A comparison of stomatal conductance responses to blue and red light between C3 and C4 photosynthetic species in three phylogenetically-controlled experiments

Introduction C4 photosynthesis is an adaptation that has independently evolved at least 66 times in angiosperms. C4 plants, unlike their C3 ancestral, have a carbon concentrating mechanism which suppresses photorespiration, often resulting in faster photosynthetic rates, higher yields, and enhanced water use efficiency. Moreover, the presence of C4 photosynthesis greatly alters the relation between CO2 assimilation and stomatal conductance. Previous papers have suggested that the adjustment involves a decrease in stomatal density. Here, we tested if C4 species also have differing stomatal responses to environmental cues, to accommodate the modified CO2 assimilation patterns compared to C3 species. Methods To test this hypothesis, stomatal responses to blue and red-light were analysed in three phylogenetically linked pairs of C3 and C4 species from the Cleomaceae (Gynandropsis and Tarenaya), Flaveria, and Alloteropsis, that use either C3 or C4 photosynthesis. Results The results showed strongly decreased stomatal sensitivity to blue light in C4 dicots, compared to their C3 counterparts, which exhibited significant blue light responses. In contrast, in C3 and C4 subspecies of the monocot A. semialata, the blue light response was observed regardless of photosynthetic type. Further, the quantitative red-light response varied across species, but the presence or absence of a significant stomatal red-light response was not directly associated with differences in photosynthetic pathway. Interestingly, stomatal density and morphology patterns observed across the three comparisons were also not consistent with patterns commonly asserted for C3 and C4 species. Discussion The strongly diminished blue-light sensitivity of stomatal responses in C4 species across two of the comparisons suggests a common C4 feature that may have functional implications. Altogether, the strong prevalence of species-specific effects clearly emphasizes the importance of phylogenetic controls in comparisons between C3 and C4 photosynthetic pathways.


Introduction
Efficient coordination of carbon gain versus water loss by stomata is achieved by a range of responses to environmental factors, most notably light and CO 2 .Light stimulates stomatal opening via at least two distinct signaling pathways with contrasting action spectra (Shimazaki et al., 2007), which are either perceived directly in the guard cells (GCs) or indirectly via signals derived from the underlying mesophyll (Mott, 2009;Lawson et al., 2014).
Blue light (BL) is a potent signal for stomatal opening, and GCs contain all the components required for the blue light signaling pathway (Shimazaki et al., 2007;Suetsugu et al., 2014).Blue light activates rapid stomatal opening via a pathway that is independent of photosynthesis and saturates at relatively low fluence rates.Blue light triggers the autophosphorylation of phototropins, PHOT1 and PHOT2 via their substrate BLUS1 (BLUE LIGHT SIGNALING 1) (Inoue et al., 2008;Takemiya et al., 2013).This sequentially triggers the activation of plasma membrane H + ATPase pump, which hyperpolarizes the plasma membrane, leading to the uptake of K + through inward-rectifying channels.The accumulation of K + and other solutes drives the influx of water into the GCs and the resulting turgor change leads to stomatal opening (Shimazaki et al., 2007).
Illumination with red light (or photosynthetically active wavelengths other than blue) also triggers stomatal opening but the quantum efficiency to stimulate opening is typically lower than blue light.An indirect effect of red light that is tied to intercellular CO 2 (C i ) is triggered by high (red) light through mesophyll photosynthesis, which reduces C i , and consequently affects g s .However, this feedback loop does not appear to be the only mesophyll-derived signal perceived by stomata.A significant response to red light was still observed in experiments where C i was kept constant (Messinger et al., 2006).In addition, epidermal peel experiments revealed that GC responses to red light are reversibly altered by the presence of the underlying mesophyll tissue.Thus, an additional link between mesophyll photosynthesis and guard cell responses may exist (Lawson et al., 2014).The identity of the mesophyll signal however remains unresolved.
Plants with C 4 photosynthesis are distinct from C 3 plants in anatomy and biochemistry.C 4 operates a CO 2 -concentrating mechanism (CCM) between two morphologically different cell types, bundle sheath cells (BSC) and mesophyll (M) cells (Sage, 2004).Having a CCM alters the relationship of photosynthesis with CO 2 in C 4 species which confers them a photosynthetic advantage over C 3 species under saturating light conditions and hot environments (Bowyer and Leegood, 1997).In these environments, C 4 species typically display higher intrinsic water use efficiency and net assimilation rates (Sage, 2004) compared to C 3 species (Lambers and Oliveira, 2019).It can therefore be argued that differential stomatal regulation is required to optimize photosynthetic capacity and water use efficiency between C 3 and C 4 species.And that the mesophyll signal argued above may be different between C 3 and C 4 species.However, not much is known about stomatal sensitivities with regard to photosynthetic types under various light and CO 2 environments.
Comparisons between C 4 and C 3 species can easily be confounded by phylogenetic distance between the species, which would lead to differences not necessarily associated with the photosynthetic pathway (Monson, 1996;Taylor et al., 2010).For instance, contrasting blue light responses in C 3 and C 4 were observed in crop species (Zhen and Bugbee, 2020), but the findings could not be directly linked to photosynthetic type alone because the species examined were separated by considerable phylogenetic distance.Closely related species or even subspecies with differing photosynthetic pathways offer a way around this issue, allowing a comparison between photosynthetic types while controlling for evolutionary distance (Monson, 1996;Huxman and Monson, 2003).
In this study, blue and red-light stomatal responses were studied in three phylogenetically-controlled experiments, using closely related species from the Cleomaceae (Gynandropsis and Tarenaya), Flaveria, and Alloteropsis, that use either C 3 or C 4 photosynthesis.The results show a striking lack of sensitivity to blue light in the C 4s Gynandropsis gynandra, and F. trinervia and F. bidentis, compared to their C 3 counterparts Tarenaya hassleriana, and F. cronquistii and F. robusta, which exhibited significant blue light responses.In contrast, in A. semialata, the blue light response was observed regardless of photosynthetic type.The quantitative red-light response varied between species, however the presence or absence of a significant stomatal red-light response was not related to differences in photosynthetic pathway.

