A computer-operated routine of gas exchange and optical measurements to diagnose photosynthetic apparatus in leaves


Correspondence: A. Laisk. Fax: + 372 7420286; e-mail: alaisk@ut.ee


Photosynthesis is a complex process whose rate is affected by many biochemical and biophysical factors. Fortunately, it is possible to determine, or at least estimate, many of the most important parameters using a combination of optical methods and gas transient analyses. We describe here a computer-operated routine that has been developed to make detailed assessments of photosynthesis at a comprehensive level. The routine comprised the following measurements: steady-state light and CO2 response curves of net CO2 assimilation at 21 and 2 kPa O2; transients from limiting to different saturating CO2 concentrations at 2 kPa O2; post-illumination CO2 fixation transient; dark–light induction of O2 evolution; O2 yield from one saturating single-turnover flash; chlorophyll fluorescence F0, Fs and Fm during the light and CO2 response curves; leaf transmission at 820 nm (P700+) during the light and CO2 response curves; post-illumination re-reduction time of P700+. The routine was executed on a two-channel fast-response gas exchange measurement system (A. Laisk and V. Oja: Dynamic Gas Exchange of Leaf Photosynthesis. CSIRO, Canberra, Australia). Thirty-six intrinsic characteristics of the photosynthetic machinery were derived, including quantum yield of CO2 fixation (YCO2), time constant of P700 re-reduction (τ′), relative optical cross-sections of PSII and PSI antennae (aII, aI), PSII and PSI density per leaf area unit, plastoquinone pool, total mesophyll resistance, mesophyll diffusion resistance, Vm, Km(CO2) and CO2/O2 specificity of Rubisco, RuBP pool at CO2 limitation (assimilatory charge). An example of the routine and calculations are shown for one leaf and data are presented for leaves of 8-year-old-trees of two birch clones growing in Suonenjoki Forest Research Station, Finland, during summer 2000. Parameters YCO2, basic τ′, aII, aI, Km(CO2) and Ks varied little in different leaves [relative standard deviation (RSD) < 7%], other parameters scattered widely (RSD typically 10–40%). It is concluded that the little scattered parameters are determined by basic physico-chemical properties of the photosynthetic machinery whereas the widely scattered parameters are adjusting to growth conditions. The proposed non-destructive routine is suitable for diagnosing the photosynthetic machinery of leaves and may be applied in plant ecophysiology and in genetic engineering of plants.


net CO2 assimilation rate


assimilatory charge


aI, a0, relative optical cross-sections of PSII and PSI antenna and of non-photosynthetic absorption

Ca, Ci, Cw, Cc, CO2 concentrations

ambient, intercellular space, cell wall liquid and carboxylation site, respectively


carbon reduction cycle


b6f, cytochrome b6f

DCMU, dichlorophenol-dimethyl urea; ETR

and J, electron transport rate


Fs, Fm, Fmd, fluorescence yields, minimum, steady state, maximum (all in the light) and maximum Fm in the dark, respectively


far-red light


CO2 compensation point


Rubisco CO2/O2 specificity


relative rate constant for regulatory non-photochemical excitation quenching


relative rate constant for photochemical excitation quenching at open PSII centres


Pm, Po, 820 nm signal difference from the dark level, steady state, maximum and corresponding to oxidizable P700


(I in equations), PFD, photon flux density, absorbed and incident, respectively




3-phosphoglyceric acid


inorganic phosphate


PSII, photosystems I and II




donor pigment of PSI


qI, non-photochemical quenching, energy dependent and inhibitory


RK, Krebs cycle CO2 evolution rate in the dark and in the light, respectively


rm, rmd, leaf diffusion resistances, in gas phase, mesophyll total and mesophyll diffusional


ribulose 1,5-bisphosphate


ribulose 1,5-bisphosphate carboxylase-oxygenase


specific carboxylation efficiency


water vapour pressure difference


white light


YF, YP, quantum yields of photosynthetic e transport, calculated from CO2 uptake, Chl fluorescence and 820 nm transmission, respectively


quantum yield of CO2 fixation.


A widely used model for analysing canopy photosynthesis Farquhar & von Caemmerer 1982) is based on Rubisco kinetic properties at limiting CO2 concentrations and on the properties of photosynthetic e transport chain and light harvesting system when CO2 is not limiting. The latter condition differentiates into light-limited state (quantum yield) and light- and CO2 saturated state where e transport is feedback-limited by Pi turnover that by itself is limited by sucrose and starch synthesis rates. The parameters used to describe the whole approach are Km for CO2 and O2, maximum rate and CO2/O2 specificity of Rubisco, the intrinsic quantum yield, and the maximum e transport rate. In ecophysiological studies these parameters are related to the investment of N into the light harvesting and e transport systems of the photosynthetic machinery (Niinemets, Kull & Tenhunen 1998; Evans & Poorter 2001). The assumption is that plants maximize their productivity by optimizing the N distribution between different parts of the photosynthetic machinery. Although the Rubisco part of the Farquhar–von Caemmerer model is sufficiently detailed, other parts are presented as a single integrative parameter. For example, the maximum e transport rate Jmax may be determined by the abundance of Cyt b6f complex, but the turnover rate of the complex is feedback-controlled by ΔpH that itself is a function of rate-limitations downstream. The intrinsic quantum yield of e transport also is a complex function of excitation distribution between the two photosystems and excitation use efficiency by each photosystem, the former being controlled by redistribution of the light-harvesting Chl between the two photosystems during state transitions (Allen & Pfannschmidt 2000), the latter being controlled by non-photochemical quenching and acceptor side reduction at PSII (Horton, Ruban & Walters 1994) and by donor side oxidation and acceptor side reduction at PSI (Genty & Harbinson 1996). Adaptation and acclimation of photosynthesis may involve changes in any, or many, of these parameters and so there is a need to estimate these parameters to fully understand how photosynthesis is affected by its environment.

Techniques of gas exchange and optical measurements allow non-destructive detection of a number of intrinsic parameters of the photosynthetic machinery non-destructively in intact leaves. By accumulating RuBP at low CO2 and O2 concentrations and using short exposures to high CO2 concentration the full kinetic curve of Rubisco with respect to CO2 at saturating RuBP has been resolved (Laisk & Oja 1974; Ruuska et al. 1998), which has allowed the determination of the Km(CO2) in intact leaves. The pool of RuBP subject to carboxylation has been found from the post-illumination CO2 fixation and the rate-regulation by competitive binding of RuBP and PGA to Rubisco has been analysed in leaves (Laisk, Oja & Kiirats 1984; Laisk et al. 1987; Ruuska et al. 1998). The CO2/O2 specificity Ks of Rubisco has been found from the CO2 photocompensation point Γ* (Laisk 1970; Brooks & Farquhar 1985). The diffusion resistance between leaf mesophyll cell wall and Rubisco active sites has been found from the discrimination against 13CO2 during photosynthesis (Evans et al. 1986) or from the assumption that all the rate of electron transport measured by fluorescence that cannot be accounted by photosynthesis is used in photorespiration (Loreto et al. 1994; Laisk & Loreto 1996).

Primary reactions of photosynthesis, including the densities of the two photosystems, excitation distribution between them and electron transport from PSII to PSI, have been analysed in intact leaves using O2 evolution, Chl fluorescence and 820 nm transmission signals. PSII density has been determined by measuring O2 evolution from single turnover flashes (Chow, Hope & Anderson 1989; Oja & Laisk 2000). PSI density has so far been measured using the bleaching of the 700 nm band (Chow, Anderson & Hope 1988; Melis 1989) or by quantifying the reduction of exogenous e acceptors (Graan & Ort 1984) in thylakoid preparations. However, this measurement has not yet been made with intact leaves. The optical cross-section of PSII antenna has been found from the speed of fluorescence induction in DCMU-poisoned chloroplasts, whereas the antenna cross-section of PSI has been determined from the speed of P700 oxidation (Melis 1989). In intact leaves excitation distribution between the photosystems (relative antenna cross-section) has been derived from the simultaneous measurements of the quantum yield of CO2 fixation, Chl fluorescence and 820 nm transmission (Eichelmann & Laisk 2000). The kinetics of e transport through the Cyt b6f complex have been analysed by measuring the post-illumination re-reduction rate of P700 in intact leaves (Harbinson & Hedley 1989; Laisk & Oja 1994, 1995).

Taken together these approaches give a relatively full picture of the photosynthetic process using only measurements on intact leaves; however, until now the practical use of these techniques in ecophysiology has been limited by the complexity of the necessary equipment and by the laborious data processing. In this report we describe a complex computer-operated routine of gas exchange and optical measurements and data processing programs that allow derivation of a number of intrinsic photosynthetic parameters. The measurement routine was executed using the fast-response system for gas exchange and optical measurements on intact leaves (Laisk & Oja 1998). Because this apparatus is not suitable for field operation, we have used a compromise approach: laboratory measurements at the field site. We assume that the basic parameters of the photosynthetic machinery are stable enough for measuring on excised leaves and believe that the laboratory measurements can be combined with in situ measurements of stomatal conductance in future, for use in a mechanistic model of canopy photosynthesis. We present data measured on 8-year-old-birch trees growing in the field. The two clones were chosen considering their further use in an open-top chamber experiment with elevated CO2 and O3.

