Elevated CO2 Enhances Dynamic Photosynthesis in Rice and Wheat

Crops developed under elevated carbon dioxide (eCO2) exhibit enhanced leaf photosynthesis under steady states. However, little is known about the effect of eCO2 on dynamic photosynthesis and the relative contribution of the short-term (substrate) and long-term (acclimation) effects of eCO2. We grew an Oryza sativa japonica cultivar and a Triticum aestivum cultivar under 400 μmol CO2 mol−1 air (ambient, A) and 600 μmol CO2 mol−1 air (elevated, E). Regardless of growth [CO2], the photosynthetic responses to the sudden increase and decrease in light intensity were characterized under 400 (a) or 600 μmol CO2 mol−1 air (e). The Aa1, Ae2, Ea3, and Ee4 treatments were employed to quantify the acclimation effect (Ae vs. Ee and Aa vs. Ea) and substrate effect (Aa vs. Ae and Ea vs. Ee). In comparison with the Aa treatment, both the steady-state photosynthetic rate (PN) and induction state (IS) were higher under the Ae and Ee treatments but lower under the Ea treatment in both species. However, IS reached at the 60 sec after the increase in light intensity, the time required for photosynthetic induction, and induction efficiency under Ae and Ee treatment did not differ significantly from those under Aa treatment. The substrate effect increased the accumulative carbon gain (ACG) during photosynthetic induction by 45.5% in rice and by 39.3% in wheat, whereas the acclimation effect decreased the ACG by 18.3% in rice but increased it by 7.5% in wheat. Thus, eCO2, either during growth or at measurement, enhances the dynamic photosynthetic carbon gain in both crop species. This indicates that photosynthetic carbon loss due to an induction limitation may be reduced in the future, under a high-CO2 world.

Plasticity in photosynthesis in response to long-term eCO 2 can be attributed to the short-term (substrate) effect, long-term (acclimation) effect, or both. The substrate effect is related to increased CO 2 supply, which is reported to enhance dynamic photosynthesis greatly (Tomimatsu and Tang, 2012;Tomimatsu et al., 2014Tomimatsu et al., , 2019Kaiser et al., 2017b). The acclimation effect is related to variations in leaf morphological, anatomical, and biochemical traits, but little information is known about its role and relative contribution.
The activation of Calvin-Benson cycle enzymes and stomatal opening regulate the photosynthetic response to a sudden increase in light intensity (Way and Pearcy, 2012;Kaiser et al., 2015) and are affected by variations in leaf chemical and morphological traits. Leaves exposed to long-term eCO 2 have a lower Rubisco activase (Rca) content (Geiger et al., 1999;Aranjuelo et al., 2011;Tomimatsu et al., 2019) and develop small stomata (Maherali et al., 2002;Zhu et al., 2018;Zheng et al., 2019). The rate of Rubisco activation is proportional to Rca content (Woodrow and Mott, 1989;Yamori et al., 2012). Stomatal morphology affects the stomatal opening kinetics in some species (Drake et al., 2013;Kardiman and Raebild, 2018;Zhang et al., 2019), although contradictory evidence has also been reported (Elliott-Kingston et al., 2016;McAusland et al., 2016). Whether and how acclimation to eCO 2 affects dynamic photosynthesis remains untested.
All studies listed in Table 1 have investigated woody species, except for one study on C 4 grass (Knapp et al., 1994). Rice and wheat are important for the global population as direct sources of food (FAO, 2016); yet, there are no available reports on the effect of long-term eCO 2 on dynamic photosynthesis in rice, wheat, or any other C 3 crop plants. The light environments for crop plants are characterized by long periods of sunlight punctuated by shadeflecks (Pearcy et al., 1990), whereas those for withincanopy woody plants are characterized by long periods of diffuse light punctuated by sunflecks (Tang et al., 1988;Chazdon and Pearcy, 1991). The slow photosynthetic induction in wheat costs at least 21% of its daily potential assimilation (Taylor and Long, 2017). Thus, it is of great importance to investigate whether acclimation to eCO 2 improves dynamic photosynthesis in C 3 crop plants.
In this study, we aimed to address: (1) how eCO 2 affects dynamic photosynthesis; (2) the relative contribution of the acclimation and substrate effects of eCO 2 on dynamic photosynthesis.

