Photosynthesis in newly developed leaves of heat-tolerant wheat acclimates to long-term nocturnal warming

Abstract

We examined photosynthetic traits of pre-existing and newly developed flag leaves of four wheat genotypes grown in controlled-environment experiments. In newly developed leaves, acclimation of the maximum rate of net CO₂ assimilation (A_n) to warm nights (i.e. increased A_n) was associated with increased capacity of Rubisco carboxylation and photosynthetic electron transport, with Rubisco activation state probably contributing to increased Rubisco activity. Metabolite profiling linked acclimation of A_n to greater accumulation of monosaccharides and saturated fatty acids in leaves; these changes suggest roles for osmotic adjustment of leaf turgor pressure and maintenance of cell membrane integrity. By contrast, where A_n decreased under warm nights, the decline was related to lower stomatal conductance and rates of photosynthetic electron transport. Decreases in A_n occurred despite higher basal PSII thermal stability in all genotypes exposed to warm nights: T_crit of 45–46.5 °C in non-acclimated versus 43.8–45 °C in acclimated leaves. Pre-existing leaves showed no change in A_n–temperature response curves, except for an elite heat-tolerant genotype. These findings illustrate the impact of night-time warming on the ability of wheat plants to photosynthesize during the day, thereby contributing to explain the impact of global warming on crop productivity.

Wheat photosynthesis and acclimation capacity are directly affected by night warming, despite the temporal separation of night warming from photosynthesis, with varying responses between newly developed and pre-existing leaves.

Introduction

Wheat (Triticum aestivum L) is the most widely cultivated crop, with >220 Mha cultivated worldwide. Wheat yield must increase by 0.83% per annum to meet the projected 44% increase in demand for it by 2050 (CGIAR Advisory Services, 2020). To achieve an environmentally sustainable increase in production of this magnitude, while also coping with rising global surface temperatures, represents a major challenge. Moreover, most research has focused on effects of elevated temperatures during the day on wheat, but rising temperatures have been more pronounced for daily minimums (i.e. night warming) than for daily maximums (Easterling et al., 1997; Alexander et al., 2006; Sillmann et al., 2013). Night warming is linked with a reduction in wheat yield in many regions including the USA, Mexico, and Argentina (Lobell et al., 2005; Prasad et al., 2008; García et al., 2015, 2016, 2018; Russell and Van Sanford, 2020; Fisher et al., 2022). Given the importance of wheat for global food production, it is vital that we develop a more complete understanding of the physiological basis for why nocturnal warming reduces wheat yields. Reflecting this, interest in wheat physiological responses to warm nights has increased, with reports on steady-state photosynthesis, acclimation of leaf and root respiration, and altered source–sink capacity at anthesis and grain filling (Fan et al., 2015, 2017; Impa et al., 2019, 2020; Bahuguna et al., 2022; Posch et al., 2022a). However, there remain key gaps in knowledge, especially in relation to the effect of warm nights on overall photosynthetic capacity, and whether key photosynthetic processes such as the critical temperature above which incipient damage to photosystem II (PSII) occurs (T_crit) acclimate to warm nights. PSII is considered the most heat-sensitive component of electron transport in the photosynthetic machinery, with heat stress resulting in the unfolding of protein complexes and loss of manganese from the oxygen-evolving complex (Enami et al., 1994).

According to the Farquhar, von Caemmerer, and Berry (FvCB model; Farquhar et al., 1980; von Caemmerer, 2000) biochemical model of photosynthesis, light-saturated A_n at a given temperature can be limited by: (i) the maximum capacity of Rubisco carboxylation (V_cmax) and internal CO₂ concentration (C_i); or (ii) the regeneration of ribulose 1,5-bisphosphate (RuBP) regeneration which relies on photosynthetic electron transport capacity (J) and irradiance. A_n can also be limited if triose phosphate use (TPU) restricts the recycling of phosphate within the chloroplast (Sharkey, 1985). V_cmax, J, and TPU can be estimated by fitting the FvCB model (Farqhuar et al., 1980) to plots of light-saturated A_n versus C_i. Light-saturated rates of A_n respond to short- and long-term changes in growth environment temperature (Atkin and Tjoelker, 2003; Way and Yamori, 2014). The thermal response of light-saturated A_n to short-term changes in growth environment temperature is curvilinear—increasing with a rise in temperature at suboptimal temperatures to a maximum rate (A_max) at the optimum temperature (T_opt), then decreasing at temperatures beyond T_opt. Longer term changes in the thermal environment cause either positive or detractive adjustments of light-saturated A_n. Positive adjustments increase A_max and/or shift T_opt towards the new, higher growth temperature, indicating acclimation, whereas detractive adjustments reduce A_max and/or T_opt (Sage and Kubien, 2007; Way and Yamori, 2014; Posch et al., 2019). Changes in the A_n–temperature response are controlled by different factors, including the temperature sensitivities of V_cmax and J (Hikosaka et al., 2005; Sage and Kubien, 2007; Lin et al., 2012), with alterations in V_cmax reflecting changes in the abundance and/or activation state of Rubisco (Scafaro et al., 2023). Another controlling factor, albeit indirectly, is stomatal conductance (g_s). At high temperatures, stomatal closure reduces water loss when the atmospheric vapour pressure deficit (VPD) rises, but this simultaneously reduces CO₂ uptake. In wheat, the A_n–temperature response to elevated daily temperatures varies from positive acclimation to detractive adjustments (Yamasaki et al., 2002; Cai et al., 2018, 2020; Coast et al., 2021a). Importantly, information on thermal acclimation of wheat photosynthesis to specifically warm nights is scarce (but see Turnbull et al., 2002 for work on a broadleaved tree species). It is likely that under night warming, wheat photosynthesis and acclimation capacity may be affected directly by constraining leaf biochemical components underlying photosynthesis (i.e. V_cmax and J), or indirectly by stomatal closure, or both. These suggestions remain untested.

The degree of physiological acclimation in leaves to warming varies substantially with the duration and intensity of warming, as well as the plant’s developmental stage. Leaves that develop under a new warming regime exhibit a higher degree of acclimation than pre-existing leaves that were formed under an earlier thermal regime and which then experience a sustained change in growth temperature (Loveys et al., 2003; Yamori et al., 2005; Rashid et al., 2020; Slot et al., 2021). Varying acclimation capacity with leaf development has also been shown in cold regimes (Hurry et al., 2005; Campbell et al., 2007). Limited reports exist about leaf development under warm conditions and the magnitude of photosynthetic acclimation in newly developed and pre-existing leaves. Determining these responses could be relevant in managing crop performance and useful in parameterizing carbon fluxes in Earth system models (Rogers et al. 2017).

