- Email: [email protected]
Calibration matters: On the procedure of using the chlorophyll fluorescence method to estimate mesophyll conductance
Journal of Plant Physiology 220 (2018) 167–172
Contents lists available at ScienceDirect
Journal of Plant Physiology journal homepage: www.elsevier.com/locate/jplph
Short communication
Calibration matters: On the procedure of using the chlorophyll fluorescence method to estimate mesophyll conductance
T
⁎
Peter E.L. van der Putten, Xinyou Yin , Paul C. Struik Centre for Crop Systems Analysis, Department of Plant Sciences, Wageningen University & Research, P.O. Box 430, 6700 AK, Wageningen, The Netherlands
A R T I C L E I N F O
A B S T R A C T
Keywords: Chlorophyll fluorescence Gas exchange Mesophyll conductance
Estimates of mesophyll conductance (gm), when calculated from chlorophyll fluorescence, are uncertain, especially when the photosystem II (PSII) operating efficiency is measured from the traditional single saturation pulse methodology. The multiphase flash method has recently been recommended to replace the single saturation pulse method, allowing a more reliable estimation of gm. Also, many researchers still directly use the PSII operating efficiency to derive linear electron transport rate J (that is required to estimate gm), without appropriate calibration using measurements under non-photorespiratory conditions. Here we demonstrate for tomato and rice that (i) using the multiphase flash method did not yield realistic estimates of gm if no calibration was conducted; and (ii) using the single saturation pulse method still gave reasonable estimates of gm when calibration based on the non-photorespiratory measurements was properly conducted. Therefore, conducting calibration based on data under non-photorespiratory conditions was indispensable for a reliable estimation of gm, regardless whether the multiphase flash or the single saturation pulse method was used for measuring the PSII operating efficiency. Other issues related to the procedure of using the chlorophyll fluorescence method to estimate gm were discussed.
1. Introduction Mesophyll conductance (gm) for CO2 transfer from intercellular airspaces to carboxylating sites of Rubisco in chloroplasts has received growing attention in studying leaf photosynthesis of C3 plants. This parameter has mainly been estimated using either the carbon isotope discrimination method (Evans et al., 1994) or the chlorophyll fluorescence method (Harley et al., 1992), although other methods have also been suggested (see reviews of Warren 2006; Flexas et al., 2008; Pons et al., 2009). The chlorophyll fluorescence method has been widely used because the technique has become routinely available in many laboratories with the advent of portable integrated fluorometer and gas exchange systems like LI-COR. The chlorophyll fluorescence method to estimate gm relies on an equation of the model of Farquhar et al. (1980) for the electron transport limited net rate of leaf photosynthesis (A):
A=
Cc − Γ* J − Rd Cc + 2Γ* 4
(1)
where Cc is the CO2 level at carboxylating sites of Rubisco in chloroplasts, J is the rate of linear electron transport supporting the Calvin cycle and photorespiration, and Γ* is the Cc-based CO2 compensation
⁎
point in the absence of day respiration (Rd), i.e. the respiratory CO2 release in the light. Data for variable A in the equation can be measured from the gas exchange system, and variable J can be derived from the simultaneously measured photosystem II (PSII) electron transport efficiency from the integrated fluorometer (ΔF/Fm′, where Fm′ is the maximum fluorescence yield during a saturating pulse of light, and ΔF is the difference between Fm′ and the steady-state fluorescence yield Fs, Genty et al., 1989); so, combining Eq. (1) with the equation for diffusion of CO2 from intercellular airspaces (Ci) to chloroplast stroma: A = gm(Ci − Cc), one can solve gm as (Harley et al., 1992):
gm =
A Ci −
Γ* [J / 4 + 2(A + R d)] J / 4 − (A + R d)
(2)
Values of gm can be calculated using Eq. (2) for each Ci or incoming irradiance (Iinc) level, at which gas exchange and chlorophyll fluorescence data are simultaneously collected, and the average value is often considered as gm for the range of Ci or Iinc involved. However, as generally recognised, the calculated gm values from Eq. (2) are very sensitive to random measurement errors in A, Ci and ΔF/Fm′ (Harley et al., 1992). Yin and Struik (2009) suggested several methods to minimise this sensitivity, and the best one is to use the equation that is obtained from solving for A from Eq. (2):
Corresponding author. E-mail address: [email protected] (X. Yin).
https://doi.org/10.1016/j.jplph.2017.11.009 Received 10 August 2017; Received in revised form 17 November 2017; Accepted 20 November 2017 Available online 26 November 2017 0176-1617/ © 2017 Elsevier GmbH. All rights reserved.
Journal of Plant Physiology 220 (2018) 167–172
P.E.L. van der Putten et al.
turnover of the PSII acceptor pools, even when using very high single saturation pulse intensities (Earl and Ennahli 2004). Neither factor will guarantee that the ratio of Φ2 to (ΔF/Fm′) would stay constant across levels of Iinc. To improve the measuring accuracy of chlorophyll fluorescencebased PSII electron transport efficiency, Loriaux et al. (2013) reported a method, referred to as the multiphase flash method, being capable of rapidly (within 1 s) describing the irradiance dependence of Fm′ and estimating Fm′ at infinite irradiance. They showed that the multiphase flash method can generate more accurate and consistent estimates of gm than the single saturation pulse method. However, they used Eq. (4) to derive the values of J for both the single saturation pulse method and the multiphase flash method. This generates two questions. First, to what extent can the single saturation pulse method still give a reasonable estimate of gm if an appropriate calibration procedure is applied? Secondly, can the multiphase flash method also yield an erroneous estimate of gm in the absence of calibration? The objective of the present communication is to address these unknowns by comparing the estimated gm based on the multiphase flash vs single saturation pulse methods, either with or without a calibration procedure. To this end, we collected combined gas exchange and chlorophyll fluorescence data for two contrasting C3 species, i.e. tomato and rice.
⎧ A = 0.5 J /4 − R d + gm (Ci + 2Γ*) ⎨ ⎩ −
[J /4 − R d + gm (Ci + 2Γ*)]2 ⎫ − 4gm [(Ci − Γ*) J /4 − R d (Ci + 2Γ*)] ⎬ ⎭
(3)
By fitting Eq. (3) to data obtained using a range of Ci or Iinc, gm within that range of Ci or Iinc can be estimated, which is considerably less sensitive to measurement errors than the average value of gm calculated using Eq. (2) for individual Ci or Iinc (Yin and Struik 2009). Obviously, the accuracy of data on J has a strong influence on the estimation of gm, regardless whether Eq. (2) or (3) is used. Values of J are calculated routinely as (Baker, 2008):
J = βρ2 I Inc (ΔF / Fm' )
(4)
where β is absorptance by leaf photosynthetic pigments, and ρ2 is the proportion of absorbed irradiance that is partitioned to PSII. Real values of β and ρ2 are hard to measure and β is often approximated to total leaf absorptance as measured by a spectroradiometer and integrating sphere, assuming the absorptance by non-photosynthetic pigments is negligible. If not measured at all, 0.84 (or 0.85) and 0.5 have frequently been assumed as the default values of β and ρ2, respectively, for healthy leaves to estimate gm (e.g. Li et al., 2009; Adachi et al., 2013) and this assumption appears to be continuously made in the literature (He et al., 2017). However, even if these represent the real values for leaves or total leaf absorptance represents the absorptance by photosynthetic pigments, Eq. (4) may be criticised because it ignores the any occurrence of alternative electron sinks. Furthermore, it is possible that the chlorophyll fluorescence-based ΔF/Fm′ does not accurately represent the true PSII electron transport efficiency of the whole leaf (Φ2), or in other words, ξ, the ratio of Φ2 to (ΔF/Fm′), may not be equal to 1. In the presence of alternative sinks, it is the total electron flux passing PSII, J2, that is equal to βρ2IincΔF/Fm′, and J2 and the linear electron flux J differ as (Yin et al., 2009):
fpseudo ⎞ ⎛ J2 J = ⎜1 − 1 − fcyc ⎟ ⎝ ⎠
2. Materials and methods 2.1. Culturing plant material An experiment was carried out in a glasshouse at Wageningen University, using tomato (Solanum lycopersicum) cv. “MoneyMaker” and rice (Oryza sativa) cv. “IR64” in four replicates. Tomato plants were grown in a nursery bed and seedlings were transplanted after seven days in 10 L pots containing potting soil. The initial nitrogen content in the soil was 0.66 g per pot. On a weekly basis a standard tomato nutrient solution was applied. In seven applications a total of 0.85 g N, 0.39 g P2O5 and 1.60 g K2O was added per pot. The glasshouse temperature was 24 ± 3 °C during the day (for 16 h) and 18 °C during the night, the relative humidity was 40–60% and the photoperiod was kept at 14 h d−1. All measurements were carried out in the seventh week after sowing (April 2015), using distal leaflets of the compound leaves that were just fully expanded, typically leaf 8 counted from below. Rice plants were grown in small pots and seedlings were transplanted after 14 days in 7-L pots containing sandy soil. The initial nitrogen content in the soil was 0.40 g per pot. With mixing granulate fertiliser through the soil 0.50 g N, 0.50 g P2O5 and 0.49 g K2O was added per pot. The rice plants were grown under submerged conditions. The glasshouse temperature was 28 ± 2 °C during the day (for 12 h) and 23 °C during the night, the relative humidity was 40–80% and the photoperiod was kept at 12 h d−1. All measurements were carried out in the eighth week after sowing (May 2015) on leaves that were just fully expanded, typically leaf 10 counted from below. About 60% of the photosynthetically active incident radiation on the greenhouse was transmitted to the plant level. During daytime supplemental light from 600 W HPS Hortilux Schréder lamps (Monster, South Holland, The Netherlands) was switched on automatically when the photon flux of the solar incident radiation dropped below 340 μmol m−2 s−1 and was switched off when it exceeded 570 μmol m−2 s−1 in the tomato greenhouse. In the rice greenhouse these thresholds were 910 μmol m−2 s−1 and 1140 μmol m−2 s−1, respectively.
