed that Kc, Ko and Γ* at 25°C are 404.9 µmol mol−1, 278.4 mmol mol−1 and 37 µmol mol−1, respectively, and the activation energies (Ea) of Kc, Ko and Γ* are 79.43, 36.38 and 23.4 kJ mol−1, respectively,.

Data sets were obtained from seven leaves of four plants (one or two leaves per plant) per growth temperature per site. When two leaves were used for an individual plant, the average of the values were used for the statistical analysis (i.e. four replications). Similar results were obtained if values of each individual leaf were used. Statistical analysis was performed with JMP (SAS Institute Inc., Cary, NC, USA).

Results

The temperature dependence of the photosynthetic rates at 370 µmol CO2 mol−1 was a parabolic curve with a maximum (Fig. 1). The optimal temperature that maximized the photosynthetic rate (Topt) was significantly higher in leaves grown at high temperature but was not different among ecotypes (Tables 1, 2). A similar tendency was found in Tmax, the maximum temperature that realizes 80% of the maximum photosynthetic rate (Tables 1, 2). However, Tmin, the minimum temperature that realizes 80% of the maximum, was significantly different not only between growth temperatures but also among ecotypes (Table 1). There was no clear difference in Tmin among ecotypes when plants were grown at 30°C, while Tmin was significantly lower in Tomakomai than in Shimada plants when plants were grown at 15°C (Table 2).

Figure 1.

Temperature dependence of the photosynthetic rate at 370 µmol CO2 mol−1 in three Plantago asiatica ecotypes grown at 15°C (a) and 30°C (b). Fitted curves are detailed in Eqn 1 (see text) using mean values of T′k,min and T′k,max. Symbols and curves are for plants from Tomakomai (squares and continuous line), Sendai (circles and dotted line) and Shimada (triangles and dashed line). Bar, mean of standard deviations.

Table 1.  Effects of ecotype, growth temperature and their interaction on photosynthetic characteristics of Plantago asiatica
 d.f.ToptTminTmaxA370(15)Ci(15)Vcmax(15)Jmax(15)J:V(15)EaVHaJRubiscoFBPaseF : R
  1. Results of the generalized linear model analysis are shown. Normal distribution was assumed in every case. χ2-values with their probabilities are shown: ***, P < 0.001; **, P < 0.01; *, P < 0.05.

  2. Topt, optimal temperature that maximizes photosynthetic rate; Tmin and Tmax, minimum and maximum temperatures where photosynthetic rate is 80% of maximum; A370(15), photosynthetic rate at 15°C under 370 µmol mol−1 CO2; Ci(15), intercellular CO2 concentration; Vcmax(15), maximum rate of carboxylation at 15°C; Jmax(15), maximum rate of electron transport at 15°C; J : V(15), the ratio of Jmax to Vcmax at 15°C; EaV, activation energy of Vcmax; HaJ, activation energy of Jmax; Rubisco, Rubisco content; FBPase, fully activated activity of stroma FBPase; F : R, the ratio of FBPase activity to Rubisco content. n = 3–4.

Ecotype2 0.1720.5*** 2.4 7.4*1.1 7.8* 4.77.8* 2.3 1.8 5.9 3.59.4**
Growth temperature112.4***38.7***25.8***12.2***0.0125.8***27.5***1.829.2***13.8***10.8**12.8***0.03
E × G2 0.4810.1** 4.0 2.10.06 1.44 0.326.73* 1.2 1.3 0.53 6.03.5
Table 2.  Photosynthetic characteristics of three Plantago asiatica ecotypes grown at 15 and 30°C
 15°C-grown30°C-grown
TomakomaiSendaiShimadaTomakomaiSendaiShimada
  1. Mean ± SD is shown (n = 3–4). Different superscript letters indicate statistical difference between ecotypes in each growth temperature at 5% level (Tukey–Kramer test).

  2. Topt, optimal temperature that maximizes photosynthetic rate; Tmin and Tmax, minimum and maximum temperatures where photosynthetic rate is 80% of maximum; A370(15), photosynthetic rate at 15°C under 370 µmol mol−1 CO2; Ci(15), intercellular CO2 concentration; Vcmax(15), maximum rate of carboxylation at 15°C; Jmax(15), maximum rate of electron transport at 15°C; EaV, activation energy of Vcmax; HaJ, activation energy of Jmax; Rubisco, Rubisco content; FBPase, fully activated activity of stroma FBPase.

