Tree Physiology 25, 361–371
© 2005 Heron Publishing—Victoria, Canada
Physiological and morphological responses to water stress in two
Acacia species from contrasting habitats
D. O. OTIENO,1,2 M. W. T. SCHMIDT,1 S. ADIKU3 and J. TENHUNEN1
1
Department of Plant Ecology, University of Bayreuth, P.O. Box 95440, Bayreuth, Germany
2
Corresponding author ([email protected])
3
Department of Science, Faculty of Agriculture, University of Ghana, Accra, Ghana
Received February 20, 2004; accepted August 22, 2004; published online January 4, 2005
Summary Container-grown seedlings of Acacia tortilis
Forsk. Hayne and A. xanthophloea Benth. were watered either
every other day (well watered) or every 7 days (water-stressed)
for 1 year in a greenhouse. Total plant dry mass (Tdm), carbon
allocation and water relations were measured monthly. Differences in leaf area (LA) accounted for differences in Tdm between the species, and between well-watered and waterstressed plants. Reduction in LA as a result of water stress was
attributed to reduced leaf initiation, leaf growth rate and leaf
size. When subjected to prolonged water stress, Acacia xanthophloea wilted more rapidly than A. tortilis and, unlike A. tortilis, lost both leaves and branches. These differences between
species were attributed to differences in the allocation of carbon between leaves and roots and in the ability to adjust osmotically. Rapid recovery in A. xanthophloea following the prolonged water-stress treatment was attributed to high cell wall
elasticity. Previous exposure to water stress contributed to water-stress resistance and improved recovery after stress.
Keywords: biomass allocation, cell wall elasticity, drought
stress, osmotic adjustment, savanna, transpiration.
Introduction
Soil water availability is a key factor in the growth, development, species composition and distribution of savanna trees
(Noy-Meir 1973, Reynolds et al. 2004). Thus, understanding
soil water uptake patterns by trees growing in the savanna, and
the associated shoot responses to water loss during drought,
will help explain differences among species in productivity,
survival and distribution, and could shed more light on the
functioning of these dryland ecosystems.
Information about root distribution and knowledge of the
basic mechanisms of soil water extraction and transport by tree
species provide one basis for assessing differences among species in habitat preferences and ecological potentials. Extensive
deep-rooting systems with a large surface area over which water absorption can take place facilitate soil water uptake; and in
trees adapted to arid habitats, both the relative allocation of
photosynthates to roots and the absolute rate of root growth
rate tend to be high (Scholz et al. 2002). Extensive root devel-
opment allows extraction of water from a large volume of soil
or from a deep water table (Jones 1992, Jackson et al. 2000).
As water becomes limiting, certain trees show a decrease in
cell sap osmotic potential, thus increasing the water potential
gradient between soil and roots, thereby allowing water uptake
to continue despite declining soil water content (Tyree and
Jarvis 1982).
However, tree performance in dry habitats cannot be evaluated without considering constraints within the plant that influence carbon gain (Ehleringer 1994). For example, trees
with an effective water supply system may lack specific adaptations for controlling water loss, resulting in low tissue water
status that affects plant performance (Kramer 1980, Levitt
1980). Low maximum stomatal conductance and high stomatal sensitivity to changes in water status may be required to
maintain leaf water potential (ΨL) above a critical threshold
and to avoid xylem cavitation (Tyree and Sperry 1989, Jones
and Sutherland 1991). In certain tree species, however, stomatal conductance declines long before there is a noticeable
change in soil water content, imposing an early restriction to
CO2 uptake (Sperry 2000). Optimally, stomatal regulation of
water loss should balance transpiration with water supply to
the leaves so that a dangerous decrease in ΨL is avoided without unnecessary restriction to carbon gain (Meinzer 2002).
Differences among species in the diurnal fluctuations in ΨL
may reveal differences in soil water uptake or the flow of water
between roots and shoots, or both, which in turn influences
stomatal responses (Jones and Sutherland 1991). Differences
in stomatal behavior depend therefore not only on differences
in sensitivity to environmental factors associated with the development of water deficit, but also on root system development (Larcher 2003). Such differences in stomatal sensitivity
between species, expressed during the development of
drought, would serve to limit transpiration and compensate for
differences in vulnerability to xylem cavitation (Tyree and
Sperry 1988, Jones and Sutherland 1991). Because gaseous
exchange, primary productivity and plant fitness are related
(Ehleringer 1994, Saliendra et al. 1995), monitoring transpiration through sap flux measurements, coupled with instantaneous measurements of gas exchange activity during the de-
362
OTIENO, SCHMIDT, ADIKU AND TENHUNEN
velopment of water stress, should provide reliable information
about species performance and ecological potentials. In certain tree species, water loss is reduced through reduction of total leaf area (LA) (Munné-Bosch and Alegre 2004). Reduction
in LA as soil water becomes limiting is achieved through reduction in leaf size, leaf rolling or leaf shedding, thus reducing
the transpiring leaf surface, but with significant negative impact on carbon gain and overall plant productivity (Jones
1992). Thus, monitoring leaf phenology may provide valuable
information about species fitness.
The arid and semi-arid savannas of Kenya are dominated by
Acacia species, most of which are drought tolerant (Oba et al.
2001). Knowledge is lacking on how different physiological
and morphological mechanisms interact to bring about water
stress tolerance as well as to maintain productivity during
drought in these species. Acacia tortilis and A. xanthophloea
have contrasting habitat preferences, with A. tortilis found in
more xeric eastern and northern provinces of Kenya, whereas
A. xanthophloea occurs in the mesic lowlands (Noad and
Birnie 1989). These distribution patterns reflect differences in
the ability to cope with water stress.
Extrapolation of seedling characteristics to other life stages
requires caution, since changes throughout development in the
ability to access resources may cause plants at different stages
of development to rely on different mechanisms for coping
with stress (Cavender-Bares and Bazzaz 2000). Nevertheless,
attributes possessed by seedlings may indicate the relative capacity of a species to survive and grow under drought conditions. Thus, controlled experiments may provide a basis for
predicting mature tree responses to changes in the environment if scaling factors are applied that account for differences
in response between seedlings and mature trees (Bazzaz et al.
