Drivers of Natural Variation in Water-Use Efficiency Under Fluctuating Light Are Promising Targets for Improvement in Sorghum

Improving leaf intrinsic water-use efficiency (iWUE), the ratio of photosynthetic CO₂ assimilation to stomatal conductance, could decrease crop freshwater consumption. iWUE has primarily been studied under steady-state light, but light in crop stands rapidly fluctuates. Leaf responses to these fluctuations substantially affect overall plant performance. Notably, photosynthesis responds faster than stomata to decreases in light intensity: this desynchronization results in substantial loss of iWUE. Traits that could improve iWUE under fluctuating light, such as faster stomatal movement to better synchronize stomata with photosynthesis, show significant natural diversity in C₃ species. However, C₄ crops have been less closely investigated. Additionally, while modification of photosynthetic or stomatal traits independent of one another will theoretically have a proportionate effect on iWUE, in reality these traits are inter-dependent. It is unclear how interactions between photosynthesis and stomata affect natural diversity in iWUE, and whether some traits are more tractable drivers to improve iWUE. Here, measurements of photosynthesis, stomatal conductance and iWUE under steady-state and fluctuating light, along with stomatal patterning, were obtained in 18 field-grown accessions of the C₄ crop sorghum. These traits showed significant natural diversity but were highly correlated, with important implications for improvement of iWUE. Some features, such as gradual responses of photosynthesis to decreases in light, appeared promising for improvement of iWUE. Other traits showed tradeoffs that negated benefits to iWUE, e.g., accessions with faster stomatal responses to decreases in light, expected to benefit iWUE, also displayed more abrupt losses in photosynthesis, resulting in overall lower iWUE. Genetic engineering might be needed to break these natural tradeoffs and achieve optimal trait combinations, e.g., leaves with fewer, smaller stomata, more sensitive to changes in photosynthesis. Traits describing iWUE at steady-state, and the change in iWUE following decreases in light, were important contributors to overall iWUE under fluctuating light.

Introduction

Water is the primary abiotic factor limiting crop productivity (Boyer, 1982), with agriculture consuming up to 85% of freshwater withdrawals (WWAP, 2015; D’Odorico et al., 2020). Breeding has almost tripled productivity of major crops over the last 60 years, without parallel improvement in the amount of water required to produce a ton of crop biomass (Sinclair et al., 1984; Ort and Long, 2014). With changing patterns of precipitation and increased drought frequency (Dai, 2013; Spinoni et al., 2018; Sun et al., 2018), rising atmospheric vapor pressure deficit due to global warming (Ort and Long, 2014), and decreasing groundwater supply around the world (Dalin et al., 2017), supplying sufficient water to crops is increasingly difficult and unsustainable (Lobell et al., 2014; WWAP, 2015). Therefore, improving crop water-use efficiency is important to achieve the crop productivity required to meet global demand (Bonsch et al., 2016; Flexas, 2016; FAO et al., 2018; Leakey et al., 2019).

The dependence of crop productivity on water supply derives from the interaction between leaf photosynthetic CO₂ assimilation (A) and transpiration. Stomatal pores on the leaf surface allow CO₂ diffusion into the leaf for A, but also allow water vapor to escape from the leaf. The inverse of the resistance to water vapor loss collectively imposed by the stomatal pores is measured as the stomatal conductance to water vapor (g_s). The ratio of A to g_s gives leaf intrinsic water-use efficiency (iWUE) (Leakey et al., 2019). Most inter- and intra-specific surveys of A, g_s and iWUE, and analyses of their limitations have concerned steady-state conditions (Galmes et al., 2007; Giuliani et al., 2013; Driever et al., 2014; Jahan et al., 2014; Sollenberger et al., 2014; Viswanathan et al., 2014; Tomeo and Rosenthal, 2017; Yabiku and Ueno, 2017; Leakey et al., 2019).

Although steady-state measurements are important to understand plant physiological performance (von Caemmerer and Farquhar, 1981; von Caemmerer, 2000), leaves in field-grown crop canopies are rarely in steady-state light conditions, and experience frequent fluctuations in incident photosynthetic photon flux density (PPFD) that cause substantial deviations from steady-state carbon and water fluxes (Pearcy, 1990; Zhu et al., 2004; Kaiser et al., 2015, 2018; McAusland et al., 2016; Vialet-Chabrand et al., 2017; Slattery et al., 2018; Acevedo-Siaca et al., 2020; De Souza et al., 2020; Wang et al., 2020). Evidence of increased water deficit stress (Sun et al., 2017), temperature stress (Leakey et al., 2003), and sensitivity of A to elevated CO₂ (Leakey et al., 2002; Tomimatsu and Tang, 2016) under fluctuating PPFD, and recent breakthroughs in improving A under fluctuating PPFD by manipulating photoprotection (Kromdijk et al., 2016; Hubbart et al., 2018), chlorophyll content (Gu et al., 2017), and stomatal light-sensing (Papanatsiou et al., 2019), have demonstrated the value of considering leaf behavior under non-steady-state lighting conditions.

A and g_s typically require several minutes to rise to a new steady-state following an increase in PPFD, as in a sun fleck within a crop canopy. The rate of increase in g_s may be mechanically limited by stomatal opening kinetics and stomatal patterning. The rate of increase in A may be limited by the kinetics of biochemical processes, in particular Rubisco activation (Soleh et al., 2016, 2017; Acevedo-Siaca et al., 2020), or the kinetics of stomatal opening and the associated mitigation of intercellular CO₂ (c_i)-limitation to A (Lawson and Blatt, 2014; McAusland et al., 2016; De Souza et al., 2020). Because the response times of A and g_s are similar following an increase in PPFD, the deviation of iWUE from steady-state is relatively modest (McAusland et al., 2016). In contrast, the response times of A and g_s are more distinct following a decrease in PPFD, as A typically declines to a new steady-state within seconds while g_s decreases over the course of several minutes as stomata close (Lawson et al., 2011; Way and Pearcy, 2012; Lawson and Blatt, 2014; McAusland et al., 2016; Lawson and Vialet-Chabrand, 2019). This lag in g_s relative to A leads to continued water loss, resulting in reduced iWUE. While the decrease in A is often simplified as a step-change from one steady-state to the next (McAusland et al., 2016; Bellasio et al., 2017), physiological processes such as post-illumination decarboxylation of photorespiratory metabolites and slow relaxation of photoprotective mechanisms may temporarily suppress A following a decrease in PPFD (Doncaster et al., 1989; Kaiser et al., 2015; Kromdijk et al., 2016; Wang et al., 2020).

Attempts to improve leaf physiology under fluctuating light have generally followed three axes: (1) accelerate induction of A following increases in PPFD (Soleh et al., 2016, 2017; Taylor and Long, 2017; Acevedo-Siaca et al., 2020; Wang et al., 2020), (2) limit the inhibition of A following decreases in PPFD (Kromdijk et al., 2016; Wang et al., 2020), (3) accelerate stomatal movement to reduce g_s response time, leading to improved A following increases in PPFD and improved iWUE following decreases in PPFD (Wang et al., 2014; Papanatsiou et al., 2019; Lawson and Matthews, 2020). An essential component of this research has been the discovery of broad diversity in A, g_s, and iWUE in fluctuating light across species (McAusland et al., 2016; Deans et al., 2019) and within C₃ species such as rice (Acevedo-Siaca et al., 2020), cassava (De Souza et al., 2020), and soybean (Soleh et al., 2016, 2017; Wang et al., 2020). In contrast, within-species diversity in C₄ species is not well understood, despite use of the C₄ pathway by some of the world’s most productive crops, including maize (Zea mays L.), sorghum (Sorghum bicolor (Lu.) Moench), and sugarcane (Saccharum officinarum L.). Further, while modification of A or g_s traits independent of one another will theoretically have a proportionate effect on iWUE, in reality these traits are inter-dependent. It is unclear how interactions between these traits affect natural diversity in iWUE, and whether some traits might be more tractable drivers for improvement of iWUE. In C₄ crops, there is the added complexity of the carbon-concentrating mechanism, which may alter the relationship between A and g_s relative to C₃ species (McAusland et al., 2016). Similarly, it is unclear whether steady-state and non-steady-state traits interact with one another and are predictive of overall leaf gas-exchange under fluctuating light.

Here, steady-state and non-steady-state iWUE were examined in the model C₄ crop sorghum. Sorghum is a key food crop for water-limited regions of the globe and fiber sorghum is among the most productive potential cellulosic bioenergy crops of the warm temperate zone, characterized by high drought-tolerance water-use efficiency and productivity (Regassa and Wortmann, 2014; Hadebe et al., 2017). This study asked three questions:

(1) Is there significant diversity in non-steady-state responses of A, g_s and iWUE to light fluctuations among diverse sorghum accessions?

(2) What are the main drivers for natural variation in iWUE in terms of steady-state A and g_s, non-steady-state A and g_s, and stomatal patterning?

(3) Are steady-state and non-steady-state gas-exchange traits predictive of leaf performance under fluctuating PPFD?

We analyzed gas exchange and stomatal patterning in 18 diverse sorghum accessions under steady-state and fluctuating PPFD conditions. Our results identify drivers for improvement of A and iWUE, including stomatal size and non-steady-state A and iWUE traits, and demonstrate that the sorghum pan-genome harbors significant potential for improving the already impressive iWUE of sorghum via breeding.

