Rapid variations of mesophyll conductance in response to changes in CO 2 concentration around leaves

Rapid variations of mesophyll conductance in response tochanges in CO2 concentration around leaves

JAUME FLEXAS1, ANTONIO DIAZ-ESPEJO2, JERONI GALMÉS1, RALF KALDENHOFF3, HIPÓLITO MEDRANO1 &MIQUEL RIBAS-CARBO1

1Grup de Recerca en Biologia de les Plantes en Condicions Mediterrànies, Departament de Biologia, Universitat de les IllesBalears. Carretera de Valldemossa Km 7.5, 07122 Palma de Mallorca, Balears, 2Instituto de Recursos Naturales yAgrobiología, CSIC & Apartado 105241080 Sevilla, Spain and 3Darmstadt University of Technology, Institute of Botany,Applied Plant Sciences, Schnittspahnstrasse 10, D-64287 Darmstadt, Germany

ABSTRACT

The effects of short-term (minutes) variations of CO2 con-centration on mesophyll conductance to CO2 (gm) wereevaluated in six different C3 species by simultaneous mea-surements of gas exchange, chlorophyll fluorescence, onlinecarbon isotope discrimination and a novel curve-fittingmethod. Depending on the species, gm varied from five- toninefold, along the range of sub-stomatal CO2 concentra-tions typically used in photosynthesis CO2-response curves(AN–Ci curves; where AN is the net photosynthetic flux andCi is the CO2 concentrations in the sub-stomatal cavity),that is, 50 to 1500 mmol CO2 mol-1 air. Although the patternwas species-dependent, gm strongly declined at high Ci,where photosynthesis was not limited by CO2, but by regen-eration of ribulose-1,5-bisphosphate or triose phosphateutilization. Moreover, these changes on gm were found to betotally independent of the velocity and direction of the Ci

changes. The response of gm to Ci resembled that of sto-matal conductance (gs), but kinetic experiments suggestedthat the response of gm was actually faster than that of gs.Transgenic tobacco plants differing in the amounts ofaquaporin NtAQP1 showed different slopes of the gm–Ci

response, suggesting a possible role for aquaporins in medi-ating CO2 responsiveness of gm. The importance of thesefindings is discussed in terms of their effects on parameter-ization of AN–Ci curves.

Key-words: AN–Ci curves; aquaporins; photosynthesis; leafinternal conductance.

INTRODUCTION

Photosynthesis requires diffusion of CO2 from the atmo-sphere into the leaf and finally to the site of carboxylationin the chloroplast stroma. From Fick’s first law of diffusion,the net photosynthetic flux (AN) can be expressed asAN = gs(Ca - Ci) = gm(Ci - Cc); where Ca, Ci and Cc are theCO2 concentrations (mmol mol-1 air) in the atmosphere,the sub-stomatal cavity and the chloroplast stroma,

respectively, with gs and gm being the stomata and meso-phyll conductances, respectively (Long & Bernacchi 2003).

Gas-exchange studies usually assumed that gm was largeand constant, that is, that Ci ª Cc (Farquhar, von Caem-merer & Berry 1980). However, there is now evidence thatgm may be sufficiently small as to significantly decrease Cc

relative to Ci, therefore limiting photosynthesis (Evans et al.1986; Di Marco et al. 1990; Harley et al. 1992; Loreto et al.1992; Warren 2006). Moreover, gm is not constant, and ithas been shown to acclimate during leaf development(Miyazawa & Terashima 2001) and senescence (Loretoet al. 1994; Grassi & Magnani 2005), as well as to the pre-vailing light (Hanba, Kogami & Terashima 2002; Niinemetset al. 2006), nutrient (Warren 2004) and watering conditions(Galmés, Medrano & Flexas 2006) during growth. Thereis also evidence of rapid variation of gm in response toleaf temperature (Bernacchi et al. 2002; Warren & Dreyer2006), water stress (Flexas et al. 2002, 2004; Grassi &Magnani 2005), salinity (Bongi & Loreto 1989; Loreto, Cen-tritto & Chartzoulakis 2003) and virus infections (Sampolet al. 2003). Recently, substantial evidence has been com-piled suggesting a role of aquaporins in the regulation of gm

(Hanba et al. 2004; Flexas et al. 2006). Aquaporin activitycan be regulated by different mechanisms, including directphosphorylation of aquaporins (Kjellbom et al. 1999), anosmotically driven cohesion/tension mechanism (Ye, Wiera& Steudle 2004), pH-dependent gating of aquaporins(Tournaire-Roux et al. 2003), and transcriptional regulationand protein stability (Eckert et al. 1999), most of themoperating in short times (seconds to hours). Therefore, asalready shown for temperature (Bernacchi et al. 2002;Warren & Dreyer 2006), it is likely that rapid variations ingm can be induced by transient changes in the most commonenvironmental conditions, including light intensity, relativehumidity, wind speed or CO2 concentration.

Regarding CO2 concentration, in their early formulationof the two most common fluorescence methods used toestimate gm, Harley et al. (1992) explicitly stated that ‘Theconstant J method worked well over a large range of CO2,but to resolve the effect of CO2 on gm required the variableJ method’. However, despite its evident interest to under-stand plant responses to climate change as well as for the

Correspondence: J. Flexas. Fax: 34-971-173184; e-mail:[email protected]

Plant, Cell and Environment (2007) 30, 1284–1298 doi: 10.1111/j.1365-3040.2007.01700.x

© 2007 The AuthorsJournal compilation © 2007 Blackwell Publishing Ltd1284

mailto:[email protected]

correct interpretation of AN–Ci curves, the response of gm tovarying CO2 has received only little attention over the past15 years. The original data by Harley et al. (1992) showedthat gm was almost halved when Ci was increased from 100to 300 mbar in Quercus rubra, but not in Eucalyptus globu-lus. Moreover, using the isotopic method, Loreto et al.(1992) demonstrated that gm was reduced at 750 mbar Ci ascompared to ambient CO2 in Q. rubra and, specially, inXanthium strumarium. Surprisingly, these authors did notdiscuss the implications of these apparent variations of gm atdifferent CO2 concentrations (Harley et al. 1992; Loretoet al. 1992). More than 10 years later, Düring (2003) was thefirst to show in grapevines that the gm estimated using thevariable fluorescence method varied as much as sixfold in arange of CO2 from 50 to 2000 mmol mol-1 air during theperformance of a typical AN–Ci curve (i.e. within minutes).On the other hand, Centritto, Loreto & Chartzoulakis(2003) showed in salt-stressed olives that, after maintainingthe leaves for an hour at very low CO2, they recovered anAN–Ci curve similar to control leaves. The effect could notbe explained by the increased stomatal conductance only,but also required a concomitant increase in mesophyll con-ductance. It was therefore concluded that the response ofgm might be as rapid and reversible as that of gs (Centrittoet al. 2003). Flexas et al. (2004) showed a similar effect indrought-stressed sunflower. Regarding long-term acclima-tion to high CO2, gm has been observed to decrease only inone species, but not in others (Singsaas, Ort & De Lucia2004; Bernacchi et al. 2005) Finally, using a novel AN–Ci

curve-fitting method to estimate gm, the values obtainedwere reported to depend somewhat on the number andrange of the specific points considered along the curve(Ethier et al. 2006), although these authors attributed it to amathematical artefact rather than to changes in gm alongthe curve.

Despite all this evidence, the possible effects of CO2 onthe regulation of gm are not usually taken into account.Actually, at least three commonly used methods for esti-mating gm, that is, the constant J method (Harley et al.1992), the slope-based variable fluorescence method(Terashima & Ono 2002) and the AN–Ci curve-fittingmethod (Ethier & Livingston 2004; Ethier et al. 2006), relyon the assumption that gm is not affected by CO2 concen-tration. On the other hand, it is now recognized that aproper analysis using the Farquhar et al. (1980) model ofphotosynthesis must be done on a Cc rather than on a Ci

basis (Long & Bernacchi 2003). However, most of thestudies transform AN–Ci curves to AN–Cc curves using asingle value of gm determined at ambient CO2 concentra-tion, that is, also neglecting the possible effects of CO2 on gm

(Flexas et al. 2002; Manter & Kerrigan 2004; Grassi &Magnani 2005).

Therefore, the objectives of the present work were (1) tocharacterize variations of gm in response to rapid (minutes)changes in CO2 concentration in different species, (2) tovalidate the results using independent methods for the esti-mation of gm and (3) to discuss how these variations wouldaffect the use of different methods to estimate gm and the

correct analysis and interpretation of the most commonlyused photosynthesis model (Farquhar et al. 1980), which isbased on the response of net photosynthesis to CO2

concentration at the site of ribulose-1·5-bisphosphatecarboxylase/oxygenase (Rubisco).

MATERIAL AND METHODS

Plant material and growing conditions

Three-year-old olive trees (Olea europaea var. europaea)and 2-year-old plants of Richter-110 (a hybrid of Vitisberlandieri ¥ Vitis rupestris) were growing outdoors undertypical Mediterranean conditions in the experimental fieldof the University of the Balearic Islands. Olive trees wererooted in a clay-calcareous soil, while Vitis plants weregrowing in 40 L pots containing a mixture of perlite, horti-cultural substrate and clay.The measurements in olives weremade during winter (early March), while in Vitis were madeduring summer (early August). The other species, includingcucumber (Cucumis sativus L.), arabidopsis [Arabidopsisthaliana (L.) Heyhn. genotype Col-0], tobacco (Nicotianatabacum L. var. Samsun), flowering tobacco (Nicotianasylvestris L.) and Limonium gibertii (Sennen) Sennen weregrown in a growth chamber in 10 L pots containing a mixtureof perlite, horticultural substrate and clay. Transformedtobacco (N. tabacum L.) plants with different levels of aqua-porin NtAQP1, differing in mesophyll conductance to CO2

(Flexas et al. 2006), were also studied. Antisense and over-expressing plants were obtained from different lines: var.Samsun for the antisense (AS) lines (Siefritz et al. 2002) andline Hö 20.20 for the overexpressing (O) lines (Uehlein et al.2003). Plants of each line with normal NtAQP1 expressionwere used as controls (CAS and CO). The environmentalconditions were set to a 12 h photoperiod (25 °C day/20 °Cnight), 40–60% relative humidity and a photon flux densityat plant height of ca. 900 mmol m-2 s-1 (halogen lamps),except in Arabidopsis (300 mmol m-2 s-1). All plants weredaily irrigated at field capacity.

