Journal of Experimental Botany, Vol. 57, No. 14, pp.

3727–3735, 2006
doi:10.1093/jxb/erl131 Advance Access publication 18 September, 2006

RESEARCH PAPER

Physiological responses of potato (Solanum tuberosum L.)

to partial root-zone drying: ABA signalling, leaf gas
exchange, and water use efficiency

Fulai Liu1,*, Ali Shahnazari1, Mathias N. Andersen2, Sven-Erik Jacobsen1 and Christian R. Jensen1
1
The Royal Veterinary and Agricultural University, Department of Agricultural Sciences, Crop Science,
Højbakkegaard Allé 13, DK-2630 Taastrup, Denmark
2
Department of Agroecology, Danish Institute of Agricultural Sciences, Research Centre Foulum,
PO Box 50, DK-8830 Tjele, Denmark

Received 4 May 2006; Accepted 20 July 2006

Abstract signals might have contributed to reduced water use

The physiological responses of potato (Solanum tuber-
osum L. cv. Folva) to partial root-zone drying (PRD)
were investigated in potted plants in a greenhouse (GH)
and in plants grown in the field under an automatic
rain-out-shelter. In the GH, irrigation was applied daily
to the whole root system (FI), or to one-half of the root
system while the other half was dried, for 9 d. In the
field, the plants were drip irrigated either to the whole
root system near field capacity (FI) or using 70% water
of FI to one side of the roots, and shifted to the other
side every 5–10 d (PRD). PRD plants had a similar
midday leaf water potential to that of FI, whereas in the
GH their root water potential (Wr) was significantly
lowered after 5 d. Stomatal conductance (gs) was more
sensitive to PRD than photosynthesis (A) particularly
in the field, leading to greater intrinsic water use
efficiency (WUE) (i.e. A/gs) in PRD than in FI plants
on several days. In PRD, the xylem sap abscisic acid
concentration ([ABA]xylem) increased exponentially
with decreasing Wr; and the relative [ABA]xylem (PRD/
FI) increased exponentially as the fraction of transpir-
able soil water (FTSW) in the drying side decreased.
In the field, the leaf area index was slightly less in
PRD than in FI treatment, while tuber biomass was
similar for the two treatments. Compared with FI,
PRD treatment saved 30% water and increased crop
water use efficiency (WUE) by 59%. Restrictions on
leaf area expansion and gs by PRD-induced ABA
and increased WUE.
Key words: Abscisic acid, leaf area, partial root-zone drying,
potato, stomatal conductance, water use efficiency.
Introduction
Potato (Solanum tuberosum L.) is one of the most important
crops in the world in terms of its use in human food and the
starch industry (Fabeiro et al., 2001). Due to its sparse and
shallow root system, potato is very sensitive to drought
stress (Jefferies, 1993), and tuber yield may be considerably
reduced by soil moisture deficits (Porter et al., 1999).
Therefore, irrigation is always needed for the production
of high-yielding crops (Fabeiro et al., 2001). However, the
increasing worldwide shortage of water resources requires
the optimization of irrigation management in order to
improve water use efficiency (WUE).
Partial root-zone drying (PRD) is a new water-saving
irrigation strategy presently being investigated in many
countries (Kang and Zhang, 2004). PRD involves irrigating
only part of the root zone, leaving the other part to dry to
a predetermined level before the next irrigation. PRD
allows the induction of the abscisic acid (ABA)-based root-
to-shoot chemical signalling system to regulate growth and
water use and thereby increase WUE (Davies et al., 2002).
PRD has been shown to be successful in grapevines (Stoll
et al., 2000) and in other fruit trees (Kang et al., 2002), and

* To whom correspondence should be addressed. E-mail: fl@kvl.dk

Abbreviations: A, net photosynthetic rate; ABA, abscisic acid; DAT, days after treatment; FC, field capacity; FI, full irrigation; FTSW, fraction of transpirable
soil water; GH, greenhouse; gs, stomatal conductance; PHC, pot-holding capacity; PRD partial root-zone drying; PWP, permanent wilting point; TDR, time
domain reflectometer; VPD, vapour pressure deficit; WUE, water use efficiency.

ª The Author [2006]. Published by Oxford University Press [on behalf of the Society for Experimental Biology]. All rights reserved.
For Permissions, please e-mail: journals.permissions@oxfordjournals.org
3728 Liu et al.
is claimed to be promising also for field crops (Kang et al.,
1998, 2000; Kirda et al., 2005) and vegetables (Kirda et al.,
2004; Dorji et al. 2005; Zegbe et al., 2006). Most recently,
a study by Saeed et al. (2005) showed that PRD could
also modify shoot growth and increase WUE in potatoes.
However, the physiological basis for improving WUE in
potatoes under PRD remains unknown.
It is well known that the PRD technique was developed
based on the hypotheses that roots in the dry soil column
sense soil drying and induce ABA-mediated signal trans-
duction pathways to reduce leaf expansion and stomatal
conductance; simultaneously, the wet soil column can
provide sufficient water to the plants and maintain a high
water status of the shoot (Davies and Zhang, 1991; Kang
and Zhang, 2004). ABA-induced partial stomatal closure
and reduced leaf area have been considered to be the main
physiological causes for saving water in plants under PRD
treatment (Davies et al., 2002). Our recent studies in
potatoes imposing progressive soil drying have shown
that root-sourced ABA reduces stomatal conductance at
mild soil water deficits and results in an increased intrinsic
WUE (Liu et al., 2005). Based on the aforementioned
studies, it is plausible to suggest that ABA signalling and
its subsequent effects on stomatal conductance and leaf
expansion may play a central role in improving WUE in
PRD-treated potatoes.
In the present study, experiments in a greenhouse (GH)
and in the field under an automatic rain-out-shelter were
done to explore the physiological basis for improving WUE
in potato under PRD irrigation. The aim of the GH
experiment was to test the basic hypothesis of PRD that
drying part of the root system could induce ABA signalling
in the absence of changing leaf water status. Thus, the
plants with a split-root system were subjected to soil drying
in only one half of the root system while the other half was
fully irrigated, and the irrigation was not shifted between
sides. However, alternating irrigation between different
parts of the root system is essential for sustaining a
continuous ABA-based chemical signal to regulate shoot
physiology (Davies and Hartung, 2004), therefore, in the
field study during the long period of PRD treatment, the irri-
gation was shifted from one side of the plants to the other
every 5–10 d in order to achieve a long-term effect of PRD
on ABA signalling, leaf gas exchange, and WUE in potatoes.
Materials and methods
Plant material and growing conditions
In the GH, potato tubers (S. tuberosum L. cv. Folva, supplied by seed
potato company Solanum Nord, Brørup, Denmark) were planted in
pots (17 cm diameter and 50 cm deep). The pots were filled with 13.3
kg of sandy soil to 47.5 cm height with a bulk density of 1.27 g dry
weight cm
3. The soil had a volumetric soil water content (h) of
17.5% at full pot-holding capacity (PHC) and 4.0% at permanent
wilting point (PWP). The pots were evenly separated into two
compartments with plastic sheets such that water exchange between
the two compartments was prevented. A piece of plastic [width
(5 cm)3height (14 cm)] was removed from the middle of the sheet
where the seed tubers were planted. Before planting, the seed
potatoes were exposed to 12–14 8C with constant dim overhead light
for sprouting. During the germination phase, only one sprout was
allowed to emerge. The roots from this sprout were evenly distributed
between the two separated compartments. TDR (time domain
reflectometer; TRASE, Soil Moisture Equipment Corp., USA) probes
(33 cm in length) were installed in the middle of each soil column to
monitor the average soil water content in the compartments. The
climate conditions in the GH were: 20/1462 8C day/night air
temperature, 15 h photoperiod and >500 lmol m
2 s
1 photosyn-
thetic active radiation supplied by sunlight plus metal-halide lamps.
After emergence, all plants were well watered daily to 95% PHC with
nutrient solution (Pioneer NPK Macro 14-3-23 + Mg combined with
Pioneer Micro; pH=5.5; EC=1.3).
The field experiment was carried out in 2005 in South Jutland,
Denmark at the Government research station, Jyndevad (latitude
54899 N, longitude 98139 E). To avoid rain, the plants were grown
under an automatic rain-out-shelter. The soil is coarse textured
meltwater sand. The soil has a h of 16.1% at field capacity (FC) and
a h of 6.3% at PWP. Fertilizers were applied at rates of 150 N, 30 P,
220 K, 30 Mg, 200 S kg ha
1 8 cm below and 12.5 cm next to the row
where the seed tubers were later placed. Potato seed tubers (cv. Folva)
were treated with Monceren against fungal diseases and pre-heated
at 14–16 8C until 1 mm sprouts had emerged. The tubers were
planted on 12 April in rows (75 cm between rows and 25 cm between
seeds). Seed tubers were ridged with 20 cm of soil in prepared
furrows.
Irrigation treatments
In the GH, plants were kept well watered at a h of 17.5% during the
first 5 weeks after planting. Thereafter, plants were subjected to two
irrigation treatments: (i) full irrigation (FI) in which both soil
compartments were watered daily at 09.00 h to 17.5% to compensate
the full water loss during the previous day; and (ii) partial root drying
(PRD) in which half of the root system was watered to 17.5% while
the other was allowed to dry until the h of the dry soil had decreased
to 7%. The treatments lasted for 9 d. In the field, the plots (eight plots,
each of 5.2534 m) were encircled by a protection area of 1 m width.
The field trial was a complete randomized design with two subsurface
drip irrigation treatments (FI and PRD) and with four replicates from
tuber bulking to tuber maturing stages (i.e. 21 June–22 August). The
FI treatment was fully irrigated and kept h close to FC. The PRD
treatment received 70% of the irrigation water volume of FI at each
irrigation event and the irrigation was shifted between the two sides
of the plants (below denoted as the right and left side) every 5–10 d.
The irrigation water use during the treatment period was 201 mm for
FI plots and 140 mm for PRD plots. The drip lines used (NETAFIM,
Tel Aviv, Israel) had a distance of 25 cm and 50 cm between two
emitters in the FI and the PRD treatments, respectively. For the FI
treatment, one drip line was placed in the ridge 10 cm below the ridge
top; each emitter was close to the seed potato with 5.33 emitters m
2.
For the PRD treatment, two drip lines were placed in parallel in the
top of the ridge positioned similarly to the FI treatment except that
each emitter was placed midway between two seed potatoes, i.e. 12.5
cm from each seed potato. Before each irrigation event, h was
measured by TDR probes (77 cm in length). Three TDR groups were
installed vertically in the top of the ridge 12.5 cm left, 12.5 cm right,
and close to the plants. Daily average h in the root zone (60 cm
in depth determined from the flat soil surface when there was no
ridges) of FI and PRD plants was calculated as the TDR-measured h
before each irrigation event plus half of the h increment by irrigation
water input.
 Physiological responses of potato to partial root-zone drying 3729
Gas exchange, plant water potentials, xylem sap collection, plant, in order to obtain a sap flow rate similar to the transpiration rate
and ABA determinations of the plant. However, as the hydraulic resistance of the shoot was
removed, the flow rate could be higher than the actual transpiration
After the onset of treatments, in both the GH and the field, net
rate, and thus the ABA concentration in the sap might have been
photosynthetic rate (A) and stomatal conductance (gs) were measured
underestimated. A 0.5–1.0 ml aliquot of sap was collected using
daily on the second fully expanded upper canopy leaflets (three
a pipette from the cutting surface into an Eppendorf vial wrapped
replicates in the GH and 12 replicates in the field) from 11.00 h to
with aluminium foil over 3–5 min in FI plants and 5–7 min in PRD
13.00 h at an ambient CO2 concentration of ; 380 ll l
1 with a LI-
plants. The sap was immediately stored at
80 8C for ABA analysis.
6200 portable photosynthesis system (Li-Cor Inc., Lincoln, NE,
In the field, the xylem sap was collected using root pressure
USA). Leaf water potential (Wl) was measured after gas exchange
according to Bahrun et al. (2002). On days 15, 29, 35, 43, 50, and 56
measurements with a pressure chamber (Soil Moisture Equipment
after the onset of treatment (DAT) (these days are close to the ends of
Corp., Santa Barbara, CA, USA) on the same leaves from 12.00 h to
the drying cycles for one side of the roots), three plants per plot were
14.00 h on eight out of nine occasions and 20 out of 57 occasions in
detopped 3–5 cm above the soil surface at 22.00 h, the cutting surface
the GH and in the field, respectively. The daily values of vapour
was immediately wiped with absorbent tissue to remove contami-
pressure deficit (VPD) during gas exchange measurements in both
nants, then silicon tubing was fitted to the stems. The tubes were
experiments are shown in Fig. 1.
sealed with parafilm and wrapped in aluminium foil to prevent
In the GH, plants were harvested every 2 d. At each harvest, xylem
evaporation and protect against dust and insects. Over 2–3 h, 2–3 ml
sap was collected by pressurizing the roots of the potted plant in
of sap were collected. The sap was stored at
80 8C for ABA
a Scholander-type pressure chamber. The entire pot was sealed into
analysis.
the pressure chamber and the shoot was detopped at 15–20 cm from
The concentration of ABA in the xylem ([ABA]xylem) was
the stem base. With the stem stump protruding outside the chamber,
analysed without further purification by an enzyme-linked immuno-
pressure was applied until the root water potential (Wr) was equalized.
sorbent assay (ELISA) using a monoclonal antibody for ABA (AFRC
The cut surface was cleaned with pure water and dried with blotting
MAC 252) according to Asch (2000). No cross-reaction of the
paper. It is known that the concentration of ABA in the sap could be
antibody with other compounds in the xylem sap was detected when
affected by the balancing pressure applied (Jokhan et al., 1996).
tested according to Quarrie et al. (1988).
Thus, the pressure was increased gradually until it equalled Wl of the
Leaf area index, plant water use, tuber yield, and WUE
3.0 These parameters were determined only in the field experiment. Plant
a water use during the experimental period was calculated based on the
2.5
amount of irrigation and TDR soil moisture measurements. At the
beginning and during the treatments (1, 15, 30, 48, and 57 DAT),
plant leaf area was measured with a leaf area meter (model 3050A;
2.0 Li-Cor Inc.) and leaf area index was calculated as leaf area per unit
VPD (kPa)