Materials and Methods

Plant Material and Growing Conditions

Seeds of 18 sorghum accessions were planted on May 28, 2017 at the University of Illinois Energy Farm near Urbana, IL, United States (40°07′N, 5 88°21′W, 228 m above sea level) in 3 m rows at 25 seeds m^–1, in plots of four rows spaced 76 cm apart. The accessions were a subset of a biomass sorghum diversity panel that was previously screened to identify transient stomatal responses to a decrease in PPFD (Pignon, 2017). Soils are deep silt loam Flanagan and silty clay loam Drummer. Plants were fertilized (101 kg/ha N) and rainfed. In August, the youngest fully expanded leaf was sampled from 3 to 5 plants per accession. Leaves were cut pre-dawn, recut with the cut-end underwater, and transferred to the laboratory for analysis.

Gas Exchange Measurements

Leaf cuvettes of two portable photosynthetic gas-exchange systems (LI-6400XT; LI-COR, Inc., Lincoln, NE, United States with LI-6400-40 Leaf-Chamber-Fluorometer) were clamped onto each side of the midrib at the middle of each leaf. PPFD was set to 2000 μmol m^–2 s^–1, block temperature to 25°C, flow rate to 700 μmol s^–1, [CO₂] in the sample cell set to 400 ppm and leaf-to-air VPD maintained <2 kPa. LEDs provided 10% blue and 90% red light. Leaves were acclimated to these conditions for an hour, then measurements began.

Because increases in PPFD are typically less impactful to iWUE than decreases in PPFD, this study focused on extracting non-steady-state traits following decreases in PPFD. For each leaf, one cuvette measured steady-state PPFD response curves, which were used to obtain A, g_s, and iWUE at steady-state, and to evaluate the transition from one steady-state to the next after decreases in PPFD. Curves were obtained by decreasing PPFD (2000, 1500, 1000, 500, 200, 100, and 0 μmol m^–2 s^–1) in 15 min steps, with data recorded every 10 s. Though useful for the purposes of quantifying steady-state and non-steady-state gas-exchange, this PPFD timecourse was highly artificial and provided limited information about repeated PPFD fluctuations, since it was only composed of lengthy step declines in PPFD. Therefore, in the second cuvette, additional measurements were used to determine whether the traits derived from steady-state PPFD response curves were predictive of performance in a repeatedly fluctuating PPFD environment. Fluctuating PPFD response curves were obtained by varying PPFD in 5.5 min steps in the following sequence: 2000, 1500, 2000, 1000, 2000, 500, 2000, 200, 2000, and 100 μmol m^–2 s^–1, with data recorded every 10 s.

Variation in environmental variables throughout gas-exchange measurements is given in Supplementary Table 1. On completion of gas-exchange the measured leaf tissue was sampled and frozen for stomatal phenotyping.

Stomatal Phenotyping

Leaf tissue was mounted on glass slides and the abaxial surface was imaged using an optical topometer (μsurf Explorer, NanoFocus, Karlsruhe, Germany) with 20× (0.8 mm² leaf surface, 20× M Plan APO, Olympus Corporation, Tokyo, Japan) and 50× air objectives (0.32 mm² leaf surface, 50× UM Plan FL N, Olympus Corporation). Reconstructions of the leaf epidermis were obtained from serial optical sections measured from surface to inside of leaf, with measurement depth of 40 μm and 20 μm for the 20× and 50× objectives, respectively (μsurf Metrology, NanoFocus). Four images at 20×, and one image at 50×, were taken per leaf. Stomata on 20× images were counted to estimate stomatal density. The size, i.e., planar surface area, of the stomatal complex was measured by outlining relevant pixels of four stomata per 50× image using ImageJ (ImageJ1.51j8, NIH, United States).

Gas-Exchange Data Analysis

Analysis of Steady-State PPFD Response Curves

Steady-state net rate of leaf photosynthetic CO₂ uptake (A₂₀₀₀), stomatal conductance to water vapor (g_{s 2000}) and intrinsic water-use efficiency (iWUE₂₀₀₀) at PPFD = 2000 μmol m^–2 s^–1 were obtained as the average for the two cuvettes (i.e., two technical reps) for each leaf over the last 40 s of the initial 1 h acclimation period. Steady-state A, g_s, and iWUE were also obtained for each PPFD in the steady-state PPFD response curve, as the average over the last 40 s of each PPFD.

After each decrease in PPFD, A declined from one steady-state to the next, often displaying a temporary inhibition as it decreased below steady-state (i.e., undershoot), then increased again to steady-state. This process was described by two traits: t_{95 A} and A_undershoot. t_{95 A} was the time required for A to come within 5% of steady-state after a decrease in PPFD, where smaller t_{95 A} indicates that A reached steady-state more rapidly. A_undershoot was the difference between steady-state A and the minimum A reached at a given PPFD, where more negative A_undershoot indicates more pronounced undershoot of steady-state. Corresponding traits were used to describe g_s (t_{95 gs}, g_{s undershoot}).

The response of iWUE to each decrease in PPFD was different from that of A and g_s, as iWUE first abruptly declined, reflecting an instantaneous loss of A while g_s remained relatively high. In the following minutes, as g_s declined, iWUE increased toward steady-state. t_{95 iWUE} was calculated analogous to t_{95 A} and t_{95 gs}, and iWUE_undershoot was calculated analogous to A_undershoot and g_{s undershoot}.

Analysis of Fluctuating PPFD Response Curves

Average A, g_s and iWUE were computed throughout the fluctuating PPFD response curves. If responses of A, g_s, and iWUE to changes in PPFD were instantaneous, then gas-exchange under fluctuating PPFD would be equivalent to steady-state gas-exchange at each PPFD. Instead, A, g_s, and iWUE showed substantial deviations from steady-state following changes in PPFD. Therefore, overall gas-exchange under fluctuating PPFD could be described as the sum of steady-state gas-exchange and the deviation of gas-exchange from steady-state following each PPFD fluctuation. To assess the importance of non-steady-state traits separate from steady-state, the deviation of gas-exchange from steady-state under fluctuating PPFD was calculated. Because steady-state and fluctuating PPFD response curves were measured on the same leaves, data from both types of curves were combined. Deviation of A from steady-state was calculated as follows: (1) A from the fluctuating PPFD response curves was normalized to A₂₀₀₀, (2) steady-state A from the steady-state PPFD response curve was obtained for each PPFD, and also normalized to A₂₀₀₀, (3) The difference between (1) and (2) was calculated at each PPFD, yielding the deviation of A from steady-state, normalized to A₂₀₀₀. This deviation is positive when A is greater than steady-state, negative when A is less than steady-state, and 0 when A is equal to steady-state. The average deviation of A from steady-state was calculated following increases and decreases in PPFD. The same method was applied to iWUE.

To test whether leaves cut from plants and measured in the laboratory might behave differently from leaves still attached to plants, an experiment was performed comparing the two types of leaves. There was no substantial effect of leaf excision on steady-state or non-steady-state A, g_s, and iWUE (Supplementary Figure 9).

Statistical Analysis

A summary of all traits analyzed in this study is given in Table 1. ANOVA was used to test the fixed effect of sorghum accession on gas-exchange and physiology traits with the aov() function in the package stats (R 3.6.1, R Core Team, 2017). There were 3–5 plants sampled per accession, with one plot per accession, such that each plant was a pseudo replicate. Time of day of measurements and leaf-to-air VPD were included as cofactors for gas-exchange measurements. If neither cofactor was significant for a given trait, then a simpler model testing only the effect of accession was used. A₂₀₀₀, g_{s 2000}, and iWUE₂₀₀₀ were the average of two technical replicates per plant, and stomatal density and size were the average of four technical replicates per plant; all other measurements had one technical replicate per plant.

TABLE 1

Table 1. Trait abbreviations and definitions.

Homogeneity of variances was tested by the Levene test using function LeveneTest() in package DescTools (Signorell, 2020) and normality of studentized residuals tested by Shapiro–Wilk using function ols_test_normality in package olsrr (Hebbali, 2020) at p = 0.01 threshold. The assumption of normality was violated for t_{95 A}, so the Kruskal–Wallis test was used to analyze this trait using kruskal.test() function in package stats (R Core Team, 2017).

Pairwise correlations were tested at p < 0.05 (significant) and p < 0.1 (marginally significant) thresholds between means per accession for the traits described above using cor.mtest() function in package corrplot (Wei and Simko, 2017).

Results

Variation Among Accessions in Steady-State Gas Exchange and in the Transition From One Steady-State to the Next Following Decreases in PPFD

Two representative accessions, PI153852 and PI152636, exemplify the genetic variation that was observed in steady-state and non-steady-state A, g_s, and iWUE (Figure 1). In PI153852, steady-state A and g_s at PPFD = 2000 μmol m^–2 s^–1 were less than in PI152636 (i.e., A₂₀₀₀ and g_{s 2000}, pink datapoints in Figures 1A–D, respectively). After each decrease in PPFD, A and g_s declined from one steady-state to the next. In PI153852, decline of A occurred over the course of several minutes, during which A decreased below steady-state (i.e., undershoot), then increased again to steady-state (Figure 1A). In contrast, in PI152636, A reached a new steady-state in under a minute, with a less pronounced undershoot (Figure 1B). In PI153852, g_s gradually decreased below steady-state (i.e., undershoot), then increased again to steady-state (Figure 1C). In contrast, in PI152636 g_s declined more rapidly to steady-state, with a slightly less pronounced undershoot (Figure 1D).

FIGURE 1

Figure 1. Timecourses of steady-state PPFD response curves in two accessions: (A,C,E) PI153852 and (B,D,F) PI152636. For each curve, leaves were acclimatized to PPFD of 2000 μmol m^–2 s^–1 for 1 h, then PPFD declined in steps every 15 min. (A,B) A, (C,D) g_s, and (E,F) iWUE were logged throughout. Each point is a mean ± s.e. of 4 plants. The final measurements at each PPFD were used to estimate steady-state A, g_s, and iWUE at each PPFD. Timecourses were also used to describe the transition from one steady-state to the next after each decrease in PPFD, including the time for A, g_s, and iWUE to come within 5% of steady-state, and the undershoot of steady-state by A, g_s, and iWUE.