Plant Materials and Growth Conditions
The study was conducted at the Agricultural Meteorology and Ecology Experimental Station, Nanjing University of Information Engineering, Nanjing, China (32 • 16 ′ N, 118 • 46 ′ E).
An Oryza sativa japonica cultivar (Nanjing 9108) and a Triticum aestivum cultivar (Yangmai 22) were used. There are many cultivars of rice and wheat since both are widely cultivated crop species with a very long cultivation history. Nanjing 9108 was nominated as "super rice" by the Chinese Ministry of Agriculture in 2015, whereas Yangmai 22 is a high-yield wheat cultivar. Currently, both are the major and most widely cultivated cultivars in the middle and lower reaches of the Yangtze River. The rice seeds were sown (rice, May 2017; wheat, October 2017) at 400 µmol CO 2 mol −1 air. Afterward, the rice seedlings were transplanted into octagonal open-top chambers (OTCs) with 400 (ambient, denoted by A) or 600 µmol CO 2 mol −1 air (elevated, denoted by E). The wheat seeds were directly sown in the soil in each OTC. The OTCs were 3 m high with a bottom area of ∼12 m 2 . Ventilation fans were installed on the inner walls of the OTCs to minimize the impact of heterogeneous temperature and CO 2 concentration on the height of the flag leaves. The plants grown within the core area of an ambient [CO 2 ] OTC and an elevated [CO 2 ] OTC were selected for the experiment. The soil was carefully plowed before transplanting the rice seedlings and sowing the wheat seeds. These practices allowed for fairly homogeneous soil conditions for the two CO 2 treatments (OTCs), which were only 8-10 m apart from each other. The soil nutrition content at a 0-20cm depth before sowing the wheat seeds was 1.25 vs. 1.34 g N kg −1 , .84 vs. .83 g P kg −1 , and 18.1 vs. 18.08 g K kg −1 for the ambient vs. elevated OTC, respectively. Atmospheric environments during the growth period were also similar between the two OTCs (Supplementary Figure 1). Furthermore, all plants in the two treatments received the same management practices, e.g., fertilization, irrigation, and pest control. All these conditions ensured that the major difference between the two treatments was growth CO 2 concentration. We then focused on the biological replication, i.e., the individual plants, for photosynthetic measurement replicates. The measurements were conducted during grain filling: rice, September 15-28, 2017; wheat, April 20-27, 2018.

Gas Exchange Measurement
In both species, flag leaves contribute an important portion of assimilates used for grain filling (Yoshida, 1981;Carmo-Silva et al., 2017). We then decided on all the measurements to be conducted with the flag leaves. Gas exchange parameters were measured on the south-facing, fully expanded flag leaves with portable infrared gas analyzers (Li-Cor 6400 and Li-Cor 6800, LI-COR Biosciences, Lincoln, NE, USA). Air temperature during grain filling was slightly higher for rice than for wheat; therefore, the block temperature was set to 32.5 • C for rice and 30 • C for wheat. Relative humidity was maintained at 60-65%. Measurements were made on three to four leaves from different plants (one leaf per plant).
To determine the responses of the net photosynthetic rate to the intercellular CO 2 concentration (P N -C i curves, equivalent to A-C i curves as in common usage), leaves were acclimated to a saturating light intensity of 1,500 µmol photons m −2 s −1 at a CO 2 concentration identical to growth [CO 2 ]. Then, the CO 2 concentration in the reference cell was varied from 50 to 1,500 µmol CO 2 mol −1 air. Altogether, 14 different CO 2 concentrations were investigated.
To characterize dynamic photosynthesis, we measured the photosynthetic time course in response to a simulated increase and decrease in light intensity. The leaves were first acclimated at 100 µmol photons m −2 s −1 for over 30 min, then the light intensity within the leaf chamber was increased to 1,500 µmol photons m −2 s −1 , and kept constant for 5 min. Afterward, the light intensity was decreased to 50 µmol photons m −2 s −1 and kept constant for 5 min. The net photosynthetic rate and stomatal conductance (g s ) were recorded every second for the entire measurement period. Photosynthetic characteristics in plants grown at 600 µmol CO 2 mol −1 air were influenced not only by the short-term but also the long-term effect of eCO 2 , in comparison with plants grown at 400 µmol CO 2 mol −1 air. To distinguish both effects, we measured dynamic photosynthesis at 400 (denoted by a) and 600 µmol CO 2 mol −1 air (denoted by e), regardless of growth [CO 2 ]. The differences between the two growth conditions [CO 2 ] under the same measurement [CO 2 ] (i.e., Ae vs. Ee and Aa vs. Ea) were considered to be caused by the acclimation effect of eCO 2 , while the differences between the two measurements [CO 2 ] from the same growth [CO 2 ] (i.e., Aa vs. Ae and Ea vs. Ee) were considered to be due to the instantaneous substrate effect of eCO 2 .

