Low-temperature photosynthetic performance of a C4 grass and a co-occurring C3 grass native to high latitudes


  • D. S. KUBIEN,

    Corresponding author
      David S. Kubien. Fax: +64 6350 5688; e-mail: dkubien@massey.ac.nz
    Search for more papers by this author
    • *

      Current address: Institute of Molecular Biosciences, Massey University, Private Bag 11 222, Palmerston North, New Zealand.

  • R. F. SAGE

    1. Department of Botany, University of Toronto, 25 Willcocks St., Toronto, ON, M5S 3B2, Canada
    Search for more papers by this author

David S. Kubien. Fax: +64 6350 5688; e-mail: dkubien@massey.ac.nz


The photosynthetic performance of C4 plants is generally inferior to that of C3 species at low temperatures, but the reasons for this are unclear. The present study investigated the hypothesis that the capacity of Rubisco, which largely reflects Rubisco content, limits C4 photosynthesis at suboptimal temperatures. Photosynthetic gas exchange, chlorophyll a fluorescence, and the in vitro activity of Rubisco between 5 and 35 °C were measured to examine the nature of the low-temperature photosynthetic performance of the co-occurring high latitude grasses, Muhlenbergia glomerata (C4) and Calamogrostis canadensis (C3). Plants were grown under cool (14/10 °C) and warm (26/22 °C) temperature regimes to examine whether acclimation to cool temperature alters patterns of photosynthetic limitation. Low-temperature acclimation reduced photosynthetic rates in both species. The catalytic site concentration of Rubisco was approximately 5.0 and 20 µmol m−2 in M. glomerata and C. canadensis, respectively, regardless of growth temperature. In both species, in vivo electron transport rates below the thermal optimum exceeded what was necessary to support photosynthesis. In warm-grown C. canadensis, the photosynthesis rate below 15 °C was unaffected by a 90% reduction in O2 content, indicating photosynthetic capacity was limited by the capacity of Pi-regeneration. By contrast, the rate of photosynthesis in C. canadensis plants grown at the cooler temperatures was stimulated 20–30% by O2 reduction, indicating the Pi-regeneration limitation was removed during low-temperature acclimation. In M. glomerata, in vitro Rubisco activity and gross CO2 assimilation rate were equivalent below 25 °C, indicating that the capacity of the enzyme is a major rate limiting step during C4 photosynthesis at cool temperatures.


C4 plants often dominate warm, dry climates where they have access to at least moderate light intensities (Sage, Wedin & Li 1999). However, the frequency of C4 species is reduced in cool climatic zones (Long 1983; Sage et al. 1999), as shown by a lower proportion of C4 grasses at higher latitude (Teeri & Stowe 1976) and elevation (Tieszen et al. 1979). Globally, C4 grasses become rare in regions in which the minimum temperature of the warmest month of the growing season is less than 8 °C (Teeri & Stowe 1976; Long 1983), and the average growing season temperatures are below 16 °C (Sage et al. 1999).

Numerous hypotheses for the rarity of C4 plants in cool climates have been proposed. Cold-induced failure of enzymes in the C4 pathway has frequently been suggested as the leading cause of the poor performance of C4 photosynthesis at low temperature (Long 1983; Edwards et al. 1985; Long 1999; Naidu et al. 2003). Several enzymes of the carbon-concentrating mesophyll reactions, notably phosphoenolpyruvate carboxylase (PEPCase, EC and pyruvate orthophosphate dikinase (PPDK, EC, can be cold labile in vitro, with dissociation occurring below approximately 10 °C (Sugiyama 1973; Krall & Edwards 1993). These enzymes are not inherently prone to failure at low temperatures, however, as there is considerable species and ecotypic variation in their thermal stability (Sugiyama & Boku 1976; Krall & Edwards 1993; Simon & Hatch 1994). Furthermore, studies have demonstrated that C4 metabolism is qualitatively unaltered by low-temperature exposure (Caldwell, Osmond & Nott 1977; Matsuba et al. 1997).

An alternative explanation for the poor performance of C4 photosynthesis at low temperature is that low Rubisco capacity limits CO2 assimilation at suboptimal temperatures. As used here, Rubisco ‘capacity’ refers to the capacity of fully activated Rubisco to consume RuBP under a given set of environmental conditions. In the context of this definition, Rubisco capacity largely reflects Rubisco content (Sage, Sharkey & Seemann 1990). In the C4 dicot Atriplex rosea, the activation energy of photosynthesis in vivo and Rubisco in vitro are similar (Björkman & Pearcy 1971). Differences in the photosynthesis rates at low temperatures in Atriplex lentiformis grown at different temperatures were correlated with changes in the maximum activity of Rubisco (Pearcy 1977). Pittermann & Sage (2000, 2001) observed that the rate of net CO2 assimilation was equivalent to the maximum activity of Rubisco in vitro in the high elevation species Bouteloua gracilis and Muhlenbergia montanum at suboptimal temperatures. Recent work with anti-RbcS Flaveria bidentis has shown that Rubisco capacity, specifically the amount of the enzyme, is an important controlling step over the rate of carbon gain in C4 plants at low temperatures (Kubien et al. 2003).