T _crit, a common measure of photosynthetic heat tolerance (Geange et al., 2020; Lancaster and Humphreys, 2020), is estimated from temperature-dependent changes in minimum Chl a fluorescence (F₀) (Schreiber et al., 1975; Schreiber and Berry, 1977). T_crit correlates with the maximum temperature at which A_n approaches zero (Downton et al., 1984). Under warmer growth conditions, an increase in T_crit is considered evidence of acclimation and indicative of higher temperature tolerance. Acclimation of T_crit to warmer growth conditions has been shown in many non-crop, mostly woody species. For example, 62 Australian species acclimated seasonally with T_crit, increasing on average from winter to summer by 0.34 °C per °C increase in growth temperature (Zhu et al., 2018). Similar reports exist for Mediterranean and temperate European oak species (Quercus spp., Daas et al., 2008), and at a global scale for 218 plant species spanning seven biomes (O’Sullivan et al., 2017). Acclimation of T_crit is assumed to increase the upper thermal threshold of photosynthesis by relaxing the limits of photosynthetic electron transport. As such, the capacity to acclimate T_crit might underlie photosynthetic performance under warm nights. However, it is not known whether T_crit acclimates in response to warm nights.

Metabolites function as regulatory agents, compatible solutes, antioxidants, and reductants in adaptation to thermal stress. Warming alters the abundance of primary metabolites [sugars, sugar alcohols, organic acids, amino acids, or tricarboxylic acid (TCA) cycle intermediates] in leaves of stress-sensitive cereals (Glaubitz et al., 2014; Impa et al., 2019; Schaarschmidt et al., 2020; Rashid et al., 2021). Increases in the abundance of organic acids and TCA cycle components in heat-sensitive genotypes under heat stress often reflect impairment to CO₂ assimilation, regulation of the TCA cycle, and amino acid biosynthesis, and these processes are central to plant photosynthesis. Deciphering changes in the metabolic phenotype of wheat in response to warm nights could aid our understanding of plant responses to climate change and provide input for developing adaptation tools for crop production in a warmer world.

To address some of the above issues, we compared four wheat genotypes (including a commercial Australian cultivar and an elite heat-tolerant genotype) under night temperatures of 15, 20, or 25 °C for different durations (5–7 d at anthesis or 9–13 weeks prior to and including anthesis). Our primary objectives were to: (i) quantify the acclimation capacity of photosynthesis and T_crit to warm nights; and (ii) assess changes in the wheat metabolic phenotype under warm nights. In addition, we set out to compare effects of warm nights on the temperature dependencies of photosynthetic parameters in newly developed versus pre-existing leaves. The results indicate that plants stressed by warm nights reduced photosynthetic performance via down-regulation of J even while exhibiting high T_crit (i.e. high PSII thermal stability). By contrast, tolerance of warm nights (in terms of photosynthesis) was marked by improved A_n linked to acclimation in V_cmax, PSII thermal stability, and an increase in leaf metabolic signatures for monosaccharides.

Materials and methods

To assess the extent to which photosynthetic capacity of pre-existing and newly developed leaves of wheat respond to warm nights, two experiments were conducted. In Experiment I, acclimation of T_crit, instantaneous temperature response functions of photosynthetic capacity [light-saturated A_n, V_cmax, J₁₅₀₀ (J measured at a photosynthetic photon flux of 1500 µmol m^–2 s^–1), and TPU], and changes in the metabolite profile of plants that developed from seedling to anthesis (9–13 weeks of night warming) were compared at three night temperatures (15, 20, and 25 °C) and a common day temperature of 26 °C. These temperature treatments are reflective of conditions in some wheat-growing regions including parts of the USA (Hein et al., 2020) and India (Bahuguna et al., 2022). In Geraldton and Northampton on the northern tip of the West Australian wheatbelt, the mean minimum temperatures between October and December (i.e. during reproductive development to harvest) range from 12 °C to 17 °C. Similar temperatures, in controlled-environment settings, enable dissection of the physiological mechanism of tolerance and application of high-throughput phenotyping platforms to screen for stress tolerance; for example, see Wang et al. (2022) with control nights of 21 °C and high nights of 28 °C. Experiment II was a repetition of Experiment I except plants with mature (pre-existing) flag leaves which had grown under 15 °C nights were exposed to a shorter period of warming (5–7 nights at 20 °C and 25 °C environments) at anthesis. As such Experiment I focused on newly developed leaves and Experiment II on pre-existing leaves. For both experiments, all measurements were taken when plants were at the same developmental stage—anthesis (i.e. between Zadok scale ZS60 and ZS69; Zadoks et al., 1974).

Plant materials, management, and temperature treatments

Four wheat (T. aestivum L.) genotypes were used for this study: Mace (pedigree Wyalkatchem/Stylet/Wyalkatchem), a well-adapted, commercial, Australian cultivar; ACIAR09PBI C38-150C-DH9 (pedigree PBW343+L24+LR28/LANG; henceforth 1704), a heat-susceptible genotype; ACIAR09PBI C27-0C-0N-3N (pedigree DBW16/ANNUELLO; henceforth 1898), also a heat-susceptible genotype; and 8:ZW11 [pedigree D67.2/P66.270//AE.SQUARROSA(320)/3/CUNNINGHAM/4/VORB; henceforth 2254], a heat-tolerant elite genotype. Mace was a benchmark variety for yield in Western Australia where it was widely grown. It accounted for 66% of plantings between 2015 and 2016 (Zaicou-Kunesch et al., 2017), although this reduced to 31% by 2018. Mace can cope with high temperatures (Bokshi et al., 2022). The two heat-susceptible genotypes (1704 and 1898) exhibit low PSII thermal stability (i.e. low basal T_crit) when grown under non-heat-stressed conditions but also the capacity to acclimate T_crit to warmer growth regimes in field conditions (Posch et al., 2022b). Genotype 2254 was developed by the International Maize and Wheat Improvement Centre (CIMMYT) and, like Mace, it is considered heat tolerant (Posch et al., 2022a). Unpublished yield data from field experiments conducted under heat stress conditions in Ciudad Obregon, Mexico and Narrabri, Australia by Professor Richard Trethowan of the University of Sydney, Australia, support the heat-tolerant classification of genotype 2254. These were chosen for their varied yield, and agronomic and physiological performance under heat stress conditions in fields across wheat-growing regions of Australia.

