Field Pea Transpiration and Leaf Growth in Response to Soil Water Deficits

Abstract

No quantitative functions exist to describe transpiration rates and leaf expansion rates of drought-stressed field pea (Pisum sativum L.) plants relative to well-watered plants. Such functions are especially important in analyses of the effects of water deficit periods on crop yield under a range of field conditions. Pot and field experiments were conducted to develop and evaluate drought-stress functions for transpiration and leaf area expansion. These functions were obtained in a dehydration cycle where plant response was monitored as the soil dried progressively. The level of water deficit was characterized as the fraction of transpirable soil water (FTSW). Transpiration and leaf area expansion did not change appreciably until FTSW reached =0.4, and then decreased linearly through FTSW equal to 0. Drought response functions obtained for transpiration and leaf area expansion in a pot study also well represented the results obtained in independent pot and field studies.

Full text

Field
Pea
Transpiration
and
Leaf Growth
in
Response
to
Soil Water Deficits
J.
Lecoeur
and T. R.
Sinclair*
ABSTRACT
No
quantitative functions
exist
to
describe transpiration
rates
and
leaf expansion
rates
of
drought-stressed
field
pea
(Pisum
sativum
L.)
plants relative
to
well-watered plants. Such functions
are
especially
important
in
analyses
of the
effects
of
water deficit
periods
on
crop
yield under
a
range
of
field
conditions.
Pot and
field experiments
were conducted
to
develop
and
evaluate
drought-stress
functions
for
transpiration
and
leaf area
expansion.
These functions were obtained
in
a
dehydration cycle where plant
response
was
monitored
as the
soil
dried
progressively.
The
level
of
water deficit
was
characterized
as
the
fraction
of
transpirable soil water (FTSW). Transpiration
and
leaf
area expansion
did not
change appreciably until FTSW reached
=
0.4,
and
then
decreased
linearly through FTSW equal
to 0.
Drought
re-
sponse
functions obtained
for
transpiration
and
leaf area expansion
in
a pot
study
also
well represented
the
results obtained
in
independent
pot and
field
studies.
S
OIL
WATER
DEFICITS
are
common
in the
production
of
most crops,
and
they
can
have substantial negative
impacts
on
growth
and
development. Plant
processes
that
depend
on
increases
in
cell volume
are
particularly
sensitive
to
water deficits.
Two
important examples
of
these sensitive processes
are
leaf
gas
exchange, which
depends
on
guard cell volume,
and
leaf area increase,
which
depends
on
cell expansion. Inhibition
of
these
processes under drought
can
result
in
substantial losses
in
yield.
Progress
in
developing quantitative response
functions
to
soil water
deficits
has
been slow. Part
of the
problem
may
result
from
the
fact
that many studies have attempted
to
characterize water deficits with thermodynamic vari-
ables (Sinclair
and
Ludlow, 1985). These thermodynamic
variables have
not
related directly
to
leaf
gas
exchange
(e.g., Bennett et al., 1987) or leaf expansion (Joly and
Hahn, 1989).
As an
alternative, Ritchie (1981) proposed
J.
Lecoeur, Laboratoire d'Ecophysiologie
des
Plantes sous Stress Environ-
nementaux
(LEPSE),
United'AgronomieINRA-ENSA
M,
34060Montpel-
lierCedex
1,
France; andT.R.
Sinclair,
USDA-ARS, Agronomy Physiol-
ogy
Lab., Bldg. 164,
Univ.
of
Florida, P.O.
Box
110840, Gainesville,
FL
32611-0840. Mention
of a
trademark
or
proprietary product does
not
constitute
a
guarantee
or
warranty
of the
product
by the
U.S. Department
of
Agriculture
and
does
not
imply
approval
or the
exclusion
of
other
products that
may
also
be
suitable. Received
10
Apr. 1995. "Corresponding
author
(aksch@gnv.ifas.ufl.edu).
Published
in
Crop Sci.
