Reliability of ion accumulation and growth componentsfor selecting salt tolerant lines in large populations of rice

Tanveer Ul HaqA,D,F, Javaid AkhtarB, Katherine A. SteeleC, RanaMunnsD,E and John GorhamC

ACollege of Agriculture, PO Box 79, Dera Ghazi Khan 32200, Pakistan.BInstitute of Soil and Environmental Sciences, University of Agriculture, Faisalabad 38040, Pakistan.CCollege of Natural Sciences, Bangor University, Deiniol Road, Bangor, Gwynedd, Wales LL57 2UW, UK.DSchool of Plant Biology, University of Western Australia, Crawley, WA 6009, Australia.ECSIRO Plant Industry, GPO Box 1600, Canberra, ACT 2601, Australia.FCorresponding author. Email: drtanveer@uaf.edu.pk

Abstract. Ion accumulation and growth under salt stress was studied in two experiments in a rice mapping populationderived from parents CO39 and Moroberekan with 4-fold differences in shoot Na+ accumulation. The 120 recombinantinbred lines (RILs) had differences up to 100-fold in Na+. Measurement of ‘salt tolerance’ (biomass production of the RILsin 100mM NaCl relative to controls) after 42 days showed a 2-fold variation in ‘salt tolerance’ between parents, with fiveRILs being more tolerant than the more tolerant parent CO39. The reliability of various traits for selecting salt tolerance inlarge populations was explored by measuring Na+, K+ and K+/Na+ ratios in leaf blades and sheaths after 7 or 21 days ofexposure to 100mM NaCl, and their correlation with various growth components and with leaf injury. The highestcorrelations were found for Na+ in the leaf blade on day 21 with injury at day 42 in both experiments (r= 0.7). Earliermeasurements of Na+ or of injury had lower correlations. The most sensitive growth components were tiller number plant–1

and shootwater content (gwater g–1 dryweight), and thesewere correlated significantlywithNa+ and, to a lesser extent, withK+/Na+. These studies showed that exposure for at least 42 daysmay be needed to clearly demonstrate the beneficial effect ofthe trait for Na+ exclusion on growth under salinity.

Additional keywords: criteria, HKT transporter, Oryza sativa, potassium, salinity, screening, sodium, tissue tolerance.

Received 23 May 2013, accepted 21 October 2013, published online 3 December 2013

Introduction

About 950million ha of the world’s crop production area hasbecome unsuitable for production due to the problem of soilsalinity and/or sodicity (Szabolcs 1989). Rice is one of the cropsmost sensitive to salt stress (Lafitte et al. 2004) but its cultivationis recommended during the process of saline soil reclamationdue to its high water requirement, which helps leach salts intodeeper zones (Bhumbla and Abrol 1978). The world’s foodsecurity demands that rice production must increase from 600to 800million tonnes by the year 2025 (Green et al. 2005). Salttolerance is a complex trait and understanding its molecularbasis is essential for breeding and transformation in cropplants (Yeo and Flowers 1986; Chinnusamy et al. 2005). Thepromising field of marker-assisted breeding can be a powerfultool to increase the efficiency of selection for stress adaptation inbreedingprograms (Witcombe et al. 2008). In rice,many attemptshave been made to understand the genetics and physiology ofsalt tolerance but progress has been slow due to the complexnature of this multigenic character. This may be because oflow selection efficiency for general agronomic traits and a lackof effective evaluation methods for assessing salt toleranceamong genotypes (Zeng et al. 2003).

The main adverse effects of salinity on plant growth are dueto osmotic effects, ionic toxicity and nutritional imbalancessuch as reduced K+ uptake (Läuchli and Epstein 1990). Highconcentrations of salt around roots affect water relations of theplant and may generate signals that reduce cell expansionrates and stomatal conductance in shoot tissues (Davies andZhang 1991; Munns 2002). If roots do not exclude most ofthe salt from the soil solution while taking up water, Na+

travelling in the xylem stream may accumulate in leaves totoxic levels causing premature plant death (Munns 2005).Chloride (Cl–) may accompany the Na+ although it is lesslikely to become toxic (Munns and Tester 2008). Potassiumtransport is usually decreased, and the K+/Na+ ratio in leavescan drop to critically low levels. Halophytes can use Na+ and Cl–

for internal osmotic adjustment to grow under saline conditions(Flowers et al. 1977), but glycophytes mainly survive salineenvironments by excludingmost of theNa+ from the transpirationstream (reviewed by Munns 2005). The success of thismechanism in glycophytes depends on the balance betweensalt influx into the plant and its cellular compartmentation.Too little accumulation of ions in leaves can lead to poorosmotic adjustment, whereas too much may result in toxic

CSIRO PUBLISHING

Functional Plant Biology, 2014, 41, 379–390http://dx.doi.org/10.1071/FP13158

Journal compilation � CSIRO 2014 www.publish.csiro.au/journals/fpb

concentrations of Na+ and/or Cl– in leaves or reproductive tissues(Flowers et al. 1986).

Exclusion of Na+ from photosynthetically active leaves is akey strategy for plants to thrive under salinity (Yeo and Flowers1986; Munns and Tester 2008), and the concentration of Na+ inleaves of crop plants has been used as one of the indicators ofsalinity tolerance (Ashraf and Harris 2004; Munns et al. 2006).Na+ exclusion was proposed as a criterion for assessing salttolerance in wheat and rice (Yeo et al. 1990; Munns et al.2006). Generally, salt tolerant lines of rice maintain lower Na+

in their leaves than salt sensitive lines when exposed to salt stress(Flowers andYeo 1981; Yeo et al. 1990; Lutts andGuerrir 1995).The decrease in K+/Na+ ratio may relate directly to a decrease inyield in someconditions (Aschet al. 2000).More recently, Plattenet al. (2013) reported a strong correlation between leaf blade Na+

concentration and salt tolerance (as assessed by leaf injurysymptoms) in the majority of accessions of two rice species.

Maintenance of a low net rate of uptake of Na+ from the soil isaccomplished by membranes that have low permeability to Na+.In rice and other cereals, Na+ can enter through K+-selectivecation transporters such as the ‘class 2’ HKT transporters (highaffinity potassium transporters) that while having a high K+/Na+

selectivity would allow uptake of Na+ under conditions of highsoil NaCl (Garciadeblás et al. 2003; Horie et al. 2012). Na+ maybe effluxed from the cells and hence to the soil by the SOS1 Na+/H+ antiporter (Shi et al. 2002), which has an orthologue in rice(Martinez-Atienza et al. 2007). The small fraction that entersthe transpiration stream can be further controlled by the ‘class 1’HKT transporters which in wheat are suggested to reduce Na+

transport to leaves by retrieval from the transpiration streamand deposition in xylem parenchyma cells (James et al. 2006).In wheat, the transporter encoded by the candidate gene forNax1, TmHKT1;4, is responsible for retrieval of Na+ from thetranspiration stream for storage in the leaf sheath tissue (Huanget al. 2006; James et al. 2006). The transporter encoded by thecandidate gene for Nax2, TmHKT1;5, plays a similar role, butfunctions primarily in root tissue (Byrt et al. 2007; Munns et al.2012). Targeted overexpression of the Arabidopsis HKT1;5homologue, AtHKT1;1, in Arabidopsis and rice has beenshown to increase Na+ exclusion from the shoot (Møller et al.2009; Plett et al. 2010). Cotsaftis et al. (2012) also suggested thatretrieval of Na+ from the xylem transpiration stream in the rootor leaf sheath is a possible mechanism to maintain low Na+ inleaf blades of rice, and explained Na+ exclusion in rice by 3Dmodelling of HKT transporters (OsHKT1;4 and OsHKT1;5).They suggested that Na+ accumulation in the leaf blade iscontrolled by regulation of complex gene expression, alternatesplicing and the protein structure of transporters under salt stress.

