ContentslistsavailableatScienceDirect
Microbiological
Research
jo u r n al ho me p a g e :w w w . e l s e v i e r . c o m / l o c a t e / m i c r e s
Long
lasting
effects
of
the
conversion
from
natural
forest
to
poplar
plantation
on
soil
microbial
communities
Francesco
Vitali,
Giorgio
Mastromei,
Giuliana
Senatore,
Cesarea
Caroppo,
Enrico
Casalone
∗DepartmentofBiology,UniversityofFlorence,ViaMadonnadelPiano6,SestoFiorentino,50019Florence,Italy
a
r
t
i
c
l
e
i
n
f
o
Articlehistory: Received3April2015
Receivedinrevisedform8September2015 Accepted10October2015
Availableonline20October2015 Keywords: T-RFLP Forestsoil Land-use Bacterialcommunities Fungalcommunities
a
b
s
t
r
a
c
t
Inthisstudy,weevaluatethelong-lastingeffectsonsoilmicrobialcommunitiesofachangewithina
singleland-usecategory,specificallytheconversionfromnaturalforesttoforestplantation.Tominimize
theeffectsofimpactsotherthanland-use(i.e.,climaticandanthropogenic),wechosethreesiteswithin
aNaturalPark,withhomogeneousorographicandsoiltexturecharacteristics.Wecomparedmicrobial
diversityinatotalof156soilsamplesfromtwonaturalmixedforestsandasimilarforestconvertedto
poplarplantationaboutthirtyyearsago.Thediversityandstructureofbacterialandfungalcommunities
wereinvestigatedbyterminalrestrictionfragmentslengthpolymorphism(T-RFLP)analysisofthe
16S-rRNAgeneandtheITS-rDNAregions,respectively.Bacterialandfungalcommunitiesfromtheforest
plantation,comparedtothosefromnaturalforestsoils,showeddifferentcommunitystructureandlower
␣-diversityvalues,consistentlywiththesignificantlyhigherpHvaluesandlowerorganicmattercontent
ofthosesoils.-diversityvalues,thenumberofmeasuredandestimateddominantOTUs,andtheir
distributionamongthethreesitesshowedthatmicrobialcommunitiesfromthetwonaturalforestswere
muchmoresimilartoeachotherthantheyweretocommunitiesfromthepoplarplantation,suggestingan
effectoftheforestconversiononthecompositionanddiversityofsoilmicrobialcommunities.␣-diversity
incultivatedforestsoilshadnarrowertemporalfluctuationsthaninnaturalforestsoils,suggestinghigher
temporalstabilityofmicrobialcommunities.Overall,wedemonstratedthattheconversionfromnatural
foresttoforestplantationalteredsoilmicrobialcommunities,changingtheirstructure,loweringtheir
diversity,andcausingaspatialandtemporalhomogenization.
©2015ElsevierGmbH.Allrightsreserved.
1. Introduction
Soilmicrobial communitiesare akey componentof the for-estecosystem;theyareinvolved infundamentalprocesses,like decomposition and nutrient cycling, and perform a link role betweenplantsandecosystemfunctions(Zaketal.,2003;Vander Heijdenetal.,2008;BardgettandWardle2010).Theinfluenceof thetreetypeonmicrobialcommunitystructureandfunctionwas supportedbya numberof differentreports (Myersetal.,2001; Hackletal.,2004;Bastiasetal.,2007;Schweitzeretal.,2008;Berg andSmalla2009;Wubetetal.,2012;Wangetal.,2013).Inturn, belowground-livingmicroorganismshavebeendemonstratedto influence, directly or indirectly, the productivity,diversity and compositionofplantcommunities(VanderHeijdenetal.,2008; Waggetal.,2011).
∗ Correspondingauthor.Fax:+390554574735. E-mailaddress:enrico.casalone@unifi.it(E.Casalone).
With200millionha(10millionofwhichinItaly),forests repre-sentthemajornaturalvegetationcoverinWestEurope(31.5%of thelandarea,5%oftheworld’sforests);aquarteroftheseforests areofthemixedtype.Despiteapositivereforestationtrend,risks forEuropeanforesthealthandvitalityseemontheincrease,mainly duetoanthropicimpact(MCPFE,2007).Despitethefundamental issueofnatureconservationandbiodiversitypreservationofforest sitesandtherecognizedimportantrolethatmicrobial communi-tiesplayinthefunctioningofforestryecosystems,veryfewstudies haveinvestigatedtheeffectsonmicrobialdiversityofforest man-agementandforestconversioninapreservednaturalenvironment (Nackeetal.,2011).Recently,thedeforestationofAmazon rain-forests(DaCJesusetal.,2009;Rodriguesetal.,2013)orofpristine forestsinthePampabiome(Suleimanetal.,2013)toobtainpasture siteshasbeenreportedtoaltermicrobialdiversityandcommunity structureofsoilmicroorganisms.Farlessattentionhasbeenspent ontheeffectthatachangewithinasingleland-usecategory,such astheconversionfromnaturalforesttoplantedforest,haveon soilmicrobialcommunities;withtheresultthatthisspecialcase http://dx.doi.org/10.1016/j.micres.2015.10.002
offorestland-usechangeisstillpoorlyunderstood(Nackeetal., 2011),especiallywithrespecttofungi.Veryrecently,twostudies reportedthatconversionfromnaturalforesttoplantationforest,in thePampabiome(Lupatinietal.,2013)andinSoutheastAsian trop-icalforest(McGuireetal.,2014),alterthebelow-groundecosystem, andultimatelyaffectthemicrobialcommunitiesresidentinthesoil. Theconversionofnaturalforeststooilpalmplantationisreported toaffectthecompositionofbothbacterialandfungalcommunities (Lee-Cruzetal.,2013;Kerfahietal.,2014).Asplantationforests intheworldaccountedforaround7%ofglobalforestcover,and areprojectedtocontinuetoincreaseintheforeseeablefuture;a wealthofsoilmicrobialdiversity,aswellastheenormousandstill untappedpoolofbiologicalresourcestheyconstitute,couldbeat risk;especiallyinhotspotofplantdiversity,suchasthe Mediter-raneanarea(Myersetal.,2001).
Themainobjectiveofthepresentworkwastostudytheeffects thata changewithina singlecategory ofsoil land-use, specifi-callya long-term(about 30 years old)conversionfrom natural mixedforesttopoplarplantation,hadonsoilinhabiting micro-bialcommunities.Duetothelongtimesincetheconversiontook place,we expecttodetectlong lasting, andalmost permanent, effects.Withthisaim,thestructure,and␣-diversity(insideeach studysite)and-diversity(betweendifferentstudysites)of bac-terialandfungalcommunitiesfromthesoilofaconvertedpoplar plantation and two natural forests were compared. The three forestsites werelocatedwithin a natural park(Migliarino–San Rossore–MassaciuccoliRegionalPark,Tuscany,Italy),notfarfrom each other, in a climatic,orographic and soil texture homoge-neouslandscape.Inthiswaynaturalandanthropiceffects,other thanthoselinkedtotheconversion,wereminimized.To charac-terizemicrobialcommunities,weusedacost-effectiveandrapid techniqueofDNAfingerprinting,theterminalrestrictionfragment lengthpolymorphism(T-RFLP) analysis.T-RFLPhasbeenwidely usedforstudyingmicrobialcommunitystructure anddynamics (Osbornetal.,2000),andhasbeenrecentlyre-evaluatedforthe estimation of microbial community diversity (Van Dorst et al., 2014). Secondarily,therelationships that microbial community structureanddiversityhadwiththephysicochemicalproperties ofthesoilsand withseasonalitywasanalyzedtorecognizethe contributionthatotherfactorshadbeyondland-useconversion.
