Fitting photosynthetic carbon dioxide and irradiance response curves for C4 leaves Authors: Russell Woodford, Maria Ermakova, Robert T Furbank, Susanne von Caemmerer, Alistair Rogers Published: 2026 ABSTRACT AbstractGas exchange measurements provide crucial insights into the complex mechanisms of photosynthesis. Responses of CO2 assimilation rate (A) to intercellular CO2 partial pressure (Ci) and irradiance (I) link gas exchange measurements to the underlying photosynthetic biochemistry of a leaf. The unique biochemistry and leaf anatomy which distinguish C4 photosynthesis make it necessary to apply models and fitting routines which appropriately parameterize and incorporate these characteristics. Here we provide updates to the C4 photosynthesis model by improving the parameterization of cyclic electron flow in C4 photosynthesis using experimentally derived values from Setaria viridis. We additionally describe two fitting routines for assessing C4 photosynthesis based on the updated model. Fitting of a CO2 response curve (A/Ci) provides estimates of maximum phosphoenolpyruvate carboxylase activity (Vpmax) and maximum Rubisco activity (Vcmax), and calculates the linear electron transport rate (J) needed to sustain the measured CO2 assimilation rate (A). Fitting of an irradiance response curve (A/I) provides estimates for the maximum linear electron transport rate (Jmax), day respiration rate (Rd), the quantum yield (ϕCO2), and light compensation point (Γlight). Values of the above output parameters are provided both at the measurement temperature and at 25 °C for ease of comparative reporting. The fitting tool has been designed in Microsoft Excel to minimize barriers to entry and enable simplicity of fitting while simultaneously catering to individuals with diverse expertise and experience in C4 gas exchange modelling.Keywords: A/Ci curves, A/I curves, C4 photosynthesis model, CO2 assimilation, gas exchange Introduction CO2 assimilation by leaves depends on the response of underlying photosynthetic biochemistry to environmental variables such as light, CO2, and temperature. The Farquhar et al. model of C3 photosynthesis has associated the gas exchange of C3 leaves with their fundamental biochemical mechanisms (Farquhar et al., 1980; von Caemmerer and Farquhar, 1981, 2025). Gas exchange measurements of CO2 response curves are now routinely used to extract biochemical parameters such as the maximum activity of Rubisco (Vcmax), maximum rate of electron transport for the given light intensity, and triose phosphate limitation (TPU) from C3 leaves using an easy curve fitting routine (Sharkey et al., 2007). Here we have developed a curve fitting routine for CO2 response and light response curves of the C4 photosynthetic pathway based on the updated model of C4 photosynthesis (von Caemmerer, 2021). We have also updated values for the fraction of cyclic electron flow (fcyc) and number of protons shuttled per cyclic electron flow (HJcyc) in C4 photosynthesis. The C4 photosynthetic pathway is complex and involves a biochemical CO2 concentrating mechanism (CCM) operating across mesophyll and bundle sheath cells. In the mesophyll, CO2 is initially fixed by phosphoenolpyruvate (PEP) carboxylase (PEPC) into C4 acids, which are then decarboxylated in the bundle sheath to supply CO2 for Rubisco. The CCM allows Rubisco to operate at high CO2 partial pressures, overcoming its low affinity for CO2, greatly inhibiting its oxygenation reaction, and reducing photorespiration rates. Both the structure of the bundle sheath cell wall (which has a low permeability to CO2) and the relative biochemical capacities of the C3 cycle in the bundle sheath and C4 cycle operating across the mesophyll–bundle sheath interface contribute to the high CO2 partial pressure in the bundle sheath. This leads to distinct CO2 response [CO2 assimilation rate (A)/intercellular CO2 partial pressure (Ci)] curves of the C4 photosynthetic pathway, characterized by a steep initial slope and saturation of the CO2 assimilation rate at lower Ci, in comparison with A/Ci curves of C3 species (von Caemmerer, 2021; von Caemmerer et al., 2012). Furthermore, due to the CCM, CO2 compensation points of C4 plants are low and independent of O2 partial pressures. C4 light response (A/I) curves can be characterized by a lack of saturation, even at irradiance >2000 µmol m−2 s−1 (Ermakova et al., 2019; Leakey et al., 2006), and the quantum yield independent of temperature (Ehleringer and Björkman, 1977). Although C4 plants are not limited by CO2 availability under ambient conditions, the electron transport reactions producing ATP and NADPH could limit the rate of C4 photosynthesis at high irradiance (Ermakova et al., 2019; von Caemmerer and Furbank, 2016). Two major electron transfer pathways differentially contribute to ATP and NADPH production: linear electron flow (J) and cyclic electron flow (Jcyc). J involves Photosystem II (PSII), Cytochrome b6f complex (Cyt b6f), Photosystem I (PSI), ferredoxin:NADP+ oxidoreductase (FNR), and ATP synthase. At most, this produces ATP and NADPH in an ATP:NADPH ratio of 3:2 (Kramer and Evans, 2011). However, for each molecule of CO2 fixed, C4 photosynthesis requires an ATP:NADPH ratio of at least 