Moderate heat stress prevented the observed biomass and yield stimulation caused by elevated CO2 in two well-watered wheat cultivars

Key message Heat stress (HS) under well-watered conditions was not detrimental to leaf photosynthesis or yield but modified the elevated CO2 response of photosynthesis and yield in two contrasting wheat cultivars. Abstract Climate change is increasing the frequency of extreme events such as heat waves, adversely affecting crop productivity. While positive impacts of elevated carbon dioxide (eCO2) on crop productivity are evident, the interactive effects of eCO2 and environmental stresses are still unclear. To investigate the interactive effects of elevated CO2 and heat stress (HS), we grew two contrasting wheat cultivars, early-maturing Scout and high-tillering Yitpi, under non-limiting water and nutrients at ambient (aCO2, 450 ppm) or elevated (eCO2, 650 ppm) CO2 and 22 °C in the glasshouse. Plants were exposed to two 3-day HS cycles at the vegetative (38.1 °C) and/or flowering (33.5 °C) stage. At aCO2, both wheat cultivars showed similar responses of photosynthesis and mesophyll conductance to temperature and produced similar grain yield. Relative to aCO2, eCO2 enhanced photosynthesis rate and reduced stomatal conductance and maximal carboxylation rate (Vcmax). During HS, high temperature stimulated photosynthesis at eCO2 in both cultivars, while eCO2 stimulated photosynthesis in Scout. Electron transport rate (Jmax) was unaffected by any treatment. eCO2 equally enhanced biomass and grain yield of both cultivars in control, but not HS, plants. HS reduced biomass and yield of Scout at eCO2. Yitpi, the cultivar with higher grain nitrogen, underwent a trade-off between grain yield and nitrogen. In conclusion, eCO2 improved photosynthesis of control and HS wheat, and improved biomass and grain yield of control plants only. Under well-watered conditions, HS was not detrimental to photosynthesis or growth but precluded a yield response to eCO2. Supplementary Information The online version contains supplementary material available at 10.1007/s11103-022-01276-7.


Introduction
Ongoing climate change is threatening the production of agricultural crops including wheat (Triticum aestivum) (IPCC 2014;Asseng et al. 2015;Mishra et al. 2021). By the end of this century, atmospheric carbon dioxide concentration ([CO 2 ]) is expected to reach 700 ppm, increasing surface temperatures by 1.1-2.6 °C (IPCC 2014). For every degree of the temperature increase, global wheat production is predicted to decrease by 6-10% (Asseng et al. 2015;García et al. 2015). Crop models are important tools for assessing the impact of climate change (Asseng et al. 2013). However, they largely lack the ability to consider genotype-specific responses to elevated [CO 2 ] (eCO 2 ) and their interaction with other environmental conditions. Hence, it is important to better understand how plants respond to eCO 2 interactions with the environment. Photosynthesis, a fundamental 1 3 process driving crop growth and yield, can partially explain the interactive effects of eCO 2 with environmental stresses and provide a mechanistic basis for crop models (Yin and Struik 2009).
Optimum temperature range for wheat growth is 17-23 °C, with a minimum of 0 °C and maximum of 37 °C (Porter and Gawith 1999). Global warming involves a gradual increase in mean temperature as well as increased frequency and intensity of heat waves. Heat can adversely affect crop growth and disrupt reproduction depending on the timing, intensity and duration (Sadras and Dreccer 2015). Higher daytime temperatures (below damaging level) increase photosynthesis up to an optimum temperature, above which photosynthesis decreases mainly due to higher photorespiration (Berry and Bjorkman 1980;Long 1991). High night time temperatures increase respiration and reduce overall photosynthetic carbon gain (Prasad et al. 2008). At the whole plant level, high temperatures accelerate growth (Fischer 1980) and shorten crop duration (Hatfield and Prueger 2015), hence reducing grain yield due to insufficient time to capture resources. Losses due to short crop duration are usually higher than benefits of growth stimulation at high temperature (Wardlaw and Moncur 1995).
The severity of the damage caused by abrupt temperature increases above the optimum range (termed heat stress, HS) depends on the magnitude and duration of HS as well as the developmental stage of the plant (Wahid et al., 2007). HS may reduce photosynthesis due to reduced chlorophyll content, impaired photosystem II and lower Rubisco activation (Berry and Bjorkman 1980;Eckardt and Portis 1997). Furthermore, HS can directly damage cells and increase grain abortion resulting in reduced growth, biomass and grain yield Nicolas 1996, 1998;Wardlaw et al. 2002;Farooq et al. 2011). Around anthesis, HS (> 30 °C) reduces seed setting due to lower pollen viability, leading to poor fertilization, abnormal ovary development and slower pollen growth (Balla et al. 2019).
The interactive effects of eCO 2 and HS on plant growth and yield can be positive, negative or neutral (Wang et al. , 2011. Plants may acclimate to changes in growth temperature and shift the optimum temperature for photosynthesis, which can maximize the photosynthetic rate at the growth temperature . Elevated CO 2 increases the temperature optimum of photosynthesis (Long 1991;Alonso et al. 2009) by reducing photorespiration and improving tolerance to photoinhibition (Hogan et al. 1991). In addition, differences in plasticity of photosynthetic parameters with respect to growth temperature have been found responsible for differences in photosynthetic temperature acclimation (Yamori et al. 2010) and respiratory temperature acclimation can generate apparent acclimation of photosynthetic processes . The impact of HS on photosynthesis will depend on whether Rubisco, electron transport or end-product synthesis is limiting at eCO 2 (Sage and Kubien 2007). Enhanced growth and leaf level intrinsic water use efficiency (iWUE) by eCO 2 may help compensate for the negative impact of HS; conversely, heat-induced shortening of the grain-filling stage and grain abortion could limit the benefits of eCO 2 (Lobell and Gourdji 2012). In addition, decreased g s under eCO 2 may limit transpirational cooling and therefore exacerbate HS. Thus, HS counteracts the positive effect of eCO 2 on yield components and may aggravate the negative effect of eCO 2 on grain quality due to the high sensitivity of wheat to temperature stress especially during anthesis and grainfilling stage (Wang and Liu 2021).
Many studies have investigated the response of wheat to eCO 2 in enclosures and in the field (Wang and Liu 2021). However, only a few studies have considered eCO 2 interaction with temperature increases in wheat (Rawson 1992;Delgado et al. 1994;Morison and Lawlor 1999;Jauregui et al. 2015;Cai et al. 2016) and rarely with HS (Coleman et al. 1991;Wang et al. 2008). Studies considering HS have addressed mainly the biomass or yield aspects and not the physiological processes such as photosynthesis (Stone and Nicolas 1994. Interactive effects of eCO 2 and HS on photosynthesis have been reported in a limited number of studies (Wang et al. , 2011Macabuhay 2016;Macabuhay et al. 2018;Chavan et al. 2019). Macabuhay et al. (2018) studied interactive effects of eCO 2 and (experimentally imposed) heatwaves on wheat (cv Scout and Yitpi) grown in a dryland cropping system and concluded that eCO 2 may moderate some effects of HS on grain yield but such effects strongly depend on seasonal conditions and timing of HS. In another glasshouse experiment on the interactive effects of severe HS (achieved by increasing RH to lower transpirational cooling) and eCO 2 in wheat (cv Scout), we found that eCO 2 mitigated the negative impacts of HS at anthesis on photosynthesis and biomass, but grain yield was reduced by HS in both CO 2 treatments (Chavan et al. 2019). However, HS can occur throughout plant growth, including during vegetative, flowering or grain filling stages. Moreover, irrigation and relative humidity during HS plays an important role in how plants handle HS and plants may cope well with HS in well-watered conditions. In addition, different crop genotypes may respond variably to the interaction of eCO 2 with HS.
Here, we build on our previous work by comparing the interactive effects of eCO 2 and HS in two commercial wheat cultivars. In this study, we investigated the impact of HS during well-watered conditions at ambient RH. Scout and Yitpi have similar genetic background but distinct agronomic features. Scout is a mid-season maturity cultivar with very good early vigor that can produce leaf area early in the season. Scout has a putative water-use efficiency (WUE) gene, which has been identified using carbon isotope discrimination (Condon et al. 2004). Yitpi is a good early vigor, freely tillering, late flowering and long maturity cultivar (Seednet 2005;Pacificseeds 2009;Bahrami et al. 2017).
Although Scout is known to be a high yielding variety with very good grain quality and high reproductive sink (Pacific seeds 2009), we hypothesized that Yitpi will produce higher grain yield due to its ability to initiate more tillers and its longer time to flower and mature (Hypothesis 1). Fast growing plants with high sink capacity show a greater eCO 2 -induced growth stimulation (Poorter 1993) and less photosynthetic acclimation (Delgado et al. 1994) compared to slow growing counterparts with low sink capacity. Consequently, we hypothesized that Yitpi will show greater photosynthetic, growth and yield response to eCO 2 due to its greater vegetative sink capacity (tillering) relative to Scout with restricted tillering (Hypothesis 2). The greater growth stimulation at eCO 2 may buffer Yitpi against HS damage compared to Scout. Thus, HS may decrease yield in Scout more than Yitpi and aCO 2 more than eCO 2 (Hypothesis 3). HS is more damaging at the reproductive relative to the vegetative developmental stage (Farooq et al. 2011). Hence, we expect less damage in plants exposed to HS at the vegetative stage relative to the flowering stage (Hypothesis 4).
To test these hypotheses, Scout and Yitpi were grown at controlled ambient or elevated CO 2 conditions and subjected to one or two heat stresses at the vegetative (HS1) and/or flowering (HS2) stage. Plant growth, biomass and leaf photosynthetic parameters were measured at different time points across the life cycle of the plants.