Gas-exchange and chlorophyllfluorescence measurements

All measurements were made on young, fully expandedleaves. Respiration in the light or ‘day’ respiration (Rd) andthe apparent CO2 photocompensation point (Ci*) weredetermined according to the method of Laisk (1977) asdescribed in von Caemmerer (2000). AN–Ci curves weremeasured, using an open gas-exchange system Li-6400 (Li-Cor Inc., Lincoln, NE, USA), at three different photosyn-thetically active photon flux densities (PPFDs) (50, 200 and500 mmol m-2 s-1) at six different CO2 levels ranging from300 to 50 mmol CO2 mol-1 air. The intersection point of thethree AN–Ci curves was used to determine Ci* (x-axis) andRd (y-axis). Ci* was used as a proxy for the chloroplasticCO2 photocompensation point (G*), according to Warren &Dreyer (2006). These values ranged from 33 mmolCO2 mol-1 air in Olea to 44 mmol CO2 mol-1 air in

CO2-induced variations of mesophyll conductance 1285

© 2007 The AuthorsJournal compilation © 2007 Blackwell Publishing Ltd, Plant, Cell and Environment, 30, 1284–1298

Arabidopsis, corresponding to Rubisco specificity factorsbetween 85 and 112, that is, in good agreement with thereported values for C3 plants (Galmés et al. 2005).

All other leaf gas-exchange parameters were determinedsimultaneously with measurements of chlorophyll fluores-cence using the open gas-exchange system Li-6400 (Li-CorInc.) with an integrated fluorescence chamber head (Li-6400-40; Li-Cor Inc.).The actual photochemical efficiency ofphotosystem II (fPSII) was determined by measuring steady-state fluorescence (Fs) and maximum fluorescence during alight-saturating pulse of ca. 8000 mmol m-2 s-1 (Fm′) follow-ing the procedures of Genty, Briantais & Baker (1989):

φPSII Fm Fs Fm= ′ −( ) ′ (1)

The electron transport rate (Jflu) was then calculated as

Jflu PSII PPFD= ⋅ ⋅ ⋅φ α β (2)

where PPFD is the photosynthetically active photon fluxdensity, a is leaf absorptance and b reflects the partitioningof absorbed quanta between photosystems II and I (PSIand PSII). a was measured using a spectroradiometer(HR2000CG-UV-NIR; Ocean Optics Inc., Dunedin, FL,USA) as described by Schultz (1996), using the light sourcefrom the Li-6400 and making the measurements inside adark chamber. In addition, the product a · b was deter-mined from the relationship between fPSII and fCO2

obtained by varying either light intensity or CO2 concentra-tion under non-photorespiratory conditions in an atmo-sphere containing less than 1% O2 (Valentini et al. 1995).An example of such relationships is shown in Fig. 1 fortobacco. No differences were observed when light intensityor CO2 concentration was changed.Therefore, only changesin light intensity were used for the other species. The

resulting a, as determined with the spectroradiometer,ranged between 0.88 (Arabidopsis) and 0.95 (Limonium),in agreement with other reports using the Li-6400 lightsource (Niinemets et al. 2005). The product a · b, as deter-mined according to Valentini et al. (1995), ranged between0.35 (Vitis) and 0.45 (Arabidopsis), depending on thespecies. The resulting value of light partitioning towardsPSII (b) ranged between 0.39 and 0.51, that is, well withinthe range of typically reported values (Laisk & Loreto1996).

CO2-response curves were performed in light-adaptedleaves of six different plants for each species. Except whenspecifically indicated, photosynthesis was induced with aPPFD of 1500 mmol m-2 s-1 (previously performed light-response curves had shown this to be sufficient light tosaturate photosynthesis in all the species analysed) and400 mmol mol-1 CO2 surrounding the leaf (Ca). The amountof blue light was set to 10–15% PPFD to maximize stomatalaperture. Leaf temperature was close to 25 °C, and leaf-to-air vapour pressure deficit was kept between 1.2 and 1.8 kPaduring all measurements. Once steady state was reached(usually 30 min after clamping the leaf), a CO2-responsecurve was performed. Gas exchange and chlorophyll fluo-rescence were first measured at 400 mmol mol-1, then Ca waseither increased stepwise until 1800 mmol mol-1 or decreasedstepwise until 50 mmol mol-1. Upon completion of measure-ments at high or low Ca, this was returned to 400 mmol mol-1

to restore the original AN. Then, Ca was either decreased orincreased (depending on the previous treatment) stepwise tocomplete the curve. The number of different Ca values usedfor the curves ranged between 10 and 12, depending on thespecies, and the time lag between two consecutive measure-ments at different Ca was restricted to 2–4 min, so that eachcurve was completed in 30–40 min.

Leakage of CO2 into and out the leaf cuvette was deter-mined for the range of CO2 concentrations used in thisstudy with photosynthetically inactive leaves of eachspecies enclosed in the leaf chamber (obtained by heatingthe leaves until no variable chlorophyll fluorescence wasobserved), and used to correct measured leaf fluxes (Flexaset al. 2007).

Estimation of gm by gas exchange andchlorophyll fluorescence

The method by Harley et al. (1992) was used to make esti-mations of gm as

g A C J A R J A Rm N i flu N d flu N d*= − ⋅ + ⋅ +( )( )( ) − ⋅ +( )( )( )Γ 8 4(3)

were AN and Ci are taken from gas-exchange measurementsat saturating light and G* and Rd were estimated using theLaisk (1977) method (see previous section).

The calculated values of gm were used to convert AN–Ci

curves into AN–Cc curves using the following equation:

C C A gc i N m= − ( ) (4)

fCO2

0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07

f PS

II

0.0

0.2

0.4

0.6

0.8

Varying light

Varying CO2

Figure 1. Example of the relationship between photochemicalefficiency of photosystem II (fPSII) and fCO2 [(An + Rd)/PPFD] intobacco leaves, obtained by varying either light intensity (emptysymbols) or CO2 concentration (filled symbols) undernon-photorespiratory conditions in an atmosphere containing lessthan 1% O2 (Valentini et al. 1995).

1286 J. Flexas et al.


From AN–Cc curves, the maximum carboxylation capacity(Vc,max) and the maximum capacity for electron transportrate (Jmax) were calculated using the temperature depen-dence of kinetic parameters of Rubisco described on a Cc

basis by Bernacchi et al. (2002), whereby net assimilationrate is given as

A A A R= { } −min c q d, (5)

with

A VC

C K o Kc c,

c

c c i o

*= −+ + ( )[ ]max

Γ1

(6)

AJ C

Cq

c

c

**

= −( )+( )

ΓΓ4 2

(7)

where Ac and Aq represent photosynthesis limited by car-boxylation and RuBP regeneration, respectively, Kc and Ko

are the Rubisco Michaelis–Menten constants for carboxy-lation and oxygenation, respectively, and oi is the leaf inter-nal oxygen concentration (assumed equal to the external).

Estimation of gm by a curve-fitting method

A new curve-fitting method (Ethier & Livingston 2004;Ethier et al. 2006) was used to estimate gm in some experi-ments. In short, the method by Ethier & Livingston (2004)fits A–Ci curves with a non-rectangular hyperbola versionof the Farquhar’s biochemical model of leaf photosynthesis.This is based on the hypothesis that gm reduces the curva-ture of the A–Ci response curve. The combination of Eqn 4with Eqns 6 and 7 yields the following quadratic equations,whose solutions are the positive roots:

Ab b ac

ac = − + −2 4

2(8)

where

a g= −1 m

b V R g C K K= −( ) + + +( )c, d m i c oOmax 1

c R C K K V C= + +( )( ) − −( )d i c o c, iO *1 max Γ

Ab b ac

aj = − + −2 4

2(9)

a g= −1 m

b J R g C= −( ) + +4 2d m i *Γ

c R C J C= +( ) − −( )d i i* *2 4Γ Γ

Values of Kc,Ko,G* and their temperature responses,werethe Cc-based in vivo values from Bernacchi et al. (2002).The

Ci cut-off point was determined based on the method pro-posed by Ethier et al. (2006).This method has been success-fully used in several studies, showing good agreement withother independent estimates of gm (Niinemets et al. 2006;Warren & Dreyer 2006). Although some of the curvesshowed a third region, with photosynthesis limited by triose-phosphate use (e.g. Limonium in Fig. 1), the curve-fittingmethod was not applied to that region, because its param-eterization is not straightforward (von Caemmerer 2000)and, depending on the equations chosen, it leads to arbitraryestimations of gm (data not shown).

Estimation of gm by carbon isotopediscrimination at near to ambient and high CO2

Instantaneous carbon isotope discrimination was measuredas previously described (Flexas et al. 2006). Leaves wereplaced in the chamber of the Li-6400 at 400 mmol mol-1, aPPFD of 1500 mmol m-2 s-1 and at 25 °C.Gas-exchange para-meters were measured as described with the Li-6400 systemunder steady-state conditions for a minimum of 45 min.

