Contrasting response of biomass and grain yield to severe drought in Cappelle Desprez and Plainsman V wheat cultivars

Plant material and growth conditions

Vernalization of one-week old seedlings was carried out for 6-week, at 4 °C in a cold chamber, under continuous dim light. Vernalized plantlets of the Cappelle Desprez and Plainsman V (Guóth et al., 2009) (http://genbank.vurv.cz/wheat/pedigree) winter wheat (Triticum aestivum L.) varieties were planted in a soil-sand-peat mixture (3:1:1, v/v/v). Plants were regularly irrigated and grown in controlled green-house conditions for two weeks before starting the drought stress treatment. Photosynthetically active radiation (PAR) within controlled environment was maintained with a 14 h photoperiod at a PPFD of 400–500 µmol m⁻² s⁻¹, 22–25 °C and ca. 45–55% relative humidity.

Drought stress treatment and measuring protocol

Drought stress was induced on the above mentioned seedlings (4–5 leaf stage) by limiting irrigation to ensure 10% field capacity of the soil using the computer-controlled water supply system of our phenotyping platform (Cseri et al., 2013) for a period of 35 days. The control well watered plants were irrigated to keep 60% field capacity of the soil. Biomass accumulation in the vegetative phase, i.e., in the first 3 weeks of the drought treatment, was monitored from the younger fully developed leaves, which are denoted as ‘secondary leaves.’ While in the reproductive grain filling phase, in the 4th–5th week of the drought treatment, the measurements were performed on the last fully developed leaf, denoted as ‘flag leaf.’ Six replicates of each treatment were used for the study in three separate experimental trials conducted in August–September 2012, April–May and July–August 2013, respectively.

Gas exchange measurements

Gas exchange parameters: CO₂ uptake rate, transpiration, stomatal conductance and intercellular CO₂ concentration were measured by using a Licor 6400 gas analyzer (LI-COR Biosciences, Lincoln, NE, USA). Two to three selected pieces of attached secondary, as well as flag leafs from plant replicates treated under respective drought regimes were inserted into the gas cuvette for individual measurements. The gas cuvette conditions were set to 400 ppm CO₂, ambient temperature and growth light intensity of photosynthetically active radiation (of 400 µmol m⁻² s⁻¹).

Water-use efficiency (WUE) was calculated from the ratio of photosynthesis (A) and transpiration (E) (Farquhar & Richards, 1984; Lovelli et al., 2007).

OJIP chlorophyll a fluorescence measurements

OJIP chlorophyll a fluorescence transients were measured by a Plant Efficiency Analyzer (Pocket Pea, Hansatech, UK). The transients were induced by red light from an LED source (627 nm, up to 3,500 µmol m⁻² s⁻¹ intensity). Prior to measurements, performed on the adaxial surface, leaves were dark adapted to for 20 min using light tight leaf-clips. The OJIP-test (Strasser, Srivastava & Tsimilli-Michael, 2000) was used to analyze the chlorophyll a fluorescence transients and the following original data were acquired: O (F_o) initial fluorescence level (measured at 50 µs), P (F_m) maximal fluorescence intensity, as well as the J (at about 2 ms) and the I (at about 30 ms) intermediate fluorescence levels. From these specific fluorescence features the following parameters of photosynthetic efficiency were calculated: maximal PSII quantum yield, F_v/F_m; The ratio of variable fluorescence to initial fluorescence, F_v/F_o where F_v = F_m − F_o; Probability of electron transport out of Q_A, (1-V_j)/V_j where V_j = (F_2ms − F_o)/F_v; Total complementary area between the fluorescence induction curve and F_m of the OJIP curve, Area; Q_A reducing reaction centers per PSII antenna chlorophyll, RC/ABS = (F_V∕F_M) ⋅ (F_J − F₀)/[4 ⋅ (F_300μs − F₀)] (Zurek et al., 2014; Campos et al., 2014; Strasser, Tsimilli-Michael & Srivastava, 2004); Performance index (potential) for energy conservation from photons absorbed by PSII to the reduction of intersystem electron acceptors, PI_Abs (Zivcak et al., 2008; Campos et al., 2014). ${\displaystyle {\mathrm{PI}}_{\mathrm{Abs}}=\frac{1-\left(F_o∕F_m\right)}{M_o∕V_j}\times \frac{F_m-F_o}{F_o}\times \frac{1-V_j}{V_j}}$ where M_o = 4∗ (F_300μs −F_o)/ (F_M − F_o) represents initial slope of fluorescence kinetics.