Plant culture and treatments
The experiment was conducted in the glasshouse facility located at the Hawkesbury campus of Western Sydney University (WSU). Seeds of commercial winter wheat cultivars Scout and Yitpi were procured from Agriculture Victoria (Horsham). Cultivars were selected based on their use in the Australian Grains Free Air CO 2 Enrichment (AGFACE) project investigating climate change impacts on wheat growth and yield (Houshmandfar et al. 2017). For germination, 300 seeds of each cultivar were sterilized using 1.5% NaOCl 2 for 1 min followed by incubation in the dark at 28 °C for 48 h in petri plates. Sprouted seeds were planted in germination trays using seed raising and cutting mix (Scotts, Osmocote®) at ambient CO 2 (aCO 2 , 400 μL L −1 ), temperature (22/14 °C day/night), relative humidity (RH, 50-70%) and natural light (midday average 500 μmol m −2 s −1 ) ( Figure S1). The growth stages are denoted by decimal code (DC) according to (Zadoks et al. 1974) along with the time points here after. Two-week-old seedlings (DC12) were transplanted to individual cylindrical pots (15 cm diameter and 35 cm height) using sieved soil collected from local site. At transplanting stage (T0) pots were distributed into two aCO 2 (400 μL L −1 ) and two eCO 2 (650 μL L −1 ) chambers ( Figure S1B). Some plants were exposed to one or two HS cycles at the vegetative (HS1, 10 weeks after planting, WAP, DC 32) and/or the flowering (HS2, 15 WAP, DC 63) stages for 3 days with temperature ramp up from 14 °C night temperature (8 pm to 6 am) to 38 °C or 33 °C during mid-day (10 pm to 4 pm) respectively at 60% daytime RH (Figures S1, S2). The two HS cycles created four sets of heat treatments at each CO 2 concentration as follows: (1) Control-plants were not exposed to HS at any stage, (2) HS1-plants were exposed to HS at vegetative (DC32) stage only, (3) HS2-plants were exposed to HS at reproductive (DC63) stage only and (4) HS1 + 2-plants were exposed to both the heat stresses HS1 and HS2 (Fig. 1).
Thrive all-purpose fertilizer (Yates) was applied monthly throughout the experiment to maintain similar nutrient supply in all treatment combinations. Pots were regularly swapped between left and right benches as well as between front and back for randomization within chamber. Pots and treatments were also swapped between the two ambient and two elevated CO 2 chambers for randomization among chambers.