Material and apparatus

Plant material and growth conditions

For the present study two clones of birch Betula pendula Roth. (V5952 and K1659, denoted below as clones 4 and 80, respectively) were selected from a field experiment that had been established in 1993 in Suonenjoki Research Station (62°05′ N, 27°00′ E, 130 m asl). The data presented below were measured during summer 2000, when the trees were 8 years old. Measurements were conducted in four rounds, first round 6–13 June; second round 4–10 July; third round 21–26 July (short-shoot leaves) and 28 July–1 August (long shoot leaves); fourth round 14–19 August (short shoot leaves) and 21–26 August (long shoot leaves). For the measurements four trees of both clones were selected. In every round of measurements one leaf was sampled from each tree, thus, measurements were repeated on leaves from four trees of each clone. A leaf was cut from a south-exposed middle-layer terminal branch and the petiole was immediately immersed in water. After 5–7 min the leaf was fitted to the chamber of the gas system and measurements commenced. With one leaf the routine lasted 2·5 h. Leaves were taken for experiment between 0600 and 1500 h and care was taken that the four measurements with one clone were distributed over the day, whereas the leaves of different clones were always measured simultaneously using two similar gas systems operated in parallel.

Gas exchange system

A two-channel fast-response leaf gas exchange measurement system (Fast-Est, Tartu, Estonia), described earlier (Laisk & Oja 1998; Eichelmann et al. 2000) was used for these measurements. Part of a leaf was enclosed in a sandwich-type round chamber (diameter 31 mm, height 3 mm) and exposed to a gas flow rate of 0·5 mmol s−1. The upper side of the leaf was sealed to the thermostatted glass window using starch gel, to stabilize leaf temperature independent of light intensity and transpiration rate. Gas exchange was possible only through the abaxial epidermis, but this did not influence the rate, because birch has no stomata in the adaxial epidermis. In all measurements the bath temperature was 22 °C and leaf temperature did not exceed 22·5 °C.

CO2 exchange

Two similar gas flow channels were mounted together in one gas exchange measurement apparatus. The leaf chamber could be connected to either of them by a computer-operated switching device. The CO2 responses of the leaf were measured by readjusting the CO2 concentration in channel 2 while a standard CO2 concentration was held in channel 1. Both channels of the gas system were fed by compressed N2, O2 and CO2 from pressure cylinders. In each channel different O2 concentrations were obtained by mixing the flows of N2 and O2, CO2 concentrations were produced by adding pure CO2 into the gas flow by adjusting CO2 pressure differences on calibrated capillaries. On the outlet of the system CO2 concentration was measured by an optical (infrared) analyser LI-6251 (LiCor, Lincoln, NE, USA), calibrated at each background CO2 concentration by creating a known concentration step by changing the pressure on the same capillary that produced the background CO2 concentration. The half-response time of the CO2 analyser in the system was 1·6 s.

Water vapour pressure was controlled by passing a fraction of gas over water warmed to 50 °C and was measured by micropsychrometers (Fast-Est) calibrated against dry air. In all measurements the water vapour pressure in the inlet gas was 1 kPa.

The CO2 concentration in intercellular space, Ci, and concentrations of dissolved CO2 in cell wall liquid, Cw, and at the carboxylation sites, Cc (µm) were calculated using the standard routine based on simultaneous measurements of net CO2 uptake, transpiration and leaf temperature (Laisk 1977; von Caemmerer & Farquhar 1981; Laisk & Oja 1998). The actual leaf temperature was not measured but calculated from the leaf energy budget. Due to the good thermal contact between the leaf and water jacket the leaf temperature rose above the temperature of the water jacket by no more than 0·5 °C. Diffusion resistances were expressed in relation to dissolved CO2 using its concentration difference as driving force. The following Eqns (1) and (2) were used to calculate solubility factors for CO2, BC and for O2, BO (leaf temperature tl in °C):



The O2 evolution from the leaf was measured in the same flow-through system with a Zr-oxide O2 analyser Ametek S-3 A (Thermox, Pittsburgh, PA, USA). The sensitivity of the analyser was calibrated using the flow of air (21 kPa O2) through the same capillary that feeds CO2 into the gas flow. Considering the different viscosity of CO2 and air and the slightly different thickness of the layer of molecular slip at the walls, the metrologies of the O2 and CO2 measurements were bound together. In O2 evolution measurements the leaf was preconditioned in channel 1 in the presence of O2, but for the measurements the leaf chamber was connected with channel 2 where O2 pressure was 1–3 Pa. Half-response time of the O2 analyser in the system was 0·8 s.

Optical measurements

Chamber illumination

The leaf chamber was illuminated through a multi-arm fibre-optic light guide from three Schott KL 1500 light sources (Heinz Walz, Effeltrich, Germany). One lamp provided actinic illumination, the second, which was equipped with a 720 nm narrow-band interference filter (Andover Corp. Salem, NH, USA) provided far-red light (FRL) (incident intensity 150 µmol m−2 s−1) and the third provided saturation pulses for fluorescence, of 10 500 µmol quanta m−2 s−1. The sources of actinic light and saturation pulses were filtered by heat-reflecting filters (Optical Coating Laboratory, Inc., Santa Rosa, CA, USA) to minimize the saturating effect of non-modulated light on fluorescence and 820 nm detectors. The absorption coefficient of the leaf for photosynthetically active quanta was measured in an integrating sphere with a spectroradiometer PS-2000 (Ocean Optics, Dunedin, FL, USA), integrating the absorption over the spectrum between 400 and 800 nm.

Chlorophyll fluorescence

Chl fluorescence was measured using fibres covering an ellipsoid of 1 × 2 cm2 individually arranged between the illumination fibres and connected to the emitter-detector unit 101ED of the PAM 101 chlorophyll fluorometer (Heinz Walz) operated at 100 kHz pulse frequency. Measurement pulses were switched on only for 5 s when photon flux density (PFD) was < 130 µmol m−2 s−1. An additional short-pass filter provided by H. Walz was used in the 101ED detector to reduce interference between simultaneous measurements of fluorescence and 820 nm transmission. This filter decreased the fluorescence signal 3·5-fold, but because the fibres were close and perpendicular to the leaf the signal : noise ratio was still good. The following corrections were considered: for cross-talk between the 650 nm excitation LED and the sensor diode (an offset for 10 and 30% of F0 in the two systems used), for overload of the detector by non-modulated emission (an increase in Fm for about 0·5%), for partial unsaturation of Fm during the saturation pulse (an increase for about 2% to the observed Fm signal) and for PSI fluorescence (Genty, Wonders & Baker 1990; Peterson, Oja & Laisk 2001). The PSI offset was dependent on the spectral window used for fluorescence measurements and was 14% of F0.

Leaf transmission at 820 nm

This signal was measured using fibres covering a circle of 1 cm2 in the centre of the leaf. Fibres arranged between the illumination fibres and connected to the emitter unit ED 800T of another PAM 101 chlorophyll fluorometer were used to provide 820 nm pulses, while a bundle of fibres at the other side of the leaf collected the transmitted 820 nm radiation and guided it to the detector of the ED 800T unit. When a leaf was enclosed in the leaf chamber, the measuring beam intensity was set to maximum and the gain of the PAM 101 used for 820 nm transmission measurements was adjusted to give a total signal of between 1·5 and 2 V, depending on leaf optical thickness. This full signal was then offset using the zeroing function of the PAM 101, and the offset signal was amplified 100 times for subsequent recording. Deflections from the dark reference line Ps in percentage of the full 820 nm signal were calculated as follows:


where Ps is the deflection from the dark level (either in percentage or volts, indicated in parentheses), G is the gain factor of the amplifier (G = 100), and S(820) is the total signal before zeroing (volts). Corrections to the 820 nm signal caused by the interference of non-modulated background light with the 820 nm signal and by the incomplete oxidation of oxidizable P700 during the saturation pulses were considered.

The signal level corresponding to oxidation of P700 under FRL, PFRL, was measured applying FRL in the dark for 3 s after different PFDs during the light response curve measurement and for 10 s at the end of the measurement of the light response curve. To obtain the signal level corresponding to completely oxidized P700, Pm, the PFRL value was corrected for the presence of PSII light in the FRL:


The correction factor was obtained from the additional oxidation of P700 when a saturating single- or multiple-turnover pulse was applied on the background of FRL.

Some PSI centres may be reduced on the acceptor side and in these P700 cannot be oxidized. To detect these centres, P700 oxidation level during a 20 ms pulse of 10 500 µmol quanta m−2 s−1 applied during steady-state photosynthesis was defined as Po (Ctrl P-procedure, below).

Single turnover flashes and millisecond front-edge pulses

Saturating single-turnover flashes (60 µmol quanta m−2, 3 µs half-width) for the measurement of active PSII pool were produced by Machine Vision Strobe MVS-7060 (EG & G Optoelectronics, Salem, MA, USA) and applied to the leaf via a branch of the fibre-optic light guide (Oja & Laisk 2000). A computer-operated electropneumatic shutter (Fast-Est) that fitted into the body of the KL 1500 light source in the slit of the slide filter holder was used for opening and closing the light beam with edges of 1·3 ms.

Computer-operation of the system

The system was operated and data were recorded by an A/D converter board ADIO 1600 (ICS Advent, San Diego, CA, USA) using a system-operation and data-recording program RECO (Fast-Est). The program mimics both a multichannel chart recorder and a recording oscilloscope. Signals from the CO2 and O2 analysers, psychrometer and PAM 101 were recorded at 5 ms intervals, but data points were plotted every 200 ms by averaging the 5 ms readings. In the oscilloscope mode, used for fast transients in P700 oxidation and reduction, the signal was recorded with 50 µs intervals.