36:331-335
(1996).
that
the
responses
of
physiological processes
to
water
deficits
could
be
evaluated
as a
function
of
available soil
water. This concept
was
refined
by
Sinclair
and
Ludlow
(1986)
to
express
these
processes
as
functions
of the
fraction
of
transpirable soil water (FTSW) remaining
in
the
soil. They defined total transpirable soil water
as the
difference
between
the
soil water content
at
field
or pot
capacity
and the
soil water content when transpiration
of
the
drought-stressed plants decreased
to 10% or
less
of
that
of
well-watered plants.
The use of
FTSW
has led to
fairly
consistent response
functions
to
soil
dehydration
across
a
range
of
conditions.
Transpiration
was
shown
to be
unaffected
by
soil drying
until
FTSW decreased
to =
0.25
to
0.35
in
several grain
legumes (Sinclair and Ludlow, 1986). This same re-
sponse pattern
for
transpiration
was
reported
for a
num-
ber of
plant species
and
experimental conditions (Ritchie
et
al., 1972; Ritchie, 1973; Meyer
and
Green, 1981;
Gollan
et
al., 1986; Rosenthal
et
al., 1987; Kuppers
et
al., 1988).
In
addition,
a
similar response pattern
was
shown
for
leaf area development
in
soybean
[Glycine
max
(L.) Merr.; Sinclair,
1986],
cowpea
[Vigna
unguicu-
lata
(L.) Walp.]
and
black gram
[Vigna
mungo (L.)
Hepper; Sinclair
et
al., 1987], sorghum
[Sorghum
bicolor
(L.) Moench]
and
cotton
(Gossypium
hirsutum
L.;
Rosen-
thal
et
al., 1987),
and
maize (Zea mays
L.;
Muchow
and
Sinclair, 1991). In these cases, leaf area expansion
started
to
decline
at
about
the
same
or
slightly higher
FTSW than transpiration,
but
leaf area expansion tended
to
decrease more rapidly with
further
soil drying
and
became zero
at a
FTSW greater than zero.
Although
field
pea
yields
are
susceptible
to
drought
(Salter, 1963; Maurer
et
al., 1968; Pumphrey
and
Schwanke,
1974), little
information
exists
on the
relation-
ship
between water deficits
and
either transpiration
or
leaf
area expansion.
The
objective
of
this study
was to
characterize
for
field
pea
these processes
as a
function
of
FTSW.
The
approach
was to
develop FTSW
functions
for
transpiration
and
leaf area development
in a
green-
house
study
and
then test these results
in
independent
pot and
field
experiments.
Abbreviations: FTSW,
fraction
transpirable soil water; TTSW,
total
transpirable
soil water; NLA, normalized leaf area expansion; NTR,
normalized
transpiration rate. "Significant
at the
0.01 probability level.

332
CROP SCIENCE, VOL. 36, MARCH-APRIL 1996
MATERIALS AND METHODS
Greenhouse Experiment
This experiment was conducted in a greenhouse in Gainesville,
FL, during December 1994 and January 1995. Greenhouse
temperature was controlled between 20 and 30°C, and relative
humidity was maintained at -~ 60 %. Two cultivars of field pea
(Frilene and Atol) were grown in 3-L pots filled with a potting
mix consisting of two-thirds potting soil and one-third 1:1
mixture of peat and vermiculite. All pots were inoculated with
N-fixing bacteria and supplied with a complete nutrient solution
during the initial growth period prior to the beginning of the
dehydration experiment.
Measurements of transpiration and leaf area expansion was
begun when the eighth or ninth leaves were fully expanded.
On the evening prior to the initiation of the measurements,
all pots were fully watered and allowed to drain overnight.
The following morning, the pots were weighed to determine
the initial pot capacity for soil water. To prevent soil evapora-
tion, the pots were enclosed in plastic bags by sealing the bag
at the base of the plant stem. For each cultivar, 12 pots
were kept well watered and 12 pots were allowed to dry.
(Observations from six of the well-watered Frilene plants were
discarded because of obvious symptoms of hypoxia in the first
few days of the experiment.) All pots were weighed daily at
= 16:00 to allow calculation of daily transpiration rates and
soil water content. The well-watered pots were irrigated to
return the soil water content to 200 g less than the pot capacity
weight each afternoon.
The total transpirable soil water (TTSW) was determined
for each pot according to the definition of Sinclair and Ludlow
(1986) as the difference between the initial pot capacity and
the weight when the daily transpiration rate decreased to < 10 %.
of the control plants. The daily value of FTSW was estimated
as the ratio between the amount of transpirable soil water still
remaining in the pot and TTSW.