Growth and biomass measurements
The full factorial experimental design included four chambers (two chambers for each CO 2 treatment) and five destructive harvests at time points T0 (2 WAP, DC12), T1 (6 WAP, DC28), T2 (10 WAP, DC35), T3 (17, DC65) and T4 (25 WAP, DC90) ( Fig. 1). Ten plants per treatment per cultivar were measured and harvested at each time point except for T4 where plants exposed to only HS1 were not harvested. Morphological parameters were measured followed by determinations of root, shoot and leaf dry mass. Samples were dried for 48 h in the oven at 60 °C immediately after harvesting. Leaf area was measured at time point T1, T2 and T3 using a leaf area meter (LI-3100A, LI-COR, Lincoln, NE, USA). Plant height, leaf number, tiller number and ear (grain inflorescence) number along with developmental stage information (booting (DC45), half-emerged (DC55) or fully emerged (DC60)) were recorded at time points T2 and T3).

Leaf gas exchange measurements
The youngest fully developed leaf (which was the flag leaf at T3) was used to measure gas exchange parameters. Steady state leaf gas exchange measurements were performed at time points T1, T2 and T3 using a portable open gas exchange system (LI-6400XT, LI-COR, Lincoln, USA) to measure light-saturated (photosynthetic photon flux density (PPFD) = 1500 µmol m −2 s −1 ) photosynthetic rate (A sat ), stomatal conductance (g s ), ratio of intercellular to ambient , leaf transpiration rate (E), dark respiration (R d ) and dark-and light-adapted chlorophyll fluorescence (Fv/Fm and Fv′/Fm′, respectively). Dark adapted leaf measurements were conducted by switching off light for 15 min. Steady state leaf gas exchange measurements were also performed during (3rd day of HS) and after heat shock (next day after HS) along with recovery stage (5 days after HS). Plants were moved to a neighboring chamber with ambient CO 2 levels for short time (20-30 min for each plant) where air temperature was separately manipulated to achieve the desired leaf temperature. The LI-COR 6400-40 leaf chamber fluorometer (LCF) was used to measure gas exchange parameters at a PPFD of 1500 μmol m −2 s −1 at two reference CO 2 (CO2R) concentrations (400 and 650 μL L −1 ) and two leaf temperatures (25 and 35 °C). Photosynthetic down regulation or acclimation was examined by comparing the measurements at common CO 2 (ambient and elevated CO 2 grown plants measured at 400 μL L −1 CO 2 partial pressure) and growth CO 2 (aCO 2 grown plants measured at 400 μL L −1 CO 2 partial pressure and eCO 2 grown plants measured at 650 μL L −1 CO 2 partial pressure). Dark respiration (R d ) was measured after a dark adaptation period of 15 min. Intrinsic water use efficiency (iWUE) was calculated as A sat (μmol m −2 s −1 )/g s (mol m −2 s −1 ). The response of A sat to variations in sub-stomatal CO 2 mole fraction (Ci) (A-Ci response curve) was measured at T3 in 8 steps of CO 2 concentrations (50, 100, 230, 330, 420, 650, 1200 and 1800 μL L −1 ) at leaf temperature of 25 °C. Measurements were taken around mid-day (from 10 am to 3 pm) on attached last fully expanded or flag leaves of the main stems. Before all leaf gas exchange measurements, the leaf was allowed to stabilize for 10-20 min until it reached a steady state of CO 2 uptake and stomatal conductance. Ten replicate plants per treatment were measured except for plants exposed only HS1 at T3 were not measured.

Mesophyll conductance and temperature response
Mesophyll conductance (g m ) was determined by concurrent gas exchange and stable carbon isotope measurements using portable gas exchange system (LI-6400-XT, LI-COR, Lincoln, NE, USA) connected to a tunable diode laser (TDL) (TGA100, Campbell Scientific, Utah, USA) for the two wheat cultivars grown at ambient atmospheric CO 2 levels. A sat and 13 CO 2 / 12 CO 2 carbon isotope discrimination were measured after T1 at five leaf temperatures (15, 20, 25, 30 and 35 °C) and saturating light (1500 µmol quanta m −2 s −1 ). Leaf temperature sequence started at 25 °C decreasing to 15 °C and then increased up to 35 °C. A-Ci response curves were measured at each leaf temperature. Dark respiration was measured by switching light off for 20 min at the end of each temperature curve.
Measurements were made inside a growth cabinet (Sanyo) to achieve desired leaf temperature. The photosynthetic carbon isotope discrimination (Δ) to determine g m was measured as follows (Evans et al. 1986): where, C ref and C sam are the CO 2 concentrations of dry air entering and exiting the leaf chamber, respectively, measured by the TDl. g m was calculated using correction for ternary effects (Farquhar and Cernusak 2012;Evans and Von Caemmerer 2013) following the next expression: where, Δ i is the fractionation that would occur if the gm were infinite in the absence of any respiratory fractionation (e = 0), Δ o is observed fractionation, Δ e and Δ f are fractionation of 13 C due to respiration and photorespiration respectively (Evans and Von Caemmerer 2013). where, The constants used in the model were as follows: E denotes transpiration rate; g t ac is total conductance to CO 2 diffusion in the boundary layer (ab = 2.9‰) and in air (a = 4.4‰); a′ is the combined fractionation of CO 2 across boundary layer and stomata; net fractionation caused by RuBP and PEP carboxylation (b = 27.3‰) (Evans et al. 1986); fractionation with respect to the average CO 2 composition associated with photorespiration (f = 11.6‰) (Lanigan et al. 2008) and we assumed null fractionation associated with mitochondrial respiration in light (e = 0).

Nitrogen and carbon estimation
Leaf discs were cut from the flag leaves used for gas exchange measurements at time points T2 and T3 which were then oven dried at 60 °C for 48 h. Leaf discs were processed for nitrogen (N) and carbon (C) content using an elemental analyzer (Dumas method). N and C were also estimated from other plant components including leaf, stem, root and grain harvested at T1, T3 and T4. Ground samples were processed for C & N with a CHN analyzer (LECO Tru-Mac CN-analyser, Leco corporation, USA) using an automated dry combustion method (Dumas method). Leaf N per unit area (N area ) was calculated as N (mmol g −1 ) × LMA (g m −2 ). Photosynthetic nitrogen use efficiency (PNUE) was calculated as A sat (μmol m −2 s −1 )/leaf N area (mmol m −2 ). Protein content was estimated using N and multiplication factor of 5.7 (Mosse 1990;Bahrami et al. 2017). N utilization efficiency was calculated as grain yield per total plant N.