(5)
where fcyc and fpseudo are fractions of total electron flux passing PSI that is used as cyclic and pseudocyclic electron transport, respectively. Here, fpseudo refers to the fraction allocated to all other noncyclic electron sinks than the Calvin cycle and photorespiration (like nitrite reduction, the Mehler reaction, malate export, and so on). It is hard to accurately measure or estimate leaf-specific values of individual parameters β, ρ2, fcyc and fpseudo, and ξ. To collectively account for them, a common protocol is to establish a calibration curve although only β and ρ2 are most commonly pointed out explicitly (Valentini et al., 1995; Gilbert et al., 2012; Martins et al., 2013; Bellasio et al., 2016; Singh and Reddy 2016). The calibration involves to conduct simultaneous gas exchange and chlorophyll fluorescence measurements under non-photorespiratory conditions (e.g. using combined low O2 and high CO2), typically at several levels of Iinc yet within the range where A is electron transport limited. The obtained parameters of calibration, typically through linear regression, are then used to calculate J under either non-photorespiratory or photorespiratory conditions. To enhance the calibration accuracy, measurements for both nonphotorespiratory and photorespiratory conditions are advised to be made on the same leaf spots. Most calibration procedures implicitly assume that parameter ξ stays constant across levels of Iinc. Two factors are known to contribute to the difference between Φ2 and ΔF/Fm′. First, chlorophyll fluorescence measurements may not sample chloroplast populations representative of the whole leaf that gas exchange data reflect (Evans 2009). Second, estimation of Fm′ by traditional single saturation pulse methodology is prone to underestimation error, which arises because complete reduction of the primary quinone acceptor in PSII may be hindered by rapid
2.2. Measurements We used the LI-COR-6400XT open gas exchange system with an integrated fluorescence chamber head enclosing 2-cm2 areas. Fully 168
Journal of Plant Physiology 220 (2018) 167–172
P.E.L. van der Putten et al.
expanded leaves (see above) were used for all measurements. The midrib and wrinkled parts of the tomato leaves were avoided to reduce the leakage of air from the leaf chamber. In the leaf chamber, leaf-to-air vapour pressure difference was kept at 0.5–1.3 kPa. Measurements were conducted on leaves from four different plants (as four replicates). We conducted measurements of both A-Iinc curves and A-Ci curves all at a leaf temperature of 25 °C. For A-Iinc curves at 21% O2 combined with Ca of 380 μmol mol−1, eight light levels were applied from high to low levels (1500, 1000, 500, 200, 150, 100, 70 and 45 μmol m−2 s−1). For A-Ci curves at 21% O2 (measured at Iinc of 1000 μmol m−2 s−1), the following Ca levels and order were applied (380, 250, 150, 100, 65, 380, 380, 380, 600, 1000, and 1500 μmol mol−1). Part of the curves were also measured under non-photorespiratory conditions for calibration: 500, 200, 150, 100 and 70 μmol m−2 s−1 for the A-Iinc curve at 2% O2 combined with Ca of 1000 μmol mol−1, and 380, 600, 1000, and 1500 μmol mol−1 of Ca for the A-Ci curve at 2% O2 and 1000 μmol m−2 s−1 Iinc. The same leaves on four different plants were used for measurements under these conditions. For the measurements at 2% O2, gas from a cylinder containing a mixture of O2 and N2 was humidified and supplied via an overflow tube to the air inlet of the LICOR 6400XT where CO2 was blended with the gas, and the IRGA calibration was adjusted for O2 composition of the gas mixture according to the manufacturer’s instructions. All CO2 exchange data were corrected for CO2 leakage into and out of the leaf cuvette (which may occur when Ca differed from the ambient CO2 level), based on measurements on boiled leaves, and Ci were then re-calculated. For each step of light or CO2 response when A reached steady-state, steady-state fluorescence (Fs) was measured from the same leaf and then a saturating light-pulse (ca 8000 μmol m−2 s−1 for ca 0.8 s) was applied to measure maximum fluorescence (Fm' ) for the single saturation pulse method. For the multiphase flash method, the flash intensity (Q′) was increased at the onset of phase 1 from ca 1500 to ca 5000 for tomato and ca 6700 μmol m−2 s−1 for rice for a duration of 300 ms, was attenuated by 40% during phase 2 for 300 ms, and was back to the initial level for phase 3 of 300 ms. The intercept of linear regression of fluorescence yields during phase 2 against 1/Q′ gives the estimate of Fm' from the multiphase flash method (Loriaux et al., 2013). A typical measurement example for a response curve was: leaves were adapted at high light and ambient CO2 levels until steady state (for at least 30 min). Then the A-Iinc curve measurements started and after 3 min a reading was taken with the single saturation pulse method, followed by a reading with the multiphase flash method 3 min later. In the A-Ci curves the measuring interval was 2 min. The sequence of measuring single saturation pulse and multiphase flash was altered in the replicates.
fpseudo ⎞ ⎛ ξ s = βρ2 ⎜1 − 1 − fcyc ⎟ ⎝ ⎠
(6)
The estimated s is then used to calculate J under either non-photorespiratory or photorespiratory conditions: J = sIincΔF/Fm′ (Yin et al., 2009). Values of Γ* required in Eq. (3) were set to be 0.5O/Sc/o, where O is the level of O2 in mbar and Sc/o is the relative CO2/O2 specificity of Rubisco in mbar μbar−1. The kinetic parameters of Rubisco may differ among diverse C3 species (e.g. Walker et al., 2013; Orr et al., 2016). However, following the method of Yin et al. (2009) based on gas exchange measurements for the linear range of A-Ci curves at contrasting O2 levels, the value of Sc/o at 25 °C was estimated to be 3.020 mbar μbar−1 for rice (Gu et al., 2012) and 2.954 mbar μbar−1 for tomato (unpublished results). These values did not differ significantly from each other and are almost identical to the value of 3.022 mbar μbar−1 for wheat estimated using membrane inlet mass spectrometry (Cousins et al., 2010). We, therefore, assumed Sc/o at 25 °C is conserved among major C3 crop species and used the value of Cousins et al. (2010) for our analysis. After Rd, J and Γ* were known, gm was estimated by non-linear curve fitting based on Eq. (3), using the scripts of Yin and Struik (2009) with the GAUSS method in the PROC NLIN of the SAS (SAS Institute Inc., Cary, NC, USA). 3. Results There were good linear relationships between A and Iinc(ΔF/Fm′)/4 for data points of irradiance response curves within the limiting irradiance levels (Iinc ≤ 200 μmol m−2 s−1) under non-photorespiratory conditions (Fig. 1). The intercept of these linear regressions gives the estimate of −Rd. There was little difference in the estimated Rd between the multiphase flash and single saturation pulse methods, but the estimated Rd was appreciably higher in tomato than in rice (Table 1). Inclusion of extra data points yet within the electron transport limited range (i.e., data points of A-Ci curves with Ca ≥ 600, 1000, 1500 μmol mol−1) under non-photorespiratory conditions did not alter significantly the linear relationships between A and Iinc(ΔF/Fm′)/4 found for levels of Iinc ≤ 200 μmol m−2 s−1 used to estimate Rd. Inclusion of these extra points allowed the calibration factor (s) to be more reliably estimated to represent Iinc and Ca levels, with which gm was estimated later. With Rd estimated in the first step as input, the estimated s, the slope of the linear regression of A against Iinc(ΔF/Fm′)/4 across all points, was similar for tomato and rice, but was higher with the single saturation pulse method than with the multiphase flash method (Table 1). This difference between the two methods was expected, given that the traditional single saturation pulse method is known to underestimate Fm′ and the apparent PSII efficiency. We did not measure leaf absorptance β for either species. Also interphotosystem excitation partitioning is hard to measure. So, we used their default values 0.84 and 0.5 to calculate J using Eq. (4) for the case without calibration. As expected, values of J based on the single saturation pulse method were lower than those based on the multiphase flash method (Fig. 2), especially at high irradiances. In contrast, if J was calculated after calibration as J = sIincΔF/Fm′, then values of J based on the single saturation pulse method were very comparable with those based on the multiphase flash method (Fig. 2). This is because the calibration factor s with the single saturation pulse method was higher (Table 1), thereby compensating for the underestimation of the apparent PSII efficiency by this method, compared with the multiphase flash method. Because the calculated J from sIincΔF/Fm′ did not differ between the single saturation pulse and multiphase flash methods, the estimate of gm based on the data of electron transport limited range did not differ much between the two methods either (Table 1). gm was considerably higher in rice (0.71–0.73 mol m−2 s−1 bar−1) than in tomato
2.3. Data analysis Values of Rd were estimated in Microsoft Excel from simple linear regressions of A against Iinc(ΔF/Fm′)/4, based on data of A-Iinc curves within limiting irradiances (Iinc ≤ 200 μmol m−2 s−1) under non-photorespiratory conditions (Yin et al., 2009, 2011). The same linear regressions of A against Iinc(ΔF/Fm′)/4, based on data of electron transport limited ranges (i.e., data points of A-Iinc curves with Iinc ≤ 200 μmol m−2 s−1 and A-Ci curves with Ca ≥ 600 μmol mol−1) under non-photorespiratory conditions, were performed to estimate the calibration factor, by forcing the intercept of the regression at Rd as estimated in the first step. This is the calibration method proposed by Yin et al. (2009), in which a linear regression of A against Iinc(ΔF/Fm′)/4 is performed using all data of electron transport limited ranges obtained under non-photorespiratory conditions. The slope of this linear regression (s) lumps β, ρ2, fractions of alternative electron transport, and the Φ2 to ΔF/Fm′ ratio (ξ) as:
169
Journal of Plant Physiology 220 (2018) 167–172
P.E.L. van der Putten et al.
Fig. 1. Net rate of photosynthesis A plotted against Iinc(ΔF/Fm′)/4 for data points within the limiting irradiance levels (Iinc ≤ 200 μmol m−2 s−1) under non-photorespiratory conditions. The intercept of the linear regression gives the estimation of day respiration Rd for the multiphase flash (A) and single saturation pulse (B) methodology. Open circles, the dashed line, and regression parameters in non-bold are for tomato, and filled triangles, the solid line, and regression parameters in bold are for rice.