In a previous study, a recombinant inbred line (RIL)population from the cross between two parents CO39 andMoroberekan with differences in leaf Na+ concentration waschosen to look for quantitative trait loci (QTL) for ionregulation, on the understanding that this would be useful inselecting for salt tolerance. CO39 has 75% lower Na+

concentration in leaves than Moroberekan (Ul Haq et al.2010). On day 7 of salt stress, two QTL each for Na+ and K+/Na+ ratio in fully expanded leaves were identified onChromosome 1 with high likelihood of odds score (LOD) andexplained up to 35 and 38% of the total phenotypic variation in

these traits respectively (Ul Haq et al. 2010). After 21 days stressthe same QTL for leaf Na+ was detected on Chromosome 1(11.56–11.72Mbp) with LOD score of 9.8 which explained24% of the total phenotypic variation. QTL were also foundfor Na+ accumulation in leaf sheaths after 21 days but not after7 days salt stress (Ul Haq et al. 2010). Fine mapping identifieda chromosomal region that was most likely the same as the SKC1QTL (Lin et al. 2004), later identified as the Na+ transporterOsHKT1;5, responsible for retrieval of Na+ in rice by beingunloaded directly from the xylem sap (Ren et al. 2005).

In this paper, which presents results of two experiments withthe same parents and population, the relation between leaf Na+

and degree of injury over time, and biomass production (%) wasmeasured. As the biomass production of the two parents differedsubstantially in the absence of salt, the first experiment included acontrol treatment where the effects of salinity on tiller numberand total biomass could be calculated, but only a small populationnumber (32 RILs) could be handled. The second experimentincluded a larger population number (120 RILs) but with salttreatment only. Here, salt injury was measured over time (days21 and 42) as well as Na+ in sheath and blade of leaf at earliertimes to investigate the reliability of leaf injury in predictingfuture growth in saline conditions and in selection for salttolerance. This allowed us to relate Na+ accumulation toinjury, to speculate at what concentration sodium becomestoxic in rice, and to relate this to growth in saline conditions.

Materials and methodsPlant material

The F9 RILs mapping population employed in this study wasdeveloped at the International Rice Research Institute (IRRI),Philippines, from the cross CO39�Moroberekan. The maternalparent variety CO39 is a salt tolerant lowland Indica cultivar withmedium height having originated from India. Moroberekan is asalt sensitive, tropical upland Japonica variety of long staturehaving its origin in West Africa. The population is skewedtowards the CO39 alleles (80%) (Champoux et al. 1995). InStudy 1, a total of 32 RILs were included, whereas in Study 2,a larger set of 120 RILs was investigated for ion accumulationand growth attributes under salt stress.

Sowing of rice nursery

The rice seeds of RILs and parent varieties (CO39 andMoroberekan) were sown in P-84 plug trays (Desch Plantpak,Malden, UK) filled with John Innes Compost No. 1 (Corwen,Clwyd, UK) at Pen y Ffridd, Research Station, BangorUniversity, Wales, UK. These trays were kept on a bench in agreenhouse (at 25�C) andwatered every day until transplantation.Twenty-three days after seed sowing, healthy and uniformseedlings of each line (at the three leaf stage) weretransplanted into 2 L plastic pots (one seedling per pot inStudy 1 and 2 seedlings per pot in Study 2) lined withhorticultural fleece and filled with John Innes No. 2 Compost(Lawrence and Newell 1939).

Setup of flood bench system

The flood bench system comprised of plastic tanks(80� 56.3� 32.5 cm) each containing 24 pots of 2 L capacity.

380 Functional Plant Biology T. Ul Haq et al.

The water reservoirs each had a capacity of more than 200 L ofwater providedwith submersible electric pumps.Tankswithplantpots were placed on an iron bench ~1m high off the greenhousefloor. Water reservoirs were placed on the floor underneath thetanks. Submersible electric pumpswere placed inwater reservoirstopumpwater into tanks.Therewere twoconnectionsof thewaterto each tank, one forwater to enter into the tank from the reservoirand the other for water to drain from the tank to the reservoir. Potswere placed into each flood bench tank in a randomised layoutwith three replications in Study 1 and two replicates in Study 2(with three replicates for CO39 and Moroberekan). The plantswere flooded once a day for at least 15min, twice on hot days andalso twicewhen plants were fully grown in size. Nutrient solutionhaving Phostrogen plant food (Phostrogen, Corwen,Wales, UK)at a rate of 1.0 g L–1 plus 0.5mLL–1 micro-nutrients (HoaglandandArnon 1950) and 0.1mLL–1 sodium silicatewere supplied tothe plants with the irrigation solution. After establishment ofseedlings (1week after transplantation, 28 days after sowing), saltstress was started in daily increments of 25mM for NaCl and1.25mM for CaCl2 to reach the final level of 100mMNaCl + 5.0mM CaCl2. The salinity level was maintained in thesolution by recording the electrical conductivity (EC) with aportable waterproof conductivity meter (Hanna Instruments,Manchester, UK) in each reservoir twice a week. Water lost bytranspiration and evaporation was replaced to keep the reservoirvolume at 200 L. Supplementary 400W high pressure sodiumvapour lamps were used to maintain a minimum photon fluxdensity of photosynthetically active radiation (400–700 nm) of350mmolm–2 s–1 during the 16 h photoperiod. The minimumtemperature was maintained at 25�C during the photoperiodand 20�C during the dark period.

Parental varieties CO39, Moroberekan and 32 F9 RILs wereevaluated for salt tolerance under abovementioned control and/orsaline conditions for Study1, and120 F9RILs alongwith parentalvarieties in the saline conditions for Study 2.

Measurements

In Study 1, after 3weeks (day 21) of salt stress, the youngest fullyexpanded leafwas sampled fromeachplant for ionic analysis.Thesamples were washed with distilled water, dried with tissue paperand stored in labelled, 1.5mL micro centrifuge tubes (SigmaChemicals, Poole, UK) at �20�C in a commercial freezer for2 weeks before sap extraction. After thawing, the tissues werecrushed with a tapered end steel rod. Two pin holes were made inthe base and cap of each Eppendorf tube. Each tube was placedinside a second Eppendorf tube and centrifuged at 8000g for10min. The sap was collected into the lower tube and remainingtissue left in the upper one was discarded (Gorham et al. 1997).The sap was either analysed immediately or frozen for lateranalysis. Sodium and potassium ions were analysed in dilutedsap with a Jenway PFP-7 flame photometer (Bibby ScientificLimited, Stone, UK). Thewhole shoot of plants fromboth controland salt treatments were harvested on day 42 of stress, and tillersplant–1, shoot FW and DW were recorded. The shoot sampleswere oven-dried at 65�C for 72 h before measuring shoot DW.

In Study 2, after 7 days of salt stress, one of each pair ofseedlings from two replicates was harvested and two tissuesamples were taken. These were: (i) a youngest fully expanded

leaf and (ii) the combined leaf sheaths. Twenty-one days afterreaching full salt stress, one tiller of the second plant washarvested and samples were taken again from the youngestfully expanded leaf and combined leaf sheaths. The sampleswere washed briefly with distilled water, dried with tissue paperand stored in 1.5mL micro centrifuge tubes at �20�C in thefreezer for 1 week. Leaf sap was extracted from each sample andions were determined as described earlier. A portable SPAD-502Meter (Minolta, Osaka, Japan)was used to record the chlorophyllin leaves. Each mean value is the average of nine SPADobservations from leaf 1, 2 and 3 (one leaf on each of threetillers) on the main tillers of the same plant. The SPAD readingsfrom each leaf were taken from the base,middle and tip, along thelength of leaf to cover the whole (or the total) leaf area.

Leaf injury scoring

Inboth studies, leaf injury scoring in the saline treatmentwasdoneon days 21 (Score 1) and 42 (Score 2) on the recently fullyexpanded leaves on a scale of 1 to 4. The injury value 1 wasassigned to leaves showing little or no damage, a value of 2 withleaves showing slight damage, a value of 3where the leaf damagewas moderate, and a value of 4 to leaves experiencing severedamage.