Statistical and temperature analysis
All data analyses and plotting were performed using R computer software (R Core Team 2022). The effect of treatments and their interactions was analyzed using linear modeling with 'anova' in R. Significance tests were performed with anova and post hoc Tukey test using the 'glht' function in the multcomp R package. Coefficient means were ranked using post-hoc Tukey test. The Farquhar-von Caemmerer-Berry (FvCB) photosynthesis model was fit to the A sat response curves to C i (A-Ci response curve) measured at T3 or chloroplastic CO 2 mole fraction (Cc), which was estimated from the g m measuremed (A-Cc response curve) after T1. We used the plantecophys R package (Duursma 2015) to perform the fits, using measured g m and R d values, resulting in estimates of maximal carboxylation rate (V cmax ) and maximal electron transport rate (J max ) for D-ribulose-1,5-bisphosphate carboxylase/oxygenase (Rubisco) using measured R d values.
Temperature correction parameter (Tcorrect) was set to False while fitting A-C i curves. Temperature response of V cmax and J max were calculated by standard Arrhenius and modified Arrhenius functions, respectively (Medlyn et al. 2002). Estimated V cmax and J max values at five leaf temperatures were then fit using nonlinear least square (nls) function in R to determine energy of activation for V cmax (EaV) and J max (EaJ) and entropy (ΔSJ). Temperature responses of V cmax and R d were fit using Arrhenius equation as follows, where, Ea is the activation energy (in J mol −1 ) and k 25 is the value of R d or V cmax at 25 °C. R is the universal gas constant (8.314 J mol −1 K −1 ) and Tk is the leaf temperature in K. The activation energy term Ea describes the exponential rate of rise of enzyme activity with the increase in temperature. The temperature coefficient Q 10 , a measure of the rate of change of a biological or chemical system as a consequence of increasing the temperature by 10 °C was also determined for Rd using the following equation: A peaked function (Harley et al. 1992) derived Arrhenius function was used to fit the temperature dependence of J max , and is given by the following equation: where, Ea is the activation energy and k 25 is the J max value at 25 °C, Hd is the deactivation energy and S is the entropy term. Hd and ΔS together describe the rate of decrease in the function above the optimum. Hd was set to constant 200 kJ mol −1 to avoid over parametrization. The temperature optimum of J max was derived from Eq. 10 (Medlyn et al. 2002) and written as follows: The temperature response of A sat was fit using a simple parabola equation (Crous et al. 2013) to determine temperature optimum of photosynthesis: where, T is the leaf temperature of leaf gas exchange measurement for A sat , T opt represents the temperature optimum and A opt is the corresponding A sat at that temperature optimum. Steady state gas exchange parameters g m , g s , C i and J max to V cmax ratio were fit using nls function with polynomial equation:

Results
Two commercial wheat cultivars Scout and Yitpi were grown under aCO 2 or eCO 2 (daytime average of 450 or 650 μL L −1 , respectively; 65% RH and 22 °C), natural sunlight and well-watered conditions ( Figure S1). Both aCO 2 and eCO 2 grown plants were exposed to two 3-day HS cycles at the vegetative (HS1, 10 WAP, DC32, daytime average of 38 °C) and flowering stage (HS2, 15 WAP, DC63, daytime average of 33.5 °C), while daytime RH was maintained at 60%. HS2 was lower in intensity relative to HS1 due to the cool winter conditions. Both HS cycles had similar overall effects on growth and yield parameters, refuting our fourth hypothesis that HS during the reproductive stage is more damaging. Hence, we mostly compare the control plants to those exposed to both heat stresses. Grain filling started 17 WAP (DC65) and final harvest occurred 25 WAP (DC90) (Fig. 1).

Photosynthetic temperature responses of the two wheat cultivars at aCO 2
A-C i curves together with g m were measured at five leaf temperatures to characterize the thermal photosynthetic responses of the two wheat cultivars grown at aCO 2 ( Fig. 2;   Fig. 2 Temperature response of photosynthetic parameters: CO 2 assimilation rate (a), mesophyll conductance (b), stomatal conductance (c) and intercellular CO 2 (d), V cmax (e), J max (f), J max / V cmax (g) and dark respiration (h) over leaf temperatures (15, 20, 25, 30 and 35 °C) in plants grown at aCO 2 . Scout and Yitpi are depicted using circles with solid cultivars and triangles with broken cultivars respectively. Data in panels (a-f) and (h) are fit using nonlinear least square (nls) function in R Table 1). Overall, both cultivars had similar photosynthetic temperature responses. A sat and g s increased with leaf temperature up to an optimum (T opt ) around 23.4 °C and decreased thereafter, while C i slowly decreased with temperature. Mesophyll conductance (g m ) increased up to 25 °C then plateaued (Fig. 2b). Scout had slightly higher A sat , g s and g m than Yitpi at most leaf temperatures (Fig. 1a-d). R d linearly increased with temperature, and both cultivars had similar Q 10 (Fig. 2h, Table 1). V cmax and J max exponentially increased with leaf temperature, but J max declined above T opt (30 °C) in both cultivars (Fig. 1e, f). There was no significant difference in V cmax , J max or their activation energies between the two wheat cultivars (Fig. 2e-g; Table 1). The ratio of J max /V cmax was similar for the two cultivars and linearly decreased with leaf temperature (Fig. 2g).

