Planta (1984)160:143 150

P l a n t a 9 Springer-Verlag 1984

Transpiration-induced changes in the photosynthetic capacity of leaves* Thomas D. Sharkey Biological Sciences Center, Desert Research Institute, University of Nevada System, P.O.B. 60220, Reno, NV 89506, U S A

Abstract. High transpiration rates were found to affect the photosynthetic capacity of Xanthium strumarium L. leaves in a manner analagous to that of low soil water potential. The effect was also looked for and found in Gossypium hirsutum L., Agathis robusta (C. Moore ex Muell.) Bailey, Eucalyptus microcarpa Maiden, Larrea divaricata Cav., the wiltyflacca tomato mutant (Lycopersicon esculentum (L.) Mill.) and Scrophularia desertorum (Munz) Shaw. Two methods were used to distinguish between effects on stomatal conductance, which can lower assimilation by reducing CO 2 availability, and effects on the photosynthetic capacity of the mesophyll. First, the response of assimilation to intercellular CO 2 pressure (Ci) was compared under conditions of high and low transpiration. Second, in addition to estimating C i using the usual Ohm's law analogy, C i was measured directly using the closed-loop technique of T.D. Sharkey, K. Imai, G.D. Farquhar and I.R. Cowan (1982, Plant Physiol, 60, 657-659). Transpiration stress responses of Xanthium strumarium were compared with soil drought effects. Both stresses reduced photosynthesis at high C i but not at low Ci; transpiration stress increased the quantum requirement of photosynthesis. Transpiration stress could be induced in small sections of leaves. Total transpiration from the plant did not influence the photosynthetic capacity of a leaf kept under constant conditions, indicating that water deficits develop over small areas within the leaf. The effect * This research was begun while the author was a Postdoctoral Research Fellow at the Australian National University, Canberra

Abbreviations and symbols: A = p h o t o s y n t h e t i c CO 2 assimilation rate; C a = ambient CO 2 partial pressure; C i = partial pressure of CO2 inside the leaf; VPD = leaf-to-air water-vapor pressure difference

of high transpiration on photosynthesis was reversed approximately half-way by returning the plants to low-transpiration conditions. This reversal occurred as fast as measurements could be made (5 min), but little further recovery was observed in subsequent hours. Key words: Photosynthesis and water stress Transpiration - Water stress - Xanthium.

Introduction

The flow of water through plants is maintained by a drop in water potential between the soil and the sites of evaporation within leaves. The water potential of leaves will depend on the water potential in the bulk soil and the water potential drop required to support the rate of transpiration that is occurring. The xylem is generally considered to be a low-resistance pathway for water (for a recent review see Passioura 1982), but the transport of water from the bulk soil to inside the roots and from the ends of the xylem vessels to the sites of evaporation in leaves (Sheriff 1982) is not well understood. Although the transport of water in the leaf is currently debated (Tyree and Yianoulis 1980; Sheriff 1982), it appears that substantial water-potential gradients can develop within leaves. Much attention has been focused on the effects of low leaf water potential that result from low soil water potential on photosynthetic metabolism, but there has been very little investigation of the effect of transpiration rate on photosynthetic metabolism. It has been shown that the first effect of soil water stress on the photosynthetic capacity of leaves can be a reduction of the COy-saturated

144

rate of photosynthesis, with little or no effect on the CO2-1imited rate of photosynthesis (von Caemmeter 1981; Sharkey and Badger 1982; Forseth and Ehleringer 1983). The quantum requirement of photosynthesis (mol photons tool-1 CO 2 fixed) increases when plants are water stressed (Mohanty and Boyer 1976). Ball (1981) first reported that leaves of mangrove exhibited reduced photosynthetic capacity at low humidity after both short- and long-term exposure. She assumed that the effect was via changes in transpiration rate. I have investigated the effects of transpiration rate on the assimilation capacity of the mesophyll (i.e. assimilation as a function of C i, the CO z partial pressure inside the leaf) in Xanthium strumarium and several other glycophytic species. I have measured C i directly in order to establish that the effects were on the mesophyll and not just the result of stomatal movements. I have also estimated Ci in some experiments using the modified Ohm's law analogy (Jarman 1974; von Caemmerer and Farquhar 1981; Leuning 1983) so that transpiration effects could be studied under normal conditions. I have found that high transpiration rates reduce the photosynthetic capacity of leaves in all species that were tested for this effect. The CO zsaturated rate of assimilation was reduced and the quantum requirement was increased by transpiration-induced stress.