5:2 (von Caemmerer and Furbank, 2016). Jcyc is an alternative pathway of electron transport used to increase ATP yield (Kramer and Evans, 2011; Yamori and Shikanai, 2016). In Jcyc, electrons from PSI bypass FNR and return to Cyt b6f via either a chloroplast NADPH dehydrogenase-like (NDH) complex-dependent pathway or a Proton Gradient Regulation 5 (PGR5)-dependent pathway, leading to the production of ATP but not NADPH (Yamori and Shikanai, 2016). To accommodate their increased ATP requirements, it has been proposed that C4 species up-regulate Jcyc (Ishikawa et al., 2016; Munekage and Taniguchi, 2016; Nakamura et al., 2013). However, previous iterations of the C4 photosynthesis model have either ignored or underestimated the contributions of Jcyc to total electron transport (von Caemmerer, 2021; von Caemmerer and Furbank, 1999; Yin et al., 2021), potentially miscalculating the influence Jcyc could have on gas exchange measurements, particularly when fitting A/I curves, resulting in inaccurate determination of electron transport rates. Gas exchange analysis Gas exchange analyses were performed using a LI-COR LI-6800 equipped with a 6800-01A fluorometer head. All measurements used 90% red:10% blue actinic light, 25 °C leaf temperature, 55% relative chamber humidity, 500 µmol m−2 s−1 flow rate, 0.1 kPa flow pressure, and 10 000 rpm fan speed. For decreasing irradiance light response curves, wild-type Setaria viridis leaves were equilibrated at 500 μmol m−2 s−1 irradiance and 400 μmol mol−1 CO2 for 15 min. CO2 assimilation was then recorded at 3 min intervals with progressively decreasing irradiance from 3000 μmol m−2 s−1 to 0 μmol m−2 s−1. The C4 photosynthetic model with additional detail of the electron transport rate The first models to capture the C4 photosynthetic biochemistry were designed by Berry and Farquhar (1978) and Peisker (1979). The Berry and Farquhar model did not provide analytical solutions but was able to predict high bundle sheath CO2 partial pressures and their dependence on bundle sheath conductance. Many of the gas exchange characteristics of C4 photosynthesis observed with intact leaves could be predicted by these models. Numerous subsequent additions have since incrementally improved and implemented the C4 photosynthesis model. Such advancements include better consideration of the energy budgets of different C4 subtypes (Yin and Struik, 2018, 2021), quantification of bundle sheath conductance (Alonso-Cantabrana et al., 2018; Bellasio et al., 2024; Yin et al., 2011b, 2016) and leakiness (Bellasio and Griffiths, 2014; Kromdijk et al., 2008; Serrano-Romero and Cousins, 2020; Wang et al., 2022), improved parameterization of enzyme activities and their responses to temperature (Boyd et al., 2015; DiMario and Cousins, 2019), application of the C4 photosynthesis model to study environmental factors influencing different C4 species (Fan et al., 2025), integration of combined gas exchange and chlorophyll fluorescence measurements into the C4 model (Bellasio et al., 2016; Yin et al., 2011b), and extension of the model into broader, Earth system and crop growth models (Harbinson and Yin, 2023; Scott and Smith, 2022; Wu et al., 2019). The 40-year-long evolution of steady-state models of photosynthesis is reviewed in Yin et al. (2021). Recently, von Caemmerer (2021) revised and updated the C4 photosynthesis model with Rubisco and PEPC kinetic parameters derived from S. viridis, an NADP-dependent malic enzyme (NADP-ME) subtype C4 model species, and incorporated expanded electron transport functions and temperature functions, improving the overall accuracy of the model. Here we have further updated the model of von Caemmerer (2021) with better parameterized electron transport by updating values for the proton:ATP production ratio of ATP synthase (h), the number of protons pumped per electron in Jcyc (HJcyc), and the fraction of total electron transport that can be attributed to Jcyc (fcyc). This is achieved through both a theoretical and an empirical approach. At a fundamental level, theoretical estimates of J, Jcyc, and fcyc depend on the number of protons translocated from the chloroplast stroma to the thylakoid lumen per electron transported and the number of protons required by chloroplast ATP synthase to produce one ATP (h). However, uncertainty remains regarding the exact values of these parameters. While the number of protons pumped per electron transported via J (HJ) is generally considered to be 3, the number of protons pumped per Jcyc (HJcyc) is variable, probably ranging from 2 to 4 depending on the relative fraction of Jcyc occurring via the NDH- and PGR5-meditated Jcyc pathways (fNDHcyc and fPGR5cyc) (Equation 1). Exclusively PGR5-mediated Jcyc is predicted to give an HJcyc of 2, while exclusively NDH-mediated Jcyc is predicted to give an HJcyc of 4, and a mixture of these pathways is predicted to result in an intermediate value (Kramer and Evans, 2011; Strand et al., 2017). where: fNDHcyc+fPGR5cyc=1 The parameter h is commonly attributed a value of either 4.67, predicted from the stoichiometric ratio of c- to β-subunits of chloroplast ATP synthase (Vollmar et al., 2009), or 4, determined from experimental measurements (Petersen et al., 2012). The