Carbon and Nitrogen Analysis
The leaves used for the gas exchange measurements were collected, brought to the laboratory, and then oven-dried at 65 • C for 48 h before grinding. The carbon and nitrogen concentrations of ground samples were determined using a CHNOS elemental analyzer (Vario EL III, Elementar Analysensysteme GmbH, Langenselbold, Germany).

Stomatal Anatomy
Five flag leaves other than those used for the gas exchange measurements from the same OTC were sampled and kept in a 2.5% glutaraldehyde solution. Images of the center of the abaxial leaf surface at ×800 magnification for rice and ×400 for wheat were acquired with a scanning electron microscope (FEI Quanta 200F, Thermo Fisher Scientific, Waltham, MA, USA). Stomatal density and guard cell length were determined using the ImageJ software (National Institute of Health, Bethesda, MD, USA) from 10-20 fields of view.

Data Analysis
To evaluate the rates of photosynthetic and stomatal induction responses, we calculated the time to reach 50 and 90% of the steady-state photosynthetic rate (P N ) (T P50% and T P90% ) and g s (T g 50% and T g 90% ) at 1,500 µmol photons m −2 s −1 . Photosynthetic induction state (IS) was calculated after Chazdon and Pearcy (1986): where P N,100 is the steady-state P N reached at 100 µmol photons m −2 s −1 and P N,1500 is the steady-state P N reached at 1,500 µmol photons m −2 s −1 . Both were calculated by averaging single values over the last half-minute of each period. We assumed that the transition between the Rubisco-limited state and the ribulose 1,5-bisphosphate (RuBP) regenerationlimited state occurs at a C i of 200-400 µmol CO 2 mol −1 air at ambient growth [CO 2 ] and a C i of 400-600 µmol CO 2 mol −1 air at elevated growth [CO 2 ] during the induction response. Then, P N -C i curves were divided into two parts and fitted separately to obtain the apparent maximum Rubisco carboxylation rate (V c,max ), electron transport rate at 1,500 µmol photons m −2 s −1 (J 1500 ), and CO 2 photo compensation point (Γ * ). Parameters estimated from the P N -C i curves were then used to correct transient P N by removing stomatal limitations. Since RuBP regeneration limitation typically relaxes within 2 min (Taylor and Long, 2017) and triose phosphate use limitation barely occurs at the CO 2 concentration investigated in this study (Long and Bernacchi, 2003), the P N was corrected only 2 min after the increase in light intensity. Under Rubisco limitation, P N at time t can be calculated after Farquhar et al. (1980): where V c,max (t) and C i (t) is V c,max, and C i at time t, respectively, K m is the Michaelis-Menten constants of Rubisco taken from Bernacchi et al. (2001), and R L is mitochondrial respiration rate in the light and assumed to be 60% of dark respiration rate (Way et al., 2019), which was determined in a preliminary test. To remove the stomatal limitation, P N was recalculated by replacing C i (t) with the final, steady-state C i (C i,f ): where P * N (t) is corrected P N (t). Combining Equations (1, 2) to eliminate the unknown variable V c,max (t): The time courses of P * N were then fitted to the model proposed by Woodrow and Mott (1989): where P * N,f and P * N,i are the final corrected and estimated initial photosynthetic rate, respectively; τ R is the apparent time constant of Rubisco activation. Diffusional and biochemical limitations were estimated after Kaiser et al. (2017a).
Accumulative carbon gain (ACG) during a period of time was calculated as where t 0 is the time when light intensity was increased. The eCO 2 may affect the steady-state P N and photosynthetic Frontiers in Plant Science | www.frontiersin.org induction, both of which result in changes in transient P N during photosynthetic induction and, thus, affect ACG (Figure 1). To distinguish the relative contribution of eCO 2 to the ACG via accelerated photosynthetic induction from the enhanced steadystate P N , we decomposed ACG into ideal carbon gain (ICG) and induction efficiency (IE). The ICG during the same period of time was defined as As Equation (7) shows, ICG is determined by steady-state P N only. Thus, IE is calculated by dividing the ACG by the ICG, after Tang et al. (1994): Induction efficiency is linearly correlated with T P90% (Supplementary Figure 2); thus, the accelerated photosynthetic induction will result in a higher IE. The relative contribution of eCO 2 on ACG via accelerated photosynthetic induction is represented as the percentage difference in ACG under the Aa treatment and estimated using IE under the Ae and Ee treatments. Additionally, post-illumination carbon gain (PICG) due to the simulated lightfleck was calculated as, where t 1 is the time when light intensity was decreased.