Relatively few studies have examined photosynthetic acclimation to low growth temperatures in C4 species that naturally occur in cool climates. C3 species show a variety of low-temperature acclimation patterns, including increases in the amount of Rubisco (Badger, Björkman & Armond 1982; Holaday et al. 1992; Hurry et al. 1995) and the enzymes of sucrose synthesis (Martindale & Leegood 1997; Savitch, Gray & Huner 1997; Stitt & Hurry 2002). The temperature acclimation responses of carbon metabolism in C4 plants are less clear. When grown at low temperature, Atriplex lentiformis increases its Rubisco content relative to warm-grown plants (Pearcy 1977). In chilling-sensitive maize the amount of Rubisco is reported to either decline or be insensitive to growth temperature (Ward 1987; Pietrini et al. 1999; Naidu et al. 2003). The chilling-tolerant C4 grass Miscanthus × giganteus maintains high photosynthetic rates when grown at low temperatures because of the ability to maintain high levels of Rubisco and PPDK; levels of these enzymes decline when chilling-intolerant maize is grown at 14 °C relative to plants grown at 25 °C (Naidu et al. 2003). The existence of a few C4 species in cool climates (Schwarz & Redmann 1988; Long 1999; Kubien & Sage 2003) indicates that C4 plants can adapt to low-temperature environments, but the reasons for their apparent inability to compete with C3 species in cool-climates remain uncertain.

In the present study, we examine the nature of the rate limitations of C3 and C4 photosynthesis at low temperatures. Specifically, we test the hypotheses that Rubisco capacity limits C4 photosynthesis at low temperatures. To address this question, the photosynthetic temperature response of the C4 grass Muhlenbergia glomerata is contrasted with that of the C3 species Calamogrostis canadensis. These species frequently co-occur in boreal Canada and have wide latitudinal ranges; M. glomerata has been reported above 60°N (Schwarz & Redmann 1988). This approach allows for direct comparisons between C3 and C4 species that occupy the same cool climates. Further, using M. glomerata as a representative C4 grass mitigates the effects of general chilling intolerance that are often associated with warm-climate C4 species, such as maize. Warm (26/22 °C) and cool (14/10 °C) growth temperature regimes were used to examine photosynthetic acclimation in these species. We measured photosynthetic gas exchange and chlorophyll a fluorescence parameters at temperatures from 5 to 35 °C, and in vitro Rubisco activity from 0 to 35 °C.


Plant growth

Rhizomes of Muhlenbergia glomerata (C4) and Calamogrostis canadensis (C3) were collected from a fen near Plevna, Ontario (45°0′ N 76°5′ W). Rhizomes were planted in 6 L pots containing 66% (v/v) Promix (Plant Products, Brampton, Canada), 17% sand and 17% plant compost. Plants were grown in two controlled environment chambers (GC-20; Enconair, Winnipeg, Canada) and maintained under a 16 h photoperiod with a maximum photosynthetic photon flux density (PPFD) of 800 µmol m−2 s−1. Plants were grown at day/night temperatures and relative humidities of 14/10 °C and 50/80% (cool-grown) or 26/22 °C and 70/80% (warm-grown), respectively. Plants were grown for 10 weeks prior to the beginning of the experiment, and growth regimes were rotated between two chambers every 2 to 3 weeks to minimize between-chamber variation. Plants were watered daily and fertilized weekly with 0.5× Hoagland's solution supplemented with 3 mm NH4NO3.

Gas-exchange measurements

The temperature response of photosynthesis was measured in an open-type leaf gas-exchange system described elsewhere (Kubien et al. 2003); all gas exchange calculations follow von Caemmerer & Farquhar (1981). Leaf temperature was measured by placing three fine wire thermocouples in contact with the abaxial surface of the leaves. Illumination was provided by a cool-light source (KL-2500; Schott, Mainz, Germany) to minimize interference with the fluorescence detector. The PPFD in the cuvette was measured using a photodiode (G1738; Hamamatsu, Bridgewater, NJ, USA) calibrated against a quantum sensor (Li-190s: LiCor Inc., Lincoln, NE, USA). Temperature responses were measured at a constant PPFD of 1300 µmol m−2 s−1, which is sufficient to saturate photosynthesis in both species at temperatures less than 25 °C (Kubien 2003). Temperature responses were measured at 200 ± 10 mbar O2 in M. glomerata and C. canadensis, and also at 20 ± 1 mbar O2 in the C3 species; CO2 was maintained at 370 ± 3 µbar. The leaf-to-air vapour pressure deficit (VPD) was maintained at 14 ± 2 mbar at temperatures greater than 10 °C; at cooler temperatures VPD was reduced.