The actinic light was computer-controlled using one D/A channel of the ADIO 1600 board, the second D/A channel was used to automatically compensate the output signal of the LI 6251, operated in absolute mode. Manostats used for adding CO2 into the gas flow were operated by digital step-motors. The electropneumatic shutters and other on-off functions of the system (switching light sources, fixed O2 pressures of 21 000, 2000 and 2 Pa) were operated using the I/O channels of the board. A meta-language based on Turbo-Pascal was created for the user-friendly programming of the measurement routine. The pre-programmed experimentation allowed us to generate transients that started and finished on exactly determined time instants and contained a fixed number of data points. This feature facilitated further data processing using templates in Microsoft Excel.

Chlorophyll concentration

The Chl was extracted and measured according to Vernon (1960).

Experimental routine

The whole routine of measurements comprised the following subsections.

  • 1Light response curve at 36 Pa CO2 and 21 kPa O2: this subroutine was used to determine photosynthetic rate under air CO2 and O2 pressures at the stomatal opening obtained in the chamber, and the relationship between e transport, Chl fluorescence and 800 nm transmission.
  • 2CO2 response curve at PFD of 760 µmol quanta m−2 s−1 and 21 kPa O2: to determine total mesophyll resistance (carboxylation plus diffusion, rm) and mesophyll diffusion resistance rmd, CO2/O2 specificity of Rubisco, Ks, and maximum CO2- and light-saturated photosynthetic rate, A25 µm.
  • 3CO2 response curve at PFD of 760 µmol quanta m−2 s−1 and 2 kPa O2: to determine total mesophyll resistance rm and the initial slope (gc = Km/Vm) of the Rubisco kinetic curve.
  • 4Rubisco kinetics at CO2 saturation: this subroutine continued the measurement of the CO2 response of Rubisco at high CO2 concentrations where steady-state photosynthesis was limited by RuBP regeneration rate, allowing determination of Vm and Km(CO2) of Rubisco.
  • 5Post-illumination CO2 fixation: to determine the maximum RuBP pool in the leaf (assimilatory charge AC) and the specific carboxylation efficiency SCE.
  • 6Light response curve at 33 Pa CO2 and 2 kPa O2: to determine the maximum intrinsic quantum yield of CO2 uptake and the relative optical cross-sections of PSII and PSI antennae aII and aI.
  • 7Dark-light induction of O2 evolution: to determine the pools of plastoquinone and PGA in the dark-adapted state and the time kinetics of light-activation of e transport.
  • 8O2 evolution from a saturating single-turnover flash: to determine the pool of O2-evolving PSII.

Leaves cut from trees about 5–7 min before, with petioles in water, were pre-adapted in the leaf chamber at PFD of 760 µmol m−2 s−1, 36 Pa CO2 and 21 kPa O2 for about 20–30 min to obtain maximum stomatal opening. Then the above-listed subroutines were executed. Below, the subroutines will be described not in their actual sequence (they were arranged in the sequence that minimally disturbed stomata), but first those that revealed parameters of light reactions and then those that were related to CO2 assimilation.

Light response at 36 Pa CO2 and 21 kPa O2

Curves were measured jumping from PFD of 760–2000 µmol m−2 s−1 and then downward, to 1300, 760, 440, 250, 140, 75, 35, 15 and 0 µmol quanta m−2 s−1. At saturating PFDs the stabilization time was 2–3 min, but at lower PFDs it was prolonged until the transients in CO2 uptake and fluorescence were completed (5 min). A typical result of this experiment is shown in Fig. 1 (□ and dotted line).

Figure 1.

Light response curves of net CO2 uptake measured in 360 µmol CO2 mol−1 and 21 kPa O2 (□ and dotted lines) and 330 µmol CO2 mol−1 and 2 kPa O2 (▪ and solid lines). The same CO2 exchange data are shown in (a) and (b) but associated with Chl fluorescence (a) and 820 nm transmission (b) data. Fluorescence data are Fm, diamonds, Fs, triangles and F0, circles; 820 nm transmission data are deflection of the signal from the dark level during steady state photosynthesis, Ps (diamonds), deflection of the signal from the dark level under far-red light, PFRL (triangles), deflection for oxidisable P700, Po from pulse-saturation measurements (crosses), and the time constant of post-illumination re-reduction of P700 measured with a 20 ms saturation pulse τ′ (circles). Calculated deflection for fully oxidized P700 Pm is presented by a dashed line. Data presented by empty symbols and dotted lines were measured under 21 kPa O2 and filled symbols and solid lines under 2 kPa O2.

Light response at 33 Pa CO2 and 2 kPa O2

A similar routine as in 21 kPa O2 was applied to measure the light response curve in 2 kPa O2. The idea of this measurement was to avoid fast photorespiratory e fluxes. This increased the precision of the calculated quantum yield of e transport that was related to fluorescence data (below). The intrinsic (maximum) quantum yield of CO2 uptake YCO2 (Table 1) was found from the initial slope of this light response curve. The full curve is shown in Fig. 1 (▪ and continuous line) together with the curve measured at 21 kPa O2.

Table 1.  Birch leaf photosynthetic parameters, related to light reactions and etransport. Average values, absolute (SD) and relative (RSD, %) standard deviations: FW, fresh weight; DW, dry weight (g m−2); Abs, leaf absorption coefficient for PAR; Chl, chlorophyll, µmol m−2; YCO2, intrinsic quantum yield of CO2 fixation (mol CO2)/(mol quanta); aII and aI, relative optical cross-sections of PSII and PSI antenna supporting CO2 fixation plus photorespiration at limiting PFD; aII and aΙ, the same, at saturating PFD; kP0 and kN, relative rate constants for photochemistry at open reaction centres and for maximum non-photochemical quenching, respectively; τmin, time constant of P700 re-reduction, minimum value with completely reduced PQ, without photosynthetic control, ms; τmax the same but with maximum photosynthetic control under O2 and CO2 limitation; τair time constant of P700 re-reduction in 21 kPa O2 and 360 µmol CO2  mol−1; PSIF, PSI density (e pool on the donor side of PSI) determined using ETR based on fluorescence, µmol m−2; PSII, the pool of active PSII evolving O2, µmol m−2; PQ, the pool of e accommodated in interphotosystem e transport chain, mostly plastoquinone, µmol e m−2; PGA, the pool of e consumed for PGA reduction in dark-adapted state of the leaf, µmol e m−2; JN, e flow for (mostly) nitrite reduction, µmol e m−2 s−1; τPGA, lag time between the beginning of illumination and the activation of PGA reduction, s; LHCII and LHCI, total antenna sizes of PSII and PSI, Chl per centre
Clone FW
(g m2)
(g m2)
m− 2)
YCO2aIIaIIaIaIaII + aIkP0kN(Fm –
Clone τ′air
e m−2)
e m−2)
m−2 s−1)

Chlorophyll fluorescence

During the light (and CO2) response measurements, at the end of each exposure under a given PFD (or Ca), a computer-operated procedure was applied to measure steady-state (Fs) and pulse-saturated (Fm) fluorescence yield at the steady state assimilation rate. The measurement pulses of PAM 101 were switched on and after 5 s a saturation pulse of 1·5 s duration was given. During this procedure the background (actinic) PFD was not manipulated. For F0 measurement, actinic light was replaced by FRL for 5 s, then FRL was removed and F0 was recorded immediately after that in the dark at 100 kHz pulse frequency. The recorded values of Fm, Fs and F0 are presented in Fig. 1A for light response curves at 21 and 2 kPa O2.

Relative rate constants for photochemistry in open PSII, kP0, and for biologically controlled non-photochemical excitation quenching kN (assuming that rate constant for fluorescence emission plus biologically uncontrolled thermal conversion kf + kd = 1) were calculated as follows (Bilger & Björkman 1990; Laisk et al. 1997)


where F0 and Fm relate to the measurement PAD, but Fmd is completely dark-adapted Fm that was measured pre-dawn for several leaves and found to be rather invariable (see also Peterson et al. 2001). Fm measured in the dark at the end of light response curve measurement was usually lower than the completely unquenched value Fmd.

Quantum yield of PSII e transport was calculated from fluorescence for each data point of light and CO2 response curves measured at 2 and 21 kPa O2 (Genty, Briantais & Baker 1989):


and from net CO2 uptake (Laisk & Loreto 1996; Peterson et al. 2001):


where JC is e transport rate supporting CO2 fixation and photorespiration, I is absorbed PFD; A, net CO2 uptake rate; RK, Krebs cycle respiration rate in the light; Ks, Rubisco specificity; Cc, calculated CO2 concentration at Rubisco sites, considering CO2 solubility (Eqn 1) and gas phase (rgw) and mesophyll diffusion resistance (rmd); Oc, O2 concentration at the Rubisco sites (calculated considering O2 solubility, Eqn 2, but neglecting with diffusion resistances). The calculated YC was plotted against YF (Fig. 2a). In some experiments the YC point measured at the lowest PFD of 15 µmol m−2 s−1 declined downward. It was due to increasing Krebs cycle respiration RK with decreasing PFD (Kok effect; Sharp, Matthews & Boyer 1984). For this lowest-PFD data point the RK value was left as measured in the dark but RK was decreased for other data points until the lowest-PFD data point did not drop out from the general relationship. This procedure revealed RK, dark respiration, as suppressed due to the Kok effect in the low light (Table 2).

Figure 2.