Once gas-exchange measurements were performed, theair exiting the cuvette was collected as follows: maintainingthe leaf in the cuvette under steady state by maintaining thesame conditions of light, CO2 concentration and tempera-ture, the exhaust tube was disconnected of the match valveand connected through a series of Swagelock tube connec-tors to a magnesium perchlorate (water trap) tube and to ahandmade 100 mL glass flask with Teflon stopcocks (Ribas-Carbo, Still & Berry 2002). Under steady-state conditions,the air passed through the desiccant and the open collectingbottle for 15 min at a flow above 150 mL min-1, ensuring 20full turnovers of air inside the collecting bottle before thestopcocks were closed and the bottle removed. In order tocollect a reference air, the same procedure was followedwith the cuvette empty.

Carbon isotope composition was determined in anisotope ratio mass spectrometer (IRMS) (Thermo DeltaXPlus, Bremen, Germany) under dual-inlet mode. CO2

from the bottles (sample and reference) was first concen-trated in a Precon loop under liquid nitrogen and thenintroduced in its corresponding fully expanded bellow. Thebellows were then compressed to increase the signal forthe m/z 44 peak to a minimum of 1000 mV to maximizethe signal/noise ratio. The dual-inlet IRMS compared theisotope ratio of the sample and reference CO2 introduced inits bellows. Firstly, the system performs a peak centre on(m/z 45), then it equilibrates the sample and referencesignal for the m/z 44 peak and then both isotope ratiosare compared 25 times. The SD of the d13C of the sampleCO2 with respect to the reference CO2 was always below0.03‰.

Carbon isotope discrimination was calculated asdescribed by Evans et al. (1986), as

Δ13 13 13 13

13 13

1000C C C CC C

obs o e o

o e

= −( ) + −([−( ))]

ξ δ δ δξ δ δ

(10)



where x = Ce/(Ce - Co), and Ce and Co are the CO2 concen-trations of the air entering and leaving the chamber, respec-tively. Because of the dual-inlet comparison method used,the value of d13Ce was equal to 0 and d13Co was the valueobtained from the isotope analysis.

Mesophyll conductance values were determined by com-paring predicted discrimination with observed discrimina-tion. Predicted discrimination (Di) was calculated from thefollowing equation by Evans et al. (1986):

Δ i i a= + −( )a b a p p (11)

where a is the fractionation occurring due to diffusion in air(4.4‰), b is the net fractionation by Rubisco and phospho-enolpyruvate carboxylase (PEPC) (29‰), and pi and pa arethe intercellular and ambient partial pressures of CO2,respectively.

Finally, gm was calculated from the equation (Evans &von Caemmerer 1996):

Δ Δi obs N m a− = −( )( )13 29 1 8. A g p (12)

where 1.8‰ is the discrimination due to dissolution anddiffusion of CO2 in water.

After collecting air from the cuvette at 400 mmolCO2 mol-1 air, Ca was increased to 1500 mmol CO2 mol-1 air,and the leaf allowed acclimating for 15 min. Then, air col-lecting and gm determinations were repeated as described.

RESULTS

Fast AN–Ci curves (30 to 45 min) were analysed in up to sixdifferent C3 species (Fig. 2). The initial portion of the curveshowed almost linear dependence of AN versus Ci, indicat-ing limitation by carboxylation. A second portion wasclearly curvilinear, indicating limitation by regeneration ofribulose-1,5-bisphosphate. In three of the species (Arabi-dopsis, Limonium and Vitis), the curves displayed a third

Figure 2. Response of netphotosynthesis (AN, large symbols) andelectron transport rate (Jflu, crosshairs) tosub-stomatal CO2 concentrations (Ci) in(a) Arabidopsis thaliana (filled circles),(b) Limonium gibertii (empty circles), (c)Nicotiana tabacum (filled upwardstriangles), (d) Vitis berlandieri ¥ Vitisrupestris (empty upwards triangles), (e)Cucumis sativus (empty downwardstriangles) and (f) Olea europaea var.europaea (open downwards triangles).Values are averages � SE of three to sixreplicates, depending on the species.

(a)

0 200 400 600 800 1000 1200 1400 1600

AN (mm

ol C

O2 m

–2 s

–1)

AN (mm

ol C

O2 m

–2 s

–1)

AN (mm

ol C

O2 m

–2 s

–1)

0

10

20

30

40(b)

0 200 400 600 800 1000 1200 1400 160050

100

150

200

250

(c)

0 200 400 600 800 1000 1200 1400 1600

0

10

20

30

40(d)

0 200 400 600 800 1000 1200 1400 160050

100

150

200

250

(e)

Ci (mmol CO2 mol–1 air) Ci (mmol CO2 mol–1

air)

0 200 400 600 800 1000 1200 1400 1600

0

10

20

30

40(f)

0 200 400 600 800 1000 1200 1400 1600

J flu (mm

ol e

– m–2

s–1

)J fl

u (mm

ol e

– m–2

s–1

)J fl

u (mm

ol e

– m–2

s–1

)

50

100

150

200

250



portion consisting of constant or slightly declining AN withincreasing Ci, indicating limitation by triose phosphate uti-lization. Jflu responded to Ci in a biphasic mode in all speciesexcept Olea, with maximum Jflu being at Ci between 200 and600 mmol CO2 mol-1 air, depending on the species (Fig. 2).Such a biphasic response was already described by Sharkey,Berry & Sage (1988) under high light intensities, and it isthought to be due to feedback limitation by triose phos-phate utilization. Because gas exchange was measured con-comitantly with chlorophyll fluorescence, it was possible toestimate gm for most Ci concentrations used, except at verylow Ci, where AN was close to zero or negative, resulting instrongly variable and often not reliable gm estimates. Atambient CO2, gm ranged from 0.127 mol CO2 m-2 s-1 in Oleato 0.315 mol CO2 m-2 s-1 in Limonium. If the bulk of themesophyll resistance is in the liquid phase, it is theoreticallymore correct to include a pressure term in the units but forthe purposes of this paper, we have used units without apressure term for easier comparison between stomatal and

mesophyll conductance values. Atmospheric pressureduring these measurements was close to 101 kPa.

Contrary to what was usually assumed, gm was not con-stant along the range of Ci used for AN–Ci curves (Fig. 3). Ithas been suggested that the reliability of gm data is question-able when dCc/dAN is lower than 10 or higher than 50 (Harleyet al. 1992).These values are shown by shaded areas in Fig. 3(left, dCc/dAN < 10; right, >50).Taking into consideration thegm values within the reliable range, it is observed that gm

decreases as Ci increases in all the species studied (Fig. 3).Athigh Ci, gm values are as low as 5 to 30% those at lower Ci.

Transgenic tobacco plants differing in the expression ofaquaporin NtAQP1 and gm (Flexas et al. 2006) also showeda similar dependence of gm on Ci (Fig. 4). In these plants, thehigher the gm at ambient CO2, the stronger its dependencyon Ci was. Consequently, the largest differences betweengenotypes were observed at low Ci and vice versa (Fig. 4).

Several tests were performed in Nicotiana to demon-strate the independence of this pattern on the experimental

Figure 3. Response of mesophyllconductance (gm) to sub-stomatal CO2

concentrations (Ci) in (a) Arabidopsisthaliana (filled circles), (b) Limoniumgibertii (empty circles), (c) Nicotianatabacum (filled upwards triangles), (d)Vitis berlandieri ¥ Vitis rupestris (emptyupwards triangles), (e) Cucumis sativus(empty downwards triangles) and (f)Olea europaea var. europaea (opendownwards triangles). Values areaverages � SE of three to six replicates,depending on the species. Un-shadedregions indicate gm data with a dCc/dAN

between 10 and 50, which are reliableaccording to Harley et al. (1992).

0 200 400 600 800 1000 1200 1400 1600

0.0

0.2

0.4

0.6

0.8

(a)

0 200 400 600 800 1000 1200 1400 1600

gm

(mo

l CO

2 m

–2 s

–1) g

m (m

ol C

O2

m–2

s–1)

gm

(mo

l CO

2 m

–2 s

–1)g

m (m

ol C

O2

m–2

s–1)

gm

(mo

l CO

2 m

–2 s

–1)

gm

(mo

l CO

2 m

–2 s

–1)

0.0

0.2

0.4

0.6

0.8

0 200 400 600 800 1000 1200 1400 1600

0.00

0.05

0.10

0.15

0.20

0.25

Ci (mmol CO2 mol–1 air) Ci (mmol CO2 mol–1

air)

0 200 400 600 800 1000 1200 1400 1600

0.00

0.05

0.10

0.15

0.20

0.25

(b)

C

0 200 400 600 800 1000 1200 1400 1600

0.0

0.1

0.2

0.3(c) (d)

(e) (f)

0 200 400 600 800 1000 1200 1400 1600

0.00

0.05

0.10

0.15

0.20

0.25

0.30



procedure. Firstly, the direction of changes in Ca was analy-sed. Therefore, curves performed by initially decreasing Ca

from 400 mmol CO2 mol-1 air to 50 mmol CO2 mol-1 air, andthen returned to 400 mmol CO2 mol-1 air and increased to1800 mmol CO2 mol-1 air were compared to curves thatwere performed the opposite way. Curves of AN, Jflu and gm

versus Ci were almost identical (data not shown).The interference of light intensity to gm was also tested

with several AN–Ci curves performed at lower light intensi-ties (1000, 750 and 250 mmol m-2 s-1). Light intensity seemedto have some effect, because the dependency of gm on Ci

was larger at higher light intensity (Fig. 5). Nevertheless, gm

declined with increasing Ci in all cases. Finally, any possibleeffect that the short time employed between two consecu-tive measurements at different Ci could produce on gm wastested. Different leaves were acclimated to different Ca untilsteady state of both AN and gs was reached (typically 45 to60 min after clamping the leaf). Data from different leavesat different Ca were plotted together to produce an ‘artifi-cial’ AN–Ci curve. The resulting plot was very similar to thatof the ‘real’ AN–Ci curves shown in Fig. 3 (data not shown).