High temperature during HS enhanced photosynthesis under eCO 2
The two HS cycles did not reduce A growth measured at 25℃ during or after HS (Fig. 3a-d; Tables S1, S2, S3). During both HS1 and HS2, eCO 2 stimulated A growth measured at 25℃ in Scout but not Yitpi. Relative to 25 °C, A growth increased at 35 °C in Scout (10-14%) and Yitpi (12-18%) grown at eCO 2 but not at aCO 2 . Immediately after the recovery from HS, A growth was upregulated in eCO 2 -grown Scout (Figs. 3a-d, S3). During both HS cycles, dark-adapted Fv/Fm measured at 25 °C tended to be lower in Yitpi grown at eCO 2 relative to aCO 2 . In both cultivars, Fv/Fm decreased at 35 °C relative to 25 °C, indicating transient damage to PSII due to HS at both CO 2 treatments (Fig. 3e-h; Tables S1, S2, S3).
Following long-term recovery from HS1 and/or HS2, the eCO 2 stimulation of A growth was still marginally apparent in all T3 plants, being the strongest in eCO 2 -grown Yitpi (Fig. 5a, b; Tables S1, S2, S3). The reduction of g s at eCO 2 was weak in all plants (Fig. 5c, d; Tables S1, S2, S3). Hence, iWUE was stimulated by eCO 2 in all treatments, while PNUE was unaffected (Fig. 5e-h; Tables S1, S2, S3). There Table 1 Summary of modelled parameters for temperature response of photosynthesis Summary of coefficients derived using nonlinear least square fitting of CO 2 assimilation rates and maximal rate of carboxylation (V cmax ) and maximal rate of RuBP regeneration (J max ) determined using A-C i response curves and dark respiration measured at five leaf temperatures 15, 20, 25, 30 and 35 °C. Values are means with standard errors. Derived parameters include temperature optima (T opt ) of photosynthesis (A opt ); activation energy for carboxylation (EaV); activation energy (EaJ)¸ entropy term (∆SJ) and T opt and corresponding value for J max with deactivation energy (Hd) assumed constant; and activation energy (EaR) and temperature coefficient (Q 10 ) for dark respiration. Letters indicate significance of variation in means 30.9 ± 1.6 a 33.2 ± 1.7 a Q 10 1.51 ± 0.03 a 1.56 ± 0.04 a was a good correlation between A growth and g s (r 2 = 0.51, p < 0.001) across all treatments (Fig. 6a). V cmax and J max were derived from A-C i response curves measured at 25 °C during the recovery stage after HS2. For control and HS plants, growth at eCO 2 marginally reduced V cmax in Scout (− 14%, p = 0.09) and Yitpi (− 15%, p = 0.06) but had no effect on J max . HS had no effect on V cmax or J max in either cultivar (Fig. 5i-l; Tables S1, S2, S3). V cmax and J max correlated well (r 2 = 0.75, p < 0.001) across treatments (Fig. 6b).

Fig. 3
Response of leaf gas exchange parameters to eCO 2 under non-HS conditions. Measurements were made at 25 °C before each harvest (T1, T2 and T3) for CO 2 assimilation rates (a, b) and stomatal conductance (c, d) in Scout (Circles) and Yitpi (Triangles). Plants were grown and measured at aCO 2 (blue solid cultivars), grown and measured at eCO 2 (red solid cultivars), and grown at eCO 2 and measured at 400 μL CO 2 L −1 (red dashed cultivars). Statistical significance levels (t-test) for the growth condition within each cultivar are shown and they are: * = p < 0.05; ** = p < 0.01: *** = p < 0.001

eCO 2 similarly stimulated wheat biomass and grain yield under non-HS conditions
The increase in plant biomass at eCO 2 depended on the growth stage (Figs. 6, 7; Tables S4, S5). However, the overall stimulation was not different between the two cultivars as evident from the non-significant eCO 2 x cultivar interaction at all harvests (Table S4). By T3 (anthesis), when both cultivars were still within the exponential growth stage, eCO 2 stimulated plant biomass of Yitpi (+ 29%, p < 0.001) and Scout (+ 9%, p < 0.001) under control conditions. The number of tillers, total leaf area, mean leaf size or leaf mass area were not significantly affected by eCO 2 in either cultivar ( Fig. 7; Tables S4, S5). eCO 2 increased allocation to stem relative to leaf biomass, particularly in Yitpi. Accordingly, there was a strong correlation across treatments between Fig. 4 Response of photosynthesis and chlorophyll fluorescence to HS in Scout and Yitpi grown at aCO 2 or eCO 2 . CO 2 assimilation rates (a-d) and dark-adapted chlorophyll fluorescence, Fv/Fm (e-h) were measured at growth CO 2 and 25 °C in Scout (Circles) and Yitpi (Tri-angles). Open and closed symbols represent control and HS plants, respectively. In addition, plants were measured at 35 °C (*) during both HS cycles 1 3 stem and leaf biomass (r 2 = 0.83, p < 0.001) and between total biomass and leaf area (r 2 = 0.83, p < 0.001) in Scout but not in Yitpi. However, the two cultivars followed common relationship for root versus shoot biomass (r 2 = 0.41, p < 0.001) and leaf area versus leaf number (r 2 = 0.82, p < 0.001) across all treatments suggesting no effect of cultivar, eCO 2 or HS on these common allometric relationships ( Figure S4; Table S5).
At the final harvest T4 (seed maturity), eCO 2 enhanced biomass and equally stimulated grain yield by increasing grain number in both cultivars (+ 64% in Scout and + 50% in Yitpi) under control conditions only (Fig. 8a-d; Tables S4, S5, S6). Harvest index was not directly affected by any treatments but showed a significant interaction (p < 0.05) between CO 2 and cultivar, such that HI was higher in Yitpi under eCO 2 (Tables S5 and S6).

eCO 2 did not stimulate the grain yield of HS plants
At T3, moderate HS (34-38 °C) applied under wellwatered conditions and 60% RH during the vegetative (HS1 applied after T1) and flowering (HS2 applied after T2) stages had no significant impact on plant biomass of either wheat cultivar or CO 2 treatment. By T4, there were significant HS × CO 2 × cultivar interactions (p < 0.01) for biomass and grain yield. HS1 + 2 reduced the biomass and grain yield of eCO 2 -grown Scout relative to aCO 2 -grown counterparts. Unlike control plants, the biomass and yield of HS plants were not enhanced by eCO 2 (Fig. 8; Tables S4, S5).

eCO 2 reduced grain N in Yitpi but not in Scout
Neither eCO 2 or HS had a significant effect on flag leaf N content in either cultivar at T2 or T3, but eCO 2 reduced aggregate leaf N content (− 18%) at T3 in Yitpi only (Cultivar × CO 2 p < 0.05) (Table S7). Yitpi had higher grain N content (+ 26%) than Scout in control plants grown at aCO 2 (Fig. 8g, h; Table S6). In control plants, eCO 2 significantly reduced grain protein content in Yitpi (− 18%, p < 0.05) but not in Scout due to significant cultivar × CO 2 interaction (p < 0.01), while HS had no effect on protein content in either cultivar (Fig. 8g, h; Table S6).