2. Materialsandmethods
2.1. Sitedescriptionandsoilsampling
The study area is located inside the Migliarino–San Rossore–Massaciuccoli Regional Park (latitude 43◦35–43◦51,
longitude 10◦15–10◦22, approximately; mean altitude 4m), which ranges along the Tyrrhenian Sea between the cities of ViareggioandLivorno(Tuscany,Italy),andbelongstothe Mediter-raneanclimatetype.Thestudywasconductedinthreedifferent field sites withinthe Park (Table1).Sites 1 (anapproximately 7000m2 large meso-hygrophilic/hygrophilic deciduous forest)
and site 3 (an approximately 118,000m2 meso-hygrophilic
deciduousforest)wereoldnaturalmixed-deciduousforeststhat untilmid-1970’sweresubjectedtocontrolledlogging,afterthat theywereleftundisturbed.Site 1and3 are8300mapart each other,in thenorthandsouthofthePark,respectively.Site2(a 28,000m2 plantedforest)wasa mature15 years old (in2010)
poplar plantation never subject to any agriculturalpractice; it is325mapartfromsite1towardnorth.Site 2wasoriginallya naturalhygrophilicmixed-deciduousforest,converted topoplar plantation;aerialphotosofthissite(includingalsosite1)placed thefirstestablishmentofapoplarplantationbetween1978and 1982(Fig.1).Plantsinthethreesiteswereidentifiedand georef-erencedIndividual soilsamples wereseasonally collectedfrom georeferenced trees, using a bulb planter (10cm wide×15cm depth),atabout20cmfromthetrunk;foreachtree,samplesat differentseasonswerecollectedatspotsnotdisturbedbyprevious sampling activities.A total of 156 soil samples were collected fromsoilsassociatedwithdifferenttrees:sevennaturalpoplars (5Populusalbaand2Populuscanescens)insite1,fromAutumn 2010toSummer2012;fourcultivatedpoplars(hybridTriploclone Populus nigra×Populus deltoids) in site 2, from Spring 2011to Summer2012;andelevenmaples(Acercampestre)insite3,from Winter2011toSummer2012.Individualsoilsampleswereplaced separatelyinsterileplasticbagsandimmediatelystoredat5◦C; thesamedaythesampleswerebroughttothelab.Eachindividual samplewas sequentiallysieved through5mmand 2mm pores sizestainlesssteelsieves.Thesievedsoilfromeachsamplewas splitintofouraliquotsthatwerestoredat4◦C,fortotalmicrobial counts,moisturecontent,pHdeterminationandlossonignition measure;a furtheraliquot was stored at −80◦C for molecular
analysis.
Rainandairtemperaturevalues,measuredbytwo meteorolog-icalstationsplacedaspartoftheLIFE08NAT/IT/00342-DEMETRA project,havebeenaveragedovera7daysperiodbeforethe sam-plingdates.
2.2. Analysisofsoilchemistry
Sampling sites were classified as follow: site 1, Humic Eutrudepts,coarsesand,mixed,thermic;site2,coarsesand,mixed, thermic;site3,FragicHapludalfs,fine,mixed,thermic.Dataforsite
Fig.1.Aerialphotography,executedbyRossi–Brescia(Italy),ofSite2in(a)1978(folderc0117,swipe29,frame412),and(b)1982(folderc0233,swipe3a,frame335). ImagesCourtesyoftheGeneralCartographicarchiveoftheTuscanyRegion.
Table1
Characteristicsofthesamplingsites.
Site Area(m2) Forestmanagement Meandistancesbetweenarea(m) Meandistancesbetweentrees(m) Sampledtreesa No.ofsamples Samplingdensityb
1 7,000 Natural 1–2(325) 51 PN 7 23.3%
2 27,745 Cultivated 2–3(8,250) 7 PC 4 N.d.
3 118,000 Natural 3–1(8,310) 217 A 11 4.12%
a PN—Populusalba;PC—Populusnigra×Populusdeltoids;A—Acercampestris.
bPercentageofsampledtreesonthetotalnumberoftreefromthesamespeciesinthatsite.
1and3werefrommapunitsdated2002;site2wasnotinthese mapsandwasdefined,atthetimeofthesamplingcampaign,for soiltextureonly.
Forphysicochemicalanalysis,allthesoilsamplesfromeachsite ineachseasonwerepooledtomakeasinglecompositesample;a totalof21compositesoilsampleswereanalysed(eightfromsite1, sixfromsite2,andsevenfromsite3).Gravimetricwatercontent, organicmatter(OM)content,andpHweredeterminedon com-positesoilsamples.Gravimetricwatercontentwasdeterminedas weightlossafterover-dryingthefreshlysievedsoilat105◦Cfor 24h.WeightpercentageOMwasdeterminedonover-driedsoils bythelossonignition(LOI)procedureinamufflefurnaceat550◦C for24h(Heirietal.,2001).pHwasmeasuredonsievedair-dried soilsamplesmixedtodeionizedwateratratioof1:2.5(w/v);the mixturewasshakentoformaslurryandleftundisturbedfor15min priortowithdraw70Lofsupernatanttobeanalysedbya micro-electrode (Ross® pHmicroelectrode,ThermoScientific;Beverly, MA,USA).
2.3. Viablecounts
To determinate the number of viable bacteria and fungi in thesoil, five grams of each individualsieved soilsample were placedinaseparatesterileplasticbagwith50mLofsaline solu-tionand processedby Stomacher® 400Circulator(Seward, UK) for3minat260rpmtoensurethedetachmentofmicroorganisms fromsoilparticles.After15minsedimentation,the suspension-supernatant from each soil was serially diluted and plated in triplicate (0.1mL) on Soil ExtractAgar Medium with cyclohex-imide(1g/mL)andonRoseBengalChloramphenicolAgarBase (Oxoid,Basingstoke,England)withchloramphenicol(0.1g/mL), forselectivegrowthofbacteriaandfungi,respectively.Plateswere incubatedforthreedaysat37◦Cand30◦C,respectively;onlyplates containingbetween30and300coloniesweretakenin consider-ationtocalculateviabilityasColonyFormingUnits(CFU/gofdry soil).Soil-extractagarmediumcontained100ml/Lsoilextract,1g/L glucose,1g/Lyeastextractand15g/Lagar.Thesoilextractwas pre-paredbymixing500gofsoilwith1Lofwater;themixturewas autoclavedat121◦Cfor1h.Thesterilizedsoilmixturewasfiltered throughgauzeandthencentrifugedfor15minat6000rpm.The supernatantwasfiltratedthrougha0.2-mmembranefilterand thepHwascorrectedto7.0.
2.4. SoilDNAextractionandpurification
TotalDNAwasextractedfrom250mgaliquotsofeachofthe156 individualsoilsamplesusingNucleoSpinSoilkit(MachereyNagel, Düren,Germany),withLysisBufferSL2and150Lof Enhancer SX.,ExtractedDNAswerefurtherpurifiedbyPowerClean® DNA Clean-UpKit(MachereyNagel,Düren,Germany).