Topt (°C)29.8 ± 0.5 a30.2 ± 0.9 a29.3 ± 2.1 a32.3 ± 1.0 a32.3 ± 2.9 a32.5 ± 1.4 a
Tmin (°C) 2.9 ± 5.0 a 9.5 ± 1.8 ab14.6 ± 1.2 b17.9 ± 3.9 a20.5 ± 1.0 a20.7 ± 2.5 a
Tmax (°C)38.4 ± 0.8 a38.4 ± 0.5 a37.5 ± 1.8 a38.8 ± 0.7 a40.7 ± 1.0 b40.4 ± 2.0 b
A370(15) (µmol m−2 s−1)12.4 ± 2.0 a15.8 ± 4.0 a16.4 ± 1.4 a10.2 ± 1.8 a11.5 ± 1.4 a11.5 ± 1.9 a
Ci(15) (µmol mol−1) 261 ± 18 a 256 ±9 a 278 ± 23 a 265 ± 15 a 269 ±5 a 259 ± 13 a
Vcmax(15) (µmol m−2 s−1)32.1 ± 2.9 a39.7 ± 7.8 a39.3 ± 4.7 a23.3 ± 5.2 a27.6 ± 2.3 a25.6 ± 3.5 a
Jmax(15) (µmol m−2 s−1) 103 ± 7.2 a 114 ± 17.4 a 101 ± 10.4 a64.5 ± 17.7 a78.7 ± 11.6 a69.4 ± 14.7 a
EaV (kJ mol−1)48.6 ± 0.1 a44.9 ± 2.2 a48.8 ± 4.0 a60.0 ± 7.2 a58.1 ± 4.7 a58.0 ± 4.0 a
HaJ (kJ mol−1)33.9 ± 6.0 a34.5 ± 2.2 a38.7 ± 4.8 a44.5 ± 3.8 a42.5 ± 3.6 a44.1 ± 8.5 a
Rubisco (g m−2)2.85 ± 0.21 a3.64 ± 0.41 a3.29 ± 0.91 a2.19 ± 0.79 a2.67 ± 0.35 a2.65 ± 0.49 a
FBPase (µmol m−2 s−1)42.0 ± 4.2 a46.7 ± 7.0 ab34.1 ± 4.7 b29.0 ± 6.4 a31.5 ± 8.4 a33.0 ± 8.3 a

The photosynthetic rate under 370 µmol CO2 mol−1 (A370) at 15°C was significantly different depending both on growth temperature and on ecotype (Table 1). It was higher when plants were grown at 15°C and tended to be low in Tomakomai plants (Table 2). The intercellular CO2 concentration at 15°C determined under a normal CO2 concentration (370 µmol mol−1) did not vary significantly (Table 1).

Figure 2 shows the temperature dependence of Vcmax and Jmax. Vcmax increased exponentially with increasing leaf temperature and the Arrhenius model fitted well. In Tomakomai plants, however, there was a deviation from the Arrhenius model at a leaf temperature of 40°C, suggesting thermal deactivation. We calculated the activation energy of Vcmax in Tomakomai plants, excluding the data from 40°C. Jmax also increased exponentially at low temperatures but there was an optimum at 30–40°C, and thus we applied a peak model.

Figure 2.

Temperature dependence of the maximum carboxylation rate (Vcmax) (a, b) and electron transport rate (Jmax) (c, d) in three Plantago asiatica ecotypes grown at 15°C (a, c) and 30°C (b, d). Fitted curves are detailed in Eqns 4 and 5 for Vcmax and Jmax, respectively, using mean values of Vcmax(25), Jmax(25) and their activation energy (EaV and HaJ). Symbols and curves are for plants from Tomakomai (squares and continuous line), Sendai (circles and dotted line) and Shimada (triangles and dashed line). Bar, mean of standard deviations.

The activation energy of Vcmax (EaV) was significantly higher in leaves grown at 30°C than those grown at 15°C, but not different among ecotypes (Tables 1, 2). Similar results were obtained for the activation energy of Jmax (HaJ). Vcmax determined at 15°C was higher in 15°C-grown leaves and tended to be low in Tomakomai plants and high in Sendai (Tables 1, 2). Jmax at 15°C was also higher in 15°C-grown leaves but there was no significant difference among ecotypes (Tables 1, 2). The Jmax : Vcmax ratio at 15°C was significantly different among ecotypes only when plants were grown at 15°C (Fig. 3); it was highest in Tomakomai plants and lowest in Shimada plants.

Figure 3.