Seeds were germinated on moist filter papers in Petri dishes. One-week-old seedlings were sown into 6 litre plastic pots (one seedling per pot) filled with Martins mix (Martins Fertilizers Ltd, Yass, NSW Australia). The Martins mix was enriched with 4 g l^–1 Osmocote® OSEX34 EXACT slow‐release fertilizer (Scotts Australia, Bella Vista, NSW, Australia) and treated at 63 °C for 1 h prior to filling pots. Potted seedlings were transferred into temperature-controlled growth chambers (Thermoline, Wetherhill Park, Australia) at the Controlled Environment Facilities of the Research School of Biology, The Australian National University, Canberra, Australia (ANU). Genotypes were arranged randomly in each chamber. Growth chambers were maintained at day/night temperatures of 26/15, 26/20. and 26/25 °C either throughout plant growth (lasting 9–13 weeks; Experiment I) or only at anthesis (lasting 5–7 d; Experiment II). For the latter conditions, day/night temperature prior to anthesis was maintained at 26/15 °C. In all growth chambers, relative humidity varied from 30% during the light period to 70% during the dark, and [CO₂] was at ambient ~400 μmol mol^–1 (38.4 Pa, considering a mean atmospheric pressure of 96 kPa at ANU). Lighting was supplied by 1000 W metal halide lamps (Multi-Vapor®; GE Lighting, Derrimut, Australia) producing photosynthetically active radiation at a plant height of 720–750 µmol m^–2 s^–1. A 12 h photothermal regime was maintained throughout plant growth. General plant management followed the established protocol of the Controlled Environment Facilities at ANU. Eight replicate plants of each genotype were assigned to each night temperature treatment.

Determination of flag leaf photosynthetic heat tolerance

T _crit was estimated according to the method of Schreiber and Berry (1977) and recently applied by Zhu et al. (2018), Arnold et al. (2021), and Coast et al. (2021b). A detailed description of the method is given in Coast et al. (2021b); briefly, discs excised during the day from the middle section of detached dark-adapted leaves were exposed to a temperature ramp at a constant rate of 1 °C min⁻¹ from 20 °C to 65 °C with simultaneous continuous measurement of F₀ taken. T_crit was calculated as the intersection point of two regression lines extrapolated from the flat and steep portion of the F₀–temperature response curve. T_crit was determined for all four genotypes in Experiment I, and for two genotypes (Mace and 1704) in Experiment II.

Gas exchange measurements

Plants in growth chambers were moved into a temperature-controlled cabinet (Thermoline Model‐1175-SD-1SL, Thermoline Scientific, Smithfield, NSW, Australia) for all gas exchange measurements. Gas exchange measurements were conducted on intact flag leaves of the main tiller identified as the first tiller to reach anthesis. Five LI-COR portable photosynthesis systems (LI-6400XT, LI-COR Inc., Lincoln, NE, USA) were used for gas exchange measurements. The LI-COR units were fitted with 6 cm² leaf chambers with a red–blue light source (6400-18 RGB Light Source, LI-COR). Leaves were exposed to saturating irradiance of 1500 μmol photons m⁻² s⁻¹ within the LI-COR leaf chamber, with both the LI-COR leaf chamber/block and the whole plant placed within the temperature-controlled cabinet. The LI-COR leaf chamber was initially set to 20 °C, reference line atmospheric [CO₂] of 400 ppm, a flow rate of 500 μmol s⁻¹, and relative humidity maintained between 40% and 75%. After flag leaves had been exposed to these conditions in the leaf chamber for at least 5 min and following equilibrium (stable readings for at least 1 min), A_n was determined. Thereafter photosynthetic [CO₂] response curves (A:C_i curves) were generated, at a constant irradiance of 1500 μmol photons m⁻² s⁻¹, by varying the [CO₂] inside the LI-COR leaf chambers as follows: 30, 50, 100, 150, 250, 400, 400, 600, 800, 1000, 1200, 1400, and 400 μmol mol⁻¹. The A:C_i curves were repeated with the leaves exposed to measurement temperatures of 25, 30, 35, 40, and 50 °C. The temperature setting of the cabinet was adjusted to enable the LI-COR leaf chamber/block to achieve the desired measurement temperatures. Plants were kept well irrigated throughout the measurement period to avoid water stress. In both Experiment I and II, four replicate plants per night temperature treatment were used. All gas measurements were taken between 08.00 h and 17.00 h during weeks 9–13 for Experiment I or days 5–7 for Experiment II. The ranges (9–13 weeks, and 5–7 d) in period reflect differences in time to anthesis of the different genotypes.

Modelling photosynthetic capacity

Model parameters for each growth and measurement temperature were estimated following the FvCB model and using the Plantecowrap package (Stinziano et al., 2018) in the R computing environment (R Development Core Team, 2021). Plantecowrap eliminates potential bias associated with manually determining the concentration of intercellular CO₂ (C_i) where one limitation transitions to another, as is the case with the commonly used Sharkey Excel spreadsheet (Sharkey et al., 2007), and allows for species-specific kinetic parameters to be pre-defined. The kinetic parameters used in modelling photosynthetic capacity for wheat were: mesophyll conductance at 25 °C (g_m=5.5 μmol m⁻² s⁻¹ Pa^–1); activation energy of mesophyll conductance (E_a=47.65 kJ mol^–1); apparent Michaelis–Menten constant for Rubisco carboxylation in 21% oxygen (K_air=772 µmol mol^–1); activation energy of K_c (93.72 kJ mol^–1); photorespiratory CO₂ compensation point or Gamma star at 25 °C (Γ*=37.74 µmol mol^–1, equivalent to µbar bar^–1); and Gamma star activation energy (24.42 kJ mol^–1). The temperature responses of V_cmax, J₁₅₀₀, and TPU were modelled using non-linear least squares fit of the Arrhenius temperature response function accounting for deactivation (Medlyn et al., 2002; Kattge and Knorr, 2007). The deactivation energy (E_d) was assumed to be 200 kJ mol^–1; the activation energy (E_a) and entropy factor (ΔS) were estimated from iterative fits of the model.