1996). Seedlings of A. tortilis and A. xanthophloea were studied to test the hypothesis that distribution patterns of trees in
the arid savanna are dependent on characteristics that improve
soil water uptake, but limit the rate of soil water depletion.
Most plant responses to water stress are elicited only after
the plant experiences or senses impending water shortage
(Ryel et al. 2004). Here, we test the hypothesis that mechanisms that lead to improved water uptake and conservation can
be induced at early stages of tree development by subjecting
seedlings to moderate water stress.
Materials and methods
Plant culture and experimental design
Seeds of Acacia tortilis and A. xanthophloea (Kibwezi provenance), previously obtained from the Kenya Forestry Research
Institute (KEFRI), Nairobi, Kenya, were germinated on May
23, 2001, in a greenhouse at the University of Bayreuth, Germany. The seeds were pretreated by immersing them in water
at 100 °C and leaving them to soak overnight as the water
cooled. The imbibed seeds were sown in moist vermiculite and
germinated at 27 °C. Most seeds germinated on the third day
after sowing. The seedlings were transplanted to plastic pots (4
× 4 × 6 cm) and grown for 1 month, with regular watering, be-
fore transfer to larger containers (18 cm high and 14 cm in diameter). On September 13, 2001, seedlings were transferred to
0.027-m3 pots containing a 2:1 mix of forest soil and sand. The
pots were arranged on a greenhouse bench in two blocks, each
comprising 24 trees or pots per species, randomly arranged
within the blocks.
Differences between blocks in the watering treatment commenced on September 16, 2001, when all pots were watered to
capacity. Subsequently, seedlings in one block (control) were
watered every other day, whereas seedlings in the other (water-stressed) block were watered at progressively longer intervals, which reached 6 days without water on October 3, 2001,
and then remained constant with watering every 7 days. The
treatments were continued for 1 year, during which time
greenhouse temperatures were 25–30 °C, and mean photosynthetic active radiation (PAR) was 500–800 µmol m – 2 s – 1 .
Measurements
Plant growth and morphology On a monthly basis, three
randomly selected seedlings from each treatment per species
were harvested and separated into leaves, stems and roots. Leaf
area was measured with a portable leaf area meter (CI-202,
CID, Camas, WA). Roots, stems and leaves were dried separately at 70–80 °C for 24 h and weighed. Total dry mass (Tdm),
leaf dry mass (Ldm), root dry mass (Rdm), leaf area (LA) and total root:shoot dry mass ratio (R:S) were determined.
Sap flux Sap flux was measured on 1-year-old plants in the
greenhouse by the stem heat balance (SHB) method described
by Sakuratani (1981, 1984). The gauges consisted of a flexible
heater mounted on a cork and attached to the stem segment.
Heat sensors (copper-constantan thermocouples) were
mounted on the cork and insulators. The heating plate was continuously supplied with a 4.5 V DC current, of which one sixth
was looped through the logger to record the current supplied to
the heaters, as needed to calculate sap flow (F). Heat dissipation in the vertical (Qv) and radial (Qr) directions was sensed by
thermocouple junctions on the mounting cork. The entire apparatus was insulated in thick polyvinyl foam and aluminum foil
to prevent solar heating. Signals from the thermocouple junctions were automatically recorded by a data logger (Delta-T
Devices, Burwell, U.K.). Data were collected every 5 min, and
30-min means stored. Sensors were installed on the main
stems, 50 cm above the stem base on two trees per species in
each treatment. The energy budget equation for the heated stem
section was expressed as:
Pin = Q r + Qv + Qflow
(1)
where Pin = electrical power to the heater (W) and Qflow = heat
loss by convection by the sap (W). Sap flow (F) was calculated as:
F =
Qflow
Cp ∆Tsap
TREE PHYSIOLOGY VOLUME 25, 2005
(2)
WATER STRESS RESPONSES IN ACACIA SEEDLINGS
where Cp = heat capacity of water (J g –1 °C – 1) and ∆Tsap = temperature difference between sap below and above the heater
(°C).
Diurnal changes in leaf water potential Diurnal courses of
leaf water potential (ΨL) were determined 1 day after re-watering with a pressure chamber (PMS Instruments, Corvallis, OR)
between 0800 and 1800 h. The days selected for measurements
were those when re-watering for water-stressed and control
plants coincided. Similar measurements were conducted on
water-stressed plants on Day 6 of the drying cycle, when maximum water stress was attained. Due to the small size of Acacia
leaves, measurements were made on young shoots bearing 2–3
leaves. Measurements were conducted between July and September when air temperatures and solar irradiances were near
the maximum (25 °C and 1200 µmol m –2 s –1, respectively).
Pressure–volume measurements Five young shoots per species in each treatment were excised under distilled deionized
water early in the morning. The shoots were left to re-saturate
in a dark chamber over a period of 24 h. During this period, they
were wrapped with a plastic film to prevent evaporative water
loss. After 24 h, the shoots were removed from water and their
fresh saturated weights and water potentials (Ψ) determined.
The shoots were then left to transpire freely under ambient conditions on a laboratory bench. At intervals (5–10 min), mass
and Ψ of each shoot were measured. On each occasion, the
pressure chamber (PMS Instruments) was pressurized slowly
until a bubble of air or water appeared at the distal end of the cut
shoot (Tyree and Jarvis 1982). Shoots were weighed immediately before and after each water potential determination and
the weights used to calculate relative water content (RWC) according to Kramer (1980). Measurements were continued until
the shoots were beyond the wilting point, i.e., there was no further change in mass. The shoots were then oven-dried at 80 °C
for 24 h and weighed. To obtain pressure–volume (P–V)
curves, reciprocals of tissue water potentials were plotted
against relative water content RWC (%) for each shoot per species.
Osmotic potential at full turgor (Π100 ), relative water content at the turgor loss point (RWCtlp) and water potential at the
turgor loss point (Ψtlp) were derived from P–V curves by considering a regression line between the inverse of the final balancing pressure points and RWC (Tyree and Hammel 1972).
Mean values for each species were statistically tested (t-test)
for significant differences between treatments and species.