Results

Effects of salinity on growth and appearance of CO39,Moroberekan and RILs, and correlations between leafinjury and biomass production

In Study 1, the two parents and 32RILswere grown in control and100mM NaCl for 42 days. The adverse effects of salinity onoverall plant appearance and biomass after 42 days at 100mMNaCl were greater onMoroberekan than on CO39.Moroberekanhadmore vigorous vegetative growth thanCO39, producing overdouble the biomass in the absence of salinity (Table 1), so thegenetic differences in ‘salt tolerance’ need to be expressed as apercentage of biomass in the control conditions. Moroberekanproduced a shoot freshweight under salt stress that was only 23%of the controls, whereas CO39 had a freshweight that was 42%ofcontrols (Table 1). Six RILs had an even higher percentage shootFW than CO39 (Fig. 1a). Shoot DW showed a similar result (seeFig. S1a, available as SupplementaryMaterial to this paper). Thevariation in shootFWandDWwashighly significant (P < 0.01) intreatments (saline and non-saline), RILs and salinity�RILinteractions as is revealed by the analysis of variance in Study1 (see Table S1, available as Supplementary Material to thispaper). The number of tillers plant–1 was the same (~12) in bothparent varieties in the control treatment (Table 1), and so could beused to assess effects of salinity on growth in the absence of acontrol treatment. Under salt stress, the number of tillers plant–1

was more in CO39 (9.0) than Moroberekan (4.7) on day 42 ofexposure to salt stress (Table 1). Out of the 32 RILs, 17 had tillersplant–1 greater than CO39 (Fig. 1b). Shoot water (g g–1 DW) ofparents and all RILs was reduced by salinity (Table 1), but theeffect of salinity�RILs interactionwas not significant (P = 0.07)(Table S1).

InStudy1, leaf injury forMoroberekanwas greater thanCO39on day 21 (Table 1). However, there was no correlation betweeninjury at day21andgrowth components onday42other than tiller

Selectable traits for salt tolerance in rice Functional Plant Biology 381

number and shoot water content (Table 2), suggesting that theinjury had until then little effect on biomass production, butmightpredict a future decrease in growth rate. Thiswas borne out by thefinding that injury at day 42 correlated negatively with all growthcomponents except for shoot DW (Table 2).

In Study 2, the parents and 120 RILs were grown in 100mMNaCl for 42 days without a control treatment. On day 42, shootFW and DW of CO39 were 167 and 104%, respectively, higherthanMoroberekan (Table 3). This confirmsCO39 as themore salttolerant cultivar as its biomass was less than Moroberekan in theabsence of salinity. The results of the first study indicated thattiller number and shoot water can be used as an index of salttolerance. The shoot water content was higher in CO39(2.8 g g–1DW) than in Moroberekan (1.9 g g–1 DW) and variedbetween 1.42 and 3.61 g g–1 in RILs population (Table 3). Tillersplant–1 were higher in CO39 (15.0) than Moroberekan (5.0). Thenumber of tillers plant–1 ranged between 4.5 and 16.5 in the RILspopulation (Table 3).

In Study 2, the overall appearance was more affected inMoroberekan than CO39 as shown by chlorophyll loss andinjury scores (Table 3). The phenotypic differences in parentalvarieties and RILs in response to salt stress can be further seen inpictures taken on day 42 (Fig. S2a, b). Chlorophyll concentrationin the upper leaves on day 42 was higher in CO39 thanMoroberekan (39 vs 13 SPAD units), and ranged between 10and42 in theRILpopulation (Table 3). Injurymeasuredonday21was higher in Moroberekan (3.7) than CO39 (1.4) and rangedbetween 1 and 3.8 for the RILs, and correlated significantly withtillers, shoot water, and even shoot FW and DW (Table 4). Injurymeasured on day 42 did not have a much higher score than onday 21 (Table 3) but the correlations with all growth componentswere higher than at the earlier time of measurement (Table 4).

Variation in ion accumulation in CO39, Moroberekanand RILs

In Study 1, Na+ accumulation in the blade of recently fullyexpanded leaves at day 21 was 75% lower in CO39 (15mM inleaf sap) than Moroberekan (64mM) in the saline treatment

Table 1. Physiological and growth attributes of CO39, Moroberekan and population of 32 recombinant inbred lines (RILs) grown incontrol (non-saline) conditions and in 100mM NaCl for 42 days (Study 1)

Growthwasmeasured onday 42 for both control and saline treatments. For the saline treatment only, ionsweremeasured onday 21 in the youngestfully expanded leaf, and injury scoresmeasured on day 21 and day 42 on a scale of 1 to 4. Each value ismean of three replicates, except for theRILs

where it is average of 96 plants

Traits CO39 Moroberekan RILs RILsControl Saline Control Saline Control Range Saline Range

Shoot FW (g) 42.1 17.5 89.8 19.5 70.6 22.8–185.5 21.1 8.7–32.5SFW % of control – 41.8 – 22.6 – – 35.9 10.6–110.3Shoot DW (g) 6.8 4.3 15.2 4.6 12.6 3.79–35.1 5.1 2.0–9.0SDW % of control – 63.5 – 30.2 – – 46.1 19.7–93.7Tillers plant–1 11.7 9.0 12.0 4.67 11.7 3.0–28.0 9.4 3.0–18.0Shoot water (g g–1 DW) 5.2 3.1 4.9 3.2 4.8 3.0–6.8 3.2 2.1–4.1Leaf Na+ (mM) on day 21 1.4 14.6 1.1 63.9 1.2 0.9–3.0 30.5 2.7–136.2Leaf K+ (mM) on day 21 363.0 336.0 263.0 245.0 277.0 173.0–397.0 283.0 166.0–369.0K+/Na+ ratio on day 21 264.5 32.9 239.7 3.8 231.8 84.5–312.8 27.3 2.0–113.1Score 1 on day 21 (top leaves) – 1.3 – 1.7 – – 1.8 1.0–3.7Score 2 on day 42 (top leaves) – 2.0 – 2.7 – – 2.5 1.0–4.0

R² = 0.440

02

0 10 20 30 40 50 60 70 80 90 100 110

4 6 8 10

Tillers plant–1 day 42

SFW% of control

12 14 16 18

1

2

3

4

5

6

7

8

9

0

2

4

6

8

10

12

14 (a)

MC

(b)

SD

W g

Num

ber

of li

nes

Fig. 1. (a) Frequency distribution for shoot fresh weight after 42 days in100mMNaCl as a percentage of control for RIL population (n= 32) in Study1. Normal curve for the distribution is also shown. The arrows indicate meanvalues for CO39 and Moroberekan. (b) Relationship between tillers plant–1

andSDWonday42 in100mMNaCl.The red andblue triangles indicatemeanvalues for CO39 and Moroberekan respectively.

382 Functional Plant Biology T. Ul Haq et al.

Table 2. Correlation matrix (Pearson’s 2-tailed) for various traits of 32 recombinant inbred lines (RILs) grown in saline (100mM NaCl) conditions(Study 1)

Na+, K+ andK+/Na+ are Leaf Na+, K+ and ratio on day 21; SFW, shoot FW (g); SDW, shoot DW (g); tiller, tillers plant–1; SWgg–1 DW, shoot water onDWbasison day 42; score 1 and score 2, leaf injury on days 21 and 42 on top leaves. Significant correlation indicated: *, P� 0.05; **, P� 0.01 levels; NS, non-significant

Traits Na+ K+ K+/Na+ SFW SDW SDW % C Tiller SWgg–1 SFW % C Score 1

K+ –0.415** 1.0 – – – – – – – –

K+/Na+ –0.652** 0.519** 1.0 – – – – – – –

SFW –0.093NS 0.346** 0.154NS 1.0 – – – – – –

SDW 0.018NS 0.372** 0.146NS 0.938** 1.0 – – – – –

SDW % C –0.145NS 0.096NS 0.019NS 0.362** 0.246* 1.0 – – – –

Tillers –0.224* 0.520** 0.489** 0.610** 0.667** 0.288** 1.0 – – –

SW g g–1 DW –0.274** –0.150NS –0.021NS –0.065NS –0.398** 0.199* –0.333 ** 1.0 – –

SFW % C –0.170NS 0.079NS 0.032NS 0.395** 0.250* 0.983** 0.266** 0.273** 1.0 –

Score 1 0.130NS 0.012NS 0.071NS 0.095NS 0.184NS –0.013NS –0.252* –0.295** –0.091NS 1.0Score 2 0.689** –0.532** –0.638** –0.307** –0.191NS –0.231* –0.332** –0.271** –0.242* 0.162NS

Table 3. Physiological and growth attributes of CO39, Moroberekan and population of 120 recombinant inbred lines(RILs) grown in 100mM NaCl for 42 days (Study 2)