Photosynthesis was measured simultaneously on two to four leaves from a single plant, and different plants were used for each replicate. All temperature response measurements were initiated at 25 °C and 200 mbar O2. Leaves were dark-adapted for 30 min before the measurement of the respiration rate (R) and Fv/Fm. This R-value was used to correct the measured net CO2 assimilation rate for respiration, following Bernacchi et al. (2001). After the respiration measurement, the leaves were allowed to equilibrate at the measurement PPFD and temperature for a minimum of 45 min before measurement of steady-state photosynthesis and fluorescence. Leaf temperature was subsequently increased in 5 °C steps to 35 °C, and then decreased to the lower temperatures; the decrease from 35 to 20 °C was performed in 5 °C steps of approximately 10 min each. At each temperature, the leaves were allowed to equilibrate for a minimum of 20 min prior to measurements. For the C3 species, the partial pressure of O2 was reduced to 20 mbar after the initial measurement at 200 mbar; photosynthesis and fluorescence were measured after a 10 min adjustment to reduced O2, at all temperatures. After the last measurement was completed, the leaves were warmed to about 15 °C, leaf area was obtained (Li-3000: LiCor Inc.), and samples (approx. 3 cm2) were frozen in liquid N2. The leaf samples were stored at −80 °C until the enzyme assays were conducted.

Chlorophyll a fluorescence was measured using a PAM-101 (Walz, Effeltrich, Germany) equipped with a blue-light emitter detector unit (ED-101BL: Walz: Kubien et al. 2003). The ratio of variable to maximal fluorescence was measured following a 30 min dark period. Reaction centre closure was achieved by applying a 0.8 s pulse of saturating light (approximately 4000 µmol m−2 s−1). Once photosynthesis had reached steady state at a given temperature, the quantum yield of photosystem II (PSII) (ΦPSII) was measured (Genty, Briantais & Baker 1989); three saturating pulses were applied at 90 s intervals, and Fm′ was estimated as the mean fluorescence peak. The rate of linear electron transport (J) was determined as J = ΦPSII(PPFDa)f, where PPFDa is the absorbed light intensity, and f is the fraction of absorbed light reaching PSII (Krall & Edwards 1992). Leaf absorbance was measured in the 370–750 nm waveband, using a dual-channel spectrometer (SD2000; Ocean Optics, Dunedin, FL, USA) and an integrating sphere (FOIS-1; Ocean Optics). Following Krall & Edwards (1992) we assumed a value of 0.5 for f at all temperatures, although in C4 species the value of f may vary depending on biochemical subtype.

Rubisco assays

Rubisco activity in vitro was assayed on samples harvested from the leaves used during gas exchange measurements. Leaf samples were rapidly ground (< 90 s) on ice in a Bicine extraction buffer (approximately 2 mL cm−2) using a glass-in-glass homogenizer (Kubien et al. 2003). Aliquots of the crude extract were used to quantify the amount of Rubisco using a 14C-carboxy-arabinitol bisphosphate (CABP) binding assay (Butz & Sharkey 1989). The radioactivity in the filter-bound Rubisco–14CABP complexes was measured by liquid scintillation spectroscopy.

A 3.15 mL aliquot of the crude leaf extract was added to 350 µL of a Rubisco activating solution (100 mm Bicine, pH 8.2). For the C4 samples, the activating solution contained 280 mm MgCl2 and 200 mm NaHCO3, giving final concentrations of 28 mm MgCl2 and 20 mm NaHCO3, respectively (Sage & Seemann 1993). For the C3 samples, the activating solution contained 200 mm MgCl2 and 100 mm NaHCO3, giving final concentrations of 20 mm MgCl2 and 10 mm NaHCO3, respectively (Butz & Sharkey 1989). The crude extract was incubated in activating solution for 20 min at room temperature to fully carbamylate Rubisco, and thereafter was kept on ice until being assayed. Rubisco activity assays follow Kubien et al. (2003).


