Plant, Cell and Environment (1998) 21, 10291038

ORIGINAL ARTICLE OA 220 EN

Control of stomatal conductance by leaf water potential in

Hymenoclea salsola (T. & G.), a desert subshrub
J. COMSTOCK & M. MENCUCCINI*

Boyce Thompson Institute for Plant Research, Tower Road, Ithaca, NY 14853, USA

ABSTRACT Several environmental parameters influencing water-

loss rates, such as temperature, humidity and windspeed,
The role of leaf water potential in controlling stomatal con-
have all been shown to affect stomatal behaviour (Jarvis &
ductance (gs) was examined in the desert subshrub
Morison 1981; Ball, Woodrow & Berry 1987; Aphalo &
Hymenoclea salsola. For plants operating at high irradi-
Jarvis 1993). Uncertainty regarding proximate mecha-
ance, stomatal closure in response to high leaf-air humidity
nisms of action is particularly notable with regard to the
gradient (D) was largely reversed by soil pressurization.
humidity response (Grantz 1990). Elegant work by Mott &
Stomatal re-opening eliminated, on average, 89% of the
Parkhurst (1991) indicated that stomatal closure due to
closure normally induced by high D. Transpiration rates
short-term humidity changes was best correlated with the
(E) reached under these conditions were far higher than
actual transpirational flux rate and not the humidity gradi-
maximal rates normally observed at any point of the D
ent between ambient air and leaf. Monteith (1995) re-eval-
response. In situ stem psychrometry indicated that water
uated a large number of humidity-response data from the
flux at all times conformed to a simple Ohms-law analogy.
literature and showed that most data are consistent with a
Under conditions of high D, E increased substantially in
linear negative relationship between gs and E. This result,
response to soil pressurization. Stomatal regulation did not
however, can be seen as consistent with either hydraulic
constrain E during this treatment, but did result in nearly
signalling based on water potential gradients linearly
constant minimum leaf water potentials.
related to E in an Ohms law analogy, or chemical sig-
nalling dependent on rates of xylem sap influx.
Key-words: epi-stomatal transpiration; hydraulic conductance;
There has been tremendous success in recent years in
rootshoot communication.
demonstrating a role for chemical signalling between root
and shoot with major influences on stomatal opening
INTRODUCTION (Dodd, Stikic & Davies 1996; Jia, Zhang & Zhang 1996;
Schurr & Schulze 1996; Tardieu 1996; Tardieu, Lafarge &
Despite decades of fruitful study, stomatal regulation in
Simonneau 1996). Such signals may reflect root water
response to water stress is still a controversial issue. For
potential and/or other aspects of the soil environment
some time, it has been recognized that stomatal closure
affecting root physiological status. The degree to which
results in a limiting resistance controlling the flow of
root-based chemical signals can explain stomatal regula-
water through the plant (van den Honert 1948). Since leaf
tion with regard to plant water status, and the degree to
water potential (l) is generally the lowest potential in the
which they may interact with separate signals generated by
plant, and is physically associated with the stomatal
leaf water status (Saliendra, Sperry & Comstock 1995;
complex, the possible role of a negative feedback loop
Fuchs & Livingston 1996), are unresolved.
between gs and l was recognized almost immediately
The use of soil pressurization (Passioura 1980; Passioura
(Cowan 1965). However, the degree to which feedback
& Tanner 1985) can help distinguish between between
models are consistent with observed behaviour remains
root-based chemical messages and hydraulic signals trans-
controversial to this day. In addition to continuous feed-
duced in the shoot. If water flux is not altered (i.e. no stom-
back models, water-potential set-points have also been
atal response), pressurizing the water in the soil raises the
proposed, where rapid stomatal closure is associated with
water potential of the shoot. Although technically the water
discrete thresholds. Such behaviour can result in a very
potential of the root is also raised, from a functional stand-
conservative minimum water potential under a wide range
point there is little effect on water-relations parameters
of environmental conditions.
such as root cell turgor or relative water content. This is
because both air and liquid phases are equally elevated in
Correspondence: Jonathan Comstock. Fax: 607 254 1242; e-mail: pressure (it is as if the definition of total = 0 MPa in the
JPC8@Cornell.edu root chamber varies to always reflect actual air pressure
*Present address: Maurizio Mencuccini, Institute of Ecology and rather than the constant 01 MPa of the standard defini-
Resource Management, University of Edinburgh, Darwin Building, tion). Root water relations are affected only indirectly and
Mayfield Road, Edinburgh EH9 3JU, UK. only if stomatal responses alter the flux of water through
1998 Blackwell Science Ltd 1029
1030 J. Comstock and M. Mencuccini