Ions are givenonday21 in theyoungest fully expanded leaf, and injury scores onday21 andday42on a scale of 1 to 4.Eachvalue ismean of three replicates� s.e. except RILs where it is an average of 240 plants

Traits CO39 Moroberekan RILs RILs range

Growth, ions and injury on day 21/42 of salt stressShoot FW (g) on day 42 29.5 ± 1.16 11.04± 1.78 24.07± 0.61 9.6–40.9Shoot DW (g) on day 42 7.78 ± 0.08 3.81 ± 0.12 6.47 ± 0.15 3.3–10.8Tillers plant–1 on day 42 15.0 ± 1.15 5.0 ± 0.58 10.14± 0.28 4.5–16.5SPAD on day 42 33.13± 0.32 13.33± 0.88 27.80± 0.63 10–41.8Shoot water (g g–1dw.) 2.80 ± 0.18 1.89 ± 0.42 2.71 ± 0.04 1.42–3.61Na+ in leaf blade 7.9 ± 2.1 67.4 ± 15.8 26.2 ± 3.0 2.1–159.4Na+ in sheath 57.6 ± 10.5 129.7 ± 15.2 75.3 ± 3.9 10.6–208.5K+ in leaf blade 317± 19.0 264± 30.0 298± 3.0 194–401.0K+ in sheath 303± 35.0 208± 15.0 258± 4.0 154–355.0K+/Na+ in leaf blade 47.5 ± 14.6 4.2 ± 0.6 36.2 ± 3.0 1.8–154.2K+/Na+ in sheath 5.7 ± 1.2 1.65 ± 0.2 6.6 ± 0.7 1.0–26.7Injury Score 1 on day 21 1.4 ± 0.3 3.7 ± 0.5 1.7 ± 0.1 1.0–3.8Injury Score 2 on day 42 2.4 ± 0.3 3.9 ± 0.0 2.3 ± 0.1 1.0–4.0

Table 4. Correlation matrix for ion accumulation in leaf blade and sheath for 120 RILs on day 7 and 21, injury scores on day 21 (Score 1) and day 42(Score 2) and growth traits on day 42 under salts stress (100mM NaCl) (Study 2)

Significant differences (Pearsoncorrelation, 2-tailed) are indicated: *,P� 0.05; **,P� 0.01 levels;NS, non-significant;LK7, leafK+ concentrationonday7;LK/Na7, leafK+/Na+ ratioonday7;LNa21,LeafNa+concentrationonday21;LK21, leafK+concentrationonday21;LK/Na21, leafK+/Na+ ratioonday21;SWgg–1

DW, shoot water on DW basis on day 42

LNa7 LK7 LK/Na7 LNa21 LK21 LK/Na21 Tillers SPAD SFW SDW SW gg–1 Score 1

LNa7 1.0 – – – – – – – – – – –

LK 7 –0.25** 1.0 – – – – – – – – – –

LK/Na7 –0.67** 0.40** 1.0 – – – – – – – – –

LNa21 0.47** –0.30** –0.43** 1.0 – – – – – – – –

LK21 –0.01NS 0.50** 0.06NS –0.19* 1.0 – – – – – – –

LK/Na21 –0.47** 0.27** 0.73** –0.57** –0.09NS 1.0 – – – – – –

Tillers –0.24** 0.35** 0.25* –0.56** 0.07NS 0.47** 1.0 – – – – –

SPAD –0.32** 0.10NS 0.27** –0.72** 0.30** 0.44** 0.48** 1.0 – – – –

SFW –0.27** 0.28** 0.19 NS –0.63** –0.13NS 0.40** 0.62** 0.67** 1.0 – – –

SDW –0.21* 0.33** 0.22* –0.48** –0.08NS 0.44** 0.62** 0.53** 0.91** 1.0 – –

SW g g–1 –0.30** 0.02NS 0.07NS –0.60** –0.14NS 0.12NS 0.23** 0.57** 0.52** 0.17NS 1.0 –

Score 1 0.08NS 0.06NS 0.06NS 0.35** 0.12NS –0.14NS –0.42** –0.54** –0.38** –0.28** –0.38** 1.0Score 2 0.32** –0.17NS –0.27** 0.71** 0.27* –0.46** –0.55** –0.87** –0.72** –0.57** –0.54** 0.55**

Selectable traits for salt tolerance in rice Functional Plant Biology 383

(Table 1). Under salinity treatment, three out of 32 RILs had Na+

concentration higher than Moroberekan, whereas in 14 RILsthe Na+ concentration was even lower than CO39, showingtransgressive segregation towards CO39, the low Na+ parentvariety (Fig. S1b). The analysis of variance revealed thattreatments, RILs and salinity�RILs interaction had highlysignificant (P < 0.01) effects on Na+ concentrations of the leafblades (Table S1). Leaf K+ was very much higher in CO39 thanMoroberekan, being 363 vs 263mM in the control treatment,and only a little less in the salt treatment (Table 1). Undersalinity, two out of 32 RILs had K+ concentration higher thanCO39, whereas, six RILs had values of K+ concentration lessthan Moroberekan, showing an approximate normal frequencydistribution for the above trait (Fig. S3a). There was a highlysignificant negative correlation of Na+with leaf K+ concentrationand K+/Na+ after day 21 of salt stress (Table 2). CO39 showedabout a 9-fold greater K+/Na+ ratio than Moroberekan undersalinity stress (Table 1). Eleven out of 32 RILs maintained

values of K+/Na+ ratio even greater than the high parentCO39 (Fig. S3b). Differences in the K+/Na+ ratio of expandedleaves were highly significant (P < 0.01) in the case oftreatments, RILs and salinity�RILs interaction (Table S1).In summary, there was a significant increase in leaf Na+ undersalinity (Table 1) but the decrease in K+ was not significant. K+

concentrations increased in many RILs after salt stress.In Study 2, both the leaf blade and leaf sheath Na+

concentrations were measured on days 7 and 21 of salt stress.The leaf sheath had a much higher Na+ concentration thanthe leaf blade, especially at the later time of sampling(Table 3; Fig. 2a, b). The increase over time of Na+ in the leafblades was negligible for CO39 but large for Moroberekan(Fig. 2a). In the leaf sheath, the Na+ increased by 2-fold inCO39, and 3-fold in Moroberekan (Fig. 2b). K+ in theexpanded leaf blades and sheaths was higher in CO39 thanMoroberekan on day 21 of salt stress (Table 3). K+ in the leafsheaths on day 21 was even higher in CO39 compared with its

160

140

120

100

80

40

20

90

80

70

60

50

40

30

20

10

0

90

80

70

60

50

40

30

20

10

0

CO39 Moro CO39 Moro

0

60

Na+

in le

af s

ap (

mM

)K

+/N

a+ in

leaf

sap

K+/N

a+ in

she

ath

sap

Na+

in s

heat

h sa

p (m

M)

160

140

120

100

80

40

20

0

60

(a) (b)

(c) (d)

Fig. 2. Change in Na+ concentration in (a) leaf blade and (b) leaf sheath. K+/Na+ ratio of (c) leaf blade and(d) sheath sap on day 7 (open bars) and day 21 (closed bars) in CO39 and Moroberekan under salt stress inStudy 2 (Error bars show s.e.). Leaf and sheath K+ concentration values are in Table 3.

384 Functional Plant Biology T. Ul Haq et al.

concentration on day 7 of salt stress (data not shown). K+/Na+

ratios decreased with exposure time in both blades and sheaths ofboth parents, but the reductions were greater in Moroberekanthan CO39 (Fig. 2c, d).

Correlations between ion accumulation, leaf injuryand growth attributes

In Study 1, there was no correlation between leaf Na+ measuredon day 21 and injury scored on that day, but the correlation withinjury scored on day 42 was highly significant (r= 0.689)(Table 2; Fig. 3). This suggests that most of the later leafinjury could have been due to the prior accumulation of toxicconcentrations of Na+, particularly in the recently fully expandedleaves. Leaf K+ had a significant negative correlation with injuryon day 42 (r= –0.532), suggesting that K+ reduced the toxiceffects of Na+ by improving K+/Na+ ratio of leaves under saltstress (Table 2). Na+ concentrations on day 21 were notsignificantly correlated with shoot FW (SFW), shoot DW(SDW), SFW%C or SDW%C (shoot fresh or dry weightrelative to weight in control); however, correlations with tillersplant–1 and shoot water were significant (Table 2). Leaf K+ andK+/Na+ ratios showed positive and highly significant correlationswith SFW, SDW and tillers plant–1 but not when SFW or SDWwas expressed as a percentage of the controls (Table 2).

