<%BANNER%>

UFIR



Nematode Spatial and Ecological Patterns from Tropical and Temperate Rainforests
www.plosone.org ( Publisher's URL )
CITATION PDF VIEWER
Full Citation
STANDARD VIEW MARC VIEW
Permanent Link: http://ufdc.ufl.edu/IR00001169/00001
 Material Information
Title: Nematode Spatial and Ecological Patterns from Tropical and Temperate Rainforests
Series Title: PLoS ONE 7(9): e44641. doi:10.1371/journal.pone.0044641
Physical Description: Journal Article
Creator: Porazinska, Dorota
Giblin-Davis, Robin
Powers, Thomas
Thomas, W. Kelley
Publisher: Public Library of Science
Place of Publication: online
Publication Date: September 11, 2012
 Notes
Abstract: Large scale diversity patterns are well established for terrestrial macrobiota (e.g. plants and vertebrates), but not for microscopic organisms (e.g. nematodes). Due to small size, high abundance, and extensive dispersal, microbiota are assumed to exhibit cosmopolitan distributions with no biogeographical patterns. This assumption has been extrapolated from local spatial scale studies of a few taxonomic groups utilizing morphological approaches. Recent molecularly-based studies, however, suggest something quite opposite. Nematodes are the most abundant metazoans on earth, but their diversity patterns are largely unknown. We conducted a survey of nematode diversity within three vertical strata (soil, litter, and canopy) of rainforests at two contrasting latitudes in the North American meridian (temperate: the Olympic National Forest, WA, U.S.A and tropical: La Selva Biological Station, Costa Rica) using standardized sampling designs and sample processing protocols. To describe nematode diversity, we applied an ecometagenetic approach using 454 pyrosequencing. We observed that: 1) nematode communities were unique without even a single common species between the two rainforests, 2) nematode communities were unique among habitats in both rainforests, 3) total species richness was 300% more in the tropical than in the temperate rainforest, 4) 80% of the species in the temperate rainforest resided in the soil, whereas only 20% in the tropics, 5) more than 90% of identified species were novel. Overall, our data provided no support for cosmopolitanism at both local (habitats) and large (rainforests) spatial scales. In addition, our data indicated that biogeographical patterns typical of macrobiota also exist for microbiota.
Acquisition: Collected for University of Florida's Institutional Repository by the UFIR Self-Submittal tool. Submitted by Robin Giblin-Davis.
Publication Status: Published
 Record Information
Source Institution: University of Florida Institutional Repository
Holding Location: University of Florida
Rights Management: All rights reserved by the submitter.
System ID: IR00001169:00001

Downloads

This item is only available as the following downloads:

Nematode Spatial and Ecological Patterns ( PDF )


Full Text

PAGE 1

NematodeSpatialandEcologicalPatternsfromTropical andTemperateRainforestsDorotaL.Porazinska1* ,RobinM.Giblin-Davis1,ThomasO.Powers2,W.KelleyThomas31 FortLauderdaleResearchandEducationCenter,UniversityofFlorida-InstituteofFoodandAgriculturalSciences,FortLauderdale,Florida,Unit edStatesofAmerica, 2 PlantPathologyDepartment,UniversityofNebraska,Lincoln,Nebraska,UnitedStatesofAmerica, 3 HubbardCenterforGenomeStudies,UniversityofNewHampshire, Durham,NewHampshire,UnitedStatesofAmericaAbstractLargescalediversitypatternsarewellestablishedforterrestrialmacrobiota(e.g.plantsandvertebrates),butnotfor microscopicorganisms(e.g.nematodes).Duetosmallsize,highabundance,andextensivedispersal,microbiotaare assumedtoexhibitcosmopolitandistributionswithnobiogeographicalpatterns.Thisassumptionhasbeenextrapolated fromlocalspatialscalestudiesofafewtaxonomicgroupsutilizingmorphologicalapproaches.Recentmolecularly-based studies,however,suggestsomethingquiteopposite.Nematodesarethemostabundantmetazoansonearth,buttheir diversitypatternsarelargelyunknown.Weconductedasurveyofnematodediversitywithinthreeverticalstrata(soil,litter, andcanopy)ofrainforestsattwocontrastinglatitudesintheNorthAmericanmeridian(temperate:theOlympicNational Forest,WA,U.S.Aandtropical:LaSelvaBiologicalStation,CostaRica)usingstandardizedsamplingdesignsandsample processingprotocols.Todescribenematodediversity,weappliedanecometageneticapproachusing454pyrosequencing. Weobservedthat:1)nematodecommunitieswereuniquewithoutevenasinglecommonspeciesbetweenthetwo rainforests,2)nematodecommunitieswereuniqueamonghabitatsinbothrainforests,3)totalspeciesrichnesswas300% moreinthetropicalthaninthetemperaterainforest,4)80%ofthespeciesinthetemperaterainforestresidedinthesoil, whereasonly20%inthetropics,5)morethan90%ofidentifiedspecieswerenovel.Overall,ourdataprovidednosupport forcosmopolitanismatbothlocal(habitats)andlarge(rainforests)spatialscales.Inaddition,ourdataindicatedthat biogeographicalpatternstypicalofmacrobiotaalsoexistformicrobiota.Citation: PorazinskaDL,Giblin-DavisRM,PowersTO,ThomasWK(2012)NematodeSpatialandEcologicalPatternsfromTropicalandTemperateRainforests.PLoS ONE7(9):e44641.doi:10.1371/journal.pone.0044641 Editor: JoshNeufeld,UniversityofWaterloo,Canada Received May2,2012; Accepted August6,2012; Published September11,2012 Copyright: 2012Porazinskaetal.Thisisanopen-accessarticledistributedunderthetermsoftheCreativeCommonsAttributionLicense,whichpermits unrestricteduse,distribution,andreproductioninanymedium,providedtheoriginalauthorandsourcearecredited. Funding: ThisworkhasbeensupportedbytheUnitedStatesDepartmentofAgriculture/CooperativeStateResearch,Education,andExtensionService–Tropica l andSubtropicalAgriculturalResearch(grants2006-04347and2008-34135-19505),theNationalScienceFoundation(DEB-0450537)-BiodiversityS urveysand Inventories.PublicationofthisarticlewasfundedinpartbytheUniversityofFloridaOpen-AccessPublishingFund.Thefundershadnoroleinstudy design,data collectionandanalysis,decisiontopublish,orpreparationofthemanuscript. CompetingInterests: Theauthorshavedeclaredthatnocompetinginterestsexist. *E-mail:dorotalp@ufl.eduIntroductionUnderstandingspatialpatternsofspeciesdiversityisimportantfor settingprioritiesforconservationandmonitoringandrestoration programs.Whilelargescalespatialpatternsarewellestablishedfor macroscopiceukaryotes(e.g.vertebratesandplants),formicroscopic eukaryotes(e.g.nematodeandmites)theyremaingreatlyuncharacterizedandunderexplored.Ithasbeenassumedthatmicrobiota exhibitcosmopolitanrandomdistributionsandlackbiogeographical patterns[1]primarilyduetotheirsmallsize,astronomical abundance,andhighdispersalrates[2–3].However,the‘‘everythingiseverywhere’’(EisE)assumptionhasbeenextrapolated predominantlyfromstudiesatlocalspatialscalesonprotozoantaxa [4–5]usingtraditionalmorphologicalapproaches.Morerecent molecularstudies,however,providestrikinglycontrastingevidence ofverylimitedcosmopolitanism[6–9]. Nematodespeciesrichnessisexpectedtoexceed1million,but lessthan4%isknowntoscience[10].Thisgapofknowledgeis commontoothereukaryoticmicroorganismsandgenerallyresults fromthedifficultyofapplyingtraditionalapproaches(morphology and/orsingleorganismPCRandsequencing)inspeciesidentification.Giventhatthesetaxaaremajorcomponentsofdetrital foodwebsandplaykeyrolesasdecomposers,predators,and parasites[11–12],itiscriticaltoexpandunderstandingoftheir biologyandecology.Knowledgeoftheirspatialpatternsisthefirst steptounderstandingtheirrolesinecosystemprocesses.Aswith protozoanspecies,theassertionofcosmopolitandistributionof nematodespeciescanbetracedbacktoseveralproblems:1) extrapolationfromobservationsatsmallspatialscales,2)useof morphologicalapproaches[13]thatprohibitidentificationathigh taxonomicresolution,3)biastowardsagriculturally-relevanttaxa andtemperateregions,4)processingoftoofewindividualsfrom toofewsamples,and5)absenceoflargespatialscalestudies. Ultrasequencingapproachesofferanopportunitytoaccelerate theknowledgeoftheglobalbiodiversityofmicroscopiceukaryotes byyieldingmoreinformationfasterandatlowercostthan traditionalapproaches.Ecometageneticshasbeensuccessfully usedtomapprokaryoticdiversity[14].Inthisstudy,weused ecometageneticstostartmappingthediversityofmicroscopic metazoans.Specifically,weconductedasurveyofnematode diversity(andothermicro-andmesofauna)withinthreevertical strataorhabitats(soil,litter,andcanopy)ofrainforestsattwo contrastinglatitudesintheNorthAmericanmeridian(temperate attheOlympicNationalForest,WA,U.S.A.andtropicalatLa PLOSONE|www.plosone.org1September2012|Volume7|Issue9|e44641