Statistical Analysis
A two-way ANOVA was performed to evaluate the contribution of the acclimation and substrate effects and their interaction. When the requirement of the normality and homogeneity of variances were met, the differences in the means of different treatments (Aa, Ae, Ea, and Ee) were then assessed by a least significant difference test. All statistical analyses were carried out with SPSS Statistics Version 26 for Windows (IBM Corp., Armonk, NY, USA) at a significance level of 0.05.

Steady-State Photosynthesis Under Different CO 2 Treatments
In both species, the P N,100 and P N,1,500 in leaves from both growth [CO 2 ] treatments were higher when the measurements were conducted under eCO 2 ( Table 2). The averaged P N,1,500 under the Ee treatment was significantly higher than that under the Aa treatment by 30.5% in rice and 37.8% in wheat. The averaged P N,50 did not differ significantly across treatments in rice. In wheat, P N,50 was significantly lower under the Ea treatment compared with the other three treatments. In both species, g s,1,500 did not differ significantly across treatments ( Table 2). In rice, FIGURE 1 | An example illustrating the substrate effect of elevated carbon dioxide (eCO 2 ) (A) and the acclimation effect of eCO 2 (B) on accumulative carbon gain (ACG) via changes in steady-state photosynthetic rates (P N ) and photosynthetic induction. In this study, we assume that the acclimation to long-term eCO 2 accelerates photosynthetic induction but decreases the steady-state P N at high light intensity. All lines were modeled with an exponential function. (A) The thin solid curve represents the time course of P N following an increase in light intensity from 100 to 1,500 µmol photons m −2 s −1 under the Aa treatment, while the thick solid curve represents that under Ae treatment. The dashed blue curve represents the time course of P N considering accelerated photosynthetic induction only, i.e., the rate of photosynthetic induction is the same as that under the Ae treatment. The dotted blue curve represents the time course of P N considering changes in steady-state P N only i.e., steady-state P N are the same as that under the Ae treatment. (B) The solid red curve represents the time course of P N under the Ee treatment. The dashed red curve represents the time course of P N considering changes in photosynthetic induction only, i.e., the rate of photosynthetic induction is the same as that under the Ee treatment. The dotted red curve represents the time course of P N considering changes in steady-state P N only i.e., the final, steady-state P N is the same as that under the Ee treatment. g s,100 did not differ significantly across treatments, however, g s,50 was higher under the Ae and Ea than under the Ee and Aa treatments. In wheat, both g s,100 and g s,50 were higher under the Aa and Ee than under the Ae and Ea treatments. The averaged C i,50 , C i,100, and C i,1,500 were higher when measurements were conducted under eCO 2 than under ambient CO 2 in both species.
In rice, the differences in P N,100 and P N,1,500 were attributable to measurement [CO 2 ] only, whereas the differences in g s,100 and g s,1,500 were attributable to the interaction of growth and measurement [CO 2 ] only. The differences in g s,1,500 were also, to a less extent, attributable to measurement [CO 2 ]. In wheat, the differences in P N and g s were attributable to measurement [CO 2 ]; the differences in P N,100 and g s,100 were also attributable to the interaction of growth and measurement [CO 2 ].

