<%BANNER%>

UFIR



Experimental test for facilitation of seedling recruitment by the dominant bunchgrass in a fire-maintained savanna
http://www.plosone.org ( Publisher's URL )
CITATION PDF VIEWER
Full Citation
STANDARD VIEW MARC VIEW
Permanent Link: http://ufdc.ufl.edu/IR00001328/00001
 Material Information
Title: Experimental test for facilitation of seedling recruitment by the dominant bunchgrass in a fire-maintained savanna
Series Title: Iacona, G. D., L. K. Kirkman, and E. M. Bruna. 2012. Experimental test for facilitation of seedling recruitment by the dominant bunchgrass in a fire-maintained savanna. PLoS ONE 7(7): e39108
Physical Description: Journal Article
Creator: Bruna, Emilio
Iacona, Gwenllian D.
Kirkman, L. Catherine
Publisher: Public Library of Science
Place of Publication: PLoS ONE
Publication Date: July 6, 2012
 Notes
Abstract: Facilitative interactions between neighboring plants can influence community composition, especially in locations where environmental stress is a factor limiting competitive effects. The longleaf pine savanna of the southeastern United States is a threatened and diverse system where seedling recruitment success and understory species richness levels are regulated by the availability of moist microsites. We hypothesized that the dominant bunch grass species (Aristida stricta Michx.) would facilitate moist seedling microsites through shading, but that the effect would depend on stress gradients. Here, we examined the environmental properties modified by the presence of wiregrass and tested the importance of increased shade as a potential facilitative mechanism promoting seedling recruitment across spatial and temporal stress gradients. We showed that environmental gradients, season, and experimental water manipulation influence seedling success. Environmental properties were modified by wiregrass proximity in a manner that could facilitate seedling success, but we showed that shade alone does not provide a facilitative benefit to seedlings in this system.
Acquisition: Collected for University of Florida's Institutional Repository by the UFIR Self-Submittal tool. Submitted by Emilio Bruna.
Publication Status: Published
Funding: Publication of this article was funded in part by the University of Florida Open-Access Publishing Fund.
 Record Information
Source Institution: University of Florida Institutional Repository
Holding Location: University of Florida
Rights Management:
This item is licensed with the Creative Commons Attribution License. This license lets others distribute, remix, tweak, and build upon this work, even commercially, as long as they credit the author for the original creation.
System ID: IR00001328:00001

Downloads

This item is only available as the following downloads:

Iacona_etal_2012_PLoSONE ( PDF )