2.5. PCRamplificationofgenomicDNAfromsoil
TotalDNAextractedfromeachofthe156individualsoil sam-ples was used for PCR amplification of the bacterial 16SrDNA gene and theITS1 and ITS2 internally transcribed spacers(ITS
region)offungi.PCRreactionswereperformedwith10ngof tem-plateDNA in a final reactionvolumeof 25LwithDreamTaq Buffer(containing20mMMgCl2—ThermoFisherScientificGmbH,
Karlsruhe, Germany), 0.2mM ofeach dNTP(EuroClone, Milano, Italy),1Mofeachprimer,1UofDreamTaqpolymerase(Thermo FisherScientificGmbH,Karlsruhe,Germany).Bacterial16SrDNA genes wereamplifiedusing thefluorescently labelled27F-FAM (5-[6FAM]- AGAGTTTGATCCTGGCTCAG-3) and the 1525R (5 -AAGGAGGTGWTCCARCC-3)primers(Osborneetal.,2005);fungi ITSregionswereamplifiedwiththefluorescentlylabelled ITS1f-(5-[6FAM]-CTTGGTCATTTAGAGGAAGTAA-3)andITS4r(5 -[PET]-TCCTCCGCTTATTGATATGC-3)primers(Gardesand Bruns1993). Bacterialamplificationreactionconsistedofaninitial denatura-tionstepat95◦Cfor5min,5cyclesat95◦Cfor30s,60◦Cfor30s, 72◦Cfor2min,then5cyclesat95◦Cfor30s,55◦Cfor30s,72◦C for2min,followedby25cyclesat95◦Cfor30s,52◦Cfor30s,72◦C for2minandafinalextensionat72◦Cfor10min.Fungal amplifi-cationreactionconsistedofaninitialdenaturationstepat94◦Cfor 5min;34cyclesat94◦Cfor1min,52◦Cfor1min,72◦Cfor2min andafinalextensionat72◦Cfor10min.16SrDNA(approximately 1500bp)andITSregion(variablesizes)PCRampliconswere puri-fiedbyWizard®SVGelandPCRClean-UpKit(Promega,Madison, Wisconsis,USA)fromgelafterelectrophoresisona0.8%agarose gelordirectlyfromtheamplificationreaction,respectively. 2.6. Geneticprofilingofmicrobialcommunities
Thediversityofthedominantmembersofbacterialand fun-galdomainswascharacterizedbygeneticprofilingusingterminal restrictionfragmentlengthpolymorphism(T-RFLP).16SrRNAPCR amplicons(0.6–1g)weredigestedwith20unitofRsaIrestriction enzyme(ThermoFisherScientific,GmbH,Karlsruhe,Germany)ina finalvolumeof20L,at37◦Cfor4h;thedigestionwasterminated byheatingat80◦Cfor20min.ITSregionPCRamplicons(0.1g) were digested with 6 unit of HinfI restriction enzyme (Roche, Basilea,Switzerland)inafinalvolumeof20L,at37◦Cfor4h;the digestionwasterminatedbyheatingat65◦Cfor20min.Thelength ofthefluorescentlylabelledterminalrestrictionfragments(T-RFs) wasdeterminedwithanAppliedBiosystems®3500SeriesGenetic AnalyzerautomatedsequencerusingLIZ500(AppliedBiosystems, FosterCity,California,USA)sizestandardasadimensional stan-dard. T-RFsprofiles were analysed with GeneMapper software (AppliedBiosystems,FosterCity,California,USA).Fixedthresholds of50and100RFUwereusedtoremove baselinenoise forblue channel (6FAM-labeled primers) and red channel (PET-labelled primer),respectively.AlignmentofT-RFspeakswasautomatically performedbythesoftwareandmanuallychecked.
2.7. Dataanalysisandstatisticalmethods
All data elaborationsand analysis werecarried out using R Statisticalsoftwareversion2.15.1(RCoreTeam,2013)withthe packageVegan(Oksanenetal.,2013)andggplot2(Wickham,2011). FortheanalysesoftheT-RFLPprofilesofbacterialandfungal communitiesinthesoilsamples,onlyT-RFpeakswithheightabove the fixed threshold were considered. Biodiversity analysis was
Fig.2.Physicochemicalcharacterizationandmicrobialabundanceofsoilsfromthethreestudysites.(a)pH(b)relativeorganicmatter(OM)content(c)relativehumidity(d) fungalviablecounts(e)bacterialviablecounts.Dataina–caremeasuredonpooledsoil,whiledataindandearemeanseasonalvalues.Boxesrepresentinterquartilerange (IQR),medianvaluesareindicatedbytheblacklines,andpointrepresentoutliers.SignificanceofT-Testbetweensitesisreportedundereachplot:**p<0.01;*p<0.05;NS notsignificant.
performedonnormalizedT-RFLPprofiles.Normalizationwasdone bymeasuringtherelativeabundanceofeachindividualT-RFina profileanddividingitspeakareabythetotalpeakareaoftheprofile. Theoverallbacterialandfungal␣-diversitywasevaluated mea-suringphylotyperichness,S=numberofT-RFs;Shannon–Wiener index,H=−
pilogepi(Shannon,1948),Simpson’sindexofdiver-sity,D=1−
pi2(Simpson,1949),wherepiistheproportionof
speciesi;andShannonevennessindex,E=H/logS,thatisequality ofphylotypeabundanceinacommunity.-diversitywas calcu-latedusingBray–Curtisdissimilarityindex,BCij=2Cij/(Si+Sj)where Cijisthesmallervalueofspeciessharedbetweensiteiandj,while CiandCjaretotalspeciesnumberinsampleiandsamplej respec-tively.TheindexwascalculatedbycomparingcumulativeT-RFLP profilesobtainedbycalculatingthemeanareaofeachT-RFinthe T-RFLPprofilesofallthesamplesineachsite;thecumulative pro-fileswerethennormalized(seeabove)andT-RFswitharelative area<0.01 wereeliminated.Toestimatetheeffects ofland-use on-diversitywithineach sitewe usedthesameprocedureas inLee-Cruzet al.(2013).Firstlywe measure-diversity asthe Bray-CurtisdistancebetweensingleT-RFLPprofiles,andthenthe betadisperfunctionintheVeganpackage wasusedtotest ifthe multivariatedispersionof-diversity(measuredasthedistance fromgroupcentroid)wasstatisticallydifferentamongdifferent landuses,using999permutations.
To estimate true phylotype richness in each site, individ-ual T-RFLP profiles were further inspected with Chao1 index, Chao=S+a12/2a2 where a1 and a2 are the number of species
occurringonlyinoneoronlyintwosites(Chao,1984),and phylo-types(T-RFs)accumulationcurves(Hughesetal.,2001).Ordination analysisofT-RFLPprofilestookintoaccountonlyqualitative infor-mationaboutpresence/absence of T-RFswithheightabovethe fixedheightthreshold.Adissimilaritymatrixwascomputedusing Sørensenindexonthepresence/absencematrixandusedasinput forordinationanalysisofmicrobialcommunitieswithUnweighted
Pair Group Method with Arithmetic mean (UPGMA) clustering andwithNon-parametricMulti-DimensionalScaling(nMDS)with 100randomstarts.Environmentalvariables(cumulativerainand airtemperature), soilchemistry variables(OM,pH,and relative humidity),andmicrobialabundancedata(log10ofviablecounts)
werefittedontothenMDSordinationusingenvfitfunctioninVegan package.Thesignificanceoffittedvectorswastestedusing999 per-mutations.TheAnalysisofSimilarities(ANOSIM)wasusedtotest spatialandseasonalvariabilityofthesoilmicrobialcommunities (Reesetal.,2004).