InStudy2, therewasno injury to leaves onday7; injury startedon day 21, and became severe on day 42 of salt stress, as shown inthe photographs in Fig. S2a, b. Na+ in the leaf blade on day 21correlated with injury on that day (r = 0.35) and more stronglywith injury on day 42 (r= 0.71), as shown in Fig. 4a, usingchlorophyll measurements with the SPADmeter as a measure ofinjury. This indicates that Na+ concentrations even when not yettoxic are predictive of subsequent injury. Likewise although Na+

on day 7 of treatment was not associated with injury on the dayof measurement, Na+ concentrations on day 7 and day 21 wereassociated with injury on day 42 (r= 0.32). Leaf K+/Na+ waspositively correlated with injury (Fig. 4b) and highly correlatedwhen K+/Na+ was below 10, suggesting a critical ratio necessaryto avoid injury. Na+ on day 7 was negatively correlated with allgrowth components, although the correlations were not as highas for Na+ on day 21 (Table 4). Fig. 5a shows the correlationof leaf Na+ with tillers plant–1. The correlation of K+/Na+ withgrowth components was not as great as for Na+ (Table 4);however, a decrease in K+/Na+ ratios below 10 was associatedwith a fall in tiller number from 10 to 4 (Fig. 5b). It is thereforelikely that injury is mainly because of accumulation of Na+ inleaves and the resultant decrease in K+/Na+, which had asignificant effect on shoot FW and DW under salinity stress.

Na+ concentrations in the sheath had a similar relation to injuryand growth components (Table S2a) as did Na+ in the leaf(Table 4); however, the sheath tissue appeared to tolerate ahigher level of Na+ concentration. Na+ in the sheath wasassociated with a decline in chlorophyll only above 100mM(Fig. 6a) whereas in the blade chlorophyll declined with Na+

concentrations above 50mM (Fig. 4a). Consistent with this,sheath ratios of K+/Na+ were associated with a significantdecline in chlorophyll only below 5 (Fig. 6b) in contrast to theleaf blade where the decline was associated with ratios below 10(Fig. 4b). K+ concentrations in the sheath at day 7 had similar

relationships to growth components as did K+ in the leaf, buton day 21 thereweremuch stronger and positive correlationswithall growth components and stronger negative correlations with

1.00 20 40 60 80 100 120

1.5

2.0

2.5

3.0

3.5

4.0

4.5

Sco

re-2

da

y 42

Leaf Na+ (mM) day 21

R² = 0.615

Fig. 3. Relationship of leaf injury Score 2 measured on day 42 with leafblade Na+ measured on day 21 of salt stress in Study 1.

0

5

10

15

20

25

30

35

40

45

0 25 50 75 100 125 150 175

0 25 50 75 100 125 150 175

Chl

orop

hyll

(SP

AD

) da

y 42

0

5

10

15

20

25

30

35

40

45

Leaf Na+ (mM) day 21

Leaf K+/Na+ day 21

R² = 0.532

R² = 0.258

(a)

(b)

Fig. 4. Relationship of (a) SPAD (Chlorophyll) with leaf blade sap Na+ and(b) SPAD with leaf blade sap K+/Na+ on day 21/42 of salt stress in Study 2.

Selectable traits for salt tolerance in rice Functional Plant Biology 385

leaf injury (Table S2a). The K+/Na+ ratio of leaf sheaths had asignificant effect on the number of tillers, SPAD and SDW butnot on SFW and shoot water on day 7 (Table S2a).

Discussion

Mechanisms of salt tolerance, and selectable traits

When plants are first exposed to salinity, the rapid reductions ingrowth are due to the osmotic effect of the salt concentration inthe root medium which changes the water status of plants, andaffects the younggrowing leaves via root signals. Later, the build-up of Na+ to toxic levels in older leaves affects growth due to lackof photosynthate supply to growing tissues (Munns 1993, 2002).This was described as a two-phase growth response to salinity(Munns 1993). In a prolonged study with various wheatgenotypes differing in the ability to exclude Na+, differencesin growth rates between genotypes were not observed until 20%of the leaves were injured, but by then growth rates were verymuch less than controls due to the osmotic effect of the salt(Munns et al. 1995). In rice, rapid decreases in leaf growthoccurred with KCl and mannitol as well as with NaCl (Yeo

et al. 1991). These authors proposed that a clear distinction canbe made between the initial effects of salinity on growth whichare due to osmotic effects and the long-term effects that resultfrom the accumulation of salt within expanded leaves. When thesalinity was low, leaf elongation rates could recover after aninitial reduction, further showing the effects had been osmoticrather than toxic (Yeo et al. 1991). Consistent with this, Luttset al. (1995) reported reductions in relative growth rates (RGR) of10 rice cultivars over the first week of salinity of ~40%. In thesecond week, RGR of six cultivars were still reduced but in four(including the salt tolerant landraces Nona Bokra and Pokkali)the RGR had recovered. Moradi and Ismail (2007) showedimmediate and sustained reductions in photosynthesis in threerice cultivars that were due to stomatal effects, suggesting waterrelations rather than ion toxicity. In the two experimentsreported here, there was no sign of leaf injury by day 7, soNa+hadnot reached toxic levels by that timeyet growth rateswerealready obviously reduced (data not shown). It was not until day21 in both studies that injury was considerable, by which timebiomass was already considerably affected. There were smalldifferences in shoot fresh and dry weights of 19 RILs which were

2

4

6

8

10

12

14

16

18

20

0 50 100 150 200

20 40 60 80 100 120 140 160

4

6

8

10

12

14

16

18

Till

ers

pla

nt–1

R² = 0.319

R² = 0.216

(a)

(b)

Leaf Na+ (mM) day 21

Leaf K+/Na+ day 21

Fig. 5. Relationship of (a) leaf blade sap Na+ with tillers plant–1 and(b) Leaf blade sap K+/Na+ with tillers plant–1 on day 21/42 of salt stress inStudy 2.

0

5

10

15

20

25

30

35

40

45

0

5

10

15

20

25

30

35

40

45

Chl

orop

hyll

(SP

AD

) d

ay 4

2

R² = 0.434

R² = 0.154

0 50 100 150 200 250

Sheath Na+ (mM) day 21

Sheath K+/Na+ day 21

0 5 10 15 20 25 30 35 40 45

(a)

(b)

Fig. 6. Relationship of (a) SPAD (Chlorophyll) with sheath sap Na+

concentration and (b) SPAD with sheath sap K+/Na+ on day 21/42 of saltstress in Study 2.

386 Functional Plant Biology T. Ul Haq et al.

included in both studies; however the RILs, which were high forshoot fresh anddryweights inStudy1,were alsohigher inStudy2and vice versa. Significant correlations were observed betweenshoot FW (r= 0.568) as well as shoot DW (r= 0.566) in Studies 1and 2.

The aim of this study was to compare different selectabletraits for reliable detection of salinity tolerant lines of rice, andin particular to quantify the relationships between Na+

accumulation, leaf injury, and growth under salt stress. Plantswere grown for a long enough time to allow the accumulationof Na+ to toxic concentrations in older leaves to affect thegrowth of young leaves via reduction in photosynthate supply.The CO39�Moroberekan RIL population was used alongsideboth parents. Although Moroberekan had a higher Na+

accumulation, it was more vigorous and produced twice thebiomass of CO39 over time – and so it was expected that theremight be segregation for the several genes (Ren et al. 2005;Walia et al. 2005; Møller et al. 2009) responsible for Na+

exclusion with combinations of genes affecting FW and DWproduction (Prasad et al. 2000; Koyama et al. 2001; Plett et al.2010).

As an index of salt tolerance for 32 RILs we measured thepercent biomass production in saline vs non-saline controlconditions after 42 days under 100mM NaCl, and assessed thecorrelationwithNa+ andK+concentrations in leaves andwith leafinjury. For a larger number of RILs (120) we measured thebiomass production and leaf injury in the presence of 100mMNaCl (without controls this time, due to lack of space to maintainplants growing at optimum growth rates for such a large numberof genotypes) and assessed the relationship of leaf blade andsheath Na+ and K+ with injury. We also measured the tillersplant–1, chlorophyll concentration with a SPADmeter, and shootwater content on a dry weight basis. We have focussed onrecognising traits which could be more useful for reliableselection from a large number of individuals with limitedresources and time.

