australis Trin. ex Steudel)

Plant,CellandEnvironment(2013)36,1860–1870doi:10.1111/pce.12095

Radialtransportofsaltandwaterinrootsofthecommonreed(PhragmitesaustralisTrin.exSteudel)

MICHAELFRITZ&RUDOLFEHWALD

InstituteofBiology,Humboldt-University,Invalidenstr.42,10115Berlin,Germany

ABSTRACT

Tounderstandtherootfunctioninsalttolerance,radialsaltandwatertransportwerestudiedusingreedplantsgrowinginbrackishhabitatwaterwithanosmoticpressure(pM)of0.63MPa.RootsbathedinthismediumexudedaxylemsapwithNaClasthemajorosmolyteanddidsoevenathighersaltconcentration(pMupto1.3MPa).ExudationwasstoppedafterasmallincreaseofpM(0.26MPa)usingpoly-ethyleneglycol600asosmolyte.Theendodermisof?nelateralrootswasfoundtobethemainbarriertoradialsolutediffusiononanapoplasticpath.ApoplasticsalttransferwasprovenbyrapidreplacementofstelarNa+byLi+inaniso-molarLiClmedium.Water?uxesdidnotexertatruesolventdragonNaCl.XylemsapconcentrationsofNaClinbasalinternodesoftranspiringculmsweremorethan?vetimeshigherthaninmedialandupperones.ItwasconcludedthattheradialNaCl?uxwasmainlydiffusionthroughtheapo-plast,andradialwatertransport,becauseoftheresistanceofthecellwallmatrixtoconvectivemass?ow,wascon?nedtothesymplast.Radialsaltpermeationinrootsreducedthewaterstressexertedbythebrackishmedium.

Key-words:apoplasmicbypass;endodermis;rootre?ectioncoef?cient;salinity;solventdrag;waterrelations;xylemtransport.

Abbreviations:D*,effectivediffusioncoef?cientofasmallsoluteinthecellwallmatrix(cm2s-1);D,diffusioncoef?cient(cm2s-1);FLR,?nelateralroot;JS,solute?ux(mols-1);JV,volume?ux(cm3s-1);LP,hydraulicconductivity(cms-1Pa-1);LPP,speci?chydraulicconductivity(cm2s-1Pa-1);a,ratiobetweentheconcentrationofasoluteinthexylemsapandthatinthemedium;pM,osmoticpressureoftherootmedium(Pa);pS,osmoticpressureofthesappressedfromtissues(Pa);pX,osmoticpressureoftherootxylemsap(Pa).

INTRODUCTION

Thecommonreed(PhragmitesaustralisTrin.exSteudel)isasalt-tolerantgraminaceouswetlandspeciesshowinghighratesoftranspiration(Stocker1928;Björk1967;Matoh,Matsushita&Takahashi1988;Lissneretal.1999).Inbrackishhabitats,rootsofthereedplantareexposedtohighsaltconcentrationsandalowwaterpotential.ThereedculmsminimizethetranslocationofNa+intothetranspiringand

Correspondence:R.Ehwald.Fax+49304495453;e-mail:rudolf.ehwald@rz.hu-berlin.de1860

assimilatingyoungleavesbyunloadingthisionfromthexylemsapanditstransfertothephloemattheculmbasis(Matohetal.1988;Matsushita&Matoh1991,1992).

Rootsrepresentarelativelysmallfractionofthetotalbiomassofthereedplant(Fiala1976).Therootdensityinareedstandgrowingin?atwaterishighestatthenodesoftheshortculminternodesclosetothegroundsurfaceinthefreewaterandintheupperlayersoftheorganiclitter.Theseculm-bornroots,whichareespeciallysigni?cantfortheuptakeofwater(Chuzhova&Arbuzova1970),differfromtheusuallysparelybranchedornon-branchedthickonesarisingfromrhizomenodeswithintheanoxicsedimentbyarelativelythinandshortmaternalaxis,aswellasbythehighdensityandsmalldiameteroftheirlaterals(Hürlimann1951;Haslam1972).Theterm?nelateralroot(FLR)usedinthefollowingreferstothethinandrelativelyshort?rst-orderbranchesoftheroots,whicharethedominatingtypeoflat-eralsinthedenserootnetworkarisingfromtheculmsclosetothegroundsurface(Fritz2012).

