742J.Zhuetal.

theimage.InPhaseII,aseriesofstepsisolatesthestelefromthecortex,allowingforseparateanalysesoftheseareas.InPhaseIII,datacollectionincludesareameasure-mentofdesiredfeatures.Correctidenti cationofthesefeaturesisbasedonnumberofpixelsandvaluethreshold-ingforedgedetection.Areasofthecross-sectionandaerenchymalacunaewereobtainedbypixelcounting.Mea-surementswereinmm2,andwerecalibratedfrompixelsusinganimageofa1mmmicrometertakenatthesamemagni cationastheanalysedimages.Thepercentageaer-enchymaasareaofthecortexwasestimatedforfourdif-ferentsectionsperreplicate.

UT,USA)spreadrandomlyacrossplotsandplacedat25and50cmdepthsinthesoilloggedbyaCR23Xdatalogger(CampbellScienti cInc.)inwaterstressandwell-wateredplotswithfourreplicatesthroughoutthegrowingseason.Leafrelativewatercontent(RWC)ofthesecondfullyexpandedleafwasdeterminedmid-dayat42DAPaccord-ingtothemethodofBarr&Weatherley(1962).

Rootdepth

Rootdevelopmentwithdepthwasmonitoredwithminirhizotrons(BTC100X,BartzTechnologyCo.,SantaBarbara,CA,USA).Oneclearacrylicminirhizotronrootobservationtubewithadiameterof5.1cmwasinstalledineachplotata30°anglefromperpendiculartoadepthofabout65cm.Rootlengthdensityandrootnumberperframewereimagedfrom10to50cmdepthinthesoilwithanintervalof10cmat42DAPandanalysedbyRoot ysoftwareversion2.0(https:///projects/root y).Destructiveharvestsat56DAPwereusedtomeasurerootlengthandnumberdensitywithdepthbysoilcoring(GiddingsMachineCo.,Windsor,CO,USA)from10to50cmdepthinthesoilwithanintervalof10cm.Thediameterofsoilcoreswas5.1cm.

Shootandrootgrowth

Rootswereseparatedfromthesoilbysubmergingcylindersinacontainer lledwithwaterandrinsingtherootscare-fullyindeionizedwater.Rootswereseparatedfromevery20cmsoildepth.Totalrootlengthinvarioussoildepthswereobtainedbyscanningwithimageanalysissoftware(WinRhizoPro,RégentInstruments,Québec,QuébecCityCanada).Maximumrootlengthofeachplantwasrecordedat35DAP.Theshootsandrootsweredriedat60°Cfor72hpriortodryweightdetermination.

Fieldstudy

Plantmaterials,experimentaldesignand eldconditions

MaizeseedsofthesixIBMRILshavingcontrastingRCAunderwaterstressinmesocosms,rsonExperimentalFarmofPenn-sylvaniaStateUniversityatRockSprings,PA(40°43′N,77°56′W)in2008.Theexperimentwasarandomizedcom-pleteblockdesignwithasplit-plotarrangementoftreat-ments.Therewerefourbiologicalreplicatesforeachgenotype,andeachreplicatehad30plantsgrowninthree2.1mrows.Theshelters(10¥30m)werecoveredwithcleargreenhouseplastic lm(0.184mm)(Grif nGreenhouse&NurserySupply,Morgantown,PA,USA)andwereauto-maticallytriggeredbyrainfalltocovertheplotsandexcludenaturalprecipitationthroughouttheentiregrowingseason.Thesheltersautomaticallyopenedquicklyafterrainfalls,exposingexperimentalplotstoambientradia-tionandtemperatureconditions.Adjacentnon-shelteredcontrolplotsweredrip-irrigatedasneededthroughoutthegrowingseasontoprovideunstressedcomparisons.ThesoilwasaMurrillsiltloam( ne-loamy,mixed,semiactive,mesicTypicHapludult).SoilpHwas6.2,andsoilnutrientlevelsofN,P,K,Ca,Mg,Zn,Cu,Swereoptimumformaizeproduc-tionasdeterminedbysoiltests.

RCA,shootdryweightat oweringstage,andindividualplantyield

Rootsandshootswereharvestedat owering.RCAwassampledfromtheoldestnodalrootinthematurationzoneat56DAPwithfoursubsamplesperreplicate.TissuesectioningandRCAquanti cationwereaspreviouslydescribed.Shootdryweightwascollectedwiththreesub-samplesperreplicate(n=4).Atmaturity,individualseednumberperplantandindividualseeddryweightwererecordedtocalculateindividualyieldwiththreesubsamplesperreplicate.

Statisticalanalysis

DatawereanalysedusingtheMinitabstatisticalpackage(MinitabInc.,UniversityPark,PA,USA).Rootdepth,RCA,shootandrootdryweight,speci crootrespiration,relativewatercontent,yield,individualseeddryweight,seednumberperplant,rootlengthandnumberperframewere rstsubjectedtoanovaformaineffectsand rst-orderinterac-tionsusingagenerallinearmodelthatincludedwaterregimeandgenotypefactors.Genotypeandwaterregimewereconsidered xedeffects,andreplicateswererandom.Fisher’sLeastSigni canceDifferencewasusedformultiplecomparisonsatP 0.05levelunderposthocanova.

RESULTS

Greenhousemesocosmstudy

WaterstressincreasedRCAofthesixlinesbyanaverageof382%at35DAPinmesocosms(Fig.1).Inwaterstress,lines34,284,and364(highRCAlines)averaged5.5%RCA,

Soilvolumetricwatercontentandleafrelativewatercontent

Soilvolumetricwatercontentwasmonitoredwithamulti-plexedTDR-100system(CampbellScienti cInc.,Logan,

©2010BlackwellPublishingLtd,Plant,CellandEnvironment,33,740–749