ATP:NADPH ratio produced by J (ATPNADPHJ) directly depends on h, with h=4.67 resulting in an ATPNADPHJ ratio of 2.6:2 (=1.29), and h=4 resulting in a ratio of 3:2 (=1.5) (Kramer and Evans, 2011). As Jcyc is proposed to only produce ATP, the h and ATPNADPHJ dictate the fcyc needed to fulfil the minimum ATP:NADPH requirement (ATPNADPHReq) of 5:2 (=2.5) for C4 photosynthesis. The fcyc needed to fulfil this minimum energetic requirement can be theoretically estimated for ATPNADPHJ and varying HJcyc using Equation 2. This provides an fcyc range of 0.33–0.50 when h=4, and an fcyc range of 0.41–0.59 when h=4.67 (Fig. 1). Dependence of the fraction of electron transport occurring through cyclic electron flow (fcyc) on the ATP:NADPH requirement of C4 photosynthesis and the number of protons per electron pumped by cyclic electron flow (HJcyc) when 4.67 protons are required by chloroplast ATP synthase to produce one ATP. (A) The fcyc required to accommodate the minimal ATP:NADPH requirement of C4 photosynthesis (values >2.5, white range) is dependent on HJcyc. Increased fcyc is required to accommodate C4 photosynthesis when fewer protons are pumped per electron. (B) The range of fcyc values required to accommodate varying ATP:NADPH across physiologically reasonable HJcyc (values between 2 and 4). Lower fcyc is required for C4 photosynthesis when higher numbers of protons per electron are pumped. The theoretical fcyc required for C3 photosynthesis (ATP:NADPH=1.5) is less than half that for C4 photosynthesis. Values of fcyc within either theoretical range are supported empirically. Treatment of S. viridis bundle sheath cells with methyl viologen, an electron acceptor from PSI competing with Jcyc, demonstrates that fcyc is 0.84 in bundle sheath cells (Ermakova et al., 2024). In C3 plants, which only have mesophyll cells and little NDH in optimal conditions, fcyc is estimated to be 0.12 (Avenson et al., 2005). Due to similarities in their electron transport chains, fcyc in C3 and C4 mesophyll cells could be assumed to be similar (Munekage and Taneguchi, 2016). However, analysis of the light-driven proton flux across the thylakoid membrane (representing Jcyc and J) and the relative PSII electron transport rate (representing J) in PGR5-deficient S. viridis suggest that PGR5-mediated Jcyc accounts for at least 20% of mesophyll electron transport (Jm) (Woodford et al., 2025). As ∼40% of total leaf chlorophyll is localized to bundle sheath cells in S. viridis (Ermakova et al., 2021a), the leaf-level fcyc can be calculated using Equation 3. Previous empirical values for fcyc determined from measurements of the quantum efficiency of oxygen evolution under limiting light align with these calculations, with fcyc values ranging from 0.32 to 0.62, and averaging 0.45, assuming h=4 (Bjorkman and Demmig, 1987; Yin and Struik, 2012). Estimates of fcyc calculated using measurements of P700+ absorbance, chlorophyll fluorescence, and oxygen evolution light response curves provided an fcyc value of 0.53 for NADP-ME species (Kiirats et al., 2010; Yin and Struik, 2012). Given the consistency in these values, we have therefore assigned an fcyc value of 0.45. By rearranging Equation 2 to give Equation 4, HJcyc can be calculated for known fcyc, resulting in HJcyc=2.39 if h=4, and HJcyc=3.39 if h=4.67. By further exploring HJcyc through rearranging Equation 1, the relative fraction of Jcyc occurring via the NDH- and PGR5-meditated Jcyc pathways on a leaf level can be calculated using Equations 5 and 6. If h=4, a 19.3% contribution of NDH and an 80.7% contribution of PGR5 to total leaf Jcyc is expected, while h=4.67 necessitates a 69% contribution of NDH and a 31% contribution of PGR5. As it is proposed that the NDH-dependent pathway provides the dominant contribution to Jcyc in C4 photosynthesis, as opposed to the PGR5-dependent pathway which has yet to be shown to contribute to ATP synthesis (Ermakova et al., 2024; Ogawa et al., 2023; Woodford et al., 2025), h=4.67 is likely to be the more biologically representative value. Given the considerations presented here, we have assigned an h value of 4.67, an fcyc value of 0.45, and an HJcyc value of 3.4. These parameters are used to determine the ratio of the rate of ATP production to linear electron flow (z), required for the calculation of light- and electron transport-limited rate equations (see C4 photosynthesis fitting routines below). While von Caemmerer (2021) previously calculated z assuming h=4, an fcyc=0.3, and HJcyc=2, calculation of z as described by Yin and Struik (2012) allows incorporation of updated h, fcyc, and HJcyc values, more accurately reflecting the contributions of cyclic electron flow to C4 photosynthesis (Equation 7). The values we assign here result in a z value of 1.24. It is necessary to recognize that the values determined for fcyc, HJcyc, and z here may vary under different conditions. fcyc could be higher to accommodate CO2 leakiness of bundle sheath cells (ϕ), H+ leakage of thylakoid membranes, or additional, unaccounted for ATP requirements, particularly during photosynthetic induction, under non-steady-state conditions or at high temperatures (Bukhov et al., 1999; Ermakova et al., 2023; Groth and Junge, 1993; Wang et