AnanalysisoftheradialpathwaysforwaterandsaltthroughtheFLRsisessentialtounderstandtherootfunctioninwaterrelationsandsalttolerance.Theserootswiththeirlargeexposedsurfaceandtheirsmalldiametershouldbeinvolvedinthecontrolofthesaltaccumulationinthisplantwhenitisexposedtobrackishwater.Thehypothesisofasigni?cantapoplasticbypassforradialwatertransportwasmuchpropagatedbyconclusionsdrawnfromstudieswiththerootpressureprobe(reviewedinSteudle1989,2000;Steudle&Peterson1998).Thesestudiesshowedthatchangesintheosmoticpressureofthemedium(pM)generatedchangesinrootpressurethatweresigni?cantlysmallerthanthechangesinpM.Itwasconcludedthatradialrootre?ectioncoef?cients(s)aretypicallybelowunitybecauseofasigni?cantnon-selectivehydraulicbypassthroughtheapoplast.Thismodel,however,hasnotbeencon?rmed,neitherbydemonstratingatruesolventdragfortheradialtransportoftheappliedosmo-lytesintherootsstudiedwiththerootpressureprobenorbythecomparisonoftherootpressurewiththeradialdifferenceinosmoticpressure(px-pM).

Whenrootsystemsofbarley,riceormaizewereincubatedinmediacontainingdifferentosmolytes(NaCl,d-mannitolandd-ribitol),atruesolventdragontheseosmolytescouldnotbedetected(Munns1985;Tsuchiyaetal.1992;Naito,Tsuchiya&Kumano1994;Fritz&Ehwald2011).Atlowratesofwateruptake,theconcentrationoftheseosmolytesintherootxylemsapofthementionedplantsreachedaconsider-ablefraction(a)ofthatappliedtothemedium,andtherewasalimitedriseintheradialsolute?ux(JS),whentheradialvolume?ux(JV)ofwaterwasincreased.However,awas

©2013JohnWiley&SonsLtd

Transportofwaterandsaltinrootsofcommonreed

reducedatincreasingJVinamannerthatistypicalforreverseosmosiswithoutcouplingbetweensoluteandvolume?uxes.Inallmentionedstudies,afelltoavalueof0.02orlower,whenJVwasincreasedbyahighhydrostaticpressuregradientorahighrateoftranspiration.Hence,theincreaseinJSwithJVwasonlyduetotheincreaseintheradialconcentrationgradientresultingfrompermeatedilu-tioninthexylemvesselsbytheincreasedvolumeofwaterandcanthereforebetermedpseudosolventdrag(Barry&Diamond1984).Theabsenceofatruesolventdragonper-meatingsolutesinrootsmightbeduetoindependentmechanismsofmembranetransportforbothwaterandtheappliedosmolytes.AlternativelyonecanassumethatNaClandotherosmolytespresentinthemediuminahighcon-centrationcandiffuseatasubstantialratethroughtheradialwallsoftheendodermis,thuscircumventingthesymplast(Frensch,Stelzer&Steudle1992;Tsuchiyaetal.1992;Naitoetal.1994;Ochiai&Matoh2002),whiletheradialvolume?uxisalmostcompletelycon?nedtotheprotoplasmicpathway(fromprotoplasttoprotoplastthroughtheplasmamembranesand/ortheplasmodesmata)assuggestedbyMichaeletal.(1997),Fritzetal.(2010)andFritz&Ehwald(2011,2012).

Inthisstudy,thestructureoftheculm-bornrootsofthecommonreedgrowinginbrackishwaterwasinvestigatedfocusingonthesuberizedandligni?edtissuesintheFLRs.Thebarrierfunctionoftheapoplastforsaltandwater?uxeswasanalysedwithrespecttoreverseosmosisandtheosmoticeffectsofNaClandlesspermeablenon-electrolytesontheradialvolume?ux.Itwasfurtherexaminedwhethertheapo-plasticpathwassigni?cantforradialsaltpermeationatthesaltconcentrationofthenaturalhabitat.Finally,?eldexperi-mentswerecarriedouttoestimatethephysiologicalsigni?-canceofradialsaltpermeationinsitu.