For screening of drought stress tolerance a further parameter, the so-called drought factor index (DFI) was used, which is derived from PI values measured after 1 or 2 weeks of drought treatment, and reflects the ability of plants to tolerate sustained drought stress conditions (Oukarroum et al., 2007) ${\displaystyle \left(\mathrm{DFI}\right)=log\left({\mathrm{PI}}_{\mathrm{week}1}∕{\mathrm{PI}}_{\mathrm{control}}\right)+2log\left({\mathrm{PI}}_{\mathrm{week}2}∕{\mathrm{PI}}_{\mathrm{control}}\right).}$

Simultaneous measurements of P700 redox state and chlorophyll fluorescence

Variable chlorophyll fluorescence from PSII and the amount of oxidized PSI primary Chl electron donor (P700⁺) was simultaneously measured using a DUAL-PAM-100 system (WALZ, Effeltrich, Germany). From the fluorescence data F_v/F_m and the effective quantum yield of photochemical energy conversion in PS II, Y(II)= (F_m′ − F)/ F_m′ (Genty, Briantais & Baker, 1989) where F_o, F_o′ are dark fluorescence yield from dark- and light-adapted leaf, respectively and F_m, F_m′ are maximal fluorescence yield from dark- and light-adapted leaf, respectively, were calculated. The P700⁺ signal (P) may vary between a minimal (P700 fully reduced) and a maximal level (P700 fully oxidized). The maximum level of P700⁺ is called P_m in analogy with F_m. It was determined by application of a saturation pulse (300 ms, 10,000 µmol m⁻² s⁻¹; 635 nm) after pre-illumination with far-red light. P_m′ is analogous to the fluorescence parameter F_m′ and was determined by applying 800 ms saturating pulse of 635 nm red light. The photochemical quantum yield of PSI, Y(I) is the quantum yield of photochemical energy conversion in PSI. It is calculated as Y(I) = (P_m′ − P)/ P_m. Y(ND) is the quantum yield of non-photochemical energy dissipation in PSI due to donor side limitation, Y(ND) =P∕P_m. Y(NA) is the quantum yield of non-photochemical energy dissipation due to acceptor side limitation in PSI, Y(NA) = (P_m − P_m′)/P_m, and Y(I) + Y(ND) + Y(NA) = 1 (Klughammer & Schreiber, 1994; Singh et al., 2014). Non-photochemical quenching NPQ (Bilger & Bjorkman, 1990), was calculated as (F_m − F_ms)/F_ms, where F_m represents the fluorescence of a dark-adapted sample and F_ms represents a fluorescence of the illuminated sample. Plants were dark-adapted for ∼20 min and kinetics were measured after repeated light pulses of 94 PPFD for 300 s. Leaves were subsequently relaxed in darkness for 240 s and fluorescence while continuously measuring and recording fluorescence (Szalonek et al., 2015).

The electron transport rates through PSII as well as through PSI were determined simultaneously (Miyake et al., 2005; Fan, 2007). The apparent rate of electron transport in higher plants were calculated as ETR(II) = Y(II) ∗ PPFD ∗ 0.5 ∗ 0.84 and ETR(I) = Y(I) ∗ PPFD ∗ 0.5 ∗ 0.84 (Genty, Briantais & Baker, 1989), where Y(II) and Y(I) are effective quantum yields of PSII and PSI respectively, PPFD is the photon flux density of incident photosynthetically active radiation and two coefficients (0.5 and 0.84 for higher plants, which imply that PSII and PSI are equally excited, and that due to leaf absorbance properties only 84% of incident irradiance will be absorbed by the photosystems, respectively (Björkman & Demmig, 1987; Schreiber, 2004).