al., 2022). For instance, a ϕ of 0.3 during the early stages of photosynthetic induction would result in increased ATP demand, driving ATPNADPHReq towards 5.9:2 (=2.95) (Wang et al., 2022), resulting in an increased fcyc requirement of 0.53 (Equation 2). For other stress-inducing or non-steady-state conditions, experimental quantification of Jcyc and fcyc in C4 species is limited, highlighting the need for additional experimental work to better inform and reinforce model parameterization. Key cyclic electron transport parameters may also vary for other C4 subtypes. NAD-ME subtype C4 photosynthesis requires the same ATP:NADPH ratio as NADP-ME, so would be expected to have similar leaf-level fcyc; however, the relative fcyc in mesophyll and bundle sheath chloroplasts is likely to differ between these subtypes due to differences in cell-specific energetic requirements (von Caemmerer and Furbank, 2016; Yin and Struik, 2018). These cell-specific considerations cannot be appropriately incorporated into the theoretical approach used here without further empirical support, which is currently lacking. Therefore, for the NAD-ME subtype using NAD-dependent malic enzyme, fcyc most probably ranges from 0.41 to 0.59 (Equation 2). ‘Pure’ PCK subtype using PEP carboxykinase requires a 4:2 ratio of ATP to NADPH, therefore requiring an fcyc between 0.29 and 0.45 (Equation 2), lower than that expected for NADP-ME and NADP-ME subtypes. Theoretically, in the pure PCK subtype, fcyc may be even closer to that of C3 species (fcyc ∼0.2 for HJcyc=2) if ATP for PCK is supplied by mitochondrial electron transport, thereby reducing the ATPNADPHReq of chloroplast electron transport (von Caemmerer and Furbank, 2016; Yin and Struik, 2021). More realistically, in a mixed decarboxylation context such as that seen in maize, NADP-ME photosynthesis with up to 25% of C4 flux via PCK would require an ATPNADPHReq of 2.375 (Ermakova et al., 2026), slightly reducing the requirement for Jcyc and resulting in an expected fcyc range of 0.39–0.56 (Equation 2). C4 photosynthesis fitting routines Using the von Caemmerer (2021) model and incorporating the improvements detailed above, we have developed an Excel fitting routine for A/Ci and A/I curves measured on C4 leaves (Supplementary Protocol S1). The fitting routines are based on Equations 8–11 and 12–15 for A/Ci and A/I curves, respectively (21–24 and 40–43 from von Caemmerer, 2021), with the essential parameters used in the fitting routine presented in Table 1. The parameters used in the routines have been derived from S. viridis, which offers the most complete parameter set. Kinetic enzyme parameters for Zea mays are presented in Crawford and Cousins (2022), which lack temperature-dependent characterization so have not been used here, but could be incorporated into fitting routines if users desire. Table 1.Photosynthetic parametersa used in the fitting routinesParameterValue at 25 °CDefinitionActivation energy(E, kJ mol−1)bReference K c 1210 μbarMichaelis constant of Rubisco for CO264.2 Boyd et al. (2015) K o 292 mbarMichaelis constant of Rubisco for O210.5 Boyd et al. (2015) K p 82 μbarMichaelis constant of PEP carboxylase for CO238.3 DiMario and Cousins (2019) γ*0.0003817 (0.5/1310)0.5/(Sc/o) half the reciprocal of Rubisco specificity31.1 Boyd et al. (2015) g bs 0.003 mol m−2 s−1 bar−1Bundle sheath conductance to CO2 Alonso-Cantabrana et al. (2018) g m 1 mol m−2 s−1 bar−1Mesophyll conductance to CO249.8 Ubierna et al. (2017) a o 0.047Bundle sheath conductance to O21.63 Farquhar (1983) R m 0.5 RdMesophyll mitochondrial respiration66.4 Farquhar et al. (1980) α0<α<1Fraction of PSII activity in the bundle sheath.Recommended values are 0.01 for NADP-ME and 0.5 for NAD-ME and PCK species x 0.4Partitioning factor of electron transport rate T opt 43 °COptimum temperature of electron transport Jmax(TL)=Jmax(To)e−(TL−ToΩ)2 June et al. (2004); Yamori et al. (2010) J max@ ToptCalculated from Jmax at 25 °CMaximum linear electron transport rate at ToptΩ26The difference in temperature from T0 at which J falls to e−1 (0.37) Yamori et al. (2010) h 4.67Number of protons per ATP H Jcyc 3.4H+ per e− for cyclic electron flow H J 3H+ per e− for linear electron flow f cyc 0.45Fraction of total electron transport attributed to cyclic electron flow Yin and Struik (2012) z 1.24Ratio of the rate of ATP production to linear electron flow (JATP/J) Yin and Struik (2012) V pmax Fitted in A/Ci curve (default value: 200 μmol m − 2 s − 1) Maximum PEP carboxylase activity 50.1 Boyd et al. (2015) V cmax Fitted in A/Ci curve (default value: 40 μmol m − 2 s − 1) Maximum Rubisco activity 78 Boyd et al. (2015) J Fitted in A/Ci Curve Linear electron transport rate required to sustain measured CO2 assimilation R d Fitted in A/I curve (default value: 1 μmol m − 2 s − 1) Leaf mitochondrial respiration 66.4 Farquhar et al. (1980) J max Fitted in A/I curve (default value: 200 μmol m − 2 s − 1) Maximum linear electron transport rate J max cyc Fitted in A/I curve Maximum cyclic electron transport rate J max 1 Fitted in A/I curve Maximum electron transport rate through PSI θ Fitted in A/I curve (default value: 0.8) Empirical curvature factor Γlight Fitted in A/I curve Light compensation point ϕCO2 