This study confirmed CO39 as a more salt tolerant ricecultivar than Moroberekan, and identified five of 32 RILs withsalt tolerance (% control biomass) greater thanCO39. In addition,several RILs were shown in the second study as having evenlower Na+ accumulation than CO39, along with very high tillersplant–1 and shoot biomass, indicating that these would have agreater salt tolerance.

Indica rice in general has the ability to exclude Na+ to avoidtoxic accumulation in leaves, and to take up more K+ to maintaina high K+/Na+ balance in the shoot than does Japonica rice(Gregorio and Senadhira 1993). The Indica variety CO39 hadthe ability to exclude Na+ from the shoot to a greater extentthan the Japonica variety Moroberekan in which Na+ levels were4-fold higher. Moreover, the CO39 also had the ability to takeup more K+ and maintain a higher K+/Na+ ratio in the shootthan Moroberekan. A previous study (Ul Haq et al. 2010) hadidentifiedfiveQTL associatedwith these traits onChromosome 1that mapped to the region containing the SKC1/Saltol locus,which has been identified as the Na+ transporter OsHKT1;5 (Renet al. 2005). This gene retrieves Na+ from the xylem in roots andso prevents toxic accumulation in leaves (Ren et al. 2005;Cotsaftis et al. 2012). Allelic variation in OsHKT1;5 is largelyresponsible for the differences in Na+ accumulation in leaves

between different rice varieties and accessions (Platten et al.2013).

Our study has shown that low Na+ accumulation in the leafblades was strongly associated with leaf injury, and is presumedto be the main cause of it. Na+ accumulation was useful inidentifying RILs with higher salt tolerance. Injury of top leaveswas useful in the absence of growth rate measurements inrelation to the growth in control conditions. However, therewas no correlation between Na+ exclusion and growth evenat day 42 in Study 1 although it was in Study 2. It is possible thatif Study 1 had been conducted over a longer period of time, thenthe same results would have been found as for Study 2. Itappears that plants might have grown faster and differences insalt tolerance become more distinct in Study 2, compared withStudy 1 CO39 reached 17.5 g with nine tillers under salt but inStudy 2 it reached 29.5 g with 15 tillers. In contrast,Moroberekan had the same tiller number (five) but a lowerSFW in the second study (Table 1). This illustrates theimportance of experimental replication when assessing aquantitative trait, as it is difficult to control all environmentalvariables within a glasshouse over different seasons, andgenotype� environment interactions are so strong.

Several earlier papers reporting work with several differentrice cultivars or genotypes have shown a correlation betweengenetic differences in the Na+ concentration in a leaf and salttolerance as expressed by % control biomass production (Ul Haqet al. 2008), plant survival (Yeo and Flowers 1984, 1986), orthe degree of injury or leaf death under salinity stress (Yeo et al.1990; Gregorio et al. 1997).

To assess the level of leaf injury, scoring can be a useful andinexpensive indicator. The score used herewas 1–4, although 1–5(Ul Haq et al. 2010) and 1–9 (Gregorio et al. 1997; Negrão et al.2013; Platten et al. 2013) have been successfully used. In thesecond experiment, chlorophyll concentration was measuredwith the SPAD meter, which might be expected to give moreprecise quantitative results, but the correlations between SPADand the sensitive growth attributes andwithNa+were very similarto thosewith injury scores (Table 4), suggesting that visual scoresare a reliable substitute for chlorophyll estimates.

What is causing the leaf injury?

Significant positive correlations of Na+ with leaf injury scores inthese studies indicates that Na+ build-up to a toxic level is themain cause of leaf senescence particularly of the older leaves. Linet al. (2004) also correlated Na+ accumulation with leaf saltdamage under high salinity in rice. The correlation of salinity-induced leaf injury with whole-plant and leaf blade Na+

concentrations was high in nearly all accessions of Oryzasativa and O. glaberrima (Platten et al. 2013). There are twoother explanations for the onset of leaf injury: that it was due totoxic accumulation in the leaf sheath, which reached a higherlevel before the blade; or that it was due to the salt around theroots, that is, osmotic stress.

Sheath Na+ concentrations were at least three times the leafblade, and approached potentially toxic concentrations earlierthan in the blade (Fig. 2b). The leaf sheath concentration wasalways higher than in the blade on a tissue water basis, andincidentally much higher again on a DW basis (the water of the

Selectable traits for salt tolerance in rice Functional Plant Biology 387

sheath being almost twice as high as the blade). Differences upto 4-fold in Na+ accumulation between sheath and leaf blade inrice were documented by Yeo and Flowers (1982). Platten et al.(2013) found sheath : blade ratios of 2–3 depending on genotype.It is possible that this leaf sheath concentration is responsible forthe onset of injury seen in the leaf blade.Basedon thevisual injuryas seen in the photographs from Study 2, when the concentrationof Na+ in the leaf blades remained below 20mM and in sheathbelow 65mM after 3 weeks under salt stress, there was nodamage to leaves and leaves remained healthy and green evenafter another 21 days of exposure. However, when the Na+ inexpanded leaves exceeded 50mM and in the sheath exceeded100mMafter 3weeks and continued to rise, it caused a significantinjury. The higher concentration of Na+ in sheath tissue than leafblademay be due to removal of Na+ from the transpiration streamand its deposition in xylem parenchyma cells in the sheath by theNa+ transporter OsHKT1;4 (Cotsaftis et al. 2012). The authors ofthat study proposed that allelic variation in OsHKT1;5 due toalternative splicing and with different levels of expression in oldversus young leaves could be important in the redistribution ofNa+ in leaf blades within the shoot.

Alternately, leaf injury may be promoted by the osmoticeffects of the salt outside the roots, accelerating leafsenescence as does water stress caused by dry soils.

Role of leaf K+ in salt tolerance

It has been often observed that the accumulation of Na+ in shoottissues is accompanied by a reduction in shootK+ concentrations,resulting in decreased K+/Na+ ratios (Asch et al. 2000). Thereduction in leaf K+ concentration found here was less in CO39than Moroberekan. We noted that K+ concentration was higher(27%) in leaves of CO39 than Moroberekan in non-salineconditions as well as 28% higher at 21 days of salt stress(Table 1). CO39 had a higher K+ concentration in the sheathaswell as leaf blade, however, differencesweremore pronouncedat 21 days than 7 days salt stress resulting in a higher K+/Na+ ratio(Fig. 2c, d). A good supply of K+ to plants canminimise injuriouseffects of high Na+ under salinity andmaintain high K+/Na+ ratioin cell cytosol when exposed to NaCl salinity (Carden et al.2003; Golldack et al. 2003; Peng and Ismail 2004). However,differences between susceptible Japonica and Indica rice typeswere highly significant for all growth parameters and shoot Na+

but not for shoot K+ concentration (Lee et al. 2003). TolerantIndica varietiesmaintained very highK+ in the shoot but Japonicavarieties showed no relation between K+ and salt tolerance,showing that shoot K+ alone did not confer salinity tolerance(Lee et al. 2003).

Useful morphological and physiological traitsfor screening in rice

The measurements to be taken in an experiment depend on theobjective of screening. If the objective is to develop varieties forgood yield on salt affected lands, absolute yield or biomass atanthesis is the appropriate criterion of selection. However, if thepurpose of selection is to identify the better sources of salttolerance to further use in breeding programs then SDWrelative to control conditions would be a more valuable indexto use for screening against salinity. However, this is not possible

for large populations. Salt tolerance measured with a largepopulation or number of genotypes is not feasible because ofthe space required to grow plants in the control treatment withenough space to allow optimal growth and sufficient replicationsto produce a meaningful statistical value. This study showedthat for experiments when genotype numbers preclude thegrowth in control conditions as well, and when growth in thatcondition varies between parents and hence between genotypes,then tiller number and shoot water content are sensitive andreliable traits. This would not be valid if the parents differed intiller number in control conditions. Then shoot water content,which reflects biomass as well as desiccation to upper leavescaused by injury, is valid. These traits are recommended as theyreflect the growth of the plant during the period of salinity andhence predictive of future biomass that is related to yield.

