JournaloftheTorreyBotanicalSociety133(2),2006,pp.240–254EffectsofatmosphericnitrogendepositionontheherbaceouslayerofacentralAppalachianhardwoodforestFrankS.Gilliam1andAnneW.HockenberryDepartmentofBiologicalSciences,MarshallUniversity,Huntington,WV25755-2510MaryBethAdamsU.S.D.A.ForestService,TimberandWatershedLaboratory,Parsons,WV26287-0404GILLIAM,F.S.,A.W.HOCKENBERRY(DepartmentofBiologicalSciences,MarshallUniversity,Huntington,WV,25755-2510),ANDM.B.ADAMS(U.S.D.A.ForestService,TimberandWatershedLaboratory,Parsons,WV,26287-0404).EffectsofatmosphericnitrogendepositionontheherbaceouslayerofacentralAppalachianhardwoodforest.J.TorreyBot.Soc.133(2):240–254.2006.—Additionsofnitrogen(N)havebeenshowntoalterspeciesdiversityofplantcommunities,withmostexperimentalstudieshavingbeencarriedoutincom-munitiesdominatedbyherbaceousspecies.Weexaminedseasonalandinter-annualpatternsofchangeintheherbaceouslayeroftwowatershedsofacentralAppalachianhardwoodforestthatdifferedinexperimentalNtreatment.ThisstudywascarriedoutattheFernowExperimentalForest,WestVirginia,usingtwoadjacentwatersheds:WS4(mature,second-growthhardwoodstand,untreatedreference),andWS3(ϳ25-yrold,treatedwith35kgNhaϪ1yrϪ1as(NH4)2SO4yrϪ1).Sevencircular0.04-hasampleplotswereestablishedineachwatershedtorepresentitsfullrangeofelevationandslopeaspect.Theherbaceouslayerwassampledbyidentifyingandvisuallyestimatingcover(%)ofallvascularplantsՅ1minheightwithin101-m2circularsub-plotsineachsampleplot,foratotalof1401-m2sub-plots.Samplingwascarriedoutinmid-Julyof1991andrepeatedatapproximatelythesametimein1992.In1994,thesesameplotsweresampledeachmonthfromMaytoOctober.Seasonalpatternsofherblayerdynamicswereassessedforthecomplete1994dataset,whereasinter-annualvariabilitywasbasedonplotdatafrom1991,1992,andtheJulysampleof1994.Therewerenosignificantdifferencesbetweenwatershedsforanysampleyearforanyoftheotherherblayercharacteristicsmeasured,includingherblayercover,speciesrichness,evenness,anddiversity.CoveronWS4decreasedsig-nificantlyfrom1991to1992,followedbynochangeto1994.Bycontrast,herblayercoverdidnotvarysignificantlyacrossyearsonWS3.Coveroftheherbaceouslayerofbothwatershedsincreasedfromearlyinthegrowingseasontothemiddleofthegrowingseason,decreasingthereafter,withnosignificantdifferencesbetweenWS3andWS4foranyofthemonthlycovermeansin1994.Similarseasonalpatternsfoundforherblayercover—andlackofsignificantdifferencesbetweenwatersheds—werealsoevidentforspeciesdiversityandrichness.Bycontrast,therewaslittleseasonalchangeinherblayerspeciesevenness,whichwasnearlyidenticalbetweenwatershedsforallmonthsexceptOctober.Seasonalpatternsforindividualspecies/speciesgroupswerecloselysimilarbetweenwatersheds,especiallyforViolarotundifoliaandViolaspp.SpeciesrichnessandspeciesdiversitywerelinearlyrelatedtoherblayercoverforbothWS3andWS4,suggestingthatspatialandtemporalincreasesincoverweremorerelatedtorecruitmentofherblayerspeciesthantogrowthofexistingspecies.Resultsofthisstudyindicatethattherehavebeennegligibleresponsesoftheherblayerto6yrofNadditionstoWS3.Keywords:easterndeciduousforest,forestecosystems,foreststrata,herbaceouslayer,nitrogendeposition,seasonalpatterns.Experimentalstudiesexaminingtheeffectsofnitrogen(N)onspeciesdiversityofplantcom-munitieshavealonghistory,withthefirstandpossiblybestknownstudybeingtheParkGrassExperimentinRothamsted,England(seeRich-ardson1938foranearlyreference).MorerecentstudiesattheCedarCreekNaturalAreainMin-nesota(e.g.,Tilman1987)confirmresultsofthisearlierwork—excessNcanprofoundlyalterspeciescompositionanddecreasespeciesdiver-sityofplantcommunities.Whereasmostsuchstudieshavebeencarriedoutonherb-dominatedE-mail:gilliam@marshall.eduReceivedforpublicationMarch7,2005,andinre-visedformSeptember26,2005.1communities,farfewerhavebeendoneinfor-estedecosystems(Bobbinketal.1998).Thisisasignificantdiscrepancy,however,consideringthatforests(1)oftenoccupyareasreceivinghighratesofatmosphericdepositionofN(e.g.,east-ernU.S.andcentralEurope),and(2)generallydisplaygreatstructuralcomplexity,withmostoftheirspeciesdiversityoccurringintheloweststratumofvegetation—theherbaceouslayer(seeGilliamandRoberts2003aforcommonly-usedsynonyms).ThetimeofresponseofplantcommunitiestoexperimentaladditionsofNcanvary,withsomeplantcommunitiesexhibitingrapidresponses.Forexample,Tilman(1987)reportedsignificantchangesinspeciescompositionofMinnesota2402006]GILLIAMETAL.:EFFECTSOFNITROGENDEPOSITIONONFORESTHERBS241Table1.CharacteristicsofstudywatershedsoftheFernowExperimentalForest,WV.VariableWS3WS4Standage(yr)StandhistoryArea(ha)AspectElevation(m)CumulativeNdeposition,1989–1994(kghaϪ1)AmbientAddedTotalTreedensity(stems/ha)Treebasalarea(m2haϪ1)ϳ25,even-agedclearcut/Nadditions34.3S735–86053210263241025.0Ͼ90,mixed-ageselectcut38.7S–SE750–8705305394840.0old-fieldstoalllevelsofNadditionswithinthefirstyearfollowinginitiationoftreatment.Infact,Ntreatmentsatthissitehaveledtohighlysignificantresponsesofspeciesineveryyearofthestudy,althoughthespecificpatternofre-sponsehasvariedinter-annually(Tilman1989).NotallN-additionstudies,however,haveshownsuchanimmediateresponse.AlthoughHubertyetal.