Fitted in A/I curve Quantum yield of CO2 assimilation a Parameters for Setaria viridis are used in the fitting routines. For kinetic enzyme parameters of Zea mays, see Crawford and Cousins (2022). b Values of parameters where activation energies have been given can be calculated at any temperature using the Arrhenius equation (Equation 16). Photosynthetic parametersa used in the fitting routines a Parameters for Setaria viridis are used in the fitting routines. For kinetic enzyme parameters of Zea mays, see Crawford and Cousins (2022). b Values of parameters where activation energies have been given can be calculated at any temperature using the Arrhenius equation (Equation 16). Enzyme-limited rate equation: where: Light- and electron transport-limited rate equation: where: For both A/Ci and A/I curve fitting, the routine starts with modelled data calculated using default output parameters (Table 1). After the user inputs measured gas exchange data, the routine determines the squared difference between values for each pair of measured and modelled data points, then sums the squared differences across all pairs of data points in the curve (sum of squares). The Excel Solver ‘Minimise’ function is then applied to iteratively adjust the output parameters to solve for values that minimize the sum of squares between measured and modelled data, providing output values that give the modelled curve the closest fit to experimental values. The root mean squared error (RMSE) is also provided. This overall approach is similar to the approach used in the Excel fitting tool developed for gas exchange measurements of C3 leaves by Sharkey et al. (2007). Fitting a CO2 response curve Before fitting an A/Ci curve, one first updates leaf temperature, atmospheric pressure, irradiance, and O2 concentration in the spreadsheet. Depending of the biochemical subtype of the species examined, the fraction of PSII active in the bundle sheath (α) is set to a value between 0 and 1. NADP-ME monocot species such as Z. mays and Sorghum bicolor have little or no PSII activity in the bundle sheath in optimal conditions, whereas NADP-ME dicots, NAD-ME, and PCK species can have high bundle sheath PSII activity (von Caemmerer, 2021, and references therein). The α value should be updated to reflect this. The CO2 and O2 concentrations are frequently expressed in mole fractions (i.e. µmol mol−1, ppm, µl l−1) and can be entered as such. These CO2 and O2 mole fractions are then converted to partial pressure with the use of the given atmospheric pressure, enabling the use of partial pressure for the units of CO2 and O2 in all calculations. This is justified as the chemical activity of a dissolved gas is proportional to its gas phase, and thus the partial pressure of a gas existing in equilibrium with solution is a better measure of its chemical activity than dissolved concentrations (Badger and Collatz, 1977; Sharkey et al., 2007). Furthermore, the mole fraction of CO2 is the same at sea level and at high altitude, but the partial pressure declines with increasing elevation as atmospheric pressure declines. To fit an A/Ci curve one enters the measured values for CO2 assimilation rate (A) and intercellular CO2 (Ci). In the fitting routine of an A/Ci curve for C3 photosynthesis, users must choose which data points are used to calculate the Rubisco-limited, electron transport-limited, or triose phosphate-limited rates (Sharkey et al., 2007; von Caemmerer and Farquhar, 1981). In the C4 photosynthesis model, such choices do not need to be made. Simply pressing the calculate button will solve for Vcmax (the maximum Rubisco activity) and Vpmax (the maximum PEPC activity) values which minimize the sum of squares between experimental and modelled data. After converging on a solution, the Vcmax and Vpmax will be the output for the modelled enzyme-limited CO2 assimilation rate, which is appropriate at high irradiance. At the same time, the electron transport rate, J, required to sustain the CO2 assimilation rate is also calculated from the enzyme-limited rate (Supplementary Protocol S2). These values are given at both the measurement temperature and calculated at 25 °C for reporting purposes using the Arrhenius equation (Equations 16, 17) and activation energies in Table 1. rearranged to: where: E is activation energy in J mol−1 and x is temperature-dependent component of Arrhenius equation: where: T is measurement temperature in °C, R is universal gas constant (8.314 J K−1 mol−1). Interpretation of the A/Ci fitting results and resolving limitations At high irradiance, the C4 photosynthetic model predicts that A is limited by Vpmax at low Ci and co-limited by Vcmax and J at high Ci. Together, Vpmax, Vcmax, and J, given in units of µmol m−2 s−1, describe C4 A/Ci curves. Vcmax and Vpmax can be compared with in vitro measurements of those activities (Cousins et al., 2007), while the electron transport rate can be related to the Cyt b6f content in leaves and to a spectrophotometric assay for Cyt f (Ermakova et al., 2019; Heyno et al., 2022; Yamori et al., 2010). Leaf nitrogen is another parameter that is a useful correlate with CO2 assimilation rate, particularly in field studies (Ghannoum et al., 2005). As Rubisco and electron