Therefore, it seems that, regardless of the procedure andconditions to modify Ci, gm showed a dependency of Ci

similar to that of gs.The velocity of response of both conduc-tances was tested in Nicotiana (Fig. 6). Firstly, AN, Jflu, Ci, gs

and gm were measured in steady state at almost ambient CO2,then Ca was suddenly increased to 1500 mmol CO2 mol-1 air,and the leaves allowed to acclimate to the new conditions forabout 1 h. AN, Jflu, Ci, gs and gm were then determined. Finally,Ca was returned to 400 mmol CO2 mol-1 air to follow gs andgm recovery (Fig. 6). Both gs and gm decreased at high CO2.

Then, when Ca was returned to lower values, both conduc-tances recovered progressively to their initial values. gm

almost fully recovered in about 15 min, while gs neededabout 45 min to fully recover.

The concomitant decrease of both gs and gm in response toincreasing CO2 strongly reduced the chloroplast CO2 con-centration (Cc) with respect to ambient (Ca), particularly athigh CO2 concentrations. Overall, the Ci/Ca ratio rangedbetween 0.50 and 0.76 at 400 mmol CO2 mol-1 air and,in mostspecies, increased at 1800 mmol CO2 mol-1 air (data notshown). By contrast, the Cc/Ca ratio ranged only between0.25 and 0.57 at 400 mmol CO2 mol-1 air and usuallydecreased at 1800 mmol CO2 mol-1 air (data not shown).

To discard any possible artefact inherent to the fluores-cence method, two totally independent methods were usedto prove the effect of Ci on gm. One experimental compari-son was performed in leaves of N. tabacum to determinegm by online carbon isotope discrimination (Evans et al.1986) at two different CO2 concentrations, 400 and1500 mmol CO2 mol-1 air (Table 1). During this experiment,a second Li-6400 with chlorophyll fluorescence chamberwas attached to the same leaves to simultaneously deter-mine gm by the Harley et al. (1992) method. Carbon isotopediscrimination and fluorescence measurements resulted insimilar estimations of gm and confirmed that, at high Ci, gm

was substantially reduced. However, these results could beaffected by differences in respiration and photorespirationat the two CO2 concentrations, because total carbon isotopediscrimination is affected by discrimination during bothprocesses (Ghashghaie et al. 2003). While the contributionof respiration may be relatively small and not very differentbetween CO2 concentrations, the rate of photorespiration isstrongly affected by CO2, and hence its associated discrimi-nation. Therefore, an additional experiment was performed

Transgenic tobacco(NtAQP1 aquaporin)

200 400 600 800 1000 1200 1400 16000.0

0.1

0.2

0.3

0.4

0.5CAS

AS

CO

O

gm

(mo

l CO

2 m

–2 s

–1)

Ci (mmol CO2 mol–1 air)

Figure 4. Response of mesophyll conductance (gm) tosub-stomatal CO2 concentrations (Ci) in transgenic Nicotianaplants overexpressing aquaporin NtAQP1 (O, empty triangles) oranti-sense (AS, empty circles) and their respective wild-type lines(CO, filled triangles, and CAS, empty triangles). Only thedeclining portion of the gm–Ci relationship is shown for clarity.Values are averages � SE of six replicates. Un-shaded regionsindicate gm data with a dCc/dAN between 10 and 50, which arereliable according to Harley et al. (1992).

0 100 200 300 400 500 600 7000.0

0.1

0.2

0.3

0.4

PPFD 1000

PPFD 750

PPFD 250

gm

(mo

l CO

2 m

–2 s

–1)

Ci (mmol CO2 mol–1 air)

Figure 5. Response of mesophyll conductance (gm) tosub-stomatal CO2 concentrations (Ci) in Nicotiana leavesilluminated at a photosynthetically active photon flux density(PPFD) of 1000 (black circles), 750 (grey circles) and 250 (whitecircles) mmol m-2 s-1. Only the data with a dCc/dAN between 10and 50 are shown. The data shown are single curves for each lightintensity.



in leaves of N. sylvestris to determine gm by online carbonisotope discrimination at two different CO2 concentrations,400 and 1000 mmol CO2 mol-1 air, in atmospheres contain-ing either 21 or 1% O2 (Table 2). The values obtained ford13Co - d13Ce and D13Cobs were similar to those in the previ-ous experiment, and did not differ significantly in theabsence of O2 (Table 2). The results confirmed a decline ofgm at high CO2, and no significant differences at P < 0.05were observed when measurements were taken at 21 or 1%O2 (Table 2), thus suggesting that discrimination duringphotorespiration can be neglected for gm estimations.

An additional independent method, consisting in a novelcurve-fitting of AN–Ci curves was applied. This methodallows solving gm at the two different regions of the curves(Ethier & Livingston 2004; Ethier et al. 2006), that is,together with Vc,max when photosynthesis is limited by car-boxylation (gm Vc,max), and with Jmax when it is limited byRuBP regeneration (gm Jmax). In three of the species (Nic-otiana, Limonium and Olea), a few additional AN–Ci curveswere performed to include a larger number of points (i.e.different Ca), allowing the application of this method. In

Arabidopsis, the method could be applied using the samecurves as for fluorescence. Clearly, this method also sup-ported that gm varies along the Ci gradient during AN–Ci

curves. In fact, averaging fluorescence-derived gm valuesfrom Fig. 3 for each Ci region in each species resulted in ahighly significant correlation with values estimated usingthe curve-fitting method (Fig. 7).

DISCUSSION

Although rapid changes of gm in response to CO2 havepreviously been suggested (Centritto et al. 2003; Düring2003), no detailed analysis of this response has yet beenperformed. The present data show, in six different species,that gm varies in the short term (minutes) by as much as five-to ninefold in response to changes in Ci between 50 and1200 mmol CO2 mol-1 air, that is, well within the timing andvalues generally used during the performance of AN–Ci

curves. gm strongly declines at high Ci, that is, when photo-synthesis is no longer limited by CO2 availability. Thisresponse has been confirmed modifying the timing and con-ditions during AN–Ci curves, as well as by two totally inde-pendent methods for gm estimation, that is, online carbonisotope discrimination method (Evans et al. 1986) and thecurve-fitting method (Ethier & Livingston 2004; Ethieret al. 2006). Recently, we have been informed that similarresults have been obtained by two other research groups(Ethier and Pepin, personal communication; Warren, per-sonal communication).

Using three totally independent methods, based on dif-ferent assumptions, which show similar results decreasesthe possibility that the observed gm variations with CO2 aredue to artefacts during gas-exchange measurements orresult from misleading estimations of the parametersinvolved in the calculations, particularly when using thefluorescence method. Several limitations of current por-table gas-exchange systems have been considered andstudied (Long & Bernacchi 2003): lateral leakage inside andoutside the chamber which was determined and used tocorrect values (Flexas et al. 2007), lateral CO2 diffusionoutside and inside the leaf between the darkened area ofthe leaf under the chamber gasket and the leaf area of thelight, although difficult to quantify, is considered to beminor especially when measurements are done at high lightintensity (Jahnke & Krewitt 2002; Galmés et al. 2006).Regarding the parameters involved in gm calculations usingthe fluorescence method, the critical assumptions are: dayrespiration (Rd), CO2 photocompensation point (G*), leafabsorptance (a), light partitioning between photosystems(b) and the absence of alternative electron-consuming reac-tions, such as the Mehler reaction. The fact that the rela-tionship between fPSII and fCO2 under low O2 was similarwhen changing light intensity or CO2 concentrations (seeFig. 1) suggests that a and b were not affected by CO2

concentration. However, in addition to photosynthesis andphotorespiration, other reactions such as the Mehler reac-tion or nitrite reduction have been shown to consume asmuch as 10% of the total Jflu (Miyake & Yokota 2000; Laisk

Time (s)

0 2000 4000 60005

10

15

20

25

30

35

40

(a)

(b)

130

140

150

160

170

180

190

200

0

200

400

600

800

1000

1200

Ca increased

Ca decreased

B

Time (s)

0 2000 4000 60000.0

0.1

0.2

0.3

0.4

0.5

0.6

0.0

0.1

0.2

0.3

0.4

0.5

0.6

g s (m

ol C

O2 m

–2 s

–1) g

m (m

ol C

O2

m–2

s–1)

AN

(mm

ol C

O2 m

–2 s

–1)

J flu (mm

ol e

– m–2

s–1

)

Ci (mm

ol C

O2 m

ol–1

air

)

Figure 6. Time-course of (a) net photosynthesis (AN, filledcircles), electron transport rate (Jflu, filled squares) andsub-stomatal CO2 concentration (Ci, empty circles), and (b)stomatal (gs, empty circles) and mesophyll conductance (gm, filledcircles) in Nicotiana subjected to changes in CO2 concentrationsin the atmosphere (Ca) from 400 to 1500 mmol CO2 mol-1 air andback (indicated with arrows in panel A). A single experimentrepresentative of a series of similar ones is shown. Un-shadedregions indicate gm data with a dCc/dAN between 10 and 50,which are reliable according to Harley et al. (1992).



et al. 2006), and their incidence can be higher at ambientthan at high CO2 concentration (Miyake & Yokota 2000).Moreover, Rd may be underestimated when using the Laiskmethod (Pinelli & Loreto 2003), and high CO2 has beensometimes reported to substantially reduce Rd (Gonzalez-Meler et al. 1996; Bruhn, Wiskich & Atkin 2007). Finally,using Ci* instead of G* may lead to further errors. Thesetwo parameters relate to each other following the equation:G* = Ci* + Rd/gm. As explained in M&M, gm was not esti-mated close to the compensation point, and hence Ci* wasused as a proxy for G*. To assess the possibility that ourconclusions were due to the interference of the possiblebias in assumed parameters, their effects on gm for severalleaves of the different species were simulated.

