Short-term elevated temperature and CO₂ promote photosynthetic induction in the C₃ plant Glycine max, but not in the C₄ plant Amaranthus tricolor

Keywords: C₄ photosynthesis, climate change, dynamic photosynthesis, fluctuating light, lightfleck, Rubisco activase, soybean, stomatal conductance.

Atmospheric CO₂ concentration has increased by half since the Industrial Revolution, and the annual global mean temperature has also risen by more than 0.6°C since 1950 (IPCC 2014). A large volume of studies have reported that elevated CO₂ concentration (eCO₂) enhances photosynthetic capacity (Drake et al. 1997; Ainsworth and Long 2005; Ainsworth and Rogers 2007), whereas the effects of elevated temperature (eT) on photosynthetic capacity are complex, depending on both the initial leaf temperature and the degree of warming (Dusenge et al. 2019). However, the stress of eCO₂ and warming does not occur in isolation. Under climate change, the two stresses occur simultaneously (Norby and Luo 2004; Luo et al. 2008). Photosynthetic responses to the two stresses may be synergistic in some circumstances, but antagonistic in others (Dieleman et al. 2012; Smith and Dukes 2013). To project photosynthesis in a future of climate change, studies on the interactive effects of eT and eCO₂ on photosynthesis are urgently needed (Xu et al. 2013, 2014).

Most studies on photosynthesis were carried out under constant light conditions, where light was well controlled. However, in natural environments, light is constantly fluctuating, leading to fluctuations in operating leaf photosynthesis (Chazdon and Pearcy 1986; Pearcy 1990; Pearcy et al. 1996; Tang 1997). Investigating photosynthesis in fluctuating light; i.e. dynamic photosynthesis, will improve our understanding of photosynthesis in natural environments.

There have been some studies addressing the effects of eT or eCO₂ on dynamic photosynthesis in C₃ plants. Below the temperature optimum point, short-term eT can promote photosynthetic induction (the time course of photosynthetic rate in response to a sudden increase in light intensity) by reducing the biochemical limitation of RuBP (ribulose-1,5-bisphosphate) regeneration and Rubisco activation (Kaiser et al. 2015). In contrast, elevating temperature above the temperature optimum point may inhibit photosynthetic induction by depressing Rubisco activation (Leakey et al. 2003; Kang et al. 2020). Short-term eCO₂ can promote photosynthetic induction by reducing the biochemical limitation of Rubisco activation and stomatal limitation (Tomimatsu and Tang 2016; Kaiser et al. 2017a, 2017b; Tomimatsu et al. 2019; Kang et al. 2021).

As far as we know, there has been no study addressing the effects of the interaction of eT and eCO₂ on dynamic photosynthesis. Elevating temperature decreases the solubility of CO₂ in water but eCO₂ can compensate this factor (Jordan and Ogren 1984; Foyer et al. 2009). Another source of uncertainty regarding the interaction of eT and eCO₂ results from the inconsistent results about the effects of eT and eCO₂ on stomatal limitation in previous studies (Naumburg et al. 2001; Leakey et al. 2002, 2003; Tomimatsu and Tang 2012; Kaiser et al. 2017a). Elevating CO₂ reduces stomatal limitation, whereas eT could increase or reduce stomatal limitation (Kaiser et al. 2015, 2017a; Wachendorf and Küppers 2017b), probably due to the trade-off between carbon gain and evaporative cooling (Moore et al. 2021). Elevating temperature generally increases post-illumination CO₂ fixation and post-illumination CO₂ burst at low and medium leaf temperature (Sun et al. 1999; Foyer et al. 2009), but eCO₂ decreases post-illumination CO₂ burst by inhibiting photorespiration (Leakey et al. 2002). Therefore, studies are needed to address the uncertainty of the effects of the interaction of eT and eCO₂ on dynamic photosynthesis.