transport rate co-limit photosynthetic flux at high Ci and high light, distinguishing these limitations might be challenging. For example, fitting of A/Ci from S. viridis plants overexpressing the Rieske FeS protein of Cyt b6f, which have increased electron transport capacity but no changes in Rubisco content (Ermakova et al., 2019), suggests that both Vcmax and J are increased compared with control (Fig. 2). If the primary limitation needs to be resolved, strategies can be employed to help disentangle these parameters. We recommend making a series of A measurements at high Ci with increasing irradiances. If increasing irradiance results in an increased CO2 assimilation rate, it is likely that electron transport is the limiting step rather than Vcmax. When J is limiting, one must be aware that Vcmax does not represent the maximum Rubisco activity, but rather the activity allowed by the electron transport rate at a given irradiance. Measurements of A/I curves on the same leaves can provide further insight into Rubisco and electron transport limitations (Fig. 3, see below). Fitting of the CO2 response curves of CO2 assimilation (A) from control and Rieske FeS overexpression S. viridis plants. Data were originally reported in Ermakova et al. (2019). Dotted lines represent the fit of modelled data to the measured data points. The fitted values for Rieske FeS overexpression (FeS-OE) plants are Vpmax=205.27 µmol m−2 s−1, Vcmax=61.69 µmol m−2 s−1, and J=197.83 µmol m−2 s−1. For control, Vpmax=239.8 µmol m−2 s−1, Vcmax=48.12 µmol m−2 s−1, and J=168.98 µmol m−2 s−1. Fitting of the light response curves of CO2 assimilation (A) from control and Rieske FeS overexpression S. viridis plants. Data were originally reported in Ermakova et al. (2019). Dotted lines represent the fit of modelled data to the measured data points. The fitted values for Rieske FeS overexpression (FeS-OE) plants are Rd=1.22 µmol m−2 s−1, Jmax=194.97 µmol m−2 s−1, θ=0.75, ϕCO2=0.054, and Γlight=30.68 µmol m−2 s−1. For control plants, Rd=1.56 µmol m−2 s−1, Jmax=185.91 µmol m−2 s−1, θ=0.65, ϕCO2=0.040, and Γlight=25.91 µmol m−2 s−1. Important A/Ci fitting considerations It is thought that in most C4 leaves, carbonic anhydrase activity is sufficiently high and does not limit photosynthetic flux (Cousins et al., 2008; Hatch and Burnell, 1990). As such, we have not included carbonic anhydrase activity as a limitation here. The pre-set value for bundle sheath conductance (gbs) is 0.003 mol m−2 s−1 bar−1, consistent with measured values for S. viridis (Table 1; Danila et al., 2021) and sufficiently low to reproduce the observed lack of oxygen sensitivity of C4 photosynthesis (von Caemmerer, 2000). It is important to note that gbs values may vary between C4 species, and the temperature sensitivity of gbs remains contentious, with variations in gbs temperature sensitivity potentially due to differences in estimation technique (Alonso-Cantabrana et al., 2018; Kiirats et al., 2002; Serrano-Romero and Cousins, 2020; Yin et al., 2016). In S. viridis, gbs is consistently ∼0.003 mol m−2 s−1 bar−1 with limited temperature sensitivity (Alonso-Cantabrana et al., 2018; Danila et al., 2021). Given this, the activation energy for gbs has been left blank in the spreadsheet by default; however, if the activation energy for gbs is known, users can easily add this to allow gbs to be calculated for the measurement temperature using a simple Arrhenius function. However, we do not recommend that bundle sheath conductance is fitted using our routine. Mesophyll conductance (gm), the ease of CO2 diffusion from the intercellular airspace to the mesophyll cytosol, is an important parameter that affects the estimate of Vpmax. We have used gm=1 mol m−2 s−1 bar−1 here, together with the temperature dependence of gm measured for S. viridis (Ubierna et al., 2017); however, a variety of values have also been reported (Barbour et al., 2016; Ermakova et al., 2021b; Osborn et al., 2016; Pathare et al., 2022). Estimating mesophyll conductance in C4 species is difficult and requires measuring C18O16O isotope discrimination. If reliable measurements of in vitro maximum PEPC activity can be obtained, Equation 18 (52 in von Caemmerer, 2021) can be used to estimate gm. This approach was used by Ubierna et al. (2017). Unless gm has been independently estimated, we recommend using the pre-set value of 1 mol m−2 s−1 bar−1. The mitochondrial day respiration rate (Rd) has been pre-set to 1 in the A/Ci parameters; this value is conventionally used in the absence of an experimentally determined estimate (von Caemmerer, 2021). For C3 species, the Rd and CO2 compensation point can be calculated as the y- and x-intercepts of the linear fit of A measurements at Ci values below 50 µbar; however, for C4 species, this method often provides unrealistic values (Yin et al., 2011a). As such, these values are not included as outputs of the tool. If a more accurate value of Rd is desired, we recommend either determining Rd from fitting A/I curves performed on the same leaf or constraining Rd to a set fraction of the dark respiration rate of the leaf (Sharkey, 2016). Fitting a light response curve It is well recognized that the light response of C4 photosynthesis