Leaf temperature and CO₂ effects on photosynthetic CO₂ assimilation and chlorophyll a fluorescence light responses during mid-ripening of Vitis vinifera cv. Shiraz grapevines grown in outdoor conditions

Keywords: carbon dioxide, chlorophyll fluorescence, leaf temperature, light response curves, photochemistry, severe climate, source effects, Shiraz vines.

Vitis vinifera cv. Shiraz grapevines are the economically most important commercially grown red wine grape cultivar in Australia (ABS 2015), with nearly 40 000 h in cultivation. In comparison with other common grapevine cultivars grown in comparable conditions, however, Shiraz vines had relatively low photosynthetic rates (10.4 μmol m⁻² s⁻¹) but had the highest shoot growth rates in a study by Rogiers et al.(2009). By contrast, Soar et al. (2009) measured light-saturated photosynthesis across the early stages of berry development and rates averaged 13–15 μmol m⁻² s⁻¹. Also early in Shiraz berry development, Caravia et al. (2016) measured light-saturated rates at 32°C at 12.5 μmol m⁻² s⁻¹ but also demonstrated that when temperatures reached approximately 40°C, photosynthesis was strongly inhibited. However, Rogiers and Clarke (2013) have shown that light-saturated photosynthesis of Shiraz leaves barely changed between approximately 19 and 38°C, averaging 11.9 μmol m⁻² s⁻¹. Although grown in greenhouse conditions at 35/30 and 25/20°C for 6 days, contradictorily, Shiraz vines measured at 25°C had average photosynthetic rates of 12.1 μmol m⁻² s⁻¹ and 10.9 μmol m⁻² s⁻¹ when measured at 35°C (Hochberg et al. 2015), suggesting some photosynthetic temperature dependency. This was confirmed for Shiraz vines grown outdoors, where light-saturated photosynthesis followed a well-known curvilinear temperature response, peaking at 30–35°C (Greer 2019). Thus the effect of temperature on Shiraz vines is becoming increasingly quantified and hence understood, which is essential in the face of a changing climate (Hall et al. 2016).

Although responsive to sink demand, photosynthesis is driven first and foremost by light interception (Allakhverdiev 2020; Brestic et al. 2021). Photosynthetic light response curves provide much information about the adaptation and acclimation of a species to the growth environment (Osmond and Chow 1988). A comparison of photosynthetic light responses in a number of horticultural species, including apple, pear, peach, grape and olive (Higgins et al. 1992), was seminal in establishing that intrinsic differences in photosynthetic light responses occurred between related species. Thus, light-saturated rates ranged between 12 and 23.5 μmol m⁻² s⁻¹ for the different species but there were also differences in the photon (quantum) yield, light compensation and light saturation.

Photosynthetic light responses can also be strongly modulated by leaf temperature. For example, Laing (1985) showed for Actinidia chinensis leaves, that light-saturated photosynthesis was strongly affected by leaf temperature, maximum rates declining from 13.1 to 7.3 μmol m⁻² s⁻¹ when leaf temperature decreased from 25 to 10°C. Similarly, when temperatures were increased from 20 to 30°C, the light-saturated rates of Annona squamosa increased from 8.0 to 14.3 μmol m⁻² s⁻¹ (Higuchi et al. 1998). More comprehensively, Greer (2015) measured photosynthetic light response in apple leaves across leaf temperatures ranging from 15 to 35°C, where rates were optimal at 25°C. Similarly, light responses of cv. Semillon grapevines at leaf temperatures from 20 to 40°C were optimal at 30°C (Greer and Weedon 2012). Photosynthetic light responses at different temperatures were also measured in cv. White Riesling grapevines (Schultz 2003). Thus, the knowledge of the temperature dependency of the photosynthetic light response is important to comprehend crop productivity in commercial horticulture and in matching productivity to the climate.

While the majority of source–sink studies have used fruit removal or crop adjustment to moderate assimilation, another approach could use CO₂ to alter assimilation without changing the sink capacity. In a unique approach, Rogiers et al. (2011) used a low growth CO₂ concentration (94 μmol mol⁻¹) to perturb assimilation to developing grape inflorescences. This, of course, had dramatic effects on light-saturated leaf photosynthesis, declining from 6.7 to 1.7 μmol m⁻² s⁻¹ when the CO₂ declined from ambient to the low concentration. Elevated CO₂ also modulates the photosynthetic light response, increasing the initial slope as well as the maximum rate, as has been shown for soybean leaves (Campbell et al. 1988, 1990). In these studies, an increase in CO₂ concentration from 330 to 660 μmol mol⁻¹ caused a 65% increase in photon yield and a 57% increase in light-saturated rates. Similar results occurred with Eucalyptus maculata leaves, where CO₂ concentrations between 140 and 1000 μmol mol⁻¹ increased the photon yield by 44%, with a 17% increase in light-saturated photosynthesis (Ögren and Evans 1993). Further support for the effect of CO₂ on the light response comes from Watson et al. (1978) and Farazdaghi and Edwards (1988).

In the present study, CO₂ assimilation and chlorophyll a fluorescence light responses were measured at a range of leaf temperatures and CO₂ concentrations when Vitis vinifera cv. Shiraz vines were in the rapid berry growth stage. The objective was to determine how the photosynthetic performance occurred during this developmental stage when a high sink demand occurred and when assimilation was manipulated by CO₂ compared with an earlier study (Greer 2019), where the assimilation was manipulated by removal of the reproductive sink. The present study was also characterised by a severe and sustained heat event, where air temperatures exceeded 45°C and a very different climate from the previous study.