After each decrease in PPFD, iWUE abruptly declined below steady-state (i.e., undershoot), reflecting an instantaneous loss of A while g_s remained relatively high. This decline was less pronounced in PI153852 (Figure 1E) than PI152636 (Figure 1F). In the following minutes, iWUE increased to reach steady-state, reflecting a decline of g_s to steady-state. This occurred slightly more rapidly in PI153852 (Figure 1E) than in PI152636 (Figure 1F). This suggests that the differences between these two accessions in A and g_s at steady-state and non-steady-state translated to differences in iWUE. Steady-state iWUE was relatively stable from PPFD = 500–2000 μmol m^–2 s^–1, but began to decline at lower PPFD (Figures 1E,F and Supplementary Figure 1).

Among all 18 accessions, there was significant variation in A₂₀₀₀ (p < 0.001, 1.63-fold variation among accessions, Figure 2A), g_{s 2000} (p < 0.001, 1.78-fold variation among accessions, Figure 2B), and iWUE₂₀₀₀ (p = 0.001, 1.29-fold variation among accessions, Figure 2C), as well as the time required for A (t_{95 A}, p = 0.010, 11-fold variation among accessions, Figure 2D) and g_s (t_{95 gs}, p = 0.041, 1.98-fold variation among accessions, Figure 2E) to come within 5% of steady-state after a decrease in PPFD. However, there was no significant variation in the time required for iWUE to come within 5% of steady-state after a decrease in PPFD (t_{95 iWUE}, p = 0.318, 1.65-fold variation among accessions, Figure 2F). There was significant variation in the undershoot of steady-state by A (A_undershoot, p = 0.010, 2.50-fold variation among accessions, Figure 2G), g_s (g_{s undershoot}, p = 0.019, 2.51-fold variation among accessions, Figure 2H), and iWUE (iWUE_undershoot, p < 0.001, 1.98-fold variation among accessions, Figure 2I) after each decrease in PPFD. A approached steady-state within seconds in most accessions (t_{95 A} < 1 min, Figure 2D), with the slowest decline of A in PI153852 (t_{95 A} = 2.86 min, Figures 1A, 2D). g_s and iWUE approached steady-state more slowly than A (t_{95 gs} and t_{95 iWUE} ranging from 2.22 to 4.40 min, Figures 2E,F).

FIGURE 2

Figure 2. Bar graphs of traits derived from steady-state PPFD response curves: (A) steady-state A at PPFD = 2000 μmol m^–2 s^–1 (A₂₀₀₀), (B) steady-state g_s at PPFD = 2000 μmol m^–2 s^–1 (g_{s 2000}), (C) steady-state iWUE at PPFD = 2000 μmol m^–2 s^–1 (iWUE₂₀₀₀), (D) time required for A to come within 5% of steady-state after a decrease in PPFD (t_{95 A}), (E) time required for g_s to come within 5% of steady-state after a decrease in PPFD (t_{95 gs}), (F) time required for iWUE to come within 5% of steady-state after a decrease in PPFD (t_{95 iWUE}), (G) undershoot of steady-state by A after a decrease in PPFD, i.e., difference between steady-state A and the minimum A reached at each PPFD (A_undershoot), (H) undershoot of steady-state by g_s after a decrease in PPFD, i.e., difference between steady-state g_s and the minimum g_s reached at each PPFD (g_{s undershoot}), (I) undershoot of steady-state by iWUE after a decrease in PPFD, i.e., difference between steady-state iWUE and the minimum iWUE reached at each PPFD (iWUE_undershoot). Bars are mean ± s.e. p-values are from ANOVA testing the fixed effect of accession on each trait.

Variation Among Accessions in Gas Exchange Under Fluctuating PPFD

Fluctuating PPFD response curves were used to determine whether the traits derived from steady-state PPFD response curves were predictive of performance in a fluctuating PPFD environment. Following decreases in PPFD, A rapidly declined to steady-state, while increases in PPFD triggered more gradual increases of A toward steady-state (Figure 3A,B). As a result, A was slightly above steady-state following decreases in PPFD, but substantially below steady-state following increases in PPFD (Figures 3C,D). As in the steady-state PPFD response curves, in accessions such as PI152636 there was little deviation of A from steady-state following decreases in PPFD, whereas in PI153852, A remained above steady-state for several minutes (Figures 3C,D).

FIGURE 3

Figure 3. Timecourses of fluctuating PPFD response curves in two accessions: (A,C,E,G) PI153852 and (B,D,F,H) PI152636. For each curve, leaves were acclimatized to PPFD of 2000 μmol m^–2 s^–1 for 1 h, then PPFD cycled between non-saturating and saturating PPFD every 5.5 min. (A,B) A, (C,D) Deviation of A from steady-state, normalized to A₂₀₀₀, (E,F) iWUE, (G,H) Deviation of iWUE from steady-state, normalized to iWUE₂₀₀₀. Each point is a mean ± s.e. of 3–5 plants. Deviation of A from steady-state was calculated as follows: (1) A from a fluctuating PPFD response curve was normalized to A₂₀₀₀, (2) the steady-state A from a steady-state PPFD curve measured on the same leaf was obtained for each PPFD, and also normalized to A₂₀₀₀, (3) The difference between (1) and (2) was calculated, yielding the deviation of A from steady-state, normalized to A₂₀₀₀, as seen in panels (C,D). This deviation is positive when A is greater than steady-state, negative when A is less than steady-state, and 0 when A is equal to steady-state. The same method was applied to iWUE, as seen in panels (G,H).

Following decreases in PPFD, iWUE abruptly declined, then gradually increased toward steady-state (Figures 3E,F). In contrast, increases in PPFD caused iWUE to rise slightly above steady-state, then return to steady-state (Figures 3E,F). In other words, iWUE was slightly above steady-state following increases in PPFD, but substantially less than steady-state following decreases in PPFD (Figures 3G,H). This resulted from the fact that A declined faster than g_s following decreases in PPFD, whereas A and g_s increased at a similar rate following increases in PPFD (Supplementary Figures 5–7). In accessions such as PI152636, there was substantial loss of iWUE relative to steady-state following decreases in PPFD, and a slight gain of iWUE relative to steady-state following increases in PPFD, compared to the less pronounced deviation of iWUE from steady-state in accessions such as PI153852 (Figures 3G,H).

Among all accessions, there was significant or marginally significant variation in average A under fluctuating PPFD (p = 0.002, 1.50-fold variation among accessions, Figure 4A), average g_s under fluctuating PPFD (p = 0.037, 1.78-fold variation among accessions, Figure 4B), and average iWUE under fluctuating PPFD (p = 0.069, 1.22-fold variation among accessions, Figure 4C), as well as the deviation of A from steady-state following increases in PPFD (p = 0.010, 2.50-fold variation among accessions, Figure 4D), the deviation of iWUE from steady-state following increases in PPFD (p = 0.088, 4.02-fold variation among accessions, Figure 4E), and the deviation of A from steady-state following decreases in PPFD (p = 0.078, 3.07-fold variation among accessions, Figure 4F). However, the experiment could not resolve significant differences among accessions for the deviation of iWUE from steady-state following decreases in PPFD (p = 0.121, 3.15-fold variation among accessions, Figure 4G). Finally, there was significant variation among accessions in stomatal density (p < 0.001, 1.72-fold variation among accessions, Figure 4H) and stomatal size (p < 0.001, 1.59-fold variation among accessions, Figure 4I) determined from optical topometry of the epidermis (Figure 5).

FIGURE 4

Figure 4. Bar graphs of traits derived from fluctuating PPFD response curves and leaf stomatal profiles: (A) Average A under fluctuating PPFD, (B) Average g_s under fluctuating PPFD, (C) Average iWUE under fluctuating PPFD, (D) Deviation of A from steady-state following increases in PPFD, expressed as a% of A₂₀₀₀, (E) Deviation of iWUE from steady-state following increases in PPFD, expressed as a% of iWUE₂₀₀₀, (F) Deviation of A from steady-state following decreases in PPFD, expressed as a% of A₂₀₀₀, (G) Deviation of iWUE from steady-state following decreases in PPFD, expressed as a% of iWUE₂₀₀₀, (H) stomatal density, (I) stomatal size, i.e., planar surface area of the stomatal complex. Bars are mean ± s.e. p-values are from ANOVA testing the fixed effect of accession on each trait.

FIGURE 5

Figure 5. Optical topometry images of leaf abaxial surfaces used for stomatal profiles, measured with: (A,B), 20× magnification, (C,D), 50× magnification. Panels (A,C) show representative examples for an accession (PI524469) with low stomatal density (73 stomata mm^–2) and large stomata (964 μm²), and (B,D) for an accession (PI329719) with high stomatal density (125 stomata mm^–2) and small stomata (734 μm²).