Plant materials
Stomatal responses were compared across closely related C 3 and C 4 species from the Cleomaceae (Cleomaceae), Asteraceae (Flaveria), and Poaceae (Alloteropsis), with considerable phylogenetic distance separating each of the controlled comparisons.From the genus Cleomaceae, Tarenaya hassleriana (C 3 ) was compared with Gynandropsis gynandra (C 4 , NAD-ME subtype).Seeds of C 3 T. hassleriana (Cleomaceae) were obtained from Prof. Julian Hibberd's group (Cambridge, UK).From the genus Flaveria, F. trinervia and F. bidentis (C4-NADP-ME) were initially compared with Flaveria pringlei as the C 3 representative.However, the F pringlei accession that most manuscripts have described and was used here, likely has seen some hybridization with Flaveria angustifolia in its history, which is a C3-C4 intermediate.We therefore decided to do a second set of experiments using F. cronquistii and F.robusta as 'true' C3 species in addition to the three aforementioned species.Seeds of F. bidentis, F. trinervia, and F. pringlei were a kind gift from Dr. Peter Westhoff (University of Dusseldorf, Germany).Clonal plants of F. cronquistii and F.robusta were a gift from Dr. Marjorie Lundgren (Lancaster University, UK) and Prof. Julian Hibberd (Cambridge University, UK), respectively.Finally, from the genus Alloteropsis, A. semialata subsp.eckloniana GMT (C 3 ) and A. semialata subsp.semialata MDG and MAJ (both C 4, mixed NADP-ME-PEPCK subtypes) were a gift from Prof. Pascal-Antoine Christin (University of Sheffield, UK).

Growth conditions
Individual seedlings of G. gynandra and T. hassleriana were grown from seeds and transferred in 10 cm H x 9 cm L x 9 cm W plastic pots with M3 compost (Levington ® , Scotts, Ipswich, UK).Seeds of C 4 F. bidentis, C 4 F. trinervia, and C 3 -like F. pringlei were also germinated on M3 compost.The seedlings and clonal plants of C 3 F. cronquistii and C 3 F.robusta propagated from cuttings were grown in 10 cm H x 9 cm L x 9 cm W plastic pots containing M3 compost.
The Cleomaceae and Flaveria species were grown in an environmentally-controlled growth room at the Plant Growth Facility (PGF), Department of Plant Sciences, University of Cambridge.Photoperiod was kept at 16 hours (6:00 to 22:00) of light supplied by fluorescent lamps with a photosynthetic photon flux density (PPFD) of 300 µmol m -2 s -1 .The growth room was maintained at day and night temperatures of 20°C and 60% RH at ambient CO 2 .
The Alloteropsis accessions were grown at the Cambridge Botanic Gardens Glasshouse.Plants were vegetatively propagated to contain 2-3 tillers per 1 L pots filled with 4:1 M3 compost: coarse vermiculite and 5 g controlled-release fertilizer (Osmocote ® , Scotts Miracle-Gro Company, Marysville, OH, USA).Supplemental light from sodium halide lamps (140-160 W m -2 ) was provided from 4:00 to 20:00.Greenhouse set point temperature was at 18°C, while nighttime temperatures were at 15°C.CO 2 and RH were at ambient.

Gas exchange measurements
An open gas exchange system (LI-6400XT, LI-COR, Lincoln, NE, USA) with an integrated fluorometer and light source (LI-6400-40, LI-COR) was used to measure stomatal conductance (g s ) and CO 2 assimilation rate (A) at a range of different conditions as specified below.
All gas exchange data were collected from the youngest expanded leaf from plants with 4-5 sets of mature leaves.

Red and blue light switchover experiment
To characterize stomatal responses to blue light, a series of switchover protocols were designed comprising sequential changes between 100% red (R) and 75% red + 25% blue light (RB), while keeping the total PPFD equal.Peak wavelengths for R and B lights in the LI-6400XT instruments are 635 and 465 nm, respectively.In Sequence R→RB→R, leaves were first acclimated to steady-state under 100% red light, and then switched to 75% red and 25% blue light, and then back to the original light environment of 100% red light.In some cases the complementary switchover experiment RB→R→RB was performed, and in this case, leaves were first acclimated to steady-state under 75% red and 25% blue light.500 µmol m -2 s -1 PPFD was maintained throughout the time course for this experiment.The choice for 75% red + 25% blue was based on the fact that 125 µmol m -2 s -1 is sufficient to saturate the stomatal blue light response, and red background light is often found to enhance the blue light response.Each switchover experiment was initiated when steady state was reached (typically after 2 h).Gas exchange measurements were logged every 60 s throughout the protocol.After logging the first 5 min (in Cleomaceae and Alloteropsis) and 10 min (in Flaveria) of the steady state condition, the light composition was switched to either red or red +blue, followed finally by a switch back to the initial light composition.During each phase gas exchange parameters were logged for 60 mins (in the Cleomaceae and Alloteropsis) and 50 mins (in Flaveria).Thus, for each measurement series, 125 mins were recorded for Cleomaceae and Alloteropsis subspecies and 100 mins for Flaveria.The initial experiment in Flaveria was performed using C 3 -like F. pringlei, C 4 F. trinervia and C 4 F. bidentis.However, because C 3 -like F. pringlei has proto-Kranz (Sage, 2016), it is not considered a canonical C 3 .Therefore, a repeat experiment in Flaveria was performed to include two true C 3 species, F. cronquistii and F. robusta.For this experiment, only sequence R→RB→R was repeated, since the previous experiments had shown that both sequences yielded very similar results.Reference CO 2 during the experiment was controlled at 400 µmol mol -1 , block temperature was kept at 25°C and average VPD was ca.1.2 kPa.All measurements were taken between 10:00 and 16:00 h.