Full Text

PAGE 1

ExperimentalTestforFacilitationofSeedling RecruitmentbytheDominantBunchgrassinaFireMaintainedSavannaGwenllianD.Iacona1,2*¤,L.KatherineKirkman1,EmilioM.Bruna2,31 J.W.JonesEcologicalResearchCenter,Newton,Georgia,UnitedStatesofAmerica, 2 DepartmentofWildlifeEcologyandConservation,UniversityofFlorida,Gainesville, Florida,UnitedStatesofAmerica, 3 CenterforLatinAmericanStudies,UniversityofFlorida,Gainesville,Florida,UnitedStatesofAmericaAbstractFacilitativeinteractionsbetweenneighboringplantscaninfluencecommunitycomposition,especiallyinlocationswhere environmentalstressisafactorlimitingcompetitiveeffects.ThelongleafpinesavannaofthesoutheasternUnitedStatesis athreatenedanddiversesystemwhereseedlingrecruitmentsuccessandunderstoryspeciesrichnesslevelsareregulated bytheavailabilityofmoistmicrosites.Wehypothesizedthatthedominantbunchgrassspecies( Aristidastricta Michx.) wouldfacilitatemoistseedlingmicrositesthroughshading,butthattheeffectwoulddependonstressgradients.Here,we examinedtheenvironmentalpropertiesmodifiedbythepresenceofwiregrassandtestedtheimportanceofincreased shadeasapotentialfacilitativemechanismpromotingseedlingrecruitmentacrossspatialandtemporalstressgradients.We showedthatenvironmentalgradients,season,andexperimentalwatermanipulationinfluenceseedlingsuccess. Environmentalpropertiesweremodifiedbywiregrassproximityinamannerthatcouldfacilitateseedlingsuccess,but weshowedthatshadealonedoesnotprovideafacilitativebenefittoseedlingsinthissystem.Citation: IaconaGD,KirkmanLK,BrunaEM(2012)ExperimentalTestforFacilitationofSeedlingRecruitmentbytheDominantBunchgrassinaFire-Maintained Savanna.PLoSONE7(7):e39108.doi:10.1371/journal.pone.0039108 Editor: KurtO.Reinhart,USDA-ARS,UnitedStatesofAmerica Received October13,2011; Accepted May17,2012; Published July6,2012 Copyright: 2012Iaconaetal.Thisisanopen-accessarticledistributedunderthetermsoftheCreativeCommonsAttributionLicense,whichpermits unrestricteduse,distribution,andreproductioninanymedium,providedtheoriginalauthorandsourcearecredited. Funding: FundingwasprovidedbytheJ.W.JonesEcologicalResearchCenter,TheRobertWoodruffFoundation,andTheUniversityofFloridaCollegeof AgriculturalandLifeSciences.Thefundershadnoroleinstudydesign,datacollectionandanalysis,decisiontopublish,orpreparationofthemanu script. CompetingInterests: Theauthorshavedeclaredthatnocompetinginterestsexist. *E-mail:giacona@utk.edu ¤Currentaddress:DepartmentofEcologyandEvolutionaryBiology,UniversityofTennessee,Knoxville,Tennessee,UnitedStatesofAmericaIntroductionFacilitativeinteractionsoccurwhenaplantenhancesthesuccess ofanother[1].Becausetheserelationshipscaninfluencepatterns ofcommunitycomposition,identifyingtheextentoffacilitation anditsunderlyingmechanismsisimportantforunderstandingthe driversofplantdiversity[2].Facilitationinplantcommunities appearstoberelativelycommon,particularlyinstressfulsystems wherethebenefitsprovidedbyaneighboringplantoutweigh competitiveinteractions[2,3,4,5].However,thenatureofthe interactionmayvaryacrossacontinuumdeterminedbyenvironmentalstressgradients,withcompetitiveeffectsdominatingin optimalconditionsandfacilitationprevailinginharsherones [2,3,6].Thesestressgradientsmaybetemporal(e.g.inter-seasonal differences[7]),orspatial(e.g.,gradientsinsoilmoisture[4],or salinity[8]). Facilitativeinteractionscaninfluenceplantcommunitycompositionbyvariousmechanismsincludingthosethatenhance seedlingsuccess.Thenutrientuptake,transpiration,andphysical presenceofa‘‘nurseplant’’changesthelocalmicroenvironment andmayimpactnearbyseedlings[1].Theseeffectscanenhance nutrientsupply[9,10],orincreasewateravailabilityforaseedling, andcanbeparticularlyimportantearlyinestablishmentorduring drought[11,12].Inaddition,nurseplantscanfacilitaterecruitmentbymodulatingtemperaturehighsandlowsbeneath theircanopy[13]. Despitetheirputativeimportance,theroleoffacilitative interactionsinthemaintenanceofspeciesdiversityisstillunclear [2].Inspecies-richsystems,facilitativehabitatameliorationor increasedresourceavailabilitycouldexpandthefundamental nicheofaspeciesandthusexplainpatternsofcoexistencethat cannotbefullyexplainedbycompetitiveexclusion[6,14].One species-richsystemthatcouldbeusedtotestforfacilitativedrivers ofspeciesdiversityisthelongleafpine( Pinuspalustris Mill.)savanna ofthesoutheasternUnitedStates[15].Previouslywehaveshown thatwateravailabilityregulatesdiversityofseedlingspeciesinthis system,whichsuggeststhatheterogeneityintheavailabilityof moistmicrositesinfluencescommunitycomposition[16].Itis possiblethatthepresenceofsuchmicrositesisaconsequenceof plant-plantinteractions.Thereiscurrentlynoevidencethat positiveplant-plantinteractionsplayaroleinpromotingsmall scalespeciesrichnessinthissystem.However,theobserved positivecorrelationbetweenunderstoryspeciesdiversityand annualnetprimaryproductivitymayindicateabeneficialrelationshipbetweenplantbiomassandseedlingestablishment[17]. Inaddition,thelongleafpinesavannaischaracterizedbyavariety ofstressesincludinglowwaterandnutrientavailability.Thesetwo factorslikelyinteractwithfrequentfiresandhighsummer temperaturestolimitcompetitiveinfluencesonseedlingrecruitment,suggestingthattheconditionsareappropriatefor facilitationtooccur[18,19]. PLoSONE|www.plosone.org1July2012|Volume7|Issue7|e39108