(b)

1861

(c)(d)

(e)

(f)(g)

Figure1.Culmbornadventitiousrootsandtheir?nelateral

MATERIALSANDMETHODSReedstand

ThestudiedreedstandislocatedneartheharbourofKloster,islandHiddensee,attheVitterBodden,abayoftheBalticSea,13°06′44″E,54°35′06″N.

Samplingoftheinvestigatedculm-bornroots

Thematernalaxisoftheculm-bornrootsstudied(Fig.1a)hadadiameterof1–2mmandalengthof80–150mm(Fritz2012).Therootsremainedattachedtotheculmwhentheywerecarriedinopen5Lcontainerswiththewateroftheirhabitattothenearbylaboratory(BiologicalStationKloster,UniversityGreifswald).Thetransportationtimedidnotexceed10min.

roots(FLRs).(a)Rootsdevelopedinthefreebrackishwater.(b)Freshhandsectionofamaternalroot.ThesteleoftheFLRandtheconnectiontothematernalstelearestainedwithEvansBluesuckedinfromthedecapitatedFLRintothematernalstele.Theslicewasstainedforlignin/suberin(red).(c,d)Freshhandsectionsofan18mmlongFLRc.9mmbehindthetip(c)andclosetothematernalroot(d).Thesliceswerestainedfor

lignin/suberin(red)andcellulose(blue).Notethestainingoftheendodermalandcorticalwalls,andthereductionofthecorticalintercellularspacesinthebasalsection.(e–f)Theendodermisasbarriertoacidfuchsin.Afterpressure-infusionofanacidfuchsinsolutionintothesteleofthematernalroot(20min),staininginthebasalandmedialpartoftheFLRswasrestrictedtothestele,whereasthedyeenteredthecortexinaregionbehindtheapex.(g):BathingoftheFLRintheacidfuchsinsolutionfor20minresultedinstainingofthecorticalparenchymawalls(medialsection).Scalebars:1cm(a);250mm(b);50mm(c,d);200mm(e);1mm(f);and25mm(g).

Collectionofthexylemsapexudedbytheisolatedrootswithoutpressureapplication

Thesteleofthematernalrootwasfreedfromthecorticaltissueonalengthofabout5mmandconnectedwithsilicone

tubingtothepolyethylenetipofanautomaticpipette.EvaporationofthecollectedsapwaspreventedbycoveringtheopenendofthetipwithPara?lm“M”;LaboratoryFilmBemisFlexiblePackaging,Neenah,WI,USA),whichwaspuncturedtopreventanincreaseingaspressure.

©2013JohnWiley&SonsLtd,Plant,CellandEnvironment,36,1860–1870

1862M.Fritz&R.Ehwald

athermostaticallycontrolledwaterbathat20°C.Themater-nalsteleofaculm-bornrootwasconnectedtoaSLG1430?owmeter(SensirionAG,Staefa,Switzerland)andthexylemsap?owraterecordedasdescribedpreviously(Fritzetal.2010).ToincreasepMinastepwisemanner,weighedamountsoftherespectiveosmolyteweredissolvedinasmallpartoftherootmediumandthesolutionwasaddedintotheaera-tionstream.Solutionsforisotonicmediumexchangewerepreparedbymixingthebrackishwaterina1:1ratiowithisotonicsolutionsofpolyethyleneglycol(PEG)600orglyc-erolmadewithdeionizedwater.Themediumexchangewasrealizedbytransferoftherootfromonemediumtotheother.Duringthetransferperiod(c.5s)recordingwaspaused.