In both species, plants grown at warm temperatures (26°/22 °C) had higher net CO2 assimilation (A) rates from 5 to 35 °C than plants that were grown at 14/10 °C (Fig. 1). Growth temperature did not affect the thermal optimum of the C3 species, with maximum A achieved at 20 °C in either case (Fig. 1a). In air, the maximum photosynthesis rates of the C3 plants were approximately 13 and 10 µmol m−2 s−1 for the warm- and cool-grown plants, respectively. In M. glomerata, plants grown at 26/22 °C reached a maximum photosynthetic rate of about 25 µmol m−2 s−1 at 30 °C. The cool-grown plants attained a maximum A of 16 µmol m−2  s−1, near 25 °C (Fig. 1b). When grown at warm temperatures, the C4 plants had considerably higher A than the C3 species above 20 °C; below 15 °C, tha value of A of the C3 species exceeded that of the C4. When grown under cool conditions, A in M. glomerata exceeded that of C. canadensis at temperatures greater than 15 °C.

Figure 1.

Temperature response of the rate of net CO2 assimilation in (a) Calamogrostis canadensis (C3), and (b) Muhlenbergia glomerata (C4). Plants were grown at 26/22 °C (•) or 14 /10 °C (□) day/night temperature. Photosynthesis was measured at 370 µbar CO2 and 200 mbar O2, under a PPFD of 1300 µmol m−2 s−1. Each point represents the mean (± 1 SE) of measurements on four individual plants. Error bars are obscured when smaller than the symbol.

Reducing the partial pressure of O2 resulted in a slight upward shift of the photosynthetic thermal optimum in C. canadensis (Fig. 2). Photosynthesis remained sensitive to O2 across the measured temperature range in C. canadensis grown at 14/10 °C (Fig. 2a). In the warm-grown C3 plants reduced O2 did not result in higher A at temperatures below 15 °C (Fig. 2b). Under low-O2 conditions, photosynthesis was stimulated 25–30% in the cool-grown C3 plants when measured from 5 to 35 °C. The stimulation by low O2 in the warm-grown C3 plants increased over four-fold across this temperature range (Fig. 2c). Photosynthesis in M. glomerata was insensitive to O2 at all temperatures, regardless of growth condition (data not shown).

Figure 2.

Temperature response of the rate of net CO2 assimilation in Calamogrostis canadensis (C3) under high and low oxygen. Plants were grown at (a) 14/10 °C, or (b) 26/22 °C day/night temperature. Photosynthesis was measured under 20 mbar O2 (◆, ◊) or 200 mbar O2 (□, •) (mean ± 1 SE., n = 4). PPFD and CO2 were 1300 µmol m−2 s−1 and 370 µbar, respectively. After leaves had reached a steady-state photosynthetic rate in air, oxygen was reduced to 20 mbar, and a new photosynthetic rate was noted after a 10 min period. The O2 sensitivity of photosynthesis (c) was calculated as 1 − A20/A200 (Sage & Sharkey 1987). The modelled sensitivity was calculated assuming that Kc and Ko were 260 µbar and 179 mbar, at 25 °C, respectively.

In air, the in vivo electron transport rate (ETR) of C. canadensis reached a maximum of about 100 µmol m−2 s−1 near the thermal optimum of 20 °C (Fig. 3a); under 20 mbar O2 the maximum ETR was about 75 µmol m−2 s−1 (data not shown). Growth temperature did not affect the relationship between ETR and measurement temperature in C. canadensis. In C. canadensis, ETR exceeded the rate required to support measured rates of photosynthesis at temperatures less than 30 °C (warm-grown plants) and 35 °C (cool-grown plants) (Fig. 3a). The in vivo electron transport rate of M. glomerata had the same thermal optimum as photosynthesis, and exceeded the theoretically required minimum at temperatures below the respective thermal optima (Fig. 3b). Below 25 °C the warm- and cool-grown C4 leaves had the same ETR. The maximum ETR in M. glomerata was 140 and 100 µmol m−2 s−1, for the warm- and cool-grown plants, respectively (Fig. 3b).

Figure 3.

Temperature responses of the rate of electron transport in (a) Calamogrostis canadensis (C3) and (b) Muhlenbergia glomerata (C4). Plants were grown at 26/22 °C (•) or 14/10 °C (□) day/night temperature. Electron transport was measured in vivo using chlorophyll a fluorescence (Krall & Edwards 1992), at 370 µbar CO2, 200 mbar O2, and a PPFD of 1300 µmol m−2 s−1 (mean ± 1 SE, n = 4, some error bars are obscured where smaller than the symbol). Lines indicate the electron transport rate theoretically required to support the measured rates of gross assimilation. A requirement of 3H+/ATP and a functional Q-cycle were assumed. In this case, the minimum requirement in a C3 plant is (3 + 3.5φ)Vc, where φ is the number of oxygenations per carboxylation, and Vc is the carboxylation rate. For the C4, the requirements for the mesophyll and bundle sheath reactions are 2Vp and 3Vc, respectively, where Vp is the PEP carboxylation rate. The requirement for the bundle sheath reactions ignores φ, as we assume that the CO2 concentration in the stroma is sufficient to eliminate the oxygenase function. All equations from von Caemmerer (2000).