ground area. At final harvests (57 DAT), tubers were collected, and
1.5 the tuber biomass was determined after oven-drying for 48 h at 80 8C.
WUE was calculated as the increment of tuber biomass divided by
1.0 the plant water use during the treatment period.

0.5
Data analysis and statistics
To facilitate data comparison between the two experiments, h in the
0.0
PRD drying side in both experiments was converted into the fraction
0 1 2 3 4 5 6 7 8 9 10 of transpirable soil water (FTSW), which was calculated by using
DAT Equation 1 (Sinclair and Ludlow, 1986):
FTSW = ðh--PWPÞ=ðPHC or FC--PWPÞ ð1Þ
1.2
b The measured values of [ABA]xylem in PRD plants were further
1.0 expressed relative to those of the FI plants yielding relative
[ABA]xylem. The dependence of the relative [ABA]xylem on FTSW
of the PRD drying side was analysed by using the NLIN regression
0.8
procedure (SAS Institute Inc., 1999–2001). Similar, the NLIN proce-
VPD (kPa)

dure was also applied to analyse the relationship between [ABA]xylem

0.6 and Wr. Data were subjected to analysis of variance (ANOVA)
procedures (SAS Institute Inc., 1999–2001). Appropriate standard
0.4 errors of the means (SE) and least significant differences (LSDs) at
P=0.05 were calculated. Tukey’s studentized range (HSD) test was
0.2 applied to detect significant effect of treatment.