Dynamic Photosynthesis Under Different CO 2 Treatments
In both species, P N increased faster when measurements were conducted under eCO 2 after the sudden increase in light intensity (Figures 2A,B). During the photosynthetic induction and postillumination period, transient P N was higher under the Ae than under the Ee treatment in rice but was higher under the Ee than under the Ae treatment in wheat. Transient IS was highest under the Ee treatment and lowest under the Ea treatment in both species (Figures 2C,D). In rice, the IS was similar between the Aa and Ae treatments, whereas, in wheat, the IS was surprisingly higher under the Aa than under the Ae treatment over the first 5 min of photosynthetic induction. In rice, transient g s was higher under the Ae and Ea than under the Aa and Ee treatments ( Figure 2E). In wheat, transient g s was higher under the Aa than under the Ae treatment but was similar between the Ea and Ee treatments ( Figure 2F). During the first 10 min of photosynthetic induction, transient C i was higher under the Ee than under the Ae treatment in both species (Figures 2G,H). Transient C i was higher under the Ea than under the Aa treatment in rice but was higher under the Aa than under the Ea treatment in wheat. During the post-illumination period, transient C i was similar across treatments in both species. In both species, the averaged T P50% and T P90% were lower under the Ee than under the Aa and Ae treatments, despite that such differences were insignificant (Figures 3A,B). In wheat plants grown at an ambient CO 2 concentration, T g 50% was significantly higher under the Ae than under the Aa treatment, whereas T g 90% was higher under the Aa than the Ae treatment (Figures 3C,D). In wheat plants grown at an eCO 2 concentration, both T g 50% and T g 90% were higher under the Ea than under the Ee treatment. However, T g 50% and T g 90% were similar in rice leaves from different growth conditions [CO 2 ] or between different measurements [CO 2 ] from the same growth [CO 2 ]. No significant differences in the IS 60s values across treatments were found in both species (Figure 3E). In both species, τ R was shorter when measurements were conducted under eCO 2 ( Figure 3F). In wheat, the differences in T P90% were attributable to measurement [CO 2 ] and the interaction of growth and measurement [CO 2 ] ( Table 3). The differences in T g 50% in both species were attributable to the interaction of growth and measurement [CO 2 ] only. In both species, the diffusional limitation was similar between the Aa and the Ea treatment, i.e., the limitation showed minimal significant differences in leaves from different growth conditions [CO 2 ] when measurements were conducted under ambient CO 2 . However, the diffusional limitation was lower under the Ee than under the Ae treatment. The biochemical limitation tended to be lower when measurements were conducted under eCO 2 , regardless of growth [CO 2 ] (Supplementary Figure 3). The percentage of diffusional limitation among the total photosynthetic limitation was lower than 10% during most of the time of the photosynthetic induction period (Supplementary Figures 3A,B). Biochemical limitation dominated over diffusional limitation for the first 3 min of induction in both species.