PAGE 2

SelvaBiologicalStation,CostaRica)usingidenticalsampling designsandsampleprocessingprotocols[15–16].Becausetwo regionsofthesamegene(5 9 -and3 9 -endsoftheSSUrDNA)are currentlyinuseinecometageneticanalysesofmarinenematodes [17],wetestedthisapproachonsamplesfromthetemperate rainforestaswell.Ourworkprovidesnosupportforcosmopolitan distributionofspeciesandinfactpointstothepresenceofpatterns typicalformacrobiota.However,thechoiceofsampledhabitats andtheprimersetsmayhaveastronginfluenceondiversity interpretation.Results PatternsforAllMesofaunaThetotalnumberofindividualnematodesinthetemperate rainforestvariedfrom8to222per100mlwithinlitter(L)and canopy(C)andfrom960to2680withinthesoil(S).Inthetropical rainforest,thepatternwassomewhatreversed,withlowest numbers(116to419)withincanopyandsoil,andhighest(887 to1490)withinthelitter.Temperaterainforestsamplesamplified fromthe3 9 -endoftheSSUdiagnosticregiongeneratedatotalof 42,023highqualitysequencingreadsfromwhich17.2%wereof nematodeorigin,41.0%wereidentifiedasothermicroscopic eukaryota,26.8%fellintoacategoryof‘‘environmentalsample’’ (sequenceswithnotaxonomicinformationintheNCBIdatabase), and15.0%weretaggedaschimeric.Usingthesamediagnostic locus,theseresultscontrastedwiththetropicalrainforestsamples (atotalof171,861highqualityreads)fornematodeandother eukaryotecategorieswith40%and26%,respectively,butwere similarforenvironmentalsamples(20%)andchimeras(14%)[16]. Thetemperatesamplesamplifiedonthe5 9 -endoftheSSU diagnosticregion,generatedatotalof41,348highqualityreads, withonly4%ofnematodeorigin,41%ofothereukaryoticorigin, 25.9%asanenvironmentalsample,and29.3%aschimeras. Usingthe3 9 -endgenerateddatasets,atotalof157micro-and mesofaunalputativespecieswereobservedacrossallhabitatsin thetemperaterainforestasopposedto323putativespeciesinthe tropicalrainforest[16]withnematodesasthemostdiverseand accountingfor44%(69species)and66%(214species)ofallthe species,respectively.Inbothforests,miteswereidentifiedasthe secondmostdiversegroupwiththeirrichnessinthetropical rainforestalmosttwiceashighasinthetemperate(Figure1).A similarpatternofdecreasingrichnessfromthetropicaltothe temperaterainforestwasobservedforallothergroupsexcept AnnelidsandCollembolans. Thetotalnumberofspeciesrecoveredfromthetemperate rainforestdependedonthechoiceofthediagnosticlocusrevealing thattheoverallspeciesrichnesswaslowerby30%whenassessed withtheuseofthe5 9 -endversusthe3 9 -endlocus.Thebiaswas mostlydirectedagainstnematodetaxa(FigureS1)with49(71%) fewernematodespecies,butasimilardecreasewasobservedfor tardigrades.PatternsforNematodesWhiletheoverall(acrossallhabitats)nematodespeciesrichness anddiversitywereconsiderablyhigherinthetropicalthaninthe temperaterainforest(Figure2A,B,FigureS2A,B),thedistribution ofspecieswithinhabitatswasforestspecific.Morethan80%ofall thespeciesandnematodeindividualsinthetemperaterainforest residedinthesoil,whereaslessthan20%inthetropics.While averagerichness,diversityanddensityweresignificantlyhigher (P 0.01)inLaSelvathanintheOlympicForestwithinthelitter andcanopyhabitats(FigureS2A,B,C),thepatternswere completelyreversedwithinthesoilhabitat,withrichnessand abundancesignificantly(P 0.01)higherattheOlympicForest thanattheLaSelvasite(FigureS2A,C). Thenematodeassemblageswerefundamentallydifferentattwo mainlevels:betweentheforestsandamonghabitatswithineach forest.LaSelvatropicalrainforestandOlympicNationalForest didnotshareevenasinglenematodespecies.Moreover,only2% ofLaSelvaspeciesand21%oftheOlympicForestspecies perfectlymatchedanexistingsequenceintheNCBIdatabase.At thescaleofeachrainforest,nematodecommunitieswerevery discretewithfewsharedspeciesbetweenhabitats(Figure2A,B). Only6(3%)outof214recognizednematodespeciesatLaSelva and10(15%)outof69intheOlympicForestweresharedamong thesoil,litter,andcanopy(FigureS3).Thesharedspeciesamong habitatswithintherespectiverainforestsbelongedtodifferent taxonomicgroups.AtLaSelva,theylargelyincludedbacterial feeders(Cephalobidae(2spp),Rhabditidae(2spp),Diplogastridae (1spp))andaplantparasite( Xiphinema (1spp))andattheOlympic Foresttheyincludedbacterialfeeders(Plectidae(3spp),Cephalobidae(2spp),Teratocephalidae(2spp))andomnivores (Dorylaimidae(3spp)).Theuniquecompositionwasfurther magnifiedbythequantitativeresponsesofthesharedspecies.For instance, Anaplectus sp.1(Species # 1inFigure2B)while completelydominatingthelittercommunity,fellintothe‘‘rare’’ speciescategorywithinsoilandcanopycommunities.Becausethe powerofthestatisticaltestwaslimitedbypatchynematode distribution(highfrequencyof0s),statisticallysignificantdifferencesweredetectedforonlytwotropicalspecies(species # 11and # 37inFigure2A).NematodeDiversityvs.TaxonomicResolutionTheleveloftaxonomicresolution(e.g.species,genus,and family)affectedthecomparabilityofthetworainforests.Atthe specieslevel,thesetwoecosystemswereentirelydifferent,withnot evenonespeciesincommonand3.4timeshigherrichnessatLa SelvathantheOlympicForest.Thetrophicrepresentationofthe specieswasalsodifferent,withthelargestdifferencesforbacterialfeeding,rootassociated,andplant-parasitictaxa.Bacterial-feeding (50%)andplant-parasitic(16%)taxacontributedthemosttothe overallspeciesrichnessatLaSelva,versusbacterial-feeders(38%) androot-associates(22%)attheOlympicForest(Figure3A).At thegenuslevel,thetwoecosystemsappearedmoresimilarby sharing22taxa(44%ofthetotalrichnessoftheOlympicForest, comparedwith23%oftotalrichnessinLaSelva),withrichness Figure1.Acomparisonoftotalnumbersofmicro-andmeiofaunalspeciesbetweentropical(LaSelvaBiologicalStation, CostaRica(LS))andtemperate(OlympicNationalForestinWA, U.S.A.(OF))rainforestsfromthe5 9 -endoftheSSUdiagnostic region. doi:10.1371/journal.pone.0044641.g001 SpatialPatternsofNematodeDiversity PLOSONE|www.plosone.org2September2012|Volume7|Issue9|e44641