Materials and methods Plant material. The following plant species (with family and source) were grown in soil in greenhouses:

Xanthiumstrumarium L. (Asteraceae), Chicago strain; seeds from J.A.D. Zeevaart, Michigan State University, East Lansing, USA Gossypium hirsutum L. (Malvaceae) Agathis robusta (C. Moore ex Muell.) Bailey (Araucariaceae); seeds from D.I. Nicholson, Qld., Australia (see Langenheim et al., in press~ for details on growth conditions) Eucalyptus mierocarpa Maiden (Myrtaceae) Scrophularia desertorum (Shaw.) Munz (Scrophulariaceae), obtained as root cuttings near Reno Helianthus petiolaris Nutt. (Asteraceae) Lycopersieon esculentum (L.) Mill. (Solanaceae), both theflacca mutant and the isogenic parent line cv. Rheinlands Ruhm, seeds from C. Rick, University of California, Davis, U S A In addition, some experiments were done with Larrea divaricata Cav. (Zygophyllaceae) growing in Death Valley, Cal., USA. Greenhouses in which Xanthium was grown had supplemental lighting (~-1 ~tmol photons m - 2 s-1) to extend the day length to 18 h. Plants were generally watered once or more per day and fertilized one to three times per week. Lycopersieon was grown under shade (~-25% of full sunlight) to reduce water stress in the flaeca mutant. All plants were used when they were between three and six weeks old, except the Eucalyptus and the Larrea which were of undetermined age.

T.D. Sharkey: Transpiration effects on photosynthesis

Measurement of gas-exchange characteristics. Air of known water vapor and CO 2 content was passed through aluminum chambers with glass windows clamped on the leaf. Each leaf surface was measured separately. About 0.81 rain -1 air was passed over 3.5 c m 2 leaf area. The humidification of the air was measured with capacitive-type Vaisala humidity sensors (Weather Measure, Sacramento, Cal., USA) or with an infrared gas analyzer (ADC 225; Analytical Development Co., Hoddesdon, UK). The depletion of CO 2 caused by the leaf was measured with infrared gas analyzers (865; Beckman Instruments, Fullerton, Cal., USA). All of the gas analyzers were fitted with optical filters by the manufacturers to reduce cross-sensitivity. Leaf temperature was controlled by circulating water through the aluminum chambers or by thermoelectric heating and cooling. The data from Eucalyptus, Lycopersieon, Larrea, Serophularia and some of the Xanthium data were obtained with a portable gas-exchange system. The rate at which dry gas flowed into a fan-stirred, aluminum chamber was controlled to establish the desired humidity. Humidity was measured by a Vaisala sensor kept at 35 ~ C or an infrared gas analyzer. Assimilation of CO 2 was measured as the amount of CO 2 injected to keep the CO 2 concentration before and after the chamber constant; CO 2 was measured with an infrared gas analyzer (Binos; Leybold-Heraeus, K t l n , FRG). Gas flows were controlled and measured with mass-flow controllers (FC 260; Tylan Corp., Carson, Cal., USA). The principles of this system are the same as those described by Field et al. (1982). The partial pressure of CO 2 inside the leaf (Ci) was measured as described by Sharkey et al. (1982). For this method the clamp-on chambers were used. One chamber was used for normal gas-exchange measurements while the other chamber was made part of a closed loop also having a gas analyzer and peristaltic pump. The CO 2 pressure in the closed loop attained an equilibrium value which is believed to be identical to the CO 2 pressure in the leaf since no net CO 2 exchange occurred across the epidermis. Calculation of gas-exchange parameters. The equations used to calculate the photosynthetic CO 2 assimilation rate (A), stomatal conductance, and Ci were those given by von Caemmerer and F a r q u h a r (1981). These equations include dilution of CO2 in the gas stream by humidification, and the diffusion of gases in the stomatal pore is treated as a ternary system consisting of air, CO 2 and H 2 0 as recommended by J a r m a n (1974). In some cases, all water was removed from the air stream prior to analysis for COz, in which case different equations were used. The level of CO 2 is expressed as partial pressure as recommended by F a r q u h a r and yon Caemmerer (1982). Partial pressure can be converted to mole fraction (~tl 1-i) by dividing by 0.95 bar, the usual atmospheric pressure in Canberra, or 0.85 bar, the usual pressure in Reno. Quantum requirement determination. The q u a n t u m requirement of photosynthesis was determined by determining the net CO 2 exchange rate at five p h o t o n flux densities between darkness and 150 ~tmol photons m -2 s -x. The inverse of the slope of the line fitted to the relationship of A as a function of light tells the n u m b e r of photons required to fix one CO 2 molecule. Measurement of fl-transmissivity, fl-Transmissivity of the leaf was measured using the technique described by Jones (1973). This method involves placing a leaf between a source offl-radiation and a Geiger tube. The degree to which the fl-rays are attenuated is a measure of the thickness of the leaf and over short time periods will reflect changes in water content of the leaf.