ACG and IE
In comparison with the Aa treatment, ACG 30min was higher by 45.5 (Ae) and 27.2% (Ee) in rice ( Figure 4A) and by 39.3 (Ae) and 46.8% (Ee) in wheat ( Figure 4B). However, ACG 30min was lower under the Ea than the Aa treatment by 10.9% in rice and by 4.7% in wheat. In comparison with the Aa treatment, PICG was higher by 45.4 (Ae) and 67% (Ee) in rice ( Figure 4C) and by 4.7 (Ae) and 4.1% (Ee) in wheat ( Figure 4D). In rice, PICG was higher under the Ea than the Aa treatment by 29.4% in rice but was lower by 26.3% in wheat.
Despite these changes in ACG 30min , the IE 30min did not differ significantly among the Aa, Ae, and Ee treatments in both species (Figures 5A,B). In both species, IE 30min was lowest under the Ea treatment. There were no significant differences observed in IE between the Aa and the Ae treatments in both species. In wheat, the averaged IE was highest under the Ee treatment throughout photosynthetic induction.
To further evaluate the contribution of accelerated photosynthetic induction under the Ae and Ee treatments FIGURE 3 | The rates of photosynthetic induction and stomatal opening in rice and wheat leaves. (A) Time required for the photosynthetic rate to reach 50% of P N,1,500 (T P 50% ). (B) Time required for the photosynthetic rate to reach 90% of P N,1,500 (T P 90% ). (C) Time required for stomatal conductance to reach 50% of g s,1,500 (T g50% ). (D) Time required for stomatal conductance to reach 90% of g s,1,500 (T g90% ). (E) The IS reached 30 s following an increase in light intensity (IS 60s ). (F) The apparent time constant of Rubisco activation (τ R ). Bars and vertical lines indicate the means and standard errors of three to 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.
to ACG 30min , we estimated the potential ACG 30min , assuming no effect of eCO 2 on steady-state P N , which was equivalent to no changes in ICG. Compared with the Aa treatment, the increase in IE 30min under Ae alone increased ACG 30min by 0.6% in rice and by 0.9% in wheat, whereas the increase in IE 30min under Ee alone increased ACG 30min by 0.4% in rice and by 3.3% in wheat (Figures 6A,B). The changes in steady-state P N under Ae alone increased ACG 30min by 44.5% in rice and 38.1% in wheat, whereas the changes in steady-state photosynthesis under Ae alone increased ACG 30min by 26.3% in rice and 42.1% in wheat.

Rubisco Carboxylation Capacity, Stomatal Anatomy, and Element Analysis
The Rubisco carboxylation capacity (V c,max ) was lower at elevated than at ambient growth [CO 2 ]; the difference was significant in wheat (Supplementary Table 1). Regardless of growth [CO 2 ], V c,max was higher in wheat than in rice. There were no significant differences in stomatal density at the two growth conditions [CO 2 ] (Supplementary Table 1). Guard cells were shorter at elevated than at ambient growth [CO 2 ]; the difference was significant in wheat. No significant differences in carbon or nitrogen content were found between ambient and elevated growth [CO 2 ] (Supplementary Table 1).

Increased CO 2 Supply Enhances Leaf Dynamic Photosynthesis Without Accelerating Photosynthetic Induction in Crops
Several studies reported accelerated photosynthetic induction in woody species at eCO 2 (Leakey et al., 2002;Holišova et al., 2012;Tomimatsu and Tang, 2012). In this study, the differences in T P between the Aa and Ae treatments were small in two crop species (Figures 3A,B), and the differences in T P50% were not related to different measurement [CO 2 ]. The modeling analysis also indicates that the changes in IE 30min under the Ae treatment alone increased ACG 30min by <1% in both species (Figure 6). On the contrary, the changes in steady-state P N alone increased ACG 30min by more than onethird in both species (Figure 6). Acevedo-Siaca et al. (2020)  Shown are Wald χ 2 statistics followed by significance symbols, which are * P < 0.05, ** P < 0.01, and *** P < 0.001, respectively.
reported that, in rice, increases in the photosynthetic carbon gain in fluctuating light are smaller compared with those in steady state. If the contribution of improved transient P N at eCO 2 on carbon gain were excluded in their study, the increased CO 2 supply may play a limited role in accelerating photosynthetic induction in crops. Thus, an increased CO 2 supply improves crop photosynthesis by improving steadystate and transient P N (Table 2, Figures 2A,B), rather than accelerating photosynthetic induction.
There are no relevant investigations on why an increased CO 2 supply imposes limited influences on the rate of photosynthetic induction in crop species, however, we hypothesized that dumbbell-shaped stomata, which respond to light faster than kidney-shaped stomata (Franks and Farquhar, 2007;McAusland et al., 2016;Harrison et al., 2020), lower the diffusion limitation imposed on photosynthetic induction in the crop species. Therefore, crop species may exhibit little changes in the rates of photosynthetic induction when there is a minimal increase in CO 2 supply. The difference in measurement [CO 2 ] in our study (200 µmol CO 2 mol −1 air) was smaller than that in the studies reporting significant effects of eCO 2 on photosynthetic induction ( Table 1).