This study was carried out using the National Wine and Grape Industry Centre growth facilities at Charles Sturt University (latitude 35.05°S and longitude 147.35°E, 212 m above sea level) in the Riverina, New South Wales, Australia during the 2018–2019 growing season. The own-rooted Vitis vinifera cv. Shiraz vines were 11 years old and grown in 52 L pots in a commercial bulk composted potting mix and fertilised with liquid fertiliser (Megamix Plus RUTEC, Tamworth, Australia). Other details of the growth conditions followed Greer (2019). Budbreak occurred in mid-September and the vines were measured in mid-summer. This study period of the growing season was characterised by air temperatures exceeding 35°C and beyond 40°C on 14 of those days, peaking at 46.4°C (see Greer 2020).

Screened air temperatures and humidities were measured (Humitter, Vaisala) at a height of 1.2 m above the ground and close by the rows of vines. The data were logged at hourly intervals on a data logger (CR1000, Campbell Scientific) over the summer. Photon flux densities (PFD) were measured each day simultaneously using a quantum sensor (QSO, Apogee) and recorded with the data logger.

On four occasions at approximately 7–10-day intervals during the rapid berry growth stage (100–130 days after budbreak), photosynthetic light responses were measured at five constant leaf temperatures from 20 to 45°C in 5°C increments on youngest fully expanded leaves. Between each temperature change, a period of 4–5 min was used to ensure the temperature had reached the set point. The LI-6400 XT gas-exchange system (LI-COR Biosciences, Lincoln, NE, USA) fitted with the LI6400 – 40 leaf chamber fluorometer attached to the cuvette system was used for these measurements. Note no measurements of fluorescence of dark-adapted samples were undertaken during these measurements. For each light response using the red–blue light from the LI6400-40 fluorometer, the PFD was initially set at 2000 μmol m⁻² s⁻¹ and when photosynthetic rates were steady, the PFD was progressively decreased in selected steps to approximately 1 μmol m⁻² s⁻¹ (dark). The CO₂ concentration was controlled at 400 μmol mol⁻¹ throughout and vapour pressure deficits were managed with a simple air humidifying system (Greer 2018b) which involved a glass flask partly filled with water, placed on a black plate and fed to the gas-exchange air supply. Over the course of the day, as the air temperature increased, the humidity of the air also increased in the flask and maintained the water vapour pressure in the gas-exchange cuvette. The whole measurement procedure was repeated 3–4 times at each temperature, with a new leaf for each light response on each occasion. The temperature and high PFD were initially turned on for 3–4 min until the set temperature was achieved and then each step of the PFD was maintained for approximately 2 min, ensuring CO₂ assimilation rates were steady, thus the whole response occurred over 25–30 min at each temperature.

Using the ‘LightCurve_MultipleCO₂’ facility in the auto programs of the LI6400 XT system, light responses as above were again conducted but with the CO₂ concentration varying in separate light responses from 200, 600 to 1000 μmol mol⁻¹. Leaf temperature was maintained constant for each light response at 25, 30 and 35°C at each of these light/CO₂ combinations. A new leaf was used for each light/CO₂/temperature combination and each response was repeated 3–4 times on each occasion and durations of each step were as above. The times required to complete these responses at each combination precluded extending the measurements to the other temperatures (20, 40 and 45°C).

During each light response, at each PFD step, the steady-state fluorescence in the light (F_s′) and the maximal fluorescence (F_m′) were measured using a red/blue flash intensity of at least 7000 μmol m⁻² s⁻¹. Then the actinic light (the prevailing PFD) was briefly turned off, the far-red light was turned on for 3 s and the light adapted minimal fluorescence (F_o′) was measured. Following Wünsche et al. (2000), the chlorophyll fluorescence attributes ΔF′/F_m′ (quantum efficiency of PSII), qP (photochemical quenching) and ETR (the electron transport rate) were determined (Greer 2019). All attributes were measured and recorded by the LiCor software at each light step.

All data were analysed using a general linear model (GLM) approach using SAS 9.3 (SAS Institute Inc.) and least squares means and standard errors were determined. A fully randomised experimental system was used. The photosynthetic and ETR data were analysed using non-linear regression with SAS to fit a hyperbolic tangent function according to Greer and Halligan (2001).

The mean air temperatures during the measurement period (mid-summer) averaged 26.7 ± 0.6°C during the day and 20.8 ± 0.5°C during the night and vapour pressure deficits, respectively, averaged 2.6 ± 0.14 and 1.4 ± 0.09 kPa (not shown). Between 14:00 and 18:00 h, the mean daily PFD averaged 1710 ± 80 μmol (photons) m⁻² s⁻¹. However, the measurements were conducted during an extended 40-day period where temperatures exceeded 35°C and beyond 40°C on 14 of those days, peaking at 46.4°C (see Greer 2020).

The CO₂ assimilation light response across all leaf temperatures at the ambient CO₂ concentration (Fig. 1) fitted the hyperbolic tangent function well (P < 0.001; r² = 0.96–0.98). The maximum rates were evident at 30°C, and the minimum rates occurred at both 20 and 45°C.