An example of such simulations is shown in Table 3 for aleaf of Arabidopsis (simulations for other species resultedin qualitatively similar results, not shown). At ambient(400 mmol CO2 mol-1 air) and high (1500 mmol CO2 mol-1

air) CO2, respectively, leaf temperature was 25.2 and25.7 °C, AN was 13.0 and 23.0 mmol CO2 m-2 s-1, Ci was 283and 1165 mmol CO2 mol-1 air, and Jflu was 112 and125 mmol e- m-2 s-1. Using these data, together withestimations of Rd (-0.46 mmol CO2 m-2 s-1) and Ci*(44 mmol CO2 mol-1 air, used as a proxy for G*) obtainedfrom the Laisk method resulted in gm estimates of 0.110 and

0.032 mmol CO2 m-2 s-1 at ambient and high CO2, respec-tively. Regarding the possible influence of alternativeelectron-consuming reactions, two simulations were per-formed (Table 3). In the first, these reactions were consid-ered to consume 10% of total Jflu (Miyake & Yokota 2000;Laisk et al. 2006), while in the second it was considered thatthese reactions were present at ambient but negligible athigh CO2 concentration (Miyake & Yokota 2000). Withrespect to respiration (Table 3), two possibilities were con-sidered: (1) that the actual Rd was about twice the estimatedone (i.e. close to RN, as suggested by Pinelli & Loreto 2003)and (2) that at high CO2, RD would be reduced by 30%(Gonzalez-Meler et al. 1996; Bruhn et al. 2007). With regardto G*, although no estimation of gm at the compensationpoint was made, judging from Fig. 3, gm at the compensationpoint could be either low (around 0.1 mmol CO2 m-2 s-1 inmost species) or high (between 0.2 and 0.6 mmol CO2 m-2 s-1,averaging 0.4). Therefore, two simulations were performed(Table 3). In the first, gm at near the compensation point wasassumed to be 0.1 mmol CO2 m-2 s-1, resulting in a G* of58.3 mmol CO2 mol-1 air. In the second, gm near the compen-sation point was assumed to be 0.4 mmol CO2 m-2 s-1, result-ing in a G* of 46.4 mmol CO2 mol-1 air.Because G* reflects anintrinsic property of Rubisco, it may be equal at ambient andhigh CO2. At ambient CO2, all the simulated values ranged

Table 1. Net photosynthesis (AN), stomatalconductance (gs), mesophyll conductance(gm), sub-stomatal (Ci) and chloroplast (Cc)CO2 concentrations, the difference in d13Cbetween the air leaving and entering theleaf cuvette (d13Co - d13Ce) and the observed13C discrimination (D13Cobs) by leavesacclimated for 15 min to either 400 or1500 mmol CO2 mol-1 air

Ca (400 mmol CO2

mol-1 air)Ca (1500 mmol CO2

mol-1 air)

AN (mmol CO2 m-2 s-1) 18.9 � 1.1 29.3 � 1.0gs (mol CO2 m-2 s-1) 0.228 � 0.021 0.128 � 0.023d13Co - d13Ce (‰) 0.86 � 0.05 0.44 � 0.04D13Cobs (‰) 12.4 � 0.7 16.7 � 1.2gm Harley (mol CO2 m-2 s-1) 0.213 � 0.018 0.053 � 0.012gm Evans (mol CO2 m-2 s-1) 0.192 � 0.010 0.088 � 0.011Ci (mmol CO2 mol-1 air) 275 � 9 1145 � 62Cc (mmol CO2 mol-1 air) 170 � 17 792 � 72

gm was estimated by two independent methods, a gas-exchange–chlorophyll fluorescencemethod (gm Harley) and online carbon isotope discrimination method (gm Evans).Values areaverages � SE of five replicates per CO2 concentration. The Cc shown was calculated usingthe gm values obtained using the Harley method.

Table 2. Net photosynthesis (AN), stomatal conductance (gs), mesophyll conductance (gm), sub-stomatal (Ci) and chloroplast (Cc) CO2

concentrations, the difference in d13C between the air leaving and entering the leaf cuvette (d13Co - d13Ce) and the observed 13Cdiscrimination (D13Cobs) by leaves of Nicotiana sylvestris acclimated for 15 min to either 400 or 1000 mmol CO2 mol-1 air, and in anatmosphere containing 21 or 1% O2

Ca (400 mmol CO2

mol-1 air) 21% O2

Ca (1000 mmol CO2

mol-1 air) 21% O2

Ca (400 mmol CO2

mol-1 air) 1% O2

Ca (1000 mmol CO2

mol-1 air) 1% O2

AN (mmol CO2 m-2 s-1) 14.0 � 0.4 22.6 � 0.8 17.9 � 1.4 22.7 � 0.7gs (mol CO2 m-2 s-1) 0.192 � 0.010 0.093 � 0.005 0.184 � 0.024 0.075 � 0.002d13Co - d13Ce (‰) 0.71 � 0.02 0.39 � 0.02 0.71 � 0.08 0.32 � 0.02D13Cobs (‰) 15.8 � 0.9 13.7 � 0.3 12.4 � 0.7 11.2 � 0.9gm (mol CO2 m-2 s-1) 0.267 � 0.034 0.177 � 0.055 0.300 � 0.037 0.152 � 0.016Ci (mmol CO2 mol-1 air) 249 � 5 545 � 35 199 � 11 451 � 14Cc (mmol CO2 mol-1 air) 192 � 9 395 � 10 137 � 11 293 � 32

gm was estimated by online carbon isotope discrimination method. Values are averages � SE of five replicates per CO2 concentration.



between 0.118 and 0.145 mmol CO2 m-2 s-1 (i.e. close to theoriginal estimation of 0.110 mmol CO2 m-2 s-1) exceptwhen a G* of 58.3 mmol CO2 mol-1 air was considered,which yielded a much higher gm (0.317 mmol CO2 m-2 s-1).However, the latter value is unlikely, because a G* of58.3 mmol CO2 mol-1 would reflect a specificity factor of only64, that is, the lowest by far ever described for a C3 species(Galmés et al. 2005). In addition, at high CO2, allthe simulated values ranged between 0.031 and0.050 mmol CO2 m-2 s-1. Therefore, although misleadingassumptions in any of the parameters involved in gm calcu-lations may lead to variations in the absolute values ofestimated gm, none of them would impair the conclusion thatgm declines at high CO2. In the case of the isotope method,the major limitation would be the interference of carbonisotope discrimination during photorespiration (Ghash-ghaie et al. 2003), which may be substantially different atambient and high CO2. However, the fact that similar gm

estimations were obtained under 21 and 1% O2 regardless ofCO2 concentration, suggests that the interference of photo-respiration does not affect gm estimations. Altogether, itprovides convincing evidence that decreased gm at high CO2

has a physiological basis,and does not result from artefacts inthe methods used for its estimation.

Except for the species-dependent differences at low Ci,the response of gm to increasing Ci resembles that known forgs (Raschke 1979). Moreover, these data suggest that gm

responds faster to varying Ci than does gs.The data shown inFig. 5 suggest that, as observed for gs, gm responds tochanges in light intensity, although further studies areneeded. Together, decreasing gs and gm as Ci increases

strongly reduces the chloroplast CO2 concentration (Cc)with respect to ambient (Ca), particularly at high CO2 con-centrations. The function of this regulation is unknown, andcan only be speculated. One possibility is that this responseserves to match CO2 availability and photosynthetic capac-ity. When photosynthesis is limited by CO2 availability, gm

tends to be higher (with some exceptions at low Ci), whichmay result in increased CO2 availability in the chloroplaststroma. Consequently, gm decreases when photosynthesis isno longer limited by CO2 availability, that is, at high Ci,which may explain also why it seems to be lower when lightis limiting photosynthesis (Fig. 5). A sustained high gm athigh Ci, when photosynthesis cannot increase the rate ofCO2 consumption, would almost double the CO2 concen-tration in the chloroplast stroma (Cc) according to Ci/Ca andCc/Ca ratios. Maintaining both gs and gm high would result inabout fourfold increase in Cc. According to the known rela-tionship between CO2 concentration in water and pH, thiscould result in a significant decrease of stromal pH (up toabout 0.3 to 0.5 units if stromal content is assumed to bepure water), which could be detrimental to photosynthesisdue to the extreme pH sensitivity of photosyntheticenzymes (Berkowitz, Chen & Gibbs 1983; Pfanz 1995).While this possibility cannot be discarded at present, it isunlikely because efficient pH regulation mechanisms havebeen shown to operate in chloroplasts (Hauser et al. 1995;Oja et al. 1999; Savchenko et al. 2000). Alternatively,perhaps the mechanism leading to a high gm requiresenergy, although this is unknown at present (see furtherdiscussion). If so, at high CO2 where photosynthesis islimited by energy availability, increasing gm would competewith photosynthesis for energy, and would be inefficient asthe added CO2 reaching the chloroplast would result inlittle increase of CO2 fixation.The opposite would be true atlow CO2, that is, energy is in excess to that required forphotosynthesis, and increased CO2 availability would bebeneficial for photosynthesis.