T.D. Sharkey: Transpiration effects on photosynthesis I

I

I

I

I

I

A

145 I

I

I

I

B

40 9 1 4 9

Q

E 30

/

E

~

20

.E_

/

lo

/

//

/

l

I

I

100

200

I

!

I

I

300

I

I

I

I

t

400 500 0 100 200 300 Intercellular CO2 pressure (IJbar) I

I

I

I

I

I

400

[

I

500

9

600

Fig. 1. A Response of CO 2 assimilation rate (A) to intercellular CO 2 pressure (Ca) in J(anthium strumarium at evaporation rates of 3.5 + 0.2 (o), 7.0_+ 0.3 (u), and then 3.5+0.2 mmol m -2 s -1 (A) (mean + SE). Evaporation was controlled by changing ambient humidity. B A control experiment where measurements were made over the same time span as in A but without the humidity changes. Leaf temperature was 29 ~ C. Intercellular CO 2 pressure was measured directly using the method of Sharkey et al. (1982). Fifty min were required for determining the first curve, 25 9 for lowering the humidity, 60 9 for the second curve, 35 min for raising the humidity, and the last curve required 55 min. The experiment was begun between 9 and 10 o'clock

9

B 9 i

40 9

o i *

9

/

E 30

-6

/

E

g

/

k~

_~ 20

/

/ o

lo

/ t

I

100

I

200

I

300

I

I

I

I

I

400 500 o lOO 200 300 Intercellular C02 pressure (labar)

Cross-sensitivity o f the CO z analyzer to water vapor. During the course of this investigation a residual cr0ss-sensitivity of the CO z analyzers to water vapor was found, despite the optical interference filters installed to reduce cross-sensitivity. The sensitivity to water vapor depended on the pressure of CO 2 in the air stream such that in CO/-free air there was no sensitivity to water vapor. However, at 1,000 labar COz, changing the water vapor content of the air going into one cell of the infrared gas analyzer did not give the calculated signal based on the analyzer sensitivity and change in CO 2 pressure that should occur because of dilution by water vapor. This effect is most easily observed by comparing the output of a differential infrared gas analyzer with two humid gas streams with the output when an ice trap is put in one of the gas streams. The resultant increase in CO z pressure (because some of the air stream has been removed) should give rise to a signal which can be calculated by knowing the analyzer sensitivity to CO/. In general, only one half of the expected signal is observed. The error is generally small but at high transpiration rate and high ambi-

I

400

I

500

600

Fig. 2. A Response of CO z assimilation rate to intercellular CO 2 pressure in Xanthium strumarium after 2 (m) and 4 (A) d without watering, and before watering was stopped (e). B Behavior of a watered plant measured on the same days. Intercellular CO 2 pressure was measured, except for the open symbols for which intercellular COz pressure was estimated in the normal way. This was necessary to obtain measurements of assimilation at high CO 2. At the end of the experiment the water-stressed leaf was detached and found to have a water potential of - 1 . 1 M P a (as measured in a pressure bomb). The differences between the before-stress data in A and the data in B reflect plant-to-plant variation

ent CO 2 pressure the assimilation rate can be underestimated by 5%. Because this error affects the data in the same direction as the evaporation effect itself in the nonportable system, the data have been corrected for this effect. When using the portable gas-exchange system the effect is in the opposite direction (because high transpiration rates occur when the humidity is low). The Eucalyptus and tomato data were not corrected for this effect. In Reno a water trap (P20/) was used so that the CO z measurements were done with dry air.

Results

Figure 1 A shows the way that the photosynthetic CO 2 assimilation rate (A) varied with measured C i at low, then high, then low transpiration rate in attached Xanthium strumarium leaves. No decline in the A - C i relationship was observed over

146

T.D. Sharkey: Transpiration effects on photosynthesis !

9

~A

I

I

I

I

I

I

I

I

~.

1.0 ? E 0.8 -5 0.6

Table 1. Apparent quantum requirement for an attached leaf of Xanthium strumarium. In a control experiment where evaporation rate was not varied, the quantum requirement was constant to within 0.1. Incident photon flux densities were used; no corrections were made for leaf absorbance. The leaf temperature was 29 ~ C Transpiration rate (mmol m -2 s - a)

Quantum requirement (mol 1Tlol- 1 C02)

4-6 8-12 5-6

16.7 20.5 18.5

"~ 0.4 o

o

0.2

A B

40

?