The instantaneous quantum cost of assimilation can be approximated by the ratio ΦPSIICO2*, where ΦPSII is the quantum yield of PSII, and ΦCO2* is the quantum yield of gross CO2 assimilation (A*) (Oberhuber & Edwards 1993). In cool-grown C. canadensis measured at 200 mbar O2, ΦPSIICO2* remained relatively constant between 16 and 18 molquanta molCO2−1 from 5 to 35 °C (Fig. 4a). This value was reduced to approximately 12 molquanta molCO2−1 when measured at 20 mbar O2. In warm-grown C. canadensis measured at 200 mbar O2, the quantum cost of assimilation increased from about 10–18 molquanta molCO2−1 as temperature increases from 5 to 35 °C (Fig. 4b). At low O2, ΦPSIICO2* was about 9 molquanta molCO2−1 in the warm-grown C3 plants, and was insensitive to measurement temperature. Growth temperature had no effect on ΦPSIICO2* in M. glomerata (Fig. 4c). At or above the thermal optimum, the quantum cost of assimilation in this C4 species was constant between 10 and 12 molquanta molCO2−1. At suboptimal temperatures, ΦPSIICO2* increased in this C4 species, regardless of growth temperature.

Figure 4.

The ratio of the quantum yield of PSII (ΦPSII) to the quantum yield of gross assimilation (ΦCO2*) as a function of temperature in Calamogrostis canadensis (a, b) and Muhlenbergia glomerata(c). Plants were grown at 26/22 °C or 14/10 °C day/night temperature. Chlorophyll a fluorescence was measured concurrently with gas-exchange. Photosynthesis was measured at 370 µbar CO2 and 200 mbar O2, under a PPFD of 1300 µmol m−2  s−1; in C. canadensis photosynthesis was also measured at 20 mbar O2. Each point represents the mean (± 1 SE) of four independent measurements.

Calamogrostis canadensis had an approximately four-fold greater Rubisco content than M. glomerata (Table 1). At 25 °C, the in vitro kcat of the C4 species was higher than in the C3 plant. The activation energy of the C3 Rubisco was nearly 10% higher than in the C4 species. In M. glomerata, the activation energy of assimilation from 5 to 20 °C was approximately 72 kJ mol−1, which was similar to the activation energy of Rubisco (approximately 66.5 kJ mol−1). In contrast, the activation energy of CO2 assimilation (measured at 200 mbar O2) in C. canadensis between 5 and 20 °C was considerably lower than that of Rubisco.

Table 1.  Biochemical characteristics of Muhlenbergia glomerata (C4) and Calamogrostis canadensis (C3)
PlantGrowth temp.
Catalytic sites
(µmol m−2)
in vitro kcat
(mol mol−1 s−1)
(kJ mol−1)x
(kJ mol−1)y
  1. The kcat and activation energy (Ea) data are values at 25 °C. The concentration of Rubisco catalytic sites was determined by the 14CABP-labelling assay described in the Methods, assuming 6.5 binding sites per Rubisco (Butz & Sharkey 1989). The Rubisco kcat values were determined by the incorporation of 14C into acid-stable products. Activation energies were calculated as described by Berry & Raison (1981). Each value represents the mean (± SE) of four to six measurements. Values with different superscripts are statistically different from each other (P < 0.05). Statistical analysis was done by anova using S + (MathSoft 1994), with growth temperature and photosynthetic pathway as main effects. xMeasured between 0° to 35 °C. yMeasured between 5° to 20 °C.

M. glomerata (C4)14/10 4.3 ± 0.2a5.21 ± 0.21a67.3 ± 0.5ab70.0 ± 1.8a
M. glomerata (C4)26/22 5.7 ± 0.4a4.57 ± 0.29ab65.6 ± 0.4a73.3 ± 4.8a
C. canadensis (C3)14/1019.5 ± 1.4b3.79 ± 0.44b73.4 ± 2.3c39.1 ± 6.8b
C. canadensis (C3)26/2220.7 ± 1.3b3.56 ± 0.17b71.3 ± 1.7bc37.6 ± 2.4b

The maximum in vitro activity of Rubisco (Vcmax) exceeded the rate of gross CO2 assimilation (A*) in C. canadensis at all temperatures (Fig. 5a & b). At the photosynthetic thermal optima, Vcmax was at least three-fold greater than A* in this C3 species. At 5 °C, Vcmax was only marginally greater than A*, and they were nearly equivalent when A* was measured at 20 mbar O2. In cool-grown M. glomerata, A* and Vcmax were equivalent between 5 and 20 °C, but Vcmax exceeded A* at warmer temperatures (Fig. 5c). Similarly, Vcmax exceeded A* at 30 and 35 °C in M. glomerata grown at 26/22 °C, but the rates were the same between 5 and 25 °C (Fig. 5d).