Despite substantial knowledge about the regulation ofstomatal opening, the mechanisms leading to its responseto Ci are still unclear. It has been hypothesized that thesemay be related to malate-induced regulation of anionchannels (Hedrich et al. 1994), to CO2-related pH andmembrane potential changes in guard cells (Hedrich et al.2001), or to CO2-mediated changes in photosynthesis inguard cells (Messinger, Buckley & Mott 2006), in a mecha-nism mediated either by zeaxanthin (Zeiger et al. 2002) orATP (Buckley, Mott & Farquhar 2003). Much less isknown about the regulation of mesophyll conductancevariations, and until recently it was assumed that leafstructural properties were causing most gm variations (vonCaemmerer & Evans 1991; Lloyd et al. 1992). One of thefew physiological bases known is the recent discovery thatsome aquaporins can be involved in the regulation of gm

(Hanba et al. 2004; Flexas et al. 2006). Actually, transgenictobacco plants differing in the amounts of aquaporinNtAQP1 differ in their slope of gm response to Ci (Fig. 2),suggesting that NtAQP1 may also be involved in thisresponse. While the responses of aquaporin physiology to

r 2 = 0.70

gm Ethier (mol CO2 m–2 s–1)

0.1 0.2 0.3 0.4

g m H

arle

y (m

ol C

O2 m

–2 s

–1)

0.1

0.2

0.3

0.4

Figure 7. Relationship between estimates of mesophyllconductance (gm) by combined gas exchange and chlorophyllfluorescence (gm Harley) and the curve-fitting method (gm

Ethier) at different regions of net photosynthesis andsub-stomatal CO2 concentration (AN–Ci) curves [maximumcarboxylation capacity (Vc,max), maximum capacity for electrontransport rate (Jmax). For the Harley method, all values of gm

shown in Fig. 2 within each interval were averaged to get an‘average’ gm for each region. Notice that estimates by the twomethods were made in the same species and Ci intervals, but notin the same leaves and measuring time.



Tab

le3.

Sim

ulat

ion

ofth

eef

fect

sof

poss

ible

erro

rsin

the

mod

elpa

ram

eter

sas

sum

ptio

nson

mes

ophy

llco

nduc

tanc

e(g

m)

esti

mat

esus

ing

the

chlo

roph

yllfl

uore

scen

cem

etho

dby

Har

ley

etal

.(19

92)

ina

leaf

ofA

rabi

dops

isth

alia

naat

400

mmol

CO

2m

ol-1

air

(am

bien

tC

O2)

or15

00mm

olC

O2

mol

-1ai

r(h

igh

CO

2)

A.E

ffec

tof

poss

ible

alte

rnat

ive

elec

tron

flow

and

its

depe

nden

ceon

CO

2.In

the

first

row

,alt

erna

tive

elec

tron

-con

sum

ing

reac

tion

sar

eas

sum

edto

use

10%

ofto

tale

lect

ron

tran

spor

tra

te(J

flu),

whi

lein

the

seco

ndro

wal

tern

ativ

eel

ectr

on-c

onsu

min

gre

acti

ons

are

assu

med

tous

e10

%of

tota

lJflu

only

atam

bien

tC

O2,

bein

gne

glig

ible

athi

ghC

O2.

J flu

Am

bien

tC

O2

(mm

ole-

m-2

s-1)

J flu

Hig

hC

O2

(mm

ole-

m-2

s-1)

g mA

mbi

ent

CO

2

(mm

olC

O2

m-2

s-1)

g mH

igh

CO

2

(mm

olC

O2

m-2

s-1)

101

113

0.14

50.

050

101

125

0.14

50.

032

B.E

ffec

tof

poss

ible

mis

lead

ing

esti

mat

esof

Rd

and

its

depe

nden

ceon

CO

2.In

the

first

row

,Rd

isas

sum

edto

beeq

ualt

ore

spir

atio

nin

the

dark

(RN).

Inth

ese

cond

row

,in

addi

tion

,Rd

athi

ghC

O2

isas

sum

edto

beon

ly30

%th

atat

low

CO

2.

Rd

Am

bien

tC

O2

(mm

olC

O2

m-2

s-1)

Rd

Hig

hC

O2

(mm

olC

O2

m-2

s-1)

g mA

mbi

ent

CO

2

(mm

olC

O2

m-2

s-1)

g mH

igh

CO

2

(mm

olC

O2

m-2

s-1)

0.92

0.92

0.11

80.

033

0.92

0.31

0.11

80.

031

C.E

ffec

tof

usin

gap

pare

ntC

O2

phot

ocom

pens

atio

npo

int

(Ci*

)in

stea

dof

chlo

ropl

asti

cC

O2

phot

ocom

pens

atio

npo

int

(G*)

.Tw

oes

tim

atio

nsar

em

ade.

Inth

efir

stro

w,g

mne

arth

eco

mpe

nsat

ion

poin

tis

assu

med

tobe

0.1

mmol

CO

2m

-2s-1

,res

ulti

ngin

aG

*of

58.3

mmol

CO

2m

ol-1

air.

Inth

ese

cond

row

,gm

near

the

com

pens

atio

npo

int

isas

sum

edto

be0.

4mm

olC

O2

m-2

s-1,r

esul

ting

ina

G*

of46

.4mm

olC

O2

mol

-1ai

r.

G*

Am

bien

tC

O2

(mm

olC

O2

mol

-1ai

r)G

*H

igh

CO

2

(mm

olC

O2

mol

-1ai

r)g m

Am

bien

tC

O2

(mm

olC

O2

m-2

s-1)

g mH

igh

CO

2

(mm

olC

O2

m-2

s-1)

58.3

58.3

0.31

70.

047

46.4

46.4

0.14

40.

037



varying CO2 have not been studied, it has been shown thata change in cytosolic pH of 0.5 to 1.0 units can induce astrong gating of aquaporins through protonation of a his-tidine residue of the protein (Tournaire-Roux et al. 2003),therefore providing a potential mechanism for theobserved CO2-induced regulation. In addition, an energy-dependent gating based on phosphorylation has beendescribed (Kjellbom et al. 1999), which would match theenergy-dependent hypothesis for gm regulation raisedearlier. An alternative mechanism, not based on aquaporinfunction, could be related to chloroplast swelling andshrinkage. Modifications in chloroplast shape, preventingclose association between chloroplast and cell surface,have been shown to alter gm in phytochrome mutants oftobacco (Sharkey et al. 1991). Moreover, gm could beaffected by chloroplast swelling or movements (Sharkey,personal communication; Tholen et al. (2007). Whetherthis occurs under high CO2 remains unknown. Clearly,further studies would be needed to elucidate whichmechanisms can mediate gm responses to CO2.

Regardless of the physiological reasons for and themechanisms involved in the regulation of the response of gm

to CO2, the observed variations during the performance oftypical AN–Ci curves may induce errors on the photosyn-thesis parameterization. These AN–Ci curves are commonlyused to develop prediction models of CO2 assimilation forcrops (Díaz-Espejo et al. 2006) and natural vegetation (Xu& Baldocchi 2003), to help predicting the effects of climatechange on photosynthesis (Rogers et al. 2001), for scalingup from leaf to whole plant and/or ecosystem carbonassimilation models (Harley & Baldocchi 1995), and toasses the influence of stresses on the photosynthetic capac-ity of plants (Centritto et al. 2003; Loreto et al. 2003).There-fore, its correct parameterization is important, and thisshould take into account mesophyll conductance to CO2, asalready highlighted (Ethier & Livingston 2004). However,AN–Ci curves are often transformed to AN–Cc curves using avalue for gm determined at ambient CO2 and assuming aconstant gm for the entire range of Ci (Flexas et al. 2002;Manter & Kerrigan 2004; Grassi & Magnani 2005; Galméset al. 2006), which may not be true according to the presentresults. The effect of neglecting changes of gm with varyingCi on parameterization of AN–Cc curves is illustrated with afew examples in Fig. 8. In some cases, when gm is large, likein Arabidopsis (Fig. 8a) or Vitis (Fig. 8b), the AN–Cc curvedetermined by concomitant gas-exchange and chlorophyllfluorescence measurements along the entire range of Ca

differs from that estimated from a single gm value atambient CO2 mostly on the maximum value of Cc attained,while the differences in parameterization of Vc,max and Jmax

are negligible. In other cases, however, when gm is largelyreduced, such as under severe water stress, the effect isexpected to be substantial, because the difference betweenCi and Cc becomes greater. This is illustrated by theresponse of a Vitis leaf subjected to severe water stress(Fig. 8c). Assuming a constant gm may have led to an esti-mation of Vc,max of only 19 mmol m-2 s-1, accompanied by avery low Jmax. However, taking into account the variations of

(a)

0 200 400 600 800 1000 1200 1400 1600

0

5

10

15

20

25

Variable gm

Vc,max

= 62.5

Jmax

= 101.2

Constant gm

Vc,max

= 65.2

Jmax

= 97.0

Variable gm

Vc,max

= 153.5

Jmax

= 125.3

Constant gm

Vc,max

= 161.3

Jmax

= 122.7

(b)

0 200 400 600 800 1000 1200 1400

0

5

10

15

20

25

(c)

0 200 400 600 800 1000

0

1

2

3

AN (mm

ol C

O2 m

–2 s

–1)

AN (mm

ol C

O2 m

–2 s

–1)

AN (mm

ol C

O2 m

–2 s

–1)

Cc (mmol CO2 mol–1 air)

Variable gm

Vc,max

= 107.0

Jmax

= n.d.