Fig. 2. The rate of linear electron transport J calculated from single saturated pulse methodology J(SP) plotted against J calculated from multiphase flash methodology J(MPF), for tomato (A) and for rice (B). The open symbols are for J calculated without calibration (0.42IincΔF/Fm′) and the closed symbols are for J calculated with calibration (sIincΔF/Fm′). The diagonal line represents the 1:1 relationship.
(0.23–0.27 mol m−2 s−1 bar−1) (Table 1). Using J calculated from Eq. (4) with the default values of β (0.84) and ρ2 (0.5), we obtained infinite gm with either method for both species (Table 1). An infinite gm seems to be unlikely; its occurrence was simply due to the default value of βρ2 (0.42) being lower than the value of calibration factor s (0.48–0.56, Table 1), even for the case when the multiphase flash method was used.
The method without calibration also did not result in a realistic estimate of gm when all data were used for curve fitting. Including all data did not alter much the estimate of gm in tomato but lowered the estimate for rice (Table 1).
Table 1 Estimated values (standard error of the estimates in brackets if available) of day respiration Rd, calibration factor s, and mesophyll conductance gm based on combined data of gas exchange and chlorophyll fluorescence with the latter assessed by the use of either the multiphase flash (MPF) method or the single saturation pulse (SP) method, for tomato and rice leaves. Tomato
Rd (μmol m−2 s−1) s gm (mol m−2 s−1 bar−1) gm (mol m−2 s−1 bar−1) se gm (mol m−2 s−1 bar−1) a b c d e
a
b
e
Calibrated J c Uncalibrated J Calibrated J c Uncalibrated J
d
d
Rice
MPF
SP
MPF
SP
2.60(0.33) 0.480(0.007) 0.265(0.035) infinite 0.270(0.015) 1.039(0.489) 0.474(0.005) 0.292(0.051)
2.11(0.31) 0.528(0.004) 0.231(0.019) infinite 0.277(0.013) infinite 0.513(0.004) 0.284(0.034)
1.69(0.65) 0.483(0.008) 0.709(0.410) infinite 0.498(0.079) infinite 0.482(0.006) 0.816(0.677)
1.72(0.44) 0.555(0.005) 0.725(0.297) infinite 0.362(0.041) infinite 0.547(0.005) 1.511(1.690)
estimated from data of the electron transport limited range only, i.e. equivalent to Iinc and Ci levels with which calibration was established. estimated from data of all Iinc and Ci levels. linear electron transport rate J was calculated from the calibration factor s, i.e. J = sIincΔF/Fm′. linear electron transport rate J was calculated from Eq. (4), i.e. J = βρ2IincΔF/Fm′ with β and ρ2 set to their default values 0.84 and 0.5, respectively. values when s and gm are fitted simultaneously to data of the electron transport limited range.
170
Journal of Plant Physiology 220 (2018) 167–172
P.E.L. van der Putten et al.
4. Discussion
parameters. The factor s can be theoretically obtained from various calibration methods that have been developed, based on the gas exchange and chlorophyll fluorescence measurements under non-photorespiratory conditions. Previous analyses in the literature mostly used the relationship either between the quantum yield for CO2 assimilation [ΦCO2 = (A + Rd)/(βIinc)] and (ΔF/Fm′) or between JA [=4(A + Rd)] and JF (the same as calculated from Eq. (4)) (Long and Bernacchi 2003). These calibration methods require Rd to be estimated beforehand, most commonly using the Laisk method (Laisk 1977). In our study we only used a relatively new calibration method described by Yin et al. (2009). The advantage of this calibration method is that Rd, a required input parameter in Eq. (1), can be estimated as the intercept of the same linear regression equation, i.e., A against Iinc(ΔF/Fm′)/4, using data collected under limiting light levels of non-photorespiratory conditions. The estimated Rd using this method is virtually insensitive to measurement errors and similar to values of Rd estimated by the more commonly used Laisk method (see Yin et al., 2011). However, not only does the Laisk method rely on a different set of gas exchange measurements (A-Ci curves of the low Ci range at two or more levels of Iinc), but also is its reliability now questioned within the framework of multiple mesophyll resistance components (Tholen et al., 2012; Walker and Ort, 2015). Within the framework of the multiple mesophyll resistance, any estimated gm based on Eq. (2) or (3) is an apparent value (Yin and Struik 2017; Tcherkez et al., 2017); however, this issue is beyond the scope of our paper. Selecting data to be included for establishing calibration factors is important to have a reliable estimation of gm. Two criteria were used in our analysis: (i) data should be within the electron transport limited range, and (ii) data should be obtained under a non-photorespiratory condition. Criterion (i) suggests that data points from very high irradiances should be excluded because A under high irradiances is most likely Rubisco limited. Including data points where A is Rubisco limited would lead to an underestimation of calibration factor s. Common procedure to obtain a non-photorespiratory condition is to use 2% O2 gas mixture. However, 2% O2 when combined with the ambient CO2 level does not create an absolute non-photorespiratory condition. Some researchers even used data points of entire A-Ci curves at 2% O2 to conduct the calibration (e.g. Flexas et al., 2007; Tomeo and Rosenthal 2017). Not only are points of the lower part of A-Ci curves Rubisco limited, but also photorespiration increases with decreasing Ci. As a result, the calibration factor obtained from entire A-Ci curves not only contains elements for β, ρ2, fractions of alternative electron transport and ξ, but is also confounded by the occurrence of photorespiration. Such confounding procedure should be avoided because of its carryover effect on estimating gm using Eqs. (1) and (3). If one’s experiment design does not permit to exclude photorespiration, it is necessary to use a one-step procedure, in which the calibration factor is not estimated beforehand, but is considered as an additional parameter to be estimated simultaneously with gm by fitting to pooled data points having varying amounts of photorespiration. Such one-step procedure for our data virtually gave nearly identical estimates of the calibration factor s for both single saturation pulse and multiphase flash measurements (Table 1). However, the resulting gm had higher standard errors, especially when gm was high. High estimation uncertainty when gm is high is a common phenomenon (Yin and Struik 2009), because in such a case, gm constrains little leaf photosynthesis, thus not allowing gm being fitted reliably. Our results in Table 1 also show that the estimated gm depended on which part of the data along response curves was used in curve fitting. We previously (Yin and Struik 2009) suggested to use the data of only electron transport limited range (i.e. lower part of A-Iinc curves and higher part of A-Ci curves) because Eq. (3), a part of the model of Farquhar et al. (1980), only applies to this range. When the full data of response curves across all Iinc and Ci levels were used, the estimated gm slightly increased in tomato and decreased in rice (Table 1). Two
Estimates of mesophyll conductance gm are very sensitive to measurement errors (Yin and Struik, 2009). Some errors are random whereas others are more structural due to the insufficient resolution of the measuring techniques. For example, Loriaux et al. (2013) reported that if gm is estimated from combined gas exchange and chlorophyll fluorescence data it is mostly overestimated by the traditional single saturation pulse fluorescence method, and this problem is overcome by employing their new multiphase flash method. Values of J were calculated as Eq. (4) with measured leaf absorptance in their analysis, assuming no absorptance by non-photosynthetic pigments. In our analysis we did not measure leaf absorptance but used a default value of 0.84 combined with an assumed equal excitation partitioning between the two photosystems (Baker, 2008). This resulted in the estimated gm being infinite regardless whether the multiphase flash method or the single saturation pulse method was used for deriving J (Table 1). However, the over-estimation of gm might become less severe if leaf absorptance was measured. Nevertheless, our analysis demonstrated that compared to using the multiphase flash method, conducting calibration based on data under non-photorespiratory conditions had more impact on estimating gm (Table 1). Such a calibration not only accounts for the uncertainty of leaf photosynthetic absorptance but also accounts for that of several other hard-to-measure parameters that are relevant to converting apparent PSII efficiency ΔF/Fm′ into linear electron transport rate J, regardless the single saturation pulse or multiphase flash method to obtain ΔF/Fm′. However, we do not want to suggest that using the multiphase flash method is not important. The calibration implicitly assumes that all the hard-to-measure parameters (β, ρ2, alternative electron fractions, and ξ) are constant across Iinc and Ca levels involved. There is no guarantee that this condition is met. In particular, the underestimation of Fm′ by the single saturation pulse method is more obvious at high than low irradiances (Loriaux et al., 2013). This is shown in Fig. 2, where even with calibrated J, the single saturation pulse-based J deviated more from the multiphase flash-based J with increasing irradiance. Also, the apparent PSII efficiency is calculated as 1-Fs/Fm′ (Genty et al., 1989), meaning that the underestimation of the PSII efficiency by the single saturation pulse method does not go proportionally with its underestimation of Fm′. Thus, mathematically, the calibration does not solve, but only alleviates, the problem of the single saturation pulse method. Values of the calibration factor s were lower with the multiphase flash method than with the single saturation pulse method (Table 1), largely because an underlying factor ξ (see Eq. (6)) is lower when using the multiphase flash method. The s values with the multiphase flash method were ca 0.48 (Table 1), still higher than the default 0.42. It is not possible to point out which factor of β, ρ2 and ξ in Eq. (6) actually contributed to the higher values of s. However, Eq. (6) allows to answer ‘what-if’ questions. The high s values mean that alternative electron transport fractions (fcyc and fpseudo in Eq. (6)) are probably negligible. The theoretical equation to calculate ρ2 is (Yin et al., 2006):
ρ 2=
1 − fcyc (1 − fcyc ) + Φ2/ Φ1
(7)
where Φ2/Φ1 is the ratio of PSII to PSI electron transport efficiency. Since this ratio is generally believed to be relatively conservative as ca 0.85, Eq. (7) would give ca 0.54 for ρ2 if fcyc = 0. Assuming that the ΔF/ Fm′ value from the multiphase flash method represents the true Φ2 (i.e., ξ = 1), one would calculate from Eq. (6) that leaf photosynthetic absorptance β is ca 0.89 for both tomato and rice leaves in our study, which appears to be a reasonable value. Likewise, one can infer what the likely value of ξ or alternative electron fraction or Φ2/Φ1 would be if β is assumed to be known. However, given the plethora of uncertain parameters underlying the factor s, we rather refrain from speculating about the range of possible variation for each of these uncertain 171
Journal of Plant Physiology 220 (2018) 167–172
P.E.L. van der Putten et al.