(1998)foundthatNadditionsincreasedplantbiomassofMichiganold-fields,therewerenosignificantchangesinspeciescompositionordiversityovertheseven-yearperiodofthestudy.Herblayerplantsaregenerallysensitivetosoilfertility,bothintermsofgrowthandfoliarnutrientcontent(Muller2003),andsomespe-cieshaveevenbeenusedasindicatorsoffertil-ity(PregitzerandBarnes1982).Herblayercom-munitieshavebeenshowntochangethroughsuccessionaltimeinresponsetochangesinsoilnutrients,includingNandcations(ChristensenandGilliam2003).Itmightbeexpected,how-ever,thattheherblayerofforestsmaydisplayalag-responsetoexperimentaladditionsofN,giventhestructuralcomplexityofforestecosys-tems,whereinresourceavailabilitycanbegreat-lyalteredbyothervegetationstrata.Indeed,Gilliametal.(1994)concludedthatexperimentaladditionsofNtoawatershedattheFernowExperimentalForest,WestVirginia,hadnosignificanteffectsontheherblayercom-position,biomass,andfoliarnutrientsafter3yroftreatment.Amongthepractical,interpretivelimitationsofthatstudy,however,arethat(1)althoughitwascarriedoutattheheightofthegrowingseason,itbasedonaone-timesamplingoftheherblayer,precludingexaminationofsea-sonalpatterns,and(2)itprovidedonlyathree-yearassessmentofNtreatmenteffects(i.e.,asinglepointintime3yrfollowinginitiationofNadditions).ThepurposeofthisstudywastocharacterizetemporalpatternsofchangeintheherbaceouslayeroftwowatershedsofFEFusingtwotem-poralscalestoassesspotentialresponsetoex-perimentaladditionsofN.Weexaminednumer-ouscharacteristicsoftheherblayer,includingcoverandspeciesdiversity,richness,andeven-ness,becausetheeffectsofNhasbeenshowninpreviousstudiestoaffectvariousaspectsoftheplantcommunitiesdifferently(Huberty1998).Thefirsttemporalscaleexaminedinter-annualchangeoverafour-yearperiod(repre-senting3to6yrfollowinginitiationofexperi-mentaladditionsofN)bycomparingsamplinginJulyof1991,1992,and1994.Thesecondscalefocusedonseasonalpatternsbasedonmonthlymeasurements(MaythroughOctober)ofasingleyear(1994).Inparticular,wewereinterestedinseeingifthereweredifferencesintemporalpatternsthatcouldberelatedtoexper-imentaladditionsofNfollowing6yroftreat-ment.MaterialsandMethods.STUDYSITE.ThisstudywascarriedoutattheFernowExperimen-talForest(FEF),whichoccupiesϳ1900haofmontanehardwoodforestsintheAlleghenyMountainsectionoftheunglaciatedAlleghenyPlateauinTuckerCounty,north-centralWestVirginia(39Њ03ЈN,79Њ49ЈW).Meanannualprecipitationisapproximately1430mmyrϪ1,withmostprecipitationoccurringbetweenAprilandSeptember(Gilliametal.2001).Twoadja-centwatershedswereselectedforthisstudy.StandsonWS3wereϳ25yr-oldatthetimeofmostrecentsamplinginthisstudy,andareeven-agedstandswhichdevelopedfollowingclear-cutting.WS4supporteduneven-agedstandsϾ90yrold(Table1).Bothwatershedssupportprimarilymixed242JOURNALOFTHETORREYBOTANICALSOCIETY[VOL.133hardwoodstands,withdominanttreespeciesvaryingwithstandage.Early-successionalspe-cies,suchasBetulalentaL.,PrunusserotinaEhrh.,andLiriodendrontulipiferaL.aredomi-nantonWS3,whereaslate-successionalspecies,suchasAcersaccharumMarshallandQuercusrubraL.,aredominantonWS4.Soils,similaramongstudywatersheds,arerelativelythin(Ͻ1mindepth),acidic,sandy-loamInceptisolsoftwoseries:Berks(loamy-skeletal,mixed,mesicTypicDystrochrept)andCalvin(loamy-skeletal,mixed,mesicTypicDystrochrept;Gilliametal.1994).Soilsofthestudywatershedsaregenerallyacidic,highinorganicmatter,andhighincationexchangeca-pacity.WS3servedasthe‘‘treatment’’watershed,re-ceivingthreeaerialapplicationsof(NH4)2SO4yrϪ1,beginningin1988.March(orsometimesApril)andNovemberapplicationsrepresentap-proximately7kgNhaϪ1;Julyapplicationsareapproximately21kgNhaϪ1,foratotalof35kgNhaϪ1yrϪ1.WS4servedasuntreatedcontrolwatershed.PreviousstudiesconfirmthatthesetreatmentshaveincreasedNavailabilityonWS3relativetoWS4,withannualnetnitrificationav-eraging144and115kgNhaϪ1yrϪ1,respective-ly,forthethree-yearperiodof1993to1995(Gilliametal.2001).CharacteristicsforbothstudywatershedsaresummarizedinTable1.FIELDSAMPLING.Theherbaceouslayerwassampledwithinsevencircular0.04-hasampleplotsineachwatershed(representingthefullrangeofelevationandslopeaspect)byidenti-fyingandvisuallyestimatingcover(%)ofallvascularplantsՅ1minheightwithin101-m2circularsub-plotsineachsampleplot,foratotalof1401-m2sub-plots.Sub-plotswerelocatedwithinsampleplotsusingastratified-randompolarcoordinatesmethod(Gaiser1951),whichwasemployedtoavoidover-samplingthecenterregionofcircularplots.Samplingwascarriedoutinmid-Julyof1991and1992.In1994,plotsweresampledmonthlyfromMaytoOctober.NomenclaturefollowsGleasonandCronquist(1991).DATAANALYSIS.Thedesignofthisstudyisanexampleofsimplepseudoreplication,witheachwatershedrepresentinganexperimentalconditionwithasamplesizeofone(Hurlbert1984);thisisacommonstatisticalproblemforfieldstudiescarriedoutatthelevelofawater-shedecosystem.Accordingly,interpretationofdatashouldbedonetakingthatintoaccount.Wearealsoawareofthepotentiallyconfoundingef-fectsoftheNtreatmentwithstandagediffer-ences.Itshouldbenoted,however,thatpub-lishedresultsofinitialsamplingoftheherblay-ershortlyafterinitiationofNadditionstoWS3havedocumentedalackofsignificantdifferenc-esbetweenWS3andWS4thatarerelatedtodifferencesinstandage(GilliamandTurrill1993).Thus,itisourcontentionthatanyeffectsthatwereportaretreatmenteffects,ratherthanpre-existingdifferencesamongwatersheds,giv-entheclosesimilaritiesamongwatershedsinseveral‘‘site’’variablessuchassoiltexture,soilorganicmatter,andextractablenutrients(Gil-liametal.1994).Datawerecomparedbetweenstudywater-shedstoexaminetwoseparatetemporalpat-terns:(1)meansforcover(totalandselectedin-dividualspecies),speciesdiversity,speciesrich-ness,andspeciesevennessforone-timesam-plingofwatersheds(inJuly)in1991and1992andtheJulysamplefrom1994;and(2)monthlymeansforthesesamevariablesoverasix-monthperiodinasinglegrowingseason.Meanswerecalculatedfromvaluesofthesevenplotsforagivensampleperiod.Speciesdiversityoftheherbaceouslayerwascalculatedonaperplotbasisusingtheln-basedShannonindex(HЈ)equation.BecauseHЈcombinesbothspeciesrichnessandevenness,richnessandevennesscanhavevaryingrelationshipswithHЈ(StirlingandWilsey2001).Accordingly,thePielou’sevennessindex(J)wasalsodeterminedforeachplot(Pielou1966).RelationshipsbetweenHЈandrichnessversuscoverweredeterminedwithlinearregression(Zar1996).Meanswerecom-paredbetweenthetwostudywatershedsusingt-testsandamongyearsofsamplingwithanal-ysisofvarianceandmultiplerangestests(Zar1996).Changesinspeciescompositionoftheher-baceouslayerwereanalyzedwithdetrendedcor-respondenceanalysis(DCA).Inter-annualpat-ternswereassessedbyrunningDCAon1991,1992,andJuly1994datacombined,followedbycalculationofcentroidsofsevenplotsperwa-tershedperyear.SeasonalpatternswereassessedsimilarlybyrunningDCAontheentire1994datasetandcalculatingcentroidsofclustersofsevensampleplotsperwatershedpermonth.Results.INTER-ANNUALCOMPARISONS.Therewerenosignificantdifferences(PϾ0.05)be-tweenwatershedsforanysampleyearforanyoftheotherherblayercharacteristicsmeasured,2006]GILLIAMETAL.:EFFECTSOFNITROGENDEPOSITIONONFORESTHERBS243includingherblayercover,speciesrichness,evenness,anddiversity.CoveronWS4,how-ever,decreasedsignificantly(PϽ0.05)from1991to1992,followedbynochangeto1994.Bycontrast,herblayercoverdidnotvarysig-nificantlyacrossyearsonWS3(Fig.1a).Speciesrichnessvariedacrosssampleyearssimilarlybetweenwatersheds,withnochangefrom1991to1992,followedbyasignificantincreaseto1994(Fig.1b).SpeciesevennessdidnotvarythroughtimeonWS4,but,followingnochangefrom1991to1992,increasedsignif-icantlyin1994onWS3(Fig.1c).Finally,spe-ciesdiversityoftheherbaceouslayerincreasedsignificantlyfrom1992to1994onWS4,withdiversityfor1991beingintermediatebetweenthesetwoextremes.OnWS3,diversitydidnotchangefrom1991to1992,butincreasedsignif-icantlyto1994(Fig.1d).Therewerefewchangesincoverofdominantspeciesoftheherbaceouslayer(Tables2and3).Detrendedcorrespondenceanalysis(DCA),usedtoassessannualchangeinoverallspeciescom-positionoftheherbaceouslayerofWS3versusWS4from1991to1994,confirmedthisquan-titatively.DCAordinationrevealedbothagen-erallyclosesimilarityofspeciescompositionbe-tweenWS3andWS4,aswellasminimalchangeincompositionovertimeforbothwa-tersheds(Fig.2).Basedonthelocationofcen-troidsinordinationspace,speciescompositionchangedlittleonWS3from1991to1992andmuchmoresofrom1992to1994.OnWS4,changeincompositionwasgreaterfrom1992to1994thanitwasfrom1991to1992(Fig.2).SEASONALPATTERNS.Coveroftheherbaceouslayerofbothwatershedsincreasedfromearlyinthegrowingseasontothemiddleofthegrowingseason,decreasingthereafter(Fig.3a).Indeed,therewerenosignificantdifferencesbetweenWS3andWS4foranyofthemonthlycovermeansin1994.Thesameseasonalpatternsfoundforherblayercover—andlackofsignif-icantdifferencesbetweenwatersheds—werealsoevidentforspeciesdiversityandrichness(Figs.3b,c).Bycontrast,therewaslittlesea-sonalchangeinherblayerspeciesevenness,whichwasnearlyidenticalbetweenwatershedsforallmonthsexceptOctober(Fig.3d).Monthlymeancoverwasalsocalculatedforselecteddominantspecies/speciesgroups,in-cludingherbaceousplants,suchasViolaspp.andRubusspp.,andseedlingsoftreespecies,suchasAcerpensylvanicumL.andPrunusser-otina.SeasonalpatternsforV.rotundifoliaMichx.andViolaspp.(agroupvariouslycom-prisingV.blandaWilld.,V.canadensisL.,V.papilionaceaPursh,V.pensylvanicaMichx.,V.sagittataAit.,andV.sororiaWilld.)weresim-ilarbetweenwatersheds,althoughtheycontrast-edwithpatternsforoverallcover(Fig.3a)inexhibitingadistinctmaximumcoverinJune(Fig.4).Rubusspp.andA.pensylvanicumweregenerallyhigherincoveronWS3thanonWS4,whereasP.serotinawasmoreseasonallydy-namiconWS3thanonWS4(Fig.4).Usingindividualplotvaluesacrossallmonth-lysamplingperiodsforeachwatershed(i.e.,combiningbothspatialandtemporalvariation),speciesrichness(S)wassignificantlyandline-arlyrelatedtoherblayercover(C)forbothWS3andWS4(Fig.5a).Thelinearmodelswerenear-lyidenticalbetweenwatersheds:Sϭ6.47ϩ0.60C,r2ϭ0.58,PϽ0.0001forWS3;Sϭ7.12ϩ0.59C,r2ϭ0.53,PϽ0.0001forWS4.Speciesdiversity(HЈ)wasalsosignificantlyandlinearlyrelatedtoherblayercover(C)forbothwatersheds(Fig.5b).Modelswerethefollow-ing:HЈϭ1.47ϩ0.051C,r2ϭ0.68,PϽ0.0001forWS3;HЈϭ1.68ϩ0.037C,r2ϭ0.29,PϽ0.0002forWS4.Similartotheanalysisofannualchange(i.e.,Fig.2),DCAwasalsousedtoassessseasonalchangeinoverallspeciescompositionoftheherbaceouslayerofWS3versusWS4.DCAre-vealedthat,althoughtheherblayerofbothwa-tershedsexhibitedsimilardegreesandpatternsofseasonalchange(i.e.,relativelylargechangefromMaytoJune,lesschangefromJunetoSep-tember,andgreatestchangefromSeptembertoOctober),thedirectionofchangecontrastedsharplybetweenwatershedsinordinationspace(Fig.6).Discussion.INTER-ANNUALCOMPARISONS.Re-sultsofourstudy,despitebeingbasedonmoreintensivesamplingcomprisingafour-yearperiod,supportthefindingsofearlierstudies,whichwerebasedonone-timesampling.Inparticular,thereare(1)remarkablesimilaritiesbetweenthesewatershedsthatdiffergreatlyinstandage,and(2)negligibleresponsesoftheherblayertotheNadditionstoWS3.Meancoverfortheherblayerdidnotdiffersignificantlybetweenwatersheds(Fig.1a),av-eraging19and23%forWS3andWS4,re-spectively.ThesevaluesarecomparabletothatfoundforHubbardBrookExperimentalForest(24%)(Siccamaetal.1970),greaterthanthe244JOURNALOFTHETORREYBOTANICALSOCIETY[VOL.133nsenpaoe—Ml.odrtenhosrce(ta4SwWnedvingaa)rraobfsdreadeayhsg—notnmeam)t5a0e.r0t(Ͻ3SPW(tnfeorerfefiydalylstunoaeccfiaibnrgeihsteohnterorr.aafer)eyDttr(eolyteelismbreaaivsraiedhvtdynhntaai,wr)oCsf(nassedseMehns.rn4eev9te9a1w,–)nB19ee(9s1wtseenV,bh)cW5i,0rt.s,0)eAroϽ(FrPlea(vttonnceer)meEifrfSei1pdxϮEylt(nnwaaocenfiiMrenFgies.1htto.fGonIeF)rraebw2006]GILLIAMETAL.:EFFECTSOFNITROGENDEPOSITIONONFORESTHERBS245Table2.HerbaceouslayerspeciesonWS3.Valuesshownareimportancevaluesbasedonrelativecover.1991SpeciesIV1992SpeciesIV1994SpeciesIVViolaspp.SmilaxrotundifoliaLycopodiumflabelliformeAcerpensylvanicumSassafrasalbidumRubusspp.PrunusserotinaPolygonumbistortaDryopterismarginalisFagusgrandifoliaPolysticumacrostichoidesA.saccharumDioscoreaquaternataFraxinusamericanaPanicumspp.UvulariaperfoliataCaryaspp.MagnoliaacuminataLaporteacanadensisCimicifugaracemosa23.719.115.912.910.04.43.12.31.61.51.51.40.70.40.40.40.30.20.10.1Violaspp.SmilaxrotundifoliaAcerpensylvanicumLycopodiumflabelliformeRubusspp.SassafrasalbidumPrunusserotinaPolygonumbistortaFagusgrandifoliaThelypterisnoveboracensisDryopterismarginalisMagnoliaacuminataA.saccharumPanicumspp.BoehmeriacylindricaRobiniapseudoacaciaUvulariasessilifoliaA.rubrumLiriodendrontulipiferaQuercusrubraFraxinusamericana26.021.612.810.69.98.02.92.11.41.00.60.60.50.40.40.40.30.10.10.10.1Violaspp.SmilaxrotundifoliaRubusspp.AcerpensylvanicumSassafrasalbidumPrunusserotinaLycopodiumflabelliformeDryopterismarginalisDiospyrosvirginianaQuercusrubraMedeolavirginianaPanicumspp.Carexspp.Asterspp.MagnoliaacuminataAcerrubrumA.saccharumFragariaspp.RobiniapseudoacaciaUvulariaperfoliataOsmorhizaclaytoniiArisaematriphyllumSolidagocaesiaBetulalentaLiriodendrontulipifera23.920.613.412.76.96.83.52.61.91.00.90.80.80.70.70.60.40.30.30.30.30.30.20.10.1Table3.HerbaceouslayerspeciesonWS4.Valuesshownareimportancevaluesbasedonrelativecover.1991SpeciesIV1992SpeciesIV1994SpeciesIVLaporteacanadensisViolaspp.AcerpensylvanicumVacciniumvacillansRubusspp.A.saccharumPrunusserotinaDryopterismarginalisSilaxrotundifoliaPolysticumacrostichoidesPolygonumbistortaAmbrosiaartemisifoliaQuercusrubraDioscoreaquaternataCastaneadentataBrachyelytrumerectumFagusgrandifoliaFraxinusamericanaLiriodendrontulipiferaCimicifugaracemosaCaryaspp.MagnoliaacuminataNyssasylvaticaUvulariaperfoliataSassafrasalbidum16.414.110.89.39.08.35.65.14.53.73.21.81.61.21.11.01.00.90.30.30.30.30.10.10.1LaporteacanadensisViolaspp.SmilaxrotundifoliaAcerpensylvanicumPrunusserotinaRubusspp.VacciniumvacillansPolysticumacrostichoidesA.saccharumCardamineangustataCarpinuscarolinianaDryopterismarginalisPolygonumbistortaFraxinusamericanaFagusgrandifoliaThelypterisnoveborcensisStellariamediaUvulariasessilifoliaCimicifugaracemosaMagnoliaacuminataMedeolavirginianaQuercusrubraSassafrasalbidumStellariapubera16.915.410.210.19.88.58.16.72.82.62.22.11.10.70.60.60.50.30.20.20.20.20.20.2Violaspp.LaporteacanadensisSmilaxrotundifoliaRubusspp.AcerpensylvanicumPrunusserotinaA.saccharumVacciniumvacillansPolysticumacrostichoidesDryopterismarginalisCastaneadentataDiospyrosvirginianaArisaematriphyllumQuercusrubraFraxinusamericanaMagnoliaacuminataTovariavirginianaFagusgrandifoliaCaulophyllumthalictroidesMedeolavirginianaSassafrasalbidumCarexspp.LiriodendrontulipiferaAsterspp.BetulalentaUvulariaperfoliata28.213.210.88.27.76.54.33.93.22.91.71.61.41.21.00.70.70.60.50.30.30.30.30.10.10.1246JOURNALOFTHETORREYBOTANICALSOCIETY[VOL.133FIG.2.DetrendedcorrespondenceanalysisofspeciescompositionoftheherbaceouslayerofWS3(treat-ment—solidsymbol)andWS4(control—opensymbol)oftheFernowExperimentalForest,WV,1991–1994.Eachpointshownrepresentsacentroidofsevensampleplotsperwatershedperyear.Centroidsareconnectedtodepicttrajectoryofchangeincompositionfrom1991to1992and1992to1994.Locationinordinationspaceisshownforprominentherblayerspecies:ACPEϭAcerpensylvanicum;ACSAϭAcersaccharum;VIOLAϭViolaspp.;PRSEϭPrunusserotina;RUBUSϭRubusspp.;SAALϭSassafrasalbidum;SMROϭSmilaxrotundifolia.10%foundforanAppalachiaoakforestbyMcEvoyetal.(1980)andthe16%ofunburnedloblollypineforests(GilliamandChristensen1986),andlessthanthe31%foraNewJerseyhardwoodforest(DavisonandForman1970).Thelackofastandageeffectonherblayercov-eratFEFsupportstheconclusionofGilliam(2002)thattheherbaceouslayerappearstore-coverandreestablishrapidlyfollowingdistur-banceintheseforestecosystems.Thesignificantincreaseinspeciesdiversityfrom1991and1992to1994(Fig.1d)appearstohavearisenfromsimultaneousincreasesinnumbersofspeciesonaplotbasis(S;speciesrichness)andinequitabilityofdominanceamongherblayerspecies(J;speciesevenness)(Fig.1b,c).Itwouldbelittlebeyondpurespec-ulationtoexplainthischange,butitcouldhaveresultedfromchangesinmeteorologicalcondi-tions,particularlysoilmoistureavailabilityearlyinthegrowingseason.ArchiveddatafromtheNa-tionalWeatherServiceforTuckerCounty,WestVirginia,(http://www.ncdc.noaa.gov/oa/climate/research/cag3/wv.html)indicatelittlevariationinbothannualmeantemperaturesandmeanmonthlytemperaturesfrom1991to1994.Bycontrast,totalprecipitationwashigherin1994thanintheoth-eryears,andwasespeciallyhigherinmidtolatespring.Thelackofsignificantdifferencebe-tweenwatershedsforanygivenyearisconsis-tentwiththeresultsforsimilarmixedhardwoodstandsofWestVirginiabyFordetal.