Correlations Between A, g_s, iWUE, and Stomatal Patterning Traits

Traits Correlated With Steady-State A, g_s, and iWUE

Steady-state g_{s 2000} was positively correlated with A₂₀₀₀ (p < 0.001, R² = 0.87, Figure 6) and negatively correlated with iWUE₂₀₀₀ (p = 0.0019, R² = 0.46, Figure 6). After a decrease in PPFD, iWUE increased toward steady-state more slowly in accessions with greater A₂₀₀₀ and g_{s 2000} (positive correlation of A₂₀₀₀ with t_{95 iWUE}, p = 0.094, R² = 0.17; positive correlation of g_{s 2000} with t_{95 iWUE}, p = 0.059, R² = 0.2, Figure 6). After a decrease in PPFD, undershoot of steady-state iWUE was more pronounced in accessions with greater A₂₀₀₀ and g_{s 2000} (negative correlation of A₂₀₀₀ with iWUE_undershoot, p = 0.03, R² = 0.26; negative correlation of g_{s 2000} with iWUE_undershoot, p = 0.074, R² = 0.19, Figure 6). Accordingly, deviation of iWUE from steady-state following decreases in PPFD was more negative in accessions with greater A₂₀₀₀ (p < 0.001, R² = 0.61, Figure 6) and greater g_{s 2000} (p < 0.001, R² = 0.52, Figure 6).

FIGURE 6

Figure 6. Pearson’s correlation coefficients (r) for traits potentially underlying variation in non-steady-state iWUE following decreases in PPFD. The size of circles gives the strength of correlation and the color gives the direction of correlation: red (negative) and blue (positive). Significance (p < 0.1 threshold) is marked by *. The corresponding pairwise correlation scatterplots are given in Supplementary Figure 8.

Traits Correlated With Non-Steady-State A

For non-steady-state A to be maximized following a decrease in PPFD, A should slowly approach steady-state (i.e., high t_{95 A}) with minimal undershoot (i.e., greater, i.e., less negative, A_undershoot). This non-steady-state A may also be related to non-steady-state g_s. Accordingly, accessions with a more positive deviation of A from steady-state following decreases in PPFD, also had greater t_{95 A} (p = 0.0088, R² = 0.36, Figure 6), greater t_{95 gs} (p = 0.0062, R² = 0.38, Figure 6), and less negative g_{s undershoot} (p = 0.078, R² = 0.18, Figure 6). In accessions in which A and g_s approached steady-state more slowly (i.e., greater t_{95 A} and t_{95 gs}), the undershoot of steady-state A and g_s was less pronounced (i.e., less negative A_undershoot and g_{s undershoot}). This was evidenced by the positive correlation of t_{95 A} with A_undershoot (p = 0.063, R² = 0.2, Figure 6) and g_{s undershoot} (p = 0.0033, R² = 0.43, Figure 6) and the positive correlation of t_{95 gs} with A_undershoot (p = 0.0081, R² = 0.36, Figure 6) and g_{s undershoot} (p < 0.001, R² = 0.54, Figure 6).

Non-steady-state A and g_s were coordinated following a decrease in PPFD. In accessions in which A approached steady-state more slowly (i.e., greater t_{95 A}) with a less pronounced undershoot (i.e., less negative A_undershoot), the same was also seen for g_s (i.e., greater t_{95 gs} and less negative g_{s undershoot}). This was evidenced by the positive correlation of t_{95 A} with t_{95 gs} (p < 0.001, R² = 0.63, Figure 6) and the positive correlation of A_undershoot with g_{s undershoot} (p < 0.001, R² = 0.66, Figure 6).

Traits Correlated With Non-Steady-State iWUE

For non-steady-state iWUE to be maximized following a decrease in PPFD, iWUE should rapidly approach steady-state (i.e., small t_95iWUE) with minimal undershoot (i.e., greater, i.e., less negative iWUE_undershoot). Accordingly, accessions with a less negative deviation of iWUE from steady-state following decreases in PPFD also had smaller t_95iWUE (p = 0.051, R² = 0.21, Figure 6), and greater iWUE_undershoot (p = 0.0074, R² = 0.37, Figure 6).

Non-steady-state iWUE traits were associated with A rather than g_s. Specifically, in accessions in which iWUE approached steady-state more slowly (i.e., greater t_95iWUE), A approached steady-state more rapidly with a more pronounced undershoot (i.e., smaller t_{95 A} and more negative A_undershoot). This was evidenced by the negative correlation of t_95iWUE with t_{95 A} (p = 0.085, R² = 0.17, Figure 6) and A_undershoot (p = 0.035, R² = 0.25, Figure 6). Additionally, accessions which displayed a less pronounced undershoot of steady-state by iWUE following decreases in PPFD (i.e., greater, i.e., less negative iWUE_undershoot) also had a slower decrease of A and g_s to reach steady-state (positive correlation of iWUE_undershoot with t_{95 A}, p < 0.001, R² = 0.65; positive correlation of iWUE_undershoot with t_{95 gs}, p = 0.0021, R² = 0.46, Figure 6). Further, in accessions which displayed a less pronounced undershoot of steady-state by iWUE, the same was seen for A and g_s (positive correlation of iWUE_undershoot with A_undershoot, p < 0.001, R² = 0.51; positive correlation of iWUE_undershoot with g_{s undershoot}, p = 0.0034, R² = 0.42, Figure 6). Finally, accessions in which iWUE approached steady-state more slowly (i.e., greater t_{95 iWUE}) had a more pronounced undershoot of steady-state iWUE (negative correlation of t_{95 iWUE} with iWUE_undershoot, p = 0.0068, R² = 0.38, Figure 6).

These surprising findings are exemplified in accessions PI153852 and PI152636 (Figure 1). PI152636 displayed faster declines in g_s after each decrease in PPFD (i.e., smaller t_{95 gs}, Figure 1D), which theoretically would lead to increased iWUE, when compared to the slower g_s response of PI153852 (Figure 1C). However, in PI153852 the slower response of g_s was also associated with slower, more gradual declines in A after each decrease in PPFD (i.e., greater t_{95 A}, Figure 1A), which would also theoretically lead to increased iWUE, when compared to the faster A response of PI152636 (Figure 1B). The net result in terms of iWUE yielded a benefit to PI153852 (Figure 1E), with a less pronounced undershoot of steady-state iWUE (i.e., less negative iWUE_undershoot) and a faster return to steady-state iWUE (i.e., smaller t_{95 iWUE}) when compared to PI152636 (Figure 1F). In other words, when comparing PI153852, with slow responses of A and g_s to decreases in PPFD, to PI152636, with fast responses of A and g_s to decreases in PPFD, PI153852 showed the greatest iWUE at non-steady-state.

Correlations With Stomatal Density and Size

Stomatal density and size were negatively correlated (p < 0.001, R² = 0.54, Figure 6). Accessions with more numerous, smaller stomata had greater steady-state g_s and more rapid responses of A and g_s to decreases in PPFD. Specifically, accessions with greater stomatal density had greater g_{s 2000} (p = 0.074, R² = 0.19, Figure 6), lower iWUE₂₀₀₀ (p = 0.032, R² = 0.26, Figure 6), smaller t_{95 A} (p = 0.073, R² = 0.19, Figure 6), smaller t_{95 gs} (p = 0.069, R² = 0.19, Figure 6), and more negative g_{s undershoot} (p = 0.059, R² = 0.21, Figure 6). Accessions with smaller stomata had smaller t_{95 A} (p < 0.001, R² = 0.51, Figure 6), smaller t_{95 gs} (p < 0.001, R² = 0.56, Figure 6), more negative g_{s undershoot} (p = 0.029, R² = 0.26, Figure 6), and more negative iWUE_undershoot (p = 0.024, R² = 0.28, Figure 6).

Overall A and iWUE Under Fluctuating PPFD: Contributions of Steady-State and Non-Steady-State Gas-Exchange

Overall gas-exchange under fluctuating PPFD could be described as the sum of steady-state gas-exchange and the deviation of gas-exchange from steady-state following each PPFD fluctuation. An additional correlation analysis was used to explore whether steady-state and non-steady-state gas-exchange were significantly associated with overall gas-exchange under fluctuating PPFD. Average A and iWUE under fluctuating PPFD were tested for correlation with steady-state A₂₀₀₀ and iWUE₂₀₀₀, and with the deviation of A and iWUE from steady-state following decreases and increases in PPFD. The majority of variation in average A under fluctuating PPFD was associated with variation in steady-state A₂₀₀₀ (p < 0.001, R² = 0.88, Figure 7A) but not with deviation of A from steady-state following decreases (p = 0.41, Figure 7B) or increases (p = 0.58, Figure 7C) in PPFD. The majority of variation in average iWUE under fluctuating PPFD was associated with variation in steady-state iWUE₂₀₀₀ (p < 0.001, R² = 0.61, Figure 7D) and there was also substantial variation associated with deviation of iWUE from steady-state following decreases (p = 0.012, R² = 0.33, Figure 7E), but not increases (p = 0.4, Figure 7F), in PPFD. This suggests a potentially important role for non-steady-state iWUE following decreases in PPFD, causing a substantial loss of iWUE under fluctuating PPFD.

FIGURE 7

Figure 7. Correlation plots of average A under fluctuating PPFD against: (A) A₂₀₀₀, (B) the average deviation of A from steady-state following decreases in PPFD, expressed as a% of A₂₀₀₀, (C) the average deviation of A from steady-state following increases in PPFD, expressed as a% of A₂₀₀₀; and correlation plots of average iWUE under fluctuating PPFD against: (D) iWUE₂₀₀₀, (E) the average deviation of iWUE from steady-state following decreases in PPFD, expressed as a% of iWUE₂₀₀₀, (F) the average deviation of iWUE from steady-state following increases in PPFD, expressed as a% of iWUE₂₀₀₀.

Discussion

Improving iWUE under fluctuating PPFD is important to sustain or further increase crop yields (Leakey et al., 2019). This study shows significant variation among sorghum accessions in steady-state and non-steady-state A and g_s and stomatal density and size, reveals how variation in these traits drives variation in iWUE, and discusses how tradeoffs between these should shape strategies for decreasing crop water use.