3. Resultsanddiscussion
3.1. Physicochemicalpropertiesandmicrobialabundanceinsoils acrossland-use
Resultsofphysicochemicalcharacterizationofcomposite sea-sonalpoolsofsoilsfromthethreesitesarereportedinFig.2.Soilsin site2showedthehighestpHvaluesandthelowestorganicmatter contentandhumidityvalues.Differencesamongsitesweretested withtwo-sampleT-test(Fig.2).pHwastheonlyparameterthat showedstatisticallysignificantdifferencesinallcomparisons(Site 1vs.Site2:t=6.623,p=<0.001;Site1vs.Site3:t=2.861,p=0.017; Site2vs.Site3:t=7.336,p=<0.001),whereasdifferencesinOM contentwere(highly)significantonlywhencomparingsite2with theothertwosites(Site1vs.Site2:t=6.421,p=<0.001;Site1vs. Site3:t=−1.804,p=0.095;Site2vs.Site3:t=−8.812,p=<0.001), andsoilrelativehumiditywassignificantlydifferentonlybetween site2and1(Site1vs.Site2:t=3.019,p=0.011;Site1vs.Site3: t=1.393,p=0.187;Site2vs.Site3:t=−1.838,p=0.095).Overall, compositesoilfromthepoplarplantationinsite2differedfrom naturalforestsoils(site1and3)morethanthelatterdifferedfrom eachother.
Fig.3. ␣-diversityanalysisofbacterialandfungalcommunitiesfromsoilsofthethreesites.LettersindicatesresultsofPosthocTukey’stest.
Microbialabundanceresults,evaluatedbyviablecountsin indi-vidualsoilsamples fromthethree sites,are reportedin Fig.2. Thenumberof cultivablebacteriawasalwaystwo-threeorders ofmagnitudegreaterthancultivablefungi.Whilebacterialviable countsweresimilarinthethreesites,fungalviablecountsinsite 1 and 3, were similar to each other, but lower than in site 2, withsite2and3showingastatisticallysignificantdifference(Site 2 vs.Site 3:t=2.342, p=0.047). Viablecounts in site 2 always showedthegreatestvariability.Toinvestigatetheoriginofthis variability,thedependencyofthecultivablecomponentof micro-bialcommunitieswithtemporalandspatialfactorswastestedwith Kruskal–Wallistest(TableS1).Thetestindicatedthat,withinthe samesite,quantitativefluctuationsofculturablecomponentsof microbialcommunitiesineachofthethreesiteswerenotlinked tospatialfactors(soilsamplesassociatedtodifferentindividual trees)but,withtheexceptionofbacterialcommunitiesinsite3,to temporalfactors(soilsamplesatdifferentseasons).
Overall,theabovereportedresultssuggestedthatthe conver-sionfromnaturalforesttoplantedforestwasassociatedtolong lasting modifications of the physico-chemicalcharacteristics of soil:poplarplantationsoilsweremorealkalineandwithreduced organicmattercontent.SoilpH,inparticular,isaparameterthat correlatesandintegrateslotsofsoilproperties,andrepresentsa well-knownfactorinfluencingcompositionanddiversityof micro-bialcommunities(Lauberetal.,2009;Rousketal.,2010).Moreover, theconversionfromnaturaltoplantedforestinfluenced,directly
orthroughthemodificationofsoilphysico-chemicalproperties, thecultivablecomponentofmicrobialcommunities,determining anincreaseincultivablefungiandahigherseasonalinstabilityof bothbacteriaandfungiinthepoplarplantation(site2)compared tonaturalforests(site1and3).
3.2. Microbialcommunitydiversityanalysis
Microbialcommunitiesdiversityandstructurewereanalyzed usingT-RFLPofamplified16SrDNAfromsoilsamples.Evenif phy-lotypesderivedfromT-RFLPmaynotbeequivalenttospecies,they doprovideabasistoestimate␣anddiversity,andcommunity structureofsoilmicrobialpopulationsthroughspaceandtime(Van Dorstetal.,2014).
3.2.1. ˛-Diversityanalysis
Results of ␣-diversity analysis of bacterial and fun-gal communities in the three sites are reported in Fig. 3. Bacterial communities from soils of the cultivated poplar in site 2 had lower ␣-diversity values (richness, S=18; Shannon Index, H=2.31)compared tosites 1(S=22;H=2.51) andsite3(S=21;H=2.39),whereasEvennessvaluesweremore similar (E=0.82, E=0.80, and E=0.79 in site 1–3, respectively). ANOVA analysisshowed that among-sites differences in bacte-rial ␣-diversity were always significant(Richness F(2,153)=4.49,
Fig.4. SpeciesaccumulationcurvesofbacterialcommunitiesinsoilsfromSite1–3.Theerrorbarsare95%confidenceinterval.HorizontallinesrepresentChao1richness estimatorvalues.
p=0.008; and Simpson F(2,153)=4.87, p=0.005). Similarly to
bacteria,fungalcommunities fromsoilsofthecultivatedpoplar insite 2 had lower ␣-diversityvalues (richness, S=9;Shannon Index,H=1.91)comparedtosites1(S=12;H=2.12)andsite3 (S=11; H=2.06), whereas Evenness values were more similar (E=0.86,E=0.86,andE=0.87insite1–3,respectively);however, onlyrichnessresultedsignificantlydifferentamongsites(Richness F(2,153)=4.00,p=0.020;EvennessF(2,153)=0.09,p=0.914;Shannon
F(2,153)=1.73, p=0.180;andSimpsonF(2,153)=0.18, p=0.835).The
relative high values of Evenness indicated equal contributions of dominantphylotypesof bacteria and fungito ␣-diversity in thethree sites. Whenanalyzed bya Post hoc Tukey’stest, the abovereporteddifferencesbetweensite2andsites1and3were particularlyevidentwithregardtobacterialrichnessandShannon index,andwerealsopresentinfungirichness(lettersinFig.3). Overall,thesedatahighlightedaneffectofland-useconversionin thesenseofadeclineinbacteriaandfungicommunityrichness. Similarreductions offungal richness wereobserved in tropical forestconverted tooilpalmplantationinBorneo(Kerfahietal., 2014)andindeadwoodsamplesofforestconvertedfromnative deciduous to coniferous species in a German study (Purahong
etal.,2014).Thislaststudyalsohighlightedthatthemodifications infungalcommunitystructureanddiversitymaybedependenton whichconiferousspecieswasintroduced.Figs.4and5show rich-nessestimationforbacterialandfungalcommunities,respectively. Richnessestimation in thenatural forestsin sites1 and 3 was almostidentical(around100bacterialand150fungalphylotypes) andhighercomparedtosite2(around50bacterialand100fungal phylotypes).Thesedifferencesinrichnessestimationdidnotseem tobeduetoundersamplingproblemsinsite2;infact,despitethe lowernumberofsamplesanalyzed,bacterialpopulationsinsite 2weretheonlyonesapproachingtheChaoIindexvalue(which representanestimationoftruerichness).Moreover,atasample sizeof24(themaximumforsite2),theerrorbars(representing 2timesthestandarddeviation)ofthephylotype(T-RF) accumu-lationcurvesinsite2neveroverlappedwiththoseofsite1and 3.This,althoughnotbeingtheresultofaproperstatisticaltest, wasa goodindicationof significantdifferencesinthemicrobial richnessbetweenthecultivatedforestin site2 andthenatural forests.Theestimatednumbersoffungalphylotypeswasalways higherthanbacterial,aresultthatapparentlyconflictedwiththe higherrichnessofbacteriacomparedtofungi(Figs.2and3),but
Fig.5.SpeciesaccumulationcurvesoffungalcommunitiesinsoilsfromSite1–3.Theerrorbarsare95%confidenceinterval.HorizontallinesrepresentChao1richness estimatorvalues.