PAGE 3

only1.9timeshigheratthetropicalthantemperaterainforest. Theproportionaldistributionofgeneraamongtrophicguilds convergedforallbutomnivoroustaxa(Figure3B).Andfinallyat thefamilylevel,thetworainforestswerenearlyidentical.The majorityoftaxa(86%ofthetotalrichnessattheOlympicForest and62%atLaSelva)wereshared,andtheproportional distributionoftaxaamongtrophicguildswassimilar(Figure3C) aswellastheestimatesofrichnesswerealsosimilar(only1.3times higheratLaSelvathantheOlympicForest).Despitethis increasingsimilarityinnematoderichnessbetweenthetwo rainforestsastheleveloftaxonomicresolutiondecreasedthe communitiesremainedfairlyuniqueinthewaydifferenttaxa contributedtotheassemblagesnumericallyevenatthefamilylevel (FigureS4).Overall,bacterial-feedingtaxaoverwhelmingly dominatedthetropicalrainforestcommunities,whilemoreeven distributionwasobservedintheOlympicForest.Allsharedtaxa hadverysitespecificnumericalresponses.Forinstance,while RhabditidaeandCephalobidae(bacterialfeeders)werethetwo mostprevalentfamiliesinthetropicalrainforest,Tylenchulidae (plantparasite)andQudsianematidae(omnivore)dominatedinthe temperateforest.Breakingthesecommunitiesintodiscretemicrohabitats(soil,litter,canopy)withineachforest,furtheremphasized theuniquenessofeachcommunityjustintermsofthesharedtaxa (Table1).Outof22sharedgenera,onlyone( Acrobeloides )was presentwithinallhabitatsinbothforestsinsimilarabundances andonlytwogenera( Pristionchus and Boleodorus )wereconsistently foundonlywithinthesoilenvironment.Whileatthefamilylevel thenumberoftaxadisplayingconsistentresponsesacrossrainforestsandhabitatsincreased(6outof24),themajorityofpatterns werestillunique.NematodeDiversityvs.DiagnosticLocusIncomparisontospeciesrichnessestimatedusingthe3 9 -end diagnosticlocus(totalspecies=63)(Figure2B),the5 9 -end underestimatedspeciesrichnessby60%(totalspecies=20) (Figure4).Forbacterialfeeders,theidentifiedspecieswerelargely parallel.Forfungal-feeding,plant-parasitic,androot-associated taxa,however,speciesrichnessandabundanceestimateswereso incongruentthatevenatthetrophicguildlevel,the3 9 -endand5 9 endrecoverednematodecommunitieshadverylittleincommon (Figure5).Inordertoexplainthisdiscrepancy,weinvestigated primingregionsofallidentifiedspeciesforwhichfulllengthSSU sequenceshavebeenpublished.Atotalof47fulllengthSSU sequences(17outof205 9 -endidentifiedspecies,and30outof62 3 9 -endidentifiedspecies)weredownloadedfromtheNCBIand alignedinMEGA5usingdefaultparameters.The3 9 -endpriming regions(bothforwardandreverse)were100%identicalacrossall 47taxa.Incontrast,the5 9 -endprimingsiteshad1–2bp differences,andthegreatmajorityofthemismatcheswere observedamongfungal-feeding,plant-parasitic,androot-associatedtylenchids.DiscussionElucidationofspatialpatternsofspeciesdiversityiscritical.Not onlydoesithelpuswithestablishingthetheoreticalmechanismsof Figure2.AveragenumberofreadsperspecieswithinSoil,Litter,andCanopyhabitats. A)tropicalrainforestatLaSelvaBiologicalStation inCostaRica(LS),B)temperaterainforestattheOlympicNationalForestinWA,U.S.A.(OF).Speciesaregroupedinto6trophicguilds:bacterial feeders(BF),fungalfeeders(FF),rootassociates(RA),plantparasites(PP),omnivores(OM),andpredators(PR)andreadsaresortedfromtheirhig hest tolowestnumbersbythelitterhabitatwithineachtrophicguild. doi:10.1371/journal.pone.0044641.g002 SpatialPatternsofNematodeDiversity PLOSONE|www.plosone.org3September2012|Volume7|Issue9|e44641

PAGE 4

diversityandtrajectoriesofevolution,butitalsohelpsinsetting practicalprioritiesforconservation,monitoring,andrestoration effortsofecosystemsandtheirfunctions.Giventhatthegreat majorityofthebiodiversityresidesnotwithmacroscopic,butwith microscopictaxa,itissurprisingthatmicrobialbiogeographystill lacksamap.Also,giventhatmicroscopictaxaplaykeyrolesin ecosystemfunctioningthroughdecompositionandnutrient mineralizationprocesses,itissurprisingthatwestilldonotknow Figure3.Anoverall(acrossallhabitats)taxonomicrichnessofLaSelva(leftpanel)andOlympicForest(rightpanel)atdifferent levelsoftaxonomicresolution. A)species,B)genus,andC)family.Ateachleveloftaxonomicresolutiontaxaweregroupedbytheirfeeding habit.BF=bacterialfeeders,FF=fungalfeeders,RA=rootassociates,PP=plantparasites,OM=omnivores,PR=predators,AP=animal parasites,AL=algivores. doi:10.1371/journal.pone.0044641.g003 SpatialPatternsofNematodeDiversity PLOSONE|www.plosone.org4September2012|Volume7|Issue9|e44641