mechanisms are behind this change. First, when data used involve the range where A is limited by mechanisms other than electron transport, additional pseudocyclic electron transport (to O2, most likely) would occur in case that non-photochemical quenching does not increase accordingly (Ort and Baker, 2002). This would result in an underestimation of gm as this additional electron transport flux is wrongly attributed to the limitation by mesophyll resistance (Yin et al., 2009). Secondly, gm has often been reported to increase with increasing Iinc (e.g. Yin et al., 2009; Douthe et al., 2012) and with decreasing Ci (e.g. Flexas et al., 2007). The first mechanism seems to explain the decreased estimate of gm in rice while the second mechanism may explain the increased gm in tomato. Because crops were shown to respond differently to the expansion of the data range, we re-state here the importance of using data of only the electron limited range for estimating gm when investigating any difference in gm among various treatments or species or cultivars.
indica × japonica rice inbred lines. Photosynth. Res. 134, 27–38. Laisk, A.K., 1977. Kinetics of Photosynthesis and Photorespiration in C3 Plant. Nauka Moscow (in Russian). Li, Y., Gao, Y.X., Xu, X.M., Shen, Q.R., Guo, S.W., 2009. Light-saturated photosynthetic rate in high-nitrogen rice (Oryza sativa L.) leaves is related to chloroplastic CO2 concentration. J. Exp. Bot. 60, 351–2360. Long, S.P., Bernacchi, C.J., 2003. Gas exchange measurements, what can they tell us about the underlying limitations to photosynthesis? Procedures and sources of error. J. Exp. Bot. 54, 2393–2401. Loriaux, S.D., Avenson, T.J., Wells, J.M., McDermitt, D.K., Eckles, R.D., Riensche, B., Genty, B., 2013. Closing in on maximum yield of chlorophyll fluorescence using a single multiphase flash of sub-saturating intensity. Plant Cell Environ. 36, 1755–1770. Martins, S.C.V., Galmes, J., Molins, A., DaMatta, F.M., 2013. Improving the estimation of mesophyll conductance to CO2: on the role of electron transport rate correction and respiration. J. Exp. Bot. 64, 3285–3298. Orr, D.J., Alcantara, A., Kapralov, M.V., Andralojc, P.J., Carmo-Silva, E., Parry, M.A.J., 2016. Surveying Rubisco diversity and temperature response to improve crop photosynthetic efficiency. Plant Physiol. 172, 707–717. Ort, D.R., Baker, N.R., 2002. A photoprotective role for O2 as an alternative electron sink in photosynthesis? Curr. Opin. Plant Biol. 5, 193–198. Pons, T.L., Flexas, J., von Caemmerer, S., Evans, J.R., Genty, B., Ribas-Carbo, M., Brugnoli, E., 2009. Estimating mesophyll conductance to CO2: methodology, potential errors, and recommendations. J. Exp. Bot. 60, 2217–2234. Singh, S.K., Reddy, V.R., 2016. Methods of mesophyll conductance estimations: its impact on key biochemical parameters and photosynthetic limitations in phosphorusstressed soybean across CO2. Physiol. Plant. 157, 234–254. Tcherkez, G., Gauthier, P., Buckley, T.N., et al., 2017. Leaf day respiration: low CO2 flux but high significance for metabolism and carbon balance. New Phytol. 216, 986–1001. Tholen, D., Ethier, G., Genty, B., Pepin, S., Zhu, X.-G., 2012. Variable mesophyll conductance revisited: theoretical background and experimental implications. Plant Cell Environ. 35, 2087–2103. Tomeo, N.J., Rosenthal, D.M., 2017. Variable mesophyll conductance among soybean cultivars sets a tradeoff between photosynthesis and water-use-efficiency. Plant Physiol. 174, 241–257. Valentini, R., Epron, D., Angelis, P.D.E., Matteucci, G., Dreyer, E., 1995. In situ estimation of net CO2 assimilation, photosynthetic electron flow and photorespiration in Turkey oak (Q. cerris L.) leaves: diurnal cycles under different levels of water supply. Plant Cell Environ. 18, 631–640. Walker, B.J., Ort, D.R., 2015. Improved method for measuring the apparent CO2 photocompensation point resolves the impact of multiple internal conductance to CO2 to net gas exchange. Plant Cell Environ. 38, 2462–2474. Walker, B.J., Ariza, L.S., Kaines, S., Badger, M.R., Cousins, A.B., 2013. Temperature response of in vivo Rubisco kinetics and mesophyll conductance in Arabidopsis thaliana: comparisons to Nicotiana tabacum. Plant Cell Environ. 36, 2108–2119. Warren, C.R., 2006. Estimating the internal conductance to CO2 movement. Funct. Plant Biol. 33, 431–442. Yin, X., Struik, P.C., 2009. Theoretical reconsiderations when estimating the mesophyll conductance to CO2 diffusion in leaves of C3 plants by analysis of combined gas exchange and chlorophyll fluorescence measurements. Plant Cell Environ. 32, 1513–1524 with corrigendum in PC&E 33: 1595. Yin, X., Struik, P.C., 2017. Simple generalisation of a mesophyll resistance model for various intracellular arrangements of chloroplasts and mitochondria in C3 leaves. Photosynth. Res. 132, 211–220. Yin, X., Harbinson, J., Struik, P.C., 2006. Mathematical review of literature to assess alternative electron transports and interphotosystem excitation partitioning of steady-state C3 photosynthesis under limiting light. Plant Cell Environ. 29, 1771–1782 (corrigendum in PCE 2006. 29: 2252). Yin, X., Struik, P.C., Romero, P., Harbinson, J., Evers, J.B., van der Putten, P.E.L., Vos, J., 2009. Using combined measurements of gas exchange and chlorophyll fluorescence to estimate parameters of a biochemical C3 photosynthesis model: a critical appraisal and a new integrated approach applied to leaves in a wheat (Triticum aestivum) canopy. Plant Cell Environ. 32, 448–464. Yin, X., Sun, Z., Struik, P.C., Gu, J., 2011. Evaluating a new method to estimate the rate of leaf respiration in the light by analysis of combined gas exchange and chlorophyll fluorescence measurements. J. Exp. Bot. 62, 3489–3499.
References Adachi, S., Nakae, T., Uchida, M., Soda, K., Takai, T., Oi, T., Yamamoto, T., Ookawa, T., Miyake, H., Yano, M., Hirasawa, T., 2013. The mesophyll anatomy enhancing CO2 diffusion is a key trait for improving rice photosynthesis. J. Exp. Bot. 64, 1061–1072. Baker, N.R., 2008. Chlorophyll fluorescence: a probe of photosynthesis in vivo. Annu. Rev. Plant Biol. 59, 89–113. Bellasio, C., Beerling, D.J., Griffiths, H., 2016. An Excel tool for deriving key photosynthetic parameters from combined gas exchange and chlorophyll fluorescence: theory and practice. Plant Cell Environ. 39, 1180–1197. Cousins, A.B., Ghannoum, O., von Caemmerer, S., Badger, M.R., 2010. Simultaneous determination of Rubisco carboxylase and oxygenase kinetic parameters in Triticum aestivum and Zea mays using membrane inlet mass spectrometry. Plant Cell Environ. 33, 444–452. Douthe, C., Dreyer, E., Brendel, O., Warren, C.R., 2012. Is mesophyll conductance to CO2 in leaves of three Eucalyptus species sensitive to short-term changes of irradiance under ambient as well as low O2? Funct. Plant Biol. 39, 435–448. Earl, H.J., Ennahli, S., 2004. Estimating photosynthetic electron transport via chlorophyll fluorometry without Photosystem II light saturation. Photosynth. Res. 82, 177–186. Evans, J.R., von Caemmerer, S., Setchell, B.A., Hudson, G.S., 1994. The relationship between CO2 transfer conductance and leaf anatomy in transgenic tobacco with a reduced content of Rubisco. Aust. J. Plant Physiol. 21, 475–495. Evans, J.R., 2009. Potential errors in electron transport rates calculated from chlorophyll fluorescence as revealed by a multilayer leaf model. Plant Cell Physiol. 50, 698–706. Farquhar, G.D., von Caemmerer, S., Berry, J.A., 1980. A biochemical model of photosynthetic CO2 assimilation in leaves of C3 species. Planta 149, 78–90. Flexas, J., Diaz-Espejo, A., Galmes, J., Kaldenhoff, R., Medrano, H., Ribas-Carbó, M., 2007. Rapid variation of mesophyll conductance in response to changes in CO2 concentration around leaves. Plant Cell Environ. 30, 1284–1298. Flexas, J., Ribas-Carbó, M., Diaz-Espejo, A., Galmes, J., Medrano, H., 2008. Mesophyll conductance to CO2: current knowledge and future prospects. Plant Cell Environ. 31, 602–621. Genty, B., Briantais, J.-M., Baker, N., 1989. The relationship between the quantum yield of photosynthetic electron transport and quenching of chlorophyll fluorescence. Biochim. Biophys. Acta 990, 87–92. Gilbert, M.E., Pou, A., Zwieniecki, M.A., Holbrook, N.M., 2012. On measuring the response of mesophyll conductance to carbon dioxide with the variable J method. J. Exp. Bot. 63, 413–425. Gu, J., Yin, X., Stomph, T.J., Wang, H., Struik, P.C., 2012. Physiological basis of genetic variation in leaf photosynthesis among rice (Oryza sativa L.) introgression lines under drought and well-watered conditions. J. Exp. Bot. 63, 5137–5153. Harley, P.C., Loreto, F., Di Marco, G., Sharkey, T.D., 1992. Theoretical considerations when estimating the mesophyll conductance to CO2 flux by analysis of the response of photosynthesis to CO2. Plant Physiol. 98, 1429–1436. He, W., Adachi, S., Sage, R.F., Ookawa, T., Hirasawa, T., 2017. Leaf photosynthetic rate and mesophyll cell anatomy changes during ontogenesis in backcrossed
172
Contents lists available at ScienceDirect
Journal of Plant Physiology journal homepage: www.elsevier.com/locate/jplph
Short communication
Calibration matters: On the procedure of using the chlorophyll fluorescence method to estimate mesophyll conductance
T
⁎
Peter E.L. van der Putten, Xinyou Yin , Paul C. Struik Centre for Crop Systems Analysis, Department of Plant Sciences, Wageningen University & Research, P.O. Box 430, 6700 AK, Wageningen, The Netherlands
A R T I C L E I N F O
A B S T R A C T
Keywords: Chlorophyll fluorescence Gas exchange Mesophyll conductance
Estimates of mesophyll conductance (gm), when calculated from chlorophyll fluorescence, are uncertain, especially when the photosystem II (PSII) operating efficiency is measured from the traditional single saturation pulse methodology. The multiphase flash method has recently been recommended to replace the single saturation pulse method, allowing a more reliable estimation of gm. Also, many researchers still directly use the PSII operating efficiency to derive linear electron transport rate J (that is required to estimate gm), without appropriate calibration using measurements under non-photorespiratory conditions. Here we demonstrate for tomato and rice that (i) using the multiphase flash method did not yield realistic estimates of gm if no calibration was conducted; and (ii) using the single saturation pulse method still gave reasonable estimates of gm when calibration based on the non-photorespiratory measurements was properly conducted. Therefore, conducting calibration based on data under non-photorespiratory conditions was indispensable for a reliable estimation of gm, regardless whether the multiphase flash or the single saturation pulse method was used for measuring the PSII operating efficiency. Other issues related to the procedure of using the chlorophyll fluorescence method to estimate gm were discussed.