About 3% of the Earth’s angiosperm species utilise the C₄ photosynthetic pathway, yet C₄ plants contribute about 25% of the net terrestrial primary productivity on Earth (Sage 2004). There have been many studies on the steady-state photosynthesis in C₄ plants (Furbank et al. 1990; Sage and Kubien 2007; Sage and Zhu 2011; Sage et al. 2012; Long and Spence 2013; von Caemmerer 2021), but rare on their dynamic photosynthesis (Furbank and Walker 1985; Horton and Neufeld 1998; Pignon et al. 2021; Wang et al. 2021). Kubásek et al. (2013) have shown that C₄ plants lost more carbon in fluctuating light than in steady light compared with C₃ plants, which they assumed was a result of an increase of bundle sheath cells (BSC) leakiness to CO₂ in C₄ plants in fluctuating light. However, Stitt and Zhu (2014) proposed that the two-cell Kranz system of C₄ photosynthesis ensures large pools of metabolites that could drive diffusion between BSCs and mesophyll cells (MC) for buffering ATP and NADPH, which may provide powerful protection against fluctuating light intensities. Recently, Li et al. (2021) have proposed that the accumulation and diffusion of metabolites in C₄ plants from MCs to BSCs take more time, which may delay the photosynthetic induction. Therefore, the differences between C₃ and C₄ plants in their dynamic photosynthesis are still under debate.

In this study, we characterised photosynthetic responses to the simulated changes of light intensity in a C₃ species and a C₄ species under four different treatments, aiming to address: (1) the effects of the interaction of short-term eT and eCO₂ on dynamic photosynthesis; and (2) the differences between C₃ and C₄ plants in their dynamic photosynthetic responses to eT and eCO₂. We hypothesised that: (1) the short-term eT and eCO₂ would promote photosynthetic induction and enhance carbon gain; and (2) the promotion of photosynthetic induction and the enhancement of carbon gain were greater for C₃ plants than for C₄ plants.

The seeds of the C₃ species, Glycine max L. and the C₄ species, Amaranthus tricolor L. (NAD-ME subtype) were sown in pots (the circular radius was 5 cm and the height was 14 cm) filled with composite soil. Each of the two species was grown within a growth chamber (E-36L1, Percival, Perry, Iowa, USA). After germination, seedlings were thinned. The plants were watered with distilled water regularly and supplied with 5 mL full concentration formula nutrient solution (nitrogen 17 g/L, phosphorus anhydride 17 g/L, potassium oxide 17 g/L, organic matter 25 g/L, and amino acid 12 g/L) every 3–5 days. These ensure that all plants can obtain sufficient nutrients and water, which helps reduce the degree of growth limitation (Poorter et al. 2012). Growth environments was kept constant and consistent throughout the experiment in both growth chambers. Photosynthetic photon flux density (PPFD) incident on the top of both canopies was about 600 μmol m⁻² s⁻¹. The photoperiod was 14 h. Day/night air temperature was set to 28/22°C, [CO₂] was the same as ambient CO₂ concentration and relative humidity was maintained at 70%.

Leaf gas exchange was measured on plants 25 days after germination using a Li-6800 portable photosynthesis system (LI-COR Biosciences, Lincoln, NE, USA) equipped with a Li-6800-01 fluorometer (90% red and 10% blue) on the most recently fully expanded leaves (n = 4). The measured plants were moved to another growth chamber 60 min before the start of the measurement to acclimate to the target temperature and [CO₂] in advance. The measured leaves were enclosed in the Li-6800 leaf chamber, and were first acclimated in the chamber to a PPFD of 100 μmol m⁻² s⁻¹ until steady-state net assimilation rate (A) and stomatal conductance for H₂O (g_sw) were visibly reached, after which PPFD was raised to 600 μmol m⁻² s⁻¹ and kept steady for 60 min. Then, PPFD was decreased to 100 μmol m⁻² s⁻¹ until A reached steady states again. Gas exchange parameters, including A, g_sw, intercellular CO₂ concentration (C_i), and transpiration rate (E), were logged every second. To avoid any swinging from correctional changes in temperature or relative humidity, the temperature of the heat exchanger (T_exchg) was controlled. All measurements were repeated under four different combinations of temperature and CO₂ concentration: (1) 400 ppm × 28°C, denoted as CT; (2) 400 ppm × 33°C, denoted as CT+; (3) 800 ppm × 28°C, denoted as C+T; and (4) 800 ppm × 33°C, denoted as C+T+. The vapour pressure difference between leaf and ambient air ranged from 0.9 kPa to 1.1 kPa at 28°C and from 1.1 kPa to 1.5 kPa at 33°C.