Table2
Comparisonofsimilarity(ANOSIM)resultsintestingthe“site”groupingfactoron cumulativeandindividualT-RFLPprofilesdata.Cumulativedatawereaveragedsite profiles. T-RFLPprofiles ANOSIM Bacteria Fungi R P R P Individual 0.07 ** 0.45 ** Cumulative 0.53 ** 0.92 ** Significance:**p<0.01.
thatcouldbeexplainedbyhigherwithin-sitediversityoffungal communities(seealsoANOSIManalysisinTable2andbelow).
Thesite-distributionofT-RFs,correspondingtodominant (T-RFswitharelativeheight>0.01)bacterialandfungalphylotypes (OTUs),isshowninFig.6.ThenumberofdominantOTUsinsite2 waslowerthaninsite1and3:41,48and46bacterialOTUsand46, 49and49fungalOTUs,respectively.ThenumberofT-RFsshared amongallsites,wasmuchlowerinfungalcommunities (10/99, 10%)thaninbacterialones(31/60,52%).Theseresultsarein con-trasttowhatobservedinastudyonnativebeechforestsandbeech forestsconvertedtoconiferforeststhatreported56–60%ofshared fungal OTUs,defined by DNA fingerprintinganalysis (Purahong etal.,2014).TheT-RFssharedamongthethreesitesofthislast studyconstitutedacoremicrobialcommunitythatdidnotseemto sufferanyinfluenceofland-use,edaphicconditions,andtree-type association.Inourstudy,theconvertedsite(Site2)hadthelower numberofbacterial(7.0%)andthehighernumberoffungal(52%) site-exclusiveT-RFscomparedtosite1(10%and38%,respectively) andsite3(17%and36%respectively),andalwayssharedagreater numberofT-RFswithsite1thanwithsite3.
Theseresultsstronglyindicatedthatthethreesitesnotonly dif-feredinthephysicochemicalcharacteristicsofsoilandinmicrobial diversity,butalsoinmicrobialcomposition.However,site2 dif-feredfromtheothertwositesmorethantheydifferedfromeach other,suggestingtheoccurrenceofamajorshiftinitsmicrobial compositionastheresultoftheconversionfromnaturalforestto poplarplantation.Purahongetal.(2014),applyinganARISADNA fingerprintinganalysistostudytheconsequencethatthe conver-sionfromnative,deciduousforesttoconiferousforestinGermany hadonthediversityand compositionoffungal communitiesin deadwood,reportedsimilaralterations,and statedthatchanges withinasingleland-usecategorycanberegardedasamajorthreat tofungaldiversityintemperateforestecosystem.Conversely,ina study,performedbyPCR-DDGE,ontheeffectsoftheconversionof naturaltropicalrainforestsinUganda,theauthorsreporteda rela-tiveresilienceofsoilmicrobialcommunitiestoforestconversionat alocalscale(Aleleetal.,2014).Differencesinthekindofland-use change,insoiltype,inanalyticalmethodusedtoinvestigatethe microbialcommunities(i.e.,onresolutionanddepthofthe imple-mentedanalysistechnique),andingeneralfeaturesofthespecific biomesubjectofthestudy,couldpartiallyexplainthecontradictory resultsofthesestudies.
Assite2differedfromsite1(wheresoilsassociatedtonatural poplarsweresampled)lessthanfromsite3(wheresoilsassociated to natural maples were sampled), a tree-type effect on phylo-typescompositioncouldbealsohypothesized.Indeed,significant plantspecificeffectsonmicrobialcommunitystructurehavebeen reportedbyanumberofstudies(Myersetal.,2001;Hackletal., 2004;Bastiasetal.,2007;BergandSmalla2009;Wubetetal.,2012; Wangetal.,2013).
3.2.2. ˇ-Diversityanalysis
Bray-CurtisdistancebetweenindividualT-RFLPprofileswithin eachsite(within-site-diversity)wascalculatedtodeterminethe
Fig.7.Distancetogroupcentroid(estimationof-diversity)ofbacterialandfungal communitiesinthethreesitesbasedonBray–Curtisdissimilarity.
effects ofland-useconversiononthespatialturnoverof micro-bialcommunities.Meanvaluesofwithin-site-diversity()for bacterialcommunitieswere:site1,ˇ=0.673(SD=0.171);site2, ˇ=0.569(SD=0.196);site3,ˇ=0.670(SD=0.186).Two-sample t-testshowedthattheonlystatisticallysignificantdifferenceswere between site 2 and sites 1 and 3 (Site 1 vs. Site 2:t= 8.253, p=<0.01;Site 1vs. Site 3:t=0.603, p=0.546;Site 2vs. Site 3: t=−8.166, p=<0.01).Meanvalues ofwithin-site -diversityfor fungal communities were: site 1, ˇ=0.838 (SD=0.113); site 2 ˇ=0.837(SD=0.151);site3ˇ=0.815(SD=0.130).Differentlyfrom bacteria,whenappliedtofungaldata,two-samplest-testshowed thattheonlystatisticallysignificantdifferencewasthatobserved betweensites1and3(Site1vs.Site3:t=5.8362, p=<0.01;Site 2vs.Site3:t=2.339,p=0.02;Site1vs.Site2:t=0.110,p=0.912). Fig.7reportsthevaluesofmultivariatedispersionofwithin-site -diversity.Forbacterialcommunities,dispersionvaluesinsite2 werelowerthaninsite 1andsite 3,whilenodifferences were foundindispersionvaluesbetweensite1and3(Site1vs.Site3: t=0.322,p=0.0770;Site1vs.Site2:t=3.126,p=0.001;Site3vs. Site2:t=2.363,p=0.016;testedwith999permutations).Onthe contrary,fungalcommunitiesdidnotshowstatisticallysignificant differencesinthedispersionvaluesbetweenthethreesites(Site1 vs.Site3:t=1.530,p=0.128;Site1vs.Site2:t=0.007,p=0.994;Site 3vs.Site2:t=−1.051,p=0.296;testedwith999permutations). Thoseresultssuggestthat,compared tothenaturalforests, the cultivatedforestshowedahigherspatialhomogeneityofbacterial communities,butnotofthefungalones.
Between-site-diversity values (calculated withBray-Curtis distance) indicatedthat cumulativeT-RFLPprofiles ofmicrobial communitiesfromthetwonaturalforestsinsite1and3weremore similartoeachother(bacteria,site1vs.site3=0.21;fungi,site 1vs.site3=0.66)thantheyweretocommunitiesfromthepoplar plantationinsite 2(bacterial,site 1vs.site2=0.26 andsite3 vs.site2=0.33;fungi,site1vs.site2=0.80andsite3vs.site 2=0.88).However,bacterialcommunitiesfromthepoplar planta-tionweremoresimilartocommunitiesfromthesoilsassociatedto naturalpoplarssampledinsites1thantothoseassociatedto nat-uralmaplesinsite3.Thosedataconfirmedthedifferencebetween thepoplarplantationandthenaturalforests,butalsoreinforced thehypothesisofatree-typeeffectonsoilmicrobialcommunities diversity;aneffectalreadysuggestedbythemicrobialphylotypes distributionanalysisinthethreesites(Fig.6).Themeanvalueof between-site–diversitywerehigherforfungi(0.78±0.11)than forbacteria(meanvalue0.27±0.06);aresultthat,inadditionto whatalreadyobservedwithineachsite(see␣-diversityanalysis), furtherhighlightedthegreatheterogeneityoffungipopulationsin thethreeforestsites.
