Photosystem II energy use,non-photochemical quenching and the xanthophyll cycle in Sorghumbicolor grown under drought and free-air CO2 enrichment(FACE) conditions


Andrew N.Webber. Fax: +1 480-965-6899; e-mail:


The present study was carried out to test the hypothesis thatelevated atmospheric CO2 (Ca) will alleviate over-excitationof the C4 photosynthetic apparatus and decrease non-photochemicalquenching (NPQ) during periods of limited water availability. Chlorophyll a fluorescencewas monitored in Sorghum bicolor plants grown under a free-aircarbon-dioxide enrichment (FACE) by water-stress (Dry) experiment.Under Dry conditions elevated Ca increased the quantum yield ofphotosystem II (φPSII) throughout the day throughincreases in both photochemical quenching coefficient (qp)and the efficiency with which absorbed quanta are transferred toopen PSII reaction centres (Fv′/Fm′).However, in the well-watered plants (Wets) FACE enhanced φPSIIonly at midday and was entirely attributed to changes in Fv′/Fm. Underfield conditions, decreases in φPSII under Dry treatmentsand ambient Ca corresponded to increases in NPQ but the de-epoxidation stateof the xanthophyll pool (DPS) showed no effects. Water-stress didnot lead to long-term damage to the photosynthetic apparatus asindicated by φPSII and carbon assimilation measuredafter removal of stress conditions. We conclude that elevated Caenhances photochemical light energy usage in C4 photosynthesisduring drought and/or midday conditions. Additionally,NPQ protects against photo-inhibition and photodamage. However,NPQ and the xanthophyll cycle were affected differently by elevatedCa and water-stress.




analysis of variance


bundle sheath cells


atmosphericCO2 concentration






chlorophyll singlet excited state


days after planting


de-epoxidationstate of the xanthophyll pool (ZA/ZAV)


free-air CO2 enrichment




Fm, minimum and maximum dark-dark-adaptedfluorescence yield


Fm′, Fs,minimum, maximum and steady-state, light-adapted fluorescence yield


maximum quantum yield of PSII photochemistry


the efficiency with which absorbed quanta aretransferred to the open PSII reaction centres






phosphoenolpyruvate (PEP) carboxylase


photosyntheticallyactive photon flux density (µmol m−2 s−1)


measured rate of CO2 assimilation


net CO2 fixation (PS + RD)


photosystem II


photochemical quenching


dark respiration








leaf-to-air vapour pressure deficits




quantum yield of CO2 fixation


quantum yield of PSII


pH gradient across the thylakoid membrane.


Anticipated increases in atmospheric CO2 concentrations (Ca)and altered precipitation patterns will certainly effect carbonassimilation and metabolism in C4 plants. It has beenshown that many C4 plants respond positively to growthunder elevated Ca in greenhouses and growth chambers (Ziska& Bunce 1997; Ghannoum & Conroy 1998; Wand et al. 1999; Watling,Press & Quick 2000). In addition, a recent study hasshown that free-air CO2 enrichment (FACE) enhances C4 photosynthesisin Sorghum bicolor in the field and increases crop yieldand biomass production during drought conditions (Cousins et al.2001; Ottman et al. 2001; Wall et al. 2001).

In C3 plants, low intercellular CO2 concentrations(Ci) increases the oxygenation of ribulose-1,5-bisphosphate (RuBP)which leads to reduced rates of photosynthesis and relative increasesin photorespiration. Photorespiration is an important electron sink,which may help protect the photosynthetic apparatus from photo-inhibitionand photodamage when CO2 assimilation is inhibited (Wingler et al. 1999). In contrast,C4 plants utilize a CO2-concentrating mechanismto maintain a near-saturating supply of CO2 to RuBP carboxylase/oxygenase(Rubisco) within the bundle sheath cells (BSC) under ambient Caconcentrations and mild drought conditions. However, C4 photosynthesiscan become inhibited during moderate to severe drought conditionsprimarily due to an inability to concentrate CO2 withinthe BSC (Lal & Edwards 1996). The inhibitionof the concentrating mechanism is likely due to either low Ci orreduced conductance of CO2 within the mesophyll cells (Lal & Edwards 1996). Under such conditions,the rate of photorespiration in C4 plants would probablyincrease as the ratio of CO2/O2 withinthe BSC decreases. However, in mature sorghum leaves (NAPDH-typeC4 plant) the relationship of linear electron transportand carbon assimilation is constant under various Ci concentrations,indicating that photorespiration and other electron sinks (i.e.Mehler reaction or pseudocyclic electron transport) are minimal evenunder low CO2 availability (Cousins et al.2001).