However, injury as a trait also has a place as it can bemeasuredeasily and non-destructively, on the top 2–3 leaves either byvisual scoring or by chlorophyll concentration with a SPADmeter. It is highly correlated with Na+ accumulation, and so isa useful trait for identifying QTL for Na+ exclusion. It can also bepredictive for future biomass production.

An aspect that must not be neglected during the process ofselection is the environmental conditions. Factors such astemperature, growth stage and RH have wide- spread effectson salt tolerance of plants. They need to be considered as integralcomponents of the selection process. The use of marker assistedselection (MAS) also provides a great potential for screening butdepends on the availability of tightly linked markers havinguniversal applicability. The one major advantage of thesemarkers over morpho-physiological markers is that they arenot affected by the growth stage and the environmentalconditions and are large in number.

In conclusion, this study confirmed that plants need to begrown for a lengthy period of time before genetic differences insalt tolerance appear, and before leaf injury is an indicator ofsalt tolerance. It showed that short-term exposure of even 21 daysmay not be enough to pick genetic differences, and that exposureof more than 42 days is required to get valuable information. Wehave also shown that much of the growth reduction, which waspreviously thought to be due to toxic ions, was due to the osmoticeffects of the salt around the roots, supporting the two phasemodel of salt stress response. The study showed that high Na+

accumulation in the leaf blades was strongly associated withleaf injury and is presumed to be the main cause of it. Injury oftop leaves was useful in predicting the ultimate salt toleranceof the RILs.

Acknowledgements

We thank John Quealy for helpful comments on the manuscript and theSchool of Plant Biology, University of Western Australia for hosting thefirst author as Visiting Scientist. Financial support for this work from HigherEducation Commission (HEC) of Pakistan and EUCommission is thankfullyacknowledged. The RILs were supplied by Dr Adam Price (University ofAberdeen) and Dr Brigitte Courtois (CIRAD, Montpellier, France).

References

Asch F, Dingkuhn M, Dorffling K, Miezan K (2000) Leaf K+/Na+ ratiopredicts salinity induced yield loss in irrigated rice. Euphytica 113,109–118. doi:10.1023/A:1003981313160

388 Functional Plant Biology T. Ul Haq et al.

Ashraf M, Harris PJ (2004) Potential biochemical indicators of salinitytolerance in plants. Plant Science 166, 3–16. doi:10.1016/j.plantsci.a2003.10.024

Bhumbla DR, Abrol IP (1978) Saline and sodic soils. I ‘Proceedings of theIRRI Symposium on Soils and Rice’. pp. 719–738. (International RiceResearch Institute: Manila, Philippines)

Byrt CS, Platten JD, Spielmeyer W, James RA, Lagudah ES, Dennis ES,Tester M, Munns R (2007) HKT1;5-like cation transporters linked toNa+ exclusion loci in wheat, Nax2 and Kna1. Plant Physiology 143,1918–1928. doi:10.1104/pp.106.093476

Carden DE, Walker DJ, Flowers TJ, Miller AJ (2003) Single-cellmeasurements of the contributions of cytosolic Na+ and K+ to salttolerance. Plant Physiology 131, 676–683. doi:10.1104/pp.011445

Champoux M, Wang G, Sarkarung S, Mackill D, O’Toole J, Huang N,McCouch SR (1995) Locating genes associated with root morphologyand drought avoidance in rice via linkage to molecular markers.Theoretical and Applied Genetics 90, 969–981. doi:10.1007/BF00222910

Chinnusamy V, Jagendorf A, Zhu JK (2005) Understanding and improvingsalt tolerance in plants. Crop Science 45, 437–448. doi:10.2135/cropsci2005.0437

Cotsaftis O, Plett D, Shirley N, Tester M, Hrmova M (2012) A two-stagedmodel of Na+ exclusion in rice explained by 3D modeling of HKTtransporters and alternative splicing. PLoS ONE 7(7), e39865.doi:10.1371/journal.pone.0039865

Davies WJ, Zhang J (1991) Root signals and the regulation of growth anddevelopment of plants in drying soil. Annual Review of Plant Physiologyand Plant Molecular Biology 42, 55–76. doi:10.1146/annurev.pp.42.060191.000415

Flowers TJ, Yeo AR (1981) Variability in the resistance of sodium chloridesalinity within rice varieties. New Phytologist 88, 363–373. doi:10.1111/j.1469-8137.1981.tb01731.x

Flowers TJ, Troke PF, Yeo AR (1977) The mechanism of salt tolerancein halophytes. Annual Review of Plant Physiology 28, 89–121.doi:10.1146/annurev.pp.28.060177.000513

Flowers TJ, Hajibagheri MA, Clipson NJW (1986) Halophytes. QuarterlyReview of Biology 61, 313–337. doi:10.1086/415032

Garciadeblás B, Senn ME, Bañuelos MA, Rodríguez-Navarro A (2003)Sodium transport and HKT transporters: the rice model. The PlantJournal 34, 788–801. doi:10.1046/j.1365-313X.2003.01764.x

Golldack D, Quigley F, Michalowski CB, Kamasani UR, Bohnert HJ (2003)Salinity stress tolerant and sensitive rice (Oryza sativaL.) regulateAKT1-type potassium channel transcripts differently. Plant Molecular Biology51, 71–81. doi:10.1023/A:1020763218045

Gorham J, Bridges J, Dubcovsky J, Dvorak J, Hollington PA, LuoMC, KhanJA (1997) Genetic analysis and physiology of a trait for enhancedK+/Na+

discrimination in wheat. New Phytologist 137, 109–116. doi:10.1046/j.1469-8137.1997.00825.x

Green RE, Cornell SJ, Scharlemann JPW, Balmford A (2005) Farming andthe fate of wild nature. Science 307, 550–555. doi:10.1126/science.1106049

Gregorio GB, Senadhira D (1993) Genetic analysis of salinity tolerance inrice (Oryza sativa L.). Theoretical and Applied Genetics 86, 333–338.

Gregorio GB, Senadhira D, Mendoza RD (1997) Screening rice for salinitytolerance. IRRI Discussion Paper Series No. 22. International RiceResearch Institute, Manila, Philippines.

Ul Haq T, Akhtar J, Gorham J, Steele KA, KhalidM (2008) Genetic mappingof QTLs, controlling shoot fresh and dry weight under salt stress in rice(Oryza sativaL.) cross between CO39�Moroberekan. Pakistan Journalof Botany 40, 2369–2381.

Ul Haq T, Gorham J, Akhtar J, Akhtar N, Steele KA (2010) Dynamic QTLfor salt stress components on chromosome 1 of rice. Functional PlantBiology 37, 634–645. doi:10.1071/FP09247

Hoagland DR, Arnon DI (1950) The water culture method for growingplant without soil. California Agriculture Experiment Station Circular347, 39.

Horie T, Karahara I, Katsuhara M (2012) Salinity tolerance mechanisms inglycophytes: an overviewwith the central focus on rice plants. Rice 5, 11.doi:10.1186/1939-8433-5-11

Huang S, Spielmeyer W, Lagudah ES, James RA, Platten JD, Dennis ES,Munns R (2006) A sodium transporter (HKT7) is a candidate for Nax1, agene for salt tolerance in durumwheat.Plant Physiology142, 1718–1727.doi:10.1104/pp.106.088864

James RA, Davenport R, Munns R (2006) Physiological characterization oftwo genes for Na+ exclusion in wheat: Nax1 and Nax2. Plant Physiology142, 1537–1547. doi:10.1104/pp.106.086538

Koyama ML, Levesley A, Koebner RMD, Flowers TJ, Yeo AR (2001)Quantitative trait loci for component physiological traits determiningsalt tolerance in rice. Plant Physiology 125, 406–422. doi:10.1104/pp.125.1.406

Lafitte HR, IsmailA, Bennett J (2004)Abiotic stress tolerance in rice forAsia:progress and the future. In ‘New directions for a diverse planet:proceedings for the 4th International Crop Science Congress, Brisbane,Australia, 26 September – 1 October 2004’. (Eds T Fischer, N Turner,J Angus, L McIntyre, M Robertson, A Borrell, D Lloyd) pp. 1–17. (TheRegional Institute Ltd: Gosford, NSW) Available at www.cropscience.org.au/icsc2004

Läuchli A, Epstein E (1990) Plant responses to saline and sodic conditions.In ‘Agricultural salinity assessment and management. ASCE ManualsandReports onEngineeringPracticeNo. 71’. (Ed.KKTanji) pp. 113–137(ASCE: New York)

Lawrence WJC, Newell J (1939) ‘Seed and potting composts.’ (Allen &Unwin: London)

Lee KS, Choi WY, Ko JC, Kim TS, Gregoria GB (2003) Salinity toleranceof japonica and indica rice (Oryza sativa L.) at the seedling stage. Planta216, 1043–1046.