Differential Effect of Elevated CO 2 on Leaf Carbon Gain in Rice and Wheat
The effect of acclimation to eCO 2 has rarely been distinguished from the effect of increased CO 2 supply because the measurement [CO 2 ] was the same as the growth [CO 2 ] in previous studies. In comparison with the Aa treatment, the acclimation effect alone resulted in a 18.3% decrease in ACG 30min in rice but a 7.5% increase in ACG 30min in wheat (Figures 4A,B), suggesting the differential influences of the acclimation effect on the photosynthetic carbon gain between the two crop species. Such differences are likely to be related to the difference in the acclimation effect on P N,1,500 , because the high IE 30min under the Ee treatment led to a smaller increase in ACG 30min than high P N,1,500 , correspondingly ICG 30min , under the Ee treatment in both species (Figure 6). Previous eCO 2 experiments also show that the enhancement of photosynthesis is greater in wheat than in rice (Long et al., 2006). Wheat allocates less leaf nitrogen to Rubisco and a greater catalytic constant of Rubisco carboxylation than rice (Supplementary Table 1), and thus may have greater benefits from eCO 2 (Makino, 2011).
In this study, IS 60s , T P90%, and IE did not differ significantly between Ee and Ae, or even between the Ee and Aa treatments (Figures 2C,D, 3B, 5), suggesting limited influence of acclimation effect on the rate of photosynthetic induction. At timescales of minutes, the rate of photosynthetic induction is mainly determined by the light-activation of Rubisco and stomatal opening (Way and Pearcy, 2012;Kaiser et al., 2015). To speed up the rate of Rubisco activation, more resources should be allocated to Rca, which is necessary for Rubisco activation (Yamori et al., 2012). This requirement is unlikely to be met at eCO 2 , as previous studies reported the decreased content of Rca in rice (Chen et al., 2005) and wheat (Zhang et al., 2009;Aranjuelo et al., 2011). This is consistent with the finding that τ R ( Figure 3F) and biochemical limitation (Supplementary Figures 3C,D) did not differ significantly between the Ae and Ee treatments in both species. On the contrary, the differences in T P between the Ae and Ee treatments may be related to the changes in the rate of stomata opening. Accumulating evidence demonstrates that species with fast stomatal opening show fast photosynthetic induction (Drake et al., 2013;Deans et al., 2019;Yamori et al., 2020). In comparison with the Ae treatment, the stomata tended to open faster (Figures 3C,D) and stomatal limitation tended to be lower (Supplementary Figures 3A,B) under the Ee treatment in both species. We hypothesized that the faster stomatal opening in the leaves acclimated to eCO 2 was achieved by a decrease in guard cell length (Supplementary Table 1), which is consistent with previous studies (Maherali et al., 2002;Zhu et al., 2018;Zheng et al., 2019). In general, small dumbbell-shaped stomata open faster than large ones within the same genus (McAusland et al., 2016). However, a negative correlation between stomatal size and T g 50% in the genus Oryza was reported recently (Zhang et al., 2019). More research is needed to address how acclimation to eCO 2 in stomatal morphology affects photosynthesis.