(2000),whofoundthatspeciesdiversity,richness,andevennessdidnotvarysignificantlyamongstandsofages15,25,50,andՆ85years.DCAdirectlyassessesdiversity,inthiscasebetweenwatershedsandamongyears,withspe-ciesturnover(i.e.,completechangeinspeciescomposition)occurringforevery‘‘unit’’(SD;averagestandarddeviationofspeciesturnover)(Gauch1982).Variationamongwatershed-yearcentroidswasapproximately0.4alongDCAaxis1andapproximately0.6alongDCAaxis2(Fig.2).Thus,wecanconcludethatspecies2006]GILLIAMETAL.:EFFECTSOFNITROGENDEPOSITIONONFORESTHERBS247—ynloartrnoofcs(d4eShWsretdanaw)neeneilwdtneabl)o5b0m.0ysϽdiPlo(st—netrneefmfitdaeyrltt(n3acSWfiinfgoisrteyoanlesrueoewcsanbareehMe.h4t9ro9f1)nDi(dyetlispremvasiddV,naW,,)tCse(rsosFenlnatenveem,i)rBe(psxsEenhwcoinrr,e)AF(erhetvfooc)yelnhitlnodmeh.snahtadneoM,lombr.3moye.slGbIaFneipraov248JOURNALOFTHETORREYBOTANICALSOCIETY[VOL.133FIG.4.MeanmonthlycoverofindividualspeciesoftheherbaceouslayerofWS3(treatment—solidsymbolandline)andWS4(control—opensymbol,dashedline)oftheFernowExperimentalForest,WV,sampledin1994:VIROϭViolarotundifolia;VIOLAϭpensylvanicumViolaspp.;;RUBUSPRSEϭϭPrunusRubusserotinaspp.;ACPE.ϭAcercompositiondidnotvaryappreciablybetweenwatershedsforanyyear;neitherdiditvaryap-preciablyamongyearsforeitherwatershed.Thisconclusionissupportedbythegeneralsimilari-tiesincovervaluesfordominantherblayerspe-ciesbetweenwatersheds(Tables2and3).SEASONALPATTERNS.Seasonalpatternsofherblayercoverwerebothpredictable(lowear-lyinthegrowingseason,increasingduringgrowingseason,decliningtowarddormantsea-son)andquitesimilarbetweentwowatershedsthatdiffergreatlyinstandage.Suchsimilaritiesconfirmthesignificanceofslopeaspect(vialightandmoistureregimes)inaffectingtheherblayer,ashasbeenfoundinotherstudies(e.g.,Goebeletal.1999,McCarthy2003),consider-ingthatWS3andWS4arebothofasouthernaspect(Table1).Inotherwords,thelackofdif-ferencesincoverbetweenwatershedsbothovertime(Fig.1a)andseasonally(Fig.3a),despitethatWS3was(1)clearcutin1969and(2)re-ceivesNadditions,maycomefromthestronginfluencethatslopeandaspecthaveonherblay-erdynamics.Similaritiesinseasonalpatternsforcoverandrichness(Figs.3a,b),andthesignificantrela-tionshipbetweenherblayercoverandrichness(Fig.5a)suggestthatseasonalincreasesandde-creasesinherblayercoverwerebroughtaboutbyappearanceanddisappearance,respectively,ofspeciesofdifferentphenologies(Goebeletal.1999,SmallandMcCarthy2002).Forexample,theMaytoJuneincreaseofanaverageϳ6%(Fig.3a)waslargelytheresultofanincreaseofanaverageof3speciesperplot(Fig.3b),ratherthansimplyincreasedgrowthofthespeciesal-readypresentontheplots,confirmingconclu-sionsofGilliam(2002).Casadoetal.(2004)foundasignificantlinearandpositiverelationshipbetweencoverandrichnessofherbaceousspeciesinMediterra-nean-typeshrublandecosystemsoftheIberianPeninsula.Theydidnotexaminerelationshipsbetweendiversityandcover,aswedid,butcon-cludedthatrichness—coverrelationshipscanbescale-dependent(Casadoetal.2004).Similari-tiesbetweenrichnessversuscoveranddiversityversuscoverinourdata(Figures5aand5b),suggestthatthenumberofspecies(richness),ratherthantheequitabilityofspecies(evenness),determinesspeciesdiversityfortheherblayeratFEF.Certainly,speciesevennessdisplayedminimalseasonalchangeintheherblayerofbothwatersheds(Fig.3d),incontrasttopro-2006]GILLIAMETAL.:EFFECTSOFNITROGENDEPOSITIONONFORESTHERBS249FIG.5.Relationshipbetweenspeciesrichness(A)andspeciesdiversity(B)versuscoverfortheherbaceouslayerofWS3(treatment—solidsymbolandline)andWS4(control—opensymbol,dashedline)oftheFernowExperimentalForest,WV,sampledin1994.Eachpointrepresentasinglesampleplot(sevenplotsperwatershedpermonth)forallmonthscombined.Forspeciesrichness(S)versuscover(C):Sϭ6.47ϩ0.60C,r2ϭ0.58,PϽ0.0001forWS3;Sϭ7.12ϩ0.59C,r2ϭ0.53,PϽ0.0001forWS4.Forspeciesdiversity(HЈ)versuscover:HЈϭ1.47ϩ0.051C,r2ϭ0.68,PϽ0.0001forWS3;HЈϭ1.68ϩ0.037C,r2ϭ0.29,PϽ0.0002forWS4.250JOURNALOFTHETORREYBOTANICALSOCIETY[VOL.133FIG.6.DetrendedcorrespondenceanalysisofspeciescompositionoftheherbaceouslayerofWS3(treat-ment—closedsymbols)andWS4(control—opensymbols)oftheFernowExperimentalForest,WV,1994.Eachpointshownrepresentsacentroidofsevensampleplotsperwatershedpermonth.CentroidsareconnectedtodepicttrajectoriesofchangeinspeciescompositionoftheherblayerfromMaytoOctober.nouncedchangesinseasonalpatternsforspeciesrichnessanddiversity(Fig.3bandc).Ourfind-ingsregardingtheserelationshipsmayhavere-sulted,atleastinpart,fromthenumberofherblayerspeciescharacteristicformixedmesophyt-icforests,suchasthosefoundatFEF.Forex-ample,StirlingandWilsey(2001)foundthatrichnessanddiversitywereweaklycorrelatedforsiteswithlownumbersofspecies(Յ10spe-cies),notcorrelatedforsiteswithhighnumbersofspecies(Ն100),buthighlycorrelatedforsiteswithintermediatenumbersofspecies(Ͼ10andϽ100).Seasonalpatternsforindividualspecies/spe-ciesgroupswerealsogenerallysimilarbetweentreatmentandcontrolwatersheds.Thiswases-peciallythecaseforViolarotundifoliaandtheViolaspp.groupasawhole(Fig.4).Thesesim-ilaritiessuggestthattheViolaspeciesfoundatFEFareadaptedtoawiderangeofenviron-mentalconditions.Indeed,studieshavedemon-stratedconsiderablevariabilitywithinandamongspeciesofViolaintermsofreproduction(Griffith1998,Culley2002)andgrowth(CurtisandKincaid1984).Somespeciesarecapableofproducingbothchasmogamousandcleistoga-mousflowers(Culley2002),andmanycombinesexualreproductionviaprolificflowerproduc-tionwithasexualreproductionviastolonsandrhizomes(Griffith1998).Thespeciesfoundinourstudy(V.blanda,V.canadensisL.,V.pap-ilionacea,V.pensylvanica,V.rotundifolia,V.sagittata,andV.sororia)areallwidelydistrib-utedinforestsoftheeasternUnitedStates,fur-thersuggestingadaptationstowidelyvaryingenvironments(GleasonandCronquist1991),in-cludingvariationinlightandsoilnutrientavail-ability(CurtisandKincaid1984,Griffith1998).RankinandTramer(2002)foundV.blandatobeaprominentcomponentoftheherblayerinhardwoodstandsboth0–5yrand65yrfollow-ingharvesting,andtohaveequallyhighcoverinbothcanopygapsandintacthardwoodcano-py.Similarly,Rubenetal.