In field pea, the appearance rate of leaves is virtually un-
affected by water deficits (Lecoeur, 1994) so that the main
effect of drought is to restrict leaf area expansion. Leaf area
expansion was determined each day by measurements of indi-
vidual leaf dimensions. The length of one of the two stipules
and one leaflet of each pair (from 1 to 3 depend on leaf position
on the stem) were measured daily for each leaf from emergence
from apical bud to the completion of its expansion. The area
of individual leaves was calculated as the sum of stipule and
leaflet areas (mm
2)
from measurements of stipule and leaflet
lengths (mm), respectively, with the following equations:
Leaflet area =
117.8(leaflet length)
2
+ 54.5 (r
2
= 0.97**) [1]
Stipule area =
75.7(stipule length)
2
+ 28.1 (r
2
= 0.98**) [2]
Equations [1] and [2] were established on 400 stipules and
400 leaflets by a LI-3000 portable area meter and a LI-3050A
transparent belt conveyer (LI-COR, Lincoln, NE). Regression
between measured and calculated leaf areas established on a
second independent set of leaves (n = 60) resulted in an
2
of 0.96** and a coefficient of variation of errors of 7.8%.
Daily individual leaf area expansion was calculated as the
difference between current leaf area and leaf area on the
previous day.
Daily transpiration and leaf area expansion for the individual
drought-stressed plants were expressed as ratios between the
values obtained for stressed plants and the means of plants
kept well watered. By calculating ratios, variations in plant
performance among days resulting mainly from differences in
solar irradiance, could be minimized in the results obtained
from the drought-stressed plants. The calculated ratios of tran-
spiration and leaf area expansion were plotted against FTSW.
However, due to variation in the initial size among individual
drought-stressed plants, there was some scatter among the
treatment plants even under well-watered conditions. To elimi-
nate this variation, the ratio data for each plant were normalized
to give a mean ratio value of 1.0 for all observations made
while the FTSW of the individual stressed plant was still >0.6.
Therefore, calculation of normalized transpiration rates (NTR)
and normalized leaf area expansion (NLA) allowed the data
from each stressed plant to be centered on a value of 1.0 when
the soil was still moist.
Montpellier Pot Experiments
Two pot experiments were conducted in a greenhouse in
Montpellier, France, from January to April 1992 and from
January to April 1993 with cultivars Frilene and Atol, respec-
tively. Six plants were grown in each of the pots (0.30
diam. and 0.70 m high). The pots were filled with a soil
mixture composed of sand, loam, clay, and organic matter in
the proportion, respectively, of 0.45, 0.27, 0.17, and 0.11 in
the 1992 experiment and 0.35, 0.28, 0.27, and 0.10 in the
1993 experiment. The nutrient availability from the soil was
high, and the soil was inoculated with N-fixing bacteria.
In both experiments, three water regimes were imposed on
12 pots each, a well-watered control and two drought-stressed
treatments. The well-watered control was maintained at a
fraction of available soil water of >0.8. In the driest drought-
stressed treatment, no water was added to the pots between
floral initiation and the beginning of flowering, which lasted
28 and 36 d for the 1992 and 1993 experiments, respectively.
In the intermediate drought-stressed treatment, the bottom
portion of the pots were well watered while the soil in the top
portion was allowed to dry. Four tubes (12 mm diam.) were
installed in each of the pots in this treatment to a depth of
0.4 m. The lower portion of these pots were watered daily
through these tubes.
Volumetric water content was calculated for each pot by
measuring soil water potential 0g~oil) with tensiometers posi-
tioned at 0.25-, 0.45-, and 0.65-m depth. The relationship
between ~g~oim and soil water content was established on soil
cores collected from the pots both before and after the experi-
ments. The tensiometers were capable of measuring ~g~o~ down
to -0.08 MPa. Only on a few days did the soil become
sufficiently dry to cause cavitation in the tensiometers and
prevent direct measurement of ~g~oil. For these sandy soil mix-
tures, it was found that only 10% of the available soil water
remained at -0.08 MPa so that the cavitation problem in the
tensiometers introduced only a soil error in the estimate of
volumetric soil water content.