3.2.3. Ordinationanalysis
ThesimilarityofindividualT-RFLPprofilesofbacterialand fun-gal communities in each samplingsite wasanalyzed by nMDS ordination.ANOSIManalysis(notshown)wasusedtotestthe sig-nificanceofwithin-sitegroupingofsoilsampleswithrespectto
Fig.6. VenndiagramillustratingthenumbersofuniqueandsharedT-RFsinthepooledbacterial(a)andfungal(b)T-RFLPprofiles.OnlyT-RFswitharelativeheight>0.01 areconsidered.
time(seasonandyear)andspace(differentindividualtrees).As regard bacteria,nMDS ordinationplots were alwayssignificant (0.19>stressvalues>0.11),and showeda tendencyofthe sam-plestoformseasonalclustersinthe2D-ordinationspace(Fig.S1), evenif ANOSIM analysisdidnotsupport anytested groupings. Unlikebacteria,nMDSordinationplotsoffungalcommunities(Fig. S2)wereclosetorandom(stressvalue>0.2),butANOSIManalysis offungalcommunitiesshowedstatisticalsignificancewhenfungi samplesweregroupedonthebasisofbelongingtothesame indi-vidualtree(site1:R=0.58,p=0.001;site2:R=0.16,p=0.024;site3: R=0.55,p=0.001),or,asinsite3only,onyearofsampling(R=0.09, p=0.003). Bacterial communities differed over space and time; whereasfungipopulationsappearedmoredifferentoverspaceand stableovertime.Theaboveresultsoutlinedageneralhigh hetero-geneityofmicrobialpopulationsinthethreesites.Highwithin-site heterogeneitymayreduce theresolution powerof theanalysis ofsimilarityofmicrobialcommunitystructurebetweendifferent sites,diminishingourcapabilitytofocusontheeffectofthe con-version,whichrepresentedthemainobjectiveofthiswork.Forthis reason,ineachofthethreesites,theT-RFLPprofilesofthe micro-bialcommunitiesfromindividualsamplesinagivenseasonwere puttogetherinsilicoandaveragedtoobtainatotalof21 cumula-tiveprofiles.Inthisway,interalia,communityprofiledatawere moredirectlycomparabletothedataobtainedfromanalysisofsoil chemistry,whichwereperformedon21correspondingseasonal compositeofsoils.ANOSIManalysiswasperformedonboth indi-vidualandcumulativeT-RFLPprofilesforcomparison.Inbothcases, differencesbetweensitesweresignificant,butcumulativeprofile analysisalwaysexplainedalargerpartofsitevariability(Table2). CumulativeT-RFLPprofileswereusedinsubsequentanalysis.
Fig.8reportsnMDSordinationplotandUPGMAclusteranalysis ofcumulativeT-RFLPprofilesofbacterialcommunitiesfromthe threesites.Theordination wassignificant(stressvalue=0.107), andoveralltheresultshighlightedaseparation ofsoilbacterial communitiesinthepoplarcultivation(site2)fromthoseofthe naturalmixedforests(site1and3),whichlargelyoverlappedwith each otherformingtwo mixed clusters. nMDSordination anal-ysisof fungal communities (Fig.9)was also significant(stress value=0.112),andshowedamoreclearseparationbetweenthe threesites,withcumulativeT-RFLPprofilesformingwellseparated anddefined clusters. Environmentalvariables (OM,pH, relative humidity,cumulativerain,andairtemperature)andbacterialand fungalviablecountswerefittedontotheordinationspaceofthe nMDSplotsinFigs.8and9.Theenvironmentaldescriptorsthatbest explaineddifferencesinassemblagestructurewerepH,OMand relativehumidity,bothinbacteria(pH:r2=0.583,p=0.003;OM:
r2=0.673,p=0.001;relativehumidity:r2=0.556,p=0.007)andin
Fig.8.Non-metricmultidimensionalscalingofT-RFLP-basedcompositionof bac-terialcommunitiesinsoilsfromSite1–3indifferentseasons.Eachpointrepresent apoolofT-RFLPprofilesfromthesamesite,season,andyear.Dottedellipsesshow resultsofUPGMAanalysisandhighlightsamplesclusteredat>75%similarity,based onSørensendissimilarityindex.Vectorsindicateonlyenvironmentalvariablesthat weresignificantlycorrelatedwiththeordination(p<0.01).
fungi(pH:r2=0.654,p=0.003;OM:r2=0.854,p=0.001;relative
humidity:r2=0.654,p=0.001).Inaddition,thefittingofthefungal
viablecountparameterinthefungiordinationwasalsosignificant (r2=0.519,p=0.005).
Fig.9.Non-metricmultidimensionalscalingordinationofT-RFLP-basedoffungal communitiesinsoilsfromSite1–3indifferentseasons.Eachpointrepresentapool ofT-RFLPprofilesfromthesamesite,season,andyear.Dottedellipsesshowresults ofUPGMAanalysisandhighlightsamplesclusteredat>50%similarity,basedon Sørensendissimilarityindex.Vectorsonlyindicateenvironmentalvariablesthat weresignificantlycorrelatedwiththeordination(p<0.01).
Table3
Coefficientofvariation(CV)of␣-diversityovertime.CVwasreportedaspercentageratiobetweenstandarddeviationandthemeanvalueofthediversityparameter.
Bacteria Fungi
Richness Evenness Simpsonindex Shannonindex Richness Evenness Simpsonindex Shannonindex
Site1 19.8% 6.5% 5.7% 10.7% 32.6% 13.7% 17.0% 23.5%
Site2 18.9% 5.1% 4.5% 10.0% 27.7% 9.8% 11.1% 18.9%
Site3 25.8% 7.8% 7.6% 13.4% 35.3% 10.9% 14.8% 22.9%
UPGMAclusteranalysisandnMDSordinationplotsclearly dif-ferentiatedthebacterialandfungalcumulativeT-RFLPprofilesof site2(poplarplantation)fromthoseofsites1and3(naturalforests) alongthefirstordinationaxis(Figs.8and9).Onlywithregardtothe fungi,sites1(soilsamplesassociatedtonaturalpoplars)andsite 3(soilsamplesassociatedtonaturalmaples)separatedalongthe secondordinationaxis.Byinterpretingtheseresults,wespeculate thatthefirstaxiswasstronglycorrelatedtosoilland-use(poplar plantationinsite2 vs.naturalforestinsite1and 3),whilethe secondaxiscouldbecorrelated,moreweaklyandonlyasregard fungi,totreetype(poplarinsite1and2vs.maplesinsite3).From thedirectionandangleofthevectorsresultingfromthefittingof soilphysicochemicalcharacteristicsonnMDSplots,pHandorganic matter(OM),andtoalesserextentsoilhumidity,werecorrelated tothefirstaxis,andhencetoland-use.
Overall,bothdiversityandordinationanalysisstrongly differ-entiatedbacteriacommunityintheconvertedforestfromthatof natural,primaryforests.Thefactthatsensibledifferencesoccurred betweenbacterialcommunitiesincultivatedpoplarsinSite2and primaryforestinSite1,despitethetwositesarelocatedinclose proximityandinbothsitessoilassociatedtopoplarswere sam-pled,stronglysuggestedthattheeffectsonbacteriapopulationsare mainlyduetoforestconversionratherthantospatialortreetype effects.Similarlossindiversityofbacterialcommunitieswere observedasaresultofforesttopastureconversionintheAmazon rainforest(Rodriguesetal.,2013),butnotofconversionfromforest tooilpalmplantationinBorneo,wherediversitywasfoundtobe higherinconvertedforestrespecttoprimaryones(Lee-Cruzetal., 2013).However,theauthorsofthislastpaper,takingintoaccount thepossibilitythattheobserveddifferencesindiversitycouldbe presentpriortotheconversion,emphasizedtheneedforlong-term studiestoassesstheeffectsoftheconversion.