Decreases in PPFD Substantially Impair iWUE Under Fluctuating PPFD, With Loss of Non-Steady-State iWUE Associated With A Rather Than g_s

Almost all variation in average A under fluctuating PPFD was associated with steady-state A₂₀₀₀ (Figures 7A–C). In contrast, much of the variation in average iWUE under fluctuating PPFD was associated with the non-steady-state loss of iWUE following decreases in PPFD in addition to steady-state iWUE₂₀₀₀ (Figures 7D–F). This points to non-steady-state iWUE as an important contributor to overall iWUE under fluctuating PPFD, with decreases in PPFD being more impactful than increases in PPFD. This could in large part be attributed to the undershoot of steady-state by iWUE following decreases in PPFD (i.e., iWUE_undershoot), which showed significant variation among accessions and so could be a promising target for improvement.

The response of iWUE to a decrease in PPFD was biphasic, with an abrupt loss of iWUE driven by A, followed by a gradual return to steady-state driven by g_s (Figure 1). Loss of iWUE was mitigated in accessions in which the undershoot of steady-state iWUE was less pronounced (i.e., greater, i.e., less negative, iWUE_undershoot) and iWUE returned to steady-state more rapidly (i.e., smaller t_{95 iWUE}, Figure 6). The finding that t_{95 iWUE} and iWUE_undershoot were associated with corresponding traits of A rather than g_s is novel, and suggests that breeding for improved non-steady-state A traits, i.e., high t_{95 A} and less negative A_undershoot, could mitigate loss of iWUE under fluctuating PPFD (Figure 6).

Non-steady-state A following decreases in PPFD is influenced by different physiological processes. Undershoot of steady-state by A following decreases in PPFD may be attributed to kinetics of protective energy-dissipating mechanisms, collectively termed non-photochemical quenching, and photorespiration (Kaiser et al., 2015, 2018), and has been observed in C₃ dicots such as Arabidopsis, French bean (McAusland et al., 2016) and tobacco (Kromdijk et al., 2016). This could explain the rapid and pronounced undershoot of steady-state A following decreases in PPFD in accessions such as PI329656 (Supplementary Figure 2G). Large pools of metabolites involved in C₄ photosynthesis could buffer energy supply and sustain a higher A for some time following a decrease in PPFD (Stitt and Zhu, 2014). For instance, a large pool of active malate carries enough reductive power to reduce CO₂ for several seconds after a light to dark transition (Slattery et al., 2018). C₄ activity could be insufficient to achieve CO₂ saturation of Rubisco and eliminate photorespiration at low PPFD, leading to reduced steady-state A (Kromdijk et al., 2010). If so, a leaf with CO₂-saturated Rubisco at high PPFD could maintain CO₂-saturation for some time after a decrease in PPFD, effectively maintaining photorespiration below its steady-state and therefore boosting A. This could explain the relatively gradual decline of A following decreases in PPFD in accessions such as PI153852 (Figure 1A), enabling it to maintain A above steady-state for several minutes following decreases in PPFD (Figure 3C). Here, we show that potential for improvement of non-steady-state A in sorghum through breeding is supported by significant variation in t_{95 A} and A_undershoot, which may result from variation in the processes above (Figure 2). This highlights the value of the methodology used in the present study to assess steady-state and non-steady-state gas-exchange traits, and demonstrates that sorghum is a relevant crop species to study diversity in these traits.

Interactions Between Non-Steady-State A and g_s Impair iWUE Following Decreases in PPFD: This Could Be Resolved in Leaves With Smaller, More Sensitive Stomata

Many efforts to improve iWUE under fluctuating light have focused on faster stomatal closure (Lawson and Blatt, 2014; Bellasio et al., 2017; Lawson and Vialet-Chabrand, 2019). In a simplified model of leaf gas-exchange following a decrease in PPFD, where A has an instant step-change from one steady-state to the next, a faster decrease in g_s would directly lead to a faster increase in iWUE, leading to overall improvement in iWUE under fluctuating PPFD. In the sorghum accessions studied here, this process was complicated by interactions between g_s and A, which may reflect precise stomatal sensing of A (reviewed: Lawson and Matthews, 2020). In other words, accessions with rapid decreases in g_s also had rapid decreases in A, negating much of the benefit to iWUE (Figures 1, 6). In fact, the undershoot of steady-state iWUE following decreases in PPFD (i.e., iWUE_undershoot) was most negative in accessions with faster stomatal responses (i.e., smaller t_{95 gs}, Figure 6). If the same coordination between non-steady state A and g_s applies across species, this could explain the observation that faster stomatal closing speed did not translate to improved water saving across diverse plant species (Deans et al., 2019).

Because of this tradeoff, fast g_s response to decreasing PPFD may be a difficult target for improvement of iWUE through breeding in sorghum, though it may be possible to bypass this tradeoff through transgenic means. The optimal leaf response following decreases in PPFD would be a slow decline in A with minimal undershoot of steady-state, paired with a rapid decline in g_s. This might be achieved in leaves with enhanced stomatal sensitivity to A, in which even a slow decline in A following a decrease in PPFD could trigger a rapid stomatal response. In particular, stomatal aperture responds to light via two separate pathways: the photosynthesis-independent and guard-cell specific blue light pathway, and the photosynthesis-dependent red light pathway. The latter is thought to be the main mechanism coordinating stomatal behavior with photosynthesis (Matthews et al., 2020). Therefore, manipulation of components involved in red light sensing, such as the redox state of the chloroplastic plastoquinone pool (Głowacka et al., 2018), could be a good target for manipulation to increase stomatal sensitivity to changes in A and improve coordination of A and g_s.

Another factor influencing the speed of change in g_s is stomatal size, with smaller stomata generally showing faster movement, possibly due to greater guard cell membrane surface area to volume ratio (Drake et al., 2013; Raven, 2014; Lawson and Vialet-Chabrand, 2019; Lawson and Matthews, 2020). Therefore, leaves with smaller stomata might allow greater sensitivity of g_s to A by enabling mechanically faster stomatal closure. However, stomatal size is usually negatively correlated with stomatal density (Hetherington and Woodward, 2003). Here, stomatal density and size were negatively correlated (Figure 6), and leaves with more numerous and smaller stomata had faster t_{95 gs} but also had reduced iWUE₂₀₀₀, more negative iWUE_undershoot, and increased g_{s 2000} (Figures 5, 6). This points to a tradeoff between leaves with more numerous, smaller stomata which show high steady-state g_s but rapid g_s responses to PPFD, and leaves with fewer, larger stomata which show low steady-state g_s but slow g_s responses to PPFD. The finding that steady-state iWUE₂₀₀₀ was mainly associated with g_{s 2000} rather than A₂₀₀₀ is consistent with prior observations in sorghum (Geetika et al., 2019) and other C₄ grasses (Leakey et al., 2019), but the tradeoff with non-steady-state g_s identified here (Figure 6) is not widely recognized.

Sorghum leaves with fewer and smaller stomata might achieve the best of both worlds with low steady-state g_s but rapid stomatal responses to PPFD. In C₄ crops such as sorghum, where photosynthesis is typically CO₂-saturated even under sub-ambient conditions, modestly reducing g_s at steady-state may not impair A, compounding benefits to iWUE (Leakey et al., 2019; Pignon and Long, 2020). Further, at a given SD, C₄ grass stomata are smaller than those of related C₃ grasses (Taylor et al., 2012). Transgenic approaches may hold potential to break the relationship between stomatal density and size (reviewed: Leakey et al., 2019).

Increases in PPFD were much less disruptive to iWUE than decreases in PPFD (Figure 3). The fact that iWUE was slightly above steady-state following increases in PPFD suggests that increase in A was faster than g_s: this is in agreement with findings from a C₄ stomata model (Bellasio et al., 2017) applied to photosynthesis induction data in maize (Chen et al., 2013). By comparison, in many other C₃ and C₄ dicots and monocots, the return of iWUE to steady-state following an increase in PPFD was much slower (e.g., >30 min), reflecting a pronounced desynchronization between A and g_s (McAusland et al., 2016). In our study, the response of g_s to an increase in PPFD occurred within seconds, whereas in C₃ species there may be a lag of up to several minutes before stomata begin to open (Lawson and Blatt, 2014). Together these findings suggest exceptional coordination of A and g_s following increases in PPFD in sorghum.

Relative to Other Species, iWUE in Sorghum Is High Both at Steady-State and Non-Steady-State

The range of natural variation among sorghum accessions tested here in iWUE₂₀₀₀ of 166–215 μmol mol^–1 (Figure 2) is similar to published variation in sorghum RILs (100–140 μmol mol^–1; Kapanigowda et al., 2014) and accessions (143–176 μmol mol^–1; Xin et al., 2009), but less variable compared to measurements in closely related NADP-ME C₄ grasses such as maize (80–140 μmol mol^–1; Yabiku and Ueno, 2017), sugarcane (100–180 μmol mol^–1; Viswanathan et al., 2014) and elephant grass (100–160 μmol mol^–1; Sollenberger et al., 2014), whereas the widest range of variation tends to be found in C₃ species such as soybean (40–115 μmol mol^–1; Tomeo and Rosenthal, 2017), wheat (25–65 μmol mol^–1; Jahan et al., 2014), and rice (50–80 μmol mol^–1; Giuliani et al., 2013).

For both steady-state and non-steady-state traits, variation in iWUE was narrower than for A and g_s (Figures 2, 4). This resulted from coordination in A and g_s, e.g., accessions with high A₂₀₀₀ also had high g_{s 2000} and accessions with high t_{95 A} also had high t_{95 gs} (Figure 6). The accessions studied here showed faster g_s responses to changes in PPFD and higher iWUE compared to diverse gymnosperms and C₃ dicots (Deans et al., 2019), C₃ monocot crops such as wheat and rice, and even closely related C₄ monocots such as maize and Miscanthus (Chen et al., 2013; McAusland et al., 2016). Understanding how sorghum maintains coordination between A and g_s to sustain high iWUE may be valuable to design strategies for improvement in species where coordination is less tight.