0.0
0 5 10 15 20 25 30 35 40 45 50 55
Results
DAT
Soil water dynamics
Fig. 1. Vapour pressure deficit (VPD) in the greenhouse (GH) (a) and in Changes of h in FI- and PRD-treated plants during the
the field (b) during gas exchange measurements in potato subjected to full
irrigation (FI) and partial root-zone drying (PRD). DAT denotes days experimental periods are shown in Fig. 2. In the GH at the
after onset of treatment. onset of PRD, h of the drying side declined quickly during
3730 Liu et al.
22 lower than that of FI plants on six out of 24 occasions,
a while on other days during the treatment period it was
19
similar to that of FI plants. In the GH, gs of PRD plants was
16
similar to that of FI plants on seven out of nine occasions;
only on two occasions (6–7 DAT) was it significantly less
θ (vol.%)

13 than that of FI plants (Fig. 3c). In the field, gs of PRD plants

was significantly less than that of FI plants on 10 out of 24
10 occasions during the treatment period (Fig. 3g). Intrinsic
FI WUE was significantly higher in PRD plants as compared
7 PRD-L with FI on two out of nine and five out of 24 occasions in
PRD-R the GH and in the field, respectively. On other days, it was
4
-1 0 1 2 3 4 5 6 7 8 9 10 similar for the two treatments except for 1–2 cases (Fig. 3d,
DAT h). Figure 4 shows the relationship between the intrinsic
WUE (A/gs) and gs, and there was a clear tendency of
20
increasing A/gs with decreasing gs.
b
18 In the GH, Wr of PRD-treated plants decreased steadily
and became significant lower than that of FI plants at
16 5 DAT (Fig. 5a). [ABA]xylem was ;60 pmol ml
1 in FI
θ (vol.%)

plants (Fig. 5b). Under PRD, [ABA]xylem increased from 5

14
12
10
8 of PRD plants on the first three out of six occasions (i.e. at
-5 0 5 10 15 20 25 30 35 40
DAT
Fig. 2. Developments of the average soil volumetric water content (h)
in pots under FI and PRD irrigation treatments in the GH (a, n=3–12)
and in the field (b, n=4). Vertical bars indicate LSDs (P=0.05) and
asterisks denote significant difference among the treatments at P <0.05.
Arrows in (b) indicate the time of shifting irrigation from one side of
the plant to the other side.
the first 4 d and became slower thereafter, and was about
7% by the end of the treatment. Although the h of the PRD
wet side was generally lower than that of the FI pots (which
maintained an average h above 16%), there was no statis-
tical difference between them. In the field, FI plants were
always irrigated at 1–1.5% less than FC (h=16.1%) (Fig.
2b); before each shift of irrigation, h of the PRD wet side
reached that of the FI except for the periods of 20–28 and
35–44 DAT. The differences of h between the PRD dry and
wet sides were significant during the first three shifts
and subsequently were smaller (Fig. 2b).
Time-courses of shoot physiology and [ABA]xylem
In the GH, Wl was basically the same for FI and PRD plants
on eight out of nine occasions during the treatment period
(Fig. 3a); while in the field, Wl of PRD plants was
significantly lower than that of FI plants on five out of 20
occasions (Fig. 3e). Figure 3b and f show the day-to-day
changes of A under FI and PRD treatments. In the GH, only
at 5 DAT was A of PRD plants significantly lower than that
of FI plants. In the field, A of PRD plants was significantly
DAT, and was ;9-fold that of the FI plants by the end of
treatment (Fig. 5b). There was a clear exponential relation-
ship between [ABA]xylem and Wr (r2=0.99, P <0.001)
(Fig. 6). In the field, [ABA]xylem of the FI plants was
;200 pmol ml
1, which was significantly lower than that
45 50 55 60
15, 29, and 35 DAT) (Fig. 7). When the [ABA]xylem in the
PRD plants was expressed relative to that of FI, a common
relationship between relative [ABA]xylem and FTSW of
the two experiments was observed, being that relative
[ABA]xylem increased exponentially (r2=0.98, P <0.001)
with decreasing FTSW (Fig. 8).
Leaf area index, biomass production, plant water use,
and WUE
These parameters were determined only in the field
experiment. During the PRD treatment period, the leaf
area index increased from about 3.1 to 4.2, and then
decreased to 3.2 at the last harvest. The leaf area index of
the PRD treatment tended to be lower than that of the FI
treatment until the fourth harvest, while it was slightly
higher than that of the FI treatment in the last harvest
(Fig. 9). Tuber biomass was recorded in the last harvest, and
no significant differences were found between the treat-
ments (Table 1). Plant water use of PRD plants during the
experimental period was 30% less than that of FI plants
(Table 1). WUE, calculated based on tuber biomass pro-
duction and plant water use during the experimental period,
was 58.7% greater in PRD than in FI plants (Table 1).
Discussion
In the GH, h of the wet side of the PRD plants was kept
close to PHC, and Wl of the plants was similar to that of FI
plants (Figs 2a, 3a). This is in line with the basic
assumption of PRD that Wl could be largely maintained
 Physiological responses of potato to partial root-zone drying 3731
-0.2
a e
* * * * **
-0.4