Determination of the red-light response under constant C i
To characterize the red-light response of stomatal conductance, the protocol by Messinger et al. (2006) was adopted.This protocol measures the stomatal response to red light, while controlling for the potentially confounding effect of the stomatal CO 2 response by maintaining a constant intercellular CO 2 concentration (C i ) at each light intensity.Briefly, the Messinger protocol involved illuminating the leaf segment clamped inside the cuvette with 800 µmol m -2 s -1 red-light to steady state.Subsequently, the PPFD was lowered stepwise from 800, 600, 400, and 200 µmol m -2 s -1 , each change was initiated once steady state of stomatal conductance was achieved.As A and g s changed in response to the new light environment, CO 2 was manually adjusted using the mixer millivolt signal to keep C i at the value observed at 800 µmol m -2 s -1 .

Stomatal morphology and density using epidermal impressions
Epidermal imprints from all species, except for Flaveria trinervia, were taken from adaxial and abaxial layers using clear nail varnish (Rimmel, London) using the same leaves used for gas exchange measurements.Dried nail varnish was lifted from the leaf using sticky tape (Sellotape) and affixed onto glass slides.Images were captured using a Nikon microscope (Olympus BX41, Japan) fitted with a CCD camera (U-TV0.5zc-3/Micropublisher3.3 RTV, Olympus, Japan) and installed with Q-Capture Pro 7 imaging software.Total magnifications of x200 and x400 (Olympus WHB 10x/20 eyepiece x Olympus PN 20x/0.40 or PCN 60x/0.80 objectives) were used for stomatal density counting and investigation of stomatal anatomical parameters, respectively, unless stated otherwise.For Flaveria trinervia the quality of nail varnish impressions was not sufficient to collect stomatal anatomical parameters, data were instead collected from epidermal peels, using the same microscopy set up with a total magnification of x400.Therefore, these data were not from the same leaf as was used for gas exchange but instead were taken from other leaves at the same developmental stage.
To assess stomatal size (SS), guard cell width (GCW) and guard cell length (GCL) were measured using FiJi software (Schindelin et al., 2012).SS was calculated using the formula for the area of an ellipse.Stomatal density (SD) was calculated as the number of stomata per unit area (mm -2 ).A stoma was counted if >50% was located within the field of view.Four random microscopic fields of view each from the abaxial and adaxial surfaces were collected from 3-5 biological replicates, giving >350 stomata per species, except C 3 F. cronquistii and C 4 F. trinervia.

Statistical analysis
For the red and red+blue light switchover experiments, pairs of C 3 and C 4 species were measured in parallel between 10:00 and 16:00 h to control for the effect of circadian rhythms on stomatal conductance.Each phylogenetically-controlled comparison was an independent experiment.To estimate the impact of light on stomatal conductance resulting from the shift from one light environment to the next, a linear mixed-effects model was used.Light was the within-subject factor and species was the betweensubject factor.Biological replicates were treated as random variables.To carry out the analysis, g s and A at the introduction or removal of blue light (t = 5 min and t =65 min, R→BR→R) or the reverse (t = 5 and t = 65, BR→R→BR), and the final g s at t = 125 min were extracted from the gas exchange time series.Planned contrasts were used to compare independent group means.Linear mixed-effects model was also fitted to the red-light response data set.Red-light PPFD and species were assigned as within-and between-subject variables, respectively, while individual biological replicates were treated as random variables.For every genus, the density of stomata on the abaxial and adaxial surfaces was analyzed using a two-way ANOVA, and planned contrasts were used to compare group means.Data on stomatal size were also analysed using ANOVA and post-hoc comparisons were carried out using Tukey's HSD at an a=0.05.The statistical tests were carried out using R, version 4.2.1 (R Core Team, 2023) and RStudio, version 2023.6.0.421 (Posit Team, 2023) and JMP Pro 17 (SAS Institute).Graphs were plotted using R, version 4.2.1 and RStudio, version 2022.07.1 and the ggplot2 package (Wickham, 2016).Figures were compiled using the ggpubr package (Kassambara, 2023).