Turgor potential was estimated as the difference between water potential (Ψ) and Π as described by Tyree and Jarvis
(1982). Osmotic adjustment was calculated as the difference in
mean Π100 between water-stressed plants and controls.
Hydraulic conductance Leaf specific hydraulic conductance
(K sL ) was estimated for each treatment per species from the reciprocals of total flow resistance (R). A simple Ohm’s Law
analogy relating diurnal changes in E and ΨL was employed to
estimate total flow resistance (R) from soil to leaf (Jones and
Sutherland 1991).
363
Imposition of severe water stress Beginning August 25, 2002,
water was withheld from sets of three plants per species of both
control and water-stressed treatments until they failed to regain
turgor overnight.
Leaf water potential during severe water stress Leaf water
potential (ΨL) was measured on Days 1, 5, 8, 12 and 14 during
the period when water was withheld. Measurements of ΨL
were conducted on the three replicate samples of each treatment per species. Leaf water potential was determined between
0600 and 0700 h, before any significant transpiration was realized, and also at midday when plants experience the lowest Ψ
values. From every treatment per species, two shoots (2–
3 leaves) were obtained from well-exposed branches of the 15month-old plants for ΨL determination.
Leaf transpiration and stomatal conductance during severe
water stress Diurnal courses of leaf transpiration (E) and
stomatal conductance (gs) were determined on duplicate
leaves of well-exposed branches of the same plants as for the
ΨL measurements. Measurements were conducted on Days 3,
4 and 5 during the time that water was withheld using a portable gas exchange system (LI-6400, Li-Cor, Lincoln, NE).
Measurements were conducted under natural light conditions.
After Day 5, leaves of control plants were wilted most of the
day and no further gas exchange measurements could be made
with them. Following measurements, the leaves were detached
and their area determined with a portable leaf area meter
(CI-202 CID).
Leaf shedding During the severe stress treatment, the course
of leaf shedding was monitored.
Post-stress recovery Acacia xanthophloea plants subjected
to severe water stress were re-watered to container capacity on
September 7, 2002, and every 2 days thereafter, during which
time their ΨL was monitored until they fully recovered, i.e., until their water potentials were about zero and they began to generate new leaves. Acacia tortilis plants subjected to severe
stress were re-watered on September 11, 2002. This was the period when A. tortilis plants had attained ΨL similar to those previously attained by severely stressed A. xanthophloea plants
and also when their leaves failed to regain turgor overnight. For
both species and in each treatment after re-watering, two
branches of approximately similar lengths, sizes, orientations
and positions on the main stems were identified and marked.
On a daily basis, the number of sprouting leaves on each of
these branches was recorded in each treatment per species.
Length, breadth, number and size of leaflets located on each
side of the rachis were determined. The mean relative expansion rate during the period with optimum water supply was determined. Examination of leaf regeneration and leaf characteristics for controls of A. xanthophloea were conducted on the
short stocks of the main stems, because the plants lost most of
the aerial shoots during the severe water stress, leaving only a
short (10–20-cm) portion of the stem alive.
On two of the recovering plants of each species per treatment, sap flux was also monitored with stem heat balance
(SHB) sensors as described above. For each species, the sensor
TREE PHYSIOLOGY ONLINE at http://heronpublishing.com
364
OTIENO, SCHMIDT, ADIKU AND TENHUNEN
guages were installed on the main stems, on the last day of severe water stress. The sensors were installed 50 cm above the
stem base on each of the measured plants. Resumption of sap
flow was monitored during the period when the plants were recovering and new leaves being formed.
Results
Plant growth
In the control treatment, A. xanthophloea and A. tortilis had
mean Tdm values of 403 ± 6.1 and 243 ± 0.74 g per plant, respectively, at the end of the experiment (Table 1). The waterstress treatment reduced dry mass by 45% in A. xanthophloea
and 40% in A. tortilis. Significant differences (P < 0.05) between treatments and species occurred 6 months after imposition of water stress.
In the control treatment, leaf areas at the end of the experiment were 0.57 and 0.34 m 2 in A. xanthophloea and A. tortilis
plants, respectively (Table 1). Water stress resulted in a 26%
reduction in LA in A. xanthophloea versus a reduction of only
15% in A. tortilis. The water-stress treatment caused significant leaf senescence and shedding, and a reduction in leaf initiation and expansion as the pots dried, which contributed to the
reduction in total LA in the water-stressed plants.
The relationship between LA and Tdm was analyzed by log
plots of mean values of accumulated LA versus Tdm. There was
a linear relationship irrespective of species between log LA
and log Tdm for both water-stressed and control plants (Figure 1), suggesting the dependence of Tdm on LA development.
There was a significant change (P < 0.001) in the slope and intercept of the relationship when seedlings were water stressed.
Carbon partitioning
The water-stress treatment significantly increased R:S ratio of
A. tortilis to 0.67 (± 0.04) at the end of the experiment compared with 0.52 (± 0.01) in the controls. Irrespective of treatment, the R:S ratio of A. tortilis increased two- and threefold
over the course of the experiment in control and water-stressed
plants, respectively. The R:S ratios of water-stressed A. tortilis
were probably even higher than measured, as this species developed many fine roots as water stress increased, some of
which were lost when the roots were harvested. No significant
difference in R:S ratio was observed between controls and wa-
Figure 1. Transformed plot of total dry mass (Tdm ) versus leaf area
(LA) of (A) controls and (B) stressed A. xanthophloea (䊏, 䊐; thick
lines) and A. tortilis (䊉, 䊊; thin lines). Controls were watered every
other day, while stressed seedlings were watered every 7 days.
ter-stressed A. xanthophloea (Table 1) seedlings. Acacia xanthophloea showed a decline in R:S ratio over the course of the
experiment of 14 and 3% in controls and water-stressed plants,
respectively.
Root dry mass to leaf area (Rdm:LA) ratio increased in both
species with time. Water-stressed A. tortilis had higher Rdm:LA
ratio (38%) than controls, but no significant difference was
Table 1. A summary of growth parameters (root dry mass (Rdm ), leaf area (LA) and total dry mass (Tdm) in control and water-stressed Acacia
xanthophloea and Acacia tortilis seedlings. Root:shoot (R:S) ratio, and root dry mass to leaf area (Rdm:LA), ratios are also shown. Data are from
the final harvest made 15 months after germination. Values are means (± SD, n = 3). Significant differences are indicated by asterisks: * = P <
0.05; and ** = P < 0.001.