Relationship between the quantum yield of e transport supporting CO2 fixation plus photorespiration YC (Eqn 8) and quantum yield of PSII e transport YF[Eqn 7, (a)] or quantum yield of PSI e transport YP[Eqn 9, (b)]. Data are from light response curves measured in 21 kPa O2 (◊) and 2 kPa O2 (◆) and from a CO2 response curve measured in 21 kPa O2 (▵). Extrapolation of the relationship in (a) to YF = 1 intersects the ordinate at aII, the relative optical cross-section of PSII antenna and extrapolation of the relationship in (b) to YP = 1 intersects the ordinate at aI, the relative optical cross-section of PSI antenna (both cross-sections supporting CO2 fixation plus photorespiration, values given in the figure for low-light and high-light data points).

Table 2.  Birch leaf photosynthetic parameters related to carbon assimilation. Average values, absolute (SD) and relative standard deviations (RSD, %): FW, fresh weight; DW, dry weight (g m−2); Chl, chlorophyll, µmol m−2; A13 µm and A25 µM, net CO2 assimilation rate at 36 and 72 Pa and 21 kPa O2, µmol m−2 s−1; Rd, RK, RK, Krebs cycle respiration in the dark, under limiting and saturating PFD, respectively, µmol m−2 s−1; rm(2 kPa), rm(21 kPa), rmd, rgw, diffusion resistances (s mm−1), respectively: total diffusion plus carboxylation in mesophyll at 2 kPa O2, the same at 21 kPa O2, diffusion resistance in liquid phase of mesophyll cells and diffusion resistance in gas phase (expressed for CO2 dissolved in water); Ks, Rubisco CO2/O2 specificity for dissolved gases; AC, assimilatory charge, µmol m−2; SCE, specific carboxylation efficiency, mm s−1 of carboxylation conductance per µmol m−2 of assimilatory charge; gc, carboxylation conductance (initial slope of Rubisco kinetic curve), mm s−1; Km(CO2) of Rubisco, µm; Vm, of Rubisco, µmol m−2 s−1.
Clone FWDWChl
(µmol m−2)
A13 µm
m−2 s−1)
A25 µm
m−2 s−1)
m−2 s−1)
m−2 s−1)
m−2 s−1)
SD 21·613·3 82 2·5 3·0 0·28 0·35 0·06
RSD 11·016·8 21·719·215·9 29·347·240
SD 21·811·4 62 2·4 2·5 0·19 0·17 0·05
RSD 10·914·3 15·417·913·4 20·1 22·536
Clone AC
(µmol m−2)
m−2 s−1)
(2 kPa)
(s mm−1)
(21 kPa)
(s mm−1)
(s mm−1)
(s mm−1)
4Average 94·4 0·059 6·1311·06893·4 0·268 0·403 0·378 0·084
SD 20·1 0·009 1·12 0·713 3·7 0·035 0·064 0·088 0·020
RSD 21·315·318·3 6·019·6 3·913·015·823·224·0
80Average100·2 0·059 6·1810·86793·5 0·271 0·419 0·322 0·093
SD 16·7 0·008 1·09 0·512 3·7 0·035 0·059 0·065 0·019
RSD 16·713·317·7 4·718·7 4·013·114·120·220·2

Theoretically, the YC versus YF relationship must be proportional, if all e transported through PSII are later used for CO2 fixation, but the actual graphs were concave (Fig. 2a). We assumed that the concavity was caused by e transport to acceptors other than CO2 (alternative ETR) that increased at light saturation, but at low PFDs there was no or very little alternative ETR. An alternative explanation would be a decrease in PSII antenna at high PFDs, e.g. due to state transition (Andrews, Bredenkamp & Baker 1993). Leaf optical density influences this curve in the opposite direction, toward convexity (Peterson et al. 2001). On the basis of these considerations a straight line was drawn through the origin of the co-ordinates and through the low light data points and extrapolated to YF = 1. The greatest weight in this procedure was given to the data points that were still light-limited but not much influenced by the Kok effect (PFD of 35 and 75 µmol m−2 s−1). The extrapolated YC value at YF= 1 (0·495 in Fig. 2a) shows the relative optical cross-section of PSII antenna aII (Laisk & Loreto 1996; Eichelmann & Laisk 2000). This graphical procedure is equivalent to calculation of aII from Eqn 1 of Eichelmann & Laisk (2000) on the basis of measured Y, Fs and Fm and assuming that nII = 4. The average value of aII was 0·492 ± 0·01 (Table 1). A straight line drawn through high-PFD data points extrapolated to a lower aII at YF = 1. The ratio aII/aII shows the relative proportion of e supporting CO2 fixation plus photorespiration in the total e flow at light saturation (or the degree of PSII antenna decrease due to state transition at the high PFDs). The ratio aII/aII was 0·87 on average (Table 1; also 0·87 in Fig. 2a).

820 nm transmission measurements in steady state

Quantum yield of PSI e transport was calculated for each data point of light and CO2 response curves from the 820 nm transmission signal


and from net CO2 uptake from Eqn 8 and YC was plotted against YP (Fig. 2b). The YC versus YP relationship must theoretically be proportional if all e transported through PSI are used for CO2 fixation, but actual graphs were even more concave than the similar graphs for PSII. We assumed that the concavity was caused by alternative linear (or pseudocyclic) e transport, as in the case of PSII, plus cyclic e flow around PSI. The pseudocyclic plus cyclic e flow increased with light saturation of photosynthesis, but, as in the case of PSII e transport, we assumed that at low PFDs there was little cyclic and pseudocyclic e flow around PSI. On the basis of this assumption a straight line was drawn through the origin of the coordinates and through the low light data points and extrapolated to YP = 1. The YC value at YP= 1 shows the relative optical cross-section of PSI antenna aI (Eichelmann & Laisk 2000). As in the case of PSII, this graphical procedure is equivalent to calculation of aI from Eqn 1 of Eichelmann & Laisk (2000) on the basis of measured Y, Ps, Po and Pm and assuming that nI= 4 at low PFD. From the experiment in Fig. 2b the aI value was 0·472, the average value was 0·485 (Table 1). A straight line drawn through high-PFD data points extrapolates to a lower aI at YP = 1. The ratio aI/aI shows the relative proportion of e supporting CO2 fixation plus photorespiration in the total e flow through PSI at light saturation. Usually the ratio was about 0·5–0·7 (0·53 in Fig. 2b). Clearly, at saturating PFDs e transport through PSI was faster than through PSII, because the concavity of the YC versus YP graph was greater than that of the YC versus YF graph. Data points from light response curves measured at 2 and 21 kPa O2 lay on one and the same dependence between YC and YF (Fig. 2a) and between YC and YP (Fig. 2b) showing that Eqn 8 used for the calculation of electron transport rate correctly considers e flow supporting photorespiration at O2 pressures from 2 to 21 kPa.

820 nm transmission transients

In the light P700 is partially oxidized and it becomes reduced when illumination is interrupted. The speed of P700 re-reduction depends on e transport rate through Cyt b6f and is an important parameter of the photosynthetic machinery. The post-illumination re-reduction of P700 was measured using the time course of 820 nm signal after illumination was rapidly interrupted by the shutter. Two procedures were used to operate the shutters.

The simpler (AltP-procedure) was designed to measure the P700 re-reduction time constant at the PQ pool that was established during steady-state photosynthesis. Illumination was interrupted and the time course of the 820 nm signal was recorded in the oscilloscope mode of the data logger (Fig. 3, upper, thin line). This procedure was successful only at higher PFDs when P700 was sufficiently oxidized and PQ was sufficiently reduced to provide e to reduce Cyt f, PC and P700. Thus, the time constant τ of P700 reduction obtained by this procedure was determined by both, e transport capacity of Cyt b6f and by the redox state of PQ pool, but it was the actual value during photosynthesis. Another (CtrlP-) procedure was designed to measure the maximum pulse-oxidizable P700 (Po) and the P700 re-reduction time constant with completely reduced PQ pool. This time constant characterizes the e transport capacity (turnover rate) of Cyt b6f not limited by the absence of e donor. In this procedure, a 20 ms pulse of 10 500 µmol quanta m−2 s−1 was applied before the leaf was darkened. The data processing program calculated an exponent with a time constant τ′ that fitted the recorded time course of P700 reduction. The values of τ′ measured in different steady states during the light response curve are presented in Fig. 1b (circles). In the particular experiment of Fig. 1 the minimum value τmin was 8 ms in the light-limited state and it increased to 13 ms in the light-saturated state due to the onset of photosynthetic control of e transport by proton back-pressure. The average value of the time constant at the completely open state (minimum H+ backpressure) τmin was 6·5 ms, during photosynthesis in air (36 Pa CO2, 21 kPa O2) it increased to 10·5 ms (the value of τ′ in air shows how much H+ backpressure controls ETR during normal photosynthesis). Over the maximum range of photosynthetic control at low CO2 and O2 concentrations τ′ increased by about 4·5 times (Table 1).

Figure 3.

Recording of the 820 nm signal corresponding to post-illumination re-reduction of P700. Thin line, illumination was interrupted in the steady state of photosynthesis (AltP-procedure); thick line, a 20 ms pulse of PFD = 10500 µmol m−2 s−1 was given before the leaf was darkened (CtrlP-procedure, see text). Time constants of fitted exponents, τ for the AltP-procedure and τ′ for the CtrlP-procedure are given in the figure; Ps is the steady-state deflection, Po is the deflection for pulse-oxidizable P700, and Pm corresponds to completely oxidized P700.