frequently does not saturate under ambient CO2 partial pressures (Cousins et al., 2006; Leakey et al., 2006). The shape of the light response curve is determined by Equation 19 (34 in von Caemmerer, 2021) which, as for the C3 photosynthetic model, remains empirical. Importantly, the model considers the optimal partitioning of electron transport capacity between the C4 and C3 cycle, as discussed in detail in von Caemmerer and Furbank (1999) (see also: Peisker, 1988; von Caemmerer, 2000). Here the partitioning has been set at x=0.4 (Table 1). The model now also better considers the contributions of Jcyc for accommodating the ATP demands of C4 photosynthesis. Using the model, A/I curves of C4 leaves can be described with the parameters Rd, Jmax (the maximum linear electron transport rate), and θ (the empirical curvature factor). The fitting of the A/I curve uses Equations 12–15 (40–43 in von Caemmerer, 2021). To fit an A/I curve, one first updates leaf temperature, atmospheric pressure, O2 concentration, and α in the spreadsheet (see details in ‘Fitting a CO2 response curve’). A, Ci, and irradiance are then entered. Pressing the calculate button will solve for the empirical curvature factor (θ), Rd, and Jmax values which minimize the sum of squares between experimental and modelled data. After converging on a solution, the θ, Rd, and Jmax will be the output for the modelled electron transport-limited CO2 assimilation rate at the measurement temperature. The maximum electron transport rates through Jcyc (Jmax cyc) and PSI (Jmax 1) are also determined from Jmax and fcyc. These electron transport rates are also calculated at 25 °C and at the optimum temperature of electron transport (Topt) to facilitate reporting (Table 1). The quantum yield of CO2 assimilation (ϕCO2) and the light compensation point (Γlight; the irradiance at which A is zero) are calculated as the slope and x-intercept of the linear fit of measurements at irradiances below 200 µmol m−2 s−1. Important A/I fitting considerations Along with the CO2 assimilation rate, stomatal conductance also varies with irradiance, although at a much slower pace. This can lead to large variations in Ci when measuring light response curves. Incorporating Ci into the model calculations helps to improve the fitting routine. However, minimizing variability in Ci over the course of an A/I measurement is preferable. To achieve this, some thought should be given to how quickly irradiance is changed to obtain a light response curve (McAusland et al., 2016). While we advise that measurements are performed on light-adapted leaves, it is important to consider whether light response curves are measured with increasing or decreasing irradiances, as this can impact parameter estimates. Measuring curves with increasing irradiances may result in lower CO2 assimilation rates due to enzyme deactivation at lower irradiances (Arce Cubas et al., 2022; Doncaster et al., 1989; Leegood and Furbank, 1984). Performing A/I curves with decreasing irradiance may help ensure enzyme activation but, in cases where plants are high light sensitive, slow relaxation of non-photochemical quenching (NPQ) or photoinhibition may lead to lower A at lower irradiances, therefore resulting in underestimated Jmax and ϕCO2. While slow relaxation of NPQ could be mitigated by allowing more time at each irradiance step, photoinhibition typically results in a long-term decrease of photosynthetic efficiency (Malnoë, 2018). Examples of A/I measurements performed on S. viridis with increasing or decreasing irradiances are provided in Fig. 4. Fitting of the curve measured with increasing irradiance demonstrates θ values that are inconsistent between biological replicates and are lower than those previously reported, but Jmax values that approach those expected (Massad et al., 2007; Sonawane et al., 2018). Comparatively, fitting of curves measured with decreasing irradiances provides θ values that are more consistent between biological replicates and are similar to those previously reported, but lower than expected Jmax values (further discussed below) (Massad et al., 2007; Sonawane et al., 2018). Fitting of light response curves of CO2 assimilation (A) measured with increasing and decreasing irradiances. Data from S. viridis leaves for increasing irradiance were originally reported in Woodford et al. (2025). Data from S. viridis leaves for decreasing irradiance are reported here. Dotted lines represent the fit of modelled data to the measured data points. For increasing irradiance, Rd=0.48 µmol m−2 s−1, Jmax=166.10 µmol m−2 s−1, θ=0.44, ϕCO2=0.057, and Γlight=21.70 µmol m−2 s−1. For decreasing irradiance, Rd=0.68 µmol m−2 s−1, Jmax=141.43 µmol m−2 s−1, θ=0.91, ϕCO2=0.065, and Γlight=26.72 µmol m−2 s−1. As with A/Ci measurements, determining the maximum electron transport rate and disentangling electron transport and Rubisco limitations can be difficult. To resolve this problem, we recommend fitting A/I curves using measurements made under ambient or higher Ci at irradiances at which electron transport is expected to be limiting (i.e. <1500 µmol m−2 s−1). Using the fitting tool shows that S. viridis lines with Rieske FeS overexpression have increased Jmax values, suggesting