/

E "6 30

g ._E 20 t~

c~ o

10

/ 9

,l

I

100

I

I

I

I

I

I

l

200

300

400

500

600

700

800

Intercellular CO 2 pressure (pbar)

Fig. 3A, B. Conductance (A) and C O 2 assimilation (B) response of Xanthium strumarium to intercellular CO 2 pressure at various ambient humidities. The leaf temperature was 30 ~ C and the ambient dew point was 2 3 ~ (e), 3 ~ ( i ) and then raised to 24 ~ C (A). Evaporation rates ranged from 14.3 to 16.1 (e), 16.0 to 24.5 (m) and 13.5 to 16.0 (A) mmol m - z s -1. Arrows indicate the first point measured under each set of conditions when ambient CO 2 was 320 gbar. The first point of the first curve was determined at 9:53 and the last point at 10:58. At 11:00 the humidity was lowered, the first point of the second curve was determined at 12:08. The last point of the second curve was determined at 13:56 whereafter the humidity was raised, The third curve was begun at 14:22 and finished at 15:25

similar time periods when the humidity was kept constant (Fig. 1 B). Because C~ was directly measured, the uncertainty often associated with the estimation of C i was avoided. The A - C~ relationship of a similar plant of Xanthium with low soil water availability is shown in Fig. 2. In both cases the rate of assimilation at high C i was reduced but at low C i little or no effect of stress was observed. These data also illustrate the large variability between leaves and so comparisons between experiments are not possible.

Since it is important to know what C~ would occur in plants under normal conditions, a similar experiment was done, but with gas exchange from both surfaces. In this case, it was necessary to estimate C i. Leaf conductance and assimilation rate are shown in Fig. 3. The C i that occurred under normal ambient CO 2 pressure (Ca) is indicated by arrows. The stomatal limitation (defined by Farquhar and Sharkey, 1982, as the relative difference between A at C i = 320 gbar and A at C a = 320 gbar CO2) was 8.5% for the high-humidity case and 10.9% for the low-humidity case. In this experiment less than 1% of the total leaf area of the plant was subjected to the changes in humidity. The quantum requirement for CO 2 assimilation was increased by increasing the evaporation rate (Table 1). As was observed for COz-saturated photosynthesis (Figs. 1, 3), the quantum requirement recovered only partially when the initial conditions were reestablished. In the experiments reported above, both transpiration rate and leaf-to-air vapor-pressure difference (VPD) changed. An experiment was conducted to distinguish between the effects of transpiration and VPD. I have exploited the fact that somata can respond to humidity to such a degree that the transpiration rate is lower at high VPD than at low. In this experiment I increased V P D and also C a to keep Ci at 200 gbar. Between the effects of increased VPD and increased CO2, the conductance declined sufficiently to reduce transpiration from the leaf. This resulted in an increase in CO 2 assimilation (Table 2). Similar results were obtained with Gossypium hirsutum. Whenever the stomata responded to VPD sufficiently to keep transpiration constant, no effect of VPD on assimilation capacity was found. To test if the change in transpiration caused a measurable change in leaf water content, the fltransmissivity of the leaf was measured (Table 3). Two problems were encountered in these experiments. First, it was impossible to keep Ci exactly

T.D. Sharkey: Transpiration effects on photosynthesis Table 2. Gas exchange characteristics of an attached leaf of Xanthium strumarium. Ambient CO 2 pressure was varied to keep Ci constant. Leaf temperature was 27 ~ C. This experiment was done in R e n t , Nev. with a portable gas-exchange system. Cross sensitivity was eliminated by drying the air before analysis for CO 2 VPD (mbar)

19 32

Transpiration rate (mmol m - z

Assimilation rate (gmol m - 2

.s-~)

.s-~)

12.4 9.0

36.6 39.2

Ci

Conductance

(gbar)

(tool m z

.s-~) 193 194

0.53 0.23

Table 3. fl-Transmissivity of an attached leaf of Xanthium strumarium. With no leaf present, 62,000 cpm were measured. The uncertainty in the counts was estimated as one division on the strip chart recorder. When the petiole of the leaf was cut, a dramatic increase in counts was observed as the leaf lost water Transpiration rate (mmol m z s - 1)

Assimilation rate (gmol m - 2 s - 1)

Ci (gbar)

fl-Transmissivity (cpm)