PAGE 5

theanswerstomostofthequestionsofwho,howmany,where, andhow.Lastly,giventhatthe21stcenturymolecularadvances haverevolutionizedallaspectofbiologicalsciences,itissurprising thatthescienceofmicroscopicbiogeographyisstilllargely resignedtoearly20thcenturyhypotheses[1,18].Massive sequencingtechnologiesnowallowforlarge-scale,standardized, andeffective(timeandcost)biodiversityassessments,challenging theconceptofEisEwithmeaningfuldata. Here,weusedpyrosequencingtodrawinsightsonspatial diversitypatternsofnematodes(butalsobyothersimilarlysized invertebrates).Evenwiththerelativelylownumberofsamples,the emergingpatternsunambiguouslycontradictedtheEisEhypothesisonseverallevels.Atthelargespatialscale,notevenasingle speciescouldbeclassifiedaswidelydistributed.Thenematode communitiesofthetworainforestswereofcompletelydifferent speciescompositions.Veryfewcommonspeciesamonghabitats withineachrainforestfurtherexemplifiedthelackofunrestricted distributionsofnematodespeciesevenatthelocalscale. Moreover,thelownumberofperfectmatchestosequences (presumablyfromallovertheworld)withinpublicdatabases, couldbetreatedasfurtherevidenceagainstEisE.Theresultsof ourstudyareinlinewiththeresultsofotherrecentstudies applyingmolecularapproaches.Wuetal.[9]examinedenvironmentalSSUrDNAsequencesofsoilanimalsfrom11locations fromdifferentbiomesatvariouslatitudes.Justasinourstudy,they foundthat95.8%ofall(2,259)OTUs(assembledat99%similarity thatisconsideredoperationallyequivalenttoaspecieslevel)were presentatjustasinglelocationandnoOTUwascommontoall locations.Intheanalysisof26 Caenorhabditis speciesusing11genes, Kiontkeetal.[19]foundthatonly C.briggsae wastruly cosmopolitan.Allotherspecieswereconfinedtoaveryspecific geographicalarea(e.g.EastAsia,WestAfrica,orSouthIndia). Usingthecox1gene,Robesonetal.[8]lookedatthediversity patternsofbdelloidrotifers,ataxonmostlikelytoexhibit cosmopolitandistributions(abundant,anhydrobiotic,asexual, notoverlyspecious)[20].Theyobservedautocorrelationatthe localscale(upto133m),butbeyondthatdistance(upto 10,000km)communitieswereextremelydissimilarpredominantly duetothepresenceofpreviouslyunrecognizedcrypticspecies. Thisrecentlyemergingpatternofhighlyendemicratherthan cosmopolitantaxanotonlyformicroscopiceukaryotesbut prokaryotesaswell[21]canbelargelyattributedtotheuseof Table1. Relativeabundanceofsharedtaxaatthegenusand familylevelsoftaxonomicresolution.Taxon LaSelvaOlympicForest SoilLitterCanopySoilLitterCanopy Genus(22) Taxacommontobothsitesandallhabitats (BF) Acrobeloides *2.1*3.1*1.3*0.3*5.2*6.6 Taxacommontobothsitesandalmostallhabitats (BF) Oscheius 22.2 10.438.50.30.9 (BF) Plectus 20.46.2 4.7 9.54.6 (BF) Teratocephalus 0.10.9 4.1 2.85.5 (FF) Tylencholaimus 0.35.9 0.2 4.15.0 (PP) Xiphinema 10.8 0.10.32.02.0 Taxacommontobothsitesbutonlyfewhabitats (BF) Anaplectus 2.0 0.1 36.6 2.6 (BF) Geomonhystera 1.7 0.54.8 (BF) Howardula *1.8 1.6 *0.3 (BF) Pristionchus *0.4*0.5 (BF) Rhabditis 6.6 0.43.0 (BF) Trypilina 1.4 4.11.1 (FF) Aphelenchoides 8.4 8.51.1 2.2 (FF) Diptherophora 0.20.8 (RA) Aglenchus 0.10.3 (RA) Boleodorus *0.2*2.2 (RA) Filenchus 0.3 0.14.4 2.2 (PP) Anguina 0.10.9 (PP) Meloidogyne 0.9 0.1 0.7 0.7 (OM) Mesodorylaimus 5.70.40.3 (PR) Mylonchulus 0.7 0.10.1 0.6 Total%47.564.256.920.971.528.4 Family(24) Taxacommontobothsitesandallhabitats (BF)Cephalobidae *5.4*10.9*32.7*0.3*12.8*13.6 (RA)Tylenchidae *0.3*3.0*0.3*7.7*3.4*5.9Taxacommontobothsitesandalmostallhabitats (BF)Monhysteridae 1.80.5 2.5 0.74.8 (BF)Plectidae 22.36.3 5.4 48.420.8 (BF)Rhabditidae28.8 12.938.73.30.9 (BF)Teratocephalidae 0.51.6 4.1 2.95.5 (BF)Trypilidae 1.44.2 0.5 1.21.1 (FF)Tylencholaimidae0.3 5.9 0.24.1 5.0 (PP)Belonolaimidae 1.60.20.22.1 1.8 (PP)Heteroderidae 0.90.00.1 0.7 0.7 (PP)Longidoridae10.8 0.00.32.02.0 (OM)Quadsianematidae 3.01.6 13.2 13.040.3 Taxacommontobothsitesbutonlyfewhabitats (BF)Allantonematidae 1.9 1.6 0.3 (BF)Diplogastridae 4.2 0.30.5 0.5 (BF)Mermithidae *4.2*10.6 (FF)Aphelenchoididae 10.2 10.11.1 2.3 (FF)Diptherophoridae0.20.8 Table1. Cont.Taxon LaSelvaOlympicForest SoilLitterCanopySoilLitterCanopy (FF)Leptonchidae0.70.2 (PP)Anguinidae6.61.10.9 (PP)Criconematidae *23.5*0.3 (PP)Ecphyadophoridae *0.2*0.6 (PP)Tylenchulidae *4.6*38.6 (OM)Dorylaimidae8.00.80.7 (PR)Mononchidae 2.2 0.00.0 3.3 Total%90.191.995.295.198.596.4 Boldindicatestaxacommontospecifichabitats. *indicatestaxacommontobothsitesandconsistenthabitats.BF=bacterial feeders,FF=fungalfeeders,RA=rootassociates,PP=plantparasites,OM= omnivores,PR=predators. doi:10.1371/journal.pone.0044641.t001 SpatialPatternsofNematodeDiversity PLOSONE|www.plosone.org5September2012|Volume7|Issue9|e44641