Steady-state A, g_sw, and C_i reached at each PPFD level were calculated by averaging the single values over the last minute of each period; T_50%A and T_90%A were defined as the time required to reach 50% and 90% of the differences between A₁₀₀ and A₆₀₀. We calculated the induction state (IS) after Chazdon and Pearcy (1986):

where A(t) is the transient A at time t.

For C₃ plants, under Rubisco limitation, A(t) can be calculated after Farquhar et al. (1980):

where V_c(t) and C_i(t) is V_c and C_i at time t, respectively. K_m is the Michaelis–Menten constants of Rubisco and Γ* is the CO₂ compensation point, they were both taken from Bernacchi et al. (2003). R_L is mitochondrial respiration rate in the light and assumed to be 40% of dark respiration rate between 25 and 35°C (Way et al. 2019).

The time courses of V_c(t) were then fitted to the model proposed by Woodrow and Mott (1989), and we use V_c(t) to replace A(t) after correction of C_i(t).

where V_c,ini and V_c,max are the initial and maximum V_c(t) after correction to 25°C (Bernacchi et al. 2003), respectively; τ_R is the apparent time constant of Rubisco activation.

The achieved carbon gain (ACG) and ideal carbon gain (ICG) were calculated after Kang et al. (2021):

where t₀ is the time when PPFD was increased. The ratio of ACG to ICG is photosynthetic induction efficiency (IE), which is calculated after Yanhong et al. (1994):

Eqn 4 indicates that ACG during photosynthetic induction can be decomposed into ICG, which is only influenced by steady-state A; and IE, which is influenced mostly by the time course of photosynthetic induction. Short-term eT and eCO₂ can influence ACG via changes in ICG and/or those in IE. We assumed that changes in IE reflected the responses of photosynthetic induction to different treatments per se.

Post-illumination carbon gain (PICG) due to the simulated lightfleck was calculated as follows:

where t₁ is the time when PPFD was decreased, A_100post is the steady-state A at the end of post-illumination period.

Intrinsic water use efficiency (iWUE) was calculated by dividing the transient A by the transient g_s (Farquhar and Sharkey 1982):

To determine the effects of measurement temperature and CO₂ concentration ([CO₂]) on gas exchange parameters for two species, when the requirement of the normality and homogeneity of variances were met, we used two-way ANOVA with temperature and [CO₂] as the main factors and temperature × [CO₂] as interaction, and Duncan test was used for post hoc multiple comparisons. When the requirement of the normality and homogeneity of variances were not met, we used a Kruskal–Wallis test to perform the same analysis. All tests were conducted using SPSS Statistics ver. 18.0 (IBM Corp., Armonk, NY, USA) and R ver. 3.6.1.

In G. max, the mean A₆₀₀ was increased by 20.0% and 61.5% under the CT+ and C+T treatments compared to the CT treatment. However, the mean A₆₀₀ was increased by 74.8% under the C+T+ treatment, which was less than the sum of the effects of eT and eCO₂ alone (Table 1). In A. tricolor, the mean A₁₀₀ and A₆₀₀ were not significantly affected by eT, eCO₂ or their interaction (Table 1).

Table 1.  Steady-state photosynthetic rate (A), stomatal conductance (g_s), and intercellular CO₂ concentration (C_i) reached under different temperature and [CO₂] treatments in G. max and A. tricolor.

In comparison with CT treatment, the transient A in G. max were higher under the other three treatments, and evident difference was observed in the first minute of photosynthetic induction (Figs 1a, 2a). The transient IS was highest under the CT+ treatment during the first 5 min of induction; after that, IS was much higher under the C+T and C+T+ treatments (Fig. 1c). In A. tricolor, the transient A was almost the same in the first minute between all treatments; after that, the transient A was highest under the C+T+ treatment and still similar in the other three treatments (Figs 1b, 2b). The transient IS was always similar between all treatments (Fig. 1d). Bars for s.e. in Figs 1 and 2 were omitted for visual clarity. In addition, iWUE was higher under the C+T and C+T+ than other two treatments during photosynthetic induction in both species (Fig. S1).