Unlikebacteria,thediversityoffungalpopulationsinourstudy seemtobealsolinkedtotreetypeeffects,inagreementwithwhat suggestedbyPurahongetal.(2014).
3.3. Temporalfluctuationanalysisofmicrobialcommunity diversity
Giventhelong-termnatureofourstudy,wewerealso inter-estedintheeffectthatseasonalityhaveonmicrobialcommunities diversityandinunderstandinghowsiteswithadifferentland-use respondtoseasonalchanges.Asabovereported,adifferentiation ofmicrobialcommunitiesonaseasonalbasewasnotsupportedby ANOSIManalysis;nevertheless,nMDSordinationsshowedsome degreeofseasonalclusterization,especiallyforthebacterial com-munities(Fig.S1).Inordertobetterclarifytherelationshipbetween seasonalityand ␣-diversityin each samplingsite, wehave cal-culateda coefficientof variation astheratio betweenstandard deviation and themean value of different ␣-diversity parame-ters(Table3).␣-diversityofbothbacteriaandfungicommunities insite2hadnarrowertemporalfluctuationscomparedtosite1 and3,indicatinghighertemporalstabilityofmicrobial commu-nitieswithin the poplarplantation soil. Other studiesreported contradictoryover-timeeffectsofland-usechangesondiversity andstructureofmicrobialcommunitiesofsoil,fromnochangesto majorchanges(DaCJesusetal.,2009;Sunetal.,2011;Rodrigues
etal.,2013;Suleimanetal.,2013;McGuireetal.,2014;and bibli-ographywithinthesepapers).Differencesingeneralfeaturesofthe study,likethekindofland-usechange,thesoiltype,theanalytical methodusedinmicrobialanalysis,maybeattheoriginofthese contradictoryresults.
4. Conclusions
Inthisstudy,acommonandwellestablishedDNA fingerprint-ingtechnique(T-RFLP)wassuccessfullyappliedtotheanalysisof bacterialandfungalpopulationsinsoilsamplesfromnativeforests andaforestconvertedtopoplarplantationtoinvestigate possi-ble effectsondiversityandcommunity structure.Withtheaim tominimizetheeffectsofgeneralclimaticdifferencesandthose ofunknownandunwantedanthropicimpacts,thestudyareawas locatedwithinanaturalpark,withhomogeneousorographicand soiltexturecharacteristics.
Overall,ourdataindicatesthatsite2,theresultofthe conver-sion(about30yearsago)fromnaturaltopoplarplantation,differed fromtwonatural(primary)forests(site1and3)byanumberof abioticandmicrobiologicalparametersofsoil.Majordifferences wereobservedonbothrichnessanddiversityofbacterialand fun-galcommunities.Thedifferencesweremuchstrongerbetweensite 2andsites1and3thanbetweensites1and3.Thesedifferences wereinterpretedastheeffectsoftheforestconversioninsite2,and becausetheywerestillvisibleafterabout30yearsfromthe conver-sion,theyshouldbeconsideredlonglastingandalmostpermanent. Interestingly,acontributionofthetreetype(poplarvs.maples)in shapingthestructureofbacterialand(particularly)fungal com-munitiesinthethreesamplingsitesemergedfrombetween-site -diversity values and nMDSanalysis. Furthermore,we cannot excludethat aplant genotype-specificcontribution(Schweitzer etal.,2008)mayhavehadsomemarginalroleindeterminingthe differencesindiversityandstructureofthemicrobialpopulations fromsoilsassociated toP.albaand P.canescens insite1and P. nigra×P.deltoidshybridinsite2.
It is evident that we are yet far from achieving general knowledgeabouttheresponseofmicrobial communitiestothe conversionwithin a singleland-usecategory, and thatno gen-eraltrendscanbeoutlined.Inanycase,fromthesestudies,forest conversionemergesasapracticethatshouldberegardedasa gen-eralthreattomicrobialcommunitiesthatstronglyaffectsmicrobial diversityandstructure.
Acknowledgments
Theauthorswouldliketothank:AlessioMengoni(Dept.of Biol-ogy,University of Florence)for hisreview of theworkand for hisprecioussuggestions,CristinaIndorato(Dept.ofBiology, Uni-versityof Florence)for hertechnicalassistance,CristinaVettori (IGV-FI,CNR)for assistancein T-RFLP analysis,EmilianoFratini (Dept. of Chemistry, University of Florence) for organic matter analysisofsoil,GiacomoCertini(DISPAA,UniversityofFlorence) forsoiltextureanalysisofsite2samples,DavideTravagliniand Francesca Bottalico (GESAAF, University of Florence) for plants georeferencing,AlessandroMaterassiandGianniFasano(IBIMET CNR,Florence)formeteorologicaldata,FrancescaLogli(Migliarino
SanRossoreMassaciuccoliregionalpark)forusefulforestry infor-mationaboutthesamplingsiteswithinthepark.Thisworkwas fundedbytheEuropeanCommunityundertheLIFE+programme (grantno.LIFE08NAT/IT/00342–DEMETRAproject).Giuliana Sena-toreandCesareaCaroppohavebeenfinanciallysupportedbygrant no.LIFE08NAT/IT/00342.
AppendixA. Supplementarydata
Supplementarydataassociatedwiththisarticlecanbefound,in theonlineversion, athttp://dx.doi.org/10.1016/j.micres.2015.10. 002.
References
Alele,P.O.,Sheil,D.,Surget-Groba,Y.,Lingling,S.,Cannon,C.H.,2014.Howdoes conversionofnaturaltropicalrainforestecosystemsaffectsoilbacterialand fungalcommunitiesintheNileRiverwatershedofUganda?PLoSOne9(8), e104818.
Bardgett,R.D.,Wardle,D.A.,2010.Aboveground-BelowgroundLinkages:Biotic Interactions,EcosystemProcesses,AndGlobalChange.OxfordUniversity Press,Oxford.
Bastias,B.A.,Anderson,I.C.,Xu,Z.,Cairney,J.W.G.,2007.RNA-andDNA-based profilingofsoilfungalcommunitiesinanativeAustralianeucalyptforestand adjacentPinuselliottiplantation.SoilBiol.Biochem.39,3108–3114.
Berg,G.,Smalla,K.,2009.Plantspeciesandsoiltypecooperativelyshapethe structureandfunctionofmicrobialcommunitiesintherhizosphere.FEMS Microbiol.Ecol.68(April(1)),1–13.
DaCJesus,E.,Marsh,T.L.,Tiedje,J.M.,deSMoreira,F.M.,2009.Changesinlanduse alterthestructureofbacterialcommunitiesinWesternAmazonsoils.ISMEJ.3 (September(9)),1004–1011,NaturePublishingGroup.
Chao,A.,1984.Nonparametricestimationofthenumberofclassesinapopulation. Scand.J.Stat.11(4),265–270.
VanDorst,J.,Bissett,A.,Palmer,A.S.,Brown,M.,Snape,I.,Stark,J.S.,etal.,2014.
Communityfingerprintinginasequencingworld.FEMSMicrobiol.Ecol.89 (August(2)),316–330.
Gardes,M.,Bruns,T.D.,1993.ITSprimerswithenhancedspecificityfor basidiomycetes—applicationtotheidentificationofmycprrhizaeandrusts. Mol.Ecol.2,113–118.
Hackl,E.,Zechmeister-Boltenstern,S.,Bodrossy,L.,Sessitsch,A.,2004.Comparison ofdiversitiesandcompositionsofbacterialpopulationsinhabitingnatural forestsoils.Appl.Environ.Microbiol.70,5057–5065.