Changes in carbon assimilation and metabolism require maintaininga balance between absorbed light energy, heat dissipation and photochemistry(Pammenter, Loreto & Sharkey 1993; Drake, Gonzales-Meler & Long 1997). Over-excitationof the photosynthetic apparatus tends to cause light-induced photo-oxidativedamage inhibiting photosynthesis and reducing crop productivity(Horton 2000). During long-term exposureto different light environments, changes in the size of chlorophyllantennae complexes associated with photosystem I and II are involvedin balancing light absorption and utilization (Anderson1986). In addition, in excess light conditions plants canharmlessly dissipate absorbed light energy as heat termed non-photochemicalquenching (NPQ) (Müller, Xiao-Ping & Niyogi2001). The amount of NPQ is primarily controlled by the magnitudeof the thylakoid ΔpH that is generated by photosyntheticelectron transport. This feed-back-­regulation of NPQ isdependent on a high ΔpH and decreased lumenal pH developedwhen the amount of ­photons absorbed exceeds the photochemicalrequirement (Niyogi 2000). Furthermore,NPQ and low lumenal pH are strongly correlated with the conversionof violaxanthin (V) to zeaxanthin (Z) via the xanthophyll cycleunder numerous environmental conditions (Demmig-Adams1990; Müller et al.2001).

The influence of elevated Ca on C4 photosynthesisand photoprotection is unclear from previous experiments and warrantsfurther investigations (Saccardy et al.1998; Ghannoum et al. 2000).The FACE system provides a unique opportunity to examine the effectsof growth at elevated Ca and other environmental conditions withoutany direct perturbation of microclimate (Hendrey& Kimball 1994). Over a two-season FACE study, elevatedCa increased total yield 15% in the Dry plots and had nosignificant effect in the Wet plots (Ottman et al.2001). In a companion study, Wall et al.(2001) found that elevated Ca enhanced C4 photosynthesisin sorghum proportionately more in the Dry plants than in the Wets;although photosynthesis was also stimulated at midday in the well-wateredplants. We hypothesize that in these sorghum plants growth underelevated CO2 will alleviate the over-excitation of thephotosynthetic apparatus and will decrease NPQ during periods oflimited water availability.

To test this hypothesis we examined how growth under elevatedCa and water-stress conditions affected photosystem II (PSII) photochemistryand photoprotection in sorghum plants grown under a FACE experiment.We have further characterized the effect of CO2 availabilityand water-stress on the relationship of NPQ and the xanthophyllcycle under field and laboratory conditions.

Materialsand methods

Free-air CO2 enrichment (FACE) experiments were conductedat the University of Arizona Maricopa Agricultural Center (MAC),Maricopa, AZ, USA in 1999 to determine the interactive effects ofelevated Ca and drought on ­Sorghum   bicolor  (L)   Moench.   See   Ottman   et al.   2001) fora comprehensive description of the sorghum FACE experiment.

CO2 treatments

The free-air CO2 enrichment (FACE) technique was used toenrich the air in circular plots within a sorghum field similarto prior experiments (Hendrey, Lewin & Nagy1993; Wechsung et al. 1995; Hunsaker et al. 1996; Kimball et al. 1999). Briefly,four replicate 25-m-diameter toroidal plenum rings constructed from0·305-m-diameter pipe were placed in the field shortlyafter planting. The mean daytime values were 566 and 373 µL L−1 andthe mean night-time values were 607 and 433 µL L−1 forFACE and control, respectively.


Each of the main circular FACE and control plots was split insemicircular halves, with each half receiving either an ample (Wet)or a water-stress (Dry) irrigation regime. The water was appliedusing flood irrigation. In 1999 two irrigations were applied tothe Dry treatments (post-plant and mid-season) compared with fivein the Wet treatments.

Leafwater potential

Leaf water potential (ψW) data wererecreated from previously published data by Wall et al.(2001). Briefly, ψW was measuredwith a pressure chamber (Model 3000; Soil Moisture Equipment Corp,Santa Barbara, CA, USA) on the upper most fully expanded leaves(See Wall et al. 2001 for furtherdetails).

