2.1 Plant materials, growth conditions, and experimental setup

The experiment was conducted in a climate chamber (size: 9 m²) located at Wageningen, the Netherlands. The chamber was equipped with dimmable LED modules (DRWFR_RSE 400 V 1.1D MP, Signify, the Netherlands), which produced a diurnal average of ~250 μmol m⁻² s⁻¹ of photosynthetically active radiation (PAR) at plant height. The daily PAR regime followed a sinusoidal pattern (16 h light period, minimum PAR of 120 μmol m⁻² s⁻¹, maximum PAR of 1200 μmol m⁻² s⁻¹) with random drops of PAR that mimicked natural irradiance fluctuations (Supporting Information S1: Figure S1). The general sinusoidal pattern and daily light sum stayed the same during the whole experiment, whereas the timing and extent of random PAR decreases changed daily. Day and night temperatures were set to 23 and 20°C respectively, and relative air humidity (RH) was set to 70% (realized temperature and RH values were given in Supporting Information S1: Figure S2).

Two CO₂ treatments were used: 450 μmol mol⁻¹ (aCO₂) and 900 μmol mol⁻¹ (eCO₂). At the start of each CO₂ treatment, tomato (cv. ‘Merlice’; Bayer, NL) and cucumber (cv. ‘HiPower’; BASF Nunhems, NL) seeds were sown in rockwool plugs (diameter: 2 cm; Grodan, NL). Chrysanthemum (cv. ‘Anastasia’; Deliflor, NL) plants that had been grown in plastic pots (~1 L, filled with potting soil) were cut back at the third or fourth node to allow for the formation of new axillary buds. One week after sowing, tomato and cucumber seedlings were transplanted to rockwool cubes (10 × 10 × 7 cm; Grodan), and treatment was started. Plants were treated for a total of 6 weeks, and all measurements were conducted in the last 2 weeks of treatment. The experiment was conducted in four consecutive runs, where each CO₂ treatment was applied twice. Plants were irrigated twice per day (at 7:00 and 19:00 h) with nutrient solution using an ebb and flow system (Supporting Information S1: Table S1).

2.2 Photosynthetic gas exchange measurements

A and g_s were measured on the youngest, fully expanded leaf using a gas exchange system (LI-6800, Li-Cor Bioscience) equipped with a 6 cm² leaf chamber fluorometer. Measurements were performed at an air temperature of 23°C, relative humidity of 70%, and a flow rate of air through the system of 500 μmol s⁻¹. Irradiance was provided by a mixture of red (90%) and blue (10%) LEDs in the fluorometer. Different types of measurements (see below) were randomized to avoid artefacts due to diurnal effects, except that measurements under the “fluctuating irradiance regime” (see below) were always performed in the morning. Once photosynthetic induction was measured on a given plant, this plant was not used for another measurement for at least 25 min, to avoid interference with previous conditions. All gas exchange measurements were conducted between 9:00 and 17:00 h.

2.3 Stomatal anatomy

Stomatal imprints were taken after the final gas exchange measurement on a given plant had been completed. Before taking imprints, plants were dark-acclimated at least for 20 min to allow for stomatal closure to minimize errors in guard cell width measurements caused by different levels of stomatal opening (Franks & Farquhar, 2007). Imprints were taken using a silicone impression material (Zhermack) with two technical replicates per leaf surface. The silicon was allowed to fully dry on the leaf before being gently removed. Clear nail polish was applied to the imprint and allowed to dry. The dry nail polish was viewed under a microscope (Leitz Aristoplan, Leica Microsystems) and photographed with a PL FLUOTAR 25X objective in chrysanthemum and tomato (NA 0.60) and 40X objective in cucumber (NA 0.70) (Axiocam 305 Color). Images were analysed with ImageJ, using the CellCounter and ObjectJ plugins (National Institute of Health). Length and width of individual stomata were measured, and the number of stomata in the field of view was counted.

