Journal of Plant Physiology 167 (2010) 1066–1075

Contents lists available at ScienceDirect

Journal of Plant Physiology

journa l homepage: www.e lsev ier .de / jp lph

Broader leaves result in better performance of indica rice under drought stress

M. Farooqa,b,∗, N. Kobayashia,c, O. Itoc, A. Wahidd, R. Serraja

a International Rice Research Institute (IRRI), DAPO Box 7777, Metro Manila, Philippinesb Department of Agronomy, University of Agriculture, Faisalabad 38040, Pakistanc Japan International Research Center for Agricultural Sciences, Tsukuba, Japand Department of Botany, University of Agriculture, Faisalabad 38040, Pakistan

a r t i c l e i n f o

Article history:Received 10 November 2009Received in revised form 11 March 2010Accepted 11 March 2010

Keywords:DroughtFraction of transpirable soil waterLeaf sizeRiceTranspiration

a b s t r a c t

Leaf growth is one of the first physiological processes affected by changes in plant water status underdrought. A decrease in leaf expansion rate usually precedes any reduction in stomatal conductance orphotosynthesis. Changes in leaf size and stomatal opening are potential adaptive mechanisms, which mayhelp avoid drought by reducing transpiration rate, and can be used to improve rice genotypes in water-saving cultivation. The indica rice cultivar IR64 and four of its near-isogenic lines (NILs; BC3-derived lines)unique for leaf size traits, YTK 124 (long leaves), YTK 127 (broad leaves), YTK 205 (short leaves) and YTK214 (narrow leaves), were compared in this study for changes in leaf growth and its water status. Theplants were subjected to two soil water regimes, well-watered and progressive soil drying measuredby the fraction of transpirable soil water (FTSW). Applied drought reduced leaf number, total leaf area,specific leaf area, plant biomass, tiller number, plant height, stomatal conductance, amount of watertranspired, leaf relative water content, and leaf water potential more in IR64 and the NILs than in therespective controls; nonetheless, transpiration efficiency (TE) was slightly higher under drought than inthe well-watered controls. NILs with broader leaves had higher biomass (and its individual components),less stomatal conductance, and higher TE under drought than NILs with narrow and shorter leaves. Underdrought, leaf number was positively correlated with tiller number and plant height; nonetheless, rootweight and total biomass, water transpired and TE, and plant height and TE were positively correlatedwith each other. However, a negative correlation was observed between stomatal conductance and the

FTSW threshold at which normalized transpiration started to decline during soil drying. Overall, the

road

I

ciTtthaui

tw

F

0d

IR64-derived lines with bdrought.

ntroduction

The development of crop genotypes consuming less water isonsidered a promising approach for sustainable crop productiv-ty in water-scarce areas (Farooq et al., 2009; Serraj et al., 2009).he development of such varieties requires a good knowledge ofhe physiological mechanisms and the genetic control of the con-ributing traits in drought resistance. In this regard, many studies

ave been conducted to analyze the genetic variation of traits, suchs leaf growth (Parent et al., in press), photosynthesis and waterse efficiency (Centritto et al., 2009), and numerous root traits

ncluding root pulling resistance (Pantuwan et al., 2002), root pene-

Abbreviations: FTSW, fraction of transpirable soil water; LWP, leaf water poten-ial; NILs, near-isogenic lines; NTR, normalized transpiration rate; RWC, relativeater contents; SLA, specific leaf area; TE, transpiration efficiency.∗ Corresponding author at: Department of Agronomy, University of Agriculture,aisalabad 38040, Pakistan. Tel.: +92 300 7108652; fax: +92 41 9200605.

E-mail address: farooqcp@gmail.com (M. Farooq).

176-1617/$ – see front matter © 2010 Elsevier GmbH. All rights reserved.oi:10.1016/j.jplph.2010.03.003

er leaves performed better than NILs with narrow and short leaves under

© 2010 Elsevier GmbH. All rights reserved.

tration ability (Ali et al., 2000; Clark et al., 2000; Fukai and Cooper,1995), deeper and thicker roots (Yadav et al., 1997), which mayinfluence the response of rice to water deficit. However, despiteresearch efforts to dissect drought resistance, the identificationand characterization of component traits (which can be transferredthrough plant breeding into cultivars with high-yielding geneticbackgrounds) ended up in limited success (Bernier et al., 2008;Serraj et al., 2009).

Rice genotypes differ greatly in their ability to withstanddrought. Plants exposed to water-limited conditions tend to stopgrowing quickly as soil dries, and leaves roll or senesce (Wopereiset al., 1996). Root architecture has been suggested, as the main basisof variation, to help maintain plant water status under prolongeddrought at vegetative stage (Nguyen et al., 1997; Parent et al., inpress). In a field trial, Mitchell et al. (1998) reported that plant size

actually defines the extent and role of the root system in plantsunder drought. In this regard, Lilley and Ludlow (1996) observedsubstantial tolerance of tissue water deficit in some rice cultivarsowing to leaf endurance during drought at the vegetative stage (DeDatta et al., 1988). Plants growing under water deficit decrease leaf

t Phy

ttciad

aRiMlPaprcs

ebiNao

ddeluwwdd

M

P

wu1laRsb((o3i1re

TN

M. Farooq et al. / Journal of Plan

ranspiration while at the same time fixing sufficient CO2 to fulfillheir energy needs. Studies have shown that transpirational lossan also be reduced even in sunny, dry environments by decreas-ng leaf size (Tardieu et al., 2000; Welander and Ottosson, 1997)nd/or controlling the response of transpiration to vapor pressureeficit (Sadok and Sinclair, 2010).

Leaf area (or leaf area index), determining canopy size, hasmajor control over water use under water-deficit conditions.

educed leaf area and small plant stature are conducive to lim-ted water use but could also lead to low productivity (Sinclair and

uchow, 2001). Botanists believe that dwarf plants with smallereaves are ideal for water-deficit areas (Geller and Smith, 1981).lants with these traits are better able to withstand drought,lthough their growth rates are relatively lower than the samelants grown in well-watered conditions. Thus, genotypes witheduced leaf area could potentially contribute to water-saving riceultivation in some specific target environments or water-deficitcenarios (Chenu et al., 2008; Tardieu et al., 2000).