Results
Stomatal sensitivity to blue light is lower in C 4 species from the genus Cleomaceae and Flaveria compared to their C 3 counterparts To compare the sensitivity of stomatal opening to blue light relative to red light of equal intensity between C 3 and C 4 species, gas exchange was measured in response to a sequence of 75% red:25% blue (RB) light to 100% red (R) with a total PAR of 500 µmol photons m -2 s -1 as well as the reverse sequence.Following steady state g s of 0.49 mol H 2 O m -2 s -1 at R light, g s in the C 3 species T. hassleriana increased gradually to 0.64 mol H 2 O m -2 s -1 upon shifting to RB light and the effect was rapidly reversed when the light was returned to R (Figure 1A and Table S1).Similarly, when moving from RB→R in the reverse sequence, C 3 T. hassleriana rapidly declined from 0.38 to 0.20 mol H 2 O m -2 s -1 , and gradually increased again by the switch from R→RB over the course of an hour (Figure 1B and Table S2).In contrast, no effects of the R→RB→R sequence, nor the reverse RB→R→RB sequence were observed on g s in the C 4 species G. gynandra, which remained invariable at 0.19 mol H 2 O m -2 s -1 (Figures 1C, D; Tables S1 and  S2).It is also worth noting that the strong stomatal movements in C 3 T. hassleriana did not correspond to parallel responses in A (Figures S1A, B; Tables S3 and S4), which instead was not significantly impacted by the changes in light composition in either of the Cleomaceae species (Figures S1C, D; Tables S1-S4).
To find out if these differences between C 3 and C 4 Cleomaceae species were representative of general differences in stomatal responses between photosynthetic pathways, the R→RB→R sequence was also performed on Flaveria, another genus which contains C 3 , C 3 -C 4 , and C 4 species.The differences in blue light response between C 3 and C 4 species were also present in Flaveria (Figure S2 and Tables S5 and S6).However, to account for the fact that the first set of Flaveria experiments used the proto-Kranz C 3like species F. pringlei as the C 3 representative, the experiment was repeated to include two other C 3 Flaveria species, F. robusta and F. cronquistii.Since the complementary sequence RB→R→RB yielded the same results as the R→RB→R sequence, only the latter R→RB→R sequence was repeated (Figure 2) on two C 3 species, F. robusta (Figure 2A) and F. cronquistii (Figure 2B), as well as the two C 4 species, F. trinervia (Figure 2C) and F. bidentis (Figure 2D) and C 3 -like F. pringlei (Figure 2E).After reaching steady state in R, g s increased significantly from 0.22 mol H 2 O m -2 s -1 to 0.28 (Figure 2A) mol H 2 O m -2 s -1 in F. robusta upon the switch to RB and from 0.30 mol H 2 O m -2 s -1 to 0.36 mol H 2 O m -2 s -1 in C 3 F. cronquistii (Figure 2B).On the other hand, g s in C 4 F. trinervia showed only a slight increase from 0.11 mol H 2 O m -2 s -1 to 0.13 mol H 2 O m -2 s -1 under RB light (Figure 2C), and similarly g s in C 4 F. bidentis remained unchanged at 0.13 mol H 2 O m -2 s -1 (Figure 2D).Finally, in C 3 -like F. pringlei, from an initial g s of 0.16 mol H 2 O m -2 s -1 it rose to 0.19 mol H 2 O m -2 s -1 under BR (Figure 2E).Further, the stimulating effects on g s of the R→RB switch in the C 3 and C 3like species were largely reversed by the switch back to R light, but again g s remained unaffected in the C 4 species (Table S7).The light composition switches marginally, but significantly affected the rate of net CO 2 assimilation among the five species with slight lower rates of A under the RB part of the experiment (Figure S3 and Table S8, p=0.0122).
C 3 and C 4 subspecies of A. semialata display the canonical blue light response Cleomaceae and Flaveria are both dicotyledonous species, but C 4 photosynthesis is also quite prevalent in the monocots (Sage, 2004).Monocot species have stomatal features distinctly different from dicots (Bertolino et al., 2019).To determine if the stomatal responses described above also extend to congeneric C 3 and C 4 species in the monocots, the R→RB→R switchover experiment was also performed on C 3 and C 4 A. semialata subspecies (Figure 3).In this case, significant stimulation of g s in response to the R→RB switch was observed across both C 3 and C 4 subspecies, increasing from 0.05 to 0.07 mol H 2 O m -2 s -1 and returning to 0.05 mol H 2 O m -2 s -1 in response to the RB→R switch.
After reaching steady state in R, g s increased significantly from 0.05 mol H 2 O m -2 s -1 to 0.07 mol H 2 O m -2 s -1 in C 3 A. semialata Time course of stomatal conductance (g s ) response in C 3 T. hassleriana (A, B) and C 4 G. gynandra (C, D) in response to a sequence of 75% red + 25% blue light and vice versa.Leaves were initially acclimated in either 100% red (A, C) or 75% red+25% blue light (B, D) until steady-state was achieved (time = 0 in the figure).Subsequently, the light environment was switched depending on the initial light condition while maintaining a photosynthetic photon flux density of 500 µmol m -2 s -1 .Subsequently, the leaves were acclimated to the new light condition for 1 h before returning it back to the original condition.Light conditions were reversed at t 6 and t 65 .Reference CO 2 was maintained at 400 µmol mol -1 , block temperature was kept at 25°C and average VPD was 1.2 kPa.Data points represent mean ± se (n=4-5).
GMT upon the switch to RB, and then returned to 0.05 mol H 2 O m -2 s -1 when the light was shifted back to R (Figure 3A; Table S9).Both C 4 A. semialata ('MDG' and 'MAJ') increased from 0.04 mol H 2 O m -2 s -1 to 0.05 mol H 2 O m -2 s -1 under RB, and back to 0.04 when the light was reversed to R (Figures 3C, D).
In the reverse sequence, g s followed a similar pattern.In C 3 A. semialata 'GMT', g s was 0.07 mol H 2 O m -2 s -1 at the start of R, which then dropped to 0.05 mol H 2 O m -2 s -1 under R, and returned to 0.07 mol H 2 O m -2 s -1 when light was changed to RB (Figure 3B; Table S10).Following steady state, the initial g s in C 4 A. semialata 'MDG' under RB was 0.08 mol H 2 O m -2 s -1 , which rapidly declined to 0.06 mol H 2 O m -2 s -1 , with a final g s of 0.06 mol H 2 O m -2 s -1 (Figure 3D).Similarly, g s in C 4 A. semialata 'MAJ' went from 0.07 to 0.05 and back to 0.07 mol H 2 O m -2 s -1 throughout of the sequence (Figure 3E).
Under R→RB→R and the reverse sequence, RB→R→RB, fitting the g s data to a linear mixed-effects model (Tables S8 and  S9) showed a highly significant effect of light (R-RB-R, p=0.0007;RB-R-RB, p=0.0060), but the main effect of species (R-RB-R, p=0.3455;RB-R-RB, p=0.8859) and its interaction with light (R-R B -R , p = 0 . 2 0 5 5 ; R B -R -R B , p = 0 .7 4 8 6 ) w e r e n o t significantly different.
Although neither sequence had a clear effect on A (Figure S4), linear mixed-effects model analysis of A under R→RB→R showed a significant main effect of light (p=0.0013)(Table S10), this same relationship could not be detected in the reverse sequence (Table pringlei in response to a sequence of 100% red and 75% red + 25% blue light.Leaves were initially acclimated under 100% red light until steady state was achieved (time = 0 in the figure).Subsequently, the light environment was switched to 75% red + 25% blue light while maintaining a photosynthetic photon flux density of 500 µmol m -2 s -1 .The leaves were acclimated to the new light condition for 1 h before returning them back to the original condition for another hour before terminating the experiment.Light conditions were reversed at t 6 and t 65 .Reference CO 2 was maintained at 410 µmol mol -1 , block temperature was kept at 25°C and average VPD was 1.2 kPa.Data points represent mean ± se (n=4-5).** indicates significant difference at alpha = 0.05; ns, not significant.
S11, p=0.1395).Both main effects in the reverse sequence were also not significant.