PAGE 2

Wiregrass( Aristidastricta Michx.)isthedominantunderstory speciesinthesoutheasternextentofthelongleafpinesavanna [20],anditformsanearcontinuousgroundcoverintheregions wherespeciesrichnessishighest[21].Thereareseveral mechanismsbywhichthepresenceofwiregrasscouldaffect seedlingsuccess.Theprimaryfacilitativebenefitmaybeenhanced seedlingwaterretentionfromreducedevapotranspirationinthe shade[11],ashasbeendemonstratedinsomearidecosystems [22,23].Alternatively,elevatedsoilcarbonpoolsbeneathclumps ofwiregrass[24]couldenhancesoilwaterholdingcapacity[25]. Finally,thenon-rhizomatousstructureofwiregrasswithscattered clumpsandnumerousarchingwiryleavesmayprotectseedlings fromhightemperaturesandotherenvironmentalextremes[26]. Here,weassessedvariationinseedlingmicrosites,andtested shadeasamechanismtoenhanceseedlingrecruitmentacross amoisturestressgradient.Wecharacterizedtheabioticenvironmentofpotentialseedlingrecruitmentmicrositeslocatedunder andbetweenwiregrassclumps.Wealsousedseedandresource manipulationstodeterminehowshade,andtheinteractionof shadeandwateravailability,influencedseedlingrecruitment.The experimentincorporatedanaturallyoccurringgradientinsoil moistureandaseasonalgradientovertwoyears.Wepredicted thatthefacilitativeeffectsofshadingshouldbemoreimportantin xericcommunitiesandafterthesummermonthswheretemperatureandmoisturestressismoreapparent.Also,weexpectedthat addingwaterwouldreducetheeffectoffacilitation,andtheeffect ofwateradditionwouldbemoreapparentinxericlocations (observableasaninteraction).Methods StudySiteIchauway(31 u 13 9 N,84 u 29 9 W)isan11,300hapropertyof theJosephW.JonesEcologicalResearchCenterthatincludes 7,500haoflongleaf-pinesavanna.Awiregrass-dominated groundcoverthatischaracterizedbynotablyhighlevelsoffine scalespeciesrichness[27]hasbeenmanagedundera2–5year prescribedfirereturnintervalforatleastthepast90years.This siteislocatedintheLowerCoastalPlainandFlatwoods(LCPF) sectionofsouthwestGeorgia,USA[28].Averageyearly temperaturesrangefrom5–17 u Cinthewinterto21–34 u Cin thesummerwithabout131cmofprecipitationevenlydistributed throughouttheyear[29].Acrossitsrange,thelongleafpinewiregrasssavannaoccupiesawidemoisturegradientthatextends fromextremelymesiclocationswithsaturatedsoilsandaperched watertabletoextremelyxericlocationsalongdeepsandridges.At Ichauway,thewiregrassdominatedunderstoryextendsfrommesic areaswithseasonallysaturatedAquicArenicPaleudultsoilsto xericareaswithTypicQuartzipsamments[29].MicrohabitatPropertiesTocompareenvironmentalconditionsinpotentialseedling recruitmentmicrositesthatdifferedintheirproximitytowiregrass, wemeasuredsoilandairtemperature,soilmoisture,relative humidity,andlightavailabilitybeneathandadjacenttonaturally occurringwiregrassclumpsfromApriltoAugust,2006.Ineach controlplot(asdescribedbelow),werandomlyselectedasite underawiregrassclump(hereafter,‘‘below’’)andidentifiedthe nearestbaresoilpatchwithnowiregrasscanopy(hereafter, ‘‘between’’).Ateachofthesemicrosites(n=16),weestimated volumetricsoilmoistureat10cmdepthusingtime-domain reflectometry(TDR)[30].Forthisprocedure,weinstalledapair of11cmstainlesssteelrodsverticallyateachmicrositetoquantify soilmoisturethreedayspermonthusingaTechtronixcabletester. Wemeasuredphotosyntheticallyactiveradiation(PAR m mol.m2 2.s2 