PAGE 6

molecularmethodsthataremuchmoresensitiveinspecies recognitionthanmethodsbasedsolelyonmorphology.Duetoa limitednumberandeasytoobservemorphologicalcharactersin microscopicorganisms,whatappearsasafewgloballydistributed morphologicalspecies,oftenturnsoutasnumerousphylogenetic speciesthatarerelativelysitespecific[22].Intheexampleofthe Caenorhabditis study[19],morphologycanbeusedtoassignspecies tothemajorsupergroups(e.g. Elegans and Drosophila )butwithin eachgroupsomespeciesareconsideredsiblingtaxaandlook identical.Moreimportantly,almostnoneofthemorphological charactersanalyzedhadanunambiguousdistributionwhen superimposedonthephylogenyreconstructedfrommolecular dataindicatingnumerousconflictsbetweenmorphologyand molecularcharacters. Theuniquenessofnematodecommunitiestorainforestsand habitattypeswithineachrainforestprovideduswithevidence corroboratingtheideathatmicro-invertebratesarenotvery differentintheirspatialpatternsfrommacroorganisms.However, evidencethatrelatestothespeciesnumbersalongalatitudinal gradientisevenmorecritical.Nematodespecies,unlikethoseof macrotaxa,havebeenpredictedtoexhibitapeakoftheirdiversity intemperateregionsratherthanattheequator.Thispeakis potentiallyartifactualforthefollowingthreereasons:1)oversamplingoftemperateandundersamplingoftropicalregions;2) dependenceonmorphologyoflowtaxonomicresolutionin nematodediagnostics;and3)theuseofsmallscalesand disconnectedstudieswithhighlyvariablemethodologiesfor inferenceaboutthelargescaledistributionpatterns.Powerset al.[15]andPorazinskaetal.[16]providedpreliminarysupport forcompliancewithlatitudinalgradientsindetailedstudiesof tropicalrainforestnematodediversity.Whetherusingtraditional molecular(singleorganismPCR/Sangersequencing)ornovel (pyrosequencing)tools,theyreportedthattheoverallnematode speciesrichnesswashigh(167and214observedspecies, Figure4.AveragenumberofreadsperspecieswithinSoil,Litter,andCanopyhabitatsinthetemperaterainforestattheOlympic NationalForestinWA,U.S.A.generatedfromtheuseofthe5 9 -enddiagnosticlocus. Foracomparisontotheresultsfromtheuseof3 9 endseeFigure2B.Speciesaregroupedintotrophicguilds:bacterialfeeders(BF),fungalfeeders(FF),plantparasites(PP),omnivores(OM)andrea ds aresortedfromtheirhighesttolowestnumbersbythelitterhabitatwithineachtrophicguild. doi:10.1371/journal.pone.0044641.g004 Figure5.Anematodecommunitycompositionatthetrophicguildleveldependingonthechoiceofthediagnosticlocus(5 9 -endor 3 9 -endoftheSSUrDNA). doi:10.1371/journal.pone.0044641.g005 SpatialPatternsofNematodeDiversity PLOSONE|www.plosone.org6September2012|Volume7|Issue9|e44641