Fig. 1.  CO₂ assimilation (mean ± s.e., N = 4–6) to increasing PFD of Shiraz leaves grown in outdoor conditions and measured at a range of leaf temperatures as indicated and at a CO₂ concentration of 400 μmol mol⁻¹. The lines are a fit to the hyperbolic tangent function according to Greer and Halligan (2001).

The CO₂ assimilation light responses at the different CO₂ concentrations and at 25 and 35°C leaf temperatures are shown in Fig. 2a, b. There was a highly significant effect of CO₂ on each light response, with clear differences in the initial light response but also in the light-saturated rates. At 25°C, for example, light-saturated photosynthesis (A_max) increased from 3.9 ± 0.2 μmol m⁻² s⁻¹ at 200 μmol mol⁻¹ CO₂ to 25.1 ± 0.5 μmol m⁻² s⁻¹ at 1000 μmol mol⁻¹ CO₂. Proportionally the biggest increase in A_max was, however, between 200 and 400 μmol mol⁻¹ CO₂ (3-fold). By contrast at 35°C (Fig. 2b), the A_max at 200 μmol mol⁻¹ CO₂ averaged 8.3 ± 0.7 μmol m⁻² s⁻¹ and only 33% higher at 400 μmol mol⁻¹ CO₂. Of note, the light-saturated photosynthetic rates were only slightly higher at the high CO₂ concentrations at 35°C compared to 25°C, but all were maximal at 1000 μmol mol⁻¹ CO₂. Rates at 30°C (not shown) were intermediate.

Fig. 2.  CO₂ assimilation (a, b) and electron transport (c, d) rates (mean ± s.e., N = 6) as responses to increasing PFD of Shiraz leaves grown in outdoor conditions and measured at a range of CO₂ concentrations as indicated and at 25°C (a, c) and 35°C (b, d) leaf temperatures. The lines in each case are a fit to the hyperbolic tangent function. Note where the electron transport rates declined at higher PFDs, these data were excluded from the function fitting. In some cases, the error bars were too small to be apparent.

There was a much smaller effect of CO₂ concentration on light-saturated electron transport rates (Fig. 2c, d), with rates at 25°C increasing from 93 ± 2 μmol m⁻² s⁻¹ at low CO₂ to 130 ± 5 μmol m⁻² s⁻¹ at 600 μmol mol⁻¹ CO₂, a 40% increase. However, there was a 15% decrease in the maximum electron transport rate at the highest CO₂ concentration to match that at 400 μmol mol⁻¹ CO₂. This effect of CO₂ was exacerbated at 35°C, where the highest electron transport rate (149 ± 5 μmol m⁻² s⁻¹) occurred at 400 μmol mol⁻¹ CO₂ and the lowest rate occurred at 1000 μmol mol⁻¹ CO₂. It was also notable that, especially at 200 and 400 μmol mol⁻¹ CO₂, the electron transport rates declined with increasingly high photon flux densities, by 5% at 200 μmol mol⁻¹ CO₂ and 17% at 400 μmol mol⁻¹ CO₂. This depreciation in ETR was also apparent at 30 and 40°C (not shown) but not apparent at high CO₂ or at any CO₂ concentration at 25°C.

Fitting the hyperbolic tangent function to the CO₂ assimilation light responses was highly significant (P < 0.001; r² = 0.95–0.98) and analysis of the function indicated (Fig. 3a–d) that there was a clear and marked interaction between leaf temperature and CO₂ concentration. For A_max, the interaction was highly significant (P < 0.001; r² = 0.95) and rates were significantly different between the CO₂ concentrations, but the effect of temperature was CO₂-dependent, with a marked temperature dependency at the low to mid CO₂ concentrations but little apparent evidence of a comparable dependency at high CO₂ concentrations, at least over the limited temperature range. However, there appeared to be a consistent response in that approximately 30°C appeared to be the optimum temperature for light-saturated photosynthesis.

Fig. 3.  Parameters (mean ± s.e., N = 4–6) fitting the hyperbolic tangent function to the CO₂ assimilation light responses shown in Fig. 2 as a function of leaf temperature: (a) light-saturated maximum rate (A_max), (b) apparent photon yield (ϕ_a), (c) PFD at light saturation (PFD_sat) and (d) dark respiration of Shiraz leaves, measured at a range of CO₂ concentrations as indicated. Note no measurements at 20, 40 and 45°C were undertaken at other than 400 μmol CO₂ mol⁻¹.

For the photon (quantum) yield (Fig. 3b), there was again a highly significant (P < 0.001; r² = 0.79) interaction between temperature and CO₂ concentration. Across the whole temperature range, the photon yield at 400 μmol mol⁻¹ CO₂ was highest at 20–30°C and declined approximately linearly to the lowest yield at 45°C, declining approximately 37% in total. There was an overall increase in photon yield with increasing CO₂, at least to 600 μmol mol⁻¹ CO₂ and across the restricted temperature range averaged approximately 0.0419 ± 0.0003 mol CO₂ mol (photons)⁻¹ at 200 μmol mol⁻¹ CO₂ increasing to 0.0609 ± 0.0011 mol CO₂ mol (photons)⁻¹ at 600 μmol mol⁻¹ CO₂. Surprisingly, the photon yield declined at the highest CO₂ concentration though approximately 5% and mostly non-significant. Consistent with A_max, there appeared to be no apparent temperature dependency of the photon yield at high CO₂ concentrations.