the plant, thus changing the magnitude of associated pres- levels. Ingoing humidity and CO2 were measured with an
sure gradients through plant tissues. In the shoot, however, IRGA (model LI-6262, LICOR Instruments, Lincoln, NE,
the pressurized water moves up the xylem into a region USA), outgoing humidity with a dewpoint monitor (model
with the air-spaces at normal atmospheric pressure. Here 2000, EG & G Moisture and Humidity Instruments,
the elevated water potential affects all water-relations Burlington, Massachusetts, USA), and CO2 differential
parameters strongly. Because of the water potential (pres- with an LI-6252 IRGA. Leaf temperature was taken as the
sure) drop associated with the movement of water from the average of seven type-E thermocouples inserted into
soil to the shoot, the water potential of the shoot may still leaves in different parts of the cuvette. All sensors were
be negative, but less so than would normally be required to scanned every 3 s. Gas exchange calculations were made
support the given flux. This permits an experimental following von Caemmerer & Farquhar (1981) and stomatal
approach in which the effects of a large transpiration rate ratios treated as described in Comstock & Ehleringer
per se can be separated from the water-potential gradient, (1993). Photosynthetic surface area was measured with a
particularly the leaf water potential normally associated leaf area meter (model LI-3200, LI-COR Instruments,
with it, in order to reveal which is more important in influ- Lincoln, NE, USA) calibrated with a paper comb
encing stomatal regulation. Further, since it is water-stress (Comstock & Ehleringer 1990).
in the shoot and not the root which is relieved by soil pres- The whole-plant photosynthesis cuvette was constructed
surization, enhanced stomatal opening under these condi- out of acrylic plastic and lined with Teflon film (Fig. 1). H.
tions would suggest a signal transduction and response to salsola has positive net photosynthesis in both leaves and
water potential in the leaves, not just transduction of root young twigs (Comstock & Ehleringer 1988). Both organ
water potential and long-range chemical signalling via the types are cylindrical with a diameter of 12 mm. Wind
xylem transpiration stream. speed in the cuvette was 05 m s1, and boundary layer
The current study was undertaken to examine (1) the rel- conductance in the cuvette was determined to be
ative importance of l versus E in controlling short-term 25 mol m2 s1 using 002 m2 wet cotton string as an evapo-
changes in stomatal aperture, (2) the relationships between rating surface with similar dimensions to the photosynthetic
transient conditions of water potential and stomatal con- organs of H. salsola (average surface area during measure-
ductance following a perturbation, and (3) whether the ments 01 m2). The shoot cuvette rested on top of a large
response of gs to l is more consistent with set-point or pressure chamber constructed at the chemistry machine
continuous feedback behaviour. shop at the University of Utah. During experimental mea-
surements, the pot containing the undisturbed root system
in soil was placed into the pressure chamber and an air-tight
MATERIALS AND METHODS seal formed around the base of the stem with neoprene gas-
Plant material and propagation kets and a steel compression plate (Fig. 1). The bottom of
the photosynthesis cuvette was a flexible Teflon film which
Hymenoclea salsola (T. & G.), a subshrub of the Mojave was fastened around the base of the stem. The completed
and Sonoran deserts of western North America, was grown installation resulted in a fully intact plant with the root sys-
from seed in the greenhouse at the Boyce Thompson tem in a pressure chamber and the shoot in a gas exchange
Institute for Plant Research in Ithaca, New York (elevation cuvette. This was similar in principle to the design used by
300 m). The plants were grown in 30 dm3 pots in a soil mix Passioura (1980) and Passioura & Tanner (1985), but rather
of 3: 1: 1 fritted clay (Turface): silica sand: pasteurized larger in scale permitting 30 dm3 root volumes and canopies
topsoil, and were watered daily with nutrient solution con- up to 05 m in diameter. A ceiling-mounted hoist was used
taining 55: 18: 55 p.p.m. N:P:K from Peters Excel. The to lift the large pots into the root pressure chamber.
photoperiod was 12 h, from combined artificial (an alter-
nating bank of 1000 Watt hi-pressure Na vapour, 1000 W
Super Metal Halide, and 150 W incandescent floodlights) Water potentials and pressure gradients
and natural lighting with a total irradiance (400700 nm) The plants were watered thoroughly with distilled water
of 44 mol m2 d1. Day/night conditions were 30/20 C, just prior to being enclosed in the root pressure chamber.
45/80 RH, and 375/390 mol mol1 mean CO2. The design of the root pressure chamber allowed for addi-
Measurements were made when plants were 4 months old, tional watering through ports in the lid even after the plant
and the main stems had extensive secondary growth. was fully installed and sealed. This was generally found to
be unnecessary, however, because of the large soil volume
relative to plant size. Watering during experiments did not
Gas exchange and root pressurization chamber
change either gas exchange behaviour or measured water
Gas exchange measurements were made with a single pass potentials. Soil water potentials were assumed to remain at
system and a whole-plant cuvette. Flow rates were mea- essentially 0 MPa throughout the measurements. The pres-
sured with mass flow controllers (model 362, Tylan sure needed in the root chamber to bring a wet film to a cut
General, Torrance, CA, USA) with a maximum flow to the surface on a terminal twig was measured and taken to be
chamber of 400 dm3 min1. Humidity and CO2 were both equal to the integrated water potential gradient for water
scrubbed out of the flow and then added back to controlled transport from soil to foliage.
1998 Blackwell Science Ltd, Plant, Cell and Environment, 21, 10291038
 Control of stomatal conductance by leaf water potential 1031

Figure 1. Schematic of the whole-plant photosynthesis cuvette with root pressure chamber. The cuvette was constructed of acrylic plastic
lined with Teflon film. Internal mixing fans generated air movement 10100 times the rate of air flow through the cuvette for gas exchange
measurements. Temperature control was achieved by both water channels in the acrylic chamber walls and small radiators in the internal air
flow pathway. The root pressure chamber was made of carbon steel and rated for pressures up to 40 MPa. The pressure chamber lid and the
compression plate were formed by two steel half-circles which could be fitted around the intact plant stem to compress a neoprene gasket.