PSI density

Knowing the time constant of P700 re-reduction and ETR when light was turned off, it was possible to calculate the pool of e vacancies being filled on the PSI donor side during the post-illumination process:


where subscript denotes whether ETR was calculated from fluorescence F or from 820 nm signal P. The corresponding e transport rates express as follows:


At low PFDs when P700 was little oxidized, both PSI numbers were equal and increased when PFD increased and P700 became more oxidized (Fig. 4). Because ETR through PSI was faster than through PSII by the cyclic e flow, the values of NIP and NIF differed at high PFDs. In this particular leaf, beginning from a certain P700 oxidation level, NIF stayed constant at 1·15 µmol e m−2, whereas NIP increased to 2 µmol e m−2. The difference of these values may be interpreted to show how many PSI turn around in the cyclic flow, on average, however, the rate of the cyclic e flow detected by this method is rather fast and it is not clear that all of it is physiologically necessary for additional ATP synthesis. Rather, it may reflect back-reaction to P700+ from the low-potential acceptor side e carriers, a kind of cyclic e flow uncoupled from H+ transport.

Figure 4.

Determining PSI density (donor side electron pool). The pool was calculated from Eqn 10 (diamonds) or Eqn 11 (squares) using data from the light response curves measured in 21 kPa (empty symbols) or 2 kPa (filled symbols) O2 and plotted against the relative oxidation state of P700, Ps/Pm.

PQ and PGA pools

These pools were determined from O2 evolution measurements. The leaf was pre-adapted in channel 1 of the gas system in the dark at 2 kPa O2 and 33 Pa CO2 for 20 min. In channel 2 O2 and CO2 concentrations were set to zero (minimum 1–3 Pa). Twenty seconds before the measurement of the dark-light induction the leaf chamber was switched to channel 2 and after the reference line stabilized the light was turned on by opening the shutter (PFD = 1300 µmol m−2 s−1). The O2 analyser recorded a complex induction transient (Fig. 5) that was analysed for the totals of e transport chain pools (the first, fast O2 burst, mainly PQ) and PGA present after the dark adaptation (the second, slower maximum in O2 evolution). The residual level of O2 evolution finally remaining in the absence of CO2 and O2 probably characterized nitrite reduction and other alternative reductions in the leaf that continued in the absence of CO2 and O2. Figure 5 shows several examples recorded with birch leaves, differing in the depth of the dark inactivation and in the rate of light activation of the ATP synthase and GAP dehydrogenase, resulting in different resolution between the first peak, reflecting the e transport chain pools and the second peak, reflecting the PGA reduction (the dark adaptation time was 20 min for all leaves). A procedure that accounted for the signal transfer function of the gas system was applied to correctly resolve the two pools, PQ and PGA. The ‘clean’ transfer function of the gas system was determined from a single turnover flash and actual recordings were deconvoluted as a sum of such elementary bell-shape functions. This procedure resulted in four parameters (Table 1): PQ and PGA pools (presented as the number of e), the lag time of the induction of PGA reduction, τPGA and the rate of nitrite (and other non-CO2) reduction, JN.

Figure 5.

Induction of O2 evolution in leaves in the absence of CO2 and O2. Leaves were dark-adapted for 20 min in 33 Pa CO2 and 2 kPa O2, then CO2 concentration was decreased to zero and O2 pressure to 20 Pa and after 20 s PFD of 1100 µmol quanta m−2 s−1 was turned on. Different traces were recorded with different birch leaves.

PSII pool

This pool was determined from O2 yield from a saturating single turnover flash. Before this measurement the leaf was pre-adapted in channel 1 at 2 kPa O2, PFD of 35 µmol m−2 s−1 and CO2 pressure of 33 Pa. For the measurement of the PSII pool the leaf chamber was switched to channel 2, where O2 pressure was 1–3 and CO2 pressure 33 Pa, and after 20 s actinic light was replaced by FRL that completely oxidized the PQ pool. One single turnover flash of 60 µmol quanta m−2 was given, and the corresponding O2 evolution recorded. The pool of active PSII was calculated as four times the O2 evolution, µmol m−2, obtained by integrating the bell-shape O2 evolution peak (Oja & Laisk 2000).

CO2 response curve at 21 kPa O2

The steady-state photosynthetic rate obtained after stomata opened at 21 kPa O2, 36 Pa CO2 and PFD of 760 µmol m−2 s−1 was recorded as A13 µm (Table 2). In the following, a CO2 response curve was measured jumping from the initial steady state to different CO2 pressures (the steady state at 36 Pa CO2 was re-established in channel 1 between changes, while the CO2 pressure was varied in channel 2 of the gas system). The CO2 pressures in channel 2 were 8, 4, 0, 20 and 72 Pa. Stabilization time at each CO2 pressure (except 72 Pa) was about 1 min. It was considered that this time was sufficient to let the CO2 fixation and photorespiratory CO2 evolution equilibrate via the pools of the glycolate pathway, but during this time the pools of the CRC were little depleted and Rubisco little decarbamylated at concentrations near the CO2 compensation point. As a result, the value of the CO2 compensation concentration obtained was suitable for the calculation of the Rubisco specificity factor, but the trend of the CO2 response was linear down to zero CO2, which allowed us to correctly calculate mesophyll diffusion resistance and to detect the inhibitory effect of O2 on the initial slope of the CO2 response. The highest CO2 pressure of 72 Pa was set to the beginning of CO2 saturation, corresponding to the maximum CO2 and light-saturated photosynthetic rate, denoted A25 µm in Table 2. The full CO2 response is shown in Fig. 6 (e and dotted lines).

Figure 6.

CO2 response curves of net CO2 uptake measured under PFD of 760 µmol quanta m−2 s−1 in 21 kPa O2 (◊ and dotted lines) and 2 kPa O2 (◆ and solid lines). The left of the two parallel curves is presented relative to carboxylation site CO2 concentration Cc, the right relative to cell wall CO2 concentration Cw on the axis of abscissa. Four lower data points in 2 kPa O2 were measured in the steady state and five upper data points were obtained from the initial carboxylation rate after transition from the steady state to the higher CO2 concentration. Inserted is an enlarged view of the initial parts of the curves, where the slopes denote gc of carboxylation and gm of total mesophyll conductance.

CO2 response curve at 2 kPa O2

This curve was measured jumping from the initial steady state at 20 Pa CO2 and 2 kPa O2 in channel 1 to CO2 pressures of 20, 10, 5 and 0 Pa in channel 2 (steady state was re-established in channel 1 between changes). The results of this experiment on the same leaf are shown in Fig. 6, with the four lowest data points of the curve denoted 2 kPa O2 (r and continuous line). Waiting time at each CO2 concentration was 50 s, to avoid depletion of the CRC pools and decarbamylation of Rubisco. This measurement revealed the slope of the CO2 response curve uninfluenced by the presence of O2 (total mesophyll conductance, gm, Fig. 6, insert). Total mesophyll resistance (rm = 1/g, Table 2) contains the components of diffusion and carboxylation resistances, which we shall separate below by calculating the diffusional component (the term carboxylation resistance is used to denote the reciprocal of the carboxylation conductance, which is the initial slope of the CO2 response curve of RuBP carboxylation rate with respect to CO2 concentration at the carboxylation sites). After the measurement at 0 Pa CO2 the leaf was again stabilized at 20 Pa in channel 1 for the subsequent measurement of the full kinetic curve of Rubisco.

Full kinetic curve of Rubisco

Rubisco Vm is usually higher than the maximum steady-state rate of photosynthesis, because the latter is limited by end product synthesis and/or e transport rate that limit RuBP resynthesis. Experimentally, the actual high capacity of Rubisco reaction could be seen only as a short peak in CO2 uptake after a transition to a high CO2 pressure. We stabilized the leaf in channel 1 at 20 Pa CO2 and 2 kPa O2 and from this steady state we made transitions to channel 2 where CO2 pressure was 210, 150, 100, 72 and 40 Pa. Measuring time at each CO2 concentration in channel 2 was only 10 s and only the initial CO2 uptake rate, extrapolated to the moment of transition, was considered to be the actual Rubisco rate at that CO2 concentration (cell wall, Cw, and chloroplast, Cc, CO2 concentrations were calculated for the extrapolated rate).

A recorded transient to 210 Pa CO2 is shown in Fig. 7 (empty diamonds). During this transient, CO2 uptake was caused by the solubilization of CO2 plus carboxylation by Rubisco. Solubilization (plus bicarbonate formation) was measured separately from a transient from 60 to 0 Pa CO2 in the dark, measuring the peak of CO2 evolution. The CO2 solubilization trace was proportionally normalized to the actual measurement concentration (Fig. 7, empty triangles) and subtracted from the CO2 uptake trace to obtain the true carboxylation rate (filled diamonds). At 210 Pa CO2 solubilization contributed about a half of the initial peak of CO2 uptake, but at lower concentrations the proportion of solubilization was smaller. The curve presented by filled diamonds was integrated to obtain cumulative amount of CO2 fixed from the beginning of the transient (solid line and right ordinate). For extrapolation to the moment where CO2 concentration was increased, the CO2 uptake rate was plotted against the cumulative pool of CO2 (Fig. 8). In this presentation empty diamonds denote the recorded data points influenced by the transfer function (inertia) of the gas exchange measurement system and filled diamonds show the true CO2 fixation rate. On the other hand, the total available pool of RuBP in the beginning of the transient was about 100 µmol m−2, as seen from the post-illumination CO2 uptake (below). Thus the linearly decreasing rate of CO2 fixation (the linearity of the response was proven by Fig. 10) was observable only in the range of filled diamonds placed to the left of the vertical line showing the pool of available AC (= RuBP), and only these data points were considered for the extrapolation of the CO2 fixation process to the moment of transition. In this procedure, the data point corresponding to 210 Pa of external CO2 was the least reliable, at lower CO2 pressures the curve was determined with a better accuracy. A full kinetic curve of Rubisco obtained from these extrapolated values is presented in Fig. 6 with the solid line and five upper filled diamonds, whereas lower data points present the CO2-limited part measured as the steady-state CO2 response curve. The curve is a Michaelis–Menten type hyperbola with Km(CO2) of 11·5 µm and Vm of 53 µmol CO2 m−2 s−1. Average values of these Rubisco parameters are presented in Table 2.