Parameter
Tdm (g)
LA (m2)
Rdm:LA ratio
R:S ratio
A. xanthophloea
A. tortilis
Water-stressed
Control
Water-stressed
Control
222 ± 2.56 *
0.42 ± 0.04 *
0.011 *
0.19 ± 0.002
403 ± 6.1
0.57 ± 0.08
0.001
0.22 ± 0.001
147 ± 0.75 **
0.29 ± 0.009 **
0.013 **
0.67 ± 0.04 **
243 ± 0.74
0.34 ± 0.04
0.008
0.52 ± 0.01 *
TREE PHYSIOLOGY VOLUME 25, 2005
WATER STRESS RESPONSES IN ACACIA SEEDLINGS
seen in R:S ratio between controls and water-stressed A. xanthophloea. The increase in Rdm:LA ratio observed in A. xanthophloea was mainly attributable to leaf shedding.
The species differed significantly in their rooting patterns.
Acacia tortilis tended to develop a stronger tap root system and
rooting depth increased significantly in the water-stressed
plants, whereas roots of A. xanthophloea were more fibrous
and lacked a well-defined tap root.
Sap flux and leaf water potential
Water stress reduced sap flux by 54 and 44% in A. xanthophloea and A. tortilis, respectively (Figure 2). Control plants
of A. xanthophloea showed higher (mean diurnal maximum =
110 g h –1) sap fluxes than A. tortilis (mean diurnal maximum =
80 g h –1). Expressing sap flux per unit LA (Figure 2B) revealed no significant differences (P < 0.05) between controls
of A. tortilis and A. xanthophloea. The water-stress treatment
significantly reduced whole-plant sap flux, with rates declining by half at the end of the water-stress cycle (not shown). Sap
flux increased more rapidly following re-watering of waterstressed plants of A. xanthophloea than of A. tortilis. Stem water storage must have been small, because of the small size of
the plants (Sakuratani 1984); therefore, we assumed that sap
flux was equivalent to transpirational water loss.
Increasing transpiration early in the day led to a decline in
ΨL (Figure 3). Controls of A. xanthophloea, which had a high
transpiration rate, underwent a steeper drop in ΨL than con-
365
trols of A. tortilis, reaching a mean diurnal minimum value of
–2.4 MPa (Figure 3A). Declines in ΨL were less in waterstressed plants than in control plants and recovered rapidly at
the end of the day. Withholding water significantly reduced
soil water availability, leading to low ΨL. Acacia xanthophloea, however, experienced much lower ΨL throughout the
day than A. tortilis (Figure 3C). A strong inhibition to the decline in water potential was observed for both species after
midday when soil water was limiting.
Hydraulic conductance
Mean leaf specific hydraulic conductances (K sL ) for controls
were 4.73 and 3.48 mmol MPa –1 m –2 s –1 in A. tortilis and
A. xanthophloea, respectively. Mean values estimated for water-stressed plants were 4.45 and 3.13 mmol MPa –1 m –2 s –1 in
A. tortilis and A. xanthophloea, respectively. In both treatments, A. tortilis exhibited greater hydraulic conductance than
A. xanthophloea. There was no significant (P < 0.05) difference in K sL between water-stressed and control plants.
Tissue water relations
Parameters derived from P–V curves are shown in Table 2. Osmotic potential at full turgor (Π100 ) of A. tortilis was significantly lower in water-stressed plants than in control plants. It
was also lower than in A. xanthophloea. Acacia tortilis had a
mean osmotic adjustment of 0.48 MPa. A slight but smaller
(0.16 MPa) osmotic adjustment also occurred in the water-
Figure 2. Mean whole-tree sap
flow (A) and whole-tree sap
flow expressed per unit leaf
area (B), in 1-year-old control
and water-stressed Acacia
xanthophloea and Acacia
tortilis seedlings. Symbols: 䊏
= A. xanthophloea (control); 䊐
= A. xanthophloea (waterstressed); 䊉 = A. tortilis (control); 䊊 = A. tortilis (water-stressed). Data represent
means of two plants from each
treatment per species.
TREE PHYSIOLOGY ONLINE at http://heronpublishing.com
366
OTIENO, SCHMIDT, ADIKU AND TENHUNEN
both species had a higher relative water content at the turgor
loss point (RWCtlp) than water-stressed plants. Bulk modulus
of elasticity (ε) was significantly greater in water-stressed
plants than in control plants of A. xanthophloea, but not of
A. tortilis.
Severe water stress
Prolonged withholding of water resulted in a faster decline in
the water status of plants that were previously well watered
(control) than of those that had been previously water stressed
(Figure 5). For example, leaves of control plants of A. xanthophloea failed to regain turgor overnight after only 3 days without watering, whereas leaves of previously water-stressed
plants maintained turgor overnight for 8 days without re-watering. Similarly, in A. tortilis, previously water-stressed plants
survived 12 days without water before losing turgor overnight,
compared with 7 days in control plants. Controls of A. xanthophloea lost turgor by midday (Table 2) after the second day
of the severe water-stress treatment, attaining a mean minimum water potential of –1.5 MPa during the day (not shown).
After reaching the Ψtlp, further declines in ΨL were accompanied by leaf senescence and branch desiccation, which was
more severe in control plants (data not shown). Previous water-stress treatment prolonged the time before reaching Ψtlp to
5 days. Control and previously water-stressed A. tortilis plants
reached Ψtlp (1.1 ± 0.11 and 1.4 ± 0.10 MPa, respectively, c.f.
Table 2) after 7 and 11 days without re-watering, respectively.
Recovery after watering was resumed occurred more slowly in
A. tortilis than in A. xanthophloea.
Leaf transpiration, stomatal conductance and water
potential
Figure 3. Diurnal changes in leaf water potential (ΨL ) of control (䊏,
䊉) and water-stressed (䊐, 䊊) (A) A. xanthophloea and (B) A. tortilis
seedlings. Plants were watered to container capacity the previous
night before measurement the following day. (C) Diurnal pattern of
(A) and (B) on Day 6 of water stress. Error bars show deviation from
the mean (n = 3).