Plant water potentials were also measured by two other through the stem combined with soil versus shoot tempera-
techniques. Transpiring leaf water potentials were taken by ture gradients. Such temperature gradients were suppressed
opening the lid of the cuvette, excising a leaf or small twig, by growing the plants in soil with buried copper tubing
and immediately measuring water potential using a connected to a circulating water bath, and controlling soil
Scholander pressure-chamber. Tissue samples were temperature to 280 C to eliminate temperature gradients
obtained within 30 s of opening the cuvette. Cut samples at the psychrometer during gas exchange measurements.
were held in slightly dampened paper towels while being This was not an abnormal soil temperature for this warm-
transferred to the Scholander chamber, which likewise held desert species. Empirical tests with calibration standards
damp towels during the measurement. indicated that the in situ psychrometers could make
Stem xylem water potentials were measured using in situ repeated readings of water potential as often as once per
temperature-compensated stem psychrometers (stem minute without loss of accuracy. Manufacturers specifica-
hygrometer, Plant Water Status Instruments, Inc, Guelph, tions indicate a liquid-vapour equilibration rate for the
Ontario, Canada) which were placed on the lower stems small psychrometric air-space of about 45 s. This made the
and monitored throughout an experiment. Although the in stem psychrometers ideal for monitoring rapidly changing
situ stem psychrometer was theoretically compensated for water potentials in response to soil pressurization and fol-
temperature gradients at the measurement point (Dixon & lowing rapid stomatal movements. The psychrometers
Tyree 1984), this correction was reliable only for very were read using a Dewpoint microvoltmeter (model HT33,
small gradients. The psychrometers were prone to exces- Wescor Inc, Logan, UT, USA). Three psychrometers were
sive temperature gradients due to transpirational water flux installed on the same plant and stem water potentials
1998 Blackwell Science Ltd, Plant, Cell and Environment, 21, 10291038
1032 J. Comstock and M. Mencuccini

reported represent average values of all three psychrome- RESULTS

ters. Total variation among the three psychrometer readings
A brief test was performed of how Scholander and stem
was always less than 0075 MPa with a consistent rank-
psychrometer readings compare on H. Salsola. A major
ing (i.e. most variation was not random but associated with
leafy branch was severed and two stem psychrometers
consistent position and possible installation effects).
were installed. The branch was then placed in a plastic bag
and allowed to equilibrate overnight. Measured water
Stomatal response to raised leaf water potential potentials were 254 014 and 258 002 MPa (mean
Twenty plants were tested to see if pressurizing the soil SD), for four leaves measured with the Scholander cham-
airspaces and raising leaf water potentials influenced the ber and the two stem psychrometers, respectively.
level of stomatal opening. Plants were exposed first to
high irradiance (18 mmol m2 s1 at chamber mid-height),
optimal temperatures (300 C), ambient CO2 of 350 mol
mol1, and low leaf-air humidity gradients to get a mea-
sure of maximal stomatal conductance. Plants were then
stressed by exposure to increasing D until substantial
stomatal closure was observed. The soil compartment was
then pressurized until a terminal twig was brought to the
balance point and formed a wet surface on exposed xylem.
During these measurements it was noticed that the balanc-
ing point was not stable, but appeared to gradually
increase with time, prompting a special experiment with
one plant.

Stability of hydraulic conductance

Capacitance in desert shrubs such as H. salsola is small
(Nobel & Jordan 1983), and steady-state water fluxes are
achieved within a few minutes under stable environmental
conditions and constant stomatal aperture. Water fluxes in
the xylem were taken to be equal to transpiration rates after
E had been stable for 10 min or more. The hydraulic con-
ductance of H. salsola was defined as the slope of water
flux versus integrated water-potential gradient (Passioura
1988). To test whether pressurization procedures were hav-
ing an adverse effect on the transport system, a selected
plant was installed in the soil pressure chamber and whole
plant cuvette as above, but changed repeatedly from low to
high D in repeated cycles for 14 h while maintaining a con-
tinuous balancing pressure. This permitted an assessment
of whether the slope or intercept of the relationship of E
versus balance pressure was changing over the course of
the day.

Dynamics of plant water potential during re-

opening Figure 2. Response of stomata of H. salsola to increasing leaf-air
humidity gradient (D), and soil pressurization to improve shoot
To better observe the dynamic behaviour of water potential
water status during whole-plant gas exchange measurements. Leaf
during soil pressurization, another plant was chosen for temperature was 300 C, Irradiance (400700 nm) 18 mmol m2
intensive study. The plant was equipped with in situ stem s1 and ambient CO2 350 mol mol1. (a) Transpiration (b)
psychrometers as described above, and was initially stomatal conductance, and (c) water potentials of stems (crosses)
exposed to high D in the same fashion as all plants tested measured by in situ psychrometry on the main trunk near ground
for stomatal response to l. At this time (1030 h) the root level, and leaf water potential (circles) measured by harvesting
chamber was still at normal ambient pressure. The pressure small twigs for Scholander pressure-chamber determination. Stem
values are the means of three psychrometers and leaf values are
in the root chamber was then raised and lowered in a series based on three replicate samples at each point. Standard errors of
of discrete increments. A full balancing pressure, which replicates were usually less than 005 MPa and are not shown.
would have flooded the psychrometers, was approached Filled symbols in all panels refer to data collected while the soil
but never fully attained. compartment was pressurized to 10 MPa.