Fig. 1.  Time courses of A (a, b), IS (c, d), g_s (e, f), and C_i (g, h) in G. max (a, c, e, g) and A. tricolor (b, d, f, h) leaves following an increase of PPFD from 100 to 600 μmol photons m⁻² s⁻¹ and then a decrease in light intensity from 600 to 100 μmol photons m⁻² s⁻¹. Values are presented as the means of four biological replicates; i.e. individual plants for each species. Bars for s.e. were omitted for visual clarity. A, photosynthetic rate; IS, induction state; g_s, stomatal conductance; C_i, intercellular CO₂ concentration.

Fig. 2.  Transient A (a, b) and g_s (c, d) in G. max (a, c) and A. tricolor (b, d) leaves during the first minute following an increase of PPFD from 100 to 600 μmol photons m⁻² s⁻¹. Values are presented as the means of four biological replicates; i.e. individual plants for each species. Bars for s.e. were omitted for visual clarity. A, photosynthetic rate; g_s, stomatal conductance.

In comparison with CT treatment, the mean T_50%A, T_90%A, τ_R, T_50%g and T_90%g in G. max were decreased under the other three treatments. However, effects on the mean T_50%A, T_90%A and τ_R under C+T+ treatment was less than the sum of the effects of eT and eCO₂ alone (Fig. 3a, b), while effects on the mean T_50%g and T_90%g under C+T+ treatment was smaller than the effects of eCO₂ alone (Fig. 3c, d). In A. tricolor, the mean T_50%A and T_90%A did not differ significantly between treatments (Fig. 3a, b). The mean T_50%g and T_90%g still had no difference between the CT, CT+ and C+T treatments, but significant higher under the C+T+ treatment (Fig. 3c, d).

Fig. 3.  The rates of photosynthetic induction and stomatal opening in G. max and A. tricolor leaves. (a) Time required for the photosynthetic rate to reach 50% of A₆₀₀ (T_50%A). (b) Time required for the photosynthetic rate to reach 90% of A₆₀₀ (T_90%A). (c) Time required for stomatal conductance to reach 50% of g_s600 (T_50%g). (d) Time required for stomatal conductance to reach 90% of g_s600 (T_90%g). Bars and vertical lines indicate the means and s.e. of four biological replicates; i.e. individual plants for each species, respectively. Different letters above error bars indicate significant differences between two treatments within each species. The absence of letters denotes the absence of significant difference.

In G. max, ACG_60 min was increased by 58.3% and 111.9% under the CT+ and C+T treatments compared to the CT treatment. However, ACG_60 min was increased by 136.4% under the C+T+ treatment, which was less than the sum of the effects of eT and eCO₂ alone (Fig. 4a). In A. tricolor, ACG_60 min under the CT treatment was also less than the other three treatments, but the promotion effect on the C+T+ treatment was larger than the sum of the effects of eT and eCO₂ alone (Fig. 4b). In comparison with CT treatment, PICG in G. max and A. tricolor were higher under the other three treatments, but the effects on the C+T+ treatment was less than the effects of eT alone (Fig. 4c, d).

Fig. 4.  Achieved carbon gain during a time period of 60 min following an increase in light intensity from 100 to 600 μmol photons m⁻² s⁻¹ (a, b) and during the post-illumination period (c, d) in G. max (a, c) and A. tricolor (b, d) leaves. Values are means of four biological replicates; i.e. individual plants for each species. Blue-filled bars lines indicate the effect of eT and eCO₂, whereas red-filled and open bars with red dotted frames indicate the interaction of eT and eCO₂. Numbers indicate the extent of the effect of eT and eCO₂ on ACG, taking ACG_60 min under the CT treatment as the base value.

In G. max, ACG and IE were higher under the other three treatments than those under the CT treatment (Fig. 5a) and ACG were always highest under the C+T+ treatment at different time points (Fig. 5c). In A. tricolor, IE did not differ significantly between all treatments at any time (Fig. 5b, d), despite there were changes in ACG_60 min in A. tricolor (Fig. 4b).

Fig. 5.  Induction efficiency (a, b) and achieved carbon gain (c, d) in G. max (a, c) and A. tricolor (b, d) leaves. Bars and vertical lines indicate the means and s.e. of four biological replicates; i.e. individual plants for each species, respectively. Different letters above error bars indicate significant differences between treatments within each species. The absence of letters denotes the absence of significant difference.