In our study, g_s declined over the course of the fluctuating PPFD timecourse, suggesting that stomatal opening during increases in PPFD was slower than stomatal closing during decreases in PPFD (Figure 3). Faster stomatal closure than opening may be a consequence of sorghum’s adaptation to dry, high-light environments, where water is more limiting than light and rapid stomatal closing can maximize iWUE (Vico et al., 2011; McAusland et al., 2016). On the contrary, species adapted to shaded environments such as a forest understory, where light is more limiting than water, typically show faster stomatal opening than closing, which can maximize A with little penalty to iWUE: these patterns suggest that dynamic stomatal traits are driven by ecological adaptation rather than evolutionary lineage (Deans et al., 2019). Increased steady-state iWUE of C₄ photosynthesis may have been a driver for evolution of this pathway (Osborne and Sack, 2012). The fact that C₄ grasses such as sorghum display fast stomatal responses to fluctuating light (Grantz and Assmann, 1991; Knapp, 1993; McAusland et al., 2016) may be an additional evolved mechanism to improve non-steady-state iWUE.

While a drought treatment was not included, we show genetic variation that may be exploited to reduce crop water demand and avoid drought (Leakey et al., 2019). An important next step will be to determine whether the trait correlations identified here are also observed in water-limited plants. Plants that develop under water-limited conditions can produce fewer leaves with fewer and/or smaller stomata to mitigate water loss. In these smaller plants, reduced canopy density may limit the prevalence of light fluctuations and alter the microenvironment including VPD and temperature. However, leaves that develop with sufficient water supply but are water-limited afterward have fewer options to acclimate. With stomatal density and size already fixed during development, stomatal closure is the main mechanism available to reduce steady-state g_s. Leaves that permanently operate in a reduced range of stomatal apertures could have an altered relationship between steady-state traits (e.g., g_{s 2000}) and non-steady-state traits (e.g., t_{95 gs}, g_{s undershoot}). This could also affect the relative association of A and g_s traits with iWUE: at low apertures, stomatal control may be less precise (Kaiser and Kappen, 2001) and so more wasteful in terms of water loss. An important next step will be to determine the degree of plasticity in steady-state and non-steady-state A, g_s and iWUE traits under different environments.

Conclusion

In this study, we show that a common measurement, the steady-state PPFD response curve, can be used to derive valuable insight into both steady-state and non-steady-state A, g_s and iWUE (Figures 1, 2). The relevance of these non-steady-state traits could be seen when leaves were exposed to a fluctuating PPFD regime, as natural diversity in traits such as t_{95 A}, A_undershoot, and iWUE_undershoot correlated with the deviation of A and iWUE from steady-state under fluctuating PPFD (Figures 3, 4, 6). Remarkably, the deviation of A and iWUE from steady-state under fluctuating PPFD was substantial even under the relatively lengthy PPFD fluctuations, spaced 5.5 min apart, of the fluctuating PPFD response curves used here (Figure 7). In a crop canopy, where most light fluctuations are more rapid (<5 s) (Kaiser et al., 2018), the non-steady-state processes quantified here, especially the photosynthetic traits t_{95 A} and A_undershoot, would likely be even more important in driving overall iWUE. Variation among accessions in steady-state and non-steady-state traits may be exploited to reduce crop water demand and avoid drought, but our results emphasize that translating this into breeding strategies will require careful consideration of emerging tradeoffs due to co-variation between traits.

Data Availability Statement

The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding author/s.

Author Contributions

CP designed and performed the research, data collection, analysis and interpretation, and wrote the manuscript. AL, SL, and JK performed the data analysis and interpretation and wrote the manuscript. All authors contributed to the article and approved the submitted version.

Funding

The information, data, or work presented herein was funded in part by the Advanced Research Projects Agency-Energy (ARPA-E), United States Department of Energy, under Award Number DE-DE-AR0000661. The views and opinions of authors expressed herein do not necessarily state or reflect those of the United States Government or any agency thereof.

Conflict of Interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Acknowledgments

The authors thank Dr. Patrick J. Brown for developing the diversity panel.

Supplementary Material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fpls.2021.627432/full#supplementary-material

Supplementary Figure 1 | Steady-state PPFD response curves for A, g_s, and iWUE in all accessions.

Supplementary Figure 2 | Timecourses of steady-state A-PPFD response curves in all accessions.

Supplementary Figure 3 | Timecourses of steady-state g_s-PPFD response curves in all accessions.

Supplementary Figure 4 | Timecourses of steady-state iWUE-PPFD response curves in all accessions.

Supplementary Figure 5 | Timecourses of fluctuating PPFD response curves for A in all accessions.

Supplementary Figure 6 | Timecourses of fluctuating PPFD response curves for g_s in all accessions.

Supplementary Figure 7 | Timecourses of fluctuating PPFD response curves for iWUE in all accessions.

Supplementary Figure 8 | Pearson’s correlation coefficients, pairwise correlation scatterplots, and density plots for traits potentially underlying variation in non-steady-state iWUE following decreases in PPFD.

Supplementary Figure 9 | Test of the effect of leaf excision on steady-state and non-steady-state leaf gas-exchange.

Supplementary Table 1 | Variation in environmental controls during gas-exchange measurements.

References

Acevedo-Siaca, L. G., Coe, R., Wang, Y., Kromdijk, J., Quick, W. P., and Long, S. P. (2020). Variation in photosynthetic induction between rice accessions and its potential for improving productivity. New Phytol. 227, 1097–1108. doi: 10.1111/nph.16454

CrossRef Full Text | Google Scholar

Bellasio, C., Quirk, J., Buckley, T. N., and Beerling, D. J. (2017). A dynamic hydro-mechanical and biochemical model of stomatal conductance for C₄ photosynthesis. Plant Physiol. 175, 104–119. doi: 10.1104/pp.17.00666

CrossRef Full Text | Google Scholar

Bonsch, M., Humpenoder, F., Popp, A., Bodirsky, B., Dietrich, J. P., Rolinski, S., et al. (2016). Trade-offs between land and water requirements for large-scale bioenergy production. Glob. Change Biol. Bioener. 8, 11–24. doi: 10.1111/gcbb.12226

CrossRef Full Text | Google Scholar

Boyer, J. S. (1982). Plant productivity and environment. Science 218, 443–448. doi: 10.1126/science.218.4571.443

CrossRef Full Text | Google Scholar

Chen, J. W., Yang, Z. Q., Zhou, P., Hai, M. R., Tang, T. X., Liang, Y. L., et al. (2013). Biomass accumulation and partitioning, photosynthesis, and photosynthetic induction in field-grown maize (Zea mays L.) under low- and high-nitrogen conditions. Acta Physiol. Plant. 35, 95–105. doi: 10.1007/s11738-012-1051-6

CrossRef Full Text | Google Scholar

Dai, A. G. (2013). Increasing drought under global warming in observations and models. Nat. Clim. Change 3, 52–58. doi: 10.1038/nclimate1633

CrossRef Full Text | Google Scholar

Dalin, C., Wada, Y., Kastner, T., and Puma, M. J. (2017). Groundwater depletion embedded in international food trade. Nature 543, 700–704. doi: 10.1038/nature21403

CrossRef Full Text | Google Scholar

De Souza, A. P., Wang, Y., Orr, D. J., Carmo-Silva, E., and Long, S. P. (2020). Photosynthesis across African cassava germplasm is limited by Rubisco and mesophyll conductance at steady state, but by stomatal conductance in fluctuating light. New Phytol. 225, 2498–2512. doi: 10.1111/nph.16142

CrossRef Full Text | Google Scholar

Deans, R. M., Brodribb, T. J., Busch, F. A., and Farquhar, G. D. (2019). Plant water-use strategy mediates stomatal effects on the light induction of photosynthesis. New Phytol. 222, 382–395. doi: 10.1111/nph.15572

CrossRef Full Text | Google Scholar

D’Odorico, P., Chiarelli, D. D., Rosa, L., Bini, A., Zilberman, D., and Rulli, M. C. (2020). The global value of water in agriculture. Proc. Natl. Acad. Sci. U.S.A. 117, 21985–21993.

Google Scholar

Doncaster, H. D., Adcock, M. D., and Leegood, R. C. (1989). Regulation of photosynthesis in leaves of C₄ plants following a transition from high to low light. Biochim. Biophys. Acta 973, 176–184. doi: 10.1016/s0005-2728(89)80419-0

CrossRef Full Text | Google Scholar

Drake, P. L., Froend, R. H., and Franks, P. J. (2013). Smaller, faster stomata: scaling of stomatal size, rate of response, and stomatal conductance. J. Exp. Bot. 64, 495–505. doi: 10.1093/jxb/ers347

CrossRef Full Text | Google Scholar

Driever, S. M., Lawson, T., Andralojc, P. J., Raines, C. A., and Parry, M. A. J. (2014). Natural variation in photosynthetic capacity, growth, and yield in 64 field-grown wheat genotypes. J. Exp. Bot. 65, 4959–4973. doi: 10.1093/jxb/eru253

CrossRef Full Text | Google Scholar

FAO, IFAD, UNICEF, WFP, and WHO. (2018). The State of Food Security and Nutrition in the World 2018. Building Climate Resilience for Food Security and Nutrition. Rome: FAO, 202.