1998 Blackwell Science Ltd, Plant, Cell and Environment, 21, 10291038
 Control of stomatal conductance by leaf water potential 1033

Stomata closed in a continuous fashion during increasing The plant was slightly more sensitive to D than average
D, with the ln(gs) versus D being strongly linear (Fig. 2b, (Fig. 2 versus Table 1 mean percentage closure), but
Comstock & Ehleringer 1993). At high D, stomatal closure showed a very typical stomatal re-opening response to soil
was sometimes great enough to cause a maximum in E pressurization. Stomatal closure resulted in a value of gs at
(e.g. Fig. 2a), although gradually increasing E was some- D = 35 only 51% of the initial value at D = 10 mmol mol1.
what more common. Pressurizing the root chamber caused The in situ stem psychrometers permitted frequent moni-
consistent and substantial stomatal re-opening, with soil toring (c. every 24 min) of water potential changes, and
pressurization reversing an average 89% of the closure oth- therefore facilitated a more detailed analysis of the tempo-
erwise caused by high D (Table 1). Since soil pressuriza- ral response of stomatal conductance to water potential
tion was performed while maintaining constant leaf fluctuations (Figs 5a & 6). Several features illustrated were
temperature and D, E under pressurized conditions was common to the response to pressurization seen for all the
also substantially increased over normal values (Fig. 2a). plants listed in Table 1. Initial stomatal responses were
It was observed in most of the plants studied that the bal- always opposite to final responses, and this was interpreted
ancing pressure needed during the stomatal re-opening to be a hydropassive response. This kind of hydropassive
phase of the experiment was not stable but slowly response has been attributed to a passive increase in turgor
increased with time. To determine whether the hydraulic by all epidermal cells as bulk water potential rises, and a
conductance was being affected by pressurization, the mechanical advantage enjoyed by the subsidiary cells
slope and intercept of balancing pressure versus E were which squeeze the stomatal pores shut. The hydropassive
measured repeatedly on one plant over 14 h (Fig. 3). Under response always occurred about 5 min after initial pressur-
constant high light and temperature, the offset increased at ization, but was rather variable in magnitude (Fig. 5a).
a steady, linear rate which was unaffected by additional
soil watering, and therefore unrelated to soil drying. The
hydraulic conductance in the restricted sense (slope)
remained relatively constant throughout the day. Non-zero,
variable offsets in such plots are not well understood, but
may be related to the build-up of solutes in the root cortex
which are excluded from the transpiration stream during
symplastic portions of the flow pathway (Stirzaker &
Passioura 1996).
One plant was selected for more detailed measurements
of the water potential dynamics during pressurization (Figs
2c, 5 & 6). Measurements on this plant were spread over
three successive days. A single linear relationship between
water potential gradient and E was seen across all three
days, and included both unpressurized and pressurized
points (Fig. 4). Approximately half (49%) of the total resis-
tance to water flow from the soil to the sites of evaporation Figure 3. Diurnal effects on the hydraulic conductance of H.
in leaves and twigs was below ground in the root system. salsola. The same plant was maintained under continuous
This included both resistance to axial flow in the xylem, saturating light for 14 h. Leaf temperature was 300 C and
and also radial flow through root cortex. For steady-state ambient CO2 350 mol mol1. The leaf-air humidity gradient (D)
conditions after stomatal re-opening, most of the applied was varied repeatedly between 15 and 35 mmol mol1 throughout
the experiment, and, for all points, soil pressurization was applied
force from soil pressurization was dissipated in increased E until a cut twig began to bleed xylem sap and the total pressure
such that for 10 MPa of soil pressurization applied, stem gradient through the plant could be measured. The soil was
and leaf water potentials had increased by only 035 and watered twice during the course of the day to ensure that no water
02 MPa, respectively (Fig. 2c). deficit could develop.

Table 1. The effect of soil pressurization on stomatal opening in H. salsola. The humidity response of stomatal conductance (gs) to gas
exchange was measured on 20 individuals. All measurements are on whole plants, and the photosynthetic surface area was a mixture of leaf
and twig organs. In all cases, substantial stomatal closure was observed as D was changed from low (10) to high (35 mmol mol1) values.
Pressurization of the soil was applied until cut twigs in the upper canopy formed a wet cut surface. This treatment, raising the water potential
of the shoot, consistently resulted in stomatal re-opening

g @ low D (mol m2 s1) g @ high D (mol m2 s1) g @ high D (mol m2 s1) Recovery (%)

Pressurized soil No No Yes

Mean 075 050 069 888
SE 0034 0031 0035 62

1998 Blackwell Science Ltd, Plant, Cell and Environment, 21, 10291038
1034 J. Comstock and M. Mencuccini

successive days (Fig. 2), and (3) levels of experimental

pressurization of the soil. Due to the difficulty of obtain-
ing samples from the sealed cuvette, leaf water potential
was verified only three times during soil pressurization.
During four successive pressurization steps, essentially
identical behaviour was seen as that discussed above, and
stomatal conductance reached very high values despite the
constant high D. While stem water potentials stabilized at
slightly higher water potentials after each pressurization
cycle, stomatal re-opening ceased when leaf water poten-
tial had returned almost exactly to its initial, prepressuriza-
tion value. After three pressurization cycles, the stomata
Figure 4. Whole-plant hydraulic resistance measured on the same were at 82% of their low D maximal value, and the open-
plant on three successive days. Open symbols are stem water ing response began to saturate. Further soil pressurization
potential gradients measured by in situ psychrometry on the main caused only small stomatal responses and stable elevation
trunk near ground level, and closed symbols are water potential
gradients from soil to leaf measured by harvesting small twigs for
Scholander pressure-chamber determination. soil is assumed to be
0 or equal to the added soil pressure when calculating the total
gradient. Stem values are the means of three psychrometers and
leaf values based on three replicate samples at each point. Standard
errors of replicates were usually less than 005 MPa and are not
shown. Line symbols represent points taken while pressurizing the
soil compartment and represent the sum of measured water
potential applied pressure. Data are adjusted for diurnal intercept
drift (Figure 3). The lines shown are fitted to the unpressurized
points only and have slopes of 0074 and 0036 MPa m2 s1
mmol1 for leaves and stems, respectively.

Within 10 min of pressurization, a strong stomatal opening

response had begun which continued for about 30 min. For
the special plant fitted with stem psychrometers, the xylem
water potential lagged behind soil pressurization by only 2
or 3 min (Fig. 6). Pressurizing the soil has little effect on E
unless there is a stomatal response. Consequently, soil
pressurization initially caused a rise in water potential
throughout the shoot. By 5 min after soil pressurization,
the measured water potential of the stem xylem had risen
018 MPa, which accounted for 90% of the applied pres-
sure. As stomata opened during the next 30 min, the water
potential dropped back towards its original value due to the
higher E and greater integrated water potential gradient
needed to supply water to the shoot. The gas exchange sys-
tem was not specifically calibrated for transient conditions. Figure 5. Demonstration of reversible stomatal response to shoot
At the high flow rates used during this experiment, the water potential manipulation. (a) Water potentials of soil, stem, and
cuvette-volume signal-buffering would have required leaf. Water potential of the stem was measured by in situ
about 3 min to reach 99% of final steady-state values. Due psychrometry on the main trunk near ground level, and leaf water
potential was measured by harvesting small twigs for Scholander
to this, the full magnitude of the brief hydropassive pressure-chamber determination. The difference between the
response and the subsequent speed of stomatal re-opening pressurized soil line and either plant tissue represents the driving
during the first minutes are likely to be slightly underesti- force for liquid-phase water transport. Only one psychrometer was
mated. Pressurized points fell on the same relationship read during periods of rapid change, but all three were read at each
between E and integrated water potential gradient as point where stomatal aperture became constant at a new value.
unpressurized points (Fig. 4). Leaf values are means of three replicates. (b) Stomatal conductance
Several steps in soil pressure were made, both increas- over the same time period. Leaf temperature was 300C, irradiance
(400700 nm) 18 mmol m2 s1 and ambient CO2 350 mol mol1.
ing and decreasing (Fig. 5a). Measured water potential Simulated water potentials for both stem and leaf positions are
changes during this experiment were compared to simula- based on current measurements of transpiration (E) and previously
tions based on (1) currently changing transpiration rates, determined hydraulic resistances (r) (Fig. 2) and offsets (Fig. 3)
(2) hydraulic conductance previously measured on three according to = Er offset + soil pressure.

1998 Blackwell Science Ltd, Plant, Cell and Environment, 21, 10291038
 Control of stomatal conductance by leaf water potential 1035