To further assess the contribution of ACG and IE on increasing ACG_60 min, we estimated the potential ICG_60 min by assuming that eT and eCO₂ had no effect on photosynthetic induction (equivalent to no changes in IE). In comparison with CT treatment, the increase in ICG_60 min enhanced ACG_60 min by 16.9%, 55.3% and 73.1% in G. max and also enhanced ACG_60 min by 27.7%, 11.4% and 64.5% in A. tricolor under the other three treatments (Fig. 6a, b). Then, we assumed no effects of eT and eCO₂ on steady-state A (equivalent to no changes in ICG) to assess the separate contribution of IE. Compared to the CT treatment, the increase in IE_60 min increased ACG_60 min by 31.5%, 31.4% and 34.3% in G. max under the other three treatments, but decreased ACG_60 min by 6.5%, 5.4% and 3.5% in A. tricolor (Fig. 6a, b). In summary, ACG_60 min in G. max was affected by both IE_60 min and ICG_60 min. eT had a similar effect on IE_60 min as eCO₂ but a lower effect on ICG_60 min than eCO₂ (Fig. 6a). However, ACG_60 min in A. tricolor was only affected by ICG_60 min and the effect of eT on it was higher than that of eCO₂ (Fig. 6b).

Fig. 6.  The effects of eT and eCO₂ on ACG_60 min via changes in photosynthetic induction and steady-state photosynthesis in G. max (a) and A. tricolor (b) leaves. The ACG under the CT treatment was calculated by integrating A over a time period of 60 min following an increase in light intensity from 100 to 600 μmol m⁻² s⁻¹. Carbon gains under other hypothetical conditions were calculated by multiplying IE_60 min and ICG_60 min measured under different three treatments, as the subscripts indicate. Values are means of four biological replicates; i.e. individual plants for each species. Blue bars indicate the effect of changing ICG and red bars indicate the effect of changing IE. Numbers indicate the percentage changes in ACG_60 min, taking ACG_60 min under the CT treatment as the base value.

We assessed the correlations between steady-state and induction parameters using Pearson’s coefficients. In G. max, ACG_60 min was not only positively correlated with A₁₀₀ (P < 0.001), A₆₀₀ (P < 0.001) and IE_60 min (P < 0.001), but also negatively related to T_50%A (P < 0.001), T_90%A (P < 0.001) and T_50%g (P < 0.01) (Fig. 7a). But in A. tricolor, ACG_60 min was only positively correlated with A₆₀₀ (P < 0.001); not related to T_50%A, T_90%A, T_50%g and IE_60 min (Fig. 7b).

Fig. 7.  The correlations between photosynthetic traits measured in G. max (a) and in A. tricolor (b). The numbers in the lower triangle of each matrix are the Pearson’s correlation coefficients for each pair of parameters, while the sizes of the circles in the upper triangle of each matrix represent the size of the correlation coefficient. *P < 0.05; **P < 0.01; and ***P < 0.001. The numbers and the circles in red indicate negative correlations whereas those in blue indicate positive correlations.

In G. max, both eT and eCO₂ promoted photosynthetic induction, but their effects were differential: eT alone imposed influences on the early stage (the first 5 min) of induction, whereas eCO₂ alone imposed significant influences on the late stage of induction (Fig. 1a, c). It is reported that among 193 genes related to photosynthesis in the KEGG database, expression of only nine genes changed significantly after sudden increases in irradiation received by rice (Oryza sativa L.) eaves (Adachi et al. 2019). This result indicates that few genes are involved in photosynthetic responses to sudden changes in light. We focused on the physiological responses of photosynthetic induction to short-term eT and eCO₂.