Implications
The effects of eCO 2 on the dynamic photosynthesis in crops have been rarely investigated until recently (Acevedo-Siaca et al., 2020;Ohkubo et al., 2020). This study showed that eCO 2 has limited influences on the rates of photosynthetic induction in two crop species but increased photosynthetic carbon gain greatly by improving steady-state and transient P N . These findings indicate that photosynthetic carbon loss due to induction limitation may be reduced in the future, under a high CO 2 world. We acknowledge that the photosynthetic responses to increasing CO 2 concentration are complex. To project crop photosynthesis and yield in the future, both field experiments and in silico modeling, spanning a wide range of CO 2 concentrations, are urgently needed (Drag et al., 2020). The findings in this study, which clarified the contribution of the substrate and acclimation effects of eCO 2 , imply that dynamic photosynthesis is likely to reduce photosynthetic induction limitation under temporally changing light environments in the future, under a high CO 2 world.
In comparison with wheat, the photosynthetic acclimation to eCO 2 in rice compromised the beneficial effect of an increased CO 2 supply on ACG 30min but further improved PICG ( Figure 4A). Such an increase in PICG may counterbalance the decrease in ACG if rice leaves receive many brief (<1 min) sunflecks. Nonetheless, photosynthetic characteristics vary greatly among rice and wheat accessions (Qu et al., 2017;Salter et al., 2019;Acevedo-Siaca et al., 2020), and the effects of eCO 2 on photosynthesis are likely to differ between them. The results presented here do not necessarily suggest an advantageous position for wheat in the future. Instead, this study may provide a potential reason for the lower enhancement of yield in rice than in wheat at eCO 2 (Long et al., 2006), though detailed assessments are needed because of the large variations in photosynthetic light utilization among them.

CONCLUSIONS
By examining dynamic photosynthesis under four different CO 2 treatments, this study showed that neither an increased CO 2 supply nor an acclimation to eCO 2 imposes large influences on the rates of photosynthetic induction in two crop species. But, an increased CO 2 supply enhances photosynthetic carbon gain FIGURE 5 | Induction efficiency (IE) in rice (A) and wheat (B) leaves. Bars and vertical lines indicate the means and SEs of three to four biological replicates, i.e., individual plants for each species, respectively. Different letters above error bars indicate significant differences across treatments within each species. The absence of letters denotes the absence of significant difference.
greatly via improving steady-state P N . The acclimation effect of eCO 2 may compromise, or slightly strengthen, the beneficial effect of an increased CO 2 supply, depending on the species and the fluctuations in light intensity. Our study suggests that the photosynthetic carbon gain in the two crop species is likely to be enhanced in a CO 2 -enriched future when photosynthetic induction limitation becomes significant for leaf carbon gain.

DATA AVAILABILITY STATEMENT
The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding author/s.

AUTHOR CONTRIBUTIONS
YT and YH conceived and designed the experiment. WS and ZH provided the experimental materials. HK, TZ, YZ, XK, and HS conducted the experiment and collected data. HK, TZ, and YZ analyzed the data and drafted the manuscript. All authors contributed to the editing and revising of the final version of the manuscript.  The ACG under the Aa treatment was calculated by integrating P N over a time period of 30 min following an increase in light intensity from 100 to 1,500 µmol photons m −2 s −1 . Carbon gains under other hypothetical conditions were calculated by multiplying IE 30min and ICG 30min measured under different CO 2 treatments, as the subscripts indicate. IE 30min , IE over the first 30 min of photosynthetic induction; ICG 30min , ideal ACG, calculated as the product of (P N,1,500 -P N,100 ) and 30 min. Values are means of three to four biological replicates, i.e., individual plants for each species. Blue bars indicate the substrate effect of eCO 2 and red bars indicate the sum of the substrate and acclimation effect of eCO 2 . Bars without any patterns indicate the effects of changes in photosynthetic induction and bars with oblique lines indicate the effects of changes in steady-state photosynthesis. Numbers indicate the percentage changes in ACG 30min , taking ACG 30min under the Aa treatment as the base value.