2.4 Measurements of individual stomatal pore opening and closure

Individual stomatal pore aperture kinetics under stepwise changes in irradiance were measured on the same leaf used for gas exchange measurements using a customized microscope placed inside the growth chamber (van den Berg et al., 2024). A green background light of 75 μmol m⁻² s⁻¹ was applied during the whole measurement for imaging. The leaf was firstly acclimated to an actinic light of 50 μmol m⁻² s⁻¹ (50% red, 50% blue) for 30 min. Then, imaging was started with an actinic light of 50 μmol m⁻² s⁻¹ for 30 min, after which the actinic light was increased to 1000 μmol m⁻² s⁻¹ with a single step that lasted 60 min, followed by a decrease of actinic light intensity to 50 μmol m⁻² s⁻¹ for 30 min. Chrysanthemum and tomato leaves were measured with a 20X objective (Mitutoyo plan apo infinity corrected long working distance objective, NA 0.42), and cucumber leaves were measured with a 50X objective (Mitutoyo plan apo HR infinity corrected objective, NA 0.75). Measurements were conducted once a minute, and at each measurement, image stacks of 90−100 images were recorded that were spaced apart 1 µm on the z-axis parallel to the imaging plane for chrysanthemum and tomato and 0.5 μm for cucumber. Image stacks were made to image the entire leaf surface within the field of view in focus at a high resolution (van den Berg et al., 2024).

2.5 Calculations

2.6 Statistical analysis

Differences between treatments were detected using one-way ANOVA (p < 0.05), taking experimental run as a block effect. Normality and homogeneity of variances were tested using Shapiro-Wilk and Levene's test, respectively. In case data were not normally distributed or inhomogeneous, log transformation was performed. When a significant difference was detected, a post hoc test was conducted for pairwise comparisons between treatments using Fisher's Protected Least Significant Difference (LSD) test (p < 0.05). Statistical analyses were conducted using R (http://www.r-ptoject.org/).

3.1 Effects of acclimation to eCO₂ on dynamic A

Chrysanthemum, tomato, and cucumber showed different acclimation responses in photosynthetic induction characteristics (Figures 1a–c and 2). Chrysanthemum grown under eCO₂ displayed smaller time constants—times needed to reach 20%, 50%, and 90% full photosynthetic induction (T_20, T₅₀ and T₉₀)—compared with plants grown under aCO₂, especially when measured at 450 μmol mol⁻¹ CO₂ (Figure 2a,d,g). In contrast, cucumber grown under eCO₂ tended to have larger time constants (especially for T₂₀ and T₅₀) compared with plants grown under aCO₂ (Figure 2c,f,i). Acclimation responses in time constants of tomato were situation dependent: tomato grown under eCO₂ tended to show larger time constants when measured at CO₂ of 450 μmol mol⁻¹, but smaller time constants when measured at 900 μmol mol⁻¹ (Figure 2b,e,h). Overall, eCO₂ acclimation responses in time constants were strongest in chrysanthemum and more subtle in tomato and cucumber (Figure 2).

The time course of V_cmax induction had similar shapes between aCO₂ and eCO₂ plants in all three crops (Supporting Information S1: Figure S5). Absolute values of maximum Rubisco carboxylation rate (V_cmax) during photosynthetic induction, however, were mostly lower in eCO₂ acclimated plants, especially during the final stages of Rubisco activation (Supporting Information S1: Figure S5). Time constants of the fast and slow phase of V_cmax induction (τ_fast and τ_low) were generally smaller in eCO₂ plants, with chrysanthemum showing the strongest response, followed by tomato. The growth CO₂ concentrations hardly affected τ_fast and τ_low in cucumber (Figure 3).

When grown under eCO₂, the net carbon gain during the post-illumination phase was lower in chrysanthemum, but was higher in tomato and cucumber, compared to plants grown under aCO₂ (Figure 4). Chrysanthemum grown under eCO₂ showed smaller values of post-illumination CO₂ fixation, but higher levels of CO₂ burst (Supporting Information S1: Figure S6a,d). In contrast, tomato and cucumber generally showed higher levels of post-illumination CO₂ fixation and reduced levels of CO₂ burst when grown under eCO₂ (Supporting Information S1: Figure S6b,c,e,f).