Gas chromatography–mass spectrometry metabolite analysis

All leaf samples used for metabolite analysis were collected during the day (between 08.00 h and 17.00 h) and within an hour of completing gas exchange measurements. Metabolite extraction was conducted using a GC-MS procedure described in Rashid et al. (2021). Derivatization was performed using the MPS2 XL-Twister autosampler (Gerstel GmbH & Co. KG, Mülheim an der Ruhr, Germany), and metabolite samples were analysed on an Agilent GC/MSD system composed of an Agilent GC 6890N gas chromatograph (Agilent Technologies) fitted with a 7683B Automatic Liquid Sampler (Agilent Technologies) and 5975B Inert MSD quadrupole MS detector (Agilent Technologies). The gas chromatograph was fitted with a 0.25 mm (i.d.), 0.25 μm film thickness, 30 m Agilent FactorFour VF-5ms capillary column with a 10 m integrated guard column (Agilent Technologies). Raw GC-MS data were converted using the GC/MS Translator (ver 1.0, Agilent Technologies, Inc., Santa Clara, CA, USA) then processed (peak detection, retention time alignment, and relative quantitation) with MS-DIAL (ver 4.60; Tsugawa et al., 2020). Metabolites were identified by comparing mass spectral features against available spectral libraries from the Golm Metabolome Database (Kopka et al., 2004). Metabolite abundance values were normalized against the averaged signal of an internal standard ([¹³C₅]valine) and the sample fresh mass, followed by weighting against the average measured signal across all samples for each compound.

Statistical analysis

Data preparation, analysis, and visualization were performed with R (R Development Core Team, 2021) using the packages tidyverse (Wickham et al., 2019), tidyr (Wickham, 2021), dplyr (Wickham et al., 2021), ggplot2 (Wickham, 2016), FactoMineR (Lê et al., 2008), factoextra (Kassambara and Mundt, 2020), ade4 (Chessel et al., 2004; Dray and Dufour, 2007; Dray et al., 2007; Bougeard and Dray, 2018; Thioulouse et al., 2018), nls2 (Grothendieck, 2013), minpack.lm (Elzhov et al., 2016), vegan (Oksanen et al., 2020), lattice (Sarkar, 2008), latticeExtra (Sarkar and Andrews, 2022), car (Fox and Weisberg, 2019), AICcmodavg (Mazerolle, 2020), and agricolae (de Mendiburu, 2021).

Experiments I and II were analysed separately. Gas exchange data were checked for outliers prior to analysis. Differences in treatment means for T_crit, gas exchange parameters (A_max and g_s), and modelled photosynthetic capacities (V_cmax and J₁₅₀₀ at a standardized temperature of 25 °C; T_opt of A_n, V_cmax, J₁₅₀₀, and TPU; and V_cmax, J₁₅₀₀, and TPU at T_opt) were tested using ANOVA and treatments separated using Fisher’s LSD or Tukey’s HSD. Comparative regressions were used to test the effects of night temperature and genotype on the temperature response curves of A_n, V_c, J, and TPU. The regression models tested fits of a common line (the simplest model) and separate lines (the advanced model) for the genotype and night temperature. The simplest model assumed a common intercept and curvature, whereas the advanced model assumed differences in intercept or curvature, or both. F-tests were used to select the model with the best fit. The A_max (i.e. A_n at T_opt) and T_opt of A_n were derived from equations of the second-order polynomials that best described the instantaneous A_n–temperature relationships. Metabolite abundance data were analysed using principal component analysis (PCA) and a non-parametric, permutation tests-based, multivariate ANOVA (Anderson, 2001). This form of analysis generates test statistics analogous to Fisher’s F-ratio, and P-values are obtained using permutations.

Results

T _crit of newly developed leaves (i.e. leaves that developed entirely under conditions where the prevailing night temperature was elevated) varied between genotypes (P<0.001) but did not significantly respond to night warming (Fig. 1A). Across night temperatures, mean T_crit of the heat-susceptible genotype 1704 (45.7 °C) was higher (P<0.05) than the T_crit of Mace (43.8 °C) and the elite genotype 2254 (44.6 °C), but not of genotype 1898 (45.2 °C, P=0.503). In pre-existing leaves (i.e. leaves that had developed under controlled-environment conditions and then exposed to 5–7 consecutive warm nights), responses of both genotypes to night warming differed: T_crit of Mace increased with warm nights whereas T_crit of 1704 did not significantly change, resulting in a significant genotype by night temperature interaction (P=0.008, Fig. 1B). Across night temperatures, Mace and 1704 differed in T_crit (P=0.015, for main effect of genotype). However, the effect of night temperature alone was not significant (P=0.107).

Fig. 1.

Photosynthetic high temperature tolerance (T_crit) of newly developed leaves of four wheat genotypes (top panel) and pre-existing leaves of two wheat genotypes (bottom panel) at night temperatures of 15, 20, and 25 °C. The four genotypes were the heat-susceptible 1704 and 1898, and the heat-tolerant Mace and 2254. Means are of 3–8 plants.

When measured at ambient CO₂, instantaneous temperature responses of light-saturated rates of A_n in newly developed wheat leaves responded to night warming (Fig. 2A; Table 1). The responses were characterized by maintaining or increasing the maximum rate of A_n (i.e. A_max at T_opt) or the optimum temperature of A_n (T_opt of A_n). The largest increase with night warming from 15 °C to 25 °C in A_max at T_opt and T_opt of A_n were in genotype 2254 and 1898, respectively. However, for genotype 1704, estimates of A_max at T_opt and T_opt of A_n for plants grown at 20 °C and 25 °C nights could not be determined due to the linear declines in A_n within the limited measurement temperature range studied. For genotype 1704 alone, comparison of regressions for the different night temperatures showed variation [F=65.07_{(4, 68)}, P<0.001, and R²=0.79], but this was due to a lower A_n–temperature response at 20 °C compared with both 15 °C and 25 °C nights. Stomatal conductance (g_s) also responded to night warming, and was generally curvilinear with increasing measurement temperature (Fig. 2B). There was a clear interaction effect of genotype by night temperature on g_s, with lower g_s for genotypes 1704 and 1898 under warmer nights, and the opposite for Mace and genotype 2254. The short-term leaf temperature response of g_s and A_n were curvilinear. Stomatal conductance was marked by an initial decline as temperature increased from 20 °C to ~30 °C, followed by an uptick at higher temperatures. There were minimal changes in photosynthesis for leaves that pre-existed when night temperatures were altered (Fig. 2C).

Fig. 2.

At ambient CO₂: instantaneous temperature response curves of light-saturated net assimilation (A_n; top panels) and stomatal conductance (g_s; bottom panels) in wheat flag leaves that developed (newly developed) at a day temperature of 20 °C and night temperatures of 15, 20, or 25 °C. Individual panels are presented for the heat-susceptible genotypes 1704 (A) and 1898 (B), and the heat-tolerant genotypes 2254 (C) and Mace (D). For pre-existing leaves that only experienced night warming after spikes emerged and flowered, the instantaneous A_n–temperature responses were unaffected by night warming (Supplementary Fig. S1). Lines are polynomial regression curves fitted to the data. The shaded regions indicate confidence intervals for the fitted polynomial curves. Means are of 2–4 plants.