Collectionofthexylemsapexudedbyisolatedrootswithapplicationofaradialpressuredifference

Rootmediawere?lledintoaplasticvesselsituatedinaScholanderchamber(SkyInstruments,Powys,UK).Thesteleofthematernalrootwasfreedforabout5mmfromthecorticalparenchyma.Anair-tightconnectionofthestelewithsiliconetubingwasobtainedbythehelpofelasticcyanoacr-ylateglue(Loctite4860;HenkelAG,Düsseldorf,Germany).Thec.1cmlongsiliconetubingsegmentwasconnectedtoPEEKtubing(UpchurchScienti?c,OakHabor,WA,USA)andthelatterwasconductedthroughtherubbersealoftheScholanderchamber.Topreventthecontaminationofthexylemsapwiththerootmediumandtoavoidliquidin?ltra-tionofthecorticalaerenchymaofthematernalroot,anyliquidwasdabbedawaywithcellulosetissuefromthecortexbelowtheseal.Atinsertionoftherootintothemediumthesealwaskeptc.1cmabovetheliquidleveltoconnecttheaerenchymaofthematernalaxispneumaticallywiththeairspaceinthechamber.Thiswaythegaspressureintheaer-enchymawaskeptequaltothepressureontheliquidmediumandin?ltrationoftherootcortexatincreasingpres-surewasprevented(Michael,Cholodova&Ehwald1999).Whenthechamberpressurewasstepwiseincreased,sam-plingofxylemsapwasstarted15minafteradjustingtherespectivepressureandlasted5–20min,dependingonthe?owrate.TodiscardresultsobtainedfromrootswithmechanicallyinjuredFLRs,EvansBluewasaddedtothemediuminaconcentrationof5gL-1aftertheexperiment,anditwascheckedataradialpressuredifferenceof0.5MPathatthisdyewasexcludedfromtheexudedsap.Thevolume?uxwasdeterminedbyweighingtheexudate.Thecollectedsapsampleswerestoredbeforechemicalanalysisat-20°C.

Collectionofxylemsapfromtheculminternodes

Combinedsamplesof10culmswereprepared.Eachculmwaschoppedoffimmediatelybeforethesapcollectionpro-cedure.Theleaveswereremovedandthestemcutoffatthegroundlevel.Thebrackishwateradheringtothebasalculmregionwaswashedoffwithdeionizedwaterandthesurfacewasblottedwith?lterpaper,beforeintactinternodeswiththeadjacentupperandlowernodeswerecutout.Thepithcavityoftheselectedinternodes(basal,medialandupper)wasopenedbyasectionc.1cmbelowtheuppernode.Theupperendoftheinternodewas?ttedinto10mmlongelasticsiliconetubingandthepithcavitywasclosedwithacone-shapedplug.Thepluggedendwas?ttedintoaconicaladapterandthexylemsapblownoutfromthexylemvesselsintoasmallvial.Gatheringthesamplesfromthreedifferentinternodes(basal,medial,andupper)requiredabout3min.Thesapvolumeyieldedfrom10internodeswasc.1mLfortherelativelylongmedialandupperinternodes(c.1or1.4maboveground)andc.0.4mLfortheshortbasalinternodes(aboveground).Thexylemsapwasstoredat-20°Cbeforeanalysis.

Pressureinfusionofanacidfuchsinsolutionintotherootxylem

Thecortexwasremovedfromthesteleatthebasal5mmlongzoneofanisolatedculm-bornrootandtransparentsiliconetubingwastightlyconnectedtothestelebythehelpofthecyanoacrylateglueasdescribedearlier.Therootwiththetubingwasinsertedinavesselwiththehabitatwater.Avesselwiththedyesolution(4gL-1)wasinsertedintheScholanderchamberandthesolutiondriventhroughaPEEKtubingoutofthechamber.ThePEEKtubingwiththedyesolutionwasconnectedtothetransparenttubingatthestelecontainingthexylemsap.Finally,thepressureonthedyesolutionwasincreased,untilthedyesolutionmovedintothestele.Thisrequiredapressureofabout0.3MPa.Afteraninfusionperiodof20minatthispressuretheFLRswerephotographed.