1)atgroundlevelateachmicrositewithaquantum linesensorceptometer(AccuparLP80,DecagonDevices,Pullman WA).Readingsabovethewiregrasscanopywereobtained simultaneouslyat50cmabovethegroundusinganexternal PARspotsensor.Foreachobservation,werecordedthree consecutivePARreadingsandallobservationsweremade betweenthehoursof10:00AMand2:00PMonsunnydays. Werecordedmicrositeairtemperatureandrelativehumidityat 5cmabovegroundlevelwithmultimodedataloggers(HOBOpro H8,OnsetComputerCorporations,BourneMA).Wealso measuredsoiltemperatureatadepthof5cmusingonefour channeldatalogger(HOBOH8,OnsetComputerCorporations, BourneMA)perplotwithtwotemperatureprobesonleadssothat measurementscouldbeobtainedatbothmicrositelocations simultaneously.Alldataloggerstookreadingsatonehourintervals for72consecutivehourseachmonth.ResourceTreatmentsOurexperimentstookadvantageofanongoingandlong-term experimentalresourcemanipulationestablishedatIchauwayin 2002.Wemanipulatedwaterandshadeatboththemesicand xericendsofthenaturalsoilmoisturegradient(meanpercent baselinevolumetricsoilmoisture8 6 0.68inxeric,and15 6 1.04in mesic).Wehadeightrandomlyassignedwateradditionplots (50m 6 50m)andeightrandomlyassignedcontrolplotssplit betweenthesoilmoistureextremesinafullyfactorial2 6 2 experimentaldesign.Irrigationwithapproximately825mmof waterperyear( 65%increaseoveryearlyprecipitation)was ongoingsince2002andmaintainedthewatertreatmentplotsat closeto40%fieldcapacity.Thislevelofwateradditionwas determinedbyengineeringlimitations,butitwasasubstantial increaseforthesystemandthuswaslikelytoremedythewater limitationsthatregulateseedlingrecruitment(R.Mitchell, personal communication )[17].Irrigationwasappliedfor24hoursonceevery eightdaysunlesssufficientrainfelltoreachthetargetfield capacity.Tominimizecationaccumulationinthesoil,waterwas treatedwithreverseosmosis. Tosimulatetheeffectsofshadingbywiregrass,andtoexamine thepotentialimportanceofinteractionsofshadeandwater availabilityonseedlingrecruitmentrates,weconstructedshade clothstructuresoverexperimentallysownseedlingcommunities replicatedwithintheplots(n=2perplot).Weusedasplit-split plotdesignthatmanipulatedshadelevelsincontrolandwatered plotsatthewetanddryextremesofthenaturallyoccurring moisturegradientfortwogrowingseasons. In2003,weinstalled4polyvinylchloride(PVC)rings(30cm diameter 6 10cmdeep)ineachofthe16plots.Theseringsservedas seedlingrecruitmentmanipulationsitesandallowedustoensurethat observedseedlingswererecruitsfromseedratherthanvegetative resprouts.ToinstallthePVC,wefirsthammereda30cmsteelring intothesoiltosevertheplantrootsthenextractedthisringwithaslittle soildisturbanceaspossible.WeinsertedthePVCringsintotheprecuttrenchsothatthetopedgeoftheringwasflushwiththeground.In 2005,weremovedtheexistingvegetationineachringwithtwo treatmentsofglyphosate,appliedinsummerandfall.Allsiteswere burnedinFebruary2004and2006. InMarch2006,weinstalledshadestructuresovertwo randomlyselectedringsineachplotandlefttworingsunshaded (n=64rings).Forthefirstyearoftheexperiment,multiplelayers ofshadeclothandscreenwerefittedtowoodenframes (55cm 6 55cm 6 30cmtall)andreducedlightatthesoilsurface by90%.Thislevelofshadewaschosentoapproximatethelight reductionprovidedbywiregrasscover(mean 6 1SD=78 6 1%,FacilitationofSeedlingsbyaBunchgrass PLoSONE|www.plosone.org2July2012|Volume7|Issue7|e39108