Indica rice cultivar IR64 is very popular among farmers in South-ast Asia (Narciso and Hossain, 2002). It is semi-dwarf and favoredy growers because of its promising yield potential, good cook-

ng quality and resistance to various biotic stresses (Khush, 1995).onetheless, it does not perform well under drought; consider-ble yield reduction has been observed, particularly when droughtccurs during the reproductive stage (Wade et al., 1999).

During previous studies, IR64 near-isogenic lines (NILs) wereeveloped at IRRI using crosses with new plant types (japonica) asonor parents to develop lines with unique agronomic traits (Fujitat al., 2009). We have identified a group of IR64 NILs unique for leafength and width characteristics, which were more productive evennder limited water conditions (Fujita et al., 2009). In this study,e explored changes in growth, phenology, transpiration and plantater status, and their interrelationships using IR64 and its NILsiffering in leaf size in order to analyze the role of leaf size in theynamics of crop water use in response to soil water deficits.

aterials and methods

lant material, experimental details and treatments

In this research, indica rice (Oryza sativa L.) genotypes usedere cultivar IR64 and its NILs (BC3-derived lines, Table 1)nique for leaf size: IR84642-8-106-7-6-2-2-4-3-2-2-B (YTK24; long leaves), IR84642-8-115-2-6-2-3-2-2-B (YTK 127; wide

eaves), IR84636-13-71-9-10-3-4-2-2-B (YTK 205; short leaves)nd IR84636-13-78-11-3-2-2-4-2-2-2-B (YTK 214; narrow leaves).ice seeds of all genotypes were surface sterilized with 0.2% HgCl2olution for 5 min and were then washed thoroughly with tap waterefore sowing. Four pre-germinated seeds were sown in each pot15 cm diameter, 75 cm high) filled with 10 kg of a soil mixturesterilized loamy soil and compost with 1:1 ratio with a bulk densityf 1.2 g cm−3) on September 18. Two weeks after sowing (October

), thinning was done to 2 plants per pot. The plants were kept

n a glasshouse under natural solar radiation (a photoperiod of4/10 h light/dark) with regulated air temperature (27 ± 3 ◦C) andelative humidity (70–80%). It was ensured that roots occupied thentire soil volume in the pot before initiating the treatment. All the

able 1ear-isogenic lines (BC3-derived lines) unique for leaf size used in this study.

Entry no. Unique trait IR desig

YTK124 Long leaves IR84642YTK127 Wider leaves IR84642YTK205 Short leaves IR84636YTK214 Narrow leaves IR84636

siology 167 (2010) 1066–1075 1067

genotypes were subjected to two water regimes: well-watered andincremental drought (dry-down). The experiment was laid out in acompletely randomized design in a factorial arrangement with 10replicates per treatment.

Plants were grown under well-watered conditions for 4 weeks(5-leaf stage; till October 21) following emergence. On the same day(October 21), water was applied to all pots to full saturation. Thesepots were then enclosed in aluminum foil around the stem to avoidsoil evaporation after overnight draining. Next day (October 22),one subset of plants underwent progressive soil drying, one wasmaintained at initial soil water status and the other one was har-vested to record plant biomass (and its individual components) forcomparison. For re-watering, a small tube was inserted in each pot.The pots were weighed after being enclosed in aluminum foil andthis value was recorded as the initial target pot weight. Thereafter,the pots were weighed every morning around 0900.

To homogenize the development of drought across replications,and to avoid too rapid imposition of stress, the decrease in soilwater was limited by partial re-watering of the stressed pots. Well-watered control plants of each genotype were maintained at initialtarget weight by recouping the daily water loss from the pots. Theexperiment was terminated when all the soil water in drought-stressed pots decreased to a level when soil was no longer availableto support transpiration (November 12). This endpoint was iden-tified when the daily transpiration rate of the drought-stressedplants had decreased to less than 10% of that of the well-wateredplants, as previously described (Ray and Sinclair, 1997; Serraj et al.,1999). In this experiment, it was reached 3 weeks after the start ofdrought treatment.

Measurement of transpiration and stomatal conductance

Daily transpiration was determined as the difference in weighton successive days and analyzed as described by Ray and Sinclair(1997). To minimize the influence of large variations in dailytranspiration across days, the daily transpiration rates of thedrought-stressed pots were normalized against the transpirationrates measured for the well-watered plants on each day. This dailynormalization for the genotypes was computed as:

normalized transpiration rate (NTR)

= transpiration of stressed plantsaverage transpiration of control plants

(1)

The comparison among the genotypes was further facilitated byexpressing the available soil water as the fraction of transpirablesoil water (FTSW) for each pot in the drought-stressed treatmenton each day:

FTSW = daily pot weight − final pot weightinitial pot weight − final pot weight

(2)

The relationship of FTSW and transpiration for each genotype was

calculated by using a non-linear regression procedure (Serraj et al.,1999) to fit the equation:

NTR = 11 + A exp(B · FTSW)

(3)

nation Donor parent

-8-106-7-6-2-2-4-3-2-2-B IR65564-22-2-3-8-115-2-6-2-3-2-2-B IR65564-22-2-3-13-71-9-10-3-4-2-2-B IR69093-41-2-3-2-13-78-11-3-2-2-4-2-2-2-B IR69093-41-2-3-2

1 nt Physiology 167 (2010) 1066–1075

Cwdrw(

sLo

A

tpm

P

ctmwilR

R

Tr

T

S

warE

R

awmllaltcda

mcawYu

es,l

eafa

rea,

spec

ific

leaf

area

and

pla

nt

biom

ass

com

pon

ents

ofIR

64an

dit

sd

eriv

edli

nes

un

der

wel

l-w

ater

ed(W

W)

and

dro

ugh

tst

ress

(DS)

con

dit

ion

s.