Fluorescenceand pigment sampling

A modulated chlorophyll fluorimeter and leaf clip (PAM 2000;Walz, Effeltrich, Germany) was used to measure variation in Fo′, Fm′, Fs,leaf temperature (TL), and PPFD. The chlorophyllfluorescence measurements were analysed as originally describedby Genty, Briantais & Baker (1989)to provide the apparent quantum yield of PSII, φPSII(Fm′ −Fs/Fm′),photochemical quenching coefficient, qp (Fm′ − Fs/Fm′ − Fo′),and the efficiency with which absorbed quanta are transferred tothe open PSII reaction centres Fv′/Fm′ (Fm′ − Fo′/Fm′).Measurements of Fo and Fm weremade following dark adaptation for 15 min to determine Fv/Fm (Fm − Fo/Fm).Non-photochemical dissipation of excitation energy (NPQ) measuredas the Stern–Volmer type of quenching was determined bythe following equation: (Fm −Fm′)/Fm′.Standard fluorescence nomenclature was used (VanKooten & Snel 1990). Measurements were made on threeuppermost fully expanded leaves from each plot and all four replicatetreatments obtained from an undisturbed portion of the canopy underambient light conditions between 20 July and 10 August 1999.

Leaves exposed to full sunlight were immediately freeze-clampedwith a liquid nitrogen-cooled clamp and stored in liquid nitrogenuntil pigment analysis was conducted. Leaf tissue (0·5–1 cm2)was ground down in a glass tissue grinder (Timbrock; Kontes GlassCo., Vineland, NJ, USA) in 500 µL of acetone-ethylacetate [3 : 2 (v/v)].The ethyl acetate extract was fractionated by high-performance liquid chromatography(Hewlett-Packard series 1100) as described by Pogson et al.(1996). The photosynthetic pigments, chlorophyll a and b (Chl),antheraxanthin (A), violaxanthin (V) and zeaxanthin (Z) were identifiedby retention time and spectrophotometric properties by using a photodiode-arraydetector and were quantified by integrating peak area (Pogson et al.1996).

Simultaneousmeasurements of carbon assimilation and fluorescence

Material was sampled prior to 0730 h to avoid any effects ofphoto-inhibition or water stress as previously described by Cousins et al. (2001). Theuppermost fully expanded leaf from three replicate treatments wasexcised below the ligule under water. The base of the leaf was keptin water and the mid-portion was placed into a 6400–06PAM2000 adapter cuvette (LiCor, Inc. Lincoln, NE, USA) which fits thefibre-optic probe of the pulse-modulated fluorometer (PAM 2000;Walz) above the leaf at an angle of 60°. Leaves were dark-adaptedfor a minimum of 1 h, after which simultaneous measurementsof chlorophyll a fluorescence and gas exchange were madeto determine the dark respiration rate (RD). Subsequently,the cuvette was illuminated with ∼ 800 µmol photon m−2 s−1 bya 400 W Agrosun Halide lamp (United Halogen Bulb, Inc.,New York, USA). The leaf temperature was maintained at 30 ± 1 °C.The leaf samples were acclimated for approximately 1 huntil steady-state photosynthesis and chlorophyll a fluorescencewas reached under ambient gas concentrations. Gas exchange and chlorophyll a measurementswere made at Ca concentrations of 75, 200, 370, 570 and 700 µL L−1.The carbon dioxide was supplied from CO2 canisters andthe LiCor 6400 computer software was used to adjust concentrations. Ateach Ca concentration, PS and fluorescence were measured in aircontaining 21% O2 and then 2% O2.Pre-mixed gasses were supplied from pressurized tanks containing21 and 2% oxygen with a nitrogen balance (Air America Liquide,Phoenix, AZ, USA) which fed directly to the intake pumps of theLiCor 6400.

At the end of the measurements the leaf tissue was allowed toreach steady-state photosynthetic rates at growth Ca. Leaf tissuewas removed from the cuvette and immediately freeze-clamped witha liquid nitrogen-cooled clamp and stored in liquid nitrogen untilpigment analysis and biochemical assays were conducted.

The quantum yield of CO2 fixation (φCO2 = PS*/absorbedPPFD) was calculated as the ratio of net CO2 fixationto photosynthetically active photon flux density (PPFD) absorbed(Oberhuber & Edwards 1993; Oberhuber, Dai & Edwards 1993). Leafabsorbance (400–1100 nm) determined using a spectroradiometer(Analytical Spectral Devices, Inc., Boulder, CO, USA) and an external-integratingsphere equipped with an incandescent light source did not differbetween treatments (data not shown).