3.2 Stomatal conductance kinetics and anatomical traits

Acclimation responses of stomatal conductance (g_s) kinetics after a stepwise increase and decrease in irradiance varied between the three crops (Figure 1d–f). Values of steady-state g_s of tomato and cucumber were hardly affected by eCO₂ acclimation, except that initial g_s of tomato under low irradiance (g_s,i) increased in eCO₂ plants (Figure 5b). g_s kinetics, however, were clearly affected: time constants of g_s increase and decrease upon irradiance changes (k_i and k_d) were significantly larger in eCO₂ plants compared to aCO₂ plants, for k_i of tomato and k_d of cucumber (Figure 5h,k,i,l). In chrysanthemum, neither steady-state nor dynamic g_s were affected by eCO₂ acclimation (Figure 5a,d,g,j).

Final steady-state pore area was not significantly different between eCO₂ and aCO₂ plants (Figure 6c,d). Time constants of pore area increase and decrease upon irradiance changes (k_{i, area} and k_{d, area}) were not significantly different between eCO₂ and aCO₂ plants (Figure 6a,b). Chrysanthemum clearly showed acclimation responses in terms of individual pore movements: there was a strong increase in Sl_i of eCO₂ plants compared to aCO₂ plants (Figure 6e), and there were trends for nonsignificant increases in Sl_d and decreases in k_{i, area} and k_{d, area} in eCO₂ plants compared to aCO₂ plants (Figure 6a,b,f). In contrast, cucumber showed opposite trends for individual stomatal pore movements: both k_{i, area} and k_{d, area} increased, whereas Sl_i and Sl_d decreased, in eCO₂ compared to aCO₂ plants (Figure 6a,b,e,f). Variations between species in terms of time constants for individual stomatal pore movements showed similar trends as g_s kinetics: chrysanthemum had the largest k_{i, area} and the smallest k_{d, area} among the three species (Figure 6a,b).

In leaves acclimated to eCO₂, stomatal density was decreased in all species, whereas responses in stomatal size varied between species. Cucumber showed the strongest reduction in stomatal density (50%−60% depending on leaf surface), whereas stomatal size in cucumber was unaffected (Figure 7; Supporting Information S1: Figure S7). In tomato and chrysanthemum there was a weaker reduction (11%−34%) in stomatal density upon eCO₂, with additional slight decreases in stomatal size (4%−9%) (Figure 7).

3.3 High CO₂ acclimation of steady-state photosynthetic parameters

Chrysanthemum showed the largest degree of acclimation in photosynthetic capacity, with strong reductions in V_cmax, J₁₈₀₀, TPU, and R_d in eCO₂ plants (Figure 8). Tomato also showed reductions V_cmax, J₁₈₀₀, and TPU in eCO₂ plants, albeit to a lesser degree, while cucumber showed no reductions in either of these parameters (Figure 8).

3.4 Net carbon gain under fluctuating irradiance

Short-term exposure to eCO₂ increased A under fluctuating irradiance (see Supporting Information S1: Figure S3 for pattern) in all three crops; although in chrysanthemum leaves that had acclimated to eCO₂, the short-term effect of eCO₂ on time-integrated A was not significant (Figure 9d). For the long-term eCO₂ acclimation effect, A of chrysanthemum under fluctuating irradiance was higher in eCO₂ acclimated plants compared with plants grown under aCO₂, when measured at both 450 and 900 μmol mol⁻¹ CO₂ (Figure 9a). A of tomato and cucumber was generally slightly decreased by eCO₂ acclimation (Figure 9b,c). Nevertheless, integrated assimilation during the 90 min fluctuating irradiance period was not significantly affected by eCO₂ acclimation. Except when measured at 900 μmol mol⁻¹ CO₂, integrated assimilation of chrysanthemum was lower in eCO₂ acclimated plants (Figure 9d–f).