PAGE 3

max99%;R.Mitchell, unpubl.data ).Duetolowoverall recruitmentin2006,weadjustedtheshadingto80%forthe secondstudyyear.BetweenMarchandMay2006,andagainin March2007,wesowed50seedsperringofeachofthefollowing species: Sorghastrumsecundum Rudbekiahirta and Desmodiumciliare .We chosethesespeciesbecausetheyarecommongrass,forb,and legumespeciesinthelongleafpineunderstory,andareknownto germinateunderfieldconditions.Allseedswereobtainedfromthe nativeplantresearchgardenatIchauway.Wedeterminedthe percentviabilityofeachspeciesusingpetridishgerminationor tetrazoliumtestsin2007( R.hirta :75%viable, S.secundum :40% viable, D.ciliare :90%viable).Toreducelossofseedsduetowind, wefittedan8cmtallringofaluminumwindowscreenaround eachunshadedPVCring.Wecountedallseedlingsofeachspecies inMayandNovember2006and2007.StatisticalAnalyses:MicrohabitatPropertiesWetestedfordifferencesinenvironmentalparametersat seedlingmicrositesthatdifferedwithinproximitytowiregrass clumpsusingamixedmodelprocedure(PROCMIXED,SAS InstituteInc.Version9.1).Themeanvalueperplotpermonth wasanalyzedforeachmonth,withplottreatedasarandomeffect, andproximitytowiregrass(beloworbetween)consideredafixed effect.Toobservefinescalevariation,weanalyzedmesicandxeric locationsbymonthtominimizethevariabilitybetweenlocations andovertime.Statisticalsignificancewasestablishedat a =0.05.StatisticalAnalyses:ResourceTreatmentsWeusedamixedmodelanalysistotestformeandifferencesin recruitmentrateattributabletoshade,wateravailability,collection period,moisturegradientlocationandforinteractionsbetween shade,wateravailabilityandmoisturegradientlocation.Dueto lownumbersofrecruits,wegroupedseedlingsofallthreespecies togetherforanalysis.Preliminaryanalysisdidnotshowany variabilityinrecruitmentresponseacrossspecies(May2006, treatment 6 speciesinteractioneffecttest;F=0.07,Pr F=0.93). Wecomparedthemeannumberofseedlingsinshadedand unshadedtreatments(tworingsofeachperplot)acrosswatered andcontrolplotsatmesicandxericgradientlocationsoverfour collectionperiodsusingasplit-splitplotdesign(n=32,PROC MIXED,SASInstituteInc.Version9.1).Shadetreatment(splitsplitplotfactor),watermanipulation(whole-plotfactor),gradient location(whole-plotfactor)andcollectionperiod(split-plotfactor) weremodeledasfixedeffects,andplotwasmodeledasarandom effect.Weincludedcollectionperiodasasplit-plotfactorrather thanusingarepeatedmeasuresframeworksincethenumberof seedlingswasassumedtoincrease,butwewereinterestedinthe interactionbetweentimeandresourceavailability.Weused plannedcontrastsofcombinationsofcollectionperiodtoexamine seedlingresponsetoseasonandyear.Themodelwasfitusing aGaussiandistributionbecausetheuseofmeanresponsevalues allowedfortheassumptionofanormaldistribution.Inaddition, examinationofthemodelresidualsindicatedadequatemodelfit. Statisticalsignificancewasassessedat a =0.05.Results MicrohabitatPropertiesThemagnitudeofdifferenceinenvironmentalfactors‘‘below’’ versus‘‘between’’wiregrassclumpsrangedfrom30%variationin PARtoalmostnovariationinvolumetricsoilmoisture(Table1). Shading(reductioninPAR)wasconsistentlygreaterunderwiregrass clumps(Figure1d).Meanwhile,differencesbetweenmicrositeswith respecttorelativehumidity,soiltemperature,andairtemperature valueswereusuallysmallanddependentongradientlocationand month.Volumetricsoilmoisturedidnotvarywithmicrosite proximity,season,orgradientlocation(Figure1e). Thevaluesofallfactorsvariedacrossmonths,buttherewasno cleartemporalpatternindifferenceformicrositesbetweenversus belowwiregrassclumps.Meanwhile,differencesbetweenmicrosite typesforsomefactorswererelatedtogradientlocation.Inthe xericlocation,percentrelativehumiditywassignificantlyor marginally(p # 0.08)greaterbelowwiregrassclumpsforthreeof thefivemonths.Nodifferenceoccurredwithproximityto wiregrassinthemesiclocation(Figure1a).Incontrast,the averagesoiltemperaturewasgreater(ormarginallygreater) betweenthanbelowwiregrassclumpsforfourofthefivemonthsat themesiclocation,butneverdifferedatthexericlocation (Figure1b).Finally,therewasnoclearpatterntovariationinair temperature.Itwasslightlyhigher(0.4C u difference)atmicrosites betweenwiregrassclumpsonlyinMayandAugustinthexeric locationbutdifferedinJuneandinAugustinthemesiclocation.ResourceTreatmentsWeobservedtwiceasmanyseedlingrecruitsinunshaded treatmentplotsversusinshadedplotsacrossallcollectionperiods (Table2,Table3),andtherewereaboutfourtimesmoreseedlings inwateredplotsthaninunwateredplots(Figure2,Table2, Table3).Aninteractionbetweenwatermanipulationandshade Table1. Meanenvironmentalfactorvaluesformicrositesbelowwiregrassclumpsandmicrositesbetweenwiregrassclumps.LocationMeasureLeastsquaremeanbelow 6 SELeastsquaremeanbetween 6 SE MesicRelativeHumidity(%)72.84 6 2.8272.63 6 2.68 SoilTemperature( u C)25.24 6 1.0126.04 6 1.15 AirTemperature( u C)24.86 6 1.3025.19 6 1.33 IncidentPAR(proportion)0.55 6 0.040.85 6 0.02 VolumetricSoilMoisture(%)6.10 6 0.766.01 6 0.76 XericRelativeHumidity(%)69.52 6 2.8267.06 6 2.82 SoilTemperature( u C)26.26 6 1.2527.20 6 1.31 AirTemperature( u C)25.47 6 1.4125.77 6 1.44 IncidentPAR(proportion)0.52 6 0.040.91 6 0.04 VolumetricSoilMoisture(%)4.19 6 0.554.19 6 0.62 doi:10.1371/journal.pone.0039108.t001 FacilitationofSeedlingsbyaBunchgrass PLoSONE|www.plosone.org3July2012|Volume7|Issue7|e39108