Figure 7.

Transients in CO2 uptake recorded after a stepwise increase of the CO2 concentration from 200 to 2100 µmol mol−1. e, original recording in the light; ▵, the same transient in the dark (presents CO2 solubilization); r, difference between the transients in the light and in the dark, the RuBP carboxylation rate. Solid line and right ordinate, integral of the RuBP carboxylation rate.

Figure 8.

Extrapolation of the RuBP carboxylation rates recorded at CO2 concentrations of 2100, 1500, 1000, 760 and 400 µmol mol−1 (from up to down) to the time of transition (zero CO2 uptake sum). Data points presented by empty symbols are distorted by the slow response of the measuring system, filled symbols are free of this distortion; vertical line indicates the pre-accumulated RuBP pool (assimilatory charge AC).

Figure 10.

Dependence of the mesophyll conductance (carboxylation and diffusion in series, thin line) and carboxylation conductance (thick line) on the residual assimilatory charge. At any point of the post-illumination CO2 uptake process the carboxylation conductance was calculated from Eqn 15 and plotted against the residual AC, integrated from the end of the curve to that point. The slope of the dependence of carboxylation conductance on AC is termed specific carboxylation efficiency (SCE) (given in the figure).

Rubisco CO2/O2 specificity

The Rubisco specificity factor Ks was calculated from the difference of CO2 compensation points at 21 and 2 kPa O2 (Sumberg & Laisk 1995) as


CO2 compensation concentrations Γ were determined from the A versus Cw response curves at 21 and 2 kPa O2 in µm, dissolved O2 concentration Ow in µm was calculated using the solubility of O2 (Eqn 2) and the external O2 concentration.


Three values of respiration rate are presented in Table 2. Dark respiration Rd was measured after the light response curve. A slightly inhibited value RK had to be used to linearize the initial part of light response curve of e transport, showing that Krebs cycle was a little inhibited at PFDs of 30–100 µmol m−2 s−1 (Kok effect). The third value (RK) was chosen such that Γ was proportional to O2 concentration (the y-offset became zero) and corrected for re-assimilation considering the diffusion and carboxylation conductances. The small value of RK shows that CO2 evolution from the Krebs cycle was considerably inhibited at high PFDs, a prerequisite for obtaining a more or less correct Ks value from this simplified procedure.

Post-illumination CO2 fixation

This experiment was designed to measure the response of CO2 fixation to the decreasing concentration of RuBP after its regeneration was stopped in the dark. The leaf was stabilized at 10 Pa CO2 in 2 kPa O2 and 760 µmol quanta m−2 s−1, illumination was interrupted and the post-illumination transient from CO2 fixation to respiratory CO2 evolution was recorded (Fig. 9). The area under the curve was integrated backwards, from the end of the process to the beginning, assuming that respiration rate was linearly increasing during the post-illumination process (in the light the initial respiration rate was assumed to be RK, about 15% of the final rate, thin continuous line). The calculation program also considered the time response of the gas system (although it was much faster than the post-illumination process) and RuBP consumed by the re-assimilation of respiratory CO2 during the post-illumination process. Total area under the curve is termed assimilatory charge, AC = 101·9 µmol CO2 m−2 for this leaf (average values in Table 2) and this parameter closely represents the RuBP pool at the moment when illumination was terminated (Ruuska et al. 1998).

Figure 9.

Recording of the post-illumination CO2 fixation. In a steady state at 100 µmol CO2 mol−1 and 2 kPa O2, PFD of 760 µmol m−2 s−1 was turned off at the vertical dotted line. Respiration rate (solid line) was assumed to increase from the light-suppressed state (15% of the final value) to the value recorded in the dark. The integral of the area between the CO2 uptake and respiration is given in the figure as assimilatory charge, AC.

The dynamic process of the post-illumination CO2 fixation was analysed considering that RuBP pool could be found at any moment of the post-illumination CO2 uptake curve as an integral of the area under the curve from the end of the process to the given time moment. Considering that while CO2 fixation rate was decreasing during the process, Cc was correspondingly increasing, we calculated the initial slope of the Rubisco kinetic curve (carboxylation conductance gc) for any time moment as


and plotted it against the remaining AC. The obtained gc versus AC curves were linear (Fig. 10, thick line) and the slope of the curves, gc per unit of AC, was termed specific carboxylation efficiency (SCE). When Cw (cell wall CO2 concentration) was used instead of Cc (carboxylation site concentration), the gm versus AC response was slightly hyperbolic (thin line).

Mesophyll diffusion resistance

This parameter was calculated using the values of ETR from fluorescence measurements at the CO2 photocompensation point Γ* in 21 kPa O2 (Appendix). This approach has the advantage that data measured at CO2 compensation are used, where CO2 response is linear and the slope is close to maximum, whereas Laisk & Loreto (1996) used the value of A at higher CO2 concentrations where the response may decline from linearity. For finding rmd, ETR through PSII was calculated from fluorescence using Eqn 12, optical cross-sections of PSII, aII and aII′ were found as described above. The resistance rmd was chosen such that JC from Eqn 8 became equal to JF from Eqn 12. The values of rmd found from this condition were 0·08–0·09 s mm−1 on average (Table 2) and were used to calculate the carboxylation site CO2 concentration Cc (Fig. 6).

Possible measurement error

With the system used, the direct measurement errors were relatively small. The random noise of the CO2 analyser signal came mostly from bubbles at the manostats (overflow tubes) of the gas mixer, the noise of the gas analyser itself was about three times smaller. In steady state a gas analyser reading was taken during 5 s, averaging 1000 inputs of the data logger. The SD of the leaf gas exchange rate from one 5 s reading was estimated to be 0·02 µmol m−2 s−1 at CO2 pressures of 20–40 Pa. This noise was important at quantum yield measurements at low PFDs when rates were about 1–5 µmol m−2 s−1. The small instrument scattering of CO2 uptake values is seen from the light response curves (Fig. 1). The noise of the psychrometer signal was about 1 mV at average signal levels of 160 mV. An estimate of the random error of calculated rm yields ± 2% at average Cw values. Measurements were not carried out when gas phase resistance rgw exceeded 1 s mm−1. The small scattering of calculated Cw and Cc values is seen from the CO2 curves in Fig. 6. Wider scattering of results (Table 2) is, thus, caused by biological variation of leaf parameters.

In O2 evolution measurements SD for the total evolution from one single turnover flash was less than ± 3%. A more important component of error was introduced by human and leaf factors. For example, the AltP and CtrlP procedures were applied in series, but the steady-state level of photosynthesis did not always re-establish between the procedures (H procedure involving the fluorescence saturation pulse was executed at the end of an exposure to a given PFD or CO2, because it considerably disturbed the steady state).

Data processing

The system-operating reco program records and saves raw data without any on-line processing. Later, an rda program is used to re-open the recorded file on the screen, to draw reference lines, take the readings at desired parts of the lines and send them to the next, synte program. As far as the measurement procedure was automatic, also data reading could be automatic, considering that changes in experimental parameters occurred at precisely predicted time moments. The post-illumination P700 reduction transients were processed by a nested program that approximated the traces by an exponent. The post-illumination CO2 fixation traces were processed by another nested program. The readings of recorded CO2, O2, H2O concentrations and optical signals and the results of the nested programs were sent to a synte program that calculated leaf temperature, CO2 and O2 concentrations, actual absorbed PADs, Ci, Cw and Cc values, etc. The final data analysis and graphing were carried out in Microsoft Excel. After obtaining the necessary skills, experimentation time and data processing time were equal, both about 2·5 h per leaf.


As an example of application of the above-described routine we present the results of measurements on birch leaves carried out during the summer of 2000 at Suonenjoki Forest Research Station, Finland. One leaf was taken from each of the chosen four trees of clone 4 and four trees of clone 80 growing on the experimental site of the station, and the measurements were repeated six times, as listed in the Methods. Some photosynthetic parameters were relatively constant during the summer and scattered little, such as relative optical cross-sections of PSII and PSI, aII and aI, also Ks and Km(CO2) of Rubisco (Fig. 11a & 12a), whereas others, such as the densities of PSII and PSI, also Rubisco Vm and leaf resistances for CO2 uptake scattered widely (Figs 11b & 12b). The resistances had visible seasonal tendency to increase.

Figure 11.

Seasonal courses of photosynthetic parameters in leaves of birch clone 4. (a) Relative cross-sections of PSII antenna aII (r) and PSI antenna aI (▵) and their sum aII + aI (□). (b) PSII (r) and PSI (□) densities

Figure 12.

Seasonal courses of photosynthetic parameters in leaves of birch clone 4. (a) Rubisco parameters: CO2/O2 specificity Ks (□), Rubisco Vm (µmol CO2 m−2 s−1, r) and Km(CO2) (µmol, ▵). (b) Leaf resistances to CO2 transport: mesophyll diffusion resistance rmd (▵), mesophyll resistance (rmd + rc) in 2 kPa O2 (r) and the same in 21 kPa O2 (□).