The estimation of TTSW in these experiments was based on
measurements of stomatal conductance. Stomatal conductance
was measured every 2 to 3 d for one leaf in each pot at midday.
The measures were made with a portable photosynthesis system
(Model LI-6200, LI-COR) in the 1992 experiment and with
a porometer (MKIII, Delta T Devices, Nottingham, UK)
the 1993 experiment. The lower limit of soil water content in
the calculation of TTSW was assumed to have occurred when
stomatal conductance reached the minimum value observed
during the soil drying cycle. With TTSW determined, FTSW
could be computed throughout the drying cycle based on soil
water content.
No direct measurement of daily increases in leaf area expan-
sion was obtained in the Montpellier pot experiments. To

LECOEUR & SINCLAIR: SOIL WATER DEFICIT EFFECTS ON FIELD PEA 333
estimate daily leaf area expansion, several approximations
were used. These approximations were based on measurements
of final stipule length for each leaf on the plant. Stipule length
was directly related to the final area of each leaf. Regression
between measured and calculated leaf areas on an independent
set of leaves (n = 150) resulted in an
2
of 0.95** and a
coefficient of variation of errors of 11.9%.
The calculation of the temporal expansion of area for each
individual leaf was based on results from the greenhouse experi-
ment at Gainesville. In that experiment, individual leaf area
expansion from 0.3 to 0.9 of final area increased linearly in
3 d for plants where FTSW was >0.2. To fully document this
progression, fraction of final leaf area was plotted against days
following emergence, as estimated from the phyllochron of
each plant (Fig. 1). Therefore, the amount of leaf area increase
on each of the three days during this period of rapid expansion
was calculated for each leaf. The temporal placement of the
expansion period for each leaf was based on dating of the
unfolded stage for each leaf. The date on which each leaf
reached 0.3 of final leaf area was calculated from the unfolded
stage based on cumulative thermal units (Maurer et al., 1966;
Lecoeur et al., 1995).
The daily measures of stomatal conductance and estimated
plant leaf area development were expressed as ratios to values
obtained from well-watered control plants. Average values for
each treatment were used because individual plant measure-
ments were not recorded to allow normalization of individual
plant differences. The relative rates of transpiration and leaf
area expansion were expressed as functions of FTSW.
tion to the beginning of flowering and from the beginning of
flowering to the end of flowering. The rain shelters, which
covered the plots only when rainfall was anticipated, were
constructed of metal archs covered with a greenhouse plastic.
The fourth treatment was designed to result in progressively
greater soil water deficits through the season. Irrigation in this
treatment was applied when the soil water potential reached
-60 kPa at increasing depths as the season progressed. Until
the beginning of flowering, the sensor depth used to initiate
irrigation was placed at 0.45 m. Sensor depth was increased
to 0.65 m for the period from the beginning of flowering to
the beginning of seed fill for the last reproductive phytomer
(Pigeaire et al., 1986) and to 0.85 m for the later stages. This
treatment resulted in a progressive decrease in irrigation so
that the severity of the soil water deficit increased slowly.
Because this treatment was not sheltered from rain, rain oc-
curring after the beginning of flowering resulted in restoration
of soil water content to the well-watered level.
Soil water potential (¥so,) was measured by positioning
tensiometers at 0.25-, 0.45-, 0.65-, and 0.85-m depths. The
relationship between ~soi~ and soil water content was established
1.2
1.0
0.8
Montpellier Field Experiment fY
A field experiment was conducted near Montpellier, France,
7
O. 6
with Frilene and Atol during the 1992 spring on a sandy loam
soil (fluvio-calcaric Cambissol). Pea seeds were sown on
0.4
Feb. at a density of 80 seed m
-2
and 0.25-m row spacing.
This multifactorial experiment was a randomized design with
four blocks. Elementary plots were 4.5 by 4.0 m.