PAGE 4

treatmentsuggestedthatshadinghadamorenegativeeffecton seedlingrecruitmentinthecontrolplotsthaninthewateredplots (Table3).Althoughthenumberofseedlingsincreasedwithtime, aninteractionwithwateravailabilityledtogreaterrecruitmentin springinirrigatedsites(Table2,Table3).DiscussionSeedlingestablishmentinthelong-leafpinesavannaisthought tobeinfluencedbytheheterogeneouswaterlimitationsthat characterizethesystem[16,17].Indeed,ourresultsshowedthat Figure1.Micrositemeasurements. Valuesrepresentenvironmentalmeasurementsatpotentialseedlingmicrositesbelowandbetweenclumps ofwiregrass;(A)relativehumidity,(B)soiltemperature,(C)airtemperature,(D)photosyntheticallyactiveradiation,(E)volumetricsoilmois ture. Differencesbetweenmicrositelocationsareindicatedby(m)or(x)formesicorxericlocationrespectively(*=p # 0.05,**=p # 0.08).RefertoTable1 formeanfactorvalues. doi:10.1371/journal.pone.0039108.g001 FacilitationofSeedlingsbyaBunchgrass PLoSONE|www.plosone.org4July2012|Volume7|Issue7|e39108

PAGE 5

seedlingrecruitmentvariedacrossstressgradientsandthatfactors thatrelievemoisturestressenhancedseedlingsuccess.Inconcert withourobservationalfindings,theseresultssuggestthat facilitationisfeasibleinthesystem.However,ourstudydidnot identifydirectfacilitativeeffectsofshadingaspredicted. Ourexpectationthatwaterstressregulatesrecruitmentsuccess wassupportedbasedonourobservationofgreaterseedling establishmentatthemesicendofthenaturalmoisturegradient andinwateringtreatments.Inaddition,thepresenceofan interactionbetweenwaterandseasonsuggestedthatatemporal effectexistedwhereadditionalwaterwasparticularlyimportant duringestablishmentinthespring.Yet,despitethelargernumber ofseedlingsinwateredplotsinthespring,neithershadenorwater wassufficienttosustainthenewseedlingsthroughthehotsummer monthsofthestudyperiod. Theresultthatshadingdoesnotfacilitateseedlingrecruitment andthatrecruitmentwasactually lower undershadingtreatments thanunderfullsunissurprisinggiventheintensesunlightandhigh soilsurfacetemperaturesinthisecosystem[31].Moisture availabilitydidappeartomodulatetheimpactofshadetosome extentgiventhattherewaslessofanegativeeffectofshadeon seedlingrecruitmentwhenwaterwasapplied.Nevertheless,our predictionofaninteractivetemporaleffectofshadeandwaterwas notsupported. Severalconditionsassociatedwithourexperimentmayhave influencedourresults.First,thelevelofshadeweprovidedwasat thehighendoftherangeobservedinthefield[31]andmayhave inhibitedseedlinggrowth.Thismaybeanexampleofanupper boundonfacilitativeeffectsinresponsetoextremeenvironmental stressassuggestedinarecenttheoreticalmodel[32].Second,itis uncleartowhatdegreetheshadestructuresmayhaveinhibited wateravailability.Otherstudiesinwaterlimitedsystemshave observedsimilarreductionsinrecruitmentsuccesswhenwater availabilitywasobstructedinconcertwithshade[33,34]. Furthermore,thepresenceoftheinteractiontermbetweenshade andwatertreatmentssuggeststhatreducedmoistureavailability mayberesponsibleforthenegativeeffectofshade.Ineithercase, Table2. Seedlingrecruitmentanalysisresults.FixedEffects EffectNumDFDenDFFValuePr F resourcetreatment11211.560.0053 collectionevent3364.770.0067 resource*collection3366.180.0017 locationongradient1120.400.54 resource*location1120.440.5181 collection*location3361.410.256 resource*collection*location3360.810.4943 shadetreatment14814.230.0004 resource*shade1483.880.0546 collection*shade3481.050.3794 resource*collection*shade3480.660.5779 location*shade1481.070.3067 resource*location*shade1480.190.6673 collection*loc*treatment3480.040.9872 resource*collection*location*shade3480.170.9173 PlannedContrasts NumDFDenDFFValuePr F springvfall1368.730.0055 2006v20071365.430.0255 Resultsofsplit-splitplotanalysisofseedlingrecruitmentacrosstwoshade treatments(shade,noshade),twowatertreatments(water,control),two gradientlocations(mesic,xeric)andfourcollectionperiods(springandfall, 2006and2007). doi:10.1371/journal.pone.0039108.t002 Table3. Seedlingrecruitmentleastsquaremeans.Leastsquaremean 6 SE resourcetreatmentControl1.54 6 0.95 Water6.09 6 0.95 locationongradientMesic3.39 6 0.95 Xeric4.23 6 0.95 shadetreatmentShade2.59 6 0.74 NoShade5.04 6 0.74 collectioneventMay20063.89 6 0.87 Nov20062.22 6 0.87 May20075.66 6 0.87 Nov20073.48 6 0.87 resource*collectionControl*May061.41 6 1.24 Control*Nov061.81 6 1.24 Control*May071.34 6 1.24 Control*May061.59 6 1.24 Water*May066.38 6 1.24 Water*Nov062.63 6 1.24 Water*May079.97 6 1.24 Water*May065.38 6 1.24 resource*shadeControl*NoShade2.13 6 1.05 Control*Shade0.95 6 1.05 Water*NoShade7.97 6 1.05 Water*Shade4.23 6 1.05 Leastsquaremeansofseedlingrecruitmentacrosstwoshadetreatments (shade,noshade),twowatertreatments(water,control),twogradientlocations (mesic,xeric)andfourcollectionperiods(springandfall,2006and2007). doi:10.1371/journal.pone.0039108.t003 Figure2.Seedlingresponsetoexperimentalshadingand wateraddition. Valuesaremeannumberofseedlingrecruitsin responsetoexperimentalmanipulationofwateravailabilityandshade. RefertoTable2foranalysisresultsandTable3formeanfactorvalues. doi:10.1371/journal.pone.0039108.g002 FacilitationofSeedlingsbyaBunchgrass PLoSONE|www.plosone.org5July2012|Volume7|Issue7|e39108