Ψl (MPa)
-0.6

-0.8

-1.0 FI
PRD
-1.2
b f
20
A (µmol m-2 s-1)

15

10

5
* * * * * ** * * * *
0
1.2
c g
gs(mol m-2 s-1)

1.0
0.8
0.6
0.4
0.2
* * * *** * * * * * * * * ** *
0.0
A/gs (mmol CO2 mol-1 H2O)

d h
36

27

18

9
* * * * ** * * * *
0
0 1 2 3 4 5 6 7 8 9 0 5 10 15 20 25 30 35 40 45 50 55
DAT DAT

Fig. 3. Changes of midday leaf water potential (Wl) (a, e), photosynthesis (A) (b, f), stomatal conductance (gs) (c, g), and intrinsic water use efficiency
(WUE, A/gs) (d, h) of potato under full irrigation (FI) and partial root-zone drying (PRD) in the GH (a, b, c, d) and in the field (e, f, g, h). Data points are
means of three and 12 replicates in the GH and in the field, respectively. Error bars indicate the SE and asterisks denote significant difference between
treatments at P <0.05.

(Sobeih et al., 2004). However, a decrease of Wl in the PRD
plants was found in the field, indicating that soil water
stress around the root system was severe enough to cause
a reduction in Wl (Figs 2b, 3e). It is notable that a low Wl
was often observed during the days when the wet side of the
root system had a low h, and which was not significantly
higher than that of the drying side (Fig. 2b). Therefore,
keeping a high h of the wet side in PRD may be essential to
maintain a high Wl of the plants. This is supported by recent
findings of Zegbe et al. (2004, 2006), who demonstrated
that the amount of water given to the wet side was impor-
tant in maintaining a high Wl in PRD-treated processing
tomatoes. In the present study, in order to maintain h of the
PRD wet side close to that of FI, in the GH ;83% water
used for FI was applied to the PRD plants; while in the field,
70% of water used for FI was applied to the PRD plants and
h of the wet side was significantly lower than that of the FI
treatment on many days during the treatment. A similar
pattern of soil water dynamics under PRD irrigation has
been reported by other authors (Kirda et al., 2004,
2005). The causes for this are not clear; therefore, in future
studies with PRD irrigation, more attention should be paid
to soil physical properties and soil water distribution in the
root zone. Moreover, Mingo et al. (2004) showed that h of
the PRD wet side could be maintained as high as in the FI
when the same amount of water of FI was applied, and Wl
was similar for PRD and FI plants. However, Sobeih et al.
(2004) observed that, with half the amount of irrigation, the
h of the PRD wet side was still kept close to that of FI and
the plants could maintain the same Wl as that of FI plants.
3732 Liu et al.
45 0.00
FI-FIELD a
PRD-FIELD
A/gs (µmol CO2 mol-1 H2O)

FI-GH
35 PRD-GH -0.05

Ψr (MPa)
-0.10
25

-0.15
15 FI
PRD * *
-0.20

5 b
600 * *
0.2 0.4 0.6 0.8 1.0 1.2 1.4

[ABA]xylem (pmol ml-1)

gs (mol m-2 s-1)
400
Fig. 4. Relationships between intrinsic WUE (A/gs) and gs in potatoes
subjected to FI and PRD irrigation in the GH and in the field. Data points
are means of three and 12 replicates in the GH and in the field,
respectively. Error bars indicate the SE. 200

The reasons for this discrepancy are not known. Putative 0

0 1 2 3 4 5 6 7 8 9 10
genotypic differences in the mechanisms of controlling
DAT
plant water use and leaf water status (e.g. isohydric versus
anisohydric behaviour) and differences in experimental Fig. 5. Development of root water potential (Wr) (a) and xylem sap
conditions may be involved. abscisic acid concentration ([ABA]xylem) (b) in potato subjected to FI and
PRD irrigation in the GH. Data points are means of three replicates. Error
In the GH, although Wl was largely maintained, Wr of the bars indicate the SE and asterisks denote significant difference between
plants decreased steadily with decreasing h of the drying treatments at P <0.05.
side in PRD treatment (Fig. 5a). A decrease of predawn Wl
(which may have a similar value to Wr) while maintaining
800
midday Wl has also been observed in PRD-treated cotton
(Gossypium hirsutum) plants (Tang et al., 2005). These
[ABA]xylem (pmol ml-1)