There is currently great interest in the role(s) of chemi-

cal signals produced by the root in controlling stomatal
regulation. Several herbaceous species have been studied
in regard to long-term soil drying, and it was found that
soil pressurization did not reverse the effects of drought
(Gollan, Passioura & Munns 1986; Schurr, Gollan &
Schulze 1992). Fuchs & Livingston (1996) did see a rever-
sal of both D and drought-induced stomatal closure in
Pseudotsuga menziesii and Alnus rubra, and suggested that
woody plants may show different responses than herba-
ceous plants. H. Salsola is a semi-woody perennial evolved
from more herbaceous ancestors. In stature it more closely
resembles the herbs, and has a small xylem volume which
will turn over every few minutes at high transpiration rates.
Stomatal closure in response to D was clearly reversible by
Figure 6. Detailed time course of one cycle of soil pressurization
from Fig. 5. Time 0 represents the time at which soil pressurization soil pressurization in this species.
was begun. Pressurization speed was 0067 MPa min1, and going While it is clear that ABA is at least one of the important
from 0 to full pressure (02 MPa) required 3 min. regulators of long-term stomatal responses to drought, it is
more uncertain what role it plays in rapid diurnal responses
to changing conditions of D, wind and temperature. Recent
of both stem and leaf water potential. The final episode work indicates that diurnal increases in transpiration rate
involved a partial depressurization of the root chamber. All may be associated with increased ABA flux from roots to
events occurred in reverse, including an initial hydro- shoots (Jarvis & Davies 1997), but the processing of the
passive stage opposite to the final adjustment. In this case, ABA signal by the leaf mesophyll is extremely complex
stomatal closure again stabilized the leaf water potential and is still imperfectly understood (Trejo, Clephan &
near its original value. Davies 1995; Wilkinson & Davies 1997). Increasing D
causes increased E and lowered water potentials through-
out the plant, including the root. Increased ABA flux, in
DISCUSSION
response to the depression of root water status, rather than
The transient water potential rise following pressurization decreased leaf water potential under these conditions could
(Figs 5 & 6) usually accounted for about 90% of the added be hypothesized to account for the stomatal closure in
soil pressure. Had there been no stomatal re-opening, response to D. Under the soil pressurization treatment,
100% of the applied pressure should have been evident in however, E reached levels almost double those exhibited
increased shoot water potential, but, given the speed of under normal conditions. Root water status and relative
stomatal response, plant capacitance effects were likely water content would have been lowered further by this
affecting the maximum value. Throughout the D response
and pressurization cycles, water potential and E at steady
state were correlated in a manner that conforms to a simple
Ohms law analogy consistent with a cohesion-tension
mechanism of water transport in the xylem (Tyree 1997).
The stomatal re-opening in response to soil pressuriza-
tion strongly supports a mechanistic interpretation based
on a response to leaf water potential, and is consistent with
other recent reports (Saliendra et al. 1995; Fuchs &
Livingston 1996). Stomatal response to the perturbation of
soil pressurization resulted in abnormally high transpira-
tion rates (Fig. 2a), while bulk leaf water potential
appeared held to a highly conserved minimum value. The Figure 7. Theoretical relationship of soil pressurization to a
data collected at high D (Fig. 5) could be consistent with a putative set-point for minimum leaf water potential. It is assumed
simple threshold model (Fig. 7), where stomatal closure is that pressurization has no effect on k, the hydraulic conductance
triggered as leaf water potential reaches a critical stress (see Figure 2). The plant is initially assumed to be operating at point
level. Such a model has a clearly defined min set-point, A, where high transpiration rates have already caused substantial
and l exhibits partial homeostatic behaviour when E is stomatal closure to avoid crossing the min line. This would be
great enough to trigger stomatal closure. However, such a analogous to the conditions at the beginning of Figs 5 and 6 where
high D has reduced g by more than 50% and produced a maximum
model does not explain the continuous response of gs to D E (Fig. 4). Pressurization of the soil would effectively shift the
(Fig. 2). Further, the near-full reversibility of stomatal clo- intercept of the E versus water potential line to left. This would
sure in response to soil pressurization argues against sepa- result in a movement from point A to point B. Stomatal opening
rate D and min set-point responses. then occurs until point C is reached as leaf approaches min.

1998 Blackwell Science Ltd, Plant, Cell and Environment, 21, 10291038
1036 J. Comstock and M. Mencuccini

increased flux. Thus ABA production and transport under

soil pressurization treatments, though not measured in
these experiments, should logically have reached maxi-
mal levels while stomata were showing an opening
response. This indicates that leaf water potential must be
invoked as an important signal independent from root
water status. The fact that the stomatal closure is
reversible by manipulating leaf water potential also
weighs against the hypothesis that increased ABA flux is
primarily responsible for closure in response to D. A
minor role of increased ABA flux at high E might be con-
sistent with the fact that root pressurization could only
reverse an average of 89% of the response to D. As
recently argued by Tardieu (1996), the conditions under
which leaf water potential is most unambiguously impor-
tant are precisely those under which it is least varying and
exhibits an apparent homeostatic behaviour. Whether the
response to leaf water potential is mechanistically part of
Figure 8. A stomatal closure model based on sensitivity to
a general ABA signal transduction system in which leaf water potential at a cuticular site which is affected by both bulk
water potential affects ABA processing in the mesophyll leaf transpiration and separate cuticular transpiration. Equation 1
or response to ABA at the guard cell, or whether it repre- was used, and a multiple regression run on just the points from
sents an entirely separate signal transduction system, is the D response.
still unclear.
Several models based on feedback mechanisms and not
homeostatic in nature can nonetheless mimic an homeo- Farquhar (1978). Attempts to validate a mechanism based
static bulk leaf water potential behaviour over a restricted on separate stomatal and cuticular fluxes have been
range of conditions. An attempt was made to model the inconclusive (Sheriff 1984; Cowan 1994; Schulze 1994).
water potential response in Figs 2 and 5 based on epistom- Most recently it has been recognized that diurnal effects
atal transpiration and a linear stomatal closure response to which contribute to low stomatal conductances later in
the water potential of a restricted cuticular site: the day may have strongly influenced most observations
of the humidity response (Franks, Cowan & Farquhar
gc
(
gs = gmax C leaf Ds
kc , ) (1) 1997). Similarly, leaf cuticle has been reported to change
its permeability to water vapour in response to changing
ambient humidity (Schnherr & Schmidt 1979;
where C is the linear coefficient for stomatal closure in Kersteins 1996; Boyer, Wong & Farquhar 1997; Hoad,
response to decreasing water potential of the cuticular Grace & Jeffree 1997) in a way that is sometimes only
sites, gc/kc is the ratio of cuticular conductance to water very slowly reversible. If the model expressed by Eqn 1 is
vapour and the liquid phase resistance between the bulk altered to assume that gc is not constant, but rather is a
leaf water pool and the cuticular evaporative sites, and Ds linear, decreasing function of Ds, the fit for pressurized
is the leaf-air humidity gradient measured at the outer leaf points is improved, though still imperfect (not shown). H.
surface. A resistance catena for this configuration was pre- salsola was not amenable to direct measurements of
sented in Jarvis & Morison (1981) and Sheriff (1984). cuticular conductance to verify assumptions regarding
Although only one signal was postulated to induce stom- the behaviour of gc. The assumption of a constant value
atal closure, Eqn 1 partitions the closure response into sep- for kc may also be suspect. kc could have changed if soil
arate bulk leaf water potential and Ds components that pressurization resulted in altered leaf water contents, or
represent a water potential drop in the common flow path- changes in the uniformity of stomatal opening across the
way of all transpired water (l soil), and the additional leaf surface. Nor is there sufficient data in this study to
drop along the further pathway to the cuticular site (c evaluate interactions of water potential with changing
l), respectively. Pressurization of the soil compartment intercellular CO2.
directly affected the value of l, but (gc/kc) Ds was affected In conclusion, short-term stomatal closure under high D
only indirectly as stomatal opening altered humidity in the can be largely reversed in H. salsola by manipulating shoot
leaf boundary layer. Running a multiple linear regression water potential. However, numerous uncertainties regard-
of gs on l and Ds provides an excellent fit to the data of ing the constancy of leaf properties such as cuticular con-
the D response, but when applied to the soil pressurization ductance or the uniformity of stomatal opening across the
points it severely underestimates the observed re-opening leaf preclude the evaluation of simple mechanistic rules
response (Fig. 8). regarding the stomatal response to water potential. A com-
Equation 1 is an hydraulically explicit expression mon mechanism which relates equally well to both mild
which conforms to the feed-forward theory formalized by and severe stress conditions is not yet obvious. None the
1998 Blackwell Science Ltd, Plant, Cell and Environment, 21, 10291038
 Control of stomatal conductance by leaf water potential 1037