0.2
The well-watered control was achieved by irrigating to keep
the fraction of available soil water above 0.7 of total available
o.0
soil water. Three water deficit treatments were imposed. Two
treatments were imlSosed by a rain shelter to exclude rainfall
at specific times during the growing season: from floral initia-
1.0
0.8
0.6
0.4
0.2
0.0
Days after leaf emergence from apical bud
A NTR = 1.05/(I+4.5*exp(-9*(FTSW-O.Oa5)))
o o
R
2
= 0.969
ATO L
I
I I
I I
I
.0 0.8
0.6 0.4
0.2 0.0
FTSW
Fig. 1. Fraction of final area of individual leaves plotted as a function
of the day of each observation for field pea cultivar Atol during
the greenhouse experiment at Gainesviile, FL.
0
2 4 6
1.0 0.8
0.6
0.4 0.2
0.0
FTSW
Fig. 2. Normalized transpiration rates (NTR) plotted as a function
of fraction transpirable soil water (FTSW) for cultivars (A) Atol
and (B) Frilene in the greenhouse experiment at Gainesville, FL.

334
CROP SCIENCE, VOL. 36, MARCH-APRIL 1996
o
1.2
o
1.0
o
O
O
0.8
0.6
ATO L
NLA -- -1 2/(+ l+exp(-7.5*FTSW))
0.2 e = o.962
CVe = 9.5%
0.0
I I I I
I
.0 0.8 0.6 0.4
0.2 0.0
FTSW
1.2
¯
¯
,. .. o
0.6
FRILENE
NLA = -1 +2/(1 +expC-9.5*(FTSW-0.05))) oi~8,~\
0.2 R
=
0.939
oe~_l
o
CVe 10.1~g
0.0
I
I I I I OI~¯I
1.0 0.8
0.6 0.4 0.2 0.0
’FTSW
Fig. 3. Normalized leaf area expansion rates (NLA) plotted as a
function of fraction transpirable soil water (FTSW) for cultivars
(A) Atol and (B) Frilene thegreenhouse experiment at Gainesville,
FL.
1
for this soil, thereby allowing FTSW to be calculated for the
soil profile as described above. Neutron probe measurements on
fully wetted soil and during the driest period of the experiments
showed that potential total transpirable soil water storage was
150 ram.
Stomatal conductance (gs) was measured with a portable
photosynthesis system (LI-6200 LI-COR). Measurements
gs were made every 2 to 3 d on three leaves in each plot or
a total of 12 leaves per treatment. Relative g~ was calculated
from the ratio of gs measured in water deficit plots to gs
measured in the well-watered control. Relative gs was expressed
as a function of FTSW. No data were collected on leaf area
expansion in the field experiment.
RESULTS AND DISCUSSION
Greenhouse Experiment
The pattern of NTR plotted against FTSW was remark-
ably consistent among the plants within each cultivar
(Fig. 2a and 2b). In each case, NTR was essentially
unchanged until the soil dried to a FTSW of 0.4. Below
0
1.2 ¯
o
0
o
1.0
o
~a
~ o
0.8
¯
~__X
°
0.6
% oN ¯
R’ = 0.616
BO
0.4 CVe = 20.8%
o v~
_~" =a ¯
0 Atol greenhouse 0 n
=
0.2
[] Atol field
0
¯ Frilene greenhouse
¯ Frilene field
0.O
I I I I I
I
.0 0.8 0.6 0.4 0.2 0.0
FTSW
Fig. 4. Estimates of relative stomatal conductance (g,) plotted as a
function of estimates of fraction transpirable soil water (FTSW)
in I~t and field experiment~ at Montpell~er, France.
FTSW of 0.4, NTR decreased linearly. The relationship
between NTR and FTSW was represented for both culti-
vars by the following logistic equation (r
z
= 0.97** for
each cultivar):
NTR = 1.05/(1 + 4.5exp [-9(FTSW - 0.085)])
These observations on the relationship between NTR
and FTSW are consistent with results obtained previously
with other species as discussed earlier.
The response pattern of NLA to FTSW was similar
to that obtained for NTR (Fig. 3a and 3b). When the
data from both cultivars were combined, the following
equation well represented (r
2
= 0.96"*, 9.5 % coefficient
of variation of errors) the relationship between NLA and
FTSW:
NLA = -1 + 2/[1 + exp (-7.1FTSW)] [4]
This equation indicates that little change in NLA occurred
with FTSW values greater than = 0.4. As soil dried to
FTSW values less than = 0.4, NLA decreased linearly.