Jmax : Vcmax ratio (a, b) and FBPase (fructose-1,6-bisphosphatase) to Rubisco (ribulose-1,5-bisphosphate carboxylase) ratio (c, d) in three Plantago asiatica ecotypes grown at 15°C (a, c) and 30°C (b, d). Mean + SD (n = 4). Different letters indicate statistical difference between ecotypes at 5% level (Tukey–Kramer test).

The fully activated stroma FBPase activity, assumed as the measure of FBPase content, and the Rubisco content tended to be higher in leaves grown at 15°C than those grown at 30°C (Tables 1, 2). When plants were grown at 15°C, the FBPase activity was lowest in Shimada plants. As in the Jmax : Vcmax ratio, the ratio of FBPase activity to Rubisco content (F : R ratio) was not significantly different when plants were grown at 30°C, but it was higher in Tomakomai than in Shimada plants when plants were grown at 15°C (Fig. 3).

CO2 dependence of photosynthesis in plants grown at 15°C was further analysed to examine the limiting step of photosynthesis after Atkin et al. (2006). Figure 4(a–c) show the ratio of A550 (photosynthetic rate at 550 µmol CO2 mol−1) to A370. Continuous lines denote the ratio of Ac (the carboxylation-limited rate of photosynthesis) at 550 µmol CO2 mol−1 (Ac550) to Ac370, which was theoretically obtained (Eqn 2), and dotted lines denote the ratio of Ar (the regeneration-limited rate of photosynthesis) at 550 µmol CO2 mol−1 (Ar550) to Ar370 (Eqn 3). Similarly, the A370 : A220 ratios are shown in Fig. 4(d–f). If photosynthesis is limited by triose-phosphate utilization (TPU), the ratio becomes 1 (Sage, 1990). In the present study, measured values of the A550 : A370 and the A370 : A220 ratio were always higher than 1. This was also the case at higher CO2 concentrations (data not shown), suggesting that the TPU limitation did not occur in the present study. In Tomakomai plants, measured values of the A550 : A370 and the A370 : A220 ratios were very close to the continuous line except for the A550 : A370 ratio at 35 and 40°C (Fig. 4a,d). This suggests that, below 35°C, photosynthesis in Tomakomai plants was limited by the RuBP carboxylation between 220 and 550 µmol CO2 mol−1. Similar results were obtained for Sendai plants, but the A550 : A370 ratio was close to the dotted line at 10°C (Fig. 4b,e). This suggests that, at 10°C in Sendai plants, the RuBP regeneration-limited photosynthesis is above 370 µmol CO2 mol−1, while the RuBP carboxylation-limited rate is below 370 µmol CO2 mol−1. In Shimada plants, measured values of the A550 : A370 ratio tended to be lower than the continuous line at any temperature (Fig. 4c), suggesting that the RuBP regeneration limited photosynthesis at higher CO2 concentrations. At 10°C, not only the A550 : A370 but also the A370 : A220 ratios were close to dotted line. This suggests that photosynthesis at 10°C was limited by the RuBP regeneration even below 370 µmol CO2 mol−1 in Shimada plants.

Figure 4.

(a–f) CO2 dependence of photosynthetic rates in three Plantago asiatica ecotypes grown at 15°C. Closed circles with bars denote mean of measured values with standard deviations. The ratio of photosynthetic rate at 550 to that at 370 µmol CO2 mol−1 (A550 : A370) (a, Tomakomai; b, Sendai; c, Shimada) and the ratio of photosynthetic rate at 370 to that at 220 µmol CO2 mol−1 (A370 : A220) (d, Tomakomai; e, Sendai; f, Shimada) are shown. Lines denote the ratio of photosynthesis assuming the limitation by ribulose-1,5-bisphosphate (RuBP) carboxylation (continuous line, see Eqn 2 in the text) or the limitation by RuBP regeneration (dotted line, see Eqn 3). A mean of measured Ci (intercellular partial pressure of CO2) values was used for the calculation at each leaf temperature. (g–i) Simulated temperature dependence of photosynthetic activities in three P. asiatica ecotypes grown at 15°C (g, Tomakomail; h, Sendai; i, Shimada). Dotted and continuous lines are simulated temperature dependence of RuBP regeneration-limited (Ar370) and RuBP carboxylation-limited (Ac370) rates of photosynthesis at 370 µmol CO2 mol−1, respectively,. Simulated curves are calculated using mean values of Ci (µmol mol−1), Vcmax(25) (µmol m−2 s−1), Jmax(25) (µmol m−2 s−1), their activation energy (EaV and HaJ) (kJ mol−1), and the entropy term (ΔS) for Tomakomai (293, 69.8, 159, 48.6, 33.9, 639), Sendai (288, 85.4, 182, 44.9, 34.5, 639), and Shimada (300, 80.6, 162, 48.8, 38.7, 640).