Leaf tissue was removed from liquid nitrogen and ground in anice-cold glass homogenizer containing 100 mm Tricine (pH 8),10 mm MgCl2,1 mm ethylenediaminetetraacetic acid,14 mm dithiothreitol (DTT),2% polyvinylpyrrollidone, 20% glycerol 1 mm phenylmethylsulfonylfluoride and1 mm NaFl at a ratio of1 cm2 leaf tissue to 1 mL buffer. Aliquotswere assayed for full activity of Rubisco using a 100 mm Tricine(pH 8) buffer containing 10 mm MgCl2, 2 mm DTT,10 mm14C-labelledsodium bicarbonate and 0·4 mm RuBP.For Rubisco full activity, the leaf homogenate was allowed to incubatewith sodium bicarbonate for 10 min before the assay. Additionalaliquots were assayed for PEPCase activity using 50 mm Hepes-KOH,5 mm MgCl2, 10 mm14C-labelledsodium bicarbonate, 10 U mL−1 MDHand 0·2 mm NADHunder optimal conditions [pH 8 and 5 mm phosphoenolpyruvate (PEP)]. Each reaction was timed for 30 sand then terminated with HCl/HCOOH (1 N/4 N).


The data were analysed as a strip-split-plot design using PROCMIXED (Little et al. 1996)for the anovas and regression analysisof variation was used to test the ­treatment effects onthe linear relationship of φPSII and φCO2.



Prior to 4 August 1999 the difference in drought conditions betweenwater treatments was mild as indicated by midday leaf water potential(ψW). The greatest difference in ψW was dueto water treatment on 4 August 1999, just prior to the second andfinal irrigation for the Dry treatment (Fig. 1a). Uponre-irrigation ψW recovered to pre-water-stresslevels.

Figure 1.

(a) Midday leaf waterpotential (ψW) of Dry and Wet irrigationtreatments in control and FACE rings between 20 July and 10 August1999 recreated from Wall et al.(2001). (b) Midday levels of φPSII measured onsunlit upper most fully expanded leaves. Means (± 1 SE)for the four replicate treatments, Control-Dry (CD), FACE-Dry (FD), Control-Wet(CW), FACE-Wet (FW). Three-way anova tested theeffect of CO2, H2O and DAP on φPSII. F-valuesand significance are: CO2 (F1,3·09 = 15·6**),water (F1,5·63 = 51·39***),DAP (F3,35·5 = 11·9***)and their interactions DAP*CO2 (F3,35·5 = 2·23 NS),DAP*H2O (F3,35·5 = 30·55***),CO2*H2O (F1,5·63 = 0·07 NS),DAP*CO2*H2O (F3,35·5 = 0·34 NS).NS denotes not significant; **P ≤ 0·05; ***P ≤ 0·01.

Plants grown under different Ca levels and water availabilityshowed distinct abilities to utilize light energy for photochemistry(Fig. 1b).Midday light levels did not differ between treatments (data notshown). However, the apparent quantum yield of photosystem II, φPSII(Fm′ − Fs/Fm′) wassignificantly enhanced by growth under elevated Ca (F1,3·09 = 15·6**).There were no significant higher-order interactions between CO2 andthe other main factors. Water treatment was significantly dependenton DAP (F3,35·5 = 30·55***)and further analyses by way of two anovasfor each sampling date indicated that the differential water treatmentreduced midday levels of φPSII only on 4 August 1999(P ≤ 0·001) (Fig. 1b)The second-order interaction between DAP, CO2 and H2Owas not significant (F3,35·5 = 0·34).The higher Ci levels reported by Wall et al.(2001) under elevated Ca conditions probably contributedto the enhanced rates of φPSII in the FACE treatments.