Constant gm

Vc,max

= 18.6

Jmax

= 13.9

Figure 8. Examples of the effects of neglecting mesophyllconductance (gm) variations with sub-stomatal CO2 concentration(Ci) on the parameterization of AN–Cc curves. AN–Cc curves weredetermined by concomitant gas-exchange and chlorophyllfluorescence measurements along the entire range of CO2

concentrations in the atmosphere (Ca, filled circles) or from asingle gm value determined at ambient CO2 (empty circles). Thecurves shown are for well-irrigated Arabidopsis (a), awell-irrigated Vitis (b) and a severely water-stressed Vitis (c). At400 mmol CO2 mol-1 air, estimated gm values were 0.200, 0.291and 0.005 mmol CO2 m-2 s-1 for Arabidopsis, irrigated Vitis andstressed Vitis, respectively. Values of maximum carboxylationcapacity (Vc,max) and maximum capacity for electron transportrate (Jmax) estimated for each curve are shown in the inset.



gm with Ci results in an estimation of Vc,max of up to107 mol m-2 s-1, that is, only slightly depressed as comparedto the irrigated plant, while Jmax cannot be determined. Thelatter result matches better biochemical evidence in water-stressed plants (Flexas et al. 2004).

CONCLUSION

In conclusion, the present study clearly demonstrates thatmesophyll conductance to CO2 changes in response tovarying CO2 even faster than does stomatal conductance,and this should be taken into account for a correct param-eterization of AN–Ci curves, particularly when gm is low as,for instance, under stress.These results are striking and urgethe need for studies regarding both the function and thephysiological and molecular basis of such regulation.

ACKNOWLEDGMENTS

This work was granted by project BFU2005-03102/BFI (PlanNacional, Spain). M. Ribas-Carbo and A. Díaz-Espejo werebeneficiaries of the ‘Programa Ramón y Cajal’ (M.E.C.).Wewould like to thank Dr Biel Martorell for his technical helpon the IRMS and all the staff at the Serveis Científico-Tècnics of the Universitat de les Illes Balears for their helpwith measurements with mass spectrometer. Helpful discus-sions with Drs T.D. Sharkey, F. Loreto and B. Vilanova aregratefully acknowledged.

REFERENCES

Berkowitz G.A., Chen C. & Gibbs M. (1983) Stromal acidi-fication mediates in vivo water stress inhibition of non-stomatal-controlled photosynthesis. Plant Physiology 72, 1123–1126.

Bernacchi C.J., Portis A.R., Nakano H., von Caemmerer S. & LongS.P. (2002) Temperature response of mesophyll conductance.Implications for the determination of Rubisco enzyme kineticsand for limitations to photosynthesis in vivo. Plant Physiology130, 1992–1998.

Bernacchi C.J., Morgan P.B., Ort D.R. & Long S.P. (2005) Thegrowth of soybean under free air [CO2] enrichment (FACE)stimulates photosynthesis while decreasing in vivo Rubiscocapacity. Planta 220, 434–446.

Bongi G. & Loreto F. (1989) Gas-exchange properties of salt-stressed olive (Olea europaea L.) leaves. Plant Physiology 90,1408–1416.

Bruhn D., Wiskich J.T. & Atkin O.K. (2007) Contrasting responsesby respiration to elevated CO2 in intact tissue and isolated mito-chondria. Functional Plant Biology 34, 112–117.

Buckley T.N., Mott K.A. & Farquhar G.D. (2003) A hydromechani-cal and biochemical model of stomatal conductance. Plant, Cell& Environment 26, 1767–1786.

von Caemmerer S. (2000) Biochemical Model of Leaf Photosynthe-sis. CSIRO Publishing, Canberra, Australia.

von Caemmerer S. & Evans J.R. (1991) Determination of theaverage partial pressure of CO2 in chloroplasts from leaves ofseveral C3 plants. Australian Journal of Plant Physiology 18,287–305.

Centritto M., Loreto F. & Chartzoulakis K. (2003) The use of low[CO2] to estimate diffusional and non-diffusional limitations of

photosynthetic capacity of salt-stressed olive samplings. Plant,Cell & Environment 26, 585–594.

Di Marco G., Manes F., Tricoli D. & Vitale E. (1990) Fluorescenceparameters measured concurrently with net photosynthesis toinvestigate chloroplastic CO2 concentration in leaves of Quercusilex L. Journal of Plant Physiology 136, 538–543.

Díaz-Espejo A., Walcroft A.S., Fernández J.E., Hafidi B., PalomoM.J. & Girón I.F. (2006) Modelling photosynthesis in oliveleaves under drought conditions. Tree Physiology 26, 1445–1456.

Düring H. (2003) Stomatal and mesophyll conductances to CO2

transfer to chloroplasts in leaves of grapevine (Vitis vinifera L.).Vitis 42, 65–68.

Eckert M., Biela A., Siefritz F. & Kaldenhoff R. (1999) New aspectsof plant aquaporin regulation and specificity. Journal of Experi-mental Botany 50, 1541–1545.

Ethier G.H. & Livingston N.J. (2004) On the need to incorporatesensitivity to CO2 transfer conductance into the Farquhar–vonCaemmerer–Berry leaf photosynthesis model. Plant, Cell &Environment 27, 137–153.

Ethier G.H., Livingston N.J., Harrison D.L., Black T.A. & MoranJ.A. (2006) Low stomatal and internal conductance to CO2

versus Rubisco deactivation as determinants of the photosyn-thetic decline of ageing evergreen leaves. Plant, Cell & Environ-ment 29, 2168–2184.

Evans J.R. & von Caemmerer S. (1996) Carbon dioxide diffusioninside leaves. Plant Physiology 110, 339–346.

Evans J.R., Sharkey T.D., Berry J.A. & Farquhar G.D. (1986)Carbon isotope discrimination measured concurrently with gasexchange to investigate CO2 diffusion in leaves of higher plants.Australian Journal of Plant Physiology 13, 281–292.

Farquhar G.D., von Caemmerer S. & Berry J.A. (1980) A biochemi-cal model of photosynthetic CO2 assimilation in leaves of C3

species. Planta 149, 78–90.Flexas J., Bota J., Escalona J.M., Sampol B. & Medrano H. (2002)

Effects of drought on photosynthesis in grapevines under fieldconditions: an evaluation of stomatal and mesophyll limitations.Functional Plant Biology 29, 461–471.

Flexas J., Bota J., Loreto F., Cornic G. & Sharkey T.D. (2004)Diffusive and metabolic limitations to photosynthesis underdrought and salinity in C3 plants. Plant Biology 6, 269–279.

Flexas J., Ribas-Carbó M., Hanson D.T., Bota J., Otto B., Cifre J.,McDowell N., Medrano H. & Kaldenhoff R. (2006) Tobaccoaquaporin NtAQP1 is involved in mesophyll conductance toCO2 in vivo. Plant Journal 48, 427–439.

Flexas J., Díaz-Espejo A., Berry J.A., Galmés J., Cifre J., KaldenhoffR., Medrano H. & Ribas-Carbó M. (2007) Leakage in leaf cham-bers in open gas exchange systems: quantification and its effectsin photosynthesis parameterization. Journal of ExperimentalBotany 58, 1533–1543.

Galmés J., Flexas J., Keys A.J., Cifre J., Mitchell R.A.C., MadgwickP.J., Haslam R.P., Medrano H. & Parry M.A.J. (2005) Rubiscospecificity factor tends to be larger in plant species from drierhabitats and in species with persistent leaves. Plant, Cell & Envi-ronment 28, 571–579.

Galmés J., Medrano H. & Flexas J. (2006) Acclimation of Rubiscospecificity factor to drought in tobacco: discrepancies between invitro and in vivo estimations. Journal of Experimental Botany 57,3659–3667.

Genty B., Briantais J.M. & Baker N.R. (1989) The relationshipbetween the quantum yield of photosynthetic electron transportand quenching of chlorophyll fluorescence. Biochimica et Bio-physica Acta 990, 87–92.

Ghashghaie J., Badeck F., Lanigan G., Nogues S., Tcherkez G.,Deleens E., Cornic G. & Griffiths H. (2003) Carbon isotopefractionation during dark respiration and photorespiration.Phytochemistry Reviews 2, 145–162.



Gonzalez-Meler M.A., Ribas-Carbo M., Siedow J.N. & Drake B.G.(1996) Direct inhibition of plant mitochondrial respiration byelevated CO2. Plant Physiology 112, 1349–1355.

Grassi G. & Magnani F. (2005) Stomatal, mesophyll conductanceand biochemical limitations to photosynthesis as affected bydrought and leaf ontogeny in ash and oak trees. Plant, Cell &Environment 28, 834–849.

Hanba Y.T., Kogami H. & Terashima I. (2002) The effect of growthirradiance on leaf anatomy and photosynthesis in Acer speciesdiffering in light demand. Plant, Cell & Environment 25, 1021–1030.

Hanba Y.T., Shibasaka M., Hayashi Y., Hayakawa T., Kasamo K.,Terashima I. & Katsuhara M. (2004) Overexpression of thebarley aquaporin HvPIP2;1 increases internal CO2 conductanceand CO2 assimilation in the leaves of transgenic rice plants. Plantand Cell Physiology 45, 521–529.

Harley P.C. & Baldocchi D.D. (1995) Scaling carbon dioxide andwater vapour exchange from leaf to canopy in a deciduousforest. I. Leaf model parametrization. Plant, Cell & Environment18, 1146–1156.

Harley P.C., Loreto F., Di Marco G. & Sharkey T.D. (1992) Theo-retical considerations when estimating the mesophyll conduc-tance to CO2 flux by the analysis of the response ofphotosynthesis to CO2. Plant Physiology 98, 1429–1436.

Hauser M., Eichelmann H., Oja V., Haber U. & Laisk A. (1995)Stimulation by light of rapid pH regulation in the chloroplaststroma in vivo as indicated by CO2 solubilization in leaves. PlantPhysiology 108, 1059–1066.

Hedrich R., Marten I., Lohse G., Dietrich P., Winter H., Lohaus G.& Heldt H.W. (1994) Malate-sensitive anion channels enableguard cells to sense changes in the ambient CO2. Plant Journal 6,741–748.