Table 1.

Estimated optimum temperature (T_opt) of light-saturated net assimilation at ambient CO₂ (A_n) in leaves of four wheat genotypes, and A_n at T_opt (i.e. maximum A_n or A_max) derived from regression models of instantaneous temperature responses of A_n

	Night temperature	Genotypesa
		1704	1898	2254	Mace
Newly developed leavesb
T opt of An	15	16.0	22.6	29.9	24.6
	20	---	22.3	26.0	25.2
	25	---	26.2	24.7	11.7
A max (An at Topt)	15	30.5	26.0	25.2	28.1
	20	---	20.5	25.6	31.9
	25	---	19.3	35.1	31.6
Pre-existing leaves
T opt of An	15–25	18.6	---	---	24.4
A max (An at Topt)	15–25	21.0	---	---	23.4

^a The heat-susceptible genotypes are 1704 and 1898, and the heat-tolerant genotypes are 2254 and Mace.

^b The final models for newly developed leaves of each genotype were based on separate regression fits for the three different night temperatures, whereas for pre-existing leaves models were common regressions across the night temperature. --- The best fit was linear, T_opt could not be determined within the limits of temperature range tested.

V _cmax of newly developed leaves increased exponentially with rising leaf temperatures before peaking then declining at measurement temperatures above 36–47 °C for V_cmax and 29–34 °C for J₁₅₀₀ (Fig. 3; Table 2), which enabled a classical Arrhenius temperature response with a deactivation component to be fit. T_opt of V_cmax was unresponsive to night warming (P=0.112, Table 2) but modelled rates of V_cmax at T_opt increased significantly, between 24% and 43% (Fig. 3; Table 2). Below leaf temperatures of 36 °C, V_cmax and its response to temperature were similar between the genotypes and not affected by night temperature. Thus, the impact of night warming was most evident as an increase in Rubisco carboxylation capacity of leaves developed under the warmest nights.

Fig. 3.

Temperature response curves of the maximum CO₂ carboxylation capacity (V_cmax), photosynthetic electron transport capacity (J₁₅₀₀), and triphosphate utilization (TPU) of leaves that developed (newly developed) at a common day temperature of 20 °C and night temperatures of 15, 20, or 25 °C for four wheat genotypes. The genotype name/ID are indicated in the panels for both the heat-susceptible genotypes 1704 (A) and 1898 (B), and the heat-tolerant genotypes 2254 (C) and Mace (D). V_cmax was iteratively fit with an Arrhenius equation with parameters given in Table 2. J₁₅₀₀ and TPU were fit with quadratic functions. Plots for pre-existing leaves of Mace and 1704 are presented in Supplementary Figure S2. Means are of 3–4 plants. Error bars represent the standard error of the mean.

Table 2.

Mean estimates (±SD) of the optimum temperature (T_opt) and maximum capacity of V_cmax, J₁₅₀₀, and TPU in newly developed leaves of four wheat genotypes at three different night temperatures

Genotypea	Night temperature (°C)	T opt of Vcmax (°C)	V cmax at Topt (μmol m–2 s–1)	T opt of J1500 (°C)	J 1500 at Topt (μmol electrons m–2 s–1)	T opt of TPU (°C)	TPU at Topt (μmol m-2 s-1)
1704
	15	38.1 ± 7.1	310 ± 32	29.4 ± 1.8	222 ± 8	29.6 ± 1.7	15 ± 2
	20	38.7 ± 0.2	217 ± 2	31.5	160	28.9	11
	25	41.6 ± 0.6	407 ± 28	32.8 ± 0.2	192 ± 17	32.6 ± 4.0	12 ± 4
1898
	15	37.3 ± 3.4	242 ± 26	39.9 ± 1.6	195 ± 15	29.5 ± 2.0	13 ± 2
	20	---	---	29.2 ± 1.4	179 ± 35	29.0 ± 0.9	12 ± 1
	25	46.1 ± 4.8	346 ± 94	33.7 ± 6.1	181 ± 14	32.5 ± 7.0	11 ± 7
2254
	15	39.3 ± 5.1	288 ± 32	29.2 ± 0.1	181 ± 37	28.5 ± 0.1	12 ± 3
	20	46.6	436 ± 2	34.1 ± 0.3	208 ± 20	32.6 ± 2.6	13 ± 2
	25	38.9 ± 2.3	498 ± 28	31.6 ± 1.5	259 ± 6	30.3 ± 2.3	16 ± 2
Mace
	15	36.2 ± 6.2	326 ± 58	29.8 ± 3.4	213 ± 49	28.5 ± 4	13 ± 4
	20	39.1 ± 5.6	389 ± 33	30.8 ± 0.8	251 ± 23	30.1 ± 0.9	16 ± 1
	25	46.0 ± 9.8	569 ± 176	32.3 ± 6.5	225 ± 31	30.7 ± 6.9	14 ± 7
Levels of significance (P-values)
	Genotype (G)	0.963	0.186	0.979	0.024	0.961	0.178
	Night temperature (NT)	0.112	0.034	0.156	0.851	0.330	0.942
	G x NT	0.568	0.796	0.854	0.047	0.922	0.078

^a The heat-susceptible genotypes are 1704 and 1898, and the heat-tolerant genotypes are 2254 and Mace. --- The model fit could not be resolved. Highlighted in bold are P-values <0.05.

J ₁₅₀₀ and TPU displayed a curvilinear response to leaf temperature (Fig. 3E, F). The genotype by night temperature effect on J₁₅₀₀ was significant (P=0.047, Table 2). This was due in part to J₁₅₀₀ decreasing with night warming for genotypes 1704 and 1898 but not being detrimentally affected by night warming in Mace and genotype 2254. J₁₅₀₀ at T_opt for Mace and 2254 increased by 5–30% with warmer nights (Table 2). By contrast, night warming reduced J₁₅₀₀ in 1704 and 1898 by 7–28% (Table 2). TPU was largely unresponsive to night warming and did not differ between genotypes. Like T_opt of V_cmax, T_opt of J₁₅₀₀ and TPU were not significantly changed under warmer nights. However, T_opt values of photosynthetic capacity were high for V_cmax (range 37–47 °C) relative to J₁₅₀₀ (range 29–34 °C) and least for TPU (range 11–16 °C). As with net assimilation rates, photosynthetic capacity in terms of V_cmax, J₁₅₀₀, and TPU were not influenced by night warming in pre-existing leaves of both genotypes (P>0.05, Supplementary Fig. S1; Supplementary Table S1).