1. Introduction Mesophyll conductance (gm) for CO2 transfer from intercellular airspaces to carboxylating sites of Rubisco in chloroplasts has received growing attention in studying leaf photosynthesis of C3 plants. This parameter has mainly been estimated using either the carbon isotope discrimination method (Evans et al., 1994) or the chlorophyll fluorescence method (Harley et al., 1992), although other methods have also been suggested (see reviews of Warren 2006; Flexas et al., 2008; Pons et al., 2009). The chlorophyll fluorescence method has been widely used because the technique has become routinely available in many laboratories with the advent of portable integrated fluorometer and gas exchange systems like LI-COR. The chlorophyll fluorescence method to estimate gm relies on an equation of the model of Farquhar et al. (1980) for the electron transport limited net rate of leaf photosynthesis (A):
A=
Cc − Γ* J − Rd Cc + 2Γ* 4
(1)
where Cc is the CO2 level at carboxylating sites of Rubisco in chloroplasts, J is the rate of linear electron transport supporting the Calvin cycle and photorespiration, and Γ* is the Cc-based CO2 compensation
⁎
point in the absence of day respiration (Rd), i.e. the respiratory CO2 release in the light. Data for variable A in the equation can be measured from the gas exchange system, and variable J can be derived from the simultaneously measured photosystem II (PSII) electron transport efficiency from the integrated fluorometer (ΔF/Fm′, where Fm′ is the maximum fluorescence yield during a saturating pulse of light, and ΔF is the difference between Fm′ and the steady-state fluorescence yield Fs, Genty et al., 1989); so, combining Eq. (1) with the equation for diffusion of CO2 from intercellular airspaces (Ci) to chloroplast stroma: A = gm(Ci − Cc), one can solve gm as (Harley et al., 1992):
gm =
A Ci −
Γ* [J / 4 + 2(A + R d)] J / 4 − (A + R d)
(2)
Values of gm can be calculated using Eq. (2) for each Ci or incoming irradiance (Iinc) level, at which gas exchange and chlorophyll fluorescence data are simultaneously collected, and the average value is often considered as gm for the range of Ci or Iinc involved. However, as generally recognised, the calculated gm values from Eq. (2) are very sensitive to random measurement errors in A, Ci and ΔF/Fm′ (Harley et al., 1992). Yin and Struik (2009) suggested several methods to minimise this sensitivity, and the best one is to use the equation that is obtained from solving for A from Eq. (2):
Corresponding author. E-mail address: [email protected] (X. Yin).
https://doi.org/10.1016/j.jplph.2017.11.009 Received 10 August 2017; Received in revised form 17 November 2017; Accepted 20 November 2017 Available online 26 November 2017 0176-1617/ © 2017 Elsevier GmbH. All rights reserved.
Journal of Plant Physiology 220 (2018) 167–172
P.E.L. van der Putten et al.
turnover of the PSII acceptor pools, even when using very high single saturation pulse intensities (Earl and Ennahli 2004). Neither factor will guarantee that the ratio of Φ2 to (ΔF/Fm′) would stay constant across levels of Iinc. To improve the measuring accuracy of chlorophyll fluorescencebased PSII electron transport efficiency, Loriaux et al. (2013) reported a method, referred to as the multiphase flash method, being capable of rapidly (within 1 s) describing the irradiance dependence of Fm′ and estimating Fm′ at infinite irradiance. They showed that the multiphase flash method can generate more accurate and consistent estimates of gm than the single saturation pulse method. However, they used Eq. (4) to derive the values of J for both the single saturation pulse method and the multiphase flash method. This generates two questions. First, to what extent can the single saturation pulse method still give a reasonable estimate of gm if an appropriate calibration procedure is applied? Secondly, can the multiphase flash method also yield an erroneous estimate of gm in the absence of calibration? The objective of the present communication is to address these unknowns by comparing the estimated gm based on the multiphase flash vs single saturation pulse methods, either with or without a calibration procedure. To this end, we collected combined gas exchange and chlorophyll fluorescence data for two contrasting C3 species, i.e. tomato and rice.
⎧ A = 0.5 J /4 − R d + gm (Ci + 2Γ*) ⎨ ⎩ −
[J /4 − R d + gm (Ci + 2Γ*)]2 ⎫ − 4gm [(Ci − Γ*) J /4 − R d (Ci + 2Γ*)] ⎬ ⎭
(3)
By fitting Eq. (3) to data obtained using a range of Ci or Iinc, gm within that range of Ci or Iinc can be estimated, which is considerably less sensitive to measurement errors than the average value of gm calculated using Eq. (2) for individual Ci or Iinc (Yin and Struik 2009). Obviously, the accuracy of data on J has a strong influence on the estimation of gm, regardless whether Eq. (2) or (3) is used. Values of J are calculated routinely as (Baker, 2008):
J = βρ2 I Inc (ΔF / Fm' )
(4)
where β is absorptance by leaf photosynthetic pigments, and ρ2 is the proportion of absorbed irradiance that is partitioned to PSII. Real values of β and ρ2 are hard to measure and β is often approximated to total leaf absorptance as measured by a spectroradiometer and integrating sphere, assuming the absorptance by non-photosynthetic pigments is negligible. If not measured at all, 0.84 (or 0.85) and 0.5 have frequently been assumed as the default values of β and ρ2, respectively, for healthy leaves to estimate gm (e.g. Li et al., 2009; Adachi et al., 2013) and this assumption appears to be continuously made in the literature (He et al., 2017). However, even if these represent the real values for leaves or total leaf absorptance represents the absorptance by photosynthetic pigments, Eq. (4) may be criticised because it ignores the any occurrence of alternative electron sinks. Furthermore, it is possible that the chlorophyll fluorescence-based ΔF/Fm′ does not accurately represent the true PSII electron transport efficiency of the whole leaf (Φ2), or in other words, ξ, the ratio of Φ2 to (ΔF/Fm′), may not be equal to 1. In the presence of alternative sinks, it is the total electron flux passing PSII, J2, that is equal to βρ2IincΔF/Fm′, and J2 and the linear electron flux J differ as (Yin et al., 2009):
fpseudo ⎞ ⎛ J2 J = ⎜1 − 1 − fcyc ⎟ ⎝ ⎠
2. Materials and methods 2.1. Culturing plant material An experiment was carried out in a glasshouse at Wageningen University, using tomato (Solanum lycopersicum) cv. “MoneyMaker” and rice (Oryza sativa) cv. “IR64” in four replicates. Tomato plants were grown in a nursery bed and seedlings were transplanted after seven days in 10 L pots containing potting soil. The initial nitrogen content in the soil was 0.66 g per pot. On a weekly basis a standard tomato nutrient solution was applied. In seven applications a total of 0.85 g N, 0.39 g P2O5 and 1.60 g K2O was added per pot. The glasshouse temperature was 24 ± 3 °C during the day (for 16 h) and 18 °C during the night, the relative humidity was 40–60% and the photoperiod was kept at 14 h d−1. All measurements were carried out in the seventh week after sowing (April 2015), using distal leaflets of the compound leaves that were just fully expanded, typically leaf 8 counted from below. Rice plants were grown in small pots and seedlings were transplanted after 14 days in 7-L pots containing sandy soil. The initial nitrogen content in the soil was 0.40 g per pot. With mixing granulate fertiliser through the soil 0.50 g N, 0.50 g P2O5 and 0.49 g K2O was added per pot. The rice plants were grown under submerged conditions. The glasshouse temperature was 28 ± 2 °C during the day (for 12 h) and 23 °C during the night, the relative humidity was 40–80% and the photoperiod was kept at 12 h d−1. All measurements were carried out in the eighth week after sowing (May 2015) on leaves that were just fully expanded, typically leaf 10 counted from below. About 60% of the photosynthetically active incident radiation on the greenhouse was transmitted to the plant level. During daytime supplemental light from 600 W HPS Hortilux Schréder lamps (Monster, South Holland, The Netherlands) was switched on automatically when the photon flux of the solar incident radiation dropped below 340 μmol m−2 s−1 and was switched off when it exceeded 570 μmol m−2 s−1 in the tomato greenhouse. In the rice greenhouse these thresholds were 910 μmol m−2 s−1 and 1140 μmol m−2 s−1, respectively.