Leaf temperature (TL) and PPFD (µmol m−2 s−1)measured at all three-measurement times; mid-morning, midday andlate afternoon, showed no treatment affects (Fig. 2i–l).On 4 August 1999, plants in the Dry plots had a significantly lower φPSIIthan the Wet plants during all three-measurement times (Fig. 2a). Only duringmidday measurements did elevated Ca increase φPSIIin the Wet plants. However, in the Drys elevated Ca significantly enhanced φPSIIover the course of the day (Fig. 2a).On 10 August 1999, shortly after the irrigation of all plots, φPSII showedlittle difference due to treatment effects (Fig. 2b). Variationin φPSII is the product of variation in photochemicalquenching, qp (Fm′ − Fs/Fm′ − Fo′),and the efficiency with which absorbed quanta are transferred tothe open PSII reaction centres (Fv′/Fm′).On 4 August 1999, the enhanced rates of φPSII dueelevated Ca in the severely water-limited plants was explained byboth an increase in qp and Fv′/Fm′.However, in the well-watered plants the midday difference in φPSIIdue to elevated Ca was explained entirely by an increase in Fv′/Fm′ (Fig. 2c–f).The increase in Fv′/Fm′ inthe FACE-Wet plants between 0900 and 1200 h is likely dueto the increase in leaf temperature (Fig. 2k)and the associated slight increase in φPSII (Fig. 2a). In addition,the midday levels of photochemical efficiency of PSII (Fv/Fm)measured after 15 min of dark adaptation were significantlylower than those for the mid-morning and late afternoon measurementson 4 August 1999 (Fig. 2g).However, there was no treatment effect on Fv/Fm oneither date (Fig. 2g &h) or a significant diurnal change on 10 August 1999 (Fig. 2h). On 4 August1999 Fv/Fm in the FACE-Wettreatment was lower then Fv′/Fm′ (Fig. 2e–g).It is not unprecedented in the literature to see Fv′/Fm′ valueslarger then Fv/Fm (e.g. Andrews, Fryer & Baker 1995). No reasonhas been given for this observation. However, one possibility inour experiment is that measurements of Fv/Fm madein the field (dark-adapted for 15 min) may not have allowedfor full Fv/Fm quenchingrelaxation. (Jahns & Miehe 1996; Saccardy et al. 1998). Thevalues of Fv/Fm measuredin the laboratory and dark-adapted for a minimum of 1 hwere substantially higher than field measured Fv′/Fm′ values(see Table 1 and Fig. 2e).

Figure 2.

φPSII(a, b), qp (c, d), Fv′/Fm′ (e,f), Fv/Fm (g, h), PPFD (I,j) and leaf temperature (TL) (k, l) measured onthe upper most fully expanded leaves at mid-morning, midday andlate afternoon. Measurements were made during peak water stress(4 August 1999) and upon re-irrigation (10 August 1999). Control-Dry(CD) closed circles; FACE-Dry (FD) open circles; Control-Wet (CW) closedtriangles; FACE-Wet (FW) open triangles. Symbols shown are the means ± 1 SE.

Table 1.  Mean(SE) carbon fixation (PS), quantum yield of PSII (φPSII),intercellular CO2 concentrations (Ci), non-photochemical quenching(NPQ), total chlorophyll content (µmol Chla+b m−2) xanthophyllcomposition (µmol ZAV m−2),de-epoxidation state of the xanthophyll pool (DPS), PEPC and Rubiscofull activities measured under laboratory conditions on 6 August1999
  1. Measurements were made at growthCa on the upper most fully expanded leaves excised under water.Plants were removed from the field pre-dawn. After photosyntheticmeasurements the leaf tissue was immediately freeze-clamped andstored in liquid nitrogen until assayed for pigment and enzyme content.The effects of CO2 (F1,4)and watertreatment (F1,2) and their interaction (F1,4)were tested for significance by two-way anova. ***P ≤0·01; **P ≤0·05; *P ≤0·1; NS, notsignificant.

PS (µmol CO2 m−2 s−1) 22·6 (3·6) 25·6 (3·7) 27·4 (1·3) 31·1 (1·5)*NSNS
φPSII 0·31 (0·07) 0·37 (0·04) 0·34 (0·05) 0·42 (0·02)NSNSNS
Ci (µL L−1)164·4 (9·5)267·5 (21·5)155·8 (19·4)293·1 (18·7)***NSNS
NPQ 2·6 (0·5) 1·6 (0·3) 2·3 (0·2) 1·5 (0·1)***NSNS
Fv/Fm 0·800 (0·001) 0·800 (0·011) 0·813 (0·003) 0·810 (0·000)NSNSNS
Chlorophyll (µmol Chla+b m−2)740 (49)731 (41)777 (104)721 (29)NSNSNS
Xanthophylls (µmol ZAV m−2) 56·4 (1·7) 53·7 (4·1) 46·7 (6·3) 43·3 (3·1)NS**NS
DPS (ZA mmol ZAV mol−1) 0·35 (0·07) 0·44 (0·09) 0·16 (0·06) 0·28 (0·1)NS*NS
PEPC (µmol CO2 m−2 s−1)530 (33)450 (69)366 (5·0)361·7 (42)NS***NS
Rubisco (µmol CO2 m−2 s−1) 23·9 (4·0) 20·2 (2·7) 17·7 (1·8) 18·5 (0·9)NSNSNS