Lin HX, ZhuMZ, YanoM, Gao JP, Liang ZW, SuWA (2004) QTLs for Na+

and K+ uptake of the shoots and roots controlling rice salt tolerance.Theoretical and Applied Genetics 108, 253–260. doi:10.1007/s00122-003-1421-y

Lutts S, Guerrir G (1995) Peroxidase activities of two rice cultivars differingin salinity tolerance as affected by proline and NaCl. Biologia Plantarum37, 577–586. doi:10.1007/BF02908842

Lutts S, Kinet JM, Bouharmont J (1995) Changes in plant response to NaClduring development of rice (Oryza sativaL.) varieties differing in salinityresistance. Journal of Experimental Botany 46, 1843–1852. doi:10.1093/jxb/46.12.1843

Martinez-Atienza J, JiangXY,Garciadeblas B,Mendoza I, Zhu JK, Pardo JM(2007) Conservation of the salt overly sensitive pathway in rice. PlantPhysiology 143, 1001–1012. doi:10.1104/pp.106.092635

Møller IS,GillihamM, JhaD,MayoGM,RoySJ, Cotes JC,Haseloff J, TesterM(2009)ShootNa+ exclusionand increased salinity tolerance engineeredby cell type-specific alteration of Na+ transport in Arabidopsis. The PlantCell doi:10.1105/tpc.108.064568

Moradi M, Ismail AM (2007) Responses of photosynthesis, chlorophyllfluorescence and ROS-scavenging systems to salt stress duringseedling and reproductive stages in rice. Annals of Botany 99,1161–1173. doi:10.1093/aob/mcm052

Munns R (1993) Physiological processes limiting plant growth in salinesoil: some dogmas and hypotheses.Plant, Cell &Environment 16, 15–24.doi:10.1111/j.1365-3040.1993.tb00840.x

Munns R (2002) Comparative physiology of salt and water stress. Plant,Cell & Environment 25, 239–250. doi:10.1046/j.0016-8025.2001.00808.x

Munns R (2005) Genes and salt tolerance: bringing them together. NewPhytologist 167, 645–663. doi:10.1111/j.1469-8137.2005.01487.x

Selectable traits for salt tolerance in rice Functional Plant Biology 389

Munns R, Tester M (2008) Mechanisms of salinity tolerance. AnnualReview of Plant Biology 59, 651–681. doi:10.1146/annurev.arplant.59.032607.092911

Munns R, Schachtman D, Condon A (1995) The significance of a two-phasegrowth response to salinity in wheat and barley.Functional Plant Biology22, 561–569.

Munns R, James RA, Läuchli A (2006) Approaches to increasing thesalt tolerance of wheat and other cereals. Journal of ExperimentalBotany 57, 1025–1043. doi:10.1093/jxb/erj100

MunnsR, JamesRA,XuB,AthmanA,ConnSJ, JordansC,ByrtCS,HareRA,Tyerman SD, Tester M, Plett D, GillihamM (2012)Wheat grain yield onsaline soils is improved by an ancestral Na+ transporter gene. NatureBiotechnology 30, 360–364. doi:10.1038/nbt.2120

Negrão S, Almadanim MC, Pires IS, Abreu IA, Maroco J, Courtois B,Gregorio GB, McNally KL, Oliveira MM (2013) New allelic variantsfound inkey rice salt-tolerance genes: an association.PlantBiotechnologyJournal 11, 87–100. doi:10.1111/pbi.12010

Peng S, Ismail AM (2004) Physiological basis of yield and environmentaladaptation in rice. In ‘Physiology and biochemistry integration for plantbreeding’. (Eds HT Nguyen, A Blum) pp. 83–140. (Marcel Dekker Inc.:New York)

Platten JD, Egdane JA, Ismail AM (2013) Salinity tolerance, Na+ exclusionand allele mining of HKT1;5 in Oryza sativa and O. glaberrima: manysources, many genes, one mechanism? BMC Plant Biology 13, 32.doi:10.1186/1471-2229-13-32

PlettD,SafwatG,GillihamM,Møller IS,RoyS,ShirleyN, JacobsA, JohnsonA, Tester M (2010) Improved salinity tolerance of rice through cell type-specific expression of AtHKT1;1. PLoS ONE 5(9), e12571. doi:10.1371/journal.pone.0012571

Prasad SR, Bagali PG, Hittalmani S, Shashidhar HE (2000) Molecularmapping of quantitative trait loci associated with seedling tolerance tosalt stress in rice (Oryza sativa L.). Current Science 78, 162–164.

Ren ZH, Gao JP, Li LG, Cai XL, Huang W, Chao DY, Zhu MZ, Wang ZY,Luan S, Lin HX (2005) A rice quantitative trait locus for salt toleranceencodes a sodium transporter. Nature Genetics 37, 1141–1146.doi:10.1038/ng1643

ShiHZ,LeeBH,WuSJ,ZhuJK(2002)Overexpressionof aplasmamembraneNa+/H+ antiporter gene improves salt tolerance in Arabidopsis thaliana.Nature Biotechnology 21, 81–85. doi:10.1038/nbt766

Szabolcs I (1989) ‘Salt affected soils.’ (CRC Press: Boca Raton, FL, USA)WaliaH,WilsonC,Condamine P, LiuX, Ismail AM,ZengL,Wanamaker SI,

Mandal J, Xu J, Cui X, Close TJ (2005) Comparative transcriptionalprofiling of two contrasting rice genotypes under salinity stress during thevegetative growth stage. Plant Physiology 139, 822–835. doi:10.1104/pp.105.065961

Witcombe JR, Hollington PA, Howarth CJ, Reader S, Steele KA (2008)Breeding for abiotic stresses for sustainable agriculture. PhilosophicalTransactions of theRoyalSocietyof London.SeriesB,BiologicalSciences363(1492), 703–716. doi:10.1098/rstb.2007.2179

Yeo AR, Flowers TJ (1982) Accumulation and localization of sodium ionswithin the shoots of rice (Oryza sativa) varieties differing in salinityresistance. Physiologia Plantarum 56, 343–348. doi:10.1111/j.1399-3054.1982.tb00350.x

Yeo AR, Flowers TJ (1984) Mechanisms of salinity resistance in rice andtheir role as physiological criteria in plant breeding. In ‘Salinity tolerancein plants’. (Eds RC Staples, GH Toenniessen) pp. 37–67. (Wiley: NewYork)

Yeo AR, Flowers TJ (1986) Salinity resistance in rice (Oryza sativa L.) and apyramiding approach to breeding for saline soils. Australian Journal ofPlant Physiology 13, 161–173. doi:10.1071/PP9860161

Yeo AR, Yeo ME, Flowers SA, Flowers TJ (1990) Screening of rice(Oryza sativa L.) genotypes for physiological characters contributingto salinity resistance, and their relationship to overall performance.Theoretical and Applied Genetics 79, 377–384. doi:10.1007/BF01186082

YeoAR,LeeKS, IzardP,BoursierPJ,FlowersTJ (1991)Short- and long-termeffects of salinity on leaf growth in rice (Oryza sativa L.). Journal ofExperimental Botany 42, 881–889. doi:10.1093/jxb/42.7.881

Zeng L, Lesch SM, Grieve CM (2003) Rice growth and yield respond tochanges in water depth and salinity stress. Agricultural WaterManagement 59, 67–75. doi:10.1016/S0378-3774(02)00088-4

390 Functional Plant Biology T. Ul Haq et al.

www.publish.csiro.au/journals/fpb