Fig. 5
Response of photosynthetic parameters to eCO 2 and HS at anthesis (T3) in Scout and Yitpi. CO 2 assimilation rate (a, b), stomatal conductance (c, d), photosynthetic water use efficiency (e, f) and photosynthetic nitrogen use efficiency (g, h) were measured at growth CO 2 . V cmax (i, j) and J max (k, l) were derived from ACi curves measured at 25 °C. Cultivars indicate means and shaded region is 95% confidence interval. Data shown for control (not exposed to any heat stress) and plants exposed to both heat stress cycles (HS1 + 2). Statistical significance levels (t-test) for the growth condition within each cultivar are shown and they are: * = p < 0.05; ** = p < 0.01: *** = p < 0.001

Two wheat cultivars with contrasting morphology and phenology, but similar photosynthesis and grain yield
The effects of future climate conditions, including eCO 2 , will depend on the environmental conditions (e.g., water and heat stress) and the crop's agronomic features. Here, we compared the interactive effects of eCO 2 and HS on two commercial wheat cultivars, Scout and Yitpi, with contrasting phenology and growth habit. Plants were grown under well-watered and fertilized conditions to remove any confounding effects of water or nutrient limitations on the eCO 2 or HS responses. RH was kept constant to minimize the negative impact of dry air during HS. Finally, we compared the effects of applying HS at the vegetative and flowering stages.
Free tillering Yitpi produced substantially more tillers, leaf area and biomass relative to the faster developing Scout. Accordingly, our first hypothesis predicted that Yitpi will have higher grain yield. The results only partially supported this hypothesis, because relative to Yitpi, Scout had higher harvest index (HI) due to its early maturing and senescing habit. While Yitpi initiated more tillers, a lower proportion of these tillers produced ears and filled grains. In contrast, Scout produced less tillers but flowered earlier which allowed enough time for all its tillers to produce ears and fill bigger grains by the final harvest. Hence, both cultivars had relatively similar yields due to bigger grain size in Scout and higher grain number in Yitpi. It is worth noting that some field trials have reported slightly higher grain yields in Scout than Yitpi (National variety trial report, GRDC, 2014). Our results are consistent with a previous study using different wheat cultivars with contrasting source-sink relationships which reported that the freely tillering cultivar "Silverstar" translated into more spikes while restricted tillering cultivar "H45" had more and heavier kernels per spike than "Silverstar" (Tausz-Posch et al. 2015). Thus, early vigor and maturity compared to high tillering capacity seem to be equally beneficial traits for high grain yield in the Australian environment.
The two wheat cultivars showed similar photosynthetic traits and response to temperature and eCO 2 . The temperature response of g m as well as the values recorded at 25 °C (0.25-0.31 mol m −2 s −1 bar −1 ) for Scout and Yitpi (Fig. 2b) were similar to what has been reported for wheat (0.39) and other crop species (cotton = 0.73, soybean = 0.49 and rice = 0.67) (Von Caemmerer and Evans 2015;Jahan et al. 2021). In contrast to our expectations that Scout would have higher WUE due to its selection based on a carbon isotope discrimination gene (Condon et al. 2004), both wheat cultivars showed similar iWUE under most measurement and growth conditions in this study (Fig. 5e, f; Table S2).

Elevated CO 2 stimulated photosynthesis but reduced photosynthetic capacity in both cultivars
Long term exposure to eCO 2 may reduce photosynthetic capacity due to lower amount of Rubisco in a process referred as 'acclimation' (Nie et al. 1995;Rogers and Humphries 2000;Ainsworth et al. 2003). Alternatively, eCO 2 may 'down-regulate' photosynthetic capacity by reducing Fig. 6 Relationships with leaf gas exchange and grain yield across treatments. CO 2 assimilation rate plotted as a function of stomatal conductance (a) (both aCO 2 and eCO 2 grown plants measured at 400 μL L −1 ), J max plotted as a function of V cmax (b) and grain protein plotted as a function of yield (c) in Scout (Circles) and Yitpi (Triangles). Ambient and elevated CO 2 are depicted in blue and red, respectively. Control and heat stressed plants depicted using open and closed symbols. Panel a depicts data for control, HS1, HS2 and both heat stresses (HS1 + 2), while panels b and cinclude only control and HS1 + 2 Rubisco activation or other regulatory mechanisms without affecting Rubisco content (Delgado et al. 1994). In the current study, eCO 2 similarly increased A growth (+ 21%) measured at growth CO 2 and reduced both A sat (− 12%) measured at common CO 2 (Fig. 3, T2) and V cmax (Fig. 5i, j) in both cultivars. In contrast, flag leaf N (and possibly Rubisco) was not significantly affected in either cultivar (Table S6). Hence, the wheat cultivars have likely undergone a photosynthetic downregulation (less rubisco activation)-rather than acclimation (reduction in rubisco content)-in response to eCO 2 (Delgado et al. 1994;Leakey et al. 2009). These results partially countered our second hypothesis suggesting that Yitpi will show less photosynthetic acclimation due to its higher sink capacity. The interaction of eCO 2 with plant traits are complex. On the one hand, eCO 2 is expected to cause less feedback inhibition on photosynthesis in plants with high sink capacity (Ainsworth et al. 2004). On the other, fastgrowing plants show a proportionally larger response to eCO 2 (Poorter and Navas 2003). Hence, high tillering in Yitpi and fast development in Scout both led to a relatively small observed photosynthetic downregulation in response to growth at eCO 2 . This allowed a sustained photosynthetic stimulation, which in turn led to a significant biomass and yield enhancement by CO 2 enrichment in both wheat cultivars (Figs. 6, 7). Photosynthetic responses of wheat in current study are in agreement with earlier enclosure studies Fig. 7 Response of plant growth and morphological traits to elevated CO 2 and HS: Total dry mass (a, b), tillers or number of tillers (c, d), leaf area (e, f), leaf number (g, h) and height (i, j) were measured at different time points across the life cycle of wheat cultivars Scout (Circles) and Yitpi (Triangles). Ambient and elevated CO 2 are depicted in blue and red color, respectively. Open symbols connected with solid cultivars and closed symbols connected with dashed cultivars represent control and HS plants, respectively. HS1 and HS2 depict the timing of HS applied at 10 and 15 weeks after planting respectively which generally have higher response to eCO 2 than the FACE studies (Kimball et al. , 1999Hunsaker et al. 1996;Osborne et al. 1998;Long et al. 2006;Cai et al. 2016).