results indicate that root hydraulic status had been affected 600
in PRD-treated plants. By decreasing Wr, the plants may be
able to maintain a substantial water potential gradient
400
between the roots and the soil, so that the rates of water
uptake might be largely sustained in response to partial soil
drying. Besides, for both experiments, it is noted that, 200
compared with Wl, gs seemed to be more sensitive to PRD. y = 51.56e-16.51x
This was particularly clear in the field experiment, where gs r 2 = 0.99***
was significantly lowered in 10 out of 24 instances during 0
0 -0.04 -0.08 -0.12 -0.16 -0.2
the PRD treatment period (Fig. 3g). It is proposed that Ψr
lowered Wl may have acted as a hydraulic signal partly
closing stomata; however, chemical signals, mainly ABA, Fig. 6. [ABA]xylem expressed as a function of Wr in potato subjected to
PRD treatment. Data points are means of three replicates and error bars
produced in the root tips of the PRD plants are most likely to indicate the SE.
have caused stomatal closure on days with similar Wl in the
two treatments. This seems to be true as in both experiments
an increase of [ABA]xylem in PRD plants was observed, the two parameters for potatoes grown under progressive
even though on only three out of six sampling occasions in soil drying was reported previously (Liu et al., 2005). Such
the field was the increase significant (Figs 5b, 7). In kinds of relationships are important in developing mecha-
agreement with this, in a previous study with the same nistic models to predict ABA production and its influence
cultivar of potato, we found that when Wr >–0.3 MPa, gs on gs in plants under different irrigation regimes. However,
was largely controlled by [ABA]xylem (Liu et al., 2005). it should be borne in mind that the relationship between Wr
Furthermore, it is shown here for the first time that and [ABA]xylem was obtained based on the data from the
[ABA]xylem increased exponentially with decreasing Wr in GH experiment in which no alternation of irrigation was
PRD plants in the GH (Fig. 6). A linear relationship between done in the PRD treatment. Therefore, it may not be
 Physiological responses of potato to partial root-zone drying 3733
600 4.5
FI FI
500 PRD PRD
[ABA]xylem (pmol ml-1)

4.0

Leaf area index

400

300 3.5

200
3.0
100
* * *
0 2.5
0 10 20 30 40 50 60 0 10 20 30 40 50 60
DAT DAT

Fig. 7. [ABA]xylem of field-grown potato subjected to FI and PRD Fig. 9. Development of the leaf area index in field-grown potato
treatments. Data points are means of four replicates and error bars indicate subjected to FI and PRD treatments. Data points are means of 12
the SE. Asterisks denote significant difference between treatments at replicates and error bars indicate the SE.
P <0.05.

10 Table 1. Effects of PRD on tuber biomass production, water

GH use, and water use efficiency (WUE) in potato grown in the field
FIELD
8 under a rain-out-shelter
Relative [ABA]xylem

Data are the means 6SE of 24 replicates. Values with the same letters
6 y = 0.86 + 40.20e-7.89x within a column are not significantly different at P <0.05 according to
r 2 = 0.98*** Tukey’s HSD test.

4 Treatment Tuber biomass Plant water use WUE (g l
1)

production (g plant
1) (l plant
1)
2 FI 231.94614.37 a 37.6660.26 a 6.1660.46 b
PRD 257.87620.65 a 26.3860.34 b 9.7960.95 a
0
1.0 0.8 0.6 0.4 0.2 0.0
FTSW