Red light response in phylogeneticallyclose pairs of C 3 and C 4 at constant C i
To evaluate the quantitative red-light stomatal response, gas exchange was measured at different red PPFDs.The depletion of CO 2 in the mesophyll can give rise to an apparent red light response (Roelfsema et al., 2002) which was not the focus of the experiments here, therefore the stomatal response to red light was evaluated under constant C i .
In Cleomaceae, C 3 T. hassleriana showed a drop in steady state g s from 0.42 mol H 2 O m -2 s -1 to 0.26 mol H 2 O m -2 s -1 when red light intensity was reduced step-wise from 800 µmol m -2 s -1 to 200 µmol red photons m -2 s -1 (Figure 4A) while C i was maintained at 288 µmol CO 2 mol -1 (Figure 4B).Stomatal conductance in C 4 G. gynandra remained consistent around 0.15 mol H 2 O m -2 s -1 at the starting light intensity of 800 µmol m -2 s -1 red light (Figure 4C), while C i was held constant at 173 µmol CO 2 mol -1 (Figure 4D) and only slightly decreased to 0.12 mol H 2 O m -2 s -1 when light was set to 200 µmol red photons m -2 s -1 .Using a linear mixed-effects model, the interaction of photosynthetic type and red PPFD on g s was found to be significant (p=0.046)(Table S14), as a result of the significantly weaker red-light response in G. gynandra compared to T. hassleriana.In both species A significantly responded to red light PPFD (p<0.001)(FigureS5 and Table S15).
In Flaveria, C 3 F. cronquistii showed an average g s of 0.36 mol H 2 O m -2 s -1 at 800 mmol red photons m -2 s -1 which decreased to The leaves were acclimated to the new light condition for 1 h before returning it back to the original light condition.Reference CO 2 was maintained at 410 µmol mol -1 , block temperature was kept at 25°C and average VPD was 1.2 kPa.Data points represent mean ± se (n=3-4).
0.27 mol H 2 O m -2 s -1 at 200 mmol photons m -2 s -1 while C i was kept constant at 296 mmol CO 2 mol -1 (Figures 5A, B).C 3 F. robusta had a lower overall g s , which started from 0.17 mol H 2 O m -2 s -1 and decreased to 0.13 mol H 2 O m -2 s -1 at a C i of 247 mol CO 2 mol -1 (Figures 5C, D).Meanwhile steady state g s in C 4 F. bidentis under 800 mmol red photons m -2 s -1 was 0.18 mol H 2 O m -2 s -1 which slowly decreased to 0.13 mol H 2 O m -2 s -1 with the decrease in light intensity (Figure 5E), but while holding C i constant at 157 mmol CO 2 mol -1 (Figure 5F).g s in C 4 F. trinervia was practically unchanged in response to red light intensity ranging between 0.12 and 0.10 H 2 O m -2 s -1 (Figures 5G, H).Finally, C 3 -like F. pringlei had a g s of 0.19 mol H 2 O m -2 s -1 to 0.15 mol H 2 O m -2 s -1 (Figure 5I) as light intensity was reduced while keeping C i at 240 mmol CO 2 mol -1 (Figure 5J).Fitting the g s data to a linear mixed-effects model showed a highly significant species (p<0.0001) and red-light PPFD (p<0.0001)main effects, and a non-significant interaction effect (p=0.6789) (Table S15).The significance of the red-light effect demonstrates that despite the limited range of the response, stomata did respond significantly to red-light.The strong significance of the species effect demonstrates that the absolute values of stomatal conductance differed significantly between Flaveria species in line with photosynthetic type.However, the presence of a non-significant interaction between species and red light demonstrates that the magnitude of the stomatal red-light response did not differ significantly between these species.Not surprisingly, CO 2 assimilation rate responded strongly to light intensity in all five species (Figure S4 and Table S16), suggesting that the significant stomatal red-light response in these species may serve to fine-tune coordination between stomatal conductance and photosynthesis, in concert with the stomatal CO 2 response.The stomatal red-light response was also characterized for C 3 A. semialata subsp.eckloniana and C 4 A. semialata subsp.semialata (Figures 6A, C).When PPFD was reduced from 800 to 200 µmol red photons m -2 s -1 for C 3 A. semialata subsp.eckloniana, g s changed very little from 0.07 to 0.05 mol H 2 O m -2 s -1 (Figures 6A, B).Similarly, the C 4 A. semialata subsp.semialata kept g s at approximately 0.07 mol H 2 O m -2 s -1 despite the decrease in red PPFD (Figures 6C, D).Curiously the C 4 subspecies (Figure 6D) had a higher C i than the C 3 subspecies at 800 µmol red photons m -2 s -1 (Figure 6C; 153 vs 193 µmol CO 2 mol -1 ).A mixed-effects model detected a marginal effect of red PPFD p=0.0525) but there was no evidence of the effect of subspecies (p=0.3884) and its interaction with red PPFD (Table S17 , p=0.1979).Similar to Cleomaceae and Flaveria, assimilation rate also responded strongly to red-light PPFD (Figure S7 and Table S18).
Altogether, these data established a significant stomatal red light opening response to red-xlight, independent of C i in Cleomaceae and most Flaveria species, but not in Alloteropsis.Although the observed reduced stomatal sensitivity to red light in C 4 G. gynandra compared to C 3 T. hassleriana is consistent with the hypothesized differences between C 3 and C 4 photosynthetic types, the differences observed between Flaveria species and the lack of these in Alloteropsis seem to suggest that these are more likely to be species-specific responses, rather than generic differences between photosynthetic pathways.