Photosynthetic induction at the early stage is widely assumed to be limited by the time lags in biochemical processes, especially RuBP regeneration and Rubisco activation (Pearcy et al. 1996; Tomimatsu and Tang 2016). We found moderately eT decrease the T_50%A and τ_R in this study (Fig. 3a; Table 2). Kaiser et al. (2015) reported a parabolic relationship between photosynthesis induction rate and temperature, with the fastest induction occurring at about 30°C. Such a parabolic relationship would be related to the activation rate of Rca (Rubisco activase) on Rubisco (CarmoSilva and Salvucci 2011). Elevating CO₂ reduces biochemical limitation by accelerating the Rubisco activation (a smaller τ_R), which may be ascribed to CO₂-stimulated Rca upregulation (Zhao et al. 2019). However, in this study, eT had a greater impact on the transient IS than eCO₂ in the first 5 min (Fig. 1c). At a longer timescale (∼60 min), photosynthetic induction is mainly limited by diffusional limitation (Way and Pearcy 2012; Kaiser et al. 2015; Lawson and Vialet-Chabrand 2019). Diffusional limitation can be alleviated more rapidly by higher initial g_s and faster stomata opening (McAusland et al. 2016; Wachendorf and Küppers 2017a). However, effects of eT and eCO₂ on g_s and stomata opening rate are conflicting between studies; increases (von Caemmerer and Evans 2015; Urban et al. 2017), decreases, (Sage and Sharkey 1987) or no changes (von Caemmerer and Evans 2015) of g_s under eT have been reported before. Elevating CO₂ generally reduces g_s but its effect on stomatal opening can be positive (Naumburg et al. 2001; Leakey et al. 2002; Kaiser et al. 2017a) or negative (Tomimatsu and Tang 2012). In this study, both eT and eCO₂ reduced diffusional limitation by accelerating stomata opening (Fig. 3c), without significant influences on the initial g_s (Table 1).

Table 2.  The apparent time constant of Rubisco activation (τ_R), the ratio of the initial and maximum V_c, and the induction state (IS) under different temperature and [CO₂] treatments in G. max and A. tricolor.

Both eT and eCO₂ enhanced carbon gain in fluctuating light by improving photosynthetic induction efficiency and photosynthetic capacity in G. max. The enhancement of ACG under eT was mainly attributable to the improved IE, while that under eCO₂ was mainly attributable to the improved photosynthetic capacity. The effects of eCO₂ were consistent with Kang et al. (2021), where ICG had a larger effect than IE on ACG during induction in wheat (Triticum aestivum L.) and rice.

In A. tricolor, photosynthetic induction was not significantly affected by eT, eCO₂ or their interaction (Table 3, Fig. 3b, d). In particular, during the first minute of photosynthetic induction, the time course of transient A under each of the four treatments almost overlapped in A. tricolor (Fig. 2b), indicating that factors insensitive to both temperature and [CO₂] are dominating photosynthesis during this time. This result is consistent with the report on photosynthetic induction at various temperatures and [CO₂] in maize (Ireland et al. 1984). Rubisco activity and stomatal conductance (Fig. 2d) change little during this period (Kaiser et al. 2015; Tomimatsu and Tang 2016), thus we propose that the enhancement of photosynthetic rate is associated with the build-up of metabolite concentration gradients. In C₄ photosynthesis, atmospheric CO₂ is first assimilated as C₄ acids in a C₄ cycle before entering the Calvin-Benson cycle, or referred to as C₃ cycle (Bräutigam and Weber 2011). The C₄ and C₃ cycle is coordinated by intracellular transport of C₄ acids and C₃ metabolites (Furbank et al. 2000), or light regulation of key enzymes (Furbank et al. 1997). Light regulation of the key enzymes takes minutes and should impose relatively small limitation on photosynthesis during the initial stage of the induction (Usuda et al. 1984). Intracellular transport of C₄ acids and C₃ metabolites is widely believed to occur by symplastic diffusion and require metabolite concentration gradients between the compartments (Sowiński et al. 2008). Isotopic labeling experiments have revealed the establishment of metabolite pools and the concentration gradients during induction (Moore and Edwards 1986a, 1986b). C₄ species have low levels of RuBP and its precursors even at high light (Borghi et al. 2022), but maintain high levels of C₄ acids at both low and high light (Moore and Edwards 1986a). Such large pools of C₄ acids facilitate a fast build-up of sufficient metabolite concentration gradients (Wang et al. 2021) and thereby provide additional abilities to buffer the redox and energy status against fluctuating environments (Stitt and Zhu 2014).

Table 3.  The influences of eT and eCO₂ on the differences in the photosynthetic characteristics of G. max and A. tricolor.

Different from G. max, the increases of ACG in A. tricolor were almost independent of the changes of IE, but were mainly attributable to the enhancement of ideal carbon gain (ICG) (Fig. 6b). The enhancement of ICG in A. tricolor was less than G. max. The transient iWUE of C₄ plants were not significantly different from those of C₃ plants under C+T+ treatment (see Supplementary Fig. S1). These findings suggest C₃ plants will benefit more from the simultaneous eT and eCO₂ under the background of future climate change than C₄ plants.