Diurnal patterns of leaf transpiration (E) and gs in control and
water-stressed A. xanthophloea and A. tortilis measured on
Day 3 of the severe water-stress treatment are shown in Figure 6. Maximum E and gs declined more than twofold in control plants compared with water-stressed plants of both species. Acacia xanthophloea was affected most, showing depressed gs and E around midday. Stomatal conductance peaked
in the morning (1000 h), reaching maximum values of 250 and
50 mmol m – 2 s –1 for water-stressed and control plants, respectively, but rapidly declined to near-zero before noon and only
resumed later in the day. A similar pattern was exhibited by E.
Previously water-stressed plants of A. tortilis maintained high
E and gs (3 mmol m – 2 s –1 and 220 mmol m – 2 s –1, respectively)
and A. tortilis control plants showed no midday depression in
either E or gs.
Post-stress recovery
stressed A. xanthophloea. Water potential at the turgor loss
point was higher in the controls of both species than in the water-stress treatment. However, Ψtlp of water-stressed A. tortilis
was significantly lower than that of control plants. It was also
lower than that of water-stressed A. xanthophloea. There was
a strong relationship (r 2 = 0.96) between Π100 and Ψtlp, irrespective of species and treatments (Figure 4). Control plants of
Plant recovery after the severe water-stress treatment was associated with leaf initiation, leaf growth and LA development.
New shoots and leaves developed rapidly after re-watering.
Previously water-stressed A. tortilis retained leaves during the
severe stress treatment, and although most of the leaves looked
wilted, they continued to transpire, as shown in Figure 7. After
re-watering, leaf initiation was much slower in previously wa-
TREE PHYSIOLOGY VOLUME 25, 2005
WATER STRESS RESPONSES IN ACACIA SEEDLINGS
367
Table 2. Water potential at the turgor loss point (Ψtlp), osmotic potential at full turgor (Π100 MPa), bulk modulus of elasticity (ε) and relative water
content at the turgor loss point (RWCtlp) in control and water-stressed A. xanthophloea and A. tortilis plants. Values are means (± SD). Significant
difference P < 0.05 between means is indicated by (*) and P < 0.001 by (**).
A. xanthophloea
Ψtlp (–MPa)
ε (MPa)
Π100 (–MPa)
RWCtlp (%)
A. tortilis
Water-stressed
Control
n
Water-stressed
Control
n
1.24 ± 0.071
9.91 ± 1.32) **
0.97 ± 0.12
85 ± 0.8 *
1.1 ± 0.16
18.7 ± 2.05
0.81 ± 0.18
88 ± 0.8
9
9
9
9
1.4 ± 0.101 **
11.2 ± 2.34
1.33 ± 0.16 **
85 ± 0.5 *
1.1 ± 0.11
10.4 ± 3.11
0.85 ± 0.06
87 ±1
10
10
10
10
Discussion
Figure 4. Relationship between osmotic potential at turgor loss point
(Πtlp) and water potential at turgor loss point (Ψtlp) of control and water-stressed A. xanthophloea and A. tortilis plants. Deviations from
the means are shown by error bars (n > 8).
ter-stressed plants than in control plants (Figure 8). Leaf initiation was also slower in A. tortilis than in A. xanthophloea. The
severe water-stress treatment reduced mean leaf area by about
60 and 80% relative to control plants in previously waterstressed A. xanthophloea and A. tortilis, respectively.
The water-stress treatment resulted in significant decreases in
Tdm and LA, and a shift in carbon allocation. Leaf area was
most sensitive to water stress, with significant differences between treatments appearing earlier after imposition of stress
than differences in other morphological characteristics. Water
stress reduced LA through reduced leaf initiation, leaf size and
leaf production rate in both species.
By the end of the experiment, water stress had caused 45 and
40% declines in Tdm in A. xanthophloea and A. tortilis, respectively. After 15 months of growth, a major decline in growth in
both species, as a result of water stress, was mediated through
LA reduction (Figure 1) and probably through reduced CO2
assimilation as well, because the rate of stomatal conductance
also declined with water stress (Figure 6). Water stress affected Tdm accumulation in the fast-growing A. xanthophloea
more than in A. tortilis, and this was in agreement with the
magnitude of decline in LA caused by water stress. The results
are consistent with other studies. For example, a fast-growing
Eucalyptus provenance with higher LA and sufficient water
supply exhibited a large proportional decrease in mean leaf
size, LA and plant biomass under drought conditions and this
reduction in growth was due to reduced foliage area (Osorio et
al. 1998, Pita and Pardos 2001).
For A. xanthophloea, a 45% reduction in Tdm caused by water stress corresponded closely with a 41% decline in LA,
whereas in A. tortilis, a 32% reduction in LA as a result of water stress was associated with a 40% reduction in Tdm. The dif-
Figure 5. Progressive changes
in predawn water potential
(Ψpd ) of control and previously
water-stressed plants of A. xanthophloea and A. tortilis during
a subsequent severe water
stress. Plants were first watered
to container capacity before
withholding water until they
were wilted overnight. Values
are means of three plants per
species per treatment. Deviations from the means are shown
by error bars.
TREE PHYSIOLOGY ONLINE at http://heronpublishing.com
368
OTIENO, SCHMIDT, ADIKU AND TENHUNEN
Figure 6. Diurnal courses of
leaf transpiration (E) and
stomatal conductance (gs ) of
A. xanthophloea (upper panel)
and A. tortilis (lower panel)
plants conducted on the third
day after withholding water
during the severe water-stress
treatment. Seedlings were initially well watered (control)
or subjected to moderate water
stress (stressed). Plants were
watered to container capacity
before completely withholding
water.
ference between species in the slope of the relationship between LA and Tdm (Figure 1) may be explained by differences
in carbon partitioning between leaves and roots (Abrams
Figure 7. Recovery of sap flux in (A) water-stressed A. xanthophloea,
(B) controls and (C) water-stressed A. tortilis after the plants were
subjected to severe water stress beyond the wilting point and then
re-watered to container capacity. Values are means of two plants per
treatment. Values for controls of A. xanthophloea are missing (see
text).