Hedrich R., Neimanis S., Savchenko G., Felle H.H., Kaiser W.M. &Heber U. (2001) Changes in apoplastic pH and membranepotential in leaves in relation to stomatal responses to CO2,malate, abscisic acid or interruption of water supply. Planta 213,594–601.

Jahnke S. & Krewitt M. (2002) Atmospheric CO2 concentration maydirectly affect leaf respiration measurement in tobacco, but notrespiration itself. Plant, Cell & Environment 25, 641–651.

Kjellbom P., Larsson C., Johanson I., Karlsson M. & Johanson U.(1999) Aquaporins and water homeostasis in plants. Trends inPlant Science 4, 308–314.

Laisk A.K. (1977) Kinetics of Photosynthesis and Photorespirationin C3 Plants (in Russian). Nauka, Moscow, Russia.

Laisk A. & Loreto F. (1996) Determining photosynthetic param-eters from leaf CO2 exchange and chlorophyll fluorescence.Ribulose-1,5-bisphosphate carboxylase/oxygenase specificityfactor, dark respiration in the light, excitation distributionbetween photosystems, alternative electron transport rate,and mesophyll diffusion resistance. Plant Physiology 110, 903–912.

Laisk A., Eichelmann H., Oja V., Rasulov B. & Rämma H. (2006)Photosystem II cycle and alternative electron flow in leaves.Plant and Cell Physiology 47, 972–983.

Lloyd J., Syvertsen J.P., Kriedemann P.E. & Farquhar G.D. (1992)Low conductances for CO2 diffusion from stomata to the sites ofcarboxylation in leaves of woody species. Plant, Cell & Environ-ment 15, 873–899.

Long S.P. & Bernacchi C.J. (2003) Gas exchange measurements,what can they tell us about the underlying limitations to photo-synthesis? Procedures and sources of error. Journal of Experi-mental Botany 54, 2393–2401.

Loreto F., Harley P.C., Di Marco G. & Sharkey T.D. (1992) Esti-mation of mesophyll conductance to CO2 flux by three differentmethods. Plant Physiology 98, 1437–1443.

Loreto F., Di Marco G., Tricoli D. & Sharkey T.D. (1994) Measure-ments of mesophyll conductance, photosynthetic electron trans-port and alternative electron sinks of field grown wheat leaves.Photosynthesis Research 41, 397–403.

Loreto F., Centritto M. & Chartzoulakis K. (2003) Photosyntheticlimitations in olive cultivars with different sensitivity to saltstress. Plant, Cell & Environment 26, 595–601.

Manter D.K. & Kerrigan J. (2004) A/Ci curve analysis across arange of woody plant species: influence of regression analysisparameters and mesophyll conductance. Journal of ExperimentalBotany 55, 2581–2588.

Messinger S., Buckley T.N. & Mott K.A. (2006) Evidence forinvolvement of photosynthetic processes in the stomatalresponse to CO2. Plant Physiology 140, 771–778.

Miyake C. & Yokota A. (2000) Determination of the rate of pho-toreduction of O2 in the water–water cycle in watermelon leavesand enhancement of the rate by limitation of photosynthesis.Plant and Cell Physiology 41, 335–343.

Miyazawa S.I. & Terashima I. (2001) Slow development of leafphotosynthesis in an evergreen broad-leaved tree, Castanopsissieboldii: relationships between leaf anatomical character-istics and photosynthetic rate. Plant, Cell & Environment 24,279–291.

Niinemets U., Cescatti A., Rodeghiero M. & Tosens T. (2005)Leaf internal diffusion conductance limits photosynthesismore strongly in older leaves of Mediterranean evergreenbroad-leaved species. Plant, Cell & Environment 28, 1552–1566.

Niinemets U., Cescatti A., Rodeghiero M. & Tosens T. (2006)Complex adjustments of photosynthetic potentials and internaldiffusion conductance to current and previous light availabilitiesand leaf age in Mediterranean evergreen species Quercus ilex.Plant, Cell & Environment 28, 1552–1566.

Oja V., Savchenko G., Jakob B. & Heber U. (1999) pH and buffercapacities of apoplastic and cytoplasmic cell compartments inleaves. Planta 209, 239–249.

Pfanz H. (1995) Apoplastic and symplastic proton concentrationsand their significance for metabolism. In Ecophysiology ofPhotosynthesis (eds E.D. Schulze & M.M. Caldwell), pp. 103–122. Springer, Berlin Heidelberg, Germany and New York, NY,USA.

Pinelli P. & Loreto F. (2003) (CO2)-C12 emission from differentmetabolic pathways measured in illuminated and darkened C3

and C4 leaves at low, atmospheric and elevated CO2 concentra-tion. Journal of Experimental Botany 54, 1761–1769.

Raschke K. (1979) Movement of stomata. In Stable Isotopes. Inte-gration of Biological, Ecological and Geochemical Processes (edsW. Haupt & M.E. Feinleb), pp. 383–441. Springer, Berlin Heidel-berg, Germany and New York, NY, USA.

Ribas-Carbo M., Still C. & Berry J.A. (2002) Automated system forsimultaneous analysis of d13C, d18O and CO2 concentration insmall air samples. Rapid Communications in Mass Spectrometry16, 339–345.

Rogers A., Ellsworth D.S., Humphries S.W. (2001) Possible expla-nation of the disparity between in vitro and in vivo measure-ments of Rubisco activity: a study in loblolly pine at elevatedpCO2. Journal of Experimental Botany 52, 1555–1561.

Sampol B., Bota J., Riera D., Medrano H. & Flexas J. (2003) Analy-sis of the virus-induced inhibition of photosynthesis in malmseygrapevines. New Phytologist 160, 403–412.

Savchenko G., Wiese C., Neimanis S., Hedrich R. & Heber U.(2000) pH regulation in apoplastic and cytoplasmic cell compart-ments of leaves. Planta 211, 246–255.

Schultz H.R. (1996) Leaf absorptance of visible radiation in Vitisvinifera L.: estimates of age and shade effects with a simple fieldmethod. Scientia Horticulturae 66, 93–102.



Sharkey T.D., Berry J.A. & Sage R.F. (1988) Regulation ofphotosynthetic electron-transport in Phaseolus vulgaris L., asdetermined by room-temperature chlorophyll a fluorescence.Planta 176, 415–424.

Sharkey T.D., Vassey T.L., Vanderveer P.J. & Vierstra R.D. (1991)Carbon metabolism enzymes and photosynthesis in transgenictobacco (Nicotiana tabacum L.) having excess phytochrome.Planta 185, 287–296.

Siefritz F., Tyree M.T., Lovisolo C., Schubert A. & Kaldenhoff R.(2002) PIP1 plasma membrane aquaporins in tobacco: fromcellular effects to function in plants. Plant Cell 14, 869–876.

Singsaas E.L., Ort D.R. & De Lucia E.H. (2004) Elevated CO2

effects on mesophyll conductance and its consequences for inter-preting photosynthetic physiology. Plant, Cell & Environment27, 41–50.

Terashima I. & Ono K. (2002) Effects of HgCl2 on CO2 dependenceof leaf photosynthesis: evidence indicating involvement of aqua-porins in CO2 diffusion across the plasma membrane. Plant andCell Physiology 43, 70–78.

Tholen D., Boom C., Noguchi K. & Terashima I. (2007) The effectsof chloroplast movement on CO2 transfer conductance inArabidopsis thaliana. Plant and Cell Physiology 48, S95–S95.

Tournaire-Roux C., Sutka M., Javot H., Gout E., Gerbeau P., LuuD.-T., Bligny R. & Maurel C. (2003) Cytosolic pH regulates rootwater transport during anoxic stress through gating of aqua-porins. Nature 425, 393–397.

Uehlein N., Lovisolo C., Siefritz F. & Kaldenhoff R. (2003) Thetobacco aquaporin NtAQP1 is a membrane CO2 transporterwith physiological functions. Nature 425, 734–737.

Valentini R., Epron D., De Angelis P., Matteucci G. & Dreyer E.(1995) In situ estimation of net CO2 assimilation, photosyntheticelectron flow and photorespiration in Turkey oak (Quercus cerrisL.) leaves: diurnal cycles under different levels of water supply.Plant, Cell & Environment 18, 631–640.

Warren C.R. (2004) The photosynthetic limitation posed byinternal conductance to CO2 movement is increased bynutrient supply. Journal of Experimental Botany 55, 2313–2321.

Warren C.R. (2006) Estimating the internal conductance to CO2

movement. Functional Plant Biology 33, 431–442.Warren C.R. & Dreyer E. (2006) Temperature response of photo-

synthesis and internal conductance to CO2: results from twoindependent approaches. Journal of Experimental Botany 57,3057–3067.

Xu L. & Baldocchi D.D. (2003) Seasonal trends in photosyntheticparameters and stomatal conductance of blue oak (Quercus dou-glasii) under prolonged summer drought and high temperature.Tree Physiology 23, 865–877.

Ye Q., Wiera B. & Steudle E. (2004) A cohesion/tension mecha-nism explains the gating of water channels (aquaporins) in Charainternodes by high concentration. Journal of ExperimentalBotany 55, 449–461.

Zeiger E., Talbott L.D., Frechilla S., Srivastava A. & Zhu J. (2002)The guard cell chloroplast: a perspective for the twenty-firstcentury. New Phytologist 153, 415–424.

Received 26 March 2007; received in revised form 8 June 2007;accepted for publication 11 June 2007



Rapid variations of mesophyll conductance in response to changes in CO 2 concentration around leaves

Documents

Transcript of Rapid variations of mesophyll conductance in response to changes in CO 2 concentration around leaves