Un

iqu

etr

aits

No.

ofle

aves

(pla

nt−1

)Le

afar

ea(c

m2

pla

nt−1

)Sp

ecifi

cle

afar

ea(c

m2

g−1)

Roo

tw

eigh

t(g

pla

nt−1

)St

emw

eigh

t(g

pla

nt−1

)Le

afw

eigh

t(g

pla

nt−1

)

WW

DS

W/S

aW

WD

SW

/Sa

WW

DS

W/S

aW

WD

SW

/Sa

WW

DS

W/S

aW

WD

SW

/Sa

–22

6a14

0.5c

d1.

6151

77a

1356

d3.

8224

5.8a

116.

5bc

2.11

8.22

a3.

18d

2.58

40.5

4a16

.67d

e2.

4321

.09a

12.3

5cd

e1.

71Lo

ng

leav

es15

3bc

137.

9cd

1.11

3596

bc12

73d

2.82

225.

7a91

.37c

2.47

7.15

ab6.

81ab

1.05

31.3

4bc

21.6

6d1.

4515

.83b

c14

.23b

cde

1.11

Wid

ele

aves

138.

2cd

112.

4d1.

2338

54b

1117

d3.

4521

8.4a

88.3

7c2.

477.

15ab

5.51

bc1.

3036

.06a

b18

.52d

e1.

9517

.33a

b12

.85c

de

1.35

Shor

tle

aves

182.

0b14

0.3c

d1.

3027

12c

1124

d2.

4223

9.0a

147.

1b1.

625.

00bc

d3.

76cd

1.33

17.1

6de

11.2

8e1.

5211

.28d

ef7.

915f

1.43

Nar

row

leav

es18

2.1b

132.

1cd

1.38

3350

bc13

04d

2.56

922

5.8a

133.

7b1.

696.

53ab

3.51

cd1.

8624

.14c

d15

.09d

e1.

6014

.81b

cd10

.29e

f1.

44

ele

tter

for

atr

ait

do

not

dif

fer

sign

ifica

ntl

y(p

0.05

).en

wel

lwat

ered

and

dro

ugh

tst

ress

ed.

068 M. Farooq et al. / Journal of Pla

omparisons of the curve generated by Eq. (3) for each genotypeere based on 95% confidence intervals of coefficients A and B. Toetermine the specific FTSW (threshold) value, at which transpi-ation began to decline, the plateau regression procedure of SASas employed (SAS Institute, Inc.) as described by Ray and Sinclair

1997).Stomatal conductance was measured at the center of the abaxial

ide of the penultimate leaf with a porometer (AP4, Delta-T Devicestd., Cambridge, UK) on a sunny day 2 weeks after the impositionf stress at 1000.

llometry

Number of tillers and plant height were measured daily afterhe imposition of drought stress. Total plant biomass and its com-onents were taken at the final harvest. Specific leaf area (SLA) waseasured as the ratio of leaf area to leaf weight.

lant water relations

Leaf water potential (LWP) was determined with a pressurehamber (Soil Moisture Equipment Corp., Santa Barbara, CA) fromhe penultimate leaf 2 weeks after the imposition of stress. To deter-

ine relative water contents (RWC), fresh leaves (0.5 g; Wf) wereeighed to obtain fresh weight. Later, these leaves were floated

n water for 4 h and saturated weight (WS) was determined. Theseeaves were dried for 24 h at 85 ◦C to determine dry weight (Wd).WC (%) were calculated as:

WC (%) = Wf − Wd

WS − Wd× 100 (4)

ranspiration efficiency (TE) was calculated using the followingelation:

E (g kg−1) = total dry matter (g)total water transpired (kg)

(5)

tatistical analysis

The data underwent analysis of variance using MSTAT-C soft-are (Freed and Scott, 1986). The least significant difference (LSD

t p = 0.05) test was used to compare the treatment means. Cor-elations were established between various traits using Microsoftxcel.

esults

Drought substantially decreased the number of leaves, leaf areand SLA in IR64 and its NILs more than in the respective well-atered controls (Table 2). Under well-watered conditions, theaximum number and area of leaves were recorded in IR64, fol-

owed by YTK205 and YTK214 for number of leaves and YTK127 foreaf area, whereas these traits had their lowest values in YTK127nd YTK205, respectively. Under drought, a minimum number ofeaves and leaf area were recorded in YTK127, which was similaro the rest of the genotypes for both traits. SLA was not statisti-ally different in all the genotypes under a control, whereas underrought this trait had a higher value in YTK205, followed by YTK214nd IR64 (Table 2).

Drought significantly reduced root, stem, leaf and total biomassore than in the respective well-watered controls. Under control

onditions, although there was no significant difference (p > 0.05)mong the NILs for root dry weight, maximum stem and leafeights and total biomass were recorded in IR64 and YTK127, whileTK124 and YTK127 displayed maximum total biomass. Likewise,nder drought, maximum root, stem and leaf dry weights and total Ta

ble

2N

um

ber

ofle

av

Gen

otyp

es

IR64

YTK

124

YTK

127

YTK

205

YTK

214

Mea

ns

wit

hsa

ma

Rat

iobe

twe

M. Farooq et al. / Journal of Plant Physiology 167 (2010) 1066–1075 1069

Table 3Total biomass, tiller number and plant height of IR64 and its derived lines under well-watered (WW) and drought stress (DS) conditions.