PAGE 7

respectively),potentiallycontradictingcurrentdogma.Boagand Yeates[13]reviewed134studiesfromdifferentecosystemsaround theworldandnotedtemperatebroadleafforestswithanaverage of67nematodespecies(morphologicallyidentified)asthemost diverse.Tropicalrainforestlaggedfarbehindwithanaverageof only33species.Sincesamplingstrategiesmostlikelywerelimited tosoil,theirresultsareremarkablyconsistentwithourdataforthe soilhabitat.ButPowersetal.[15]andPorazinskaetal.[16] showedthatthediversityinthetropicalrainforestreachedfrom thebelowgroundintothecanopystratum,suggestingthat estimatesofterrestrialnematodediversityexclusivelybasedon surveysofsoilhabitatswereinadequate.Theyhypothesizedthatin humidtropicalecosystems,suitabletemperatureandmoisture conditionsarenotrestrictedtotheconfinesofthesoilenvironment butextendintotheabovegroundandthatthisverticaldistribution ofnematodehabitatsmaybecomecompressedwithincreasing latituderestrictingnematodepresencetotheconventionalsoil environment.Themostlogicalnextsteptotestthishypothesiswas toreplicatethesamplingdesignandsampleprocessingprotocols inatemperaterainforest,anecosystemmostequivalentin structureandfunctiontothetropicalrainforest.Bothecosystems receive 4mofrainperyear,andextensivegrowthofplant-life (e.g.plants,ferns,mosses)expandsfromthefloorintothetree trunksandbranches.Withthismoreappropriatecomparison,not onlyoverallnematoderichness,butalsorichnessofother meiofaunaltaxonomicgroupsfollowedthelatitudinalgradient withrichness300%higherinthetropicalthantemperate rainforest.However,aswehypothesized,thedistributionof speciesamonghabitatswasofopposingpatterns.Eightypercentof thespeciesresidedinthesoilinthetemperaterainforest,whereas only20%inthetropics.Ifwehadonlysampledthesoil environment,80%ofthetropicalnematodediversitywouldhave goneundetected,corroboratingyetagainthepresumeddiversity peakinthetemperateregion.Thisresultissignificantbecauseit illustrateshowunintentionallybiasedsamplingdesignscan ultimatelyskewourconclusions.Nematodes,althoughtraditionallyconceptualizedassoilorganisms,arereallyaquatic[23]and theywillthrivewhereverafilmofwatercansupportthem.In grasslandsordesserts,theyaremostlikelytobeconfinedtothesoil environment,butinmorethree–dimensionalecosystemswith higherabovegroundstructureandcomplexity,theirhabitatscan beextendedabovethesoillayer.Inthecaseofourstudy,the reversepatternofdiversityamonghabitatswaslikelydrivenbysoil andlitterproperties.Withnoharddatainhand,wecouldonly speculatethathighlyorganicsoilsinthetemperaterainforestdid whatathickanddiverselitterlayerdidinthetropicalrainforest: providedamplefoodresourcestosupportadiversenematode community.Incontrast,limitedfoodresourceswithinpoorsoils (minimalorganicmatter)inthetropics[24]andscantlitterlayer (oftenovergrownbyasinglemossspecies)inthetemperate rainforestrestrictedthenematodediversityintherespective habitats.Again,temperatureislikelythemainunderlyingfactor settingdifferencesintheratesofsoilorganicmatterdecomposition andlitterlayeraccumulation,andconsequentlyindiversity patternswithinspecifichabitats.Overallnematodediversity, however,wasprobablyinfluencedbyacombinationoftemperatureandplantdiversity,butremainstobetested. Thisstrictadherencetosoilsamplinghaspossiblycontributed tothelackofobservanceoflatitudinalpatternsinthemostrecent studyofworldwidedistributionanddiversityofsoilanimals[9]. Forinstance,ifsampledaboveground,theCostaRicantropical rainforest(thesameonewesampled)andthePeruviantropical rainforestwouldbeprobablymoreonparwiththeKenyan grassland,themostdiverseecosystemrightattheequator.We oftenstriveforstandardizedmethodologies,butitisclearthat identicalsamplingacrossstructurallydivergentecosystemsdoes notequatetoappropriatesamplingstrategy.Kiontkeetal.[19] elegantlyshowedtheslowrateofdiscovery(22speciesover 120 years)ofapresumablysoilinhabiting Caenorhabditis species,but onceitwasrealizeditisarottingfruitinhabitant,therate increasedwith16newspeciesjustwithinthelast6years.Clearly, describingdiversitywillrequireexpandingourrepertoireof samplingstrategies. Asmuchassamplingstrategiesreflectinganecosystem’s complexityandstructurewillplayaroleinadequateassessments ofmeiofaunaldiversity,thediagnosticlociselectedforthe assessmentswillbejustasimportant.Fortheassessmentofthe diversitywithintheOlympicForest,weexpandedourprevious worktousetwoloci,5 9 -and3 9 -endoftheSSU[25].Whilethe5 9 partoftheSSUmightbemoredesirablebecauseofhigher sequencedivergenceandresolvingpower[25],itturnedouttobe inferiortothe3 9 -partasthelessconservedprimingregionfailedto amplifymanyTylenchinaknownasfungal-feedersandplantparasites/associates[26].Abiaslikethisnotonlymisrepresentsthe diversity,butalsodistortsinferencesaboutecosystemfunctioning byomittingwholegroupsoftaxa.Clearly,alocuswithaconserved andstableprimingregionforamplifyingacrossalltaxonomic groups,eveniflessresolving,ismoreappropriate.Inastudyofthe diversitypatternsofmarinenematodes,Biketal.[17]usedthe sametworegionswithoutanyobservablebiases.However, tylenchidsconstituteagroupofnematodesderivedfrommarine ancestorsthatinvadedterrestrialhabitats[27],andanapparent divergencewithintheprimingregionmakesthe5 9 sectionofthe SSUrDNAsub-optimalforterrestrialnematodediversitystudies. Analternativeto5 9 and3 9 sectionsoftheSSUrDNA,amid sectionwasusedintheworldwidestudyofsoilanimals[9]. BecausetheestimatesofrichnessfortheCostRicanrainforest weresomewhatlowerthanours,webrieflyinvestigatedthe conservationoftheirprimersusingthesamedatasetofthe47full SSUsequencesusedtocompare5 9 -vs.3 9 -rDNAdiagnosticloci forourtemperaterainforeststudy.Assuspected,weobserved1– 3bpmismatchesfor60%ofnematodesspeciesacrossall phylogeneticclades.Incontrast,theprimingregionsofthe3 9 enddiagnosticlocusthatweusedinbothrainforeststudieswere extensivelytested( 2,000NCBISSUeukaryoticsequences coveringallphyla)andreportedtobeuniquelyconserved, particularlywithinNematoda[28].Whilethedevelopmentand availabilityofalternativeprimerstodiagnosticregionsisurgently needed,thesealternativeprimersmustbethoroughlyvettedto avoidgrosstaxonomicbiases. Asmentionedabove,methodologicalbiasesnotonlycanaffect theperceptionsofthegeneraldiversitypatterns,butmostimportantlycanaffecttheperceptionsofawidevarietyof ecologicalconceptsrangingfromcommunitycomposition,toroles ofspecificspeciesinecosystemfunctioning,tospeciesredundancy, orrelationshipsbetweendiversityinthebelowgroundandthe aboveground.Asshowninourstudy,thecompositionand structureofnematodecommunitiesdifferednotonlybetweenthe rainforestsbutalsoatthelocalscaleofindividualhabitats.Because oftheverylowoverlapofspeciesamonghabitatsandbetweenthe forests,itappearsthattherolesthatspeciesplaywithineach communityareverynarrowandaredefinedbythespecific conditionsofeachenvironment.Italsosuggeststhatspecies,even whenconsideredfunctionalequivalents(e.g.bacterial-feeders), maynotbefunctionallyidenticalandthusnot,asoftenassumed, redundant.Inthetropicalrainforest,forexample,adifferentsetof specieswerepartofthedecompositionprocess. Oscheius sp.A, Cephalobus sp.A,B,C(BF)and Aphelenchoides sp.A,B,C(FF)SpatialPatternsofNematodeDiversity PLOSONE|www.plosone.org7September2012|Volume7|Issue9|e44641