(1999),comparingherblayercomposition25and60yrpostharvestinnorthernhardwoodforests,classifiedV.ro-tundifoliaasaninsensitivespeciesinresponsetoharvest-mediateddisturbance.Theyfoundittobeofhighrelativecoverinharvestedplotsindependentofstandage(Rubenetal.1999).Alsonotableamongherblayerspeciescom-paredbetweenwatershedsisRubusspp.,which2006]GILLIAMETAL.:EFFECTSOFNITROGENDEPOSITIONONFORESTHERBS251wasgenerallyhigheronWS3thanonWS4(Fig.4).Thisdifference,althoughnotlargeoverall,mayberelatedtosuccessionalageofthetwowa-tersheds(Table1),consideringthatRubusspp.representsagroupofdisturbance-maintainedspecies(RobertsandDong1993).KochenderferandWendel(1983)foundsubstantialchangeincompositionanddominanceoftheherblayerimmediatelyfollowingharvestonWS3in1969.Rubusspp.increasedfromapproximately20%ofherbcoverinyear1followingharvestingtojustunder40%byyear5,decliningsharplybyyear10(KochenderferandWendel1983).An-nualmeancoverforRubusspp.wasapproxi-mately1–2.5%forbothWS3andWS4duringtheperiodofthisstudy(1991–1994),indicatingthatcoverhasremainedlowsincelastreportedbyKochenderferandWendel(1983).Multivariateanalysisoftheseasonalherblay-erdata(Fig.6)supportconclusionsbasedonempiricaldataoncover,richness,anddiversity(Fig.3a,b,c).Thatis,therewasgreaterchangefromMaytoJuneandSeptembertoOctoberthanduringtheperiodJunetoSeptember.How-ever,aswiththeresultsofDCAforinter-annualcomparisons(Fig.2),thegradientlengthsweregenerallyshortalongbothaxes,suggestingthattherelativeamountchangeinspeciescomposi-tionwassmallforbothwatersheds.INFLUENCEOFSTANDDEVELOPMENTANDN-DEPOSITION.Thelackofsignificantdifferencesinherblayercharacteristicsbetweenwatershedsatthebeginningofoursamplingperiod(i.e.,1991)bothsupports,notsurprisingly,conclu-sionsofpreviousstudies(e.g.,GilliamandTur-rill1993)andprovidesabaselineforfurtherevaluationofeffectsofNadditionsontheherblayerofWS3.Thislackofdifferenceisrelevanttotheon-goingdebateregardingtheresponseoftheherbaceouslayerofforestecosystemstodis-turbances(seeBattlesetal.2001,RobertsandGilliam2003,andRoberts2004forrecentre-views).Naturaldisturbances,suchcanopygapformation,increaseinfrequencyduringsecond-arysuccessioninwaysthatcaninfluenceherblayerdynamics(Goldblum1997,Schumannetal.2003).ReaderandBricker(1992)foundherblayerspeciesofadeciduousforestinsouthernOntariotobesensitivetogapsize.However,forestmanagementincludestreatmentsthatrep-resentagradientofdisturbanceintensity(Gil-liamandRoberts1995),virtuallyallofwhicharegreaterthanthatofgapformation.Themoreintensivepracticesarerelatedtoplantationfor-estry,oftencoupledwiththinningandfertiliza-tion(Thomasetal.1999,RamovsandRoberts2003).Apotentiallyimportantvariableinfluencingherblayerdynamicsislight.Althoughlightpen-etrationtotheforestfloorisgenerallyassumedtodecreaselinearlywithstandage,NeufeldandYoung(2003)demonstratedthatthisismostof-tennotthecase.Indeed,BrownandParker(1994)measuredtransmittanceoflight(aspho-tonfluxdensity;PFD)inhardwoodstandsrang-ingfrom10toϾ340yrold.Theyfoundnodif-ferencesinpercentPFDbetweenstandagesrep-resentedbyWS3andWS4usedinthisstudy,25andϳ100yr,respectively.Thus,wefeelthatlight,thoughnotmeasuredhere,playsaminimalroleindeterminingvariationinherblayerdy-namicsinthisstudy.Althoughresponseoftheherblayertohar-vestingistypicallyrapid,withanearlydomi-nanceofdisturbance-maintainedspecies(Rob-ertsandZhu2002),severalstudieshavedem-onstratedthatherblayerrecoveryfollowingdis-turbancecanoccurinalittleas20–30years(Reiners1992,OliveroandHix1998,Freder-icksonetal.1999,Fordetal.2000).HalpernandSpies(1995)foundthatchangesinherblay-erdiversitywereshort-livedfollowingclearcut-tingandslashburningofDouglas-firforestsofwesternOregonandWashington,andthatherbdiversityreturnedtopre-harvestconditionsbe-forecanopyclosure(10–20yr).Becausehard-woodstandsonWS3fallwithintheserangesoftime,itislikelythatthelackofsignificantdif-ferencesinherblayercomposition,cover,anddiversitybetweenWS4andWS3—especiallyfor1991and1992—resultedfromsimilarrapidrecoveryoftheherblayeronWS3by20yearsfollowingclearcutting.Thelackresponseoftheherbaceouslayertotheadditionof35kgNhaϪ1yrϪ1,whichhadbeencarriedoutfor6yrbythetimeofthe1994sampling,isparticularlynotableinitssharpcon-trasttoresultsofotherstudiesthatshowedmoresensitiveresponsesofherblayerspeciestoex-perimentaladditionsofN.Hurdetal.(1998)foundthatcoverofprominentherbspeciesinhardwoodforestsoftheAdirondackMountainsdeclinedsignificantlyafteronly3yroftreatmentwith(NH4)2SO4atratesof14and28kgNhaϪ1yrϪ1.Thisdeclinewasaccompaniedbyincreasesinrelativecoveroffernspecies,suggestingthatsomeofthedeclineinherbcovermayhavere-sultedfromincreasedshadingbytheferns,whichhavebeenshowntoreducemidsummer252JOURNALOFTHETORREYBOTANICALSOCIETY[VOL.133lightlevelsbyasmuchas70%(GeorgeandBazzaz2003).WorkinginHarvardForest,Rai-neyetal.(1999)reporteddeclinesindensityandbiomassofherblayerspeciesof80%andϳ90%,respectively,after7yrofadditionsofNH4NO3atratesof50and150kgNhaϪ1yrϪ1.Strengbometal.