To characterize the parameters we bring average values and absolute SD and relative (RSD) standard deviations of the photosynthetic parameters from measurements in July, when seasonal trends were not strong, altogether 12 measurements, from round 2 (short shoot) and round 3 (short and long shoots, Tables 1 and 2). The intrinsic quantum yield of CO2 fixation YCO2 was close to 0·09. It, as well as the relative optical cross-section of PSII antenna aII had relatively small RSD of 2% or less. Although aI varied more, it is important that the sum of the optical cross-sections aII + aI was close to 0·97–0·98 and exceeded 1·0 only in a few individual measurements (Fig. 12), which shows the correct calibration of CO2 and PFD measurements. The residual cross-section a0 of 0·02–0·03 is serving nitrate reduction and absorption in non-photosynthetic structures, including PSIIβ and other inactive PSII forms. The density of PSII was 1·15 µmol m−2 on average, that of PSI was 1·45 µmol m−2. Thus, the population of PSI was 1·26 times denser than that of PSII, but both varied more than the relative antenna cross-sections. The average PSII and PSI antenna sizes, calculated considering total Chl m−2 and the fractions aII and aI were 170 Chl at PSII and 120–140 Chl at PSI. Assuming that most of the initial peak in the induction of O2 evolution (Fig. 5) was attributable to PQ reduction, the ratio was 10–11 PQ molecules per PSII (PQ in Table 1 is given as the number of e carried by this compound).

Maximum variable fluorescence (Fm − F0)/Fm was 0·82 and varied very little, resulting in the calculated relative rate constant for photochemistry kP = 5·18 and 4·98 for the clones (the value shows how many times photochemistry is faster than physical excitation quenching by fluorescence emission and thermal conversion). The maximum relative rate constant for non-photochemical quenching kN was 3·88 (3·55) in the same relative units, but varied considerably. The basic e transport parameter, the fastest Cyt b6f turnover time τmin was shorter and varied less than its value during photosynthesis in air, τair.

The Rubisco-related parameters Vm and gc (and rm) were variable (RSD > 15%), as expected if Rubisco content and activity were adjusted to the growth conditions of an individual leaf. The variability of rmd was probably mostly caused by its measurement method, as a difference of two big resistances. The actual photosynthetic rate at 36 Pa CO2 was mainly influenced by stomatal resistance, as it adjusted in the leaf chamber, and it was rather variable. The rate at 72 Pa CO2 was also variable, showing that sucrose and starch synthesis rates, probably limiting this maximum rate of CO2 fixation, differed in individual leaves. The RSD of Ks and Km(CO2) of 4–6% is clearly lower than that of other parameters.

No other significant differences between the clones could be detected, except that the dark-light induction of PGA reduction τPGA tended to be faster and stomata tended to be more open (rgw smaller) in clone 80.


In this work we described a computer-operated routine of condensed experimental procedures for the measurement of a number of intrinsic photosynthetic parameters in intact leaves, suitable for diagnosing the state of the photosynthetic machinery of leaves in ecophysiological investigations. The parameters are calculated from non-destructive, kinetic measurements of CO2 uptake, O2 evolution, Chl fluorescence and leaf transmission at 820 nm.

Parameters of light reactions

Our system detected the maximum quantum yield of CO2 fixation YCO2 of 0·090, similar to the value reported by Long, Postl & Bolhar-Nordenkampft (1993), but higher than that of Ehleringer & Björkman (1977). The quantum yield for CO2 fixation is lower than that of O2 evolution, of 0·106 (Demmig & Björkman 1987; Lal & Edwards 1995), probably because some alternative reductions (e.g. nitrite) are visible as O2 evolution but not as CO2 uptake (Lal & Edwards 1995). The quantum yield of photosynthesis is determined by the efficiency and co-operation conditions of the two light reactions of photosynthesis (Eichelmann & Laisk 2000). Our data (Fig. 2a & Table 1) show that at low PFDs the maximum PSII efficiency (Fm − F0)/Fm was 0·82, which is very close to frequently reported values of this parameter in not or little photo-inhibited leaves (e.g. Baker et al. 1994 Cornic 1994). In Table 1 we also present the values of relative rate constants kP0 and kN (Eqns 5 and 6) based on the Stern–Volmer model of fluorescence, scaled to the basic excitation decay rate constant kf + kd = 1 (Laisk et al. 1997). The rate constant kP0 (Table 1) was calculated for the dark state after the measurement of the light response curve and it shows that photochemistry of open PSII centres was maximally about five times faster than the basic physical decay of excitation. In overnight dark-adapted sunflower leaves kP0 was 6·7 (Peterson et al. 2001). The lower value in birch probably indicates a slight photo-inhibition after the experimental treatment, rather than a difference in the excitation capture processes. The regulatory non-photochemical rate constant kN was calculated for the maximum PFD and it was 3·7 times faster than the basic physical decay of excitation. Thus, in these leaves the non-photochemical quenching was not exactly complementary to the photochemical quenching, although the two processes still had quite comparable rate constants, as also shown earlier (Laisk et al. 1997). It is possible that the wider variation of kN compared to kP0 is caused by different equilibrium levels of xanthophylls in individual leaves dependent on their actual light environment during growth.

Unexpectedly, the pulse oxidation method detected also a significant PSI acceptor side closure at low PFDs, due to which excitation losses at PSI were about equal to losses at PSII. As a result of the equal losses, the calculated antenna cross-sections aI and aII became also equal, aII = 0·492 and aI = 0·488 (0·481 in clone 80; Table 1), aII/aI = 1·01. Earlier we found that in tobacco losses at PSI were smaller than at PSII, resulting in the ratio aII/aI of 1·25 (Eichelmann & Laisk 2000). An even higher value of aII/aI = 1·32 was calculated on the basis of the absorption spectra of the chlorophyll-protein complexes in high light-grown pea leaves (Evans 1986), whereas an average ratio of 0·54/0·38 = 1·42 of chlorophyll associated with PSIIα and PSI was found by Melis (1989). Our present result shows not the Chl distribution but the ratio of the actual optical cross-sections, which is influenced by Chl packing, the distribution of Chl a/Chl b and spectral distribution of light (Melis 1989). The pulse-oxidation measurements of P700 detected a considerable fraction of acceptor-side closed PSI (Klughammer & Schreiber 1994; Laisk & Oja 1995), whereas earlier we assumed no PSI acceptor side closure occurred at low PFDs. The about equal PSII and PSI antenna cross-sections obtained in this work justify the frequent intuitive assumption that excitation is equally shared by the two photosystems (e.g. Loreto et al. 1994; Brestic et al. 1995), at least when plants are grown under daylight. If the pulse illumination itself created closed PSI during a few milliseconds, the method could overestimate the portion of acceptor side-closed PSI, leading to overestimated PSI antenna size.

Contrary to the relative antenna size, the density of photosystems per leaf area was rather variable. No trend in the PSI density could be detected, and that of the PSII tended to be lower in young leaves. The range of the measured PSII pools extended from 0·6 to 1·6 µmol m−2, in general agreement with published values (Evans & Terashima 1987; Lee & Whitmarsh 1989; Backhausen et al. 2000). Our method showed a little more PSI than PSII, though the lower, fluorescence-based ETR was used to calculate the number of PSI (Table 1). This is in agreement with some reports (De la Torre & Burkey 1990; Burkey 1993) but other authors have found equal numbers of PSII and PSI or even a smaller PSI than PSII density (Graan & Ort 1984; Evans & Terashima 1987; Lee & Whitmarsh 1989; Backhausen et al. 2000). Interpreting the cross-sections aII and aI as relative distribution of Chl between the photosystems, we calculated the antenna sizes, LHCII = 170 and LHCI = 120–140 Chl. This average PSII antenna size corresponds to the core antenna plus three LHCII trimers (van Amerongen, Valkunas & van Grondelle 2000), suggesting that the optical cross-sections were still rather close to Chl distribution between the photosystems.

The dark re-reduction kinetics of P700 were close to exponential, provided that PQ was sufficiently reduced to ensure the availability of e. The shortest τmin = 6·3 ± 0·4 ms, measured with saturation pulse pre-illumination, scattered relatively little, suggesting that the parameter is a basic enzyme constant characteristic to the Cyt b6f complex. This τmin value in birch was close to the minimum τ measured in sunflower (Laisk & Oja 1994) and pea (Harbinson & Hedley 1989) and the control range τmax/τmin = 4·5 was also similar to that in our earlier experiments with sunflower (Laisk & Oja 1994). In the latter work we concluded that photosynthetic control did not change τ over the range of limiting PADs, but we did not consider that at low PADs the little-reduced PQ level could become rate-limiting, increasing τ for P700 re-reduction. The new method applied in this work, using pre-illumination with a saturation pulse, showed that τ′ still increased when photosynthesis increased over the light-limited part of the light response curve. Thus, even the ΔpH accompanying the relatively free e transport under atmospheric CO2 and O2 concentrations induces significant down-regulation of the turnover of Cyt b6f.