PAGE 8

appearedtobethekeyplayersinthecanopy,while Plectusparvus Oscheius sp.A, Myolaimus sp.A(BF)and Aphelenchoides sp.A,D,and Tylencholaimus sp.A(FF)werethekeyplayerswithinthelitter. Noneoftheabovespecies,however,participatedintheprocess withinthesoilenvironmentwith Oscheius sp.B(BF)fillingthisrole. Aswemovetoatemperaterainforest,thepatternwassimilarsuch thatthreecompletelynewsetsofspeciesweremostsignificant. Importantly,theguildsofspecieswerenotmerelysimple,closelyrelatedreplacements,butinsteadwerephylogeneticallydiverged lineagesthathaveevolvedoverpotentiallylongperiodoftimeto filltheseroles. Withnewmolecularapproaches,wearejustscratchingthe surfaceoftheecologyofmicroscopiccommunitiesandtheir significancetoecosystemfunctioning.Atemporalcomponentin ourstudywouldprobablyfurtherhighlightthespecificityandthe importanceofeachspeciesindifferentplacesatdifferenttimes indicatingthatdominanceandraritycanbefluid.Isbelletal.[29] studiedplantspeciesin17biodiversityexperimentsandwhile speciesgenerallyappearedredundantwhenconsideredunderone setofenvironmentalconditionsinthecontextofoneecosystem function,85%ofspecieswereneededtomaintainmultiple functionsatmultipleplacesandmultipletimes.Thissamepattern islikelytoemergefrommicroscopiccommunities.However,the relianceonadequatesamplinganddiagnosticlocicannotbe overemphasized.Theinverserelationshipbetweensoilorganisms andabovegroundplantdiversitysuggestedbyWuetal.[9]wasa likelyartifactofproblemsassociatedwithbothsamplingerrorand theselectivenatureoftheirprimers.Itisnosurprisetoobservelow diversityofsoilcommunitiesinhighdiversityconservationareas suchasCostaRicanrainforest,where80%ofmeiofaunalspecies residenotinthesoilbutintheaboveground.Fromourownstudy, Caenorhabditisbriggsae ,ararespecieswithknownwidespread distribution[19],wasexpectedtobefoundinbothforests.While itwasdetectedinasinglesoilsampleatLaSelva,itwasentirely absentfromtheOlympicForest.Knowingnowthatrottingfruit, notsoil,isthepreferredhabitat,itsabsenceinmostofoursamples isnotanenigma.Theseexamplesillustratethatecological concepts,suchasrelationshipsbetweenbelowgroundandabovegrounddiversity,havetotakeintoaccountthestructureand complexityofastudiedecosystem. Anothernotabledifferencebetweentherainforestswasthatin contrasttothetemperateregion,almosteverynematodegenusin thetropicalsystem(particularlywithinlitterandcanopy)was representedbyseveral(possiblyclosely-related)speciespotentially pointingout,aspredicted,toahigherresilienceofthetropicsthan temperateregionstoenvironmentaldisturbances.Thetraditional useofmorphologicalcharacterswouldlikelyfailtodistinguish thesesubtledifferencesandultimatelyresultinunderestimatesof speciesrichnessaswellasaninabilitytorecognizetheuniqueness ofeachcommunity.High-throughputsequencingallowsusto executediversityassessmentsfasterandcheaper,butmost importantlytoexaminethediversityofmicroscopicorganismsat thespecieslevelofresolution.Onlyatthespecieslevel,canwe appreciatethecommonalityofendemismvs.rarityofcosmopolitanism.Predictably,astheresolutiondeclines,communities becomemoreandmoresimilarandthepatternofcosmopolitanismfalselyappears.Similarrecentobservationsweremadefor nematode,rotifer,tardigrade,andfungaltaxa[8,19,22,30–31] wherecosmopolitan‘‘phenotypicspecies’’wereactuallyphylogeneticspeciescomplexes,andwhenfinallyindividuallyrecognized, theyshowedsignificantendemism.Whilethe3 9 -partofSSU performedreasonablywelltouncoverspeciesdiversityinour study,itundoubtedlyunderestimatedthetrueextentofendemism. TheSSUDNAhasoftenbeenshowntoofferlimitedresolutionfor closelyrelated/crypticspecies[32].Aswedevelopprimersof greatertaxonomicdiscrimination,e.g.COIprimers[33],anduse theminparalleltoSSU,wearelikelytoreinforcethemain conclusionsofthisstudy.Methods SamplingandExtractionInordertobeabletomakeadirectcomparisontoourresults fromCostaRicantropicalrainforest(allnecessarypermitswere obtainedtothisfieldstudy,seeacknowledgments),wefollowed similarprotocolsfornematodesamplingandextractions,DNA extraction,amplification,andsequencing,aswellassequencing tagprocessing.ExplicitdetailscanbefoundinPorazinskaetal. [16].Briefly,inSeptember2010wecollectedsoil,litterand canopysamplesfromatemperaterainforestattheOlympic NationalForestneartheLakeQuinault,WA.Sampleswere collectedat4locations(replicates)separatedfromeachotherby approximately100meters.Withineachlocation,notlargerthan 1500m2,4randomcanopytreesand4randomunderstorytrees wereselectedassamplingpoints(atotalof8/replicate).Onesoil (15cmdepth)andonelitter(anyorganicmaterialoverlyingthe soil)samplewascollectedfroma15cm 6 15cmareawithin1–2m awayfromthecanopyandtheunderstorytrees.Alleightsamples werecombinedtomakeuponecompositesoilandonecomposite littersamplepereachsamplinglocation.Acanopysamplewas madeupofepiphyticmaterial(e.g.lichen,moss,algae)presenton thesurfaceofstemsofcanopyandunderstorytrees.Eachtreewas sampledatthreeverticalpoints(baseofthetree,1mand2m abovethesoil)froma15 6 15-cmarea.Atotalof24subsamples(3 verticalpoints 6 8trees)werepooledtogethertoformone compositecanopysamplepereachsamplinglocation.Nospecific permitswererequiredforthisfieldstudy.Sampleswerestoredina coolerandtransportedtoOregonStateUniversityandUSGSin Corvallis,ORforimmediateprocessing. Toensuremaximumrecoveryofnematodes(andothersimilarly sizedfauna)fromdifferenthabitats(non-buoyantsoilvs.buoyant organicandplantmaterial),weusedtwodifferentextraction methodologies.Ahundredmlofsoilsubsamples(equivalentof 70–80g)wasprocessedusingsugarflotationandcentrifugation (basedonpassiveseparationduetodensitydifferencesof nematodesandsoilparticles)[34],and100mloflitterand canopymaterial(equivalentof15–30g)wereextractedusing Baermannfunnels(basedonactivemigrationofnematodes)[35]. Priortobeingplacedinfunnels,litterandcanopymaterialwas firstcutintosmallerpieces,mixed,and100mlsubsampleswere choppedinablenderin150mLofdeionizedwaterfor10s. Nematodeswerecollectedafter48-hr.Allextractednematodes werecountedimmediatelyforabundanceatthetrophicgroup levelunderaninvertedmicroscope,reducedto0.5ml,transferred intoZRBashingBeadLysisTubes(ZymoResearchCorp,Santa Ana,CA)andtransportedtotheUniversityofFloridaforDNA processing.DNAExtraction,Amplification,andSequencingZRTubeswereprocessedatmaximumspeedfor2minutesona Mini-BeadBeater(BioSpecProducts,Inc.Bartlesville,OK). GenomicDNAwasextractedusingZRSoilMicrobeDNAkit accordingtothemanufacturer’sprotocol.Similarlytothetropical samples,elutedDNAwasusedasaPCRtemplatefor amplificationofa 400bpdiagnosticregionwithin3 9 -partof SSUrDNA:NF1/18Sr2b[28,36].Inadditionto3 9 -partofthe SSU,adiagnosticregionwithinthe5 9 -partofSSUrDNA:F04/ R22[37]wasamplifiedusingthesameDNAtemplate(tropicalSpatialPatternsofNematodeDiversity PLOSONE|www.plosone.org8September2012|Volume7|Issue9|e44641