The effects of simultaneously eT and eCO₂ on T_50%A, T_90%A, ACG_60 min and IE_60 min in G. max were lower than the sum of the sole effect of eT and eCO₂, indicating that the effects of eT and eCO₂ on photosynthetic induction were partially offset. We hypothesised that this finding was a result of an offset in the effects of eT and eCO₂ on stomatal behaviour. Elevating CO₂ decreased the g_s600 but promoted the rate of the increases in g_s during induction, both of which shortened the time required for stomatal opening (Fig. 1e). The effects of eCO₂ on stomatal behaviour observed in this study were consistent with previous reports (Kaiser et al. 2017b). In contrast, eT increased the g_s600 and promoted the rate of the increases in g_s during induction; yet the effect of the latter dominated over that of the former. We were not clear on the mechanism that underlies the increases of g_s600 under eT. An increases of the guard cell metabolic activity or of the evaporative demand under eT could drive the increases of g_s600. The interaction of eT and eCO₂ had no influences on T_50%A, T_90%A and IE_60 min in A. tricolor.

The effects of simultaneously eT and eCO₂ on A₆₀₀ in G. max were in close proximity to the sum of the sole effect of eT and eCO₂ and the effects of simultaneously eT and eCO₂ on A₆₀₀ in A. tricolor are higher than the sum of the sole effect of eT and eCO₂. These findings are different from some studies, where eT inhibited the positive effect of eCO₂ on steady-state photosynthetic rate and photosynthetic efficiency (Lambreva et al. 2005; Cai et al. 2016). In contrast, Sage and Kubien (2003) found that the effects of eCO₂ on C₃ and C₄ photosynthesis were greater at warmer than at cooler temperatures. Some studies have also found the enhancement of photosynthesis by eCO₂ is larger at higher temperature (Long 1991; Morison and Lawlor 1999). That is because eCO₂ suppresses photorespiration and mitochondrial respiration in C₃ plants, expanding the photosynthetic thermal optimum range (Long 1991; Way et al. 2015).

In general, the steady state photosynthetic rate of C₄ plants is higher than that of C₃ plants. A lower A₆₀₀ in A. tricolor than that in G. max in this study may result from the low growth light intensity. Due to biochemical and energetic requirement (Furbank et al. 1990; Ubierna et al. 2011), C₄ plants are more suitable for growing in high light. Photosynthetic rate was reduced to a greater extent in low light in the six C₄ grasses relative to the two C₃ species, and C₄ grasses also tended to have a lower stomatal conductance and stomatal aperture than C₃ species (Israel et al. 2022).

However, C₄ plants grown in low light or medium light still have a large metabolites pool because BSC leakiness was found to be similar for C₄ plants grown in different light intensity (Pengelly et al. 2010; Bellasio and Griffiths 2014; Ma et al. 2017). Above researches suggest that the CCM is still robust and the biochemical efficiency of the C₄ cycle does not decrease for C₄ plants grown in low light or medium light. Therefore, the low growth light intensity may have little influences on the effects of eT, eCO₂, and their interaction on photosynthetic induction in A. tricolor observed in our study.

By examining dynamic photosynthesis under four different temperature and [CO₂] treatments, this study showed that for G.max, the T_50%A and T_90%A were significantly affected by eT and eCO₂; whereas for A. tricolor, they were almost unaffected by eT or eCO₂. This study suggests that the effects of eT and eCO₂ on photosynthetic induction were partially offset in C₃ plants and greater enhancement of photosynthesis in fluctuating light for C₃ plants than for C₄ plants in a warming and CO₂-enriched future. More research is needed to address how the interaction of eT and eCO₂ influences dynamic photosynthesis in future experiments.

Supplementary material is available online.

The original contributions presented in the study are included in the article and Supplementary material, further enquiries can be directed to the corresponding author.

Author has no conflicts of interest.

This study was supported by the Key Research of Plant Functional Ecology Program of Peking University (No. 7101302307), National Natural Science Foundation of China (Grant No. 41530533), and Key Laboratory of College of Urban and Environmental Sciences (No. 7100602014).