Elevated CO 2 stimulated grain yield similarly in both wheat cultivars
In disagreement with our second hypothesis, eCO 2 similarly stimulated plant biomass and grain yield in early-maturing Scout and high tillering Yitpi (Figs. 6,7;Table S4,S5). In Scout, the biomass stimulation was associated with increased tillering (one extra tiller per plant). In contrast, Yitpi produced many tillers at aCO 2 and the additional fixed carbon at eCO 2 was allocated to the existing tillers. At seed maturity, eCO 2 stimulated grain yield similarly in both cultivars as a result of the trade-off between grain yield components (Dias de Oliveira et al. 2015). In particular, eCO 2 stimulated grain number in both cultivars, while grain size increased in Scout only ( Fig. 8; Table S6). Generally, eCO 2 stimulates grain yield by increasing the number of tillers and consequently, ears per plant (Zhang et al. 2010;Bennett et al. 2012) as a result Fig. 8 Response of total plant dry mass and grain parameters to growth at eCO 2 and HS at maturity (T4): Total dry mass (a, b), grain dry mass (c, d), grain number (e, f) and grain nitrogen (g, h) were measured at the final harvest. Cultivars indicate means and shaded region is 95% confidence interval. Ambient and elevated CO 2 are depicted in blue and red color respectively. Heat stress levels include plants not exposed to any heat stress (control) and both heat stresses (HS1 + 2). Statistical significance levels (t-test) for the growth condition within each cultivar is shown and they are: * = p < 0.05; ** = p < 0.01: *** = p < 0.001 1 3 of increased carbon supply due photosynthetic stimulation by eCO 2 , which has also been reported in FACE studies (Högy et al. 2009;Tausz-Posch et al. 2015;Fitzgerald et al. 2016). However, in our study, the increase in grain yield at eCO 2 was mainly due to the increase in the number of grains per ear. In line with our results, Tausz-Posch et al. (2015) reported comparable grain yield stimulation by eCO 2 in two different wheat cultivars with contrasting source-sink relationships. Moreover, grain yield of twenty wheat cultivars that differed in tillering propensity, water soluble carbohydrate accumulation, early vigor and transpiration efficiency responded similarly to eCO 2 in glasshouse settings (Ziska et al. 2004;Bourgault et al. 2013).

Elevated CO 2 reduced grain N in Yitpi only
Overall, there is a negative relationship between grain yield and quality Pleijel and Uddling 2012). Hence, increased grain yield at eCO 2 results in lower grain N and hence protein content (Seneweera and Conroy 1997;Bahrami et al. 2017). In our study, eCO 2 reduced grain N in Yitpi under control conditions. Scout was characterized by having larger grains which accumulated less N than Yitpi. Moreover, eCO 2 reduced total leaf N (− 18%) at T3 and grain N (− 17%) at T4 in Yitpi but not in Scout. This is consistent with the results from FACE study with same cultivars which reported − 14% reduction in N content by eCO 2 in above ground dry mass in Yitpi but not in Scout under well-watered conditions (Bahrami et al. 2017). The higher biomass accumulation in free tillering Yitpi may have exhausted the nutrient supply or nitrogen uptake may have slowed down (Carreras Navarro et al. 2020), such that further biomass stimulation by eCO 2 lead to a significant dilution in N content (Taub and Wang 2008).
Wheat cultivars with early vigour such as Scout have greater root biomass accumulation as well as greater early N uptake which may have avoided a negative effect of eCO 2 on leaf and grain N (Liao et al. 2004;Bahrami et al. 2017). Accordingly, Scout maintained a higher N utilization efficiency (grain yield per total plant N) relative to Yitpi under all treatments (Table S6). Increased grain yield is strongly associated with higher grain number per unit area (Zhang et al. 2010;Bennett et al. 2012) which dilutes the amount of N translocated per grain. Quality deterioration due to lower protein via reduced N is of critical concern in future high CO 2 climate considering that even additional supply of N does not prevent N dilution in grain under eCO 2 ). In addition, eCO 2 has strong detrimental effect on other nutrient availability and remobilization from leaves to grains (Tcherkez et al. 2020).