(2003)conductedasynopticsurveyof557Swedishconiferstandsandcon-cludedthatevenlowratesofenhancedNde-positioncanalterherblayerspecies.Strengbometal.(2001)foundthattheeffectsofincreasedNdepositionontheherblayerofmanagedSwedishforestscanbelong-lived.Sucheffects,includingdeclinesinericaceousspeciesandin-creasesinnitrophilousgrasses,remainedaslongas30yraftercessationoftreatmentswithNfertilizers.Wesuggestthatthelackofmeasurablere-sponseoftheherbaceouslayerto6yrofaerialapplicationsof35kgNhaϪ1yrϪ1toWS3re-sultedfromtwofactors:(1)highambientat-mosphericdepositionofNand(2)Nsaturationstatus.WetdepositionofN(NH4-NϩNO3-N)averagesϳ10kgNhaϪ1yrϪ1atFEF,isevenhigherathigherelevations,andincreasesmark-edlyduringthegrowingseason(Gilliametal.2001).Thus,theNadditiontoWS3representsalowerrelativeadditionthanitwouldatasitewithlowerambientinputsofN.Thisisconsis-tentwithresultsofHurdetal.(1998),whofoundthatplantresponsestoNadditionstohardwoodforestsoftheAdirondackMountainswasgreatestatsiteswithlowatmosphericinputsofN.SeveralwatershedsofFEF,includingWS3andWS4,arewell-documentedtobeNsaturat-ed(Peterjohnetal.1996,Gilliametal.2004).Gilliametal.(2001)foundratesofnetNmin-eralizationandnitrificationforuntreatedwater-sheds(includingWS4)tobeϳ125andϳ114kgNhaϪ1yrϪ1,respectively;forWS3thiswas135and141kgNhaϪ1yrϪ1.Thus,theaerialappli-cationofNinthisstudyisonlyanincreaseinavailableNof25%relativetotheamountthatisgeneratedbynetNmineralization.Itislikely,then,thattheNtreatmentonWS3representsacomparativelysmalladditionofanessentialnu-trientthatisnolongergrowth-limiting.FutureworkatFEFwillincluderepeatedsamplingoftheseplots,usingdatasummarizedinthispaperasbaselineforfurtheranalysis.LiteratureCitedBATTLES,J.J.,A.J.SHLISKY,R.H.BARRETT,R.C.HEALD,ANDB.H.ALLEN-DIAZ.2001.TheeffectsofforestmanagementonplantspeciesdiversityinaSierranconiferforest.For.Ecol.Manage.146:211–222.BOBBINK,R.,M.HORNUNG,ANDJ.G.M.ROELOFS.1998.Theeffectsofair-bornenitrogenpollutantsonspeciesdiversityinnaturalandsemi-naturalEu-ropeanvegetation.J.Ecol.86:717–738.BROWN,M.J.ANDG.G.PARKER.1994.Canopylighttransmittanceinachronosequenceofmixed-spe-ciesdeciduousforests.Can.J.For.Res.24:1694–1703.CASADO,M.M.,I.CASTRO,L.RAMIREZ-SANZ,M.COS-TA-TENORIO,M.DEMIGUEL,ANDF.D.PINEDA.2004.HerbaceousplantrichnessandvegetationcoverinMediterraneangrasslandsandshrubs.PlantEcol.170:83–91.CHRISTENSEN,N.L.ANDF.S.GILLIAM.2003.Temporalandspatialpatternsofherbaceouslayercommuni-tiesontheNorthCarolinaPiedmont,p.224–237.InF.S.GilliamandM.R.Roberts[eds.],TheHer-baceousLayerinForestsofEasternNorthAmeri-ca,OxfordUniversityPress,NewYork,NY.CULLEY,T.M.2002.Reproductivebiologyandde-layedselfinginViolapubescens(Violoceae),anunderstoryherbwithchasmogamousandcleistog-amousflowers.Int.J.PlantSci.163:113–122.CURTIS,W.F.ANDD.T.KINCAID.1984.Leafconduc-tanceresponsesofViolaspeciesfromsunandshadehabitats.Can.J.Bot.62:1268–1272.DAVISON,S.E.ANDR.T.T.FORMAN.1970.Herbandshrubdynamicsinamatureoakforest:athirty-yearstudy.Bull.TorreyBot.Club109:64–73.FORD,W.M.,R.H.ODOM,P.E.HALE,ANDB.R.CHAPMAN.2000.Stand-age,standcharacteristics,andlandformeffectsonunderstoryherbaceouscommunitiesinsouthernAppalachiancove-hard-woods.Biol.Conserv.93:237–246.FREDERICKSEN,T.S.,B.D.ROSS,W.HOFFMAN,M.L.MORRISON,J.BEYEA,B.N.JOHNSON,M.B.LESTER,ANDE.ROSS.1999.Short-termunderstoryplantcommunityresponsestotimber-harvestingintensityonnon-industrialforestlandsinPennsylvania.For.Ecol.Manage.116:129–139.GAISER,R.N.1951.Randomsamplingwithincircularplotsbymeansofpolarcoordinates.J.For.49:916–917.GAUCH,H.G.1982.Multivariateanalysisincommu-nityecology.CambridgeUniversityPress,Cam-bridge.GEORGE,L.O.ANDF.A.BAZZAZ.2003.Theherba-ceouslayerasafilterdeterminingspatialpatterninforesttreeregeneration,p.265–282.InF.S.Gil-liamandM.R.Roberts[eds.],TheHerbaceousLayerinForestsofEasternNorthAmerica,OxfordUniversityPress,NewYork,NY.GILLIAM,F.S.2002.Effectsofharvestingonherba-ceouslayerdiversityofacentralAppalachianhard-woodforest.For.Ecol.Manage.155:33–43.GILLIAM,F.S.ANDN.L.CHRISTENSEN.1986.Herb-layerresponsetoburninginpineflatwoodsofthelowerCoastalPlainofSouthCarolina.Bull.TorreyBot.Club113:42–45.GILLIAM,F.S.ANDM.R.ROBERTS.1995.Impactsofforestmanagementonplantdiversity.Ecol.Appl.5:911–912.GILLIAM,F.S.ANDM.R.ROBERTS.2003.Introduction:conceptualframeworkforstudiesoftheherbaceous2006]GILLIAMETAL.:EFFECTSOFNITROGENDEPOSITIONONFORESTHERBS253layer,p.3–11.InF.S.GilliamandM.R.Roberts[eds.],TheHerbaceousLayerinForestsofEasternNorthAmerica,OxfordUniversityPress,NewYork,NY.GILLIAM,F.S.ANDN.L.TURRILL.1993.HerbaceouslayercoverandbiomassinayoungversusamaturestandofacentralAppalachianhardwoodforest.Bull.TorreyBot.Club120:445–450.GILLIAM,F.S.,B.M.YURISH,ANDM.B.ADAMS.2001.Temporalandspatialvariationofnitrogentransfor-mationsinnitrogen-saturatedsoilsofaCentralAp-palachianhardwoodforest.Can.J.For.Res.31:1768–1785.GILLIAM,F.S.,M.B.ADAMS,D.A.DICK,ANDM.L.KERR.2004.EffectsofsilviculturalpracticesonsoilcarbonandnitrogeninanitrogensaturatedCentralAppalachianhardwoodforestecosystem.Environ.Manage.33:S108–S119.GILLIAM,F.S.,N.L.TURRILL,S.D.AULICK,D.K.EVANS,ANDM.B.ADAMS.1994.HerbaceouslayerandsoilresponsetoexperimentalacidificationinacentralAppalachianhardwoodforest.J.Env.Qual.23:835–844.GLEASON,H.A.ANDA.CRONQUIST.1991.ManualofVascularPlantsoftheNortheasternUnitedStatesandAdjacentCanada,2nded.NewYorkBotanicalGarden,NewYork,NY.GOEBEL,P.C.,D.M.HIX,ANDA.M.OLIVERO.1999.Seasonalground-florapatternsandsitefactorre-lationshipsofsecond-growthandold-growthsouth-facingforestecosystems,southeasternOhio,USA.Nat.AreasJ.19:12–19.GOLDBLUM,D.1997.TheeffectsoftreefallgapsonunderstoryvegetationinNewYork,USA.J.Veg.Sci.8:125–132.GRIFFITH,C.JR.1998.TheresponseofViolablandaWilld.