To determine whether light-saturated A_n becomes RuBP limited in newly developed leaves, as suggested by the decline in J₁₅₀₀, A_n was plotted against C_i for the different night temperatures. The curvilinear responses were fitted using the modelled carboxylation and RuBP regeneration limitation rates (Fig. 4). At C_i corresponding to an ambient atmospheric CO₂ concentration of 400 µmol mol^–1, light-saturated A_n was reduced in 1704 and 1898 in response to a greater RuBP regeneration limitation at 20 °C and 25 °C nights (see black outlined boxes in Fig. 4). Genotypes 2254 and Mace either maintained or increased their RuBP regeneration capacity with night warming and consequently A_n at 400 µmol mol^–1 across night treatments.

Fig. 4.

A–C_i curves and corresponding C₃ photosynthesis model fits for newly developed leaves of wheat genotypes 1704 (A, heat-susceptible), 1898 (B, heat-susceptible), 2254 (C, heat-tolerant), and Mace (D, heat-tolerant). Plants were treated to a common day temperature of 20 °C and night temperatures of 15 (purple lines and shapes), 20 (green lines and shapes), or 25 °C (yellow lines and shapes). Solid, dashed, and dotted curves represent predicted carboxylation-limited A_n rates, RuBP regeneration-limited A_n rates, and triphosphate utilization- (TPU) limited A_n rates, respectively. The black outlined box encapsulates the values corresponding to ambient CO₂ concentration of 400 μmol mol^–1. Values are the means of pooled measurements at leaf temperatures of 25 °C and 30 °C and model parameters standardized to 28 °C. Means are of 4–6 plants.

Across the four genotypes and night temperatures of Experiment I, a total of 162 metabolites were identified in newly developed leaves. PCA determined relationships among individual metabolites and across treatments. Metabolites were assigned to classes as listed in Fig. 5 (metabolite relative abundance and classification are presented in Supplementary Table S2). The response of metabolites generally followed a similar pattern for a particular class, and often a class responded independently of other classes. Most notably, organic acids and amino acids clustered in the right half of the PCA (positive Dim1), while many carbohydrates clustered in the opposite left half. Thus, organic and amino acid abundance were in general positively related to one another and negatively correlated with many carbohydrates. In terms of night treatment, there was a shift to a more distinct separation of metabolite abundance in genotypes that were more physiologically responsive and tolerant to the warmer nights (Fig. 5). For example, the most sensitive and non-responsive genotype 1704 had little separation of metabolite profiles among night temperatures—evident in overlapping night temperature-dependent metabolite distribution clouds in Fig. 5. By contrast, the two most heat-responsive and tolerant genotypes (Mace and 2254) had a distinct special separation of the 25 °C night treatment metabolites, represented by special separation at negative Dim1 and Dim2 values (i.e. the lower left corner of PCA graphs). Genotype 1898 had intermediate separation of metabolite profiles in response to night warming, shifting to negative Dim1 values. Interestingly, many of the same metabolites positively correlated with warmer nights for the three genotypes (1898, 2254, and Mace) that responded to warmer nights (Fig. 6). Of these metabolites that responded positively, most were monosaccharides, namely 3,6-anhydro-d-hexose, fructose, mannose, mannose-6-phosphate, and tagatose (metabolites 14, 55, 113, 114, and 149, respectively). The other metabolites that consistently responded to the warmest night of 25 °C were the saturated 18:0 and 16:0 fatty acids octadecanoic acid and palmitic acid (metabolites 124 and 129).

Fig. 5.

Principal component analysis (PCA) decomposition of metabolite responses in newly developed leaves of wheat genotypes to three different night temperature. The genotype name/ID are indicated in the panels for both the heat-susceptible genotypes 1704 and 1898, and the heat-tolerant genotypes 2254 and Mace. Metabolites are colour-coded by functional class. The direction and length of vector arrows provide information on the contribution each metabolite makes to dimensions 1 (Dim 1) and 2 (Dim 2), and the inter-relatedness of metabolites among themselves (opposing directions represent strong negative correlations while arrows of similar direction represent positive correlations), as well as in relation to night temperatures (the latter represented by blue, green, and red shaded areas). Arrow numbers correspond to individual metabolites which are listed in Supplementary Table S1, along with their relative abundance.

Fig. 6.

Individual metabolites that consistently responded positively to the warmest night of 25 °C in the genotypes 1898 (open circles; heat-susceptible), 2254 (open squares; heat-tolerant), and Mace (open triangles; heat tolerant). Metabolites presented are monosaccharides (A–D) and fatty acids (E, F). The genotype 1704 was excluded from this analysis as its metabolite profile showed no significant response to night temperature, evident in the PCA plots presented in Fig. 5. The metabolite numbers refer to their corresponding labels in the PCA plots.

Discussion

In this study, we showed that wheat photosynthesis and acclimation capacity are directly affected by night warming, with varying genotypic responses and differences between newly developed and pre-existing leaves. For most genotypes, metabolites for monosaccharides and saturated fatty acids had the strongest positive correlation to warmer nights. Osmotic regulation and changes in fatty acid saturation with warming, which is consistent with changing membrane structural integrity (Rudolph and Goins, 1991; Harishchandra et al., 2010; Marček et al., 2019; Siddiqui et al., 2020), may be linked to the heat tolerance potential of wheat to night warming. Difference between newly developed and pre-existing leaves were consistent with previous physiological reports of plants demonstrating a far greater acclimation potential in tissue that developed under a particular thermal regime rather than simply experiencing thermal change post-development (Armstrong et al., 2006; Campbell et al., 2007; Rashid et al., 2020). However, we note that our results are based on a small number of genotypes, with a focus on photosynthesis-related traits measured at a single growth stage (anthesis) in controlled-environment conditions. Thus, possible conclusions about mechanisms of high night temperature tolerance are limited and deserve further attention in future studies.

Heat-tolerant wheat genotypes can sustain V_cmax at higher leaf temperatures when grown at warmer nights

Two heat-tolerant genotypes (Mace, a locally adapted commercial cultivar; and 2254, an elite heat-tolerant genotype) were able to either maintain or increase CO₂ fixing capacity at hotter leaf temperatures in response to warmer nights (Fig. 3). This was most evident in the significantly greater V_cmax reached at its T_opt when grown under warmer nights (Table 2; Fig. 3). An increase in photosynthetic capacity through an increase in V_cmax and its T_opt, in response to both day and night warming, is evident across a range of plant species (Kattge and Knorr, 2007; Smith and Dukes, 2017). The fact that we observed these acclimation responses in wheat genotypes due to increases in night temperature alone is interesting considering that the night temperature was not expected to have a direct influence on photosynthesis due to the temporal separation of the night warming from photosynthesis in C₃ species. Night-time respiration is more likely to be a major contributor to acclimation than photosynthesis. Although we have focused on photosynthesis and not respiration, we expect that night warming will alter the ratio of dark respiration to A_n, similar to previous studies (Turnbull et al., 2002; Bahuguna et al., 2017; Impa et al., 2019).