Genotypes Unique trait Total biomass (g) Tiller number Plant height (cm)

WW DS W/Sa WW DS W/Sa WW DS W/Sa

IR64 – 72.49a 32.21bcd 2.25 21.30a 16.60bc 1.28 40.8a 29.5bc 1.38YTK124 Long leaves 61.88a 42.70b 1.45 19.1b 16.20bc 1.18 27.6c 25.2c 1.09YTK127 Wider leaves 64.70a 36.89bc 1.75 14.70c 11.00d 1.34 23.3d 20.0d 1.16YTK205 Short leaves 34.26bcd 22.96d 1.49 18.00b 12.70cd 1.42 31.2b 26.0c 1.20

M

bwb

htwobn

F

YTK214 Narrow leaves 45.48b 28.89cd 1.57

eans with same letter for a trait do not differ significantly (p 0.05).a Ratio between well watered and drought stressed.

iomass were noted in YTK124, while minimum root dry weightas noted in IR64, and minimum stem and leaf weight and total

iomass were recorded in YTK205 (Tables 2 and 3).Likewise, a substantial decrease in tiller number and plant

eight was observed in drought-stressed IR64 and its NILs morehan in the respective well-watered controls (Table 3). Under well-

atered conditions, maximum tiller number and plant height were

bserved in IR64, followed by YTK124 and YTK205 for tiller num-er, and YTK205 and YTK214 for plant height, while minimum tillerumber was observed in YTK214, followed by YTK127, and mini-

ig. 1. Normalized transpiration rate of (a) IR64, (b) YTK124, (c) YTK127, (d) YTK205 and (e

14.00c 13.00cd 1.07 31.1b 22.0d 1.41

mum plant height was noted in YTK127 (Table 3). Similarly, underdrought, maximum tiller number and plant height were recorded inIR64, followed by YTK124 for tiller number and YTK205 and YTK124for plant height (Table 3). Minimum number of tillers and plantheight were observed in YTK127 (Table 3).

Although stomatal conductance and amount of water transpired

decreased substantially, TE of YTK124 and YTK127 was higherunder drought than in well-watered conditions (Table 4). Underwell-watered conditions, the highest stomatal conductance wasmeasured in YTK205, followed by YTK214 and YTK124; however,

) YTK214 at different levels of fraction of transpirable soil water. *Threshold values.

1070 M. Farooq et al. / Journal of Plant Physiology 167 (2010) 1066–1075

Tab

le4

Stom

atal

con

du

ctan

ce,t

otal

wat

ertr

ansp

ired

,tra

nsp

irat

ion

effi

cien

cyan

dw

ater

rela

tion

sof

IR64

and

its

der

ived

lin

esu

nd

erw

ell-

wat

ered

(WW

)an

dd

rou

ght

stre

ss(D

S)co

nd

itio

ns.

Gen

otyp

esU

niq

ue

trai

tSt

omat

alco

nd

uct

ance

(mm

olm

−2s−1

)W

ater

tran

spir

ed(k

g)To

talw

ater

add

ed(k

g)Tr

ansp

irat

ion

effi

cien

cy(T

E;g

kg−1

)Le

afw

ater

pot

enti

al(M

Pa)

Leaf

rela

tive

wat

erco

nte

nts

(RW

C;

%)

WW

DS

W/D

aW

WD

SW

/Da

WW

DS

W/D

aW

WD

SW

/Da

WW

DS

W/D

aW

WD

SW

/Da

IR64

–30

5.4b

34.8

4c8.

766

15.0

0a6.

36cd

2.36

15.0

0a3.

26e

4.60

4.54

b4.

49b

1.01

1.41

c3.

62a

0.38

989

.67a

71.5

0e1.

25Y

TK12

4Lo

ng

leav

es35

1.0a

b32

.94c

10.6

513

.63a

6.74

cd2.

0213

.63a

3.67

e3.

944.

32b

5.88

a0.

731.

24c

3.59

a0.

345

91.4

2a77

.75c

d1.

17Y

TK12

7W

ider

leav

es28

7.0b

30.8

9c9.

2913

.82a

6.78

cd2.

0313

.82a

b3.

60e

3.84

4.44

b4.

88ab

0.91

1.38

c3.

53ab

0.39

192

.62a

77.5

9cd

e1.

19Y

TK20

5Sh

ort

leav

es39

5.0a

36.8

5c10

.72

8.00

bc4.

77d

1.68

8.00

d1.

66g

5.21

4.53

b4.

33b

1.05

1.31

c3.

04ab

0.43

188

.12a

b82

.64b

c1.

06Y

TK21

4N

arro

wle

aves

380.

6a65

.5c

5.81

10.2

8b6.

04cd

1.71

10.2

8c2.

88f

3.95

4.74

ab4.

46b

1.06

1.28

c2.

84b

0.45

181

.87b

c71

.68d

e1.

14

Mea

ns

wit

hsa

me

lett

erfo

ra

trai

td

on

otd

iffe

rsi

gnifi

can

tly

(p0.

05).

aR

atio

betw

een

wel

lwat

ered

and

dro

ugh

tst

ress

ed.

Fig. 2. Relationship between normalized transpiration rate threshold values andtranspiration efficiency. Means ± s.e. of 10 replicates for TE.

there was no difference among the genotypes under drought. Incontrast, IR64, followed by YTK124 and YTK127, transpired a max-imum amount of water under well-watered conditions. Underdrought, on the other hand, all the genotypes transpired the sameamount of water. Contrarily, in well-watered pots, water transpiredwas added back in respective pot. Nonetheless, under drought,maximum water was applied in YTK124 followed by YTK127 andIR64 while minimum amount was added in YTK205. A maxi-mum TE was recorded in YTK124 and YTK127 under drought, butin YTK 214 under well-watered conditions (Table 4). Similarly,drought lowered LWP and leaf RWC in all genotypes comparedwith the respective controls (Table 4). Data indicated maximumLWP in IR64, although not differing from other genotypes underwell-watered conditions. However, under drought IR64 displayedminimum LWP followed by YTK124, YTK127 and YTK205 (Table 4).Likewise, higher RWC were recorded in YTK127 but were mini-mum in YTK214 under well-watered conditions, while minimumRWC were observed in IR64 but were maximum in its derived NILYTK215 under drought (Table 4).

In all the genotypes, NTR decreased with a decrease in FTSW.The maximum NTR threshold value was recorded in YTK205 andIR64, whilist a minimum one in YTK214 (Fig. 1). A strong inverserelationship was recorded between NTR threshold values and TE(Fig. 2).

Plant height was substantially reduced by drought than well-watered controls, though response of various genotypes variedconsiderably (Fig. 3). For IR64, plant height increased steadily untilday 6 after the imposition of drought, although it continued toincrease under well-watered conditions (Fig. 3a). There was nosignificant difference in the plant heights of YTK124 and YTK127after the imposition of drought, thought it was higher during lastthree days under well-watered condition (Fig. 3b and c). Droughtdid not influence plant height in YTK205 until 17 days of drought;however, it declined later on (Fig. 3d). For YTK214, plant heightwas similar under well-watered and drought conditions until day12 after the imposition of drought. Afterwards, plants were tallerunder well-watered conditions (Fig. 3e).