The PFD at which light saturation of photosynthesis occurred (PFD_sat) was strongly dependent on the CO₂ × temperature interaction (Fig. 3c), which was highly significant (P < 0.001; r² = 0.71). At the ambient CO₂ concentration, there was a general tendency for PFD_sat to increase curvilinearly with increasing leaf temperature, from approximately 600 ± 24 μmol (photons) m⁻² s⁻¹ at 20°C to a maximum of 965 ± 37 μmol (photons) m⁻² s⁻¹ at 40°C. At the lowest CO₂ concentration, PFD_sat appeared to follow the same response at 400 μmol mol⁻¹ CO₂. Whereas at the higher CO₂ concentrations, although there were increases in PFD_sat, both appeared maximal at 30°C, 985 ± 36 and 1164 ± 42 μmol (photons) m⁻² s⁻¹ at 600 and 1000 μmol mol⁻¹ CO₂, respectively. Also, at the highest CO₂ concentrations, there appeared to be a different temperature response, with an apparent peak, but more data were needed to confirm this.

For dark respiration (Fig. 3d), the CO₂ × temperature interaction was highly significant (P < 0.001) but with r² = 0.29, hence the interaction was not especially strong. This was evident in that at CO₂ concentrations above 400 μmol mol⁻¹, dark respiration tended to increase curvilinearly with increasing leaf temperature to a maximum rate at 40°C and CO₂ had no apparent effect. By contrast, respiration at the low CO₂ concentration increased more steeply with increasing temperature such that rates at 35°C were approximately 30% higher compared with 400 μmol mol⁻¹.

Fitting the hyperbolic tangent function to the ETR response to light was highly significant (P < 0.001; r² = 0.93–0.98). For the light-saturated maximum rates of electron transport (Fig. 4), there was a highly significant (P < 0.001; r² = 0.57) temperature × CO₂ interaction. At 400 μmol mol⁻¹ CO₂, rates of electron transport increased curvilinearly with increasing leaf temperature to reach a maximum of 155 ± 8 μmol (electrons) m⁻² s⁻¹ at 35–40°C, and a marked depreciation in the rates occurred at 45°C (Fig. 4a). At the higher CO₂ concentrations, rates of electron transport were broadly similar to those at ambient CO₂ and at 25–30°C, but there was a highly significant (P < 0.01) decrease in the maximum rates at 35°C, at both lower and higher CO₂ concentrations. However, across all leaf temperatures at 200 μmol mol⁻¹ CO₂, the rates of electron transport were significantly (P < 0.01) lower compared to all other CO₂ concentrations but apparently peaked at 30°C.

Fig. 4.  Parameters (mean ± s.e., N = 4–6) fitting the hyperbolic tangent function to the electron transport light responses shown in Fig. 2 as a function of leaf temperature: (a) light-saturated maximum rate, (b) initial light limited slope, (c), PFD at light saturation (PFD_sat) and (d), dark rate of electron transport of Shiraz leaves, measured at a range of CO₂ concentrations as indicated.

Although there was a highly significant temperature × CO₂ interaction (P < 0.001; r² = 0.63) on the initial, light limited slope of the ETR light response (Fig. 4b), by and large the effect of temperature was relatively weak across all CO₂ concentrations. However, there was an apparent effect of CO₂ concentration, where the slope increased from 0.279 ± 0.057 to 0.295 ± 0.037 mol (electrons) mol (photons)⁻¹ from 200 to 400 μmol mol⁻¹ CO₂ whereas at the higher CO₂ concentrations, the average slope decreased non-significantly from 0.247 ± 0.042 at 600 μmol mol⁻¹ CO₂ to 0.241 ± 0.057 mol (electrons) mol (photons)⁻¹ at 1000 μmol mol⁻¹ CO₂.

For the PFD at which the ETR light response was saturated (Fig. 4c), there was a highly significant temperature × CO₂ interaction (P < 0.001; r² = 0.53) and at 400 μmol mol⁻¹ CO₂, there was a bell-shaped temperature response in that PFD_sat increased from 800 ± 15 μmol (photons) m⁻² s⁻¹ at 20°C to a maximum of 1385 ± 35 μmol (photons) m⁻² s⁻¹ 35°C then declined at the higher temperatures to 1205 ± 35 μmol (photons) m⁻² s⁻¹ at 45°C. Apart from a few deviations, the PFD_sat at the higher CO₂ concentrations conformed to the temperature-induced increase as occurred at ambient CO₂. At the low CO₂ concentration, there was little effect of temperature on light saturation and PFD_sat averaged 947 ± 12 μmol (photons) m⁻² s⁻¹ and significantly lower than the PFD_sat at the higher CO₂ concentrations. It was noteworthy that the PFD_sat for ETR was generally similar to those for photosynthesis (cf. Fig. 3).