Stomatal morphology and density
Stomatal density and size were determined in conjunction with the gas exchange measurements described in the previous sections.In Cleomaceae, the overall stomatal density was significantly higher in C 3 T. hassleriana (590 stomata mm -2 ) than in C 4 G.gynandra (200 stomata mm -2 ) (p<0.0001).The distribution of stomata on both leaf surfaces was also found to be significantly different, but in a species-dependent manner (Figure 7A, p=<0.0001).The SD on the adaxial surface was almost two times higher than the abaxial surface in C 3 T. hassleriana (Figures S8A, B).In contrast, in C 4 G.gynandra, SD was 25% higher on the adaxial compared to the abaxial surface (Figures S8C, D).While density varied depending on leaf surface in Cleomaceae, stomatal size was significantly larger on the abaxial than the adaxial surface (Leaf surface, p <0.0001) (Figure S9A).This SS difference was similar in both Cleomaceae species.
SD among Flaveria species (Figures 7C and S9A-J) were significantly different (p<0.0001),but the trend in SD was not consistent with the proposed progression of larger and fewer stomata in C 4 Flaveria than in C 3 Flaveria (Zhao et al., 2022).Instead, F. cronquistii, one of the two C 3 species, had the lowest SD and the highest SD was found for F. bidentis, one of the C 4 species.Thus, these differences do appear to be species-dependent, but not determined by photosynthetic pathway.SD was similar between both leaf surfaces in both C 4 Flaveria species (Figure S9C) as well as in the C 3 F. cronquistii, whereas the abaxial surface had higher SD in the C 3 -like F. pringlei and the C 3 F. robusta (significant for the latter, p<0.05).SS was higher on the abaxial side for all Flaveria species except for C 3 F. cronquistii resulting in an interaction effect of species x leaf surface which was highly significant (p<0.0001, Figure S7D).
Meanwhile, there was no statistically significant difference in SD detected between the Alloteropsis subspecies (p = 0.338) (Figures 7  and S10), while SS was larger on the abaxial surface (p <0.05, Figure 7F).Altogether, stomatal density, distribution and anatomical traits appeared to be largely determined by species effect, rather than photosynthetic pathway, although notably, all C 4 species had at least equal SD on the adaxial and abaxial surfaces, whereas the more commonly found bias of SD towards the abaxial surface was only found in some of the C 3 species.

Discussion
Leaf-level g s responses to red light quantity or red/blue spectral composition were compared between C 3 and C 4 species from Cleomaceae, Flaveria, and Alloteropsis genera.All three groups are well-studied models in the quest to understand evolution of C 4 photosynthesis.Combining experimental comparisons within all three genera allowed for a global comparison of stomatal light responses between C 3 and C 4 photosynthetic types while controlling for the effects of phylogenetic distance within each genus.
C 4 species from the genus Cleomaceae and Flaveria are less sensitive to blue light than the C 3 counterparts BL promotes diverse physiological plant responses ranging from phototropism, chloroplast photorelocation movement, leaf flattening, leaf positioning, and stomatal opening (Christie, 2007;Inoue et al., 2008;Goh, 2009).Blue-light induced stomatal opening is probably the most well-characterized among these responses.
However, the stomatal opening response to BL is not universal.Studies in model and non-model species provide strong evidence of the species-specificity of the BL stomatal opening response (Vialet-Chabrand et al., 2021).Differences in BL-dependent stomatal opening may exist.A recent study reported that C 4 crop species (Zhen and Bugbee, 2020) displayed diminished stomatal opening to BL compared to C 3 crops.Indeed, in this study, C 4 dicots were found to be less sensitive to blue light-induced stomatal opening than close relatives operating the C 3 pathway.This observation demonstrates a clear difference between C 3 and C 4 opening response, as hypothesized.In contrast, both the C 3 and C 4 accessions from the monocot A. semialata displayed blue lightdependent stomatal opening, which meant that the C 4