Preparationofpressedsapfromtheleaves,rhizomesandroots

Theleafsampleswerepreparedfromthebladesofthetwoyoungestalreadyunfoldedleaves.Pooledsamplesofleavescomprisingmaterialfromatleast?veculmswereenclosedinpolypropylenetubesinthe?eld.Rhizomeswerewashedinthe?eldwithdeionizedwaterandblottedwithpulpbeforepiecesoftherhizomewallwereenclosedinthesamplingtubes.Sampleswerefrozenat-20°C.Sapwaspressedfromthethawedsamplesinanaluminium-cylinderpressbymeansofabenchvice.Pressedsapofrootswaspreparedinaspecialwaytoremovethesaltofthehabitatwaterandtoenableacorrectionforthevolumeofthesurface?lm.Rootedbasalculmsegmentswerecollectedinthe?eldandtransferredtothelaboratorytogetherwiththehabitatwaterasdescribedearlier.Thebrackishmediumwasremovedfromthefreespaceofrootsamples(1.5–2gfreshweight)byshortblottingon?lterpaperanda5minwashin500mLofasolution

Recordingofxylemsapexudationbytherootswithhightimeresolution

Rootexudationwasrecordedwithrootsincubatedin500mLoftheaeratedmedium.Thevesselwiththerootswaskeptin

©2013JohnWiley&SonsLtd,Plant,CellandEnvironment,36,1860–1870

Transportofwaterandsaltinrootsofcommonreed

ofPEG(10000,c.10kDa,concentration1mM)madewithdeionizedwater.Thewashedrootswereblottedonpulpagain,weighedandfrozen.Pressedsapwaspreparedfromthethawedtissueandthevolumefractionof?lmwaterinthesapwasdeterminedfromthePEGconcentrationinthesapasdescribedpreviously(Fritz&Ehwald2011).

1863

Osmoticpressureandchemicalanalysis

Theosmoticpressurewasdeterminedusingafreezingpointosmometer(Dr.KnauerGmbH&Co.KG,Berlin,Germany).Theconcentrationsofthecationsweredeter-minedafterappropriatedilutionbymeansofa?amepho-tometer(JenwayLtd.,Essex,UK).Chloridewasdeterminedbytitrationofsmallde?nedvolumesofthebrackishwater,pressedsaporcollectedxylemsapwith0.1MAgNO3usingamicro-syringe(50or100mL)using200mLof40mMpotas-siumdichromateasindicator(Mohr’smethod,compareSträhle&Schweda2005).ThePEGconcentrationwasdeter-minedbythemethodofChilds(1975).

of9.3?2.6mmandameandiameterof170?25mm(500FLRs,20culm-bornroots).Lessthan1%ofthelateralswerethicker(200–300mm).ThebasesoftheFLRswereburiedwithintheradialseptaoftheaerenchymainthematernalroot(Fig.1b).WhenanEvansBluesolutionwassuckedthroughadecapitatedFLRintothemainaxis,stainingwascon?nedtothestelarconnection(Fig.1b).Thetwoorthreeligni?edxylemvesselsintheFLRsweresituatedintheperipheryofthestele.Inspiteoftheirsmalldiameter(10.9?2.5mm)thevesselsweretracheaasprovenbyaxialmovementofinkparticleswithadiameterof0.5–1mmalongtheirwholelength(notshown).Notfarfromthetip,wherethewallsoftheendodermiswerenotyetthickened,theradialandtangentialwallsoftheepidermis/hypodermisalreadycontainedligninand/orsuberin(Fig.1c).AtthebaseoftheFLRsclosetothematernalrootalsothewallsofthecorticalparenchymacellswerethickenedandligni?ed/suberized.Heretheintercellularspaceswereabsentorextremelysmall(Fig.1d).

Radialpermeationofacidfuchsin

Microscopy

Freshsectionswerestainedforcellulose(Astralblue),lignin(Neufuchsin,red)andsuberin(Chrysoidin,yellow/orange)inanacidi?edaqueoussolutionaccordingtoKremer(2002).

Whenappliedtotherootmediumacidfuchsindiffusedintothecortexalongthewholerootlength(Fig.1g).Radialdif-fusionofthedyefromthemediumintotherootapoplastwasmostrapidintheapicalandsubapicalzones(Fritz2012).Inthexylem-infusionexperiments,theacidfuchsinsolutionmovedintothesteleofmostFLRsatapressureofabout0.3MPa.InthebasalandmedialpartoftheFLRsstainingwascon?nedtothestele,whileinasub-apicalzonecompris-ingthesecondandthirdmillimetrealsothecortexbecamestained.Thestainingdidnotreachthenon-fasciculartissueoftherootapex(Fig.1e,f).