In IR64 and its derived NILs, tillering was substantially morereduced by drought than in well-watered controls (Fig. 4). In IR64,

tiller number was the same under both conditions till 10 days afterdrought, but increased in well-watered plants toward the end ofthe experiment (Fig. 4a). In YTK124 till day 14 after the drought,the number of tillers was similar under both conditions; afterward,

M. Farooq et al. / Journal of Plant Physiology 167 (2010) 1066–1075 1071

F der w

tdt(ntdnsd

wwbtL(tdwt

ig. 3. Plant height of (a) IR64, (b) YTK124, (c) YTK127, (d) YTK205 and (e) YTK214 un

his number was higher under well-watered conditions than inrought (Fig. 4b). Similarly, in YTK127 number of tillers continuedo increase under well-watered conditions but not under droughtFig. 4c). In YTK205 till day 12 after the drought imposition, theumber of tillers was similar under both conditions; afterward,iller number was higher under well-watered conditions than inrought (Fig. 4d). In YTK214, there was no difference (p > 0.05) inumber of tillers under both conditions till day 14 after stress impo-ition, but the number increased under well-watered conditions tillay 21 after stress imposition (Fig. 4e).

Correlations drawn between various traits indicated that, underell-watered conditions, leaf number was positively correlatedith tiller number, LWP and plant height; SLA with tiller num-

er and LWP; whereas leaf area was positively correlated withiller number, leaf weight, root weight, stem weight, total biomass,WP, stomatal conductance, water transpired, TE and plant height

Table 5). Likewise, leaf weight, root weight, stem weight andotal biomass were positively correlated with LWP, stomatal con-uctance, TE and plant height. LWP was negatively correlatedith stomatal conductance, and was positively correlated with

he amount of water transpired, TE and plant height. Likewise,

ell-watered (WW) and drought stress conditions (DS). Means ± s.e. of 10 replicates.

RWC were negatively correlated with stomatal conductance, andpositively with water transpired (Table 5). Similarly, stomatal con-ductance was negatively correlated with the amount of watertranspired, TE and plant height. The amount of water transpiredwas negatively correlated with TE; however, there was a positivecorrelation between plant height and TE (Table 5).

Under drought, however, leaf number was positively correlatedwith tiller number and plant height (Table 6). A negative corre-lation of SLA was evident with leaf, stem and total dry weight,LWP and water transpired, while SLA was positively correlatedwith plant height. Likewise, leaf area was positively correlated withtiller number, and negatively with RWC. Leaf weight was positivelycorrelated with biomass components, LWP, water transpired andTE. Similarly, root weight and total biomass were positively corre-lated with each other, water transpired and TE. LWP was negativelycorrelated with stomatal conductance and positively correlated

with water transpired, while RWC were positively correlatedwith plant height. Also, there was a positive correlation of plantheight and TE, but a negative correlation existed between stom-atal conductance and threshold value for normalized transpiration(Table 6).

1072 M. Farooq et al. / Journal of Plant Physiology 167 (2010) 1066–1075

F 214 un

D

mttsNlwmStg(iasiwIadc

ig. 4. Number of tillers in (a) IR64, (b) YTK124, (c) YTK127, (d) YTK205 and (e) YTK

iscussion

Absolute and derived plant growth traits can provide directeasurements of the ultimate tendency of genotypic responses

o drought conditions (Hunt, 1978). Our results confirmed thathough drought treatment influenced growth traits, there was sub-tantial genetic variation in the response of IR64 and its derivedILs under drought. Although the highest values of leaf number,

eaf area, SLA, leaf, stem and root dry weights, and total biomassere observed in IR64 under well-watered conditions, the perfor-ance of the genotypes was variable under drought. The highest

LA was observed in line YTK205 (Table 2), which seemed to behe result of minimum leaf weight under drought. Likewise, thereatest root, stem and total dry weights were recorded in YTK124Tables 2 and 3). Furthermore, all these growth parameters werenterdependent as evident from the strong positive correlationsmong them (Tables 5 and 6). The decline in SLA, root weight,tem weight and total biomass under drought compared with thatn well-watered controls was highest in IR64 (Tables 2 and 3),

hich confirmed its high sensitivity to drought (Lafitte et al., 2006).mpaired cell division and plant water relations have been associ-ted with reduced plant height, leaf area and crop growth underrought (Nonami, 1998), which is a likely reason for the observedhanges in this study.

der well-watered (WW) and drought (DS) conditions. Means ± s.e. of 10 replicates.

Interestingly, total plant biomass under drought was observedin YTK124, which had broader leaves, while this trait was mini-mum under both well-watered and drought conditions in YTK205and YTK214 having shorter and narrower leaves (Table 3). Further-more, the number of leaves was lesser in IR64 NILs YTK124 andYTK127 having longer and broader leaves than in lines YTK205 andYTK214 with shorter and narrower leaves, albeit the leaf area washigher in YTK124 and YTK127 (Table 2). Although SLA was simi-lar in all genotypes under well-watered conditions, it was higherin the lines with smaller leaf size (YTK205 and YTK214) than inthose with larger leaves (YTK124 and YTK127) under drought. Thisseemed due to a lower leaf weight of the former NILs (Table 2).Leaf area also indicated a strong positive correlation with leaf, stemand root weights, and total biomass under well-watered condi-tions (Table 5) but not under drought (Table 6). This is importantin view of the fact that smaller leaf size offers lesser opportuni-ties for the interception of solar radiation and thus implicatingreduced dry matter production (Sinclair and Muchow, 2001). Geno-types adapted to lowland conditions tend to cease growth as soil

water declines, and leaves roll or senesce (Parent et al., in press;Wopereis et al., 1996). As evident from previous reports, a vigorousroot system and more tillers are a key factor in the maintenanceof higher leaf water status (Nguyen et al., 1997; Parent et al., inpress).