In the field on 4 August 1999, midday non-photochemical dissipationof excitation energy measured as NPQ (Fm −Fm′/Fm′)was greater (F1,5·43 = 20·82***)in the water-stressed plants than in the well-watered plants (Fig. 3a–d).In addition, on this date the elevated Ca reduced NPQ in both water treatments(F1,1·94 = 9·74*).The treatment effect on NPQ was less significant prior and afterthe dry-down cycle. Freeze-clamped samples taken directly from thefield on this date showed significantly higher levels of the de-epoxidation stateof the xanthophyll pool, DPS (ZA/ZAV) than at other datesmeasured (F3,16 = 4·68*).However, on each of the days measured there were no significantdifferences among treatments; CO2 (F1,16 = 0·0),H2O (F1,16 = 0·18)or their interaction CO2*H2O (F1,16 = 0·49)(Fig. 3e & f).

Figure 3.

Non-photochemicalquenching (NPQ) determined in the field at midday (a–d).Measurements were made on the same leaf tissue and at the same timesas measurements in Fig. 1.The de-epoxidation state of the xanthophyll pool (DPS) from fullsun leaves shown in (e, f). Antheraxanthin (A), Zeaxanthin (Z) and Violaxanthin(V). Leaf tissue used to determine the xanthophyll composition wereimmediately freeze-clamped and stored in liquid nitrogen until measurementswere made.

Controlledenvironment measurements

Rates of carbon assimilation and Ci were significantly enhancedby elevated Ca whereas the values of φPSII measuredat growth Ca concentrations showed no major differences betweenplants from the different growth treatments (Table 1). Rates ofphotosynthesis measured under these conditions were lower then middayfield rates due to lower light levels and leaf temperatures (Wall et al. 2001). There wereno differences in the rates of NPQ due to the prior water-stresstreatment on 6 August 1999; however, NPQ was reduced by elevatedCa in both water treatments (Table 1).The value of Fv/Fm measuredafter 1 h of dark adaptation did not differ due to priortreatment conditions, reconfirming that high light and drought conditionsdid not permanently damage the photosynthetic apparatus. Pigmentanalysis on the same leaf tissue, which was rapidly frozen and storedin liquid nitrogen until measured, showed no treatment effect onthe total chlorophyll content (µmol Chla+b m−2)(Table 1). However,the DPS of the xanthophyll cycle pigments was approximately 57% higher(P < 0·01) in the Dry versusthe Wet plants (Table 1).Additionally, there was no CO2 effect on the total contentof the xanthophyll cycle pigments on a leaf area basis (µmol ZAV m−2);however, the amount of ZAV (µmol ZAV m−2)was higher in the Dry versus the Wet plants (Table 1).

There was no effect of growth Ca on the total PEPC activity;however, water-stressed plants had a PEPC enzyme activity that was35% higher then the wets (P ≤ 0·01)(Table 1). Growthconditions had no effect on the total activity of Rubisco. Rateswere lower then midday field PS measurements due to differencesin assay temperatures and field leaf temperatures (Pittermann& Sage 2001; Wall et al.2001). The ratio of total activity of PEPC to total activityof Rubisco (P/R) was 15% higherin the water-stressed plants due to increased activity of PEPC and unchangedlevels of Rubisco (Table 1).

Regression analysis of variance of φPSII as afunction of φCO2, measured at variousCa, on mature leaves showed a positive intercept but neither theslope nor the intercept was effected by elevated Ca, water treatmentor their interaction (Fig. 4).

Figure 4.

Therelationship between φPSII and φCO2 measuredat various Ca levels. Measurements were made on the uppermost fullyexpanded leaves from all treatments. No significant effect of growthCa, water treatment or their interaction was found on the relationshipbetween φPSII and φCO2.Slope: CO2 (F1,2 = 0·2 NS), water(F1,4 = 0·24 NS)and their interactions CO2*Water (F1,4 = 0·34 NS).Intercept: CO2 (F1,2 = 0·1 NS),water (F1,4 = 0·09 NS) andtheir interactions CO2*Water (F1,4 = 0·16 NS).Symbols are as in Fig. 2.