There was a highly significant temperature × CO₂ interaction (P < 0.001; r² = 0.76) for the rates of electron transport measured in the dark (Fig. 4d). At 400 μmol mol⁻¹ CO₂, the dark rates of electron transport increased in a linear pattern with increasing leaf temperature to an apparent peak of 2.72 ± 0.03 μmol (electrons) m⁻² s⁻¹ at 40°C and a marked decrease occurred at 45°C. There was a matching response at 600 μmol mol⁻¹ CO₂ with very little difference in the dark rates at each temperature. The temperature response at 1000 μmol mol⁻¹ CO₂, was generally comparable between 25 and 30°C to that at ambient CO₂ with similar dark rates, however, at 35°C, the response deviated downwards, declining by approximately 40%. The effect of this was it appeared that at high CO₂ concentrations, dark rates of electron transport were not dependent on temperature. A similar conclusion was apparent at the lowest CO₂ concentration, where there was little effect of temperature, and the dark rates of electron transport were very low averaging 1.05 ± 0.07 μmol (electrons) m⁻² s⁻¹ compared to rates at the higher CO₂ concentrations.

For each chlorophyll fluorescence attribute, the interaction between temperature × CO₂ was not significant, however, there was a weak effect of leaf temperature (P < 0.01; r² = 0.197–0.323). This was apparent with the decrease in each attribute with increasing PFD (Fig. 5) at each temperature, with only small differences in each attribute. The most consistent effect of temperature was that for each attribute, the values were lowest at 20°C and highest between 35 and 40°C. In all cases, the effect of PFD was highly significant (P < 0.001; r² = 0.95–0.98) and all responses declined in an exponential pattern.

Fig. 5.  Chlorophyll fluorescence attributes (mean ± s.e., N = 4–6) of outdoor-grown Shiraz leaves as a function of PFD, averaged over all CO₂ concentrations and measured at a range of leaf temperatures as indicated; (a) quantum efficency of PSII (ΔF′/F_m′) and (b) photochemical quenching (qP).

The quantum efficiency of PSII (ΔF′/F_m′) was markedly reduced across all leaf temperatures at the saturating light (Fig. 6a), with rates ranging from 0.081 ± 0.011 to 0.251 ± 0.013 at 20 and 40°C, respectively. The quantum efficiency increased in a linear fashion between these limits, with a sharp decrease in efficiency at 45°C but mostly there was only a small response to leaf temperature. These data suggested a relatively small proportion of absorbed light was being used in PSII light reactions at these high PFD intensities.

Fig. 6.  PSII quantum efficiency (closed symbols) and photochemical quenching (closed symbols) (mean ± s.e., N = 8–12) of outdoor-grown Shiraz leaves as a function of leaf temperature. In each case the values were determined as averages at saturating PFDs and across the range of CO₂ concentrations investigated. In some cases, the error bars were too small to be apparent.

Photochemical quenching (qP) at saturating light had the steepest response to temperature, with photochemical quenching decreasing from 20 to 40°C in a linear pattern, from 0.231 ± 0.003 to 0.582 ± 0.013; but note the increase in photochemical quenching at 45°C was less than occurred at all temperatures below 35°C. These results confirmed PSII reaction centres were closing at the warmest temperatures, in keeping with the lower quantum efficiencies.

To confirm that CO₂ has no effect on these fluorescence attributes, the mean values across all PFDs and leaf temperatures are plotted as a function of CO₂ in Fig. 7. This shows that the slopes of both attributes were more or less flat, although there was a slight weakly significant but positive slope (P = 0.04; r² = 0.87) for photochemical quenching; but the indications were that all reaction centres were open, irrespective of CO₂.

Fig. 7.  PSII quantum efficiency (open symbols) and photochemical quenching (closed symbols), (mean ± s.e., N = 8–12) of outdoor-grown Shiraz leaves as a function of CO₂ concentration. In each case the values were determined as averages across all temperatures and PFDs.

In previous research (Greer 2020), it was shown that fruiting Shiraz grapevines grown in a hot climate were well acclimated to these conditions, with the photosynthetic process biased towards high temperatures (>30°C) and not towards low (<25°C) temperatures. This was determined by measuring CO₂ assimilation light responses across different leaf temperatures with fruiting and non-fruiting vines. In the present study, CO₂ assimilation light responses were also measured but with an interaction between leaf temperature and CO₂ to perturb the source rather than the sink as in the previous study. Elevated CO₂ is known to enhance photosynthesis in the short-term (Laing et al. 2002) and likewise, a low CO₂ concentration will limit photosynthesis. As shown by Rogiers et al. (2011), growing vines in a low CO₂ concentration of 94 μmol mol⁻¹ markedly depreciated photosynthesis and consequently inhibited biomass accumulation in the fruit sink. In the present study, a range of short-term exposures of CO₂ from 200 to 1000 μmol mol⁻¹, were used to manipulate photosynthesis during the period of rapid fruit growth (mean fruit biomass accumulation increased from 170 to 443 g vine⁻¹ over the course of this study) and to determine if the acclimation potential of the Shiraz grapevines to temperature was enhanced or reduced at the different CO₂ concentrations.

Across the whole temperature range, the CO₂ assimilation light responses were comparable with those for fruiting Shiraz vines (Greer 2020), in that the optimum temperature was between 30°C and 35°C. Similarly, rates at 20°C matched those at 45°C. However, there were differences in the rates of photosynthesis, in that the maximum rate in the present study averaged 16 μmol m⁻² s⁻¹ compared to 11.8 μmol m⁻² s⁻¹ in the earlier study. This appeared to result from an overall increase in rates across the whole temperature range in the current season, as rates at the temperature extremities were 2–3 μmol m⁻² s⁻¹ higher as well. This increase in rates across the board in the present study was probably associated with seasonal variation in climate across the different growth seasons. Although the mean daily temperatures on days when measurements were done were similar between the two studies, heat events (>40°C) were markedly evident in the present study and absent in the former study. An increase in capacity for CO₂ assimilation may have occurred in these extreme growth conditions, supporting the conclusion that this grapevine cultivar was inherently tolerant of high temperature regimes.