M.Farooq

etal./JournalofPlant

Physiology167

(2010)1066–1075

1073

Table 5Correlation coefficients of important traits in IR64 NILs under well-watered conditions (n = 10).

Characteristics Leaf area Tiller number Leaf weight Root weight Stem weight Total biomass Water Potential RWC Stomatal conductance Water transpired TE Plant height

Leaf number 0.510ns 0.992*** 0.359ns 0.235ns 0.124ns 0.080ns 0.696** −0.264ns 0.091ns 0.165ns 0.156ns 0.594*

SLA 0.355ns 0.559* 0.174ns 0.033ns 0.0005ns −0.023ns 0.626* 0.0439ns 0.157ns 0.441ns −0.027ns 0.341nsLeaf area 0.559* 0.982*** 0.928*** 0.916*** 0.890*** 0.868*** 0.286ns −0.763*** 0.757** 0.923*** 0.957***

Tiller number 0.410ns 0.302ns 0.186ns 0.155ns 0.703** −0.197ns 0.0563ns 0.231ns 0.208ns 0.616*

Leaf weight 0.971*** 0.964*** 0.941*** 0.788*** 0.289ns −0.833*** 0.805*** 0.977*** 0.940***

Root weight 0.958*** 0.961*** 0.627* 0.257ns −0.789*** 0.805*** 0.969*** 0.874***

Stem weight 0.991*** 0.680** 0.472ns −0.924*** 0.953*** 0.988*** 0.821***

Total biomass 0.610* 0.504ns −0.890*** 0.994*** 0.969*** 0.771***

Water potential 0.273ns −0.625* 0.642** 0.685** 0.858***

RWC −0.587* 0.904*** 0.342ns 0.017nsStomatal conductance −0.808*** −0.907*** −0.665**

Water transpired 0.873*** 0.468nsTE 0.870***

ns: non-significant.* Significant at p < 0.1.

** Significant at p < 0.05.*** Significant at p < 0.01.

Table 6Correlation coefficients of important traits in IR64 NILs under drought stress conditions (n = 10).

Characteristics Leaf area Tiller number Leaf weight Root weight Stem weight Totalbiomass

Waterpotential

RWC Stomatalconductance

Watertranspired

TE Plant height Threshold forNTR

Leaf number 0.517ns 0.878*** −0.284ns −0.325ns −0.282ns −0.305ns −0.142ns −0.023ns 0.102ns −0.476ns −0.047ns 0.059ns 0.260nsSLA 0.014 ns 0.582 * −0.934*** −0.789*** −0.928*** −0.944*** −0.800*** 0.0857ns 0.527ns −0.906*** −0.754** −0.375ns 0.159nsLeaf area 0.582 * 0.338ns −0.263ns 0.274ns 0.197ns 0.123ns −0.834*** 0.369ns 0.3341ns 0.139ns −0.486ns −0.258nsTiller number −0.101ns −0.421ns −0.192ns −0.218ns 0.201ns −0.185ns −0.163ns −0.281ns −0.144ns −0.311ns 0.547nsLeaf weight 0.679** 0.979*** 0.969*** 0.808*** −0.335ns −0.393ns 0.952*** 0.795*** 0.231ns −0.225nsRoot weight 0.790*** 0.833*** 0.507ns 0.374ns −0.469ns 0.952*** 0.915*** 0.856*** −0.263nsStem weight 0.996** 0.722** −0.237ns −0.340ns 0.921*** 0.890*** 0.388ns −0.342nsTotal biomass 0.739** −0.155ns −0.400ns 0.899*** 0.905*** 0.451ns −0.302nsWater potential −0.046ns −0.815*** 0.673** 0.545ns 0.179ns 0.390nsRWC −0.501ns −0.491ns 0.088 ns 0.734** 0.395nsStomatal conductance −0.200ns −0.333ns −0.412ns −0.859***

Water transpired 0.649** 0.054ns −0.376nsTE 0.716** −0.353nsPlant height −0.376ns

ns: non-significant.* Significant at p < 0.1.

** Significant at p < 0.05.*** Significant at p < 0.01.

1 nt Phy

ftlTbia

citutn

imptuiat(adiadehwdadcrB1v2

owgmwcfii

ImtcwbtnGbtbdd

074 M. Farooq et al. / Journal of Pla

Under both well-watered and drought conditions, there wereewer tillers in IR64-derived lines than in IR64 itself (Table 3). Geno-ypes with a new plant type were used as donor parents for theseines, which showed reduced tillering capacity (Peng et al., 1994).he lines with smaller leaves had more tillers than those withroader leaves under well-watered conditions (Table 3; Fig. 4). This

s consistent with the findings that tillering regulation mechanismsre dependent on leaf size and shape (Jaffuel and Dauzat, 2005).

IR64 displayed maximum plant height under well-wateredonditions, but minimum plant height under drought. Anothernteresting finding is that plants of YTK124 and YTK205 were tallerhan those of other IR64-derived lines under drought; nonetheless,nder well-watered conditions, YTK205 and YTK214 were tallerhan other IR64-derived lines (Table 3; Fig. 3), the reasons for whicheed to be explored.

Stomatal conductance was drastically reduced under droughtn all genotypes, possibly as a strategy to curtail water loss and

atain plant water status (Lo Gullo et al., 2003; Parent et al., inress). A negative correlation of stomatal conductance with NTRhreshold further supports this assumption (Table 6). Althoughnder well-watered conditions stomatal conductance was higher

n lines with smaller leaves (YTK205 and YTK214) than in IR64nd NILs with larges leaves (YTK124 and YTK127), more water wasranspired in the latter ones seemingly because of higher leaf areaTables 2 and 3). The range of TE values measured in this experimentre slightly higher than those reported in similar pot studies con-ucted between 1994 and 2006 at the International Rice Research,

.e., between 2.2 and 4.0 g dry matter per liter transpired (Haefele etl., 2009). However, the differences in total biomass accumulationue to root systems may explain such discrepancy. The most inter-sting result in this experiment was observed for TE, which wasigher in NILs with smaller leaves under well-watered conditions,hile broad leaf NILs behaved in a similar way as compared underrought (Table 4). This is most likely due to reduced SLA (Table 2)s evident from the negative correlation between SLA and TE underrought (Table 6), but no relationship occurred under well-wateredonditions (Table 5). This finding is in conformity with previouseports for switchgrass (Byrd and May, 2000), peanut (Brown andyrd, 1997; Wright et al., 1994) and pearl millet (Brown and Byrd,997). LER is a major determinant of final leaf area at both indi-idual and whole-plant scales (Bultynck et al., 2004; Chenu et al.,008).