5 8

33 27

233 217

22,000 + 150 22,000_+150

Table 4. Time course of recovery of assimilation rate from transpiration stress. Humidity was lowered at 10:55 and raised at 13:07. The data are for attached Xanthium leaves. Leaf temperature was 32 ~ C. Ca was 320 gbar Time

A ( g m o l m 2s 1)

Ci (l~bar)

Transpiration ( m m o l m - 2 s -1)

10:04 12:03 13:27 16:18

40.2 35.5 38.7 40.0

198 238 215 228

11.4 22.7 9.5 12.5

constant. U p to one half of the difference is assimilation rate in Table 3 can be accounted for by the change in C i that occurred. Second, some condensation of water vapor occurred inside the chamber, causing the measured transpiration rate to be low, especially at the low VPD. Nevertheless, this experiment showed that when the humidity around a leaf was changed sufficiently to change the assimilation rate by 10% or more, the change in t-transmission was less than 1%. Both the t-source and the Geiger tube had to be positioned very near the leaf and probably prevented substantial transpiration from the section of leaf being measured, so that if only highly localized water deficits developed, say within the areoles, they would not have been detected. The time course of recovery was investigated by measuring assimilation as a function of C~ before, during and for several hours after transpira-

147

tion stress. Data obtained at Ca=320 gbar are listed in Table 4. In this and other experiments, the reduction in assimilation response to stress was apparent as soon as measurements could be made. Measurements were, on occasion, made within 5 rain and so the effect develops in less than 5 min. Recovery also occurred more quickly than measurements could be made. Little further recovery was observed in the next 3 h. (Although the assimilation rate was the same in the last measurement as in the first, Ci was substantially higher. The full A - C i relationship showed that recovery of the mesophyll was not complete.) The next day the assimilation rate was lower than the initial recovered rate, which I interpret as resulting from the stress of being in the gas-exchange chamber overnight. The degree to which high transpiration could reduce the photosynthetic capacity of leaves in nature was tested by lowering the humidity around an attached Xanthiurn leaf at normal C, and comparing the assimilation rate and C i value with the A - C i relationship determined before stress was induced. If no effect on the photosynthetic capacity of the leaf occurred, but stomatal closure occurred, the assimilation rate would be lower but still on the original A - C i relationship. Deviation from this relationship indicates effects on photosynthesis independent of stomatal movements. The results (Fig. 4) show that a substantial difference between the initial A - - C i relationship and measurements made on stressed leaves can occur under natural conditions. The reduction in mesophyll capacity for photosynthesis adversely affects the water-use efficiency (tool H / O m o l - 1 COa) of leaves. It has been proposed that stomata "optimize" the water cost of carbon gain by keeping the marginal water cost of carbon gain, ~E/~A, constant through the day (Cowan 1977; Cowan and Farquhar 1977) Calculations of 8E/OA are often made using graphs of transpiration rate as a function of assimilation rate with conductance the implicit variable. Such a plot is shown in Fig. 5. In this experiment ! lowered the humidity in one step. The stomata closed over 30 min and as they closed, and hence the transpiration rate was reduced, the assimilation rate recovered even though C i declined. In the portable system used for this experiment, as the evaporation rate decreased, VPD increased; thus, these results confirm those in Table 2. Further points were determined by calculating transpiration rates at lower conductances and reading assimilation rates from the A - Ci plot. This line ( + , e) shows that if stomata open too widely the transpiration rate continues to rise but the assimilation rate may begin

148

T.D. Sharkey: Transpiration effects on photosynthesis I

I

}

'

I

8

A ,e

? E -~ 2O E

~

E

0

o/ ~/5.9 (15.4)

E

6

7.5 (20.3)

JJ

4

.1 (15.01

0 r,

2

E

+x

< 0

Fig. 4. The A -- C i relationship of an attached leaf of Xanthium and low evaporation rate; the flow rate of dry air was then increased so that the VPD was increased in steps. The numbers next to the data points (A) are evaporation rate (VPD). Afterwards the original conditions were reestablished and A recovered (m). Data points for the curve were determined beginning at 12: 09, the last point was determined at 12:47. As humidity was progressively decreased points were determined at 15 : 30, 16: 25, and 17 : 00. Initial conditions were reestablished and a final point was determined at 17:30

I

I

I

I

I

I

. / ~

I

I

[

I

Oqi

100

200

......