VanderHeijden,M.G.,Bardgett,R.D.,vanStraalen,N.M.,2008.Theunseen majority:soilmicrobesasdriversofplantdiversityandproductivityin terrestrialecosystems.Ecol.Lett.11(March(3)),296–310.
Heiri,O.,Lotter,A.F.,Lemcke,G.,2001.Lossonignitionasamethodforestimating organicandcarbonatecontentinsediments:reproducibilityandcomparability ofresults.J.Paleolimnol.25,101–110.
Hughes,J.B.,Hellmann,J.J.,Ricketts,T.H.,Bohannan,B.J.M.,2001.Countingthe uncountable:statisticalapproachestoestimatingmicrobialdiversity.Appl. Environ.Microbiol.67(10),4399–4406.
Kerfahi,D.,Tripathi,B.M.,Lee,J.,Edwards,D.P.,Adams,J.M.,2014.Theimpactof selective-loggingandforestclearanceforoilpalmonfungalcommunitiesin Borneo.PLoSOne9(11),e111525.
Lauber,C.L.,Hamady,M.,Knight,R.,Fierer,N.,2009.Pyrosequencing-based assessmentofsoilpHasapredictorofsoilbacterialcommunitystructureat thecontinentalscale.Appl.Environ.Microbiol.75,5111–5120.
Lee-Cruz,L.,Edwards,D.P.,Tripathi,B.M.,Adams,J.M.,2013.Impactofloggingand forestconversiontooilpalmplantationsonsoilbacterialcommunitiesin Borneo.Appl.Environ.Microbiol.79(23),7290–7297.
Lupatini,M.,Jacques,R.J.S.,Antoniolli,Z.I.,Suleiman,A.K.A.,Fulthorpe,R.R.,Roesch, L.F.W.,2013.Land-usechangeandsoiltypearedriversoffungalandarchaeal
communitiesinthePampabiome.WorldJ.Microbiol.Biotechnol.29(Feburary (2)),223–233.
McGuire,K.L.,D’Angelo,H.,Brearley,F.Q.,Gedallovich,S.M.,Babar,N.,Yang,N., etal.,2014.Responsesofsoilfungitologgingandoilpalmagriculturein SoutheastAsiantropicalforests.Microb.Ecol.(August(23)).
MCPFE,UNECE,FAO.StateofEurope’sforests2007.MCPFERep.Sustain.For. Manag.Eur.2007.
Myers,R.T.,Zak,D.R.,White,D.C.,Peacock,A.,2001.Landscape-levelpatternsof microbialcommunitycompositionandsubstrateuseinuplandforest ecosystems.SoilSci.Soc.Am.J.,359.
Nacke,H.,Thürmer,A.,Wollherr,A.,Will,C.,Hodac,L.,Herold,N.,etal.,2011.
Pyrosequencing-basedassessmentofbacterialcommunitystructurealong differentmanagementtypesinGermanforestandgrasslandsoils.PLoSOne6 (2).
OksanenJ.,BlanchetG.L.,KindtR.,LegendreP.,MinchinP.R.,O’HaraB.R.,etal. 2013.vegan.CommunityEcologyPackage.CommunityEcol.Packag. Osborn,A.M.,Moore,E.R.B.,Timmis,K.N.,2000.Anevaluationof
terminal-restrictionfragmentlengthpolymorphism(T-RFLP)analysisforthe studyofmicrobialcommunitystructureanddynamics.Environ.Microbiol.2.
Osborne,C.A.,Galic,M.,Sangwan,P.,Janssen,P.H.,2005.PCR-generatedartefact from16SrRNAgene-specificprimers.FEMSMicrobiol.Lett.248,183–187.
Purahong,W.,Hoppe,B.,Kahl,T.,Schloter,M.,Schulze,E.,Bauhus,J.,etal.,2014.
Changeswithinasingleland-usecategoryaltermicrobialdiversityand communitystructure:molecularevidencefromwood-inhabitingfungiin forestecosystems.J.Environ.Manage.139,109–119,ElsevierLtd.
RCoreTeam,2013,R:ALanguageandEnvironmentforStatisticalComputing.R Found.Stat.Comput.Vienna,Austria:RFundationforstatisticalComputing. Rees,G.N.,Baldwin,D.S.,Watson,G.O.,Perryman,S.,2004.Ordinationand
significancetestingofmicrobialcommunitycompositionderivedfrom terminalrestrictionfragmentlengthpolymorphisms:applicationof multivariatestatistics.AntonievanLeeuwenhoek86,339–347.
Rodrigues,J.L.M.,Pellizari,V.H.,Mueller,R.,Baek,K.,Jesus,E.D.C.,Paula,F.S.,etal., 2013.ConversionoftheAmazonrainforesttoagricultureresultsinbiotic homogenizationofsoilbacterialcommunities.Proc.Natl.Acad.Sci.U.S.A.110 (January(3)),988–993.
Rousk,J.,Bååth,E.,Brookes,P.C.,Lauber,C.L.,Lozupone,C.,Caporaso,J.G.,etal., 2010.SoilbacterialandfungalcommunitiesacrossapHgradientinanarable soil.ISMEJ.4(October(10)),1340–1351.
Schweitzer,J.A.,Bailey,J.K.,Fischer,D.G.,LeRoy,C.J.,Lonsdorf,E.V.,Whitham,T.G., etal.,2008.Plant-soilmicroorganisminteractions:heritablerelationship betweenplantgenotypeandassociatedsoilmicroorganisms.Ecology89 (March(3)),773–781.
Shannon,C.E.,1948.Amathematicaltheoryofcommunication.BellSyst.Tech.J. 27,379–423.
Simpson,E.H.,1949.Measurementofdiversity.Nature163,688.
Suleiman,A.K.A.,Manoeli,L.,Boldo,J.T.,Pereira,M.G.,Roesch,L.F.W.,2013.Shifts insoilbacterialcommunityaftereightyearsofland-usechange.Syst.Appl. Microbiol.36(March(2)),137–144,ElsevierGmbH.
Sun,B.,Dong,Z.X.,Zhang,X.X.,Li,Y.,Cao,H.,Cui,Z.L.,2011.Ricetovegetables: short-versuslong-termimpactofland-usechangeontheindigenoussoil microbialcommunity.Microb.Ecol.62,474–485.
Wagg,C.,Husband,B.C.,Green,D.S.,Massicotte,H.B.,Peterson,R.L.,2011.Soil microbialcommunitiesfromanelevationalclinedifferintheireffecton coniferseedlinggrowth.PlantSoil,491–504.
Wang,M.,Qu,L.,Ma,K.,Yuan,X.,2013.Soilmicrobialpropertiesunderdifferent vegetationtypesonMountainHan.Sci.ChinaLifeSci.56(6),561–570,SP ScienceChinaPress.
WickhamH.,2011,ggplot2.WileyInterdiscipRevComputStat.3:180–5. Wubet,T.,Christ,S.,Schöning,I.,Boch,S.,Gawlich,M.,Schnabel,B.,etal.,2012.
DifferencesinsoilfungalcommunitiesbetweenEuropeanbeech(Fagus sylvaticaL.)dominatedforestsarerelatedtosoilandunderstoryvegetation. PLoSOne7(January(10)),e47500.
Zak,D.R.,Holmes,W.E.,White,D.C.,Peacock,A.D.,Tilman,D.,2003.Plantdiversity, soilmicrobialcommunities,andecosystemfunction:arethereanylinks? Ecology84,2042–2050.