LWP and RWC also declined under drought, though the extentf the reduction was different in various genotypes. Under well-atered conditions, LWP did not differ appreciably among the

enotypes, whereas, under drought, the value of this trait wasaximum in YTK214 (Table 4). Likewise, minimum RWC valuesere observed in YTK214 under both well-watered and drought

onditions (Table 4). This trend of changes in both traits poses dif-culty for defining clear physiological mechanisms associated with

mproved drought tolerance in these NILs.A maximum NTR threshold value was recorded in YTK205 and

R64, which indicated that both genotypes reduced transpirationuch earlier than all other genotypes although water was available

o support the transpiration, while YTK214, YTK124 and YTK127ontinued stomatal activity although the amount of availableater for transpiration was meager (Fig. 1). A negative correlation

etween stomatal conductance and NTR threshold value substan-iated this finding (Table 6). This is also affirmed by the strongegative correlation between NTR threshold values and TE (Fig. 2).enotypes with larger and broader leaves exhibited higher TE

y using water more judiciously as evident from the higher NTRhreshold values (Fig. 1). Plant height continued to increase underoth well-watered and drought conditions, although there was noifference in plant height under drought and well-watered con-itions in lines with bigger leaves (YTK124 and YTK127; Fig. 3).

siology 167 (2010) 1066–1075

However, in IR64 and YTK214 the rate of increase in plant heightand tillering were lower under drought. Despite no difference forincrease in plant height under drought and well-watered condi-tions in lines with bigger leaves (YTK124 and YTK127), IR64 andYTK214 displayed a lower rate of increase in plant height underdrought (Fig. 3).

Substantial research efforts have been previously devoted toexploring the mechanisms of dry matter production in lowland riceunder drought conditions (Mackill et al., 1996; Mitchell et al., 1998;Serraj et al., 2009). Recent studies suggest that rice is in fact notdrought tolerant, rather an avoider of drought (Fukai et al., 2009).Furthermore, a comparative analysis of the responses of water rela-tions and leaf elongation rate to soil water deficit and evaporativedemand in contrasting lowland and upland genotypes revealed anisohydric behaviour in rice and a relatively low sensitivity of leafelongation rate water deficit, concluding that the main reason ofrice drought sensitivity may be its poor root architecture and/orhydraulic properties (Parent et al., in press). However, the presentstudy shows that variation in leaf size traits may also influence sig-nificantly plant responses to water deficit. Hence, development ofgenotypes with broader leaves might be an important strategy forimproving CO2 assimilation, dry matter accumulation and seed set-ting (Bouchabké et al., 2006). A similar strategy of avoiding droughtemerged in rice cultivar IR64 and its derived NILs in this study, asevident from a close relationship of leaf area to dry matter yieldunder drought (Table 6).

In summary, leaf size seems to be an important trait forgenotype improvement suitable for water-saving rice cultivation.IR64-derived lines with smaller (short and narrow) and biggerleaves performed better under well-watered and drought condi-tions, respectively. Nonetheless, further studies are required toconfirm these findings and analyze in more detail the physiologicalprocesses and genetic control of these leaf architecture traits.

References

Ali ML, Pathan MS, Zhang J, Bai G, Sarkarung S, Nguyen HT. Mapping QTLs for roottraits in a recombinant inbred population from two indica ecotypes in rice. TheorAppl Genet 2000;101:756–66.

Bernier J, Atlin GN, Serraj R, Kumar A, Spaner D. Breeding upland rice for droughtresistance. J Sci Food Agric 2008;88:927–39.

Bouchabké O, Tardieu F, Simonneau T. Leaf growth and turgor in growing cells ofmaize (Zea mays L.) respond to evaporative demand under moderate irrigationbut not in water-saturated soil. Plant Cell Environ 2006;29:1138–49.

Brown RH, Byrd GT. Relationships between specific leaf weight and mineral concen-tration among genotypes. Field Crops Res 1997;54:19–28.

Bultynck L, Ter Steege MW, Schortemeyer M, Poot P, Lambers H. From individual leafelongation to whole shoot leaf area expansion: a comparison of three Aegilopsand two Triticum species. Ann Bot 2004;94:99–108.

Byrd GT, May PA. Physiological comparisons of switchgrass cultivars differing intranspiration efficiency. Crop Sci 2000;40:1271–7.

Centritto M, Lauteri M, Monteverdi M, Serraj R. Leaf gas exchange, carbon isotopediscrimination and grain yield in contrasting rice genotypes subjected to waterdeficits during reproductive stage. J Exp Bot 2009;60:2325–39.

Chenu K, Chapman SC, Hammer GL, McLean G, Ben Haj Salah H, Tardieu F. Short-term responses of leaf growth rate to water deficit scale up to whole-plantand crop levels: an integrated modelling approach in maize. Plant Cell Environ2008;31:378–91.

Clark LJ, Aphale SL, Barraclough PB. Screening the ability of rice roots to overcomethe mechanical impedance of wax layers: importance of test conditions andmeasurement criteria. Plant Soil 2000;219:187–96.

De Datta SK, Malabuyoc JA, Aragon EL. A field screening technique for evaluatingrice germplasm for drought tolerance during the vegetative stage. Field CropsRes 1988;19:123–34.

Farooq M, Wahid A, Kobayashi N, Fujita D, Basra SMA. Plant drought stress: effects,mechanisms and management. Agron Sust Dev 2009;29:185–212.