Using the model of Farquhar et al. (1980) with obtained parameter values, we reconstructed the temperature-response curve of photosynthesis of 15°C-grown plants (Fig. 4g–i). The photosynthetic rates are given as the minimum of the RuBP carboxylation-limited rate (Ac) and the RuBP-regeneration limited rate (Ar). In Tomakomai plants, Ac was lower than Ar at every temperature, indicating that photosynthesis was always limited by RuBP carboxylation. In Shimada plants, Ac and Ar crossed at 15°C; photosynthesis was limited by RuBP regeneration below 15°C and by RuBP carboxylation at other temperatures. Sendai plants were intermediate between Tomakomai and Shimada plants. These results are consistent with the results of the CO2-dependence analyses (Fig. 4a–f).

Discussion

Difference in temperature dependence of photosynthesis among ecotypes

We found significant intraspecific variations in temperature dependence of photosynthesis among populations. While there was no significant difference in Topt and Tmax among ecotypes, Tmin was different when plants were grown at 15°C (Tables 1, 2, Fig. 1). This difference can be explained by the limiting step of photosynthesis. In Shimada plants, photosynthesis below 15°C was limited by the RuBP regeneration process, while in Tomakomai plants it was always limited by RuBP carboxylation (Fig. 4). Since the temperature dependence of Ac is smaller than that of Ar below 15°C, Tomakomai plants maintained relatively high photosynthetic rates at low temperatures and had a lower Tmin.

Differences in the limiting step of photosynthesis among ecotypes are attributable to nitrogen partitioning within the photosynthetic apparatus (Hikosaka, 1997, 2004, 2005; Hikosaka et al., 2006). Tomakomai plants had higher Jmax : Vcmax and F : R ratios than Shimada plants (Fig. 3), suggesting that they invested more nitrogen in the RuBP regeneration process (electron transport, ATP synthesis and Calvin cycle other than RuBP carboxylation) than in Rubisco.

What is the ecological significance of the among-population difference in the temperature response of photosynthesis? In Shimada plants, at a leaf temperature of 15°C, Ac370 was very close to Ar370 (Fig. 4i), indicating that the photosynthetic rate was colimited by carboxylation and regeneration of RuBP. This is optimal in terms of nitrogen use in the photosynthetic apparatus because no proteins are excessive for maintaining photosynthetic rates (Hikosaka & Terashima, 1995, Hikosaka, 1997). In Tomakomai plants, on the other hand, proteins related to RuBP regeneration may be somewhat over-invested at 15°C (Fig. 4g). However, a higher capacity of RuBP regeneration may contribute to the tolerance to low-temperature stress. At lower temperatures, photosystem II is more susceptible to high light (Huner et al., 1998, Tsonev & Hikosaka, 2003), leading to decreases in RuBP regeneration capacity and thus in photosynthesis. If the RuBP regeneration capacity is enhanced, a small decrease in RuBP regeneration may not lead to a decrease in photosynthetic rates.

Recently it has been suggested that photosynthesis at high temperatures is limited by the RuBP regeneration (Wise et al., 2004). The present results on CO2-dependence and model simulation suggest, however, that A370 was limited by the RuBP carboxylation at 40°C (Fig. 4). As Jmax abruptly decreased with increasing leaf temperature above 35°C (Fig. 2), we might observe a limitation by the RuBP regeneration if we analysed photosynthesis above 45°C (Fig. 4).

Tomakomai plants showed a thermal deactivation of RuBP carboxylation at 40°C (Fig. 2). This is in accordance with previous studies showing that mild heat stress decreases the Rubisco activation state (Salvucci & Crafts-Brandner, 2004a,b; Yamori et al., 2006). Salvucci & Crafts-Brandner (2004a) showed that the optimal temperature of Rubisco activation in Antarctic hairgrass (Deschampsia antarctica) was 10°C lower than that in some thermal-tolerant species such as cotton. This difference was ascribed to the difference in heat stability of Rubisco activase (Salvucci & Crafts-Brandner, 2004a). Thus, Rubisco activase might be unstable in Tomakomai plants at 40°C, although its effect on the temperature dependence of photosynthesis was negligible (Table 1). On the other hand, Shimada and Sendai plants did not exhibit deactivation of RuBP carboxylation at 40°C (Fig. 2), which may be related to their success in high temperature regimes.