Figure 5.

The temperature response of the rates of in vitro Rubisco activity (Vcmax) and gross CO2 assimilation (A*) in Calamogrostis canadensis (C3) and Muhlenbergia glomerata (C4). Plants were grown at 14/10 °C (a, c) or 26 /22 °C (b, d). CO2 assimilation was measured at 200 mbar oxygen in both species, and at 20 mbar oxygen in the C3. The in vitro activity of Rubisco (Δ) was determined by the incorporation of 14C into acid stable products, following carbamylation of the enzyme. Enzyme assays were conducted on leaves that had been used for gas-exchange. Each value represents the mean (± 1 SE, error bars obscured where smaller than the symbol) of measurements on four individual plants.


In this study, we present data supporting the hypothesis that Rubisco capacity, as determined by Rubisco content, is the principle limitation on C4 photosynthesis at suboptimal temperatures. Rubisco capacity largely reflects Rubisco content at low temperatures, as indicated by the similarity between Vcmax and gross CO2 assimilation. Because Rubisco operates in a high CO2 environment in C4 plants, gross CO2 assimilation should reflect Vcmax, and thus Rubisco content (Sage 2002). At low temperatures, photosynthesis (A) in warm-grown C. canadensis appears to be limited by the ability to regenerate Pi in the chloroplast, but growth at 14/10 °C appears to alleviate this limitation. In both species, electron transport in vivo exceeds the theoretical requirement at temperatures below the thermal optimum, indicating that photosynthesis is not limited by light-harvesting processes in the thylakoid at low temperatures. If Rubisco capacity limits the rate of C4 photosynthesis, then leakage of CO2 from the bundle sheath should increase, and this is consistent with the observed increase in ΦPSIICO2* at low temperatures.

Rate limitations on C4 and C3 photosynthesis at low temperatures

Between 5 and 20 °C, the activity of Rubisco in vitro (Vcmax) and the rate of gross CO2 assimilation (A*) in vivo are equivalent in M. glomerata grown at 14/10 °C, indicating that Rubisco capacity limits C4 photosynthesis at low temperatures. The similarity between the activation energy (Ea) of photosynthesis and Vcmax in M. glomerata is also consistent with a Rubisco limitation of C4 photosynthesis. Similar results have been reported in the C4 grasses Bouteloua gracilis and Muhlenbergia montanum, where Vcmax and net photosynthesis rate are equivalent below 17 to 22 °C (Pittermann & Sage 2000, 2001). When grown at 26/22 °C, the Rubisco limitation of photosynthesis in M. glomerata extended to 25 °C, probably due to a rise in the electron transport capacity during acclimation to the warmer growth temperature. There was no difference in the amount of Rubisco in cool- and warm-grown M. glomerata, whereas the in vivo electron transport rate in warm-grown M. glomerata was 25% greater than that of the cool-grown plants. A rise in electron transport capacity would increase the capacity to regenerate RuBP and PEP. Increased capacity of these potentially rate-limiting steps, arising from growth at warmer growing conditions, allows the Rubisco control over photosynthesis to extend to warmer temperatures.

In C. canadensis, there is little evidence that Rubisco limits A below the thermal optimum in plants grown at 26/22 °C. Instead, the capacity to regenerate orthophosphate (Pi) appears to be the primary control over A, as indicated by the low level of O2 sensitivity between 5 and 20 °C. A 90% reduction in O2 should enhance A in C3 plants, as the rate of photorespiration is suppressed. This enhancement generally declines with temperature, reflecting the inhibition of Rubisco oxygenase activity at cooler temperatures (Fig. 2c; Sage & Sharkey 1987). When O2-sensitivity is reduced at constant temperature and gas concentrations, the capacity to regenerate Pi in the chloroplast generally exerts high control over A (Sharkey 1985; Stitt & Hurry 2002). Pi-regeneration limitations are common in C3 species when A is measured below the growth temperature (Sage & Sharkey 1987; Sage, Sharkey & Seemann 1989). Acclimation to low temperature is known to enhance the Pi-regeneration capacity and thereby reduce the extent of a Pi-regeneration limitation in C3 plants grown at low temperature (Stitt & Hurry 2002). Consistent with this, Pi-regeneration limitations are not apparent in cool-grown C. canadensis, as shown by the relatively high O2 sensitivity. The response shown for C. canadensis indicates that this species is prone to Pi-regeneration-limited A when grown at warmer temperature, but can overcome this limitation upon acclimation to cool conditions. The precise mechanism of low-temperature acclimation in C. canadensis is uncertain, but probably involves increases in the capacity of enzymes associated with carbon metabolism, such as sucrose phosphate synthase and cytosolic fructose-1, 6-bisphosphatase. Such changes have been previously noted in C3 grasses (Hurry et al. 1995) and Arabidopsis thaliana (Stitt & Hurry 2002).