1994) or in osmotic adjustment (Munns 1988, Blake et al.
1991), either of which could improve soil water uptake at the
expense of growth (Osonubi and Davies 1978, Abrams 1994).
Acacia xanthophloea grew faster than A. tortilis, which we attribute to its ability to allocate a greater proportion of dry mass
to LA, thereby maximizing carbon gain under favorable conditions.
Control plants of A. xanthophloea with higher LA experienced greater water use compared to those of A. tortilis (100
and 80 g h –1, respectively; see Figure 2). On a long-term basis,
reduction in transpiration as a result of the water-stress treatment was associated with LA reduction. However, mechanisms such as stomatal regulation, leaf folding, root characteristics and root-to-shoot water transport were involved in controlling the daily plant water budget. Under conditions of
ample water supply, A. xanthophloea experienced higher transpiration rates and lower ΨL (–2.0 and –1.4 MPa, respectively)
at the end of the day than A. tortilis, suggesting a poor balance
between transpirational water loss and shoot water supply. At
midday, when ΨL was low, plants showed strong regulation of
water loss through reduced stomatal conductance, resisting
further decline in ΨL (Figure 3). In A. xanthophloea, however,
the resulting low transpiration rates were not accompanied by
an immediate recovery of ΨL as observed in A. tortilis, suggesting an interrupted supply of water from roots (Tyree and
Sperry 1989, Meinzer and Grantz 1990, Sperry and Pockman
1993, Sperry 2000).
Hydraulic conductance is associated with water uptake at
the root surfaces, root:leaf surface area ratio or inherent absorption capacity, root permeability and an effective water
transport system (Tyree and Sperry 1988, Reich and Hinkley
1989, Ni and Pallardy 1990). Control plants of A. tortilis exhibited higher hydraulic conductance than control plants of
A. xanthophloea (4.73 and 3.48 mmol MPa –1 m –2 s –1, respectively). At high soil water, soil resistance to water flow is small
and any observed differences in hydraulic conductance should
TREE PHYSIOLOGY VOLUME 25, 2005
WATER STRESS RESPONSES IN ACACIA SEEDLINGS
Figure 8. Recovery and development of (A) leaf area and (B) leaf
initiation of controls and previously water-stressed A. xanthophloea
and A. tortilis after the plants were subjected to severe water stress
beyond the wilting point and then re-watered to container capacity.
Measurements were taken from four branches of two trees per treatment. Data for controls of A. xanthophloea are missing in B.
be largely attributed to differences in plant resistance (Ni and
Pallardy 1990). From our results, it follows that A. tortilis has a
more robust water-conducting system than A. xanthophloea.
Generally, hydraulic conductance is expected to decline with
increasing water stress as a result of increasing resistances
along the conducting pathway (Blizzard and Boyer 1980). The
decline in K sL observed in the water-stressed plants was, however, insignificant: 0.2 and 0.3 mmol MPa –1 m –2 s –1 for A. tortilis and A. xanthophloea, respectively. Two possible reasons
for this are that (1) a more competent transport system developed following the imposition of water stress and (2) there was
an improved balance between the absorbing root and transpiring leaf surface area in the water-stressed plants.
Slower cell expansion during periods of limited water availability, associated with a high carbohydrate concentration,
permits rapid primary and secondary wall growth, leading to
formation of cells with smaller pit-membrane pores, which are
less vulnerable to cavitation (Tyree and Sperry 1989). Thus,
the water-stress treatment may have led to the development
of better-adapted water-conducting vessels. Water-stressed
plants continually shed leaves between Days 4 and 6 following
each watering event. This may have improved or maintained
the balance between absorbing root and transpiring leaf surface area (shown as Rdm:LA ratio), and is likely to have improved root-to-shoot hydraulic conductance (see Meinzer and
Grantz 1990).
There was an absolute increase in root growth in waterstressed plants. The results, therefore, show a strong link between water supply from the soil by roots and the transpiring
369
leaf surface area. Furthermore, they show that carbon allocation is one way by which plants determine the effectiveness of
water supply from roots to shoots. Increased resource allocation to the roots compared to leaves leads to high hydraulic
conductance, and this represents a vital adaptation to drought
stress (Kramer 1980). This means that, for a given rate of transpiration, a species like A. tortilis, which has a high hydraulic
conductance, will undergo less reduction in ΨL during the day.
The capacity to dampen the decline in ΨL during water stress
ensures prolongation of photosynthesis during drought periods and has implications for plant survival and productivity in
dry environments (Cochard et al. 2002). Acacia xanthophloea,
however, exhibited an exaggerated dehydration avoidance
mechanism associated with stomatal closure that must be
linked to a low K sL, and with the disadvantage that declining
ΨL may have approached the threshold for xylem cavitation
(Tyree and Sperry 1989). That water-stressed plants of both
species were able to restore ΨL when they were re-watered
during the moderate water-stress treatment, however, indicates
that severe xylem cavitation did not occur during drought and
that effective water transport to the shoots was restored when
drought was relieved. However, cavitation occurred in control
plants of A. xanthophloea during severe water stress, inhibiting resumption of water transport and restoration of ΨL, and
causing the death of aerial shoots in the control plants.
Water-stressed A. tortilis showed a significant (P < 0.05) reduction in Π100 compared with controls, which may have been
largely due to solute accumulation in the cells, as determination of cell osmotic potential at full turgor eliminates increased
solute concentration as a result of decreased cell volume
(Jones 1996). The observed solute increase, therefore, constitutes osmotic adjustment (Hsiao 1973, Osonubi and Davies
1978). As a result of the 6-day water stress cycle, A. tortilis adjusted osmotically by 0.48 MPa, whereas A. xanthophloea adjusted an insignificant 0.16 MPa. Osmotic adjustment may
constitute an adaptation to drought stress when it causes a decrease in Ψtlp (Morgan 1984), i.e., maintenance of turgor down
to lower values of ΨL. Osmotic adjustment significantly improves soil water uptake under dry conditions (Tyree and
Jarvis 1982), and allows maintenance of open stomata with
larger apertures and a higher stomatal conductance and net
photosynthesis rate down to lower values of ΨL (Myers and
Landsberg 1989).