PAGE 6

itcouldbearguedthatourtreatmentcombinationswereoutside therangeofenvironmentalameliorationwherefacilitationcan counteractdroughtstressinseedlings[19]Additionally,although nospatialgradienteffectswereindicatedbythemixedmodel results,inferenceinthiscomponentofthestudyislimitedbythe clumpedconfigurationofplotswithinthegradientlocations.Itis alsoimportanttonotethattherewereextremedroughtconditions duringthestudyperiod(NCDCaverageannualprecipitationfor GAis127cm,yet2006–2007averagewas101cm).Insuchdry, hotconditions,waterlimitationmayhaveimposedsimilar constraintsonseedlingsurvivalinmesicasinxericlocations,thus potentiallymaskinganeffectofthenaturalmoisturegradient. Finally,thereisapossibilitythatseedpredatorscouldbeattracted totheshadestructuresanddisproportionallyreducedthenumber ofseedlingsinshadedlocations[35].Determiningtheinfluenceof theseconfoundingfactorsrequiresadditionaltesting. Ourresultssuggestthatheavyshadingalonedoesnotpromote facilitationeventhoughwaterisparamountforrecruitmentinthis system.Thisevidencesuggeststhatthepreviouslyobserved relationshipbetweenwiregrassproductivityandsystemspecies diversitymayalternativelyhingeoncommunitywideeffectsof wiregrasspresenceratherthantheimpactofindividualclumps. Weshowthatmicrositesincloseproximitytowiregrassclumps tendtobeslightlycoolerandmorehumidthanthosewithout anearbywiregrassclump.Whenthesesmalldifferencesare extrapolateduptothecommunityscale,onecouldarguethatan areawithadensewiregrassunderstorycouldcontainmore micrositessuitableforseedlingrecruitmentthananareathathad asparseunderstory.Inthissituation,theoveralleffectof facilitationoncommunityspeciescompositioncouldbesignificant (eg.[36]).Theseresultsalsohaveimplicationsforassessingthe importanceofpositiveplant-plantinteractionsinothersystems. Theysuggestthattheremaybeconsiderablevariationinthegrain ofspatialandtemporaleffectsandthusamuchwiderrangeof stressgradientconditionsmustbeconsidered.Thismayexplain whyapredictablerelationshiphasyettobeobservedbetween stressorsandfacilitativeeffectsovertimeandspace[32,33].We concludethatidentificationofpossiblefacilitativeeffectson recruitmentinthissystemwillrequireexaminationofawider rangeofconditionsthaninourstudy,particularlywithregardsto agradientofwiregrasscoverandduringperiodsofgreaterrainfall.AcknowledgmentsWethankthemembersoftheIchauwayplantlabforassistanceinthefield includingM.Kaeser,M.Creech,K.Stuble,S.Wiggers,K.Coffey,and manyothers.AssistancewithstatisticalanalyseswasprovidedbyM. BrennanattheUniversityofFloridaInstituteofFoodandAgricultural ScienceStatisticsDepartment.W.Cropper,P.Armsworth,C.Dumoulin, andseveralanonymousreviewersprovidedhelpfulcommentsonearly versionsofthismanuscript.WealsoacknowledgethecontributionsofR. Mitchellasacollaboratorintheimplementationofthelongtermresource manipulationexperiment.TheUniversityofFlorida’sDepartmentof WildlifeEcologyandConservationandtheJ.W.JonesEcologicalResearch Centerprovidedlogisticalsupportthroughouttheproject.AuthorContributionsConceivedanddesignedtheexperiments:GDILKKEMB.Performedthe experiments:GDI.Analyzedthedata:GDI.Contributedreagents/ materials/analysistools:LKKEMB.Wrotethepaper:GDILKKEMB.References1.CallawayRM(1995)Positiveinteractionsamongplants.BotanicalReview61: 306–349. 2.BrookerRW,MaestreFT,CallawayRM,LortieCL,CavieresLA,etal.(2008) Facilitationinplantcommunities:thepast,thepresent,andthefuture.Journal ofEcology96:18–34. 3.FreestoneAL(2006)Facilitationdriveslocalabundanceandregional distributionofarareplantinaharshenvironment.Ecology87:2728–2735. 4.GreenleeJT,CallawayRM(1996)Abioticstressandtherelativeimportanceof interferenceandfacilitationinmontanebunchgrasscommunitiesinwestern montana.TheAmericanNaturalist148:386–396. 5.HackerSD,BertnessMD(1995)Morphologicalandphysiologicalconsequences ofapositiveplantinteraction.Ecology76:2165–2175. 6.BrunoJF,StachowizJJ,BertnessMD(2003)Inclusionoffacilitationinto ecologicaltheory.TrendsinEcologyandEvolution18:119–125. 7.KikvidzeZ,KhetsurianiL,KikodzeD,CallawayRM(2006)Seasonalshiftsin competitionandfacilitationinsubalpineplantcommunitiesofthecentral Caucasus.JournalofVegetationScience17:77–82. 8.HackerSD,BertnessMD(1999)Experimentalevidenceforfactorsmaintaining plantspeciesdiversityinaNewEnglandsaltmarsh.Ecology80:2064–2073. 