A B D C FIGURE 7
Stomatal density (SD, stomata mm -2 ) and stomatal size (SS, µm 2 ) on the abaxial and adaxial leaf surfaces of Cleomaceae (A, B), Flaveria (C, D).Data were collected from 1-4 random microscopic fields of view from abaxial and adaxial surfaces from 3-5 biological replicates.In each species, at least 350 stomata per surface were measured for SS, with the exception of C 3. F. cronquistii and C 4 F. trinervia, for which at least 141 and 170 stomata per surface were measured, respectively.Violin plots with embedded box plots in the same panel, with the same letters are not significantly different at a = 0.05 using Tukey's HSD test.
photosynthetic pathway per se does not determine this reduced blue light sensitivity.
The decreased sensitivity of C 4 dicots to blue light may be a result of decreased guard cell expression of phototropins, the major blue light photoreceptor in plants.Analysis of transcript abundance in different cell types of C 4 G. gynandra showed that the transcript abundance of PHOT1and PHOT2, as well as another blue light photoreceptor, CRY2, were lower in mesophyll and guard cells of C 4 G. gynandra, than in C 3 T. hassleriana Aubry et al. (2016).In Arabidopsis, phototropins are expressed in almost every plant part Sakamoto and Briggs (2002), however the strongest expression in the epidermis is detected in guard cells, where phots are associated with the plasma membrane ( (Sakamoto and Briggs, 2002)).PHOT1 and 2 have partially overlapping roles.PHOT1 has been suggested to respond to lower PPFD (0.1-50 µmol m -2 s -1 ), while PHOT2 responds to PPFD up to 250 µmol m -2 s -1 in Arabidopsis [15].Several hypotheses exist with regards to the function of the stomatal blue light response.One putative role is to stimulate photosynthesis via enhanced stomatal opening in the morning hours when blue light is more prevalent.In the results presented here, PPFD was kept equal during the spectral changes, and CO 2 assimilation rate was invariable, despite significant changes in stomatal conductance in C 3 T. hassleriana, C 3 -like F. pringlei, C 3 F. robusta, C 3 F. cronquistii and both A. semialata subspecies.These observations do not seem consistent with an important role for blue light induced stomatal opening for photosynthetic carbon gain.
Could there be other benefits to blue sensitivity of stomatal movements?An alternative hypothesis to explain the differential sensitivity to blue light could be related to leaf thermoregulation.Stomatal opening in response to blue light has been suggested to work as a proxy for high-intensity sunlight (Zhao et al., 2022) and function primarily to cool the leaf via transpiration (Taylor et al., 2012).Furthermore, photorespiration increases with temperature in C 3 species but does not change in C 4 species due to CCM (Lawson et al., 2008;Arrivault et al., 2017).Hence, net CO 2 assimilation rate in C 4 species has a higher optimum temperature than in C 3 species (Xu et al., 2016), and a slight elevation of leaf temperature can substantially increase the photosynthesis rate in C 4 , whereas the temperature response between 20°C and 30°C in C 3 species is almost negligible.Based on these differences, the presence of stomatal blue light response may help to cool the leaves of the C 3 dicot species studied here, and thus keep photorespiration lower, while the lack of the stomatal blue light response in the C 4 dicots may help elevate leaf temperature and thereby stimulate CO 2 assimilation rate.
Interestingly, this would also be consistent with the putative involvement of diminished PHOT expression, since PHOTs also perform thermosensory roles Fujii et al. (2017) and promote evapotranspiration and leaf cooling at high temperatures in Arabidopsis (Kostaki et al., 2020).
Differences in stomatal red-light response are determined by species, rather than photosynthetic pathway The mechanism underlying the red-light response remains unresolved.Unlike the blue light response, red light-induced stomatal opening does not seem to involve direct signal transduction from red light photoreceptors in the guard cells but rather relies on the decrease in C i through the consumption of CO 2 via mesophyll photosynthesis (Mott, 1988;Roelfsema et al., 2002;Lawson et al., 2008).However, g s was also observed to increase with light despite constant or high C i or after achieving steady-state photosynthesis (Matrosova et al., 2015), which implies other signals than C i could also be involved (Lawson et al., 2014;Flu¨tsch and Santelia, 2021).
C 4 photosynthesis operates in a two-cell compartment system, such that a strong metabolite gradient is essential to run it efficiently (Leegood and von Caemmerer, 1988;Arrivault et al., 2017).Precise coordination between the C 3 and C 4 cycles is crucial under varying irradiances and C i to avoid excessive CO 2 leakage from the bundle sheath (Kromdijk et al., 2014).C 3 photosynthesis, on the other hand, is less complex.We therefore hypothesized that species with these contrasting mesophyll photosynthesis characteristics may also be expected to show a different stomatal response to red light.
However, the comparison between the red-light response in C 3 and C 4 dicot species showed that although significant variation in the stomatal red-light response was found between species, no structural differences between C 3 and C 4 species were found across the three phylogenetically controlled comparisons.Instead, a distinct red-light response was found in C 3 T. hassleriana, and a weaker, but significant red-light effect was also found across all Flaveria species, suggesting that these species support the coordination between stomatal conductance and photosynthesis via stomatal responses to both C i and red light.In contrast, C 4 G. gynandra as well as both A. semialata subspecies appear to rely solely on the C i response.
One aspect of the red-light response debate is whether stomatal guard cells can independently respond to red light.Indeed, a recent metabolomics study reported a direct guard cell response to red light (Zhu et al., 2020).This finding could be consistent with a putative mechanism involving the phosphorylation of the guard cell plasma membrane H + ATPase leading to stomatal opening in intact leaves (Ando and Kinoshita, 2018).Consistent with this mechanism, it was demonstrated that H + ATPase activation by red light was dependent on fluence rate (Ando and Kinoshita, 2019).If this mechanism can be shown to play a significant role, it may offer an alternative explanation for the observed species differences in the red light response, not necessarily dependent on a mesophyllderived signal.

Role of stomatal morphology and density on stomatal movements in response to light
Stomatal density and morphology are well-known to impact responses to light (Harrison et al., 2020).These traits varied significantly between species studied here (Figures 7, 8).For most dicots, such as Cleomaceae and Flaveria, stomata are defined by a pair of kidney-shaped guard cells (Figures S8 and S9).In contrast, stomata in the monocot A. semialata subspecies are dumbbellshaped with additional subsidiary cells flanking the guard cells (Figure S10).The morphology and size of stomata in the grasses are often suggested to facilitate more rapid responses to short-term environmental perturbations (Chen et al., 2017), such as fluctuating light or acute changes to temperature or VPD (Bauer et al., 2013).However, there was no evidence in this study linking stomatal morphology nor size to blue light-induced opening or quantitative red-light responses.
Within the dicots, C 3 T. hassleriana (Cleomaceae) had denser and smaller stomata than the C 4 species G. gynandra, which was consistent with earlier observations Aubry et al. (2016).The higher SD on the adaxial surface could be responsible for the pronounced red-light response, independent of C i in C 3 T. hassleriana.However, measurements on intact leaves do not make it possible to derive whether these responses rely on signal perception in the guard cell, or in the underlying mesophyll.
Furthermore, in Flaveria, SD and SS varied between species.Unlike in the Cleomaceae species, SD or SS had no evident relationship to the red-light response in Flaveria.Additionally, the observed between-species variability in SD and SS in Flaveria was inconsistent with the proposed trajectory of stomatal density and guard cell length (and therefore, size) changes during C 4 evolution, where C 3 Flaveria species tended to have smaller but more stomata, while C 4 Flaveria show increased stomatal size and decreased density (Zhao et al., 2022).In fact, C 3 -like F. pringlei and C 3 F. cronquistii, were found to have larger stomata (Figure 7D), whereas both C 4 Flaveria had smaller stomatal sizes.The discrepancy between this study and Zhao et al. (2022) might have stemmed from very low sample sizes used by the latter.Stomatal parameters in Zhao et al. (2022) were measured from a minimum of 5 individual stoma, to at most 10 stomata per species, which, as shown in Figure 7D, is insufficient to capture size differences among Flaveria species.
Another possible reason for these discrepancies is the potential impact of ecological adaptation in some of the Flaveria species.C 3 F. cronquistii had equal stomatal distribution on both leaf surfaces like the C 4 Flaveria species, but with twice larger stomatal size on either leaf surface.In terms of leaf shape, C 3 F. cronquistii has linear, elongated leaves, whereas the other Flaveria representatives display elliptic or ovate leaf shapes.At an ecological standpoint, narrower leaves such as those in C 3 F. cronquistii perhaps reflect an adaptation in minimizing excessive heat load (Leigh et al., 2017) whilst maximizing net carbon gain (Michaletz et al., 2016) and together with a uniform stomatal distribution and a larger proportion of large stomata allow it to sustain higher g s rates to effect cooling.