In the current study, A. tortilis maintained higher ΨL and
stomatal conductances when the soil was drying. Although
these observations could be explained by osmotic adjustment,
an alternative explanation is that they were the result of reduced LA. Water-stressed A. tortilis had limited LA, resulting
in less water loss from the pots. This would mean delayed pot
dehydration, hence the plants experienced favorable soil water
status for longer. Another explanation might relate to root system characteristics: A. tortilis had a higher R:S ratio, had more
roots distributed over the entire soil mass and may have had a
higher water permeability across the root surfaces (as shown
by high K sL). Thus, repeatedly water-stressed plants were able
to draw water from a large soil mass, but were required to supply only a relatively small transpiring leaf surface.
TREE PHYSIOLOGY ONLINE at http://heronpublishing.com
370
OTIENO, SCHMIDT, ADIKU AND TENHUNEN
In A. xanthophloea, repeated water stress resulted in a reduction in ε from 18.7 to 9.9 MPa. Elastic modulus reflects the
mechanical properties of the cell wall such that fully turgid tissue is expected to possess the greatest apparent ε (Blake et al.
1991). Bulk modulus of elasticity will, therefore, decline with
decreasing cell water content (Tyree and Jarvis 1982). A more
elastic tissue, as denoted by a lower modulus of elasticity, indicates that turgor potential declines less rapidly per unit loss of
water (Blake et al. 1991). For a given Π, increased elasticity
facilitates turgor maintenance over a greater range of water
contents, hence improving water uptake from a drying soil
(Tyree and Jarvis 1982). Ability to improve cell wall elasticity
during water stress as observed in A. xanthophloea may account for its increased water-stress tolerance during severe
stress and might also have contributed to its quick recovery
when water stress was alleviated, by allowing greater carbon
utilization in cell repair processes and more rapid growth after
the relief of water stress (Blake et al. 1991). Active solute accumulation, however, shifts photosynthates away from growth
towards cell turgor regulation (Dale and Sutcliffe 1986,
Munns 1988), and could have accounted for the reduced
growth rate and slow recovery observed in water-stressed
A. tortilis after re-watering.
Response patterns shown by greenhouse plants of the two
Acacia species when subjected to different degrees of water
stress reflected those observed in mature trees growing under
natural conditions (Otieno et al. unpublished). Although the
water-stress treatment was limited to watering every seventh
day, the root mass of the experimental plants was confined to a
limited soil volume, compared with the volume that may be
explored by plants in the field, hence their tissue water status
dropped to an extent that might not be experienced by fieldgrown trees until several months without rainfall. Acacia xanthophloea trees are associated with marshes and swamps
(Noad and Birnie 1987), where water availability is high. They
grow rapidly during rainy seasons but lose a large number of
aboveground shoots during drought (Otieno et al. unpublished
observations). Similar observations were made for seedlings
of the same species subjected to different watering regimes in
the greenhouse. The results provide detailed information on
the physiological adaptations possessed by the two Acacia
species and help to explain the pattern of distribution exhibited
by mature trees growing naturally in the Kenyan savanna.
Conclusions
This study showed that behavior patterns exhibited by mature
trees growing under natural conditions are reflected closely in
young greenhouse-grown plants subjected to experimental
manipulation. The results therefore help to explain the distribution pattern observed in these species in their natural environment.
Plants previously subjected to water stress maintained a favorable ΨL and survived longer than well-watered control
plants when subjected to severe water stress, suggesting that
preconditioning improves the ability of seedlings to survive
water stress. However, the capacity of a species to overcome
drought was an overriding factor in water-stress resistance, as
shown by the differences between A. tortilis and A. xanthophloea. Plants of both species were better able to develop
water-stress resistance characteristics when subjected to prolonged moderate water stress than when severe water stress developed rapidly.
References
Abrams, M.D. 1994. Genotypic and phenotypic variation as stress adaptations in temperate tree species: a review of several case studies.
Tree Physiol. 14:833–842.
Bazazz, F.A., S.L. Bassow, G.M. Berntson and S.C. Thomas. 1996.
Elevated CO2 and terrestrial vegetation: implications for and beyond the global carbon budget. In Global Change and Terrestrial
Ecosystems. Eds. B. Walker and W. Stefan. Cambridge University
Press, Cambridge, pp 43 –76.
Blake, T.J., E. Bevilacqua and J.J. Zwiazek. 1991. Effects of repeated
stress on turgor pressure and cell elasticity changes in black spruce
seedlings. Can. J. For. Res. 21:1329–1333.
Blizzard, W.E. and J.S. Boyer. 1980. Comparative resistance of soil
and the plant water transport. Plant Physiol. 66:809–814.
Cavender-Bares, J. and F.A. Bazzaz. 2000. Changes in drought response strategies with ontogeny in Quercus rubra: implications for
scaling from seedlings to mature trees. Oecologia 124:8–18.
Cochard, H., L. Coll, X. Le Roux and T. Ameglio. 2002. Unraveling
the effects of plant hydraulics on stomatal closure during water
stress in walnut. Plant Physiol. 128:282–290.
Dale, J.E. and J.F. Sutcliffe. 1986. Water relations in plant cells. In
Plant Physiology. A Treatise. Vol. 9. Water and Solutes in Plants.
Ed. F.C. Steward. New York Academic Press, New York, 26 p.
Ehleringer, J.R. 1994. Variation in gas exchange characteristics among desert plants. Ecol. Stud. 100:361–392.
Hsiao, T.C. 1973. Plant response to water stress. Annu. Rev. Plant
Physiol. 24:519–570.
Jackson, R.B., J.S. Sperry and T.E. Dawson. 2000. Root water uptake
and transport: using physiological processes in global predictions.
Trends Plant Sci. 5:482–488.
Jones, H.G. 1992. Plants and microclimate: a quantitative approach to
environmental plant physiology. 2nd Edn. Cambridge Univ. Press,
Cambridge, 428 p.