9.ShumwaySW(2000)Facilitativeeffectsofasandduneshrubonspeciesgrowing beneaththeshrubcanopy.Oecologia124:138–148. 10.VintonMA,BurkeIC(1995)Interactionsbetweenindividualplant-speciesand soilnutrientstatusinshortgrasssteppe.Ecology76:1116–1133. 11.HolmgrenM,SchefferM,HustonMA(1997)Theinterplayoffacilitationand competitioninplantcommunities.ecology78:1966–1975. 12.ZouCB,BarnesPW,ArcherS,McMurtryCR(2005)Soilmoisture redistributionasamechanismoffacilitationinSavannatree-shrubclusters. Oecologia145:32–40. 13.ShreveF(1931)Physicalconditionsinsunandshade.Ecology12:96–104. 14.HackerSD,GainesSD(1997)Someimplicationsofdirectpositiveinteractions forcommunityspeciesdiversity.ecology78:1990–2003. 15.KeddyPA,SmithL,CampbellDR,ClarkM,MontzG(2006)Patternsof herbaceousplantdiversityinsoutheasternLouisianapinesavannas.Applied VegetationScience9:17–26. 16.IaconaG,KirkmanLK,BrunaEM(2010)Effectsofresourceavailabilityon seedlingrecruitmentinafire-maintainedsavanna.Oecologia163:171–180. 17.KirkmanLK,MitchellRJ,HeltonRC,DrewMB(2001)Productivityand speciesrichnessacrossanenvironmentalgradientinafire-dependentecosystem. AmericanJournalofBotany88:2119–2128. 18.PecotSD,MitchellRJ,PalikBJ,MoserEB,HiersJK(2007)Competitive responsesofseedlingsandunderstoryplantsinlongleafpinewoodlands: separatingcanopyinfluencesaboveandbelowground.CanadianJournalof ForestResearch-RevueCanadienneDeRechercheForestiere37:634–648. 19.HolmgrenM,SchefferM(2010)Strongfacilitationinmildenvironments:the stressgradienthypothesisrevisited.JournalofEcology98:1269–1275. 20.PeetRK,AllardDJ(1993)Longleafpinevegetationofthesouthernatlanticand easterngulfcoastregions:apreliminaryclassification;;Tallahassee,FL,USA. 21.MitchellRJ,KirkmanLK,PecotSD,WilsonCA,PalikBJ,etal.(1999)Patterns andcontrolsofecosystemfunctioninlongleafpine-wiregrasssavannas.I. Abovegroundnetprimaryproductivity.CanadianJournalofForestResearchRevueCanadienneDeRechercheForestiere29:743–751. 22.AbdElRahmanAA,BatanounyKH(1965)Transpirationofdesertplantsunder differentenvironmentalconditions.TheJournalofEcology53:267–272. 23.MacMahonJA,SchimpfDJ(1981)WaterasafactorinthebiologyofNorth Americandesertplants.WaterinDesertEcosystems:114–171. 24.WestJB,DonovanLA(2004)EffectsofindividualbunchgrassesonpotentialC andNmineralizationoflongleafpinesavannasoils.JournaloftheTorrey BotanicalSociety131:120–125. 25.OlnessA,ArcherD(2005)Effectoforganiccarbononavailablewaterinsoil. SoilScience170:90–101. 26.FulbrightTE,KutiJO,TiptonAR(1995)Effectsofnurse-plantcanopy temperaturesonshrubseedgerminationandseedlinggrowth.ActaOecologicaInternationalJournalofEcology16:621–632. 27.DrewMB,KirkmanLK,GholsonAKJ(1998)ThevascularfloraofIchauway, BakerCounty,Georgia:aremnantlonleafpine/wiregrassecosystem.Castanea 63:1–24. 28.McNabWH,AversPE(1994)Ecologicalsub-regionsoftheUnitedStates: sectiondescriptionsWashington,DC.:USDAForestServiceAdministration. AdministrativepublicationWO-WSA-5AdministrativepublicationWO-WSA5. 29.GoebelPC,PalikBJ,KirkmanLK,WestL(1997)Fieldguide:landscape ecosystemtypesofIchauway.JosephW.JonesEcologicalResearchCenterat Ichauway,Newton,Ga.Reportnumber97–1Reportnumber97–1. 30.ToppGC,DavisJL,AnnanAP(1980)Electromagneticdeterminationofsoil watercontent:MeasurementsinCoaxialTransmissionLines.WaterResources Research16:574–582. 31.IaconaG(2008)Seedlingrecruitmentasadriverofspeciesrichnessinthe understoryofthelongleafpinesavanna[M.S.Thesis].Gainesville:Universityof Florida.78p. 32.ButterfieldBJ(2009)Effectsoffacilitationoncommunitystabilityanddynamics: synthesisandfuturedirections.JournalofEcology97:1192–1201.FacilitationofSeedlingsbyaBunchgrass PLoSONE|www.plosone.org6July2012|Volume7|Issue7|e39108

PAGE 7

33.TielborgerK,KadmonR(2000)TemporalEnvironmentalVariationTipsthe BalancebetweenFacilitationandInterferenceinDesertPlants.Ecology81: 1544–1553. 34.WeedonJT,FacelliJM(2008)Desertshrubshavenegativeorneutraleffectson annualsattwolevelsofwateravailabilityinaridlandsofSouthAustralia. JournalofEcology96:1230–1237. 35.MansonRH,StilesEW(1998)Linksbetweenmicrohabitatpreferencesandseed predationbysmallmammalsinoldfields.Oikos82:37–50. 36.CavieresLA,BadanoEI(2009)Dofacilitativeinteractionsincreasespecies richnessattheentirecommunitylevel?JournalofEcology97:1181–1191.FacilitationofSeedlingsbyaBunchgrass PLoSONE|www.plosone.org7July2012|Volume7|Issue7|e39108