Under water stress or during midday conditions low stomatal conductancemay reduce Ci concentrations leading to limited rates of CO2 fixation.Under these conditions, growth under elevated Ca will likely increaserates of C4 photosynthesis. In our experiment, middaylevels of φPSII measured on the upper most fullyexpanded sorghum leaves was higher in plants grown under elevatedCa regardless of water treatment (Fig. 1b).Our data suggest that growth under elevated Ca enhanced PSII activityand linear electron transfer in sorghum leaves at midday even inwell-watered plants. This is consistent with Wall et al. (2001),who showed that elevated Ca stimulated photosynthesis rates independentof water treatment. (Table 2, p. 242 in Wall et al.(2001)]. Furthermore, in the Wets there was a generaltrend of higher field photosynthesis rates (although not alwayssignificant) under elevated Ca. The fluorescence data reported herehave less variability due to a larger sampling number.

Increases in midday φPSII by elevated Ca in theWets corresponded to increases in Fv′/Fm′,but qp was unchanged (Figs 2a,c & e). This implies that the variation in electron flowthrough PSII was the result of changes in the efficiency with   which   absorbed   quanta   was   transferred   from   the light-harvestingantenna to the reaction centre and not the ­availabilityof electron acceptors downstream of PSII. ­Additionally,the low midday rates of φPSII corresponded to higherlevels of NPQ in the Control-Wet plants versus FACE-Wet as morelight energy was dissipated harmlessly as heat and not used forphotochemistry in the Control-Wet plants (Fig. 3). However,in the Dry treatments the increases in φPSII dueto growth under elevated Ca corresponded to an increase in bothelectron acceptors downstream of PSII and a reduction in the dissipationof light energy as heat (Figs 2a.c & e). Elevated CO2 decreased non-photochem­icalquenching as well as increased electron flux through PSII in water-stressedplants as would occur when the demand for NADPH changes with changesin carbon ­assimilation (Hymus et al.1999). In contrast, changes in φPSII in theWet plants were not mediated by changes in electron flux but bychanges in the redox-regulated non-photochemical quenching, implyingthat photosynthesis in these plants was less substrate limited (Wall et al. 2001).


Excess light energy can lead to photo-inhibition and photodamage,thereby reducing rates of photosynthesis and plant productivity.Adaptation of the photosynthetic apparatus to withstand droughtconditions is also important for maintaining high rates of photosynthesisunder stress conditions and to avoid irreversible damage to thephotosynthetic capacity (Niyogi 2000).The fact that rates of φPSII recovered to pre-stressedlevels upon re-irrigation (Figs 1b &2b) indicates that limited CO2 fixation and highlight associated with the drought conditions in our experiment (Wall et al. 2001) did nothave a long-term effect on the photosynthetic apparatus. In addition,prior water treatments had no effect on the rates of φPSII, Fv/Fm orPS measured under non-stress conditions on leaves of plants removedfrom the field on 6 August 1999 (Table 1).

The values of φPSII and φCO2 measuredsimultaneously at various levels of Ca in a controlled laboratoryenvironment maintained a constant relationship confirming that in maturesorghum leaves carbon assimilation is the primary electron sinkand photorespiration is minimal even at low Ci levels (Fig. 4). This supportsour assertion that rates of φPSII measured in thefield are dependent on the rate of CO2 fixation and thatphotorespiration and/or electron flux to O2 hasa minimal role in protecting the photosynthetic apparatus in grainsorghum against over excitation.

Reduced levels of φPSII were primarily attributedto increased rates of NPQ as well as an increased reduction stateof PSII (Figs 2& 3). Under elevated Ca, increases in φPSIIand concurrent reductions in NPQ during drought and midday conditionsare probably due to increased CO2 availability (Lal & Edwards 1996; Saccardy et al.1998). This would increase rates of linear electron transferas well as dissipate the large ΔpH developed across thethylakoid membrane as rates of photosynthesis increase. The lackof treatment effect on Fv/Fm (Fig. 2g & h) duringsevere drought conditions indicates that dissipation of excess light energyvia NPQ was sufficient to protect the photosynthetic apparatus againstphotodamage.