less, it is increasingly clear that for many plants, maximum forward response of stomata to air vapour pressure deficit: infor-
transpiration rates are limited by water potential responses mation revealed by different experimental procedures with two
that mimic a set-point-like behaviour under those prevailing rainforest trees. Plant, Cell and Environment 20, 142145.
Fuchs E.E. & Livingston N.J. (1996) Hydraulic control of stomatal
conditions. Cavitation vulnerability previously measured conductance in Douglas fir [Pseudotsuga menziesii (Mirb)
on H. salsola (Mencuccini & Comstock 1997) indicates Franco] and alder [Alnus rubra (Bong)] seedlings. Plant, Cell
that at high D, E was sufficiently high for cavitation in the and Environment 19, 10911098.
proximal root (closest to the stem where xylem water Gollan T., Passioura J.B. & Munns R. (1986) Soil water status
potential was measured) to approach 50%. Such a high affects the stomatal conductance of fully turgid wheat and sun-
level of cavitation indicates that rapid responses to water flower leaves. Australian Journal of Plant Physiology 13,
459464.
potential fluctuations may often be essential to avoid catas-
Grantz D.A. (1990) Plant response to atmospheric humidity. Plant,
trophic collapse in the water transport system (Tyree & Cell and Environment 13, 667679.
Sperry 1988; Sperry et al. 1998). Hoad S.P., Grace J. & Jeffree C.E. (1997) Humidity response of
cuticular conductance of beech (Fagus sylvatica L.) leaf disks
maintained at high relative water content. Journal of
ACKNOWLEDGMENTS Experimental Botany 48, 19691975.
van den Honert T.H. (1948) Water transport in plants as a catenary
We wish to thank John Passioura for helpful discussions process. Discussions of the Faraday Society 3, 146153.
while designing the root pressurization chamber, John Jarvis A.J. & Davies W.J. (1997) Whole-plant ABA flux and the
Sperry for advice regarding hydraulic measurements and regulation of water loss in Cedrella odorata. Plant, Cell and
use of the in situ stem psychrometers, and John Laurence Environment 20, 521527.
for comments on the manuscript. The work was supported Jarvis P.G. & Morison J.I.L. (1981) The control of transpiration and
by NSF grants IBN 9119560 and IBN 9496093. photosynthesis by the stomata. In Stomatal Physiology (eds P.G.
Jarvis & T.A. Mansfield), pp. 247279. Cambridge University
Press, New York.
Jia W., Zhang J. & Zhang D.-P. (1996) Metabolism of xylem-deliv-
REFERENCES
ered ABA in relation to ABA flux and concentration in leaves of
Aphalo P.J. & Jarvis P.G. (1993) The boundary layer and the appar- maize and Commelina communis. Journal of Experimental
ent responses of stomatal conductance to wind speed and to the Botany 47, 10851091.
mole fractions of CO2 and water vapour in the air. Plant, Cell Kersteins G. (1996) Cuticular water permeability and its physiolog-
and Environment 16, 771783. ical significance. Journal of Experimental Botany 47,
Ball J.T., Woodrow I.E. & Berry J.A. (1987) A model predicting 18131832.
stomatal conductance and its contribution to the control of photo- Mencuccini M. & Comstock J. (1997) Vulnerability to cavitation in
synthesis under different environmental conditions. Progress in populations of two desert species, Hymenoclea salsola and
Photosynthetic Research 4, 221228. Ambrosia dumosa, from different climatic regions. Journal of
Boyer J.S., Wong S.C. & Farquhar G.D. (1997) CO2 and water Experimental Botany 48, 13231334.
vapor exchange across leaf cuticle (epidermis) at various water Monteith J.L. (1995) A reinterpretation of stomatal responses to
potentials. Plant Physiology 114, 185191. humidity. Plant, Cell and Environment 18, 357364.
von Caemmerer S. & Farquhar G.D. (1981) Some relationships Mott K.A. & Parkhurst D.F. (1991) Stomatal responses to humidity
between the biochemistry of photosynthesis and the gas in air and helox. Plant, Cell and Environment 14, 509515.
exchange of leaves. Planta 153, 376387. Nobel P.S. & Jordan P.W. (1983) Transpiration stream of desert
Comstock J.P. & Ehleringer J.R. (1988) Contrasting photosynthetic species: resistances and capacitances for a C3, a C4, and a CAM
behavior in leaves and twigs of Hymenoclea salsola, a green- plant. Journal of Experimental Botany 34, 13791391.
twigged, warm desert shrub. American Journal of Botany 75, Passioura J.B. (1980) The transport of water from soil to shoot in
13601370. wheat seedlings. Journal of Experimental Botany 31, 333345.
Comstock J. & Ehleringer J. (1990) Effect of variations in leaf size Passioura J.B. (1988) Water transport in and to roots. Annual
on morphology and photosynthetic rate of twigs. Functional Review of Plant Physiology and Plant Molecular Biology 39,
Ecology 4, 209221. 245265.
Comstock J. & Ehleringer J. (1993) Stomatal response to humidity Passioura J.B. & Tanner C.B. (1985) Oscillations in apparent
in common bean (Phaseolus vulgaris): Implications for maxi- hydraulic conductance of cotton plants. Australian Journal of
mum transpiration rate, water-use efficiency and productivity. Plant Physiology 12, 455461.
Australian Journal of Plant Physiology 20, 669691. Saliendra N.Z., Sperry J.S. & Comstock J.P. (1995) Influence of
Cowan I.R. (1965) Transport of water in the soil-plant-atmosphere leaf water status on stomatal response to humidity, hydraulic
system. Journal of Applied Ecology 2, 221239. conductance, and soil drought in Betula occidentalis. Planta 196,
Cowan I.R. (1994) As to the mode of action of the Guard Cells in 357366.
dry air. In Ecophysiology of Photosynthesis (eds E.D. Schulze & Schnherr J. & Schmidt H.W. (1979) Water permeability of plant
M.M. Caldwell), pp. 205229. Springer-Verlag, New York. cuticles. Planta 144, 391400.
Dixon M.A. & Tyree M.T. (1984) A new stem hygrometer, Schulze E.D. (1994) The regulation of plant transpiration: interac-
corrected for temperature gradients and calibrated against the tions of feedforward, feedback, and futile cycles. In Flux Control
pressure bomb. Plant, Cell and Environment 7, 693697. in Biological Systems (eds E.D. Schulze), pp. 203235.
Dodd I.C., Stikic R. & Davies W.J. (1996) Chemical regulation of Academic Press, San Diego.
gas-exchange and growth of plants in drying soil in the field. Schurr U., Gollan T. & Schulze E.D. (1992) Stomatal response to
Journal of Experimental Botany 47, 14751490. drying soil in relation to changes in the xylem sap composition of
Farquhar G.D. (1978) Feedforward responses of stomata to humid- Helianthus annuus. 2. Stomatal sensitivity to abscisic acid
ity. Australian Journal of Plant Physiology 5, 787800. imported from the xylem sap. Plant, Cell and Environment 15,
Franks P.J., Cowan I.R. & Farquhar G.D. (1997) The apparent feed- 561567.
1998 Blackwell Science Ltd, Plant, Cell and Environment, 21, 10291038
1038 J. Comstock and M. Mencuccini