In addition to temperature differences, there were also differences in the irradiance over the two growing seasons, with approximately 35% higher intensities in the current season, which may have also enhanced the photosynthetic capacity. This conclusion is consistent with studies where photosynthetic light responses were determined in different canopy positions (Jurik et al. 1988; Campbell et al. 1992; Mierowska et al. 2002) or treated with different light exposures (Gardiner and Krauss 2001). Furthermore, light saturation of photosynthesis of the Shiraz vines occurred at PFDs above 800 μmol (photons) m⁻² s⁻¹, especially at the higher temperatures during the current growing season whereas in the former growing season, light saturation was markedly lower (Greer 2020). The increase in saturation PFDs in the present growing season and the response to temperature were well in keeping with exposure to the high temperature–high light regime of the current season.

For the Shiraz vines, the CO₂ assimilation light responses were both highly temperature- and CO₂-dependent, but the CO₂ had a greater effect. Across the whole temperature range, A_max increased by 1.5-fold whereas for the increase in CO₂ concentration, A_max increased more than 3-fold. However, there were strong interactions between temperature and CO₂, and the effect of CO₂ on light-saturated photosynthesis was much greater at 25°C compared with that at 35°C. Similarly, temperature had a greater effect on A_max at low compared to high CO₂, where there was no apparent evidence of temperature sensitivity at the two highest CO₂ concentrations, albeit over a restricted temperature range. The effect of CO₂ on the photosynthetic light response has been well established (Ludlow and Wilson 1971; Farazdaghi and Edwards 1988; Ögren and Evans 1993; Zhou et al. 2011), for example, that the light-saturated rates of apple trees increased 6-fold with CO₂ increasing from 195 to 1580 μmol mol⁻¹ (Watson et al. 1978). In the present study, light-saturated rates at the optimum temperature increased 4-fold with increasing CO₂ and were thus well in keeping with these other studies.

By contrast, there is much less understanding of the interaction between leaf temperature and CO₂ concentration on the CO₂ assimilation light responses. In the Campbell et al. (1990) study, there were no marked interactions between two CO₂ concentrations and three leaf temperatures ranging between 22 and 28°C, possibly too small a range to effect an interaction. By contrast, with CO₂ assimilation light responses of apple leaves, Greer (2018a) demonstrated a highly significant interaction between CO₂ concentrations and leaf temperatures ranging from 20 to 38°C. In that study, increased CO₂ enhanced the effect of temperature on light-saturated photosynthesis, with a greater sensitivity at high (600 μmol mol⁻¹) compared to low (300 μmol mol⁻¹) CO₂ concentrations (see also Greer 2017). In the present study, however, the range of temperatures was too restricted to detect any change in temperature sensitivity on light-saturated photosynthesis, although there were some indications that greater sensitivity occurred at the low compared to the high CO₂ concentrations. This may be attributed to high rates of photorespiration at low CO₂ (Farazdaghi and Edwards 1988; Ögren and Evans 1993) and also to the highly temperature-dependent nature of this process (Greer 2015). Another possible effect is low Rubisco activity, induced by the low substrate supply, although Campbell et al. (1990) failed to detect any effect of the growth CO₂ or temperature treatments on measured total Rubisco activity or Rubisco activation. Further study is warranted to determine if high CO₂ does suppress the sensitivity of light-saturated photosynthesis to temperature, as this may have climate change implications.

For the Shiraz vines, there was also a marked increase in the apparent photon yield with increasing CO₂, although the response was curvilinear and appeared to maximise at 600 μmol mol⁻¹ (see also Greer 2018a). This is well in keeping with maximal photon yields occurring at saturating CO₂ as determined by Björkman and Demmig (1987). The range in apparent photon yields determined here were consistent with that for Eucalyptus maculata measured at different CO₂ concentrations (Ögren and Evans 1993) and other Eucalyptus species (Wong and Dunin 1987). Similar results occurred also with apple leaves (Watson et al. 1978). However, there was a highly significant temperature × CO₂ interaction for the apparent photon yields of the Shiraz leaves. At 400 μmol mol⁻¹, photon yields declined progressively by approximately 1.5-fold between 20 and 45°C (see also Greer 2019) while at 200 μmol mol⁻¹ there was an apparent peak at 30°C. However, temperature had no marked effect on the photon yields at elevated CO₂. When measured at ambient CO₂, the photon yields of apple leaves grown in a cool climate also barely altered over a comparable temperature range and a similar response occurred at different CO₂ concentrations (Greer 2015). However, for cv. Braestar apple leaves (Pretorius and Wand 2003) photon yields also declined with increasing temperature (see also Harley et al. 1985; Gardiner and Krauss 2001; Gindaba and Wand 2007). It remains uncertain if the apparent lack of effect of temperature on photon yields at high CO₂ of the Shiraz vines was sustainable, given the narrow range of temperatures tested, or was an effect of the extreme growth climate.