(Violaceae)tophosphorusfertilizationandshading.J.TorreyBot.Soc.125:194–198.HALPERN,C.B.ANDT.A.SPIES.1995.PlantspeciesdiversityinnaturalandmanagedforestsofthePa-cificNorthwest.Ecol.Appl.5:913–934.HUBERTY,L.E.,K.L.GROSS,ANDC.J.MILLER.1998.Effectsofnitrogenadditiononsuccessionaldy-namicsandspeciesdiversityinMichiganold-fields.J.Ecol.86:794–803.HURD,T.M.,A.R.BRACH,ANDD.J.RAYNAL.1998.ResponseofunderstoryvegetationofAdirondackforeststonitrogenadditions.Can.J.For.Res.28:799–807.HURLBERT,S.H.1984.Pseudoreplicationandthede-signofecologicalfieldexperiments.Ecol.Monogr.54:187–211.KOCHENDERFER,J.N.ANDG.W.WENDEL.1983.Plantsuccessionandhydrologicrecoveryonadeforestedandherbicidedwatershed.For.Sci.29:545–558.MCCARTHY,B.C.2003.Theherbaceouslayerofeast-ernold-growthdeciduousforests,p.163–176.InF.S.GilliamandM.R.Roberts[eds.],TheHerba-ceousLayerinForestsofEasternNorthAmerica,OxfordUniversityPress,NewYork,NY.MCEVOY,T.J.,T.L.SHARIK,ANDD.W.SMITH.1980.VegetativestructureofanAppalachianoakforestinsouthwesternVirginia.Amer.Midl.Nat.103:96–105.MULLER,R.N.2003.Nutrientrelationsoftheherba-ceouslayerindeciduousforestecosystems,p.15–37.InF.S.GilliamandM.R.Roberts[eds.],TheHerbaceousLayerinForestsofEasternNorthAmerica,OxfordUniversityPress,NewYork,NY.NEUFELD,H.S.ANDD.R.YOUNG.2003.Ecophysiol-ogyoftheherbaceouslayerintemperatedeciduousforests,p.38–90.InF.S.GilliamandM.R.Roberts[eds.],TheHerbaceousLayerinForestsofEasternNorthAmerica,OxfordUniversityPress,NewYork,NY.OLIVERO,A.M.ANDD.M.HIX.1998.Influenceofaspectandstandageongroundfloraofsoutheast-ernOhioforestecosystems.PlantEcol.139:177–187.PETERJOHN,W.T.,M.B.ADAMS,ANDF.S.GILLIAM.1996.Symptomsofnitrogensaturationintwocen-tralAppalachianhardwoodforests.Biogeochem.35:507–522.PIELOU,E.C.1966.Themeasurementofdiversityindifferenttypesofbiologicalcollections.J.Theor.Biol.13:131–144.PREGITZER,K.S.ANDB.V.BARNES.1982.TheuseofgroundfloratoindicateedaphicfactorsinuplandecosystemsoftheMcCormickExperimentalFor-est,UpperMichigan.Can.J.For.Res.12:661–672.RAINEY,S.M.,K.J.NADELHOFFER,W.L.SILVER,ANDM.R.DOWNS.1999.Effectsofchronicnitrogenadditionsonunderstoryspeciesinaredpineplan-tation.Ecol.Appl.9:949–957.RAMOVS,B.V.ANDM.R.ROBERTS.2003.Understoryvegetationandenvironmentresponsestotillage,forestharvesting,andconiferplantationdevelop-ment.Ecol.Appl.13:1682–1700.RANKIN,W.T,ANDE.J.TRAMER.2002.UnderstorysuccessionandthegapregenerationcycleinaTsu-gacanadensisforest.Can.J.For.Res.32:16–23.READER,R.J.ANDB.D.BRICKER.1992.Responseoffivedeciduousforestherbstopartialcanopyre-movalandpatchsize.Am.Midl.Nat.127:149–157.REINERS,W.A.1992.Twentyyearsofecosystemre-organizationfollowingexperimentaldeforestationandregrowthsuppression.Ecol.Monogr.62:503–523.RICHARDSON,H.L.1938.Thenitrogencycleingrass-landsoils:withespecialreferencetotheRothamst-edParkGrassExperiment.J.Agric.Sci.28:73–121.ROBERTS,M.R.2004.Responseoftheherbaceouslay-ertodisturbanceinNorthAmericanforests.Can.J.Bot.82:1273–1283.ROBERTS,M.R.ANDH.DONG.1993.Effectsofsoilorganiclayerremovalonregenerationafterclear-cuttinganorthernhardwoodstandinNewBruns-wick.Can.J.For.Res.23:2093–2100.ROBERTS,M.R.ANDF.S.GILLIAM.2003.Responseoftheherbaceouslayertodisturbanceineasternfor-ests,p.302–320.InF.S.GilliamandM.R.Roberts[eds.],TheHerbaceousLayerinForestsofEasternNorthAmerica,OxfordUniversityPress,NewYork,NY.ROBERTS,M.R.ANDL.ZHU.2002.Earlyresponseoftheherbaceouslayertoharvestinginamixedco-niferous-deciduousforestinNewBrunswick,Can-ada.For.Ecol.Manage.155:17–31.RUBEN,J.A.,D.T.BOLGER,D.R.PEART,ANDM.P.AYRES.1999.Understoryherbassemblages25and254JOURNALOFTHETORREYBOTANICALSOCIETY[VOL.13360yearsafterclearcuttingofanorthernhardwoodforest,USA.Biol.Conserv.90:203–215.SCHUMANN,M.E.,A.S.WHITE,ANDJ.W.WITHAM.2003.Theeffectsofharvest-createdgapsonplantspeciesdiversity,composition,andabundanceinaMaineoak-pineforest.For.Ecol.Manage.176:543–561.SICCAMA,T.G.,F.H.BORMANN,ANDG.E.LIKENS.1970.TheHubbardBrookecosystemstudy:pro-ductivity,nutrientsandphytosociologyoftheher-baceouslayer.Ecol.Monogr.40:389–402.SMALL,C.J.ANDB.C.MCCARTHY.2002.Theinflu-enceofspatialandtemporalvariabilityonunder-storydiversityinaneasterndeciduousforest.PlantEcol.164:37–48.STIRLING,G.ANDB.WILSEY.2001.Empiricalrelation-shipsbetweenspeciesrichness,evenness,andpro-portionaldiversity.Amer.Nat.158:286–299.STRENGBOM,J.,A.NORDIN,T.NA¨SHOLM,ANDL.ER-ICSON.2001.Slowrecoveryofborealforesteco-systemfollowingdecreasednitrogeninput.Func.Ecol.15:451–457.STRENGBOM,J.,M.WALHEIM,T.NA¨SHOLM,ANDL.ER-ICSON.2003.RegionaldifferencesinoccurrencesofunderstoreyforestspeciesreflectdifferencesinNdeposition.Ambio32:91–97.THOMAS,S.C.,C.B.HALPERN,D.A.FALK,D.A.LIGUORI,ANDK.A.AUSTIN.1999.Plantdiversityinmanagedforests:understoryresponsestothin-ningandfertilization.Ecol.Appl.9:864–879.TILMAN,D.1987.Secondarysuccessionandthepat-ternsofplantdominancealongexperimentalnitro-gengradients.Ecol.Monogr.57:189–214.TILMAN,D.1989.Ecologicalexperimentation:strengthsandconceptualproblems,p.136–157.InG.E.Likens[ed.],Long-TermStudiesinEcology:ApproachesandAlternatives,Springer-Verlag,NewYork,NY.ZAR,J.H.1996.Biostatisticalanalysis,3rded.Prentice-Hall,EnglewoodCliffs,NJ.662p.