Thermal imaging

Thermal images were taken by using a Thermo Varioscan (Jenoptik, Laser optik, Systeme, GmbH) camera as described by Kana & Vass (2008). Thermal images of wheat cultivars under various drought stress treatments were analyzed by using ImageJ software to select and measure areas based on color. Images were threshold using Hue, Saturation and Brightness (HSB) color space and converted to binary values by defining a color scale cutoff point. Values of evaporative cooled area, represented by pixels below the ambient temperature, become black and those in above become white.

Digital imaging

Digital images of seedlings were taken by using a Nikon D80 camera equipped with an AF-S DX Zoom-NIKKOR 18–135 mm objective (f/3.5–5.6G ED-IF Lens) and close-up rings. Digital images of plant replicates under various drought stress treatments were analyzed for green biomass area using ImageJ software. We used colour thresholding to select just the plant green area and exclude the stand, pot, shadows and yellowish leaves (http://rsbweb.nih.gov/ij/).

Statistical analysis

The comparison of traits of plants of the same variety, which were grown under different watering conditions was based on the two-sample Student’s t-Tests (http://www.physics.csbsju.edu/stats/t-test.html). Levels of significance (P values) in differences from means of well watered and drought stressed plants are indicated in the figure legends.

Phenotyping for biomass accumulation and grain yield

Growth of wheat plants was monitored by digital photography by recording green pixel-based shoot surface area of wheat plants, which was performed during the whole growth period once a week. According to our previous data the green pixel-based shoot surface area correlates with green biomass (Fehér-Juhász et al., 2014). Digital RGB imaging of leaf/shoot area showed that the Cappelle Desprez cv. produces larger above ground green biomass than Plainsman V not only under conditions of water availability but also under water scarcity (Fig. 1). This conclusion is supported by direct measurement of green biomass (Fig. S1).

The grain yield data showed an opposite trend compared to the biomass accumulation, i.e., although grain yield under well watered conditions was higher in the Cappelle Desprez cv., the Plainsman V produced more grains under water limitation, and showed lower grain yield loss (24%) than the Cappelle Desprez (74%) (Table 1), see also Fig. S2.

Carbon fixation, stomatal functions and water-use efficiency

From gas exchange measurements, we could observe that the net CO₂ uptake rate and other gas exchange parameters were not affected in the first week of drought stress. After the second week the net CO₂ uptake decreased in the secondary leaves of water limited plants both in the case of drought sensitive Cappelle Desprez and drought tolerant Plainsman V cv. (Fig. 2A). Interestingly, in the grain filling period only Cappelle Desprez cv. responded with decreased CO₂ uptake to the decreased soil water content in case of the flag leaves (Fig. 2A).

The response of stomatal conductance (Fig. 2B) and evaporation rate (Fig. 2C) shows similar pattern to that of CO₂ uptake. Interestingly, in case of secondary leaves both parameters decreased under drought stress in both cultivars, while in the flag leaves only the Cappelle Desprez cv. showed significant decrease of stomatal conductance and evaporation rate under drought conditions relative to their well watered controls. The Plainsman cv. did not close its stomata in the flag leaves and did not decrease its evaporation rate during early grain filling phase (21st day). However, in the later phases (29th day) of the drought stress the stomatal conductance and evaporation showed a decreasing tendency also in the Plainsman V cv. (Figs. 2B and 2C).

The ability of leaves to achieve optimal photosynthesis relative to the amount of used water, i.e., to conserve water under drought conditions, is characterized by the water-use efficiency (WUE), which is given by the ratio of the rates of net photosynthesis and transpiration. In the Cappelle Desprez cv. the WUE (A/E) parameter increased both in the secondary and flag leaves, while in the Plainsman cv. WUE increased only slightly in the secondary leaves and also in the flag leaves during the grain filling phase in the drought stressed plants (Fig. 2D), which is consistent with sustained photosynthesis and flag leaf transpiration in order to maintain high grain yield.