The PFD saturating photosynthesis (PFD_sat) was strongly affected by a significant interaction between leaf temperature and CO₂. This was apparent in that at low to ambient CO₂ concentrations, there was a curvilinear increase in PFD_sat with increasing temperature to a maximum at 40°C whereas at elevated CO₂, there appeared to be a near comparable response, yet with a maximum at 30°C. Across the whole range of temperatures, PFD_sat increased progressively from 200 to 1000 μmol mol⁻¹ CO₂, by approximately 2-fold. In many studies (Man and Lieffers 1997; Schultz 2003; Greer and Weedon 2012), the saturation PFD (at ambient CO₂) increased with temperature and, therefore, conformed to the present study. What was striking about the PFD_sat of the Shiraz leaves, however, was that increasing temperature increased the saturation PFD by 1.5-fold but the increase in CO₂ concentration approximately doubled the saturation PFD (see also Wong and Dunin 1987). It was also apparent that PFD_sat increased at all temperatures with the increased CO₂ concentration (cf. Kakani et al. 2008), however, the effect was greatest at 25°C and lowest at 35°C, hence the interaction between these two factors. The effect of CO₂ on PFD_sat appeared to conform with the effect on A_max, suggesting that driving photosynthesis with increased substrate CO₂ appeared to induce a greater demand for light interception (Fig. 2a, c) although these two processes were seemingly not well matched. It therefore remains necessary to re-evaluate the whole temperature × CO₂ interaction to fully comprehend how light saturation of photosynthesis responds, as this will be useful in modelling photosynthesis.

Although there was a highly significant temperature × CO₂ interaction on dark respiration, for most part CO₂ had a relatively small effect at the different temperatures. At ambient CO₂, dark respiration increased exponentially with increasing temperature up to 40°C, in keeping with the well-established exponential rise in respiration with temperature (Atkin et al. 2006; Way and Sage 2008; Greer 2017). Notably, the same pattern of increase in respiration with temperature also occurred at the higher CO₂ concentrations, with few differences in the rates, suggesting little dependency of respiration on CO₂ concentrations above ambient, as was also concluded by Greer (2017) and Kakani et al. (2008). However, the temperature response deviated upwards when measured at 200 μmol mol⁻¹ CO₂, suggesting a steeper rise in the exponential pattern occurred. This conforms partially to Amthor et al. (1992) who demonstrated respiration decreased with increasing CO₂, but also that the effect was diminished at high (25°C) compared to low (15°C) temperatures. By contrast, Greer (2018a) observed an increase in respiration rates with increasing CO₂ concentration. Unfortunately, there appear to be too few studies to compare the effects of temperature and CO₂ on respiration. However, for the Shiraz leaves at low CO₂, respiration appeared strongly temperature dependent but notably highest rates of respiration occurred at 30 and 35°C, which may have partly reflected increased photorespiration at low CO₂ and high temperatures (Farazdaghi and Edwards 1988). Certainly, estimates of rates of photorespiration at these temperatures for apple leaves (Greer 2015) were in the range of 3–3.5 μmol m⁻² s⁻¹, thus consistent with the high apparent respiration rates. Hence the results generally confirm the exponential response of respiration to temperature (Man and Lieffers 1997; Greer 2015) and the lack of a response to elevated CO₂ shown for Chardonnay and Merlot vines (Greer 2017) and for Shiraz vines (Fig. 4d); but the explanation for a steeper temperature response at low CO₂ was most probably related to photorespiration.

The response of the electron transport rate to light followed similar saturating kinetics as assimilation, as shown for Meconopsis horridula (Zhang and Yin 2012), Camelina sativa (Cendrero-Mateo et al. 2015) and Actinidia deliciosa (Greer and Halligan 2001). However, the interaction with temperature and CO₂, although highly significant, was very different in that CO₂ assimilation increased consistently with increasing CO₂ whereas electron transport rates were maximal at 600 μmol mol⁻¹ at 25°C but at 400 μmol mol⁻¹ at 35°C. The light-saturated maximum ETR at ambient CO₂ concentration had a curvilinear response to temperature and rates were maximal at 40°C. In contrast, for the more abridged temperature response at the other CO₂ concentrations, the maximum electron transport rates occurred at 30°C and were markedly decreased at 35°C but inversely related to CO₂. By contrast, for apple leaves there was a consistent effect of temperature on the maximum electron transport rates, rising from 15 to 35°C (Greer 2015). Although not explicitly as a function of temperature, light-saturated electron transport increased from winter to spring in three conifer species, in keeping with rising air temperatures (Robakowski 2005). Thus, there is some consensus that light-saturated electron transport rates increase with leaf temperature. By contrast, the effect of CO₂ on the response of electron transport rates to PFD has not been well addressed, although Ögren and Evans (1993) demonstrated an increase in light-saturated electron transport occurred with increasing CO₂. While there was some response of electron transport rates to rising CO₂ of the Shiraz vines, what was notable was that elevated CO₂ appeared to impede electron transport, especially at the highest temperature and in marked contrast to the CO₂ assimilation response. It remains uncertain how different CO₂ concentrations perturbed both the temperature dependency and rates of electron transport, especially as the PSII quantum efficiency was largely unaffected by leaf temperature and CO₂. Similarly, there was no change in non-photochemical quenching as determined by qN (Supplementary Fig. S1) in response to CO₂. It has been proposed (Greer 2022) that PSII may be more sensitive to high temperature–high light conditions than CO₂ assimilation as a means of fine tuning photosynthesis to these growth conditions (see also Allakhverdiev 2020) to ensure CO₂ fixation continued.