Jones, H.G. and R.A. Sutherland. 1991. Stomatal control of xylem
embolism. Plant Cell Environ. 11:111–121.
Kramer, P.J. 1980. Drought stress and the origin of adaptations. In Adaptation to Water and High Temperature Stress. Eds. N.C. Turner
and P.J. Kramer. John Wiley and Sons, New York, pp 7–20.
Larcher, W. 2003. Physiological plant ecology. Ecophysiology and
stress physiology of functional groups. 4th Edn. Springer-Verlag,
Berlin, 513 p.
Levitt, J. 1980. Response of plants to environmental stresses. Vol. 2:
Water, radiation, salt and other stresses. 2nd Edn. New York Academic Press, New York, 606 p.
Lüttge, U. 1997. Physiological ecology of tropical plants. SpringerVerlag, Berlin, 384 p.
Meinzer, C.F. 2002. Co-ordination of vapor and liquid phase water
transport properties in plants. Plant Cell Environ. 25:265–274.
Meinzer, P.C. and D.G. Grantz. 1990. Stomatal and hydraulic conductance in growing sugar cane: stomatal adjustment to water transport
capacity. Plant Cell Environ. 13:383–388.
Morgan, J.M. 1984. Osmoregulation and water stress in higher plants.
Annu. Rev. Plant Physiol. 35:299–319.
Munné-Bosch, S. and L. Alegre. 2004. Die and live: leaf senescence
contributes to plant survival under drought stress. Funct. Plant Biol.
31:203–216.
TREE PHYSIOLOGY VOLUME 25, 2005
WATER STRESS RESPONSES IN ACACIA SEEDLINGS
Munns, R. 1988. Why measure osmotic adjustment? Aust. J. Plant
Physiol. 15:717–726.
Myers, B.J. and J.J. Landsberg. 1989. Water stress and seedling
growth of two Eucalypt species from contrasting habitats. Tree
Physiol. 5:207–218.
Ni, B. and S.G. Pallardy. 1990. Response of liquid flow resistance to
soil drying in seedlings of four deciduous angiosperms. Oecologia
84:260– 264.
Noad, T. and A. Birnie. 1989. Trees of Kenya. Nairobi Press, Nairobi,
Kenya, 169 p.
Noy-Meir, I. 1973. Desert ecosystems: environment and producers.
Annu. Rev. Ecol. Syst. 4:25–41.
Oba, G., N. Inger, N.C. Stenseth, J. Stave, C.S. Bjora, J.K. Muthondeki and W.K.A. Bii. 2001. Growth performance of exotic and
indigenous tree species in saline soils in Turkana, Kenya. J. Arid
Environ. 47:499–511.
Osonubi, O. and W.J. Davies. 1978. Solute accumulation in leaves
and roots of woody plants subjected to water stress. Oecologia 32:
323–332.
Osorio, J., M.L. Osorio, M.M. Chaves and J.S. Pereira. 1998. Water
deficits are more important in delaying growth than in changing
patterns of carbon allocation in Eucalyptus globulus. Tree Physiol.
18:363–373.
Pita, P. and J.A. Pardos. 2001. Growth, leaf morphology, water use
and tissue water relations of Eucalyptus globulus clones in response to water deficit. Tree Physiol. 21:599–607.
Reich, P.B and T. M. Hinkley. 1989. Influence of predawn water potential and soil to leaf hydraulic conductance on maximum daily
diffusive conductance in two oak species. Funct. Ecol. 3:719–726.
Reynolds, J.F., P.R. Kemp, K. Ogle and R.J. Fernández. 2004. Modifying the pulse reserve paradigm for deserts of North America: precipitation pulses, soil water and plant responses. Oecologia 141:
194–210.
Ryel, R.J., A.J. Leffler, M.S. Peek, P.Y. Ivans and M.M. Caldwell.
2004. Water conservation in Artemisia tridentata through redistribution of precipitation. Oecologia 141:335–345.
371
Sakuratani, T. 1981. A heat balance method for measuring water flux
in the stem of intact plants. J. Agric. Meterol. 37:9–17.
Sakuratani, T. 1984. Improvement of the probe for measuring water
flow rate in intact plants with the stem heat balance method. J.
Agric. Meteorol. 40:273–277.
Saliendra, N.Z., J.S. Sperry and J.P. Comstock. 1995. Influence of
leaf water status on stomatal responses to hydraulic conductance,
atmospheric drought and soil drought in Betula occidentalis. Planta
196:357–366.
Sperry, J.S. 2000. Hydraulic constraints on plant gas exchange. Agric.
For. Meteorol. 104:13–23.
Sperry, J.S. and W.T. Pockman 1993. Limitation of transpiration by
hydraulic conductance and xylem cavitation in Betula occidentalis.
Plant Cell Environ. 16:279–287.
Scholz, F.G., S.J. Bucci, G. Goldstein, F.C. Meinzer and A.C. Franco.
2002. Hydraulic redistribution of soil water by neotropical savanna
trees. Tree Physiol. 22:603–612.
Tyree, M.T. and H.T. Hammel. 1972. The measurement of turgor
pressure and the water relations of plants by the pressure bomb
technique. J. Exp. Bot. 23:267–282.
Tyree, M.T. and P.J. Jarvis. 1982. Water in tissues and cells. In Encyclopedia of Plant Physiology 12B. Physiological Plant Ecology II.
Eds. O.L. Lange, P.S. Nobel, C.R. Osmond and H. Ziegler. Springer-Verlag, Berlin, pp 36–77.
Tyree, M.T. and J.S. Sperry. 1988. Do woody plants operate near the
point of catastrophic xylem dysfunction caused by dynamic water
stress? Answers from a model. Plant Physiol. 88:574–580.
Tyree, M.T. and J.S. Sperry. 1989. Vulnerability of xylem to cavitation and embolism. Annu. Rev. Plant Phys. Plant Mol. Biol. 40:
19–38.
Walker, B.H. and A.N. Gillison. 1982. Australian savannas. Ecol.
Stud. 42:6–23.
Zimmermann, M.H. 1983. Xylem structure and the ascent of sap.
Springer-Verlag, Berlin, 139 p.
TREE PHYSIOLOGY ONLINE at http://heronpublishing.com