The low-temperature performance of Rubisco

It is important to identify the potential for Rubisco limitations in M. glomerata and C. canadensis, as this highlights the fundamental differences between the function of Rubisco in C4 and C3 plants. To differentiate between the effects of enzyme capacity versus CO2 availability, we have modelled the temperature dependencies of the Michaelis–Menten constant for CO2 (Kc), and the amount of CO2 required for the rate of Rubisco carboxylation (Vc) to be 80% of Vcmax (Fig. 6); above this concentration, the increase in enzymatic reaction rates with increasing substrate concentration slows considerably, and the amount of the enzyme becomes the principal limitation. For Rubisco from a C4 species, Kc ranges from about 135–1500 µbar from 5 to 30 °C (Fig. 6a). Using the model of von Caemmerer & Furbank (1999) and assuming a bundle-sheath conductance to CO2 of 3 mmol m−2 s−1, we estimate that the partial pressure of CO2 in the bundle sheath of M. glomerata exceeds 4 mbar between 5 and 30 °C, which is at least 2.5 times Kc. As shown in Fig. 6a, Rubisco is effectively CO2 saturated in the bundle sheath of M. glomerata at temperatures below at least 20 °C, because the CO2 level required for Vc to achieve 80% of Vcmax is below the predicted 4 mbar substrate concentration in the stroma. These simulations indicate that Vc will be limited by the amount of the enzyme (Vcmax) at low temperatures, assuming that the stromal concentration of RuBP is saturating for Rubisco.

Figure 6.

Temperature effects on the amount of CO2 required in order for Rubisco carboxylation (Vc) to proceed at 80% of Vcmax. (a) A C4 Rubisco, assuming that Kc is 650 µbar and Ko is 450 mbar, at 25 °C (von Caemmerer 2000; p. 100). In this case the partial pressure of O2 in the stroma (Os) is modelled at 200 mbar. (b) A C3 Rubisco, with a Kc of 260 µbar (von Caemmerer 2000; p. 45) and a Ko of 397 mbar (Jordan & Ogren 1984); this value was selected to minimize the oxygenation function in this simulation. The Os is modelled at 200 mbar () and 20 mbar (– –). The dashed line in each panel represents the current atmospheric CO2. The temperature response of Kc (····) is shown in each panel, modelled using a Q10 of 2.24 (von Caemmerer 2000; p. 45). Q10 values of 2.21 and 1.63 were assumed for Vcmax and Ko, respectively (von Caemmerer 2000; p. 45).

In C3 plants the situation is different; Kc increases from about 40–400 µbar between 0 and 30 °C (conservatively assuming a value of 260 µbar at 25 °C, Fig. 6b). Assuming that the partial pressure of O2 in the chloroplast stroma is close to atmospheric (e.g. Os = 200 mbar), then the partial pressure of CO2 required for Vc to be 80% of Vcmax is higher than atmospheric between 0° and 30 °C (Fig. 6b). In C. canadensis at 5 °C, Vc would be about 70% of Vcmax based on the measured Ci of 325 µbar (Kubien 2003). If Rubisco capacity is limiting, the CO2 deficiency is manifested by the inhibition from the oxygenase activity of Rubisco and by the deficiency of CO2 as a substrate. If RuBP regeneration capacity is limiting, the CO2-deficiency is manifest only by oxygenase inhibition (Sage & Reid 1994). If the oxygenase function of Rubisco is reduced by lowering Os to 20 mbar, then atmospheric levels of CO2 could effectively CO2-saturate C3 photosynthesis at temperatures up to 10 °C, and the nature of the Rubisco limitation would be more like that seen in C4 plants.