Overall, these results of leaf area expansion for field
pea are consistent with previous observations of leaf
area expansion with other grain legumes (Sinclair, 1986;
Sinclair et al., 1987).
Montpellier Experiments
Greater variability in both transpiration (Fig. 4) and
leaf area expansion (Fig. 5) rates occurred in the Mont-
pellier experiments than the previous pot experiment.
Much of this variability probably was due to differences
in experimental technique. In the Montpellier experi-
ments soil water content was calculated based on mea-
surements of water potential at various depths in the
soil. Also, the greenhouse experiment at Gainesville was
conducted with smaller pots so that roots fully exploited
the soil volume. In both the pot and field experiments
at Montpellier, unexploited water remained at the bottom
of the soil profile, which resulted in a gradient of soil
water content.
In addition, temperature was less stable in the Montpel-

LECOEUR
&
SINCLAIR:
SOIL
WATER
DEFICIT
EFFECTS
ON
FIELD
PEA
335
1.2
1.0
0.8
0.6
0.4
0.2
0.0
-
4&^
O
Atol
greenhouse
•
Frilene
greenhouse
_2
'
Atol
0.775
CVe
Alo|
=
9.BX
1.0
0.8
0.6 0.4
FTSW
0.2
0.0
Fig.
5.
Estimates
of
normalized leaf area expansion rates (NLA)
plot-
ted
as a
function
of
estimates
of
fraction transpirable soil
water
(FTSW)
in pot
experiments
at
Montpellier, France.
Her
experiments, especially
in the
field
experiment, than
in
the
previous experiment. Consequently, temperature
extremes might have
influenced
both
NTR and
NLA.
Estimation
of NTR
from
stomatal conductance measure-
ments probably introduced variability
in the NTR
mea-
surements at Montpellier because stomatal conductance
is
an
instantaneous measure while daily weighing
of
pots
yields an integrated
estimate
of NTR. Indirect measure-
ment of leaf area expansion may have increased the
variability
in the NLA
data.
Nevertheless,
in
both
the pot
experiment
and the
field
experiment,
the
same general trends
in
relative
g
s
re-
sponse
to
FTSW (Fig.
4) and NTR
response
to
FTSW
were observed.
The
same logistic equation
as
obtained
previously
was
found
to fit the
data
(r
2
=
0.62**).
The
trend
previously observed
for NLA was
also evident
in
the data
from
the pot experiments at Montpellier (Fig.
5).
Equation
[4]
well represented these independent data
from
Montpellier (r
2
=
0.78**,
coefficient
of variation
of
errors
=
9.5%).
Overall,
the NTR and NLA
response
to
FTSW
is
consistent with results obtained with other crop
species.
In
particular, little change
in the
behavior
of
water
deficit
sensitive
processes
was
observed with soil drying across
a
wide range
of
FTSW.
In
field
pea, both
NTR and
NLA
decreased
for
FTSW <0.4. Reports
with
other
species
indicated
NTR
declined
at
FTSW
of
=0.25
to
0.35
(e.g.,
Sinclair
and
Ludlow,
1986).
Our
results indicated a strong consistency in the rela-
tionship of NTR and NLA to FTSW across a diversity
of
experimental conditions
and
measurement techniques.
The
logistic
functions
obtained with these
two
cultivars
provide
an
effective
method
for
describing
the
response
of
transpiration
and
leaf area expansion
to
soil water
deficits.
These
results represent a relatively easy ap-
proach
to
simulate
the
effects
of
drought
on
field
pea
behavior
and
ultimately
on
crop
yield under
a
range
of
field
conditions. Predictions of soil water balance could
provide estimates
of
FTSW
and
consequently, estimates
of the
response
of
field
pea to
seasonal changes
in
soil
water deficits.
ACKNOWLEDGMENTS
We are
grateful
to the DPE
(Ministere
de
1'Agriculture
et
de la
Foret),
the
Region
Languedoc-Roussillon,
the
Union
Nationale
Interprofessionnelle
des
Plantes
riches
en
proteines
and
the
Federation
Nationale
des
Agriculteurs
Multiplicateurs
de
Semences
for
their
financial
support
to J.
Lecoeur.