300

,

,

ll6

Fig. 6A, B. Response of C O 2 assimilation rate to intercellular CO 2 pressure in Eucalyptus microcarpa. A The leaf being measured was subjected to 5 ( . ) or 25 (m) mbar vapor pressure difference between leaf and air. The rest of the plant was kept under constant conditions. B The leaf being measured was kept under constant humidity while the rest of the plant was subjected to 21 (=) or 8 (e) mbar water-vapor pressure in a controlled-environment cabinet

10

//

,

I

'E "

I

B

Jr

15

I

Fig. 5. Relationship between evaporation rate and A when conductance is the implicit variable, e, Points determined as stomata closed after large VPD was imposed; + , theoretical points determined assuming VPD remaining high and constant and an A - - C i relationship determined after the high evaporation rates had occurred; x , theoretical points determined using the A - Ci relationship determined before the high evaporation had occurred. Ca was 700 gbar and leaf temperature was 25 ~ C. Plant material was Scrophularia desertorum

strumarium at 25~

A

,

4 8 12 C02 assimilation rate (pmol m-2 s -1)

200 400 600 Intercellular CO 2 pressure (pbar)

l

./

400 500 0 100 200 300 Intercellular C02 pressure (pbar)

to fall. In addition, I determined the relationship between transpiration rate and assimilation rate assuming the metabolism was as had been measured before any stress had occurred. This curve shows that excessive transpiration rates affect both the instantaneous water-use efficiency and the potential water-use efficiency for the rest of the day. While the preceding data document the transpiration effect on assimilation capacity in Xanthium, it is important to know whether this phenomenon is widespread in nature. In these experiments, the following criteria were used to establish that there was a transpiration effect: 1) an increase in transpiration rate of at least 20% had to decrease assimilation at some Ci by at least 10% ; 2) the assimila-

........

400

500

600

tion rate had to increase somewhat when the original conditions were restored to ensure that the e f fect was not merely a consequence of general senescence of the leaf. These criteria have been fulfilled for Larrea divaricata, a drought-hardy plant measured in Death Valley, California; for Agathis robusta, a rainforest plant; for Scrophularia desertorum, a desert plant growing near Reno, Nevada; for the wilty tomato mutant Jlacca (for more details see Bradford et al. 1983), for Gossypium hirsu; tum, Helianthus petiolaris and Eucalyptus micro -~ carpa. (These data are all essentially like Fig. 1.) The effect of humidity on the parts of the plant not enclosed in the gas-exchange chamber was also tested in Eucalyptus. The assimilation capacity ap-

T.D. Sharkey: Transpiration effects on photosynthesis

peared unaffected when the humidity around the plant parts not being measured was varied between 5 and 25 mbar (Fig. 6). Discussion

High rates of transpiration from leaves resulted in a reduced mesophyll capacity for A, the photosynthetic CO 2 assimilation. This was first observed in the halophyte Avicennia marina (Ball 1981) and now in Xanthium strumarium, Gossypium hirsutum,

Agathis robusta, Eucalyptus microcarpa, Larrea divaricata, Scrophularia desertorum, Helianthus petiolaris and Lycopersicon esculentum (this study). Forseth and Ehleringer (1983) recently found the effect also in Malvastrum rotundifolium and the effect in Agathis was also found by Langenheim et al. (in press). It is quite likely that high transpiration rates can induce water stress in most plant species. The effect of transpiration-induced stress on CO 2 assimilation is similar to the effects observed after withholding water from plants (Fig. 2). Specifically, assimilation is more affected at high than at low C~ (Fig. 1), and the quantum requirement is increased (Table 1). Schulze et al. (1972) demonstrated that stomata respond to atmospheric humidity, rather than to bulk leaf water status, by showing that a decrease in humidity could result in a higher leaf water content, but a lower transpiration rate. Table 2 and Fig. 5 show that transpiration rate, rather than humidity, is responsible for the reduced photosynthetic capacity. The most likely effect of transpiration on assimilation capacity is the development of low water potentials at the sites of evaporation. The transpiration rate through the parts of the plant not being measured had no effect on the photosynthetic capacity of a leaf in a gas-exchange chamber (Fig. 6; Schulze and Kfippers 1979). When only small sec-. tions of a leaf were subjected to changes in transpiration rate, large effects on the assimilation capacity were found. Bradford et al. (1983) found that the tomato mutantflacca developed interveinal necrosis, indicating that water transport within the leaf was not sufficient to compensate for evaporative water loss. I therefore propose that large water-potential gradients within the areoles of leaves, between the xylem and the sites of evaporation, are responsible for the decline in photosynthesis observed after transpiration stress. This also explains why no change in leaf thickness was measured in the fl-gauge experiment (Table 3), since transpiration was prevented in the small section of leaf being measured.