Google Scholar

Flexas, J. (2016). Genetic improvement of leaf photosynthesis and intrinsic water use efficiency in C₃ plants: why so much little success? Plant Sci. 251, 155–161. doi: 10.1016/j.plantsci.2016.05.002

CrossRef Full Text | Google Scholar

Galmes, J., Medrano, H., and Flexas, J. (2007). Photosynthetic limitations in response to water stress and recovery in Mediterranean plants with different growth forms. New Phytol. 175, 81–93. doi: 10.1111/j.1469-8137.2007.02087.x

CrossRef Full Text | Google Scholar

Geetika, G., van Oosterom, E. J., George-Jaeggli, B., Mortlock, M. Y., Deifel, K. S., McLean, G., et al. (2019). Genotypic variation in whole-plant transpiration efficiency in sorghum only partly aligns with variation in stomatal conductance. Funct. Plant Biol. 46, 1072–1089. doi: 10.1071/fp18177

CrossRef Full Text | Google Scholar

Giuliani, R., Koteyeva, N., Voznesenskaya, E., Evans, M. A., Cousins, A. B., and Edwards, G. E. (2013). Coordination of leaf photosynthesis, transpiration, and structural traits in rice and wild relatives (Genus Oryza). Plant Physiol. 162, 1632–1651. doi: 10.1104/pp.113.217497

CrossRef Full Text | Google Scholar

Głowacka, K., Kromdijk, J., Kucera, K., Xie, J., Cavanagh, A. P., Leonelli, L., et al. (2018). Photosystem II Subunit S overexpression increases the efficiency of water use in a field-grown crop. Nat. Commun. 9:868.

Google Scholar

Grantz, D. A., and Assmann, S. M. (1991). Stomatal response to blue light: water-use efficiency in sugarcane and soybean. Plant Cell Environ. 14, 683–690. doi: 10.1111/j.1365-3040.1991.tb01541.x

CrossRef Full Text | Google Scholar

Gu, J. F., Zhou, Z. X., Li, Z. K., Chen, Y., Wang, Z. Q., and Zhang, H. (2017). Rice (Oryza sativa L.) with reduced chlorophyll content exhibit higher photosynthetic rate and efficiency, improved canopy light distribution, and greater yields than normally pigmented plants. Field Crops Res. 200, 58–70. doi: 10.1016/j.fcr.2016.10.008

CrossRef Full Text | Google Scholar

Hadebe, S. T., Modi, A. T., and Mabhaudhi, T. (2017). Drought tolerance and water use of cereal crops: a focus on sorghum as a food security crop in sub-Saharan Africa. J. Agron. Crop Sci. 203, 177–191. doi: 10.1111/jac.12191

CrossRef Full Text | Google Scholar

Hebbali, A. (2020). olsrr: Tools for Building OLS Regression Models.

Google Scholar

Hetherington, A. M., and Woodward, F. I. (2003). The role of stomata in sensing and driving environmental change. Nature 424, 901–908. doi: 10.1038/nature01843

CrossRef Full Text | Google Scholar

Hubbart, S., Smillie, I. R. A., Heatley, M., Swarup, R., Foo, C. C., Zhao, L., et al. (2018). Enhanced thylakoid photoprotection can increase yield and canopy radiation use efficiency in rice. Commun. Biol. 1:22.

Google Scholar

Jahan, E., Amthor, J. S., Farquhar, G. D., Trethowan, R., and Barbour, M. M. (2014). Variation in mesophyll conductance among Australian wheat genotypes. Funct. Plant Biol. 41, 568–580. doi: 10.1071/fp13254

CrossRef Full Text | Google Scholar

Kaiser, E., Morales, A., and Harbinson, J. (2018). Fluctuating light takes crop photosynthesis on a rollercoaster ride. Plant Physiol. 176, 977–989. doi: 10.1104/pp.17.01250

CrossRef Full Text | Google Scholar

Kaiser, E., Morales, A., Harbinson, J., Kromdijk, J., Heuvelink, E., and Marcelis, L. F. M. (2015). Dynamic photosynthesis in different environmental conditions. J. Exp. Bot. 66, 2415–2426. doi: 10.1093/jxb/eru406

CrossRef Full Text | Google Scholar

Kaiser, H., and Kappen, L. (2001). Stomatal oscillations at small apertures: indications for a fundamental insufficiency of stomatal feedback-control inherent in the stomatal turgor mechanism. J. Exp. Bot. 52, 1303–1313. doi: 10.1093/jexbot/52.359.1303

CrossRef Full Text | Google Scholar

Kapanigowda, M. H., Payne, W. A., Rooney, W. L., Mullet, J. E., and Balota, M. (2014). Quantitative trait locus mapping of the transpiration ratio related to preflowering drought tolerance in sorghum (Sorghum bicolor). Funct. Plant Biol. 41, 1049–1065. doi: 10.1071/fp13363

CrossRef Full Text | Google Scholar

Knapp, A. K. (1993). Gas-exchange dynamics in C₃ and C₄ grasses - consequences of differences in stomatal conductance. Ecology 74, 113–123. doi: 10.2307/1939506

CrossRef Full Text | Google Scholar

Kromdijk, J., Glowacka, K., Leonelli, L., Gabilly, S. T., Iwai, M., Niyogi, K. K., et al. (2016). Improving photosynthesis and crop productivity by accelerating recovery from photoprotection. Science 354, 857–861. doi: 10.1126/science.aai8878

CrossRef Full Text | Google Scholar

Kromdijk, J., Griffiths, H., and Schepers, H. E. (2010). Can the progressive increase of C₄ bundle sheath leakiness at low PFD be explained by incomplete suppression of photorespiration? Plant Cell Environ. 33, 1935–1948. doi: 10.1111/j.1365-3040.2010.02196.x

CrossRef Full Text | Google Scholar

Lawson, T., and Blatt, M. R. (2014). Stomatal size, speed, and responsiveness impact on photosynthesis and water use efficiency. Plant Physiol. 164, 1556–1570. doi: 10.1104/pp.114.237107

CrossRef Full Text | Google Scholar

Lawson, T., and Matthews, J. (2020). Guard cell metabolism and stomatal function. Ann. Rev. Plant Biol. 71, 273–302. doi: 10.1146/annurev-arplant-050718-100251

CrossRef Full Text | Google Scholar

Lawson, T., and Vialet-Chabrand, S. (2019). Speedy stomata, photosynthesis and plant water use efficiency. New Phytol. 221, 93–98. doi: 10.1111/nph.15330

CrossRef Full Text | Google Scholar

Lawson, T., von Caemmerer, S., and Baroli, I. (2011). “Photosynthesis and stomatal behaviour,” in Progress in Botany 72, Vol. 72, eds U. Luttge, W. Beyschlag, B. Budel, and D. Francis (Berlin: Springer-Verlag), 265–304. doi: 10.1007/978-3-642-13145-5_11

CrossRef Full Text | Google Scholar

Leakey, A. D. B., Ferguson, J. N., Pignon, C. P., Wu, A., Jin, Z., Hammer, G. L., et al. (2019). Water use efficiency as a constraint and target for improving the resilience and productivity of C₃ and C₄ crops. Ann. Rev. Plant Biol. 70, 781–808. doi: 10.1146/annurev-arplant-042817-040305

CrossRef Full Text | Google Scholar

Leakey, A. D. B., Press, M. C., and Scholes, J. D. (2003). High-temperature inhibition of photosynthesis is greater under sunflecks than uniform irradiance in a tropical rain forest tree seedling. Plant Cell Environ. 26, 1681–1690. doi: 10.1046/j.1365-3040.2003.01086.x

CrossRef Full Text | Google Scholar

Leakey, A. D. B., Press, M. C., Scholes, J. D., and Watling, J. R. (2002). Relative enhancement of photosynthesis and growth at elevated CO₂ is greater under sunflecks than uniform irradiance in a tropical rain forest tree seedling. Plant Cell Environ. 25, 1701–1714. doi: 10.1046/j.1365-3040.2002.00944.x

CrossRef Full Text | Google Scholar

Lobell, D. B., Roberts, M. J., Schlenker, W., Braun, N., Little, B. B., Rejesus, R. M., et al. (2014). Greater sensitivity to drought accompanies maize yield increase in the US midwest. Science 344, 516–519. doi: 10.1126/science.1251423

CrossRef Full Text | Google Scholar

Matthews, J. S. A., Vialet-Chabrand, S., and Lawson, T. (2020). Role of blue and red light in stomatal dynamic behaviour. J. Exp. Bot. 71, 2253–2269. doi: 10.1093/jxb/erz563

CrossRef Full Text | Google Scholar

McAusland, L., Vialet-Chabrand, S., Davey, P., Baker, N. R., Brendel, O., and Lawson, T. (2016). Effects of kinetics of light-induced stomatal responses on photosynthesis and water-use efficiency. New Phytol. 211, 1209–1220. doi: 10.1111/nph.14000

CrossRef Full Text | Google Scholar

Ort, D. R., and Long, S. P. (2014). Limits on Yields in the Corn Belt. Science 344, 483–484.

Google Scholar

Osborne, C. P., and Sack, L. (2012). Evolution of C₄ plants: a new hypothesis for an interaction of CO₂ and water relations mediated by plant hydraulics. Philos. Transact. R. Soc. B Biol. Sci. 367, 583–600. doi: 10.1098/rstb.2011.0261

CrossRef Full Text | Google Scholar

Papanatsiou, M., Petersen, J., Henderson, L., Wang, Y., Christie, J. M., and Blatt, M. R. (2019). Optogenetic manipulation of stomatal kinetics improves carbon assimilation, water use, and growth. Science 363, 1456–1459. doi: 10.1126/science.aaw0046

CrossRef Full Text | Google Scholar

Pearcy, R. W. (1990). Sunflecks and photosynthesis in plant canopies. Ann. Rev. Plant Physiol. Plant Mol. Biol. 41, 421–453. doi: 10.1146/annurev.pp.41.060190.002225

CrossRef Full Text | Google Scholar

Pignon, C. P. (2017). Strategies to Improve C₄ Photosynthesis, Water and Resource-Use Efficiency Under Different Atmospheres, Temperatures, and Light Environments. Urbana, IL: University of Illinois at Urbana-Champaign.