The variation in the relationship of NPQ and DPS may be due totreatment effects on the electron transport capabilities and theredox state of the photosynthetic membrane (Förster,Osmond & Boynton 2001). Excess light energy leads toa decrease in the lumenal pH, which is attributed to the inductionof NPQ through the protonation of LHC proteins associated with PSII(PsbS) and activation of the xanthophyll cycle (Müller et al.2001). DPS has been highly correlated with NPQ in a largenumber of plants under a number of environmental conditions (Demmig-Adams 1990). Furthermore, studiesusing inhibitors such as DTT (Horton, Ruban &Walters 1994), or mutants of Arabidopsis and Chlamydomonas,that impair the accumulation of Z and A, show lower levels of NPQ(for review see Niyogi 1999). However,in our experiment, with field-grown sorghum under a FACE experiment,there were large changes in the magnitude of NPQ that did not correspondto changes in DPS. This suggests that under such conditions, NPQmaybe influenced by factors other then the DPS. Interestingly, ithas been shown that Arabidopsis mutants constitutively accumulatingZ still require a low lumenal pH to induce NPQ (Niyogi1999). This indicates that the lumenal pH has an additionalrole in regulating NPQ in addition to the xanthophyll cycle (Niyogi 2000; Muller et al. 2001).Furthermore, time-resolved fluorescence studies have demonstratedthat both lumenal pH and DPS can influence NPQ and that even verylow levels of Z + A can lead to significant NPQ (Gilmore 1997). It is possible that the increasedphotosynthetic rate under elevated Ca was able to alleviate ΔpHsufficiently enough to reduce rates of NPQ, but was insufficientto reverse the xanthophyll cycle in the direction of violaxanthin.

Rubiscoand PEPC activity

Our experiment with mature FACE-grown plants showed no CO2 effecton the total amount of PEPC and Rubisco activity per leaf area (Table 1). However,drought conditions significantly increased the ratio of PEPC toRubisco as previously seen in other experiments (see: Saliendra et al.1996; Ghannoum et al. 2000).During periods of limited CO2 availability and high lightintensity, increases in the ratio of PEPC/Rubisco activitypresumably allow for a greater capacity to concentrate CO2 withinthe BSC, thus maintaining high rates of CO2 fixation(von Caemmerer & Furbank 1999). Underfield conditions, increases of Ca to levels predicted to occur withinthe next 50 years (ambient + 200 p.p.m.)were insufficient to alter key components of the C3 andC4 cycle even when CO2 availability was further limitedby drought conditions. The lack of PEPC and Rubisco response togrowth under elevated Ca in our field-grown plants is inconsistentwith previously published work carried out in growth chambers. Rubiscobut not PEPC activity was reduced in mature maize plants grown under three-timesambient levels of Ca concentrations (Maroco, Edwards& Ku 1999). In another growth chamber experiment, thetotal amount of PEPC but not Rubisco was reduced in sorghum plantsgrown under double Ca levels (Watling et al.2000). The variation in response to elevated Ca is probablyattributed to differences in chamber versus field-growth conditionsas well as differences in light levels and Ca in each of the experiments.


Midday rates of φPSII in sorghum leaves werehigher under elevated Ca in both water treatments. However, mid-­morningand late afternoon rates of φPSII were increased onlyin the Dry treatment. NPQ was higher in the water-stressed plantsbut was reduced by elevated Ca in both water treatments. Higherlevels of NPQ protected against photo-inhibition when φPSIIwas low during water-stress conditions. Additionally, NPQ and pigmentcompositions were affected differently by elevated Ca and waterstress. Low NPQ under elevated Ca did not correlate with DPS, whichmay be attributed to a reduced thylakoid ΔpH that is sufficientto alleviate NPQ but not DPS.


A.B.C. acknowledges support from a NSF Graduate Research TrainingGrant (DGE-9553456). The research was supported by Interagency AgreementNo. DE-AI03–97ER62461 between the Department of Energy,Office of Biological and Environmental Research, Environmental SciencesDivision and the USDA Agricultural Research Service (B.A.K.); byGrant no. 97-35109-5065 from the USDA Competitive Grants Programto the University of Arizona (S.W.L.); and by the USDA, AgriculturalResearch Service as part of the DOE/NSF/NASA/USDA/EPAJoint Program on Terrestrial Ecology and Global Change (TECO III).This work contributes to the Global Change Terrestrial Ecosystem(GCTE) Core Research Programme, which is part of the InternationalGeosphere-Biosphere Programme (IGBP). We also acknowledge the helpfulco-operation of Dr Robert Roth and his staff at the Maricopa AgriculturalCenter. Portions of the FACE apparatus were furnished by BrookhavenNational Laboratory, and we are grateful to Mr Keith Lewin, Dr JohnNagy, and Dr George Hendrey for assisting in its installation andconsulting about its use.

Received 5 March 2002;received inrevisedform 17 June 2002;accepted for publication 17 March 2002