Fig. 8. Exponential relationship between relative [ABA]xylem (defined

as the ratio between [ABA]xylem of the PRD plants and that of the FI
plants) and the fraction of transpirable soil water (FTSW) of the PRD
drying side of potatoes grown in the GH and in the field. The calculation
of FTSW is defined in Equation 1. Data points are means of three and 12
replicates in the GH and in the field, respectively. Error bars indicate
the SE.
applicable in the field study where irrigation was shifted
several times between the two sides of the root system.
Nevertheless, according to this relationship, it is considered
that the increase of h (Fig. 2b) and thus Wr in the latter stages
of the PRD treatments in the field may explain the
convergence of [ABA]xylem between the two treatments.
It is noteworthy that [ABA]xylem of FI plants in the field
was much higher than that observed in the GH. It is
suggested that the method for collecting the xylem sap may
be critical. In the field, the sap was collected using natural
root pressure from 22.00 h to 24.00 h in the absence of
transpiration, the flow rate of the sap was very slow, and the
sap may therefore be concentrated, resulting in a relative
high [ABA]xylem; whereas in the GH, the sap was collected
by pressurizing the roots at a pressure equivalent to Wl
during the day, the flow rate of the sap was much higher
than that during night under natural root pressure, so
that the sap might have been diluted, leading to a relative
low [ABA]xylem. However, when [ABA]xylem of PRD
plants was expressed relative to that of FI plants and h of
the PRD drying side was converted into FTSW, there was
a common exponential relationship between the two
parameters (Fig. 8). This relationship clearly indicates
that ABA production from the root tips under PRD
treatment was directly related to the soil water availability
in the drying side. Accordingly, the time of alternating of
irrigation between the two sides may be defined as when the
FTSW of the drying side had decreased to a certain level
that may induce a sustainable ABA signal. However, one
should be aware that the roots may extract a substantial
amount of water from the PRD wet side as shown in the
GH, and ABA produced by the dry roots might have
been diluted. Therefore, it was considered that [ABA]xylem
detected from the base of the stem may be associated much
more closely with the integrated soil water deficits in the
whole root zone rather than with the soil water status in part
of the root system alone.
Moreover, gs is controlled not only by the soil drying-
induced ABA signalling, but climate conditions, plant
growth stage, and genotypic factors are also involved. In
the present study, gs might have been partly regulated by
VPD, particularly in the GH. It was clear that the lower the
VPD during the day, the smaller the differences of gs were
3734 Liu et al.
between the treatments (Figs 1, 3c, g). Similar results have
been reported by Maurel et al. (2004) in PRD-treated
chestnut (Castanea sativa) saplings.
It is obvious that, in both experiments, A was less sensitive
to PRD treatment than gs (Fig. 3b, c, f, g). As a consequence
of this, intrinsic WUE (i.e. A/gs) was increased under PRD.
This was the case as shown in Fig. 3d and h, where intrinsic
WUE was significantly higher in PRD plants as compared
with FI on two out of nine and five out of 24 occasions in
the GH and in the field, respectively. It is well known that,
due to the non-linearity relationship between A and gs,
partial stomatal closure would lead to an increase in
intrinsic WUE (Jones, 1992). We have previously shown
a linear increase of intrinsic WUE with decreasing gs in
potatoes exposed to progressive soil drying (Liu et al.,
2005). In the present study, data from both experiments
showed a clear trend of increasing intrinsic WUE when gs
decreased (Fig. 4), confirming earlier findings with the
same cultivar (Liu et al., 2005). Based on these results, it is
believed that PRD-induced ABA signalling reduces gs,
resulting in an increase of intrinsic WUE, which may
ultimately lead to an increase in crop WUE.
A large body of evidence has shown that reduction in
leaf size is the first morphological response to soil water
deficits in potatoes (Jefferies and MacKerron, 1989). In the
present study, mild soil water deficits developed in the pot
and in the field under PRD treatments (Fig. 2a, b), and the
leaf area index was slightly less in the PRD treatment
during most of the season (Fig. 9). Similarly, Saeed et al.
(2005) reported that PRD treatment significantly reduced
plant leaf area in potato. As Wl was maintained on most
days during the treatment, it was speculated that soil drying
signals including ABA under PRD irrigation might partly
account for the reduced leaf area index. A similar
explanation for the restriction of leaf expansion in PRD-
treated plants has been proposed in grapevine (Stoll et al.,
2000), cotton (Tang et al., 2005), and other crops (Kang
and Zhang, 2004). A smaller green leaf area will not only
reduce crop transpiration, but also reduce crop photosyn-
thetic rate due to less light interception. However, in
potatoes, a leaf area index of 3 is sufficient for the
interception of 90–95% of the incoming radiation, which
can maintain the maximum growth rate (Khurana and
McLaren, 1982). In the present study in the field, the
average leaf area index during the PRD treatment was ;3.3
(Fig. 9), which may significantly reduce crop transpira-
tion but can still maintain crop photosynthesis and growth
rate. This may partly explain why PRD treatment could
maintain tuber yield (Table 1) even though the irrigation
water was reduced by 30% (Table 1). As a result of yield
maintenance and reduced plant water use, crop WUE of
PRD treatment was 59% higher than that of FI treatment
(Table 1).
Collectively, the present results suggest that restrictions
on stomatal opening and leaf expansion by PRD-induced
ABA signals might have contributed to the reduced water
use and improved WUE in potatoes. However, beside these
effects, other physiological modifications caused by PRD
such as an enhanced root system and an increased ability to
take up soil nutrients (e.g. nitrogen) might also play a role
in improving WUE (Kirda et al., 2005). Despite this, it can
be concluded that PRD is a promising water-saving
irrigation strategy for potato production in areas with
limited water resources. However, to optimize the PRD
irrigation strategy for the crop, more issues should be