(5)
where fcyc and fpseudo are fractions of total electron flux passing PSI that is used as cyclic and pseudocyclic electron transport, respectively. Here, fpseudo refers to the fraction allocated to all other noncyclic electron sinks than the Calvin cycle and photorespiration (like nitrite reduction, the Mehler reaction, malate export, and so on). It is hard to accurately measure or estimate leaf-specific values of individual parameters β, ρ2, fcyc and fpseudo, and ξ. To collectively account for them, a common protocol is to establish a calibration curve although only β and ρ2 are most commonly pointed out explicitly (Valentini et al., 1995; Gilbert et al., 2012; Martins et al., 2013; Bellasio et al., 2016; Singh and Reddy 2016). The calibration involves to conduct simultaneous gas exchange and chlorophyll fluorescence measurements under non-photorespiratory conditions (e.g. using combined low O2 and high CO2), typically at several levels of Iinc yet within the range where A is electron transport limited. The obtained parameters of calibration, typically through linear regression, are then used to calculate J under either non-photorespiratory or photorespiratory conditions. To enhance the calibration accuracy, measurements for both nonphotorespiratory and photorespiratory conditions are advised to be made on the same leaf spots. Most calibration procedures implicitly assume that parameter ξ stays constant across levels of Iinc. Two factors are known to contribute to the difference between Φ2 and ΔF/Fm′. First, chlorophyll fluorescence measurements may not sample chloroplast populations representative of the whole leaf that gas exchange data reflect (Evans 2009). Second, estimation of Fm′ by traditional single saturation pulse methodology is prone to underestimation error, which arises because complete reduction of the primary quinone acceptor in PSII may be hindered by rapid
2.2. Measurements We used the LI-COR-6400XT open gas exchange system with an integrated fluorescence chamber head enclosing 2-cm2 areas. Fully 168
Journal of Plant Physiology 220 (2018) 167–172
P.E.L. van der Putten et al.
expanded leaves (see above) were used for all measurements. The midrib and wrinkled parts of the tomato leaves were avoided to reduce the leakage of air from the leaf chamber. In the leaf chamber, leaf-to-air vapour pressure difference was kept at 0.5–1.3 kPa. Measurements were conducted on leaves from four different plants (as four replicates). We conducted measurements of both A-Iinc curves and A-Ci curves all at a leaf temperature of 25 °C. For A-Iinc curves at 21% O2 combined with Ca of 380 μmol mol−1, eight light levels were applied from high to low levels (1500, 1000, 500, 200, 150, 100, 70 and 45 μmol m−2 s−1). For A-Ci curves at 21% O2 (measured at Iinc of 1000 μmol m−2 s−1), the following Ca levels and order were applied (380, 250, 150, 100, 65, 380, 380, 380, 600, 1000, and 1500 μmol mol−1). Part of the curves were also measured under non-photorespiratory conditions for calibration: 500, 200, 150, 100 and 70 μmol m−2 s−1 for the A-Iinc curve at 2% O2 combined with Ca of 1000 μmol mol−1, and 380, 600, 1000, and 1500 μmol mol−1 of Ca for the A-Ci curve at 2% O2 and 1000 μmol m−2 s−1 Iinc. The same leaves on four different plants were used for measurements under these conditions. For the measurements at 2% O2, gas from a cylinder containing a mixture of O2 and N2 was humidified and supplied via an overflow tube to the air inlet of the LICOR 6400XT where CO2 was blended with the gas, and the IRGA calibration was adjusted for O2 composition of the gas mixture according to the manufacturer’s instructions. All CO2 exchange data were corrected for CO2 leakage into and out of the leaf cuvette (which may occur when Ca differed from the ambient CO2 level), based on measurements on boiled leaves, and Ci were then re-calculated. For each step of light or CO2 response when A reached steady-state, steady-state fluorescence (Fs) was measured from the same leaf and then a saturating light-pulse (ca 8000 μmol m−2 s−1 for ca 0.8 s) was applied to measure maximum fluorescence (Fm' ) for the single saturation pulse method. For the multiphase flash method, the flash intensity (Q′) was increased at the onset of phase 1 from ca 1500 to ca 5000 for tomato and ca 6700 μmol m−2 s−1 for rice for a duration of 300 ms, was attenuated by 40% during phase 2 for 300 ms, and was back to the initial level for phase 3 of 300 ms. The intercept of linear regression of fluorescence yields during phase 2 against 1/Q′ gives the estimate of Fm' from the multiphase flash method (Loriaux et al., 2013). A typical measurement example for a response curve was: leaves were adapted at high light and ambient CO2 levels until steady state (for at least 30 min). Then the A-Iinc curve measurements started and after 3 min a reading was taken with the single saturation pulse method, followed by a reading with the multiphase flash method 3 min later. In the A-Ci curves the measuring interval was 2 min. The sequence of measuring single saturation pulse and multiphase flash was altered in the replicates.
fpseudo ⎞ ⎛ ξ s = βρ2 ⎜1 − 1 − fcyc ⎟ ⎝ ⎠
(6)
The estimated s is then used to calculate J under either non-photorespiratory or photorespiratory conditions: J = sIincΔF/Fm′ (Yin et al., 2009). Values of Γ* required in Eq. (3) were set to be 0.5O/Sc/o, where O is the level of O2 in mbar and Sc/o is the relative CO2/O2 specificity of Rubisco in mbar μbar−1. The kinetic parameters of Rubisco may differ among diverse C3 species (e.g. Walker et al., 2013; Orr et al., 2016). However, following the method of Yin et al. (2009) based on gas exchange measurements for the linear range of A-Ci curves at contrasting O2 levels, the value of Sc/o at 25 °C was estimated to be 3.020 mbar μbar−1 for rice (Gu et al., 2012) and 2.954 mbar μbar−1 for tomato (unpublished results). These values did not differ significantly from each other and are almost identical to the value of 3.022 mbar μbar−1 for wheat estimated using membrane inlet mass spectrometry (Cousins et al., 2010). We, therefore, assumed Sc/o at 25 °C is conserved among major C3 crop species and used the value of Cousins et al. (2010) for our analysis. After Rd, J and Γ* were known, gm was estimated by non-linear curve fitting based on Eq. (3), using the scripts of Yin and Struik (2009) with the GAUSS method in the PROC NLIN of the SAS (SAS Institute Inc., Cary, NC, USA). 3. Results There were good linear relationships between A and Iinc(ΔF/Fm′)/4 for data points of irradiance response curves within the limiting irradiance levels (Iinc ≤ 200 μmol m−2 s−1) under non-photorespiratory conditions (Fig. 1). The intercept of these linear regressions gives the estimate of −Rd. There was little difference in the estimated Rd between the multiphase flash and single saturation pulse methods, but the estimated Rd was appreciably higher in tomato than in rice (Table 1). Inclusion of extra data points yet within the electron transport limited range (i.e., data points of A-Ci curves with Ca ≥ 600, 1000, 1500 μmol mol−1) under non-photorespiratory conditions did not alter significantly the linear relationships between A and Iinc(ΔF/Fm′)/4 found for levels of Iinc ≤ 200 μmol m−2 s−1 used to estimate Rd. Inclusion of these extra points allowed the calibration factor (s) to be more reliably estimated to represent Iinc and Ca levels, with which gm was estimated later. With Rd estimated in the first step as input, the estimated s, the slope of the linear regression of A against Iinc(ΔF/Fm′)/4 across all points, was similar for tomato and rice, but was higher with the single saturation pulse method than with the multiphase flash method (Table 1). This difference between the two methods was expected, given that the traditional single saturation pulse method is known to underestimate Fm′ and the apparent PSII efficiency. We did not measure leaf absorptance β for either species. Also interphotosystem excitation partitioning is hard to measure. So, we used their default values 0.84 and 0.5 to calculate J using Eq. (4) for the case without calibration. As expected, values of J based on the single saturation pulse method were lower than those based on the multiphase flash method (Fig. 2), especially at high irradiances. In contrast, if J was calculated after calibration as J = sIincΔF/Fm′, then values of J based on the single saturation pulse method were very comparable with those based on the multiphase flash method (Fig. 2). This is because the calibration factor s with the single saturation pulse method was higher (Table 1), thereby compensating for the underestimation of the apparent PSII efficiency by this method, compared with the multiphase flash method. Because the calculated J from sIincΔF/Fm′ did not differ between the single saturation pulse and multiphase flash methods, the estimate of gm based on the data of electron transport limited range did not differ much between the two methods either (Table 1). gm was considerably higher in rice (0.71–0.73 mol m−2 s−1 bar−1) than in tomato
2.3. Data analysis Values of Rd were estimated in Microsoft Excel from simple linear regressions of A against Iinc(ΔF/Fm′)/4, based on data of A-Iinc curves within limiting irradiances (Iinc ≤ 200 μmol m−2 s−1) under non-photorespiratory conditions (Yin et al., 2009, 2011). The same linear regressions of A against Iinc(ΔF/Fm′)/4, based on data of electron transport limited ranges (i.e., data points of A-Iinc curves with Iinc ≤ 200 μmol m−2 s−1 and A-Ci curves with Ca ≥ 600 μmol mol−1) under non-photorespiratory conditions, were performed to estimate the calibration factor, by forcing the intercept of the regression at Rd as estimated in the first step. This is the calibration method proposed by Yin et al. (2009), in which a linear regression of A against Iinc(ΔF/Fm′)/4 is performed using all data of electron transport limited ranges obtained under non-photorespiratory conditions. The slope of this linear regression (s) lumps β, ρ2, fractions of alternative electron transport, and the Φ2 to ΔF/Fm′ ratio (ξ) as:
169
Journal of Plant Physiology 220 (2018) 167–172
P.E.L. van der Putten et al.
Fig. 1. Net rate of photosynthesis A plotted against Iinc(ΔF/Fm′)/4 for data points within the limiting irradiance levels (Iinc ≤ 200 μmol m−2 s−1) under non-photorespiratory conditions. The intercept of the linear regression gives the estimation of day respiration Rd for the multiphase flash (A) and single saturation pulse (B) methodology. Open circles, the dashed line, and regression parameters in non-bold are for tomato, and filled triangles, the solid line, and regression parameters in bold are for rice.