The stable to higher V_cmax at hotter leaf temperature in 2254 and Mace with warming nights was probably a result of more active Rubisco rather than a change in its abundance or kinetics (Scafaro et al., 2023). It is well established, including in wheat, that a proportion of Rubisco is inactive, and this inactivity increases with leaf temperature (Law and Crafts-Brandner, 1999; Sharwood et al., 2016; Perdomo et al., 2017). The decline in V_cmax at the hotter leaf temperatures is consistent with Rubisco becoming inactive. Declines in Rubisco activity with heat are not due to Rubisco per se which is a thermally stable protein (Salvucci et al., 2001), but rather due to the heat-labile nature of Rubisco activase (Rca), its accessory protein (Salvucci and Crafts-Brandner, 2004). The stable to higher V_cmax of the heat-tolerant genotypes at hot leaf temperatures—and only when grown at warmer nights—suggests that night warming is inducing changes in Rca and its control over Rubisco. Recently, a thermally stable isoform of Rca (Rca1β) was identified in wheat, and its expression and abundance increase with exposure of wheat to heat stress (Scafaro et al., 2019; Degen et al., 2021; Perdomo et al., 2021). Night warming-induced expression of Rca1β in 2254 and Mace but not in 1704 and 1898 would confirm this hypothesis.

The shift in V_cmax with night temperature could potentially be due to temperature-dependent shifts in the kinetic properties of Rubisco. However, there are limited indications that Rubisco kinetics may alter in response to growth conditions. One example is spinach (Spinacia oleracea), when grown at the hotter day/night temperature of 30/25 °C compared with 15/10 °C (Yamori et al., 2006). A small but significant increase in the thermal stability of spinach Rubisco implied acclimation-induced alterations in Rubisco functionality. A wider study, which included wheat, demonstrated that Rubisco properties such as its CO₂ substrate affinity can differ depending on growth temperature (Orr et al., 2016). More studies are needed to conclusively determine if Rubisco isoforms and associated kinetics within a species can shift in response to growth temperature. What is most unlikely is that night warming increased Rubisco abundance, as this would manifest as an increase in V_cmax across all measured leaf temperatures, but night warming did not influence V_cmax at leaf temperatures below 36 °C (Fig. 3). Furthermore, growth of wheat under both hotter days and nights does not appear to change Rubisco abundance (Perdomo et al., 2017).

CO₂ conductance and electron transport limit photosynthesis in the heat-susceptible genotypes with night warming

Newly developed leaves of the genotype 1898 and the heat-susceptible genotype 1704 (known to be susceptible in terms of yield and growth in the field under hotter growth conditions) exhibited reduced photosynthesis rates when night temperature was warmer than 15 °C. The detrimental effects of night warming on the photosynthetic activity could be attributed to a decline in CO₂ substrate availability, electron transport capacity, and TPU under the warmer nights (Figs 2–4). Declines in g_s with night warming are associated with a decline in the C_i/C_a ratio. A decline in C_i/C_a (as a result of decreased g_s) without an increase in A_n is a strong indication of water stress and CO₂ substrate limitations on photosynthetic rates (Morison and Gifford, 1983; Condon et al., 2004). Thus, it is likely that these wheat genotypes had a sensitive response to perceived limitations in water availability triggered by night warming, or adopted reduced g_s by increasing abscisic acid in direct response to heat stress (Rodriguez and Davies, 1982; Li et al., 2020). It may also have been due to the link between temperature or VPD and stomatal conductance. For the well-adapted Mace and heat-tolerant elite genotype 2254, the strong coupling of g_s with A_n (Lawson et al., 2011), and other mechanisms linked to water status (e.g. increase in xylem and mesophyll hydraulic conductance; Urban et al., 2017a), would explain the parallel increases in g_s and A_n at higher nocturnal temperature. The assumption of a fixed g_s–A_n relationship is central to many models of stomatal control of photosynthesis at different scales: leaf, plant, ecosystem, and global circulation models (Farquhar and Wong, 1984; Leuning 1995; Buckley et al., 2003; Verhoef and Egea, 2014). However, this g_s–A_n relationship can be decoupled under extreme conditions such as heatwaves (Ameye et al., 2012; von Caemmerer and Evans, 2015) or at extremely high leaf temperatures (Urban et al., 2017b).

Night warming above 15 °C reduced the photosynthetic electron capacity and/or the T_opt of J₁₅₀₀ of genotypes 1704 and 1898 (Fig. 3; Table 2). This reduction in J₁₅₀₀ led to a modelled RuBP regeneration limitation in A_n at current ambient CO₂ concentrations (Fig. 4). Despite the reductions in J₁₅₀₀ and/or the T_opt of J₁₅₀₀ with night warming, PSII was found to be thermally stable (higher T_crit) than the heat-tolerant Mace and 2254 (Fig. 1A). The mismatch between T_crit and J₁₅₀₀ demonstrates that the upper temperature limit of electron transport is not necessarily aligned with susceptibility of electron transport rates under the prevailing growth temperature. In other words, photosynthesis of 1898 and 1704 may have an ability to withstand a more severe heat shock despite being more susceptible to loss of function under milder but more sustained heating. Genotype 2254 was able to increase J₁₅₀₀ and had relatively larger increases in T_crit with night warming than 1704 and 1898, possibly meaning that the extent of change in T_crit is more informative than the upper limit of T_crit when assessing the potential for electron capacity to adjust to temperature. Pre-existing leaves of Mace, but not 1704, responded strongly to warmer nights. This suggests a greater role for older leaves in heat stress resilience, even though these leaves are assumed to have less capacity to acclimate to high temperatures.

In contrast to V_cmax, the response of J₁₅₀₀ and TPU to leaf measuring temperatures did not follow an expected exponential rise that could be fit with an Arrhenius equation (Fig. 3). A similar finding has previously been observed in wheat for J (Silva-Perez et al., 2017). This is likely to be linked to V_cmax being driven by the kinetics of Rubisco enzyme and its exponential response to temperature (apart from the aforementioned decline at hotter leaf temperatures due to Rca loss of function). J₁₅₀₀ is set by more complex protein- and membrane-dependent interactions, some of which may not respond as dynamically to short-term rises in leaf temperature. TPU, on the other hand, largely comes into play at high CO₂ levels, beyond the ambient conditions under which these experiments were conducted.