PAGE 9

samples3 9 -partofSSU,temperatesamples3 9 -and5 9 -partof SSU).PCRamplificationswereperformedfollowingprotocols describedelsewhere[17]usingMID-tagged(10nucleotides)fusion primersasopposedto2nucleotideMIDtagsthatwereusedfor tropicalsamples[16].AlltemperaterainforestmetageneticSSU samplesweresequencedontwo(toaccommodatetwodiagnostic regions)GenomeSequencesTitanium(Roche/454LifeSciences) half-plates(alongwithothersamples)attheInterdisciplinary CenterforBiotechnologyResearch(ICBR)attheUniversityof Florida,Gainesville,FL(tropicalsampleswererunonanearlier versionGSFLX).Earlierexperimentswithartificiallyassembled nematodecommunitiesestablishedthattheuseofasinglePCR reactionandasingleemulsionPCRandpyrosequencingrunwere sufficientforbothqualitativeandquantitativeanalysisofthe nematodecommunitycompositionandstructure[36].MetageneticSequenceProcessingGeneratedsequenceswereprocessedusinganOCTUPUS ( O perational C lustered T axonomic U nitsfor P arallel-tagged U ltra S equencing)bioinformaticspipeline[25]thathasbeenbenchmarkedagainstotherpipelinesusedforprokaryotes[17,38]. OCTUPUSscannedsequencesforqualityusingLucy-trimwith defaultparameters[39]andscreenedthemforaminimumlength of200bp,andthenbinnedthembytheirMIDtags.Sequences werethenclusteredtoOCTUs(OperationalClusteredTaxonomicUnits)at99%similarityusingMEGABLAST[40]and MUSCLEtogeneratealistof‘‘fixed’’OCTUs(anOCTU consensusachievedwhenanadditionofasequencingreadtoan OCTUgroupdoesnotresultanymoreinachangeoftheOCTU consensus).Thelevelof99%withinOCTUsimilaritywas determinedtobethemostappropriateforrecognizingthe relationshipbetweenOCTUsandputativespecies[41].Fixed OCTUswereblast-matched[42]againsttheNCBIdatabase, expandedbythenematodereferencesequencesfromourcontrol experiments[28,36]andnematodereferencesequencesfrom CostaRica[15].Thereferencesequencesweregeneratedby singlenematodePCRfollowedbySangersequencing.The similaritycut-offforidentifyingOCTUswassettonolessthan 90%.AllOCTUswereanalyzedforthepresenceofputative chimerasusingafrequencyandlengthdependantalgorithm incorporatedintotheOCTUPUSpipeline.Chimerataggingis referencedatabaseindependentandinsteadcomparesOCTUs againsteachother.Itisbasedontheassumption(asinother algorithmslikePerseusorUCHIME)thatchimericsequencesare lessfrequentthantheirparentalsequences.OCTUsequencesare comparedalongtheirtotallengths.Achimericsequenceis detectedwhentwosequencesinitiallymatchathighidentityon the5 9 -endbutdiffergreatlyonthe3 9 -endresultinginincomplete lengthmatch.Basedontheanalysisofcontroldatasetsfrom artificially-assemblednematodecommunities[43],allOCTUs withincompletelengthmatchof $ 10bpwereflaggedaschimeric. Consequently,allOCTUsflaggedaschimericwereremovedfrom theanalysisofnematodeOCTUs.AnalysesBecauseOCTUsgeneratedfromSSUrDNAbyultrasequencingarenotequivalenttospecies,OCTUspersewerenotusedfor theanalysisofbiodiversity.Instead,allhighqualitynonchimeric OCTUswerelinkedbacktoputativespeciesbyusingHead-Tail patternsidentifiedanddescribedfromartificiallyassembled nematodecommunities[41].Briefly,inmetageneticdatasets generatedfromSSUrDNAbyultrasequencing,asinglespeciesis usuallyrepresentedbyseriesofOCTUsandeachOCTUby multiplesequencingreads[41].ThemostfrequentOCTUofa species,Head,ischaracterizedbythehighestbioinformaticsscores resultingfromblast-matchingittothedatabasereference sequence,andlessabundantOCTUs,Tail,withslightlyvariant sequencingreadsbylowerscorestothesamematchingreference sequence.Whensortedbythescores,predictableHead-Tail patternsemerge.Thepresenceoftwo-threeHeadsofsimilar scoresandreadfrequency,ontheotherhand,indicatesthe presenceofcloselyrelatedorcrypticspecies[41].Toinferabout quantitativerelationships,allreadswithineachOCTU(Headand Tail)linkedtoaputativespeciesweresummeduptogenerate abundanceperspeciespersample.Nematodespecieswere groupedintolessresolvedtaxonomicgroupingssuchasgenera andfamilies,butalsointofunctionalguilds(bacterial-feeders, fungal-feeders,omnivores,plant-parasites,predators,rootassociates,andanimalparasites)followingYeates etal .[26].EstimateS [44]wasusedtocomputespeciesrichness(expectedandtotal predicted)[45],anddiversity(Shannon-Weaver)[46].Forrichness anddiversityestimateswithinhabitats,inputdataintoEstimateS consistedofamatrixofthelistofspeciesandtheirabundancesper everyreplicatewithineachhabitat(e.g.N=4forsoilinthe tropicalrainforest).Fortotalrichnessanddiversityacrossall habitatswithineachrainforest,allspeciesinallhabitatsinall replicateswereused(e.g.N=12fortropicalrainforest).EstimateS derivedrichnessanddiversityforeachsamplewerethenaveraged acrosseachhabitat(e.g.N=4forsoilintheTropicalrainforest)or acrosstheentirerainforest(N=12forthetropicalrainforest). Becauseofthepresenceofnoreadsformanyspeciesandveryhigh readnumbervariation,calculationsinEstimateSwereperformed ontransformed/normalizeddata(numbersofsequencingreads pereachputativespeciesweretransformedintonumbersof nematodeindividualsperputativespeciesusingguidelinesfrom controlexperimentswithartificially-assemblednematodecommunities)[36].Two-wayanalysisofvariance(ANOVA)wasusedto detectstatisticaldifferencesbetweenrainforestsandamong habitatsinspeciesrichness,diversity,andthenumberofreal individuals(density).Forsharedspeciesamonghabitatswithin eachrainforest(6intropicaland10intemperate),one-way analysisofvariancewasusedtodetectdifferencesinabundance. Becauseofhighvariance,abundancewaslog(x + 1)transformed priortoanalysis.TheStatistiXLdataanalysispackageasanAddIntoExcel2007wasusedforbothclusterandANOVAanalyses. Fromthe12metagenetictemperaterainforestsamples,three samples(L1,C2,C4)thatwereamplifiedonthe3 9 -endoftheSSU and2samples(L3andC3)thatwereamplifiedonthe5 9 -endof theSSUgeneratednooronlyafewsequencingreadsandwere thereforeremovedfromanalyses. Whileoursampling,extraction,andmetageneticsmethodologiesarefine-tunedfornematodetaxa,theyarenotselective againstothermicroscopiceukaryotes(e.g.mites,tardigrades, springtails).Because 50%ofthemetageneticdataconsistedof non-nematodesequences,theyarepresentedinthispaperaswell, althoughthesecouldbesubjecttosampling,extraction,and amplificationbiases.Allmethodsandanalyses,includingbioinformatics,werethesameasfornematodes.DataAccessibilityRaw454readdataalongwithmetadatadescribingspecific primersandMID-tagshavebeendepositedattheShortRead ArchiveattheNCBIunderthefollowingsubmissionnumbers: Study:SRSPO14451,Sample1:SRS350224(diagnosticlocus coveringthe3 9 -partofthe18S)andSample2:SRS350225 (diagnosticlocuscoveringthe5 9 -partofthe18S).SpatialPatternsofNematodeDiversity PLOSONE|www.plosone.org9September2012|Volume7|Issue9|e44641

PAGE 10

SupportingInformationFigureS1Acomparisonoftotalnumbersofmicro-and meio-faunalspeciesbetween3 9 -and5 9 -enddiagnostic lociinthetemperate(OlympicNationalForestinWA, U.S.A)rainforest. (XCF)FigureS2Averagediversityandabundancewithinsoil, litterandcanopyhabitatsandacrossallhabitats(Total) inthetropicalrainforestatLaSelvaBiologicalStationin CostaRica(LS),andthetemperaterainforestatthe OlympicNationalForestinWA,U.S.A.(OF). A)Richness (numberofspecies),B)diversity(Shannon),andC)Abundance (numberofnematodeindividualsper100cc).Barsindicate standarderrors. (TIF)FigureS3Percentofsharedspeciesamonghabitats (soil=S,litter=L,andcanopy=C)inthetropical rainforestatLaSelvaBiologicalStationinCostaRica (LS),andthetemperaterainforestattheOlympic NationalForestinWA,U