Google Scholar

Pignon, C. P., and Long, S. P. (2020). Retrospective analysis of biochemical limitations to photosynthesis in 49 species: C₄ crops appear still adapted to pre-industrial atmospheric [CO₂]. Plant Cell Environ. 43, 2606–2622. doi: 10.1111/pce.13863

CrossRef Full Text | Google Scholar

R Core Team (2017). R: A Language and Environment for Statistical Computing. Vienna: R Foundation for Statistical Computing.

Google Scholar

Raven, J. A. (2014). Speedy small stomata? J. Exp. Bot. 65, 1415–1424. doi: 10.1093/jxb/eru032

CrossRef Full Text | Google Scholar

Regassa, T. H., and Wortmann, C. S. (2014). Sweet sorghum as a bioenergy crop: literature review. Biomass Bioenergy 64, 348–355. doi: 10.1016/j.biombioe.2014.03.052

CrossRef Full Text | Google Scholar

Signorell, A. (2020). DescTools: Tools for Descriptive Statistics.

Google Scholar

Sinclair, T. R., Tanner, C. B., and Bennett, J. M. (1984). Water-use efficiency in crop production. Bioscience 34, 36–40.

Google Scholar

Slattery, R. A., Walker, B. J., Weber, A. P. M., and Ort, D. R. (2018). The Impacts of fluctuating light on crop performance. Plant Physiol. 176, 990–1003. doi: 10.1104/pp.17.01234

CrossRef Full Text | Google Scholar

Soleh, M. A., Tanaka, Y., Kim, S. Y., Huber, S. C., Sakoda, K., and Shiraiwa, T. (2017). Identification of large variation in the photosynthetic induction response among 37 soybean Glycine max (L.) Merr. genotypes that is not correlated with steady-state photosynthetic capacity. Photosynth. Res. 131, 305–315. doi: 10.1007/s11120-016-0323-1

CrossRef Full Text | Google Scholar

Soleh, M. A., Tanaka, Y., Nomoto, Y., Iwahashi, Y., Nakashima, K., Fukuda, Y., et al. (2016). Factors underlying genotypic differences in the induction of photosynthesis in soybean Glycine max (L.) Merr. Plant Cell Environ. 39, 685–693. doi: 10.1111/pce.12674

CrossRef Full Text | Google Scholar

Sollenberger, L. E., Woodard, K. R., Vendramini, J. M. B., Erickson, J. E., Langeland, K. A., Mullenix, M. K., et al. (2014). Invasive populations of elephantgrass differ in morphological and growth characteristics from clones selected for biomass production. Bioenergy Res. 7, 1382–1391. doi: 10.1007/s12155-014-9478-9

CrossRef Full Text | Google Scholar

Spinoni, J., Vogt, J. V., Naumann, G., Barbosa, P., and Dosio, A. (2018). Will drought events become more frequent and severe in Europe? Int. J. Climatol. 38, 1718–1736. doi: 10.1002/joc.5291

CrossRef Full Text | Google Scholar

Stitt, M., and Zhu, X. G. (2014). The large pools of metabolites involved in intercellular metabolite shuttles in C₄ photosynthesis provide enormous flexibility and robustness in a fluctuating light environment. Plant Cell Environ. 37, 1985–1988. doi: 10.1111/pce.12290

CrossRef Full Text | Google Scholar

Sun, H. M., Wang, A. Q., Zhai, J. Q., Huang, J. L., Wang, Y. J., Wen, S. S., et al. (2018). Impacts of global warming of 1.5 degrees C and 2.0 degrees C on precipitation patterns in China by regional climate model (COSMO-CLM). Atmosph. Res. 203, 83–94. doi: 10.1016/j.atmosres.2017.10.024

CrossRef Full Text | Google Scholar

Sun, J. L., Zhang, Q. Q., Tabassum, M. A., Ye, M., Peng, S. B., and Li, Y. (2017). The inhibition of photosynthesis under water deficit conditions is more severe in flecked than uniform irradiance in rice (Oryza sativa) plants. Funct. Plant Biol. 44, 464–472. doi: 10.1071/fp16383

CrossRef Full Text | Google Scholar

Taylor, S. H., Franks, P. J., Hulme, S. P., Spriggs, E., Christin, P. A., Edwards, E. J., et al. (2012). Photosynthetic pathway and ecological adaptation explain stomatal trait diversity amongst grasses. New Phytol. 193, 387–396. doi: 10.1111/j.1469-8137.2011.03935.x

CrossRef Full Text | Google Scholar

Taylor, S. H., and Long, S. P. (2017). Slow induction of photosynthesis on shade to sun transitions in wheat may cost at least 21% of productivity. Philos. Transact. R. Soc. B Biol. Sci. 372:9.

Google Scholar

Tomeo, N. J., and Rosenthal, D. M. (2017). Variable mesophyll conductance among soybean cultivars sets a tradeoff between photosynthesis and water-use-efficiency. Plant Physiol. 174, 241–257. doi: 10.1104/pp.16.01940

CrossRef Full Text | Google Scholar

Tomimatsu, H., and Tang, Y. H. (2016). Effects of high CO₂ levels on dynamic photosynthesis: carbon gain, mechanisms, and environmental interactions. J. Plant Res. 129, 365–377. doi: 10.1007/s10265-016-0817-0

CrossRef Full Text | Google Scholar

Vialet-Chabrand, S. R. M., Matthews, J. S. A., McAusland, L., Blatt, M. R., Griffiths, H., and Lawson, T. (2017). Temporal dynamics of stomatal behavior: modeling and implications for photosynthesis and water use. Plant Physiol. 174, 603–613. doi: 10.1104/pp.17.00125

CrossRef Full Text | Google Scholar

Vico, G., Manzoni, S., Palmroth, S., and Katul, G. (2011). Effects of stomatal delays on the economics of leaf gas exchange under intermittent light regimes. New Phytol. 192, 640–652. doi: 10.1111/j.1469-8137.2011.03847.x

CrossRef Full Text | Google Scholar

Viswanathan, R., Chinnaraja, C., Malathi, P., Gomathi, R., Rakkiyappan, P., Neelamathi, D., et al. (2014). Impact of Sugarcane yellow leaf virus (ScYLV) infection on physiological efficiency and growth parameters of sugarcane under tropical climatic conditions in India. Acta Physiol. Plant. 36, 1805–1822. doi: 10.1007/s11738-014-1554-4

CrossRef Full Text | Google Scholar

von Caemmerer, S. (2000). “Modelling C₄ photosynthesis,” in Biochemical Models of Leaf Photosynthesis, Vol. 2, ed. S. Von Caemmerer (Collingwood,VIC: CSIRO Publishing), 91–122.

Google Scholar

von Caemmerer, S., and Farquhar, G. D. (1981). Some relationships between the biochemistry of photosynthesis and the gas exchange of leaves. Planta 153, 376–387. doi: 10.1007/bf00384257

CrossRef Full Text | Google Scholar

Wang, Y., Burgess, S. J., de Becker, E. M., and Long, S. H. P. (2020). Photosynthesis in the fleeting shadows: an overlooked opportunity for increasing crop productivity? Plant J. 101, 874–884. doi: 10.1111/tpj.14663

CrossRef Full Text | Google Scholar

Wang, Y., Noguchi, K., Ono, N., Inoue, S., Terashima, I., and Kinoshita, T. (2014). Overexpression of plasma membrane H⁺-ATPase in guard cells promotes light-induced stomatal opening and enhances plant growth. Proc. Natl. Acad. Sci. U.S.A. 111, 533–538. doi: 10.1073/pnas.1305438111

CrossRef Full Text | Google Scholar

Way, D. A., and Pearcy, R. W. (2012). Sunflecks in trees and forests: from photosynthetic physiology to global change biology. Tree Physiol. 32, 1066–1081. doi: 10.1093/treephys/tps064

CrossRef Full Text | Google Scholar

Wei, T., and Simko, V. (2017). R package “corrplot”: Visualization of a Correlation Matrix.

Google Scholar

WWAP (2015). The United Nations World water Development Report 2015: Water for a Sustainable World. Paris: UNESCO.

Google Scholar

Xin, Z. G., Aiken, R., and Burke, J. (2009). Genetic diversity of transpiration efficiency in sorghum. Field Crops Res. 111, 74–80. doi: 10.1016/j.fcr.2008.10.010

CrossRef Full Text | Google Scholar

Yabiku, T., and Ueno, O. (2017). Variations in physiological, biochemical, and structural traits of photosynthesis and resource use efficiency in maize and teosintes (NADP-ME-type C₄). Plant Product. Sci. 20, 448–458. doi: 10.1080/1343943x.2017.1398050

CrossRef Full Text | Google Scholar

Zhu, X. G., Ort, D. R., Whitmarsh, J., and Long, S. P. (2004). The slow reversibility of photosystem II thermal energy dissipation on transfer from high to low light may cause large losses in carbon gain by crop canopies: a theoretical analysis. J. Exp. Bot. 55, 1167–1175. doi: 10.1093/jxb/erh141

CrossRef Full Text | Google Scholar