Other variables that are potentially responsible for the temperature dependence of photosynthesis, Ci, EaV and HaJ, were not significantly different among ecotypes (Table 1). Ci may be related to environmental factors other than temperature, such as water stress. Although EaV and HaJ are known to have large interspecific variations (Medlyn et al., 2002b; Hikosaka et al., 2006), its ecological significance is not known. Identical EaV and HaJ values among three ecotypes would suggest that the interspecific variation in these traits is constrained by phylogeny rather than by adaptation to the habitat environment.

Difference in acclimation potential among ecotypes

It is well known that the temperature-response curve of photosynthesis changes depending on the growth temperature. In most species the optimal temperature of photosynthesis increases with increasing growth temperature (e.g. Berry & Björkman, 1980, Hikosaka et al., 2006). P. asiatica also increased Topt from 29.3 to 32.4°C irrespective of ecotypes (Table 2, Fig. 1). This change is ascribed to the change in the temperature dependence of Rubisco activity because photosynthesis around the Topt was limited by RuBP carboxylation in all cases (Fig. 4). It was found that the activation energy of Vcmax (EaV) increased with increasing growth temperature for all ecotypes (Table 2). An increase in EaV increases the carboxylation rate more at higher temperatures, leading to an increase in Topt (Hikosaka et al., 2006). From a literature survey, Hikosaka et al. (2006) found that it is a general response of C3 plants to increase EaV in response to an increase in growth temperature.

It is also known that phenotypic plasticity in temperature dependence of photosynthesis differs among species. Cunningham & Read (2002) studied the dependence of Topt on growth temperature in four temperate and four tropical species. In one of the temperate species (Eucryphia lucida), the Topt was independent of growth temperature. The other seven species showed a significant dependence but the slope of Topt on the growth temperature differed among species from 0.10 to 0.48°C °C−1. In the present study, there was no difference in Topt among ecotypes (Tables 1 and 2), but significant interactive effects of ecotype and growth temperature were found on the Jmax : Vcmax ratio (Table 1). Jmax : Vcmax and F : R ratios were significantly different among ecotypes when plants were grown at 15°C but not different at 30°C (Fig. 3). Furthermore, Tomakomai plants increased the Jmax : Vcmax ratio by 17% with decreasing growth temperature while Shimada plants did not alter this ratio. These results indicate that phenotypic plasticity was different among ecotypes.

It has been shown that the response of the Jmax : Vcmax ratio to growth temperature differs among species: the Jmax : Vcmax ratio increased with increasing growth temperature in Quercus myrsinaefolia (Hikosaka et al., 1999), Polygonum cuspidatum (Onoda et al., 2005a) and Spinacia oleracea (Yamori et al., 2005), but not in Pinus pinaster (Medlyn et al., 2002a), Fagus crenata (Onoda et al., 2005b) and Quercus crispula (Hikosaka et al., 2007). Recently Atkin et al. (2006) showed that two lowland Plantago species (P. lanceolata and P. major) altered the Jmax : Vcmax ratio depending on growth temperature, but an alpine congeneric, Plantago euryohylla, did not. In the present study it was found that even within a species, the response of the Jmax : Vcmax ratio differed among ecotypes. From a viewpoint of acclimation potential, our result is contrasting to that of Atkin et al. (2006): the ecotype from a habitat with lower temperature had more plasticity to growth temperature in our study (Fig. 3). From a viewpoint of the balance in the photosynthetic apparatus, on the other hand, results of the two studies are consistent: in plants from low-temperature conditions (high altitude or latitude), photosynthesis tended to be limited by RuBP carboxylation at any temperature, while in plants from high-temperature conditions there was a colimitation of carboxylation and regeneration of RuBP at a certain temperature (Fig. 4). Thus high Jmax : Vcmax ratios may be important for surviving under low-temperature conditions.

The present study showed that nitrogen partitioning in the photosynthetic apparatus and its response to growth temperature were different among ecotypes. These differences altered the temperature dependence of photosynthesis depending on ecotypes, and may contribute to adaptation to respective habitat environment in terms of efficient use of nitrogen or stress tolerance. Such genetic differentiation may enable this species to spread its distribution across a large range of habitat temperatures.

Acknowledgements

We thank R. Oguchi, Y. Yasumura, H. Yasumura, J. Yasumura and N. Onoda for plant collection; K. Sato and I. Nakamura for technical support; and O. Muller for helpful comments. This study was supported in part by grants from the Japan Ministry of Education, Culture, Sports, Science and Technology and by the Global Environment Research Fund (F-052) from the Japan Ministry of the Environment.

Ancillary