Monosaccharide and saturated fatty acid accumulation are key responses of metabolite to warm nights

The most striking change in metabolites was a build-up of monosaccharide soluble sugars in newly developed leaves of the more heat-tolerant genotypes when grown at warmer nights. The build-up of monosaccharides may be related to maintenance of the osmotic potential of cells (Marček et al., 2019) and regulation of stomatal conductance, which is also linked to photosynthesis (Flütsch et al., 2020). Soluble sugars are potent contributors to cell osmotic potential and regulation (Moinuddin et al., 2005; Blum, 2017). It is noted that osmolyte adjustments contribute to drought and heat stress avoidance in wheat (Blum and Pnuel, 1990). Specific to night warming, the disaccharide trehalose accumulated in response to night warming in wheat spike tissue (Impa et al., 2019). Thus, the accumulation of soluble sugars in the heat-responsive genotypes but not in 1704 under warmer nights may have contributed to greater osmotic protection during warmer nights. The assumed decline in osmotic potential of the heat-responsive 2254 and Mace leaves may have contributed to its increase in g_s with warmer nights, as opposed to 1704 which had a reduction in g_s. Solutes, including soluble sugars, have also been shown to provide protection to membrane structural integrity during stress through greater hydration of the lipid surfaces (Rudolph and Goins, 1991; Harishchandra et al., 2010). The higher abundance of simple soluble carbohydrates together with night warming may have contributed to the maintenance of electron transport (maintained and even increased J₁₅₀₀) and ability to acclimate T_crit to a greater extent than 1704. In agreement with our postulation that osmolytes are protecting J₁₅₀₀ in 2254 and Mace but not 1704, supplementing Populus tremula L. leaves with increasing amounts of the monosaccharide sucrose substantially reduced leaf osmotic potential and led to an increase in stability of photosynthetic electron transport at higher T_leaf (Hüve et al., 2006). That we observed a consistent increase in the saturated fatty acids octadecanoic acid and palmitic acid in the more photosynthetically heat-tolerant genotypes with night warming is a further indication of membranes adjusting to be more stable in response to warmer temperatures. Saturation of fatty acids in membranes is known to be a key heat tolerance response of plants (Larkindale et al., 2005; Zhu et al., 2018). Our results suggest that soluble sugar and saturated fatty acid contents in wheat leaves are key metabolites linking night warming and alterations in day processes. The connection between night temperature, carbohydrate and saturated fatty acid accumulation, and CO₂ conductance and electron transport capacity needs further exploration. Having identified a specific heat tolerance metabolic profile, high-throughput metabolic approaches such as hyperspectral imaging could be used to build models to predict this metabolic profile as has been achieved for amino acids in maize (Shu et al., 2022). These high-throughput measures could then be deployed by wheat populations through both ground and aerial sensing applications to determine heat tolerance.

Conclusions

Night warming has significant implications on the photosynthetic performance of wheat leaves that have developed under the prevailing night temperature. The heat-tolerant genotype (2254 and Mace) had higher light-saturated A_n when grown at warmer nights, in contrast to the other genotypes (1704 and 1898) which had reduced A_n. These night temperature-dependent differences in A_n between genotypes of wheat could in part be attributed to how night warming influenced photosynthetic capacity and CO₂ conductance. The thermally tolerant genotypes had a more stable V_cmax at higher leaf temperatures when grown at warmer nights, which is likely to be due to night-dependent alteration to the temperature sensitivity of the activation state of Rubisco. The thermally susceptible genotypes had reduced g_s and J₁₅₀₀ when grown at warmer nights, which accounts for the corresponding reduction in A_n. The stability of electron transport and CO₂ conductance in the heat-tolerant genotypes with night warming may be linked to the greater accumulation of monosaccharides and saturated fatty acids in its leaves, balancing osmotic pressure alterations and cell membrane integrity on exposure to warm nights. The lack of a night warming response in pre-existing leaves, as well as differences between the two wheat genotypes, demonstrates the divergence of strategies needed to improve wheat performance under future climate trends. Improving CO₂ conductance and photosynthetic capacity may be beneficial in some genotypes and potentially achieved through altering monosaccharide and saturated fatty acid contents of cells. In other genotypes with no detrimental impact of night warming on photosynthetic performance (exemplified by the heat-tolerant 2254 and Mace), it might be best to focus on sink tissue limitations and other physiological processes detrimentally affected by heat and not photosynthetic capacity. An important physiological process that clearly deserves investigating is mitochondrial respiration in the dark.

Supplementary data

The following supplementary data are available at JXB online.

Fig. S1. Instantaneous temperature response curves of light-saturated net assimilation (A_n) and stomatal conductance (g_s) in pre-existing flag leaves of four wheat genotypes at ambient CO₂.

Fig. S2. Temperature response curves of the maximum CO₂ carboxylation capacity (V_cmax), photosynthetic electron transport capacity (J₁₅₀₀), and triphosphate utilization (TPU) of pre-existing wheat leaves at night temperatures of 15, 20, or 25 °C.

Table S1. Mean estimates (±SD) of the optimum temperature (T_opt) and maximum capacity of V_cmax, J₁₅₀₀, and TPU in pre-existing leaves of two wheat genotypes at three different night temperatures.

Table S2. Metabolite relative abundance and classification.

Author contributions

OC and OKA: conceptualization; OC and APS: data curation, formal analysis, visualization, and writing—original draft: OC, OKA, and HB: funding acquisition; OC: investigation; OC, APS, and NT: methodology; OC, APS, HB, and OKA: project administration; OKA, resources and supervision; OC, APS, HB, NT, and OKA: writing—review and editing.

Conflict of interest

No conflict of interest declared.

Funding

This work was supported by grants from the ARC Centre of Excellence in Plant Energy Biology (CE140100008), the Australian Grains Research and Development Corporation (GRDC) Postdoctoral Fellowship: ‘Photosynthetic acclimation to high temperature in wheat’ (US1904-003RTX–9177346), and project ‘A national approach to improving heat tolerance in wheat through more efficient carbon allocation’ (US00080). OC was supported by Research England’s Expanding Excellence in England (E3)-funded Food and Nutrition Security Initiative of the Natural Resources Institute.

Data availability

All primary data to support the findings of this study are openly available in the Dryad Digital Repository at doi: 10.5061/dryad.fqz612jx7 (Coast et al., 2023).