studied in the future: (i) how can [ABA]xylem be predicted
with a partial decrease of h in the root zone; (ii) is there
a causal relationship between [ABA]xylem and gs under
PRD; and (iii) how does PRD affect root growth and soil
nutrient availability?
Acknowledgements
FL thanks the Danish Research Council (SJVF, 23-03-0208) for the
financial support for his postdoctoral research. This study was also
partly supported by the European Commission (WATERWEB, EU,
INCO-CT-2004-509163 and SAFIR, EU, FOOD-CT-2005-023168).
References
Asch F. 2000. Determination of abscisic acid by indirect enzyme
linked immunosorbent assay (ELISA). Technical Report. Taastrup,
Denmark: Laboratory for Agrohydrology and Bioclimatology,
Department of Agricultural Sciences, The Royal Veterinary and
Agricultural University.
Bahrun A, Jensen CR, Asch F, Mogensen VO. 2002. Drought-
induced changes in xylem pH, ionic composition, and ABA
concentration act as early signals in field-grown maize (Zea
mays L.). Journal of Experimental Botany 53, 251
263.
Davies WJ, Hartung W. 2004. Has extrapolation from biochemistry
to crop functioning worked to sustain plant production under water
scarcity? In: Proceedings of the 4th International Crop Science
Congress, 26 September–1 October 2004, Brisbane, Australia.
Davies WJ, Wilkinson S, Loveys BR. 2002. Stomatal control by
chemical signalling and the exploitation of this mechanism to
increase water use efficiency in agriculture. New Phytologist 153,
449–460.
Davies WJ, Zhang JH. 1991. Root signals and the regulation of
growth and development of plants in drying soil. Annual Review of
Plant Physiology and Plant Molecular Biology 42, 55–76.
Dorji K, Behboudian MH, Zegbe-Domı́nguez JA. 2005. Water
relations, growth, yield, and fruit quality of hot pepper under deficit
irrigation and partial rootzone drying. Scientia Horticulturae 104,
137–149.
Fabeiro C, Martı́n de Santa Olalla F, de Juan JA. 2001. Yield and
size of deficit irrigated potatoes. Agricultural Water Management
48, 255–266.
Jefferies RA. 1993. Responses of potato genotypes to drought. 1.
Expansion of individual leaves and osmotic adjustment. Annals of
Applied Biology 122, 93–104.
Jefferies RA, Mackerron DKL. 1989. Radiation interception and
growth of irrigated and drought-stressed potato (Solanum tuber-
osum). Field Crops Research 22, 101
112.
Jokhan AD, Else MA, Jackson MB. 1996. Delivery rates of abscisic
acid in xylem sap of Ricinus communis L. plants subjected to part-
drying of the soil. Journal of Experimental Botany 47, 1595–1599.
 Physiological responses of potato to partial root-zone drying
Jones HG. 1992. Plants and microclimate: a quantitative approach
to environmental plant physiology, 2nd edn. Cambridge:
Cambridge University Press.
Kang SZ, Hu XT, Goodwin I, Jerie P, Zhang J. 2002. Soil water
distribution, water use and yield response to partial rootzone
drying under flood-irrigation condition in a pear orchard. Scientia
Horticulturae 92, 277–291.
Kang SZ, Liang ZS, Hu W, Zhang JH. 1998. Water use efficiency
of controlled alternate irrigation on root-divided maize plants.
Agricultural Water Management 38, 69–76.
Kang SZ, Liang ZS, Pan YH, Shi PZ, Zhang JH. 2000. Alternate
furrow irrigation for maize production in an arid area. Agricultural
Water Management 45, 267–274.
Kang SZ, Zhang JH. 2004. Controlled alternate partial root-zone
irrigation: its physiological consequences and impact on water use
efficiency. Journal of Experimental Botany 55, 2437–2446.
Khurana SC, McLaren JS. 1982. The influence of leaf area, light
interception and season on potato growth and yield. Potato
Research 25, 329–342.
Kirda C, Cetin M, Dasgan Y, Topcu S, Kaman H, Ekici B,
Derici MR, Ozguven AI. 2004. Yield response of greenhouse-
grown tomato to partial root drying and conventional deficit
irrigation. Agricultural Water Management 69, 191–201.
Kirda C, Topcu S, Kaman H, Ulger AC, Yazici A, Cetin M,
Derici MR. 2005. Grain yield and N-fertilizer recovery of maize
under deficit irrigation. Field Crops Research 93, 132–141.
Liu F, Jensen CR, Shahnazari A, Andersen MN, Jacobsen SE.
2005. ABA regulated stomatal control and photosynthetic water
use efficiency of potato (Solanum tuberosum L.) during pro-
gressive soil drying. Plant Science 168, 831–836.
Maurel M, Robin C, Simonneau T, Loustau D, Dreyer E,
Desprez-Loustau M-L. 2004. Stomatal conductance and root-to-
shoot signalling in chestnut saplings exposed to Phytophthora cin-
namomi or partial soil drying. Functional Plant Biology 31, 41–45.
Mingo DM, Theobald JC, Bacon MA, Davies WJ, Dodd IC. 2004.
Biomass allocation in tomato (Lycopersicon esculentum) plants
3735
grown under partial rootzone drying: enhancement of root growth.
Functional Plant Biology 31, 971–978.
Porter GA, Opena GB, Bradbury WB, McBurnie JC, Sisson JA.
1999. Soil management and supplemental irrigation effects on
potato. I. Soil properties, tuber yield, and quality. Agronomy
Journal 91, 416–425.
Quarrie SA, Whitford PN, Appleford NE, Wang TL, Cook SK,
Henson IE, Loveys BR. 1988. A monoclonal antibody to (S)-
abscisic acid: its characterisation and use in a radioimmunoassay
for measuring abscisic acid in crude extracts of cereal and lupin
leaves. Planta 173, 330–339.
Saeed H, Grove IG, Kettlewell PS, Hall NW. 2005. Potato root
and shoot growth under different water management strategies.
Aspects of Applied Biology 73, 85–91.
SAS Institute Inc. 1999–2001. Release 8.2 edn. SAS Inst., Inc.,
Cary, NC.
Sinclair TR, Ludlow MM. 1986. Influence of soil water supply on
the plant water balance of four tropical grain legumes. Australian
Journal of Plant Physiology 13, 329–341.
Sobeih WY, Dodd IC, Bacon MA, Grierson D, Davies WJ. 2004.
Long-distance signals regulating stomatal conductance and leaf
growth in tomato (Lycopersicon esculentum) plants subjected to
partial root-zone drying. Journal of Experimental Botany 55,
2353–2363.
Stoll M, Loveys BR, Dry P. 2000. Hormonal changes induced by
partial root-zone drying of irrigated grapevine. Journal of Experi-
mental Botany 51, 1627–1634.
Tang LS, Li Y, Zhang JH. 2005. Physiological and yield responses
of cotton under partial rootzone irrigation. Field Crops Research
94, 214–223.
Zegbe JA, Behboudian MH, Clothier BE. 2004. Partial rootzone
drying is a feasible option for irrigating processing tomatoes.
Agricultural Water Management 68, 195–206.
Zegbe JA, Behboudian MH, Clothier BE. 2006. Responses of
‘Petopride’ processing tomato to partial rootzone drying at
different phenological stages. Irrigation Science 24, 203–210.