RadialdiffusionofacidfuchsinfromthemediumintothecortexoftheFLRs

Intactrootswereimmersedfor20mininasolutionofacidfuchsin(4gL-1)madeupwiththebrackishhabitatwater.Therootswerebrie?ywashedinthehabitatwater,beforehandsectionsoftheFLRswerepreparedandphotographsweretaken.

Effectsofexternalosmolytesonthexylemsap?owintheabsenceofaradialhydrostaticforce

Whenisolatedculm-bornrootswereincubatedinthehabitatwaterwithanosmoticpressure(pM)of0.63MPa,theyexudedxylemsapwithanosmoticpressure(pX)of0.68–0.83MPa(Table1).TheconcentrationsofNa+andCl-intheexudedsapreachedmorethantwo-thirdsoftheexternalconcentration,whereastheconcentrationofK+inthexylemsapwasseveraltimeshigherthaninthemedium.AscanbeconcludedfromthesecondandthirdexperimentsshowninTable1,therewasadeclineintheexudationratewithinatimespanofseveralhours,wherebytheK+/Na+ratiooftheexudedsapdecreasedanditsCl-concentrationincreased.Thesap?owrecordedwiththe?owsensorunderstepwiseadditionofNaClorPEG600tothebrackishhabitatwatershowedatransientreversalofthe?owdirectionimmediatelyaftertheadditionofeitherosmolyte(Fig.2a,b).Out?http://wendang.chazidian.comingPEG600,anincreaseinpMbyc.0.23MPawassuf?cienttopreventxylemsapout?owpermanently.AmuchstrongerincreaseinpM(c.0.83MPa)wasrequiredtopreventsapout?owperma-nentlybyadditionofNaCl(Fig.2c).

Visualizationofthexylemsap?owfroma?nelateraltothematernalroot

Awholeculmbornrootwasconnectedwiththematernalsteletosilicontubingasdescribedearlier.

AfewoftheFLRsattachedtothematernalrootweredecapitated,therootwasimmersedinanEvansBluesolu-tion(10gL-1)orinaninkparticlesuspension(Fritz&Ehwald2012)bothmadeupwiththegrowthmedium,andthetubingconnectedtoavacuumpump,beforephotographsofhandslicesweretaken.

RESULTS

Structureoftheculm-bornroots

Denselybranchedadventitiousrootswithadiameterofthemainaxisof1–2mmwerepresentingreatquantitiesbetweentheshortbasalinternodesoftheculminthefreewaterandintheoxygen-containingsurfacelayerofthesedi-ment.Thesurfaceareaoftheculm-bornrootswasdominatedbyFLRsofthe?rstorder(Fig.1a).Theyhadameanlength

©2013JohnWiley&SonsLtd,Plant,CellandEnvironment,36,1860–1870

1864M.Fritz&R.Ehwald

Table1.Osmoticpressure(pX)andionconcentrations(C)inthexylemsapexudedbytheculmbornrootsinthebrackishhabitatwaterwithoutpressureapplication

C(mM)

Exudationrate(mLh-1)

June,2006(n=20)0–4hafterexcisionMay,2007(n=24)0–4hafterexcision4–16hafterexcisionJune,2009(n=9)

0–1.5hafterexcision1.5–16.5hafterexcision

3?29?53?125?105?2

pX(MPa)0.68?0.010.80?0.060.83?0.120.69?0.060.69?0.17

Na+84?3109?4105?1092?13113?2

K+33?328?511?213?815?3

Cl-73?1071?8115?991?11115?5

n,numberofadventitiousroots.Meanvalueswithstandarddeviation.Habitatwatercontained120mMNa+,125mMCl-and2.5mMK+,pMwas0.63MPa.

WhenanexudingrootwastransferredfromthebrackishhabitatwatertoamediumobtainedbymixingthiswaterwithanisotonicsolutionofglycerolorPEG600intheratio1:1(Fig.3),thesapout?owshowedastrongtransientincreaseanddeclinedrapidlytoavaluethatwaslowerthaninthepurehabitatwater(glycerol)orclosetozero(PEG600).Thesechangeswerereversible.IntheglycerolmediumthetransientstimulationdisappearedinashortertimethaninthePEGmedium.

theirconcentrationinthehabitatwater(Table2).ThemeanvalueforpXinthebasalinternodesreachedabout40%ofpM.InthemedialandupperinternodespXwasmarkedlyreducedbecauseofastrongdecreaseintheconcentrationsofNa+andCl-.Themeanconcentrationofbothionsinthexylemsapofthebasalinternodeswas?vetosixtimeshigherthaninthemedialonesandseventoninetimeshigherthanintheupperones.

Dilutionofsodiumandchlorideionsinthexylemsapbyincreasingtheradialvolume?uxofwater

TheratiobetweentheconcentrationofNa+andCl-inthexylemsapandthatofthemedium(a)decreasedwhentheradialvolume?uxofwater(JV)wasincreasedbyraisingtheradialhydrostaticpressuredifferenceacrossthecortexofexudingculm-bornroots.IncreasingJVfrombelow1mLmin-1perroottoabout3–4mLmin-1perrootresultedinasteepdecreaseinaforbothNa+andCl-.AtevenlargervaluesofJVobtainedatpressuresupto1MPa,awasreducedtovaluesbelow0.01(Fig.4).

Ionconcentrationsandosmoticpressure(pS)ofsappressedfromdifferentorgansofthereedplant

PressedsappreparedfromtherootscontainedNa+andCl-inconcentrationsthatexceededtheconcentrationofK+andexplainedmorethanhalfofpS,whichwashigherthanpMofthebrackishhabitatwaterbymorethan0.3MPa.Inpressedsapspreparedfromrhizomesandyoungleavestheconcen-trationofCl-wasmuchhigherthanthatofNa+,butbelowthesumoftheK+andNa+concentrations.TheyoungleavesandrhizomebudsshowedaveryhighK+/Na+ratioandarela-tivelylowCl-concentration(Table3).

IonconcentrationintheexudedxylemsapinNaClandLiClmedia

Aftertransferringarootfrom120mMNaCltoamediumcontaining120mMLiCl(Fig.5),theconcentrationofNa+intheexudedxylemsapwasdecliningfromc.100mMto2.5mM,whereasatthesametime,Li+reachedaconcentra-tionofc.60mM.AftertransferoftherootbacktotheNaClmedium,Li+inthexylemsapwasreplacedbyNa+withasimilartimecourse.AtanytimetheCl-concentrationinthexylemsapwasapproximatelyequaltothesumoftheLi+andNa+concentrations.

DISCUSSION

Theapoplasticbypassforradialsolute

transportintheFLRsanditssigni?canceforradialNaCltransfer

InthebasalandmedialzoneoftheFLRsasigni?cantdiffu-sionofacidfuchsinfromthemediumintothecortexwasfound(Fig.1g),buttherewasnoleakageofthedyefromthesteleintothecortexinthisregion,whenthexylemwaspressure-infusedwithanacidfuchsinsolution(Fig.1e,f).ThisshowedthatthesizeexclusionlimitofradialwallsinthefullydevelopedendodermiswasbelowtheStokes’diameteroftheacidfuchsinmolecule(about2nm),whichisahydrophilicmembrane-impermeabletrivalentanion.Clearly,thematureendodermiswasthemaindiffusionbarrierinthecorticalapoplastoftheFLRs(compareDeRufzdeLavison1910,1911;Alassimoneetal.2012),althoughtheepidermisandhypodermisshowedamarkedligni?cationorsuberization

Ioncompositionandosmoticpressureofthexylemsapinbasal,medialandupperculminternodes

InthexylemsapofthebasalinternodesoftranspiringculmsconcentrationsofNa+andCl-werebetween20and30%of

©2013JohnWiley&SonsLtd,Plant,CellandEnvironment,36,1860–1870