Schurr U. & Schulze E.-D. (1996) Effects of drought on nutrient relations between leaf water potential and stomatal conductance
and ABA transport in Ricinus communis. Plant, Cell and in anisohydric species. Plant, Cell and Environment 19, 7584.
Environment 19, 665674. Trejo C.L., Clephan A.L. & Davies W.J. (1995) How do stomata
Sheriff D.W. (1984) Epidermal transpiration and stomatal read abscisic acid signals? Plant Physiology 109, 803811.
responses to humidity: some hypotheses explored. Plant, Cell Tyree M. & T. (1997) The Cohesion-Tension theory of sap ascent:
and Environment 7, 669677. current controversies. Journal of Experimental Botany 48,
Sperry J.S., Adler F.R., Campbell G.S. & Comstock J. (1998) 17531765.
Hydraulic limitation to flux and pressure in the soil-plant con- Tyree M.T. & Sperry J.S. (1988) Do woody plants operate near the
tinuum: results from a model. Plant, Cell and Environment 21, point of catastrophic xylem dysfunction caused by dynamic
347359. water stress? Plant Physiology 88, 574580.
Stirzaker R.J. & Passioura J.B. (1996) The water relations of the Wilkinson S. & Davies W.J. (1997) Xylem sap pH increase: a
rootsoil interface. Plant, Cell and Environment 19, 201208. drought signal received at the apoplastic face of the guard cell
Tardieu F. (1996) Drought perception by plants: do cells of that involves the suppression of saturable abscisic acid uptake by
droughted plants experience water stress? Plant Growth the epidermal symplast. Plant Physiology 113, 559573.
Regulation 20, 93104.
Tardieu F., Lafarge T. & Simonneau T. (1996) Stomatal control by Received 4 February 1998; received in revised form 5 June 1998;
fed or endogenous xylem ABA in sunflower: interpretation of cor- accepted for publication 5 June 1998

1998 Blackwell Science Ltd, Plant, Cell and Environment, 21, 10291038