Conclusion
Here we used three phylogenetically-controlled comparisons to assess differences between stomatal anatomy and stomatal

A B
Stomatal density (SD, stomata mm -2 ) and stomatal size (SS, µm 2 ) on the abaxial and adaxial leaf surfaces of Alloteropsis (A, B).Data were collected from 1-4 random microscopic fields of view from abaxial and adaxial surfaces from 3-5 biological replicates.In each species, at least 350 stomata per surface were measured for SS.Violin plots with embedded box plots in the same panel, with the same letters are not significantly different at a = 0.05 using Tukey's HSD test.
responses to red and blue light.The 4 species from the genus Cleomaceae and Flaveria in this study did not have a detectable blue light stomatal response, unlike their C 3 counterparts.However, perhaps surprisingly, the results demonstrate that the impact of photosynthetic pathway and stomatal morphology and distribution were not as strong as initially hypothesized, but instead varied between genera.Similarly, the quantitative red-light response showed significant species variation but no association with photosynthetic pathway.Altogether, the findings suggest that the evolution of C 4 photosynthesis in the dicots may have led to a change in light-regulated stomatal movements, challenge the general nature of previously observed stomatal morphological differences between C 3 and C 4 species (Zhao et al., 2022) and demonstrate the importance of controlling for evolutionary distance.
FIGURE 2Time course of stomatal conductance (g s ) response in C 3 F.robusta (A), C 3 F. cronquistii (B), C 4 F. trinervia (C), C 4 F. bidentis (D), and (E) C 3 -like F. pringlei in response to a sequence of 100% red and 75% red + 25% blue light.Leaves were initially acclimated under 100% red light until steady state was achieved (time = 0 in the figure).Subsequently, the light environment was switched to 75% red + 25% blue light while maintaining a photosynthetic photon flux density of 500 µmol m -2 s -1 .The leaves were acclimated to the new light condition for 1 h before returning them back to the original condition for another hour before terminating the experiment.Light conditions were reversed at t 6 and t 65 .Reference CO 2 was maintained at 410 µmol mol -1 , block temperature was kept at 25°C and average VPD was 1.2 kPa.Data points represent mean ± se (n=4-5).** indicates significant difference at alpha = 0.05; ns, not significant.
FIGURE 3Time course of stomatal conductance (g s ) response in C 3 A. semialata subsp.eckloniana 'GMT' (A, B), C 4 A. semialata subsp.semialata 'MDG' (C, D) and C 4 A. semialata subsp.semialata 'MAJ' (E, F) in response to a sequence of 75% red + 25% blue light and vice versa.Leaves were initially acclimated in either 100% red (A, C, E) or 75% red+25% blue light (B, D, F) until steady-state was achieved (time = 0 in the figure).Subsequently, the light environment was switched depending on the initial light condition while maintaining a photosynthetic photon flux density of 500 µmol m -2 s -1 .The leaves were acclimated to the new light condition for 1 h before returning it back to the original light condition.Reference CO 2 was maintained at 410 µmol mol -1 , block temperature was kept at 25°C and average VPD was 1.2 kPa.Data points represent mean ± se (n=3-4).
FIGURE 4Response of stomatal conductance (g s ) to red light photosynthetic photon flux density (PPFD) in C 3 T. hassleriana (A) and C 4 G. gynandra (B) at constant C i (C, D).A linear mixed-effects model analysis was carried out to test if photosynthetic type influences the response to red light in congeneric species belonging to Cleomaceae.Data points represent mean ± se (n=3-4).
FIGURE 5Response of stomatal conductance (g s ) to red light photosynthetic photon flux density (PPFD) at constant intercellular CO 2 concentration (C i ) in C 3 F. cronquistii (A, B, n=4) and C 3 F. robusta (C, D, n=5), C 4 F. bidentis (E, F, n=8), C 4 F. trinervia (G, H, n=5), and C 3 -like F. pringlei (I, J, n=5).A linear mixed-effects model analysis was carried out to test if photosynthetic type influences the response to red light in congeneric species belonging to Flaveria.Data points represent the mean ± se.
FIGURE 6Response of stomatal conductance (g s ) at constant intercellular CO 2 concentration (C i ) in C 3 A. semialata subsp.eckloniana 'GMT' (A, B, n=4) and C 4 A. semialata subsp.semialata'MDG' (C, D, n=5).Linear mixed-effects model analysis was carried out to test if photosynthetic type influences the response to red light in C 3 and C 4 subspecies belonging to Alloteropsis.Each data point represents the mean ± se.