HS had little effects on wheat photosynthesis or yield at aCO 2
One of the key finding of this study was that the application of HS events (HS1, HS2 or HS1 + 2) was not detrimental to aCO 2 -grown wheat plants (Figs. 5, 7; Tables S4, S5, S6). Thus, our hypothesis that HS will reduce photosynthesis, biomass and yield at aCO 2 was rejected. This finding is in contrast to previously reported studies where HS reduced the grain yield and negatively affected the growth and development in wheat Nicolas 1996, 1998;Farooq et al. 2011;Coleman et al. 1991). During heat waves in the field, the vapor pressure deficit (VPD) increases and soil moisture decreases leading to lower stomatal conductance and consequently lower transpiration rate. Thus, plants are unable to cool down and leaf temperatures rise beyond optimum levels causing damage. The negligible effect of HS in our study could be explained by the ability of well-watered plants to maintain leaf temperature below damaging levels due to transpirational cooling (Perera et al. 2019;Deva et al. 2020) even with air temperatures reaching up to 38 °C. At moderate (~ 60%) relative humidity, there is sufficient water vapour gradient to sustain high transpiration rates when soil water is available, as was the case in our experiment. In most cases, g s was not significantly affected (Tables S1 and S2), and even slightly higher at T3 in HS-pants relative to the control (Fig. 5c, d) Well-watered crops can maintain grainfilling rate, duration and size under HS (Dupont et al. 2006), and high temperatures can increase crop yields if not exceeding critical optimum growth temperature (Welch et al. 2010). Also, in the current study, the night temperatures were not increased during HS which favors plant growth by reducing respiratory losses (Prasad et al. 2008).
In particular, HS did not elicit a direct negative impact on photosynthesis or chlorophyll fluorescence in either cultivar or CO 2 treatment. During HS, high temperature transiently reduced maximum efficiency of PSII (Fv/Fm) in both cultivars and CO 2 treatments ( Fig. 4e-h). However, unchanged Fv/Fm measured at 25 °C confirmed that photosynthesis did not suffer long-term damage during or after HS. Moreover, HS was not severe enough to negatively affect A growth measured at 25 °C. These results are corroborated by the insensitivity of V cmax and J max to HS (Fig. 5i-l), but contrast with previously reported studies where HS reduced photosynthesis in wheat at the vegetative ) and the flowering (Chavan et al. 2019;Balla et al. 2019) stages. HS lowers membrane thermostability by inducing reactive oxygen species (ROS) and altering the membrane protein structures, which lead to changes in the fluidity of the thylakoid membrane and separation of light harvesting complex from the photosystems (Wahid et al. 2007;Poudel 2020). We were unable to measure leaf temperatures in the current study, but we speculate that, in well-watered wheat plants growing at moderate RH, leaf temperatures might not have increased beyond damaging levels to the membranes during the HS events.
Repeated HS may result in priming which involves preexposure of plants to a stimulating factor such as HS (Wang et al. 2017) and enable plants to cope better with later HS events (Balla et al. 2021). However, there was no difference between HS applied at the vegetative (HS1) and/or flowering stage (HS2) in rejection of our fourth hypothesis, and this may additionally be due to the short term duration of the two HS cycles (3 days each). Hence, our study demonstrated the benign effect that HS has on crop yield when separated from water stress and plants are able to transpire.

HS precluded an eCO 2 response in biomass and grain yield
In our study, the impact of HS depended on the wheat cultivar and growth CO 2 (Tables S1 and S4). Elevated CO 2 and temperature interactions can be complex, dynamic and difficult to generalize as they can go in any direction depending on plant traits and other environmental conditions (Rawson 1992). Plant development is generally accelerated by increased temperature; eCO 2 can accelerate it further in some instances or may have neutral or even retarding effects in other cases (Rawson 1992).
While eCO 2 stimulated wheat biomass and grain yield under control (non-HS) conditions, HS precluded a yield response to eCO 2 in Yitpi and reduced biomass and yield in eCO 2 -grown Scout relative to aCO 2 -grown counterparts (Fig. 8). These results are in contrast with previous studies that reported similar wheat yield reduction at ambient or elevated CO 2 in response to severe (Chavan et al. 2019) or moderate HS (Zhang et al. 2018). The results also partially refuted our third hypothesis that HS may decrease yield more at aCO 2 than eCO 2 , while partially agreeing that HS will have a more negative impact on Scout relative to Yitpi, albeit for different reasons than what we originally suggested. The negative effect of HS on Scout biomass and grain yield at eCO 2 occurred despite the eCO 2 stimulation of A growth under HS (T3, Fig. 5). However, over the long term, A growth was stimulated in eCO 2 -grown Yitpi and not Scout (Fig. 5).
Lack of a biomass stimulation despite high photosynthetic rates during HS under eCO 2 could be due to the short duration of HS (3 days), which may not have been long enough to stimulate biomass gain. In addition, nutrient limitation at eCO 2 may have restricted the eCO 2 growth response. Typically, eCO 2 studies show reduced N content in wheat and other crops (Taub and Wang 2008;Leakey et al. 2009;Bahrami et al. 2017). Hence, the wheat plants may have exhausted available nutrients due to increased demand by growing sinks at eCO 2 , which may limited further stimulation by high temperature. HS may be more damaging at eCO 2 due to reduce transpirational cooling as a result of reduced g s at eCO 2 , leading to higher leaf temperatures. However, A growth increased in response to high temperature (35 °C) under eCO 2 but not under aCO 2 during HS (Fig. 4a-d), which refutes the suggestion of HS-damage to photosynthesis.
Higher g s during HS at moderate RH in well-watered conditions may increase A growth by increasing C i in both aCO 2 and eCO 2 grown plants. Furthermore, lower photorespiration under eCO 2 allows additional increase in A growth with temperature when measured at 35 °C relative to 25 °C (Long 1991). Under aCO 2 , photorespiration increases with temperature reducing A growth measured at 35 °C relative to 25 °C. Our results also point to a shift in T opt of photosynthesis (~ 24 °C at aCO 2 ) to higher temperatures for plants grown at eCO 2 (Sage and Kubien 2007). This would come about as a result of lower photorespiration at eCO 2 as well as the slight upregulation of photosynthetic rates observed in eCO 2 -grown Scout at the recovery stage of HS (Fig. 4a, c). However, at T3, A growth was similar between aCO 2 and eCO 2 grown plants (Fig. 5a) indicating the short-term nature of this photosynthetic upregulation.

Conclusions
The two wheat cultivars, Scout and Yitpi differed in growth and development but produced similar grain yield. Under control conditions, eCO 2 stimulated biomass and yield similarly in both cultivars. HS was not damaging to photosynthesis, growth, biomass or grain yield under wellwatered and moderate RH conditions. However, HS interacted with eCO 2 , leading to similar or lower biomass and grain yield at eCO 2 relative to both aCO 2 in plants exposed to HS. This interactive effect precluded the positive effects of eCO 2 in HS-plants. eCO 2 improved photosynthetic rates in control and HS plants. Also, high temperature stimulated photosynthesis under eCO 2 but not under aCO 2 during HS which suggests increased optimum temperature of photosynthesis at eCO 2 . We speculate that in the current study, HS plants were able to cool down using high transpiration which helped to maintain lower leaf temperatures despite high air temperatures during HS. The current study provides important insights into the effect of short-term moderate temperature increases in well-watered conditions under future elevated CO 2¸ potential role of transpirational cooling during HS and interactions between HS and eCO 2 which will be useful in breeding cultivars for future climate and improving crop model accuracy to predict crop performance in future high CO 2 environment with frequent heat waves.