Thermal imaging

Leaf temperature can be conveniently monitored by thermal imaging in the infrared spectral range, and drought stress induced leaf temperature changes can be studied. The status of stomata influences not only CO₂ uptake, but also the efficiency of evaporating water from the leaf tissue, which in turn affects the temperature of the leaves. Thermal images were quantified based on pixel numbers calculated from thresholded evaporative cooled area relative to ambient background temperature (Fig. 3A). The data show that the leaf and shoot area, which is cooler than the surrounding air, i.e., cooled by evaporation via transpiration, is small in drought stressed Cappelle Desprez plants indicating low transpiration rate due to stomatal closure (Fig. 3A). In contrast, the Plainsman V cv. has larger cooled leaf area not only in the well watered, but also in the drought stressed plants (Fig. 3A). Evaporative cooling is especially pronounced in the spikes of Plainsman V plants (Figs. 3B–3D), which is in agreement with the open status of the stomata allowing efficient CO₂ uptake and large net photosynthesis rate for a sustained period during grain filling leading to increased grain yield.

Simultaneously recorded fast Chl fluorescence and P700 redox kinetics

We observed drought induced changes in photosynthetic electron transport rate from the analysis of fast chlorophyll and P700 kinetics measured in dark adapted leaves (Fig. 4). The effects of drought on the initial processes of photosynthesis which take place in the Photosystem II complex were characterized by measuring variable chlorophyll fluorescence induction transients, the so called OJIP curves. The measurements were performed both on the secondary leaves, which are expected to reflect photosynthetic activity that is responsible for overall green biomass growth, as well as on flag leaves, which primarily support grain development. The variable fluorescence transients show clear differences in the I-P amplitude, reflects the size of the electron acceptor pool of PSI (Schansker, Tóth & Strasser, 2005; Tsimilli-Michael & Strasser, 2008), and was correlated with a higher PSI activity due to an increased PSI/PSII ratio (Ceppi et al., 2012). In case of secondary leaves, the Cappelle Desprez cv. responded to water withdrawal with only minor changes of the OJIP fluorescence and P700 kinetic transients in the first 2 weeks of the drought treatment (Fig. 4A). After 3 weeks, however, faster rise in Chl fluorescence and P700 oxidation and re-reduction transient in normalized curve was observed in drought stressed Cappelle cv. showing reduction of ferredoxin and increased PSI content (ca. 20%, Fig. 4B) (Schansker, Tóth & Strasser, 2005; Ceppi et al., 2012). In drought stressed Plainsman V we could observe a faster rise of the J-I phase, PSI oxidation was prolonged and PSI re-reduction did not reach normal level. Faster decay of drought stressed Plainsman V shows less functional PSII activity and smaller PSI content. Well-watered controls of both cultivars showed comparable responses in Chl fluorescence and P700 redox kinetics (Fig. 4B).

During the early phase of spiking (21st and 28th day of drought), 10% W Cappelle Desprez cv. showed similar trend of higher PSI content in the flag leaves with an increase of I-P amplitude in fast chlorophyll fluorescence and of P700⁺/P700 ratio kinetics (Figs. 4C and 4D). In drought-stressed Plainsman V we could observe a slower rise of J-I phase and I-P amplitude with increased P700 oxidation ending in faster re-reduction decay kinetics (Figs. 4C and 4D). This indicates higher functional PSII activity for drought stressed Plainsman V in the flag leaves during grain filling period.

Chl fluorescence parameters reflect plant responses to drought stress

Various biophysical parameters derived from Chl a fluorescence transient measurements help to understand the energy flow through PSII and provide useful indicators of the development and severity of stress effects (Kenny et al., 2011), including drought. One of the useful calculated parameters is the so called performance index (PI), which combines the three main functional steps taking place in PSII (light energy absorption, excitation energy trapping, and conversion of excitation energy to electron transport), and was used as measure of drought stress tolerance (Strauss et al., 2006). PI, which provides useful and quantitative information about the physiological conditions and the vitality of plants, revealed differences among varieties under conditions of drought stress (Oukarroum et al., 2007). Other calculated parameters like Area and RC/ABS also responded for secondary and flag leaves under severe drought stress (Campos et al., 2014).