149

The sites of evaporation within the leaf are generally believed to be near the guard cells and for the most part on the epidermis (Meidner 1975; Cowan 1977; Tyree and Yianoulis 1980). Wylie (1943) suggested that the path of water from the xylem vessels to these sites of evaporation was via the epidermis. Recent work supports this hypothesis (Tyree and Yianoulis 1980; Sheriff 1982). Both analyses show that a water-potential gradient as high as 2 MPa can develop when 5 mmol water m - z s-1 is being lost from one surface of a leaf. Much smaller water-potential drops within the mesophyll could explain the observed transpirationinduced water-stress effects. Large water-potential gradients in rapidly transpiring leaves seem probable, considering that rapidly growing, but nontranspiring, bean hypocotyls can have a water-potential gradient of 0.2 MPa (Molz and Boyer 1978). It is generally assumed that the function of stomata is to prevent development of deleterious rates of water loss. During this investigation I often observed that stomata closed in response to increased VPD, often enough to keep the evaporation rate constant. Separating the effects of stomatal closure from the effects on the mesophyll capacity for photosynthesis is difficult and it is theoretically impossible to calculate a number which describes the stomatal contribution relative to the mesophyll contribution to a decline in photosynthesis. This is illustrated in Fig. 3. Because assimilation rate at the operational Ci was on the original A - C a relationship, stomatal closure alone could have caused the entire reduction in photosynthesis. However, if stomata had not responded to humidity at all, the assimilation rate would still have been reduced substantially by the changes in mesophyll capacity that occurred. No single number can convey this information. Rather, it is necessary, when evaluating whole-leaf responses, to know the relationship between assimilation rate and Ci, and the C i at which the plants operate. F r o m these curves (e.g., Fig. 3), we see that the mesophyll capacity is comprised of two components. At low Ci, the mesophyll capacity is insensitive to transpiration rate or VPD changes. This limitation is considered to reflect the activity of ribulose-l,5-bisphosphate carboxylase (Farquhar and von Caemmerer 1982). At high C~, the mesophyll capacity is sensitive to changes in transpiration rate. This limitation has been attributed to ribulose-l,5-bisphosphate regeneration capacity (Farquhar and von Caemmerer 1982) and to triose-phosphate utilization capacity (Sharkey and Badger 1984; Badger et al. 1984). To complete the picture, there is the stomatal limitation which is sensitive to changes in VPD.

150

Stomatal responses to VPD allow them to guard against excessive water loss. This has been interpreted as a way by which plants "optimize" their water use (Farquhar et al. 1980; Hall and Schulze 1980). The data reported here show that the stomatal responses to humidity can also guard against a more immediate problem, that of reduced mesophyll capacity for photosynthesis. Since stomata act to prevent transpiration stresses from occurring, it is not surprising that the demonstration of these stresses is difficult and sometimes requires nonphysiological conditions. Nevertheless, transpiration stess can be observed under natural conditions (Fig. 4). Because plants can suffer transpiration-induced stress, investigations of the effects of water stress on photosynthesis are important for plants in environments well supplied with water as well as plants in deserts. I thank Marilyn Ball for her persistence when those of us who knew better assured her that this effect would not occur, and Drs. M.C. Ball, G.D. Farquhar, I.R. Cowan (all of the Australian National University, Canberra) and S.D. Smith (BSC, Desert Research Institute, Reno) for their comments on this manuscript.

References Badger, M.R., Sharkey, T.D., von Caemmerer, S. (1984) The relationship between steady-state gas exchange of bean leaves and the levels of carbon reduction cycle intermediates. Planta, in press Ball, M.C. (1981) Physiology of photosynthesis in two mangrove species : responses to salinity and other environmental factors. Ph.D. thesis, Australian National University, Canberra Bradford, K.J., Sharkey, T.D., Farquhar, G.D. (1983) Gas exchange, stomatal behavior and 613C values of the flaeea tomato mutant in relation to abscisic acid. Plant Physiol. 72, 245-250 Cowan, I.R. (1977) Stomatal behaviour and environment. Adv. Bot. Res. 4, 117~28 Cowan, I.R., Farquhar, G.D. (1977) Stomatal function in relation to leaf metabolism and environment. Syrup. Soc. Exp. Biol. 31, 471-505 Farquhar, G.D., Sehulze, E.-D., Kfippers, M. (1980) Responses to humidity by stomata of Nicotiana glauca L. and Corylus avellana L. are consistent with the optimization of carbon dioxide uptake with respect to water loss. Aust. J. Plant Physiol. 7, 315-327 Farquhar, G.D., von Caemmerer, S. (1982) Modelling of photosynthetic response to environmental conditions. In: Encyclopedia of plant physiology, N.S., vol. 12B: Physiological plant ecology II: Water relations and carbon assimilation, pp. 549-587, Lange, O.L., Nobel, P.S., Osmond, C.B., Ziegler, H., eds. Springer, Berlin Heidelberg New York Earquhar, G.D., Sharkey, T.D. (1982) Stomatal conductance and photosynthesis. Annu. Rev. Plant Physiol. 33, 317 345 Field, C., Berry, J.A., Mooney, H.A. (1982) A portable system for measuring carbon dioxide and water vapour exchange of leaves. Plant Cell Environ. 5, 17%186 Forseth, I.N., Ehleringer, J.R. (1983) Ecophysiology of two