Fig. 2. The rate of linear electron transport J calculated from single saturated pulse methodology J(SP) plotted against J calculated from multiphase flash methodology J(MPF), for tomato (A) and for rice (B). The open symbols are for J calculated without calibration (0.42IincΔF/Fm′) and the closed symbols are for J calculated with calibration (sIincΔF/Fm′). The diagonal line represents the 1:1 relationship.
(0.23–0.27 mol m−2 s−1 bar−1) (Table 1). Using J calculated from Eq. (4) with the default values of β (0.84) and ρ2 (0.5), we obtained infinite gm with either method for both species (Table 1). An infinite gm seems to be unlikely; its occurrence was simply due to the default value of βρ2 (0.42) being lower than the value of calibration factor s (0.48–0.56, Table 1), even for the case when the multiphase flash method was used.
The method without calibration also did not result in a realistic estimate of gm when all data were used for curve fitting. Including all data did not alter much the estimate of gm in tomato but lowered the estimate for rice (Table 1).
Table 1 Estimated values (standard error of the estimates in brackets if available) of day respiration Rd, calibration factor s, and mesophyll conductance gm based on combined data of gas exchange and chlorophyll fluorescence with the latter assessed by the use of either the multiphase flash (MPF) method or the single saturation pulse (SP) method, for tomato and rice leaves. Tomato
Rd (μmol m−2 s−1) s gm (mol m−2 s−1 bar−1) gm (mol m−2 s−1 bar−1) se gm (mol m−2 s−1 bar−1) a b c d e
a
b
e
Calibrated J c Uncalibrated J Calibrated J c Uncalibrated J
d
d
Rice
MPF
SP
MPF
SP
2.60(0.33) 0.480(0.007) 0.265(0.035) infinite 0.270(0.015) 1.039(0.489) 0.474(0.005) 0.292(0.051)
2.11(0.31) 0.528(0.004) 0.231(0.019) infinite 0.277(0.013) infinite 0.513(0.004) 0.284(0.034)
1.69(0.65) 0.483(0.008) 0.709(0.410) infinite 0.498(0.079) infinite 0.482(0.006) 0.816(0.677)
1.72(0.44) 0.555(0.005) 0.725(0.297) infinite 0.362(0.041) infinite 0.547(0.005) 1.511(1.690)
estimated from data of the electron transport limited range only, i.e. equivalent to Iinc and Ci levels with which calibration was established. estimated from data of all Iinc and Ci levels. linear electron transport rate J was calculated from the calibration factor s, i.e. J = sIincΔF/Fm′. linear electron transport rate J was calculated from Eq. (4), i.e. J = βρ2IincΔF/Fm′ with β and ρ2 set to their default values 0.84 and 0.5, respectively. values when s and gm are fitted simultaneously to data of the electron transport limited range.
170
Journal of Plant Physiology 220 (2018) 167–172
P.E.L. van der Putten et al.
4. Discussion
parameters. The factor s can be theoretically obtained from various calibration methods that have been developed, based on the gas exchange and chlorophyll fluorescence measurements under non-photorespiratory conditions. Previous analyses in the literature mostly used the relationship either between the quantum yield for CO2 assimilation [ΦCO2 = (A + Rd)/(βIinc)] and (ΔF/Fm′) or between JA [=4(A + Rd)] and JF (the same as calculated from Eq. (4)) (Long and Bernacchi 2003). These calibration methods require Rd to be estimated beforehand, most commonly using the Laisk method (Laisk 1977). In our study we only used a relatively new calibration method described by Yin et al. (2009). The advantage of this calibration method is that Rd, a required input parameter in Eq. (1), can be estimated as the intercept of the same linear regression equation, i.e., A against Iinc(ΔF/Fm′)/4, using data collected under limiting light levels of non-photorespiratory conditions. The estimated Rd using this method is virtually insensitive to measurement errors and similar to values of Rd estimated by the more commonly used Laisk method (see Yin et al., 2011). However, not only does the Laisk method rely on a different set of gas exchange measurements (A-Ci curves of the low Ci range at two or more levels of Iinc), but also is its reliability now questioned within the framework of multiple mesophyll resistance components (Tholen et al., 2012; Walker and Ort, 2015). Within the framework of the multiple mesophyll resistance, any estimated gm based on Eq. (2) or (3) is an apparent value (Yin and Struik 2017; Tcherkez et al., 2017); however, this issue is beyond the scope of our paper. Selecting data to be included for establishing calibration factors is important to have a reliable estimation of gm. Two criteria were used in our analysis: (i) data should be within the electron transport limited range, and (ii) data should be obtained under a non-photorespiratory condition. Criterion (i) suggests that data points from very high irradiances should be excluded because A under high irradiances is most likely Rubisco limited. Including data points where A is Rubisco limited would lead to an underestimation of calibration factor s. Common procedure to obtain a non-photorespiratory condition is to use 2% O2 gas mixture. However, 2% O2 when combined with the ambient CO2 level does not create an absolute non-photorespiratory condition. Some researchers even used data points of entire A-Ci curves at 2% O2 to conduct the calibration (e.g. Flexas et al., 2007; Tomeo and Rosenthal 2017). Not only are points of the lower part of A-Ci curves Rubisco limited, but also photorespiration increases with decreasing Ci. As a result, the calibration factor obtained from entire A-Ci curves not only contains elements for β, ρ2, fractions of alternative electron transport and ξ, but is also confounded by the occurrence of photorespiration. Such confounding procedure should be avoided because of its carryover effect on estimating gm using Eqs. (1) and (3). If one’s experiment design does not permit to exclude photorespiration, it is necessary to use a one-step procedure, in which the calibration factor is not estimated beforehand, but is considered as an additional parameter to be estimated simultaneously with gm by fitting to pooled data points having varying amounts of photorespiration. Such one-step procedure for our data virtually gave nearly identical estimates of the calibration factor s for both single saturation pulse and multiphase flash measurements (Table 1). However, the resulting gm had higher standard errors, especially when gm was high. High estimation uncertainty when gm is high is a common phenomenon (Yin and Struik 2009), because in such a case, gm constrains little leaf photosynthesis, thus not allowing gm being fitted reliably. Our results in Table 1 also show that the estimated gm depended on which part of the data along response curves was used in curve fitting. We previously (Yin and Struik 2009) suggested to use the data of only electron transport limited range (i.e. lower part of A-Iinc curves and higher part of A-Ci curves) because Eq. (3), a part of the model of Farquhar et al. (1980), only applies to this range. When the full data of response curves across all Iinc and Ci levels were used, the estimated gm slightly increased in tomato and decreased in rice (Table 1). Two
Estimates of mesophyll conductance gm are very sensitive to measurement errors (Yin and Struik, 2009). Some errors are random whereas others are more structural due to the insufficient resolution of the measuring techniques. For example, Loriaux et al. (2013) reported that if gm is estimated from combined gas exchange and chlorophyll fluorescence data it is mostly overestimated by the traditional single saturation pulse fluorescence method, and this problem is overcome by employing their new multiphase flash method. Values of J were calculated as Eq. (4) with measured leaf absorptance in their analysis, assuming no absorptance by non-photosynthetic pigments. In our analysis we did not measure leaf absorptance but used a default value of 0.84 combined with an assumed equal excitation partitioning between the two photosystems (Baker, 2008). This resulted in the estimated gm being infinite regardless whether the multiphase flash method or the single saturation pulse method was used for deriving J (Table 1). However, the over-estimation of gm might become less severe if leaf absorptance was measured. Nevertheless, our analysis demonstrated that compared to using the multiphase flash method, conducting calibration based on data under non-photorespiratory conditions had more impact on estimating gm (Table 1). Such a calibration not only accounts for the uncertainty of leaf photosynthetic absorptance but also accounts for that of several other hard-to-measure parameters that are relevant to converting apparent PSII efficiency ΔF/Fm′ into linear electron transport rate J, regardless the single saturation pulse or multiphase flash method to obtain ΔF/Fm′. However, we do not want to suggest that using the multiphase flash method is not important. The calibration implicitly assumes that all the hard-to-measure parameters (β, ρ2, alternative electron fractions, and ξ) are constant across Iinc and Ca levels involved. There is no guarantee that this condition is met. In particular, the underestimation of Fm′ by the single saturation pulse method is more obvious at high than low irradiances (Loriaux et al., 2013). This is shown in Fig. 2, where even with calibrated J, the single saturation pulse-based J deviated more from the multiphase flash-based J with increasing irradiance. Also, the apparent PSII efficiency is calculated as 1-Fs/Fm′ (Genty et al., 1989), meaning that the underestimation of the PSII efficiency by the single saturation pulse method does not go proportionally with its underestimation of Fm′. Thus, mathematically, the calibration does not solve, but only alleviates, the problem of the single saturation pulse method. Values of the calibration factor s were lower with the multiphase flash method than with the single saturation pulse method (Table 1), largely because an underlying factor ξ (see Eq. (6)) is lower when using the multiphase flash method. The s values with the multiphase flash method were ca 0.48 (Table 1), still higher than the default 0.42. It is not possible to point out which factor of β, ρ2 and ξ in Eq. (6) actually contributed to the higher values of s. However, Eq. (6) allows to answer ‘what-if’ questions. The high s values mean that alternative electron transport fractions (fcyc and fpseudo in Eq. (6)) are probably negligible. The theoretical equation to calculate ρ2 is (Yin et al., 2006):
ρ 2=
1 − fcyc (1 − fcyc ) + Φ2/ Φ1
(7)
where Φ2/Φ1 is the ratio of PSII to PSI electron transport efficiency. Since this ratio is generally believed to be relatively conservative as ca 0.85, Eq. (7) would give ca 0.54 for ρ2 if fcyc = 0. Assuming that the ΔF/ Fm′ value from the multiphase flash method represents the true Φ2 (i.e., ξ = 1), one would calculate from Eq. (6) that leaf photosynthetic absorptance β is ca 0.89 for both tomato and rice leaves in our study, which appears to be a reasonable value. Likewise, one can infer what the likely value of ξ or alternative electron fraction or Φ2/Φ1 would be if β is assumed to be known. However, given the plethora of uncertain parameters underlying the factor s, we rather refrain from speculating about the range of possible variation for each of these uncertain 171
Journal of Plant Physiology 220 (2018) 167–172
P.E.L. van der Putten et al.