Freed RD, Scott DE. MSTAT-C. MI, USA: Crop and Soil Sci. Dept., Michigan StateUniversity; 1986.

Fujita D, Santos RE, Ebron LA, Telebanco-Yanoria MJ, Kato H, Kobayashi S, et al. Devel-opment of introgression lines of an Indica-type rice variety, IR64, for uniqueagronomic traits and detection of the responsible chromosomal regions. FieldCrops Res 2009;114:244–54.

Fukai S, Cooper M. Development of drought resistant cultivars using physio-morphological traits in rice. Field Crops Res 1995;40:67–86.

t Phy

F

G

HH

J

K

L

L

M

L

M

N

N

N

P

M. Farooq et al. / Journal of Plan

ukai S, Basnayake J, Makara O. Drought resistance characters and variety develop-ment for rainfed lowland rice in Southeast Asia. In: Serraj R, Bennett J, HardyB, editors. Drought frontiers in rice—crop improvement for increased rainfedproduction. Singapore: World Scientific Publishing; 2009.

eller GN, Smith WK. Influence of leaf size, orientation, and arrangement on tem-perature and transpiration in three high-elevation, large-leafed herbs. Oecologia1981;53:227–34.

unt R. Plant growth analysis. UK: Edward Arnold; 1978.aefele SM, Siopongco JDLC, Boling AA, Bouman BAM, Tuong TP. Transpiration effi-

ciency of rice (Oryza sativa L.). Field Crop Res 2009;111:1–10.affuel S, Dauzat J. Synchronism of leaf and tiller emergence relative to position and

to main stem development stage in a rice cultivar. Ann Bot 2005;95:401–12.hush G. Modern varieties: their real contribution to food supply and equity. Geo J

1995;35:275–84.afitte HR, Li ZK, Vijayakumar CHM, Gao YM, Shi Y, Xu JL, et al. Improvement of

rice drought tolerance through backcross breeding: evaluation of donors andselection in drought nurseries. Field Crops Res 2006;97:77–86.

illey JM, Ludlow MM. Expression of osmotic adjustment and dehydration tolerancein diverse rice lines. Field Crops Res 1996;48:185–97.

ackill D, Coffman W, Garrity D. Rainfed lowland rice improvement. Manila, Philip-pines: International Rice Research Institute; 1996.

o Gullo M, Nardini A, Trifilo P, Salleo S. Changes in leaf hydraulics and stomatalconductance following drought stress and irrigation in Ceratonia siliqua (Carobtree). Physiol Plant 2003;117:186–94.

itchell JH, Siamhan D, Wamala MH, Risimeri JB, Chinyamakobvu E, Henderson SA,et al. The use of seedling leaf death score for evaluation of drought resistance ofrice. Field Crops Res 1998;55:129–39.

arciso J, Hossain M. World Rice Statistics (Online). Available from IRRI atwww.irri.org/science/ricestat (posted 2002); 2002.

guyen HT, Babu RC, Blum A. Breeding for drought resistance in rice: physiologyand molecular genetic considerations. Crop Sci 1997;37:1426–34.

onami H. Plant water relations and control of cell elongation at low water poten-tials. J Plant Res 1998;111:373–82.

antuwan G, Fukai S, Cooper M, Rajatasereekul S, O’Toole JC. Yield response ofrice (Oryza sativa L.) genotypes to different types of drought under rain-fed

siology 167 (2010) 1066–1075 1075

lowlands-Part 2. Selection of drought resistant genotypes. Field Crops Res2002;73:169–80.

Parent B, Suard B, Serraj R, Tardieu F. Rice leaf growth and water poten-tial are resilient to evaporative demand and soil water deficit oncethe effects of root system are neutralized. Plant Cell Environ; in press,doi:10.1111/j.1365-3040.2010.02145.x.

Peng S, Khush GS, Cassman KG. Evaluation of a new plant ideotype for increasedyield potential. In: Cassman KG, editor. Breaking the yield barrier. Proceed-ings of a workshop on rice yield potential in favorable environments. Los Banos(Philippines): International Rice Research Institute; 1994. p. 5–20.

Ray JD, Sinclair TR. Stomatal closure of maize hybrids in response to drying soil. CropSci 1997;37:803–7.

Sadok W, Sinclair TR. Transpiration response of ‘slow-wilting’ and commercial soy-bean (Glycine max (L.) Merr.) genotypes to three aquaporin inhibitors. J Exp Bot2010;61:821–9.

Serraj R, Allen HL, Sinclair TR. Soybean leaf growth and gas exchange response todrought under carbon dioxide enrichment. Global Change Biol 1999;5:283–92.

Serraj R, Kumar A, McNally KL, Slamet-Loedin I, Bruskiewich R, Mauleon R, et al.Improvement of drought resistance in rice. Adv Agron 2009;103:41–98.

Sinclair TR, Muchow RC. System analysis of plant traits to increase grain yield onlimited water supplies. Agron J 2001;93:263–70.

Tardieu F, Raymond M, Hamadi P, Granier C, Muller B. Spatial distribution of expan-sion rate and cell size in maize leaves: a synthesis of the effects of soil waterstatus, evaporative demand and temperature. J Exp Bot 2000;51:1505–14.

Wade LJ, McLaren CG, Quintana L. Genotype by environment interactions acrossdiverse rainfed lowland rice environments. Field Crops Res 1999;64:35–50.

Welander NT, Ottosson B. Influence of photosynthetic photon flux density on growthand transpiration in seedlings of Fagus sylvatica. Tree Physiol 1997;17:133–40.

Wopereis MCS, Kropff MJ, Maligaya AR, Tuong TP. Drought-stress responses of two

lowland rice cultivars to soil water status. Field Crops Res 1996;46:21–39.

Wright GC, Nageswara RRC, Farquhar GD. Water-use efficiency and carbon isotopediscrimination in peanut under water deficit conditions. Crop Sci 1994;34:92–7.

Yadav R, Courtois B, Huang N, McLaren G. Mapping genes controlling root morphol-ogy and root distribution in a doubled-haploid population of rice. Theor ApplGenet 1997;94:619–32.