For all the fluorescence attributes, the temperature × CO₂ interaction was not significant, whereas the leaf temperature was highly significant (P < 0.01) but only accounted for 20–30% of the variance, with PFD accounting for much of the variance. The PFD induced decrease in the PSII quantum efficiency of the Shiraz vines was fairly severe at all temperatures, but mostly at 20°C (90%) and least at 40°C (80%), although similar decreases have occurred in a number of studies (Urban et al. 2004; Zhang and Yin 2012; Cendrero-Mateo et al. 2015). For Shiraz vines in a previous study (Greer 2019), the quantum efficiencies across a similar range of temperatures were much higher, peaking at 0.35 at 35°C and only approached the efficiencies in the present study at 20°C. Thus, the PFD induced decreases in quantum efficiency in the present study were more severe, in keeping with the high ambient heat load of the current growing season and again consistent with fine tuning of the light reactions of photosynthesis. However, while not strongly evident here, the PSII quantum efficiency was highly temperature dependent, consistent with this conclusion. For example, Weston and Bauerle (2007) demonstrated that the quantum efficiency of Acer rubrum cuttings was curvilinearly dependent on temperature, with an optimum at approximately 30°C and a marked and decreasing efficiency as temperatures increased to beyond 45°C. Similar results were demonstrated for field-grown grapevines cv. Kékfrankos, again with a broad optimum at approximately 30–35°C (Zsófi et al. 2009) and with outdoor-grown Shiraz vines, although the optimum was distinctly at 35°C (Supplementary material in Greer 2019). Thus, the severe reduction in quantum efficiency across all temperatures for the Shiraz vines, while common in extent to other studies and different species, was probably related to the severe prevailing climate during growth.

Photochemical quenching increased with increasing PFD (Kalaji et al. 2017), but the effect here was highly temperature dependent, with the most quenching occurring at 20°C and the least quenching at 40°C and consistent with more PSII reaction centres being open (Murchie and Lawson 2013). Notably, at 20°C, photochemical quenching of the Shiraz vines was increased by 77% at steady state, while the comparable figure at 40°C was 45%. These responses clearly support the interpretation that the Shiraz leaves, and PSII photochemistry in particular, were highly tolerant of the high temperatures. By comparison, field-grown Camellia sinensis plants (Mohotti and Lawlor 2002) had a 30% increase in photochemical quenching when exposed to comparable PFDs. Similarly, for Mangifera indica trees (Urban et al. 2004), a 56% increase in photochemical quenching occurred when exposed to saturating PFDs. Both Greer and Halligan (2001) and Zhang and Yin (2012) have shown photochemical quenching is dependent on the growth irradiance, with increased quenching as the growth light decreases. Similarly, Bigras (2000), Haldimann and Feller (2004), and Greer (2019) have all shown for different species that the extent of photochemical quenching is highly temperature dependent and mostly optimal (least quenching) at approximately 35°C, although in the present study, this was at 40°C. The markedly greater photochemical quenching at 20°C in the Shiraz vines in the 2018/2019 growing season (0.23) compared to that for Shiraz vines in the 2017/2018 growing season (0.61; Greer 2019), was consistent with marked PSII reaction centre closure, and again probably related to the severe climate in the present study but also reflected the high sensitivity to low temperatures.

This study confirmed that a strong interaction between temperature and CO₂ affected the photosynthetic response of the Shiraz vines to light. While elevating CO₂ had an expected response on increasing CO₂ assimilation, what was less evident was that the light-saturated maximum assimilation rate, the apparent photon yield, and light saturation were all apparently more sensitive to temperature at low to moderate CO₂ compared with elevated CO₂. The interaction was also supported by the PSII electron transport rate, which followed comparable kinetics to increasing PFD as assimilation. However, in contrast to CO₂ assimilation, light-saturated ETR increased with elevated CO₂ to a maximum that was highly temperature dependent, at 600 μmol mol⁻¹ at 25°C and at 400 μmol mol⁻¹ at 35°C, and 1000 μmol mol⁻¹ CO₂ concentration appeared inhibitory to electron transport at all temperatures. Despite the interactive effect on PSII electron transport, there was no interaction between temperature and CO₂ on chlorophyll a fluorescence, largely because CO₂ concentration had no effect whereas temperature had marked, but well-established, effects on the PFD induced decline in the fluorescence attributes. In contrast to previous work where crop removal had dramatic effects on Shiraz leaf assimilation and PSII quantum efficiency, especially response to temperature, the current study did not demonstrate any change in temperature dependency with increasing substrate supply and clearly suggested the high sink demand of ripening grape bunches had a greater effect on assimilation than on chloroplast light interception. However, several attributes of fluorescence and photosynthesis indicated the severe climate that the vines were exposed to during berry ripening in the current growing season most probably compromised vine performance.

Supplementary material is available online.

The data that support this study will be shared upon reasonable request to the corresponding author.

The author declares no conflicts of interest.

This project had no funding attached to the author.