The interpretation of the 820 nm signal is complicated by its multiple nature. When the PSI donor side is completely oxidized, then about one-third of the signal is theoretically generated by PC+, whereas two-thirds originate from P700+ (Harbinson & Hedley 1989). Because the titrated Em of P700 is for about 0·1 V more positive than that of PC (Klughammer & Schreiber 1991), the equilibrium between the two compounds predicts that when the signal is gradually increasing (e.g. during the increasing PFD), then the first about 20% of the amplitude are caused by PC+ only, then a range of mixed signals follows, and only the last third of the signal, close to the complete oxidation, is a pure P700+ signal. Considering this, one should interpret the fast decrease of PSI quantum yield that happened during the light-limited part of the response curve, when the 820 nm signal was still small (Fig. 2b), as completely caused by the acceptor side closure, because P700 was not yet oxidized at all. Although the pulse method indicated more acceptor side closure at low PFDs, we are still reluctant to state that acceptor side closure is the only cause for reduced PSI quantum yield that occurs at the low rates at limiting PFDs. Similarly, if the exponential shape of the post-illumination P700 rereduction (Fig. 3) was only apparent, caused by the smooth transition from the P700 signal in the beginning of the trace to the PC signal at the end of the trace, then the calculated time constants were overestimated, as well as the PSI densities based on these time constants. Thus, the problem of quantitative interpretation of the 820 nm signal remains open and the investigations to deconvolute the signal in its components are in progress in our laboratory.

Rubisco and CO2-limited photosynthesis

Under CO2 limitation Rubisco specificity determines the balance between gross CO2 fixation (RuBP carboxylation) and CO2 evolution from photorespiration. The Rubisco specificity factor Ks = 93·5 scattered little and the value may be considered a kinetic constant of Rubisco. Values of Ks in leaves, of 102 (wheat), 94·1 (spinach) (Brooks & Farquhar 1985; Cornic & Briantais 1991) and on the extracted enzyme in vitro, 89·9 (wheat) 78–82 (spinach) (Jordan & Ogren 1981, 1984; Kane et al. 1994) have been reported. Differences can be explained by measurement temperatures, but even at a constant temperature we have found that Ks may vary between 85 and 96 at 23 °C in different species (Laisk & Sumberg 1994; Sumberg & Laisk 1995). Differences between the reported Ks values exceed the measurement error and it would be interesting to know whether Rubisco properties are really different or leaf structure (different relative placement of chloroplasts and mitochondria, different respiration in the light RK) can influence the result obtained using the gas exchange method. Knowing the Ks value allows one to calculate the proportions of RuBP carboxylation and oxygenation at any combination of CO2 and O2 concentrations, provided that diffusion resistances for calculating the CO2 concentration at the carboxylation site Cc are known. This relationship allowed us to reliably calculate the total ETR that supports photosynthesis and photorespiration at 2, as well as at 21 kPa O2 (Eqn 8). The result will allow the measurement routine to be shortened in future by omitting the light response curve at 2% O2.

The gas phase diffusion resistance rgw was the most widely scattering parameter. This shows that stomatal opening under the constant conditions in the leaf chamber was still strongly influenced by the prehistory of the leaf. The mesophyll diffusion resistance rmd may contain a fraction of the gas phase diffusion resistance in intercellular spaces, but mostly it reflects CO2 transport through cell wall, cell membrane, a thin layer of cytosol, chloroplast membranes and a part of stroma, where the Rubisco active sites are situated (Evans & Loreto 2000). In this work we used the e transport method to determine the mesophyll diffusion resistance rmd. ETR was calculated from Chl fluorescence measurements and the mesophyll diffusion resistance was chosen such that the reassimilated flux of photorespiratory CO2 was sufficiently fast to make the two e transport rates equal, one calculated from the net CO2 uptake considering photorespiration and the other calculated from fluorescence (Loreto et al. 1992; Harley et al. 1992; Laisk & Loreto 1996). The average mesophyll diffusion resistance determined this way, rmd = 0·08–0·09 s mm−1, made up about 20% of the total mesophyll resistance (0·4 s mm−1) in air containing 21 kPa O2 and it tended to increase in older leaves (Fig. 12). Under 2 kPa O2rmd was 30% of the total rm, because the carboxylation resistance decreased. The role of rmd in the total resistance for CO2 transport was still small, considering also the stomatal component (about 0·3 s mm−1) in the series of resistances. This result is in agreement with our earlier measurements on different species where the same method was used (Laisk & Loreto 1996). Thus, although rmd may form a relatively big part of the mesophyll resistance in some woody species (Lloyd et al. 1992; Epron et al. 1995; Syvertsen et al. 1995), it is not critically rate-limiting the CO2 fixation rate in birch.

Carboxylation resistance, associated with binding CO2 to RuBP, still dominates in the liquid phase of mesophyll cells (we use the term ‘resistance’ to emphasize that rate limitation by Rubisco is the best understood as one component in the series of resistances). In this work we applied the CO2 pulse method to measure the full kinetic curve of Rubisco (Laisk & Oja 1974; Oja 1985; Ruuska et al. 1998). The gas system, specially designed for fast-response measurements (Laisk & Oja 1998) allowed the CO2 concentration in the leaf chamber to be changed within 0·5 s and to measure the initial CO2 uptake rate after 1·6 s from the concentration change. During this time a part of pre-accumulated RuBP was already consumed, but we extrapolated the reaction rate to the initial moment of transition, to find the rate that corresponded to the maximum pool of RuBP. Measurements showed that the response was a Michaelis–Menten type hyperbolic curve with Km(CO2) = 11 µm. This Km value well agrees with in vitro measurements (Yokota & Kitaoka 1985) and may be considered an enzyme constant. Consequently, Vm and the initial slope (carboxylation conductance gc = Vm/Km) are proportional, what also has been shown on Rubisco-deficient transgenic tobacco (Ruuska et al. 1998). The maximum rate Vm varied relatively more than Km(CO2), as expected if Rubisco activity was adjusted to growth conditions in individual leaves. The average value, Vm = 68 µmol m−2 s−1, exceeded the maximum CO2 and light-saturated photosynthetic rate A25 µm = 18·7 µmol m−2 s−1 by 3·6 times and was similar to Vm of Rubisco in wild-type tobacco (Ruuska et al. 1998). This confirms that the maximum steady-state photosynthetic rate is not limited by Rubisco but by the RuBP regeneration capacity (Laisk & Oja 1974; von Caemmerer & Farquhar 1981). The latter may be limited by the capacity of e transport chain (Cyt b6f) or by sucrose and starch synthesis rates that limit the turnover of Pi, which then limits the rate of ATP synthesis (Laisk & Walker 1986; Sivak & Walker 1986).

Kinetics of post-illumination CO2 fixation indicate mainly the relationships between the RuBP carboxylation rate and RuBP concentration (Laisk, Oja & Kiirats 1984; Ruuska et al. 1998), although some RuBP may be regenerated during the post-illumination period from the pools of ATP and triosephosphates (Sharkey, Seemann & Pearcy 1986). The measured AC pool in birch leaves was about 100 µmol m−2, typical for leaves of C3 plants with relatively slow photosynthetic rates, but smaller than in sunflower that has faster photosynthesis (about 200 µmol m−2 (Laisk et al. 1984; Osmond, Oja & Laisk 1988). Under low CO2 and O2 concentrations the directly measured RuBP pool is similar to typical AC values (Badger, Sharkey & van Caemmerer 1984; von Caemmerer & Edmondson 1986; Ruuska et al. 1998). The proportional relationship between gc and AC can be explained by competition between RuBP, PGA and Pi for free enzyme (Badger & Lorimer 1981; Jordan, Chollet & Ogren 1983), which linearizes the kinetics when RuBP concentration is decreasing and PGA concentration is simultaneously increasing during the post-illumination process (Laisk et al. 1987). As a result, the slope of the gc versus AC curve, termed specific carboxylation efficiency (SCE) is not a parameter specific to Rubisco only, but, rather, it characterizes the ratio of Rubisco activity to chloroplast Pi pool. Total AC characterizes the Pi pool (= Pi/2), provided that at the low CO2 and O2 concentration most of the carbon reduction intermediates were converted to RuBP and little Pi was trapped in hexosephosphates, which sometimes may happen (Eichelmann & Laisk 1994).

Concluding, despite some interpretation problems, this work opens the practical prospects of parallel measurements of CO2 uptake, O2 evolution, chlorophyll fluorescence and 820 nm transmission as tools for the analysis of the state of the photosynthetic machinery by determining its intrinsic parameters in intact leaves. Differences between the studied birch clones were minor. Such equality in most photosynthetic parameters but different sensitivity to O3 (Pääkonen, Holopainen & Karenlampi 1997) makes these clones suitable for the comparative investigation of the effects of elevated O3 and CO2 on the photosynthetic machinery, however, the wide variation of adjustable parameters in individual leaves makes detecting the effects of these atmospheric pollutants very difficult, unless the effect will exceed 20–30%.


This work was supported by grant ERB IC15 CT98 0102 from the European Commission and partially by grant 3907 from Estonian Science Foundation and by project 180517s98 from Estonian Ministry of Education

Received 7 September 2001;received inrevised form 14 February 2002;accepted for publication 26 February 2002


At CO2 photocompensation concentration the following relationships hold true.

Carboxylation rate and photorespiration rate are equal,


Carboxylation rate is expressed as


Where Γ* is the CO2 photocompensation concentration (at which Eqn A1 holds true) and rc is carboxylation resistance, reciprocal of the carboxylation conductance (efficiency) gc. ETR at Γ* is the sum of ETR supporting carboxylation and photorespiration


where JC denotes ETR supporting CO2 assimilation, i.e. carboxylation plus photorespiration.

From Eqn A3



  • (for aII and aII see text), and


On the other hand, the CO2 photocompensation point

  • And, thus,


where Oc and JF correspond to Γ*. Total mesophyll diffusion resistance rm was calculated in a conventional way, using CO2 uptake, transpiration and leaf temperature, and mesophyll diffusion resistance was found as the difference