that an increase of assimilation seen in A/Ci at high Ci is due to increased electron transport capacity in the transgenic plants (Figs 2, 3). The Jmax values reported by our tool may be lower than previous literature values due to the suggested increase in fcyc (Fig. 5). To complement this change, we have also included a parameter for the rate of maximum electron transport through Jcyc which is determined from Jmax and fcyc using Equation 20. Following the derivations by von Caemmerer (2021), it is possible to sum Jmax and Jmax cyc using Equation 21 to obtain a maximum electron transport rate through PSI (Jmax 1) (Fig. 5B). The proportion of electron transport occurring through cyclic electron flow (fcyc) influences the output parameters of light response curve (A/I) fitting. (A) Fitting S. viridis A/I curve measured with decreasing irradiance (from Fig. 4) with varying fcyc values. (B) The output values for the maximum electron transport rates through linear electron flow (Jmax), cyclic electron flow (Jmax cyc), and PSI (Jmax 1) obtained by fitting data from (A) with varying fcyc values. (C) The output values for the curvature factor (θ) obtained by fitting data from (A) with varying fcyc values. The values for θ determined by our tool are influenced by the choice of fcyc (Fig. 5C). While θ values >0.7 have been previously reported for both C3 and C4 A/I curves (Evans, 1989; Sonawane et al., 2018), the θ values derived by our tool may be <0.7. This is because both θ and fcyc act as curvature factors that influence the shape of the fit and are generally inversely proportional (Fig. 5C). As a consequence, the potential interaction between θ and fcyc should also be considered when interpreting outputs. It is possible that changes in θ could be reflective of changes in Jcyc; however, the complexity of the relationship between θ and fcyc requires further investigation. Importantly, the relationship between J and irradiance remains empirical and should not be overinterpreted. Conclusion The fitting of A/Ci and A/I curves to modelled data provides insight into the underlying biochemistry of C­4 photosynthesis. By keeping fitting routines as simple as possible, our fitting tool provides streamlined, meaningful biochemical characterization of A/Ci and A/I curves performed on C4 leaves. This should simplify reporting of the C4 gas exchange measurements and facilitate comparison between different experiments and species, allowing users to explore the variation in biochemistry of C4 photosynthesis. Combining A/Ci and A/I curves with estimates of Rubisco and Cyt b6f abundance and/or activity could help resolve limitations at high light and high Ci. The fitted electron transport rate through PSII can be compared with estimates of electron transport rate from chlorophyll fluorescence, now frequently measured concurrently with gas exchange in many portable gas exchange systems. Supplementary Material Contributor Information Russell Woodford, School of Biological Sciences, Monash University, Melbourne, Victoria 3800, Australia; Centre of Excellence for Translational Photosynthesis, Division of Plant Science, Research School of Biology, The Australian National University, Acton, ACT 2600, Australia. Maria Ermakova, School of Biological Sciences, Monash University, Melbourne, Victoria 3800, Australia; Centre of Excellence for Translational Photosynthesis, Division of Plant Science, Research School of Biology, The Australian National University, Acton, ACT 2600, Australia. Robert T Furbank, Centre of Excellence for Translational Photosynthesis, Division of Plant Science, Research School of Biology, The Australian National University, Acton, ACT 2600, Australia. Susanne von Caemmerer, Centre of Excellence for Translational Photosynthesis, Division of Plant Science, Research School of Biology, The Australian National University, Acton, ACT 2600, Australia. Alistair Rogers, Lawrence Berkeley National Laboratory, USA. Supplementary data The following supplementary data are available at JXB online. Protocol S1. Excel fitting tool for A/Ci and A/I curves measured on C4 leaves. Protocol S2. Equations for calculating the electron transport rate, J, required to sustain the CO2 assimilation rate for A/Ci measurements. Author contributions SvC, RW, and ME: designed the research; ME, SvC, and RTF: supervised the research; RW: updated model parameterization and developed the Excel fitting tool; RW, SvC, and ME: wrote the paper; all authors discussed the results and contributed to the final manuscript. Funding This work was supported by a Research Scholarship from the Grains Research and Development Corporation to RW (UMO2501-001RSX) and the Australian Research Council’s Discovery Project to ME and RTF (DP230100175). Data availability Data presented in this manuscript were originally reported in Ermakova et al. (2019) and Woodford et al. (2025). The fitting tool described in this work is available in Supplementary Protocol S1. References Associated Data Supplementary Materials Data Availability Statement Data presented in this manuscript were originally reported in Ermakova et al. (2019) and Woodford et al. (2025). The fitting tool described in this work is available in Supplementary Protocol S1.