T.D. Sharkey: Transpiration effects on photosynthesis solar tracking desert winter annuals. III. Gas exchange responses to light, CO z and VPD in relation to long-term drought. Oecologia 57, 344-351 Hall, A.E., Schulze, E.-D. (1980) Stomatal response to environment and a possible interrelation between stomatal effects on transpiration and CO 2 assimilation. Plant Cell Environ. 3, 467474 Jarman, P.D. (1974) The diffusion of carbon dioxide and water vapour through stomata. J. Exp. Bot. 25, 927-936 Jones, H.G. (1973) Estimation of plant water status with the beta gauge. Agric. Meteorol. 11, 345-355 Langenheim, J.H., Osmond, C.B., Brooks, A., Ferrar, P.J. (1983) Photosynthetic responses to light in seedlings of selected Amazonian and Australian rainforest tree species. Oecologia, in press Leuning, R. (1983) Transport of gases into leaves. Plant Cell Environ. 6, 181-194 Meidner, H. (1975) Water supply, evaporation, and vapour diffusion in leaves. J. Exp. Bot. 26, 666-673 Mohanty, P., Boyer, J.S. (1976) Chloroplast response to low leaf water potentials. IV. Quantum yield is reduced. Plant Physiol. 57, 704-709 Molz, F.J., Boyer, J.S. (1978) Growth-induced water potentials in plant cells and tissues. Plant Physiol. 62, 423-429 Passioura, J.B. (1982) Water in the soil-plant-atmosphere continuum. In: Encyclopedia of plant physiology, N.S., vol. 12B: Physiological plant ecology II: Water relations and carbon assimilation, pp. 5 34, Lauge, O.L., Nobel, P.S., Osmond, C.B., Ziegler, H., eds. Springer, Berlin Heidelberg New York Schulze, E.-D., Kfippers, M. (1979) Short-term and long-term effects of plant water deficits on stomatal response to humidity in Corylus avellana L. Planta 146, 319-326 Schulze, E.-D., Lange, O.L., Buschbom, U., Kappen, L., Evenari, M. (1972) Stomatal responses to changes in humidity in plants growing in the desert. Planta 108, 259-270 Sharkey, T.D., Badger, M.R. (1982) Effects of water stress on photosynthetic electron transport, photophosphorylation and metabolite levels of Xanthium strumarium mesophyll cells. Planta 156, 199 206 Sharkey, T.D., Badger, M.R. (1984) Factors limiting photosynthesis as determined from gas exchange characteristics and metabolite pool sizes. Proc. Vlth Int. Congr. Photosyn., Brussels 1983. Nijhoff/Dr. Junk, The Hague Sharkey, T.D., Imai, K., Farquhar, G.D., Cowan, I.R. (1982) A direct confirmation of the standard method of estimating intercellular partial pressure of CO2. Plant Physiol. 69, 657-659 Sheriff, D.W. (1982) The hydraulic pathways in Nicotania glauca (Grah.) and Tradescantia virginiana (L.) leaves, and water potentials in leaf epidermes. Ann. Bot. (London) 50, 535-548 Tyree, M.T., Yianoulis, P. (1980) The site of water evaporation from sub-stomatal cavities, liquid path resistances and hydroactive stomatal closure. Ann. Bot. (London) 46, 175-193 von Caemmerer, S. (1981) On the relationship between chloroplast biochemistry and gas exchange of leaves. Ph.D. thesis, Australian National University, Canberra von Caemmerer, S., Farquhar, G.D. (1981) Some relationships between the biochemistry of photosynthesis and the gas exchange of leaves. Planta 153, 376-387 Wylie, R.B. (1943) The role of the epidermis in foliar organization and its relations to the minor venation. Am. J. Bot. 30, 273-280 Received 6 June; accepted 14 October 1983