mechanisms are behind this change. First, when data used involve the range where A is limited by mechanisms other than electron transport, additional pseudocyclic electron transport (to O2, most likely) would occur in case that non-photochemical quenching does not increase accordingly (Ort and Baker, 2002). This would result in an underestimation of gm as this additional electron transport flux is wrongly attributed to the limitation by mesophyll resistance (Yin et al., 2009). Secondly, gm has often been reported to increase with increasing Iinc (e.g. Yin et al., 2009; Douthe et al., 2012) and with decreasing Ci (e.g. Flexas et al., 2007). The first mechanism seems to explain the decreased estimate of gm in rice while the second mechanism may explain the increased gm in tomato. Because crops were shown to respond differently to the expansion of the data range, we re-state here the importance of using data of only the electron limited range for estimating gm when investigating any difference in gm among various treatments or species or cultivars.
indica × japonica rice inbred lines. Photosynth. Res. 134, 27–38. Laisk, A.K., 1977. Kinetics of Photosynthesis and Photorespiration in C3 Plant. Nauka Moscow (in Russian). Li, Y., Gao, Y.X., Xu, X.M., Shen, Q.R., Guo, S.W., 2009. Light-saturated photosynthetic rate in high-nitrogen rice (Oryza sativa L.) leaves is related to chloroplastic CO2 concentration. J. Exp. Bot. 60, 351–2360. Long, S.P., Bernacchi, C.J., 2003. Gas exchange measurements, what can they tell us about the underlying limitations to photosynthesis? Procedures and sources of error. J. Exp. Bot. 54, 2393–2401. Loriaux, S.D., Avenson, T.J., Wells, J.M., McDermitt, D.K., Eckles, R.D., Riensche, B., Genty, B., 2013. Closing in on maximum yield of chlorophyll fluorescence using a single multiphase flash of sub-saturating intensity. Plant Cell Environ. 36, 1755–1770. Martins, S.C.V., Galmes, J., Molins, A., DaMatta, F.M., 2013. Improving the estimation of mesophyll conductance to CO2: on the role of electron transport rate correction and respiration. J. Exp. Bot. 64, 3285–3298. Orr, D.J., Alcantara, A., Kapralov, M.V., Andralojc, P.J., Carmo-Silva, E., Parry, M.A.J., 2016. Surveying Rubisco diversity and temperature response to improve crop photosynthetic efficiency. Plant Physiol. 172, 707–717. Ort, D.R., Baker, N.R., 2002. A photoprotective role for O2 as an alternative electron sink in photosynthesis? Curr. Opin. Plant Biol. 5, 193–198. Pons, T.L., Flexas, J., von Caemmerer, S., Evans, J.R., Genty, B., Ribas-Carbo, M., Brugnoli, E., 2009. Estimating mesophyll conductance to CO2: methodology, potential errors, and recommendations. J. Exp. Bot. 60, 2217–2234. Singh, S.K., Reddy, V.R., 2016. Methods of mesophyll conductance estimations: its impact on key biochemical parameters and photosynthetic limitations in phosphorusstressed soybean across CO2. Physiol. Plant. 157, 234–254. Tcherkez, G., Gauthier, P., Buckley, T.N., et al., 2017. Leaf day respiration: low CO2 flux but high significance for metabolism and carbon balance. New Phytol. 216, 986–1001. Tholen, D., Ethier, G., Genty, B., Pepin, S., Zhu, X.-G., 2012. Variable mesophyll conductance revisited: theoretical background and experimental implications. Plant Cell Environ. 35, 2087–2103. Tomeo, N.J., Rosenthal, D.M., 2017. Variable mesophyll conductance among soybean cultivars sets a tradeoff between photosynthesis and water-use-efficiency. Plant Physiol. 174, 241–257. Valentini, R., Epron, D., Angelis, P.D.E., Matteucci, G., Dreyer, E., 1995. In situ estimation of net CO2 assimilation, photosynthetic electron flow and photorespiration in Turkey oak (Q. cerris L.) leaves: diurnal cycles under different levels of water supply. Plant Cell Environ. 18, 631–640. Walker, B.J., Ort, D.R., 2015. Improved method for measuring the apparent CO2 photocompensation point resolves the impact of multiple internal conductance to CO2 to net gas exchange. Plant Cell Environ. 38, 2462–2474. Walker, B.J., Ariza, L.S., Kaines, S., Badger, M.R., Cousins, A.B., 2013. Temperature response of in vivo Rubisco kinetics and mesophyll conductance in Arabidopsis thaliana: comparisons to Nicotiana tabacum. Plant Cell Environ. 36, 2108–2119. Warren, C.R., 2006. Estimating the internal conductance to CO2 movement. Funct. Plant Biol. 33, 431–442. Yin, X., Struik, P.C., 2009. Theoretical reconsiderations when estimating the mesophyll conductance to CO2 diffusion in leaves of C3 plants by analysis of combined gas exchange and chlorophyll fluorescence measurements. Plant Cell Environ. 32, 1513–1524 with corrigendum in PC&E 33: 1595. Yin, X., Struik, P.C., 2017. Simple generalisation of a mesophyll resistance model for various intracellular arrangements of chloroplasts and mitochondria in C3 leaves. Photosynth. Res. 132, 211–220. Yin, X., Harbinson, J., Struik, P.C., 2006. Mathematical review of literature to assess alternative electron transports and interphotosystem excitation partitioning of steady-state C3 photosynthesis under limiting light. Plant Cell Environ. 29, 1771–1782 (corrigendum in PCE 2006. 29: 2252). Yin, X., Struik, P.C., Romero, P., Harbinson, J., Evers, J.B., van der Putten, P.E.L., Vos, J., 2009. Using combined measurements of gas exchange and chlorophyll fluorescence to estimate parameters of a biochemical C3 photosynthesis model: a critical appraisal and a new integrated approach applied to leaves in a wheat (Triticum aestivum) canopy. Plant Cell Environ. 32, 448–464. Yin, X., Sun, Z., Struik, P.C., Gu, J., 2011. Evaluating a new method to estimate the rate of leaf respiration in the light by analysis of combined gas exchange and chlorophyll fluorescence measurements. J. Exp. Bot. 62, 3489–3499.
References Adachi, S., Nakae, T., Uchida, M., Soda, K., Takai, T., Oi, T., Yamamoto, T., Ookawa, T., Miyake, H., Yano, M., Hirasawa, T., 2013. The mesophyll anatomy enhancing CO2 diffusion is a key trait for improving rice photosynthesis. J. Exp. Bot. 64, 1061–1072. Baker, N.R., 2008. Chlorophyll fluorescence: a probe of photosynthesis in vivo. Annu. Rev. Plant Biol. 59, 89–113. Bellasio, C., Beerling, D.J., Griffiths, H., 2016. An Excel tool for deriving key photosynthetic parameters from combined gas exchange and chlorophyll fluorescence: theory and practice. Plant Cell Environ. 39, 1180–1197. Cousins, A.B., Ghannoum, O., von Caemmerer, S., Badger, M.R., 2010. Simultaneous determination of Rubisco carboxylase and oxygenase kinetic parameters in Triticum aestivum and Zea mays using membrane inlet mass spectrometry. Plant Cell Environ. 33, 444–452. Douthe, C., Dreyer, E., Brendel, O., Warren, C.R., 2012. Is mesophyll conductance to CO2 in leaves of three Eucalyptus species sensitive to short-term changes of irradiance under ambient as well as low O2? Funct. Plant Biol. 39, 435–448. Earl, H.J., Ennahli, S., 2004. Estimating photosynthetic electron transport via chlorophyll fluorometry without Photosystem II light saturation. Photosynth. Res. 82, 177–186. Evans, J.R., von Caemmerer, S., Setchell, B.A., Hudson, G.S., 1994. The relationship between CO2 transfer conductance and leaf anatomy in transgenic tobacco with a reduced content of Rubisco. Aust. J. Plant Physiol. 21, 475–495. Evans, J.R., 2009. Potential errors in electron transport rates calculated from chlorophyll fluorescence as revealed by a multilayer leaf model. Plant Cell Physiol. 50, 698–706. Farquhar, G.D., von Caemmerer, S., Berry, J.A., 1980. A biochemical model of photosynthetic CO2 assimilation in leaves of C3 species. Planta 149, 78–90. Flexas, J., Diaz-Espejo, A., Galmes, J., Kaldenhoff, R., Medrano, H., Ribas-Carbó, M., 2007. Rapid variation of mesophyll conductance in response to changes in CO2 concentration around leaves. Plant Cell Environ. 30, 1284–1298. Flexas, J., Ribas-Carbó, M., Diaz-Espejo, A., Galmes, J., Medrano, H., 2008. Mesophyll conductance to CO2: current knowledge and future prospects. Plant Cell Environ. 31, 602–621. Genty, B., Briantais, J.-M., Baker, N., 1989. The relationship between the quantum yield of photosynthetic electron transport and quenching of chlorophyll fluorescence. Biochim. Biophys. Acta 990, 87–92. Gilbert, M.E., Pou, A., Zwieniecki, M.A., Holbrook, N.M., 2012. On measuring the response of mesophyll conductance to carbon dioxide with the variable J method. J. Exp. Bot. 63, 413–425. Gu, J., Yin, X., Stomph, T.J., Wang, H., Struik, P.C., 2012. Physiological basis of genetic variation in leaf photosynthesis among rice (Oryza sativa L.) introgression lines under drought and well-watered conditions. J. Exp. Bot. 63, 5137–5153. Harley, P.C., Loreto, F., Di Marco, G., Sharkey, T.D., 1992. Theoretical considerations when estimating the mesophyll conductance to CO2 flux by analysis of the response of photosynthesis to CO2. Plant Physiol. 98, 1429–1436. He, W., Adachi, S., Sage, R.F., Ookawa, T., Hirasawa, T., 2017. Leaf photosynthetic rate and mesophyll cell anatomy changes during ontogenesis in backcrossed
172