4.1 Photosynthetic acclimation to eCO₂ differed substantially between three horticultural crops

Acclimation of leaf photosynthesis to eCO₂ has received a lot of attention in the past decades in the context of global climate change (Becklin et al., 2021). Most studies have focused on eCO₂ acclimation of leaf photosynthetic capacity (in the steady state), and biochemical and anatomical traits underlying it (Ainsworth & Long, 2005; Ainsworth & Rogers, 2007; Leakey et al., 2009). Upon acclimation to eCO₂, many species reduce leaf photosynthetic capacity as well as g_s, stomatal density and size (e.g. Cai et al., 2018; Tjoelker et al., 1998; Zhu et al., 2018). We found that chrysanthemum, tomato, and cucumber clearly showed eCO₂ acclimation of both dynamic and steady-state photosynthesis, but the three species differed in which traits were affected. Chrysanthemum acclimated to eCO₂ by reducing photosynthetic capacity (e.g. J₁₈₀₀ and V_cmax), whereas cucumber did not (Figure 8). In contrast, cucumber mostly responded to eCO₂ via decreasing steady-state g_s and slowing down g_s kinetics, whereas g_s of chrysanthemum hardly responded to eCO₂ (Figure 5). Tomato did a bit of both, by lowering leaf photosynthetic capacity and changing g_s responses, but these responses were smaller than those of either chrysanthemum (Figure 8) or cucumber (Figure 5).

Compared to the vast amount of research on eCO₂ acclimation of steady-state A, very few studies (Kang et al., 2021; Leakey et al., 2002; Tomimatsu et al., 2019) have focused on acclimation of dynamic A and stomatal traits to eCO₂. In early studies, photosynthetic induction was measured under the same CO₂ concentration as the growth environment, and it was generally found that eCO₂ acclimation reduced the time needed for photosynthetic induction (Košvancová et al., 2009; Leakey et al., 2002; Tomimatsu & Tang, 2012). However, given that the speed of photosynthetic induction typically increases when measurements are conducted at high CO₂ (Kaiser et al., 2015, 2017; Kang et al., 2021), comparisons in many early studies (e.g. on Populus, Shorea leprosula, Fagus sylvatica) have mixed the eCO₂ acclimation effect with its transient effect (Košvancová et al., 2009; Leakey et al., 2002; Tomimatsu & Tang, 2012). Recent studies (e.g. in rice, wheat and Populus) have separated the acclimation versus transient effect of eCO₂ on photosynthetic induction and show that acclimation to eCO₂ hardly changes the speed of photosynthetic induction (Kang et al., 2021; Tomimatsu et al., 2019). We found that in tomato and cucumber, effects of eCO₂ acclimation on photosynthetic induction were mostly nonsignificant, whereas acclimation to eCO₂ in chrysanthemum substantially reduced the time needed for photosynthetic induction, when measured at ambient CO₂ (Figure 2); this could be because in this case, A only needed to reach to a low level at high irradiance (Figure 1a), which may further connect with smaller time constants of V_cmax induction that possibly resulted from a lower steady-state V_cmax (Figures 3 and 7a). Tomato also showed smaller time constants of V_cmax induction in eCO₂ acclimated leaves (Figure 3), but this did not speed up photosynthetic induction, possibly because g_s responses to irradiance changes tended to be slower in eCO₂ acclimated leaves (Figures 2b,e,h and 5h,k). In cucumber, neither the steady-state value nor the induction rate of V_cmax were affected by eCO₂ acclimation (Figures 3 and 7a), however, the speed of photosynthetic induction tended to be slower, possibly caused by slower g_s responses to irradiance changes (Figure 5). The slower g_s responses in eCO₂ acclimated tomato may have resulted in a higher net carbon gain during the post-illumination phase (Figure 4b,c), since photosynthesis could function under a relatively high g_s for a longer period when irradiance dropped to a low level (allowing for a high C_i). Overall, these results suggest that the three important horticultural crops acclimated to eCO₂ with very different strategies.

4.2 All crops exhibited eCO₂ acclimation in stomatal anatomy, but this did not always affect individual pore movement dynamics and leaf-level g_s

Stomatal density and aperture are often reduced when plants grow in high CO₂, often leading to reduced g_s (Chater et al., 2015). We also found significant decreases in stomatal density in all three horticultural crops grown at eCO₂, as well as decreases in stomatal size (except for cucumber; Figure 7). However, acclimation in stomatal anatomy under eCO₂ did not translate into changes in operational g_s in the steady state: values of g_s at both low and high irradiance (initial and final g_s) were hardly affected by eCO₂ acclimation (Figure 5a–f). Nevertheless, theoretical maximum g_s decreased, especially in cucumber and tomato, mostly due to decreases in stomatal density (Supporting Information S1: Figure S8). In rice, genotypic variation in stomatal morphology was found to correlate with steady-state g_s to a small extent only, but more strongly with dynamic g_s responses to an irradiance increase (Xiong et al., 2022). We also found relatively stronger acclimation responses in dynamic compared to steady-state g_s traits (Figure 5). Moreover, it seems that crops with higher stomatal density showed stronger acclimatory responses to eCO₂ in terms of g_s kinetics (Figures 5 and 7).