In case of Cappelle Desprez cv., the secondary leaves responded to water withdrawal with only minor changes of the deduced electron transport parameters in the first 2 weeks of the drought treatment (Fig. 5A). In contrast, the Plainsman V showed a clear tendency for decreasing the PI, the Area reflecting the size of oxidized PQ pool, and the RC/ABS parameters indicating a decreased photosynthetic performance at the level of secondary leaves (Fig. 5A). As regards deduced fluorescence parameters, there was no significant effect of drought stress in either of the developmental stages of the secondary leaves of the Cappelle Desprez cv. with the exception of a small increase of PI after 2 weeks of drought stress (Fig. 5B). In contrast, the Plainsman V cv. showed a marked increase of the PI parameter with the progress of drought stress with some increase also in the Area, RC/ABS, and F_v∕F_m parameters after 4 weeks of drought stress (Fig. 5B). These data show that the photosynthetic efficiency of flag leaves of the Plainsman V cv. is affected by drought stress conditions to a smaller extent than of the Cappelle Desprez variety.

For the characterization of drought tolerance on the basis of Chl fluorescence data Strasser and coworkers have introduced the so called drought factor index (DFI). This parameter represents the relative decrease of the performance index (PI) during water scarcity and reflects the ability of plants to tolerate long-term drought stress (Oukarroum et al., 2007). A large positive value of DFI indicates drought tolerance, while a large negative value is characteristic for drought sensitivity. According to Table 2, in the periods of severe drought and senescence the secondary leaves of the Cappelle Desprez and Plainsman V cv. show positive and negative DFI, respectively, which is in agreement with the larger green leaf area and biomass of Cappelle Desprez as compared to that of Plainsman V Interestingly, the DFI values of the flag leaves show an opposite trend with positive DFI for the Plainsman and negative DFI for the Cappelle Desprez cv. Again, this behavior is consistent with the higher grain yield of Plainsman V as compared to the Cappelle Desprez cv. under drought stress.

Drought induced cyclic electron flow around PSI

A characteristic response of drought stressed wheat plants is the induction of cyclic electron flow (CEF) around Photosystem I (Johnson, 2011; Zivcak et al., 2013; Zivcak et al., 2014), which directs electrons from the acceptor side of PSI back to the PQ pool. This effect is considered as a defense mechanism against oxidative stress that develops under conditions of limited availability of CO₂ as final electron acceptor due to stomatal closure (Golding & Johnson, 2003; Zulfugarov, Mishra & Lee, 2010). Evidence for cyclic flow at high light exists in all the plants to a greater or lesser extent (Kono, Noguchi & Terashima, 2014).

Since under conditions of cyclic electron flow part of the electrons, which arise from PSII circulate around PSI the rate of electron flow though PSII (ETR(II)) and PSI (ETR(I)) are different. One way to assess the efficiency of cyclic electron flow is to follow the changes in the ETR(I)/ETR(II) ratio. Calculation of ETR(I) and ETR(II) requires the directly measured quantum yield of PSI (Y(I)) and PSII (Y(II)), light intensity and estimated constant factors (see Mat. Meth.). From this definition of ETR it follows that instead of the calculated ETR(I)/ETR(II) ratio the directly measured Y(I)/Y(II) can be used. As shown in Fig. 6, the Y(I)/Y(II) ratio does not show significant differences between the different wheat cultivars either in the control or in drought stress conditions in the biomass accumulation period at various grow light regimes (94–849 µmol photons m⁻² s⁻¹). Only in the well-watered Plainsman V. was the Y(I)/Y(II) ratio somewhat lower under medium and high light regimes (Figs. 6B and 6C). However, the Y(I)/Y(II) ratio increased in drought-stressed Cappelle Desprez cv. during grain filling period at medium (363 µmol photons m⁻² s⁻¹) and high (849 µmol photons m⁻² s⁻¹) light regimes, respectively (Figs. 6B and 6C).