Temperature effects on the leakage of CO2 from the bundle sheath of C4 leaves

The increase in ΦPSIICO2* at cooler temperatures reflects increased leakage of CO2 (φ) from the bundle sheath cells in M. glomerata. In the C4 dicot Flaveria bidentis, genotypes with antisense Rubisco constructs (anti-RbcS) show higher ΦPSIICO2*, and higher estimates of φ, than wild-type plants at warm and intermediate temperatures (Kubien et al. 2003). Both wild-type and anti-RbcS plants show a rise in leakiness with reduced temperature, and converge on common leakiness estimates at low temperatures where the Rubisco flux control coefficient indicates complete limitation by Rubisco capacity (Kubien et al. 2003). Increased CO2 leakiness reduces ΦCO2*, but not ΦPSII; hence ΦPSIICO2* should increase as Rubisco increasingly limits A. The values of ΦPSIICO2* in M. glomerata at warm temperatures (> 20 °C) are similar to those reported by Oberhuber & Edwards 1993) for a range of C4 grasses representing each biochemical subtype. The stability of ΦPSIICO2* at warm temperatures reflects a constant stoichiometry between the mesophyll and bundle sheath reactions (Kubien et al. 2003). As Rubisco becomes a principal control over the rate of photosynthesis in M. glomerata at cool temperatures (< 20 °C) this stoichiometry is altered, allowing CO2 to accumulate in the bundle sheath. The capacity of Rubisco to consume CO2 is increasingly exceeded by the capacity of the C4 pump as temperature declines. This results in higher bundle sheath CO2 levels, increasing the diffusion gradient between the bundle sheath and intercellular airspace, and hence increasing φ (Henderson, von Caemmerer & Farquhar 1992; von Caemmerer & Furbank 1999). Increased leakiness cannot arise from a limitation in the mesophyll reactions (e.g. PPDK or PEPCase), as this would reduce the delivery of CO2 to the bundle sheath and so should have the opposite effect.

Consequences for C4 plants in cool climates

The CO2-concentrating functions of the mesophyll give C4 species numerous photosynthetic advantages over their C3 competitors, particularly at warm temperatures. At low temperatures, the Rubisco content of C4 plants imposes a low ceiling on the maximum rate of carbon gain. Whereas in some C4 species the amount and activity of enzymes such as PPDK and PEPCase may affect low-temperature performance, the inferior photosynthetic performance at low temperatures of many C4 plants occurs because they either do not, or cannot, hold as much Rubisco as C3 species (Pittermann & Sage 2000). It is well established that they do not; C3 plants typically have three to six times as much Rubisco as C4 species (Ku, Schmitt & Edwards 1979; Long 1999). A possible low-temperature acclimation strategy would be to increase the amount of Rubisco, but such a response does not appear to commonly occur in C4 plants (Pearcy 1977; Pittermann & Sage 2000, 2001; Naidu et al. 2003). C4 species may have little capacity to increase Rubisco content. This could be due to a biochemical constraint, whereby synthesis of photosynthetic proteins is restricted by a need to maintain specific stoichiometric ratios required for effective co-ordination of the C3 and C4 cycles. The ability to increase Rubisco content may be constrained physically, because the localization of Rubisco to the bundle sheath chloroplasts places an upper limit on the amount of the protein a leaf may contain. In C4 grasses bundle sheath tissues occupy less than 30% of the leaf volume; by contrast, the Rubisco-containing mesophyll tissues occupy more than 50% of a C3 leaf (Dengler et al. 1994). Furthermore, unlike C3 plants in which Rubisco-containing chloroplasts can occur on all axis of the cell, in C4 plants, Rubisco-containing chloroplasts are usually restricted to an inner or outer pole of the bundle sheath cells, so the actual leaf volume available to compartmentalize Rubisco is 40 to 70% less then total bundle sheath volume (Dengler & Nelson 1999). If tissue compartmentalization is a limiting factor, then C4 plants would have to increase the size and/or amount of bundle sheath tissue to accommodate additional Rubisco. As an acclimation response this is unlikely, as it would require fundamental changes to the specific vascular patterning and associated intercellular transport that are required to ensure efficient co-ordination between the carbon-concentrating mechanism and the bundle sheath reactions (Dengler & Nelson 1999). Acclimation to low temperature via changes in these relationships would likely be detrimental at warmer temperatures, where the efficiency of C4 photosynthesis is facilitated by high bundle sheath CO2 concentration. Low Rubisco content does not impair carbon gain at warm temperatures in C4 plants, and is beneficial for their nitrogen economy. However, this characteristic is detrimental when low-temperature limits the turnover capacity of Rubisco.


The authors wish to thank Bruce Hall and Andrew Petrie for assistance with plant growth, Barbara and Garnett Sproule for permission to collect M. glomerata and C. canadensis, and the help of two anonymous reviewers. This work was funded by an NSERC grant (OGP0154273) to RFS.