Several recent studies have highlighted the effects of eCO₂ acclimation on g_s dynamics (De Souza et al., 2020; Wall et al., 2023): eCO₂ acclimation tends to decrease the speed for g_s increase upon an irradiance increase in rice and wheat (Kang et al., 2021; Wall et al., 2023). We found that eCO₂ acclimation increased the time needed for g_s increase (k_i) in tomato (Figure 5h), similarly to other species (Kang et al., 2021; Wall et al., 2023). The slower g_s response in eCO₂ acclimated plants could partly be explained by reductions in stomatal density, requiring individual stomatal pores to open to a larger extent to reach a given g_s (Zhang et al., 2022). With a custom-made microscope that could follow the opening and closure of individual stomata in situ, we quantified eCO₂ acclimation effects on individual pore movement dynamics in the growth environment (van den Berg et al., 2024). We found that eCO₂ acclimation responses of individual pores were quite different from leaf-level g_s: for example, individual pore opening in eCO₂ acclimated Chrysanthemum leaves was faster (Figure 6e,f). However, despite faster responses of individual pores, g_s responses were hardly affected by eCO₂ acclimation in chrysanthemum, possibly due to decreased stomatal density (Figures 5 and 7). Nevertheless, it should be noted that microscope measurements were conducted with a larger boundary layer and less tightly controlled microclimate than in the LI-6800, which may have caused some of the differences observed. We conclude that the apparent g_s response is a complex integrated result from responses of stomatal anatomy at leaf level and pore movement at individual stomata level, and it is necessary to quantify eCO₂ acclimatory responses at different integration levels to understand plant photosynthesis and transpiration under elevated CO₂.

4.3 eCO₂ acclimation only marginally affected A under fluctuating irradiance

Traits pertaining to dynamic photosynthesis are often investigated by measuring photosynthetic induction, using a single-step change from a low to a (much) higher irradiance and keeping that higher irradiance steady until a new steady state is reached. However, under natural irradiance fluctuations, irradiances often change at high frequencies (Durand et al., 2021; Kaiser et al., 2018), which causes photosynthesis to transiently change between non-steady states. In other words: while photosynthesis is busy adjusting to a change in irradiance, it is already exposed to another irradiance and may be exposed to yet another irradiance 0.1 s later. This raises a question: how to extrapolate photosynthetic behavior under fluctuating irradiances from photosynthetic induction measurements? Our results showed that despite the eCO₂ acclimation in dynamic and steady-state photosynthesis discussed above, effects on time-integrated A under fluctuating irradiance were nonsignificant in most cases (Figures 9 and 10). How can these results be explained? We hypothesize that different acclimatory responses affected A in opposite directions, leading to a canceling-out of these effects. For example, we found lower steady-state A in high CO₂ acclimated chrysanthemum and tomato, and slower g_s responses to irradiance changes in high CO₂ acclimated tomato and cucumber; such responses reduced A under fluctuating irradiance in previous studies (Kimura et al., 2020; Ohkubo et al., 2020) (Figure 10). In contrast, faster photosynthetic induction was found in high CO₂ acclimated chrysanthemum, which by itself should increase A under fluctuating irradiance (Adachi et al., 2019), but at the same time steady-state A in chrysanthemum was reduced (Figure 10). An overall higher net carbon gain during the post-illumination phase in eCO₂ tomato and cucumber was found in our study (Figure 4), which should increase A under fluctuating irradiance (Figure 10). By integrating the stomatal morphology with dynamic photosynthesis models that consider relevant biochemical and diffusional processes (e.g. Morales et al., 2018), the relative importance of each individual acclimatory response on dynamic A can be quantified. This could further help to identify promising traits for crop breeding under future eCO₂ conditions. Our results suggest the potential importance of testing dynamic photosynthesis traits estimated from conventional induction measurements under fluctuating irradiance to provide a more accurate evaluation of plant performance under naturally fluctuating light conditions.