The relationship between ETR (I) and ETR (II) can also be analyzed by looking at the linearity of ETR (II) as a function of ETR (I). When only linear electron flow dominates there is a linear relationship between ETR (II) and ETR (I). However, after the onset of cyclic flow ETR (I) increases faster than ETR (II) and the linear relationship breaks down at high light (Johnson, 2011). During the early days of drought (13th day) cyclic flow sets in under similar conditions, ETR (I) ≈ 60 µmol/m² s, in all treatments (Fig. 7A), but as the drought progresses they show divergent responses. For 21 days of drought there is a tendency for the maximum ETR(I) to increase while the maximum ETR(II) does not change much in the Cappelle cv. (Fig. 7B) implying that linear electron transport is not much affected by the drought treatment while cyclic electron flow around PSI is increasing in the secondary leaves. A Similar tendency is seen in the flag leaves after 22 and 28 days of drought stress (Figs. 7C and 7D). In contrast, drought stressed Plainsman showed a tendency of smaller cyclic flow in the later phases of drought stress (Figs. 7C and 7D). Thus we could observe that as the drought progresses cyclic electron flow is losing in Plainsman, while it is enhancing in Cappelle.

Light responses of the PSI electron transport parameters

Characterization of electron transport through Photosystem I is important for the identification of the sites where limitation of electron flow occurs under conditions of drought stress. Light responses of PSI parameters obtained from P700 signals were further analyzed (Fig. 8). The fraction of overall P700 that is kept in the oxidized state, Y(ND), was significantly increased with the increase of PPFD in secondary and flag leaves of drought stressed Plainsman V from the end of the second week of water withdrawal with respect to Cappelle cv in both 60% and 10% field capacity (Figs. 8B and 8C). Higher values of Y(ND) indicates that a major fraction of overall P700 is in the oxidized state during illumination due to limitation of electron flow from PSII towards PSI under severe drought condition. Y(NA) represents the fraction of overall P700 that cannot be oxidized by a saturation pulse in a given state due to a lack of oxidized PSI acceptors (Singh et al., 2014). A substantial increase of Y(NA) was observed in the flag leaves of drought stressed Cappelle Desprez cv. (Fig. 8C), which correlates with the lower CO₂ uptake rate of these leaves, that creates a limitation of reduced acceptors at the acceptor side of PSI at lower light intensities (Zivcak et al., 2013). Y(NA) of Cappelle Desprez cv. was greater than Y(ND) at PPFDs <250 µmole photons m⁻² s⁻¹, while, above this level, Y(NA) decreased and Y(ND) of drought-stressed Plainsman V increased. Higher effective quantum yield of PSI, Y(I) was observed for drought-stressed Cappelle Desprez cv. in both biomass accumulation as well as grain filling period in comparison to plants of all other treatment conditions (Figs. 8A–8D).

Concluding remarks

Non-invasive photosynthetic measurements provide highly useful tools for making reliable predictions of physiological traits of wheat and other plants. Our findings demonstrate that the agronomically highly important traits of biomass and grain yield are not necessarily correlated in wheat and possibly in other cereal crops. Therefore, phenotyping of biomass responses alone is not sufficient for predictions of grain yield changes. As a consequence, phenotyping protocols should include grain yield assessment when the aim is the optimization of grain yield and grain yield stability under stress conditions.

Our results support the importance of cyclic electron flow in drought stress tolerance. We can also conclude that changes in physiological parameters show different responses to drought stress depending on the developmental stage of leaves in the case of the studied two cultivars. Flag leaves, which serve as grain supporting leaves show similar response in their CO₂ fixation, drought factor index, and electron transport parameters as the grain yield, whereas the secondary leaves, which support overall green biomass growth show similar responses as biomass accumulation. These findings are warranted by the presented results for the Cappele Deprez and Plainsman V cultivars studied here. However, the general validity of the differential response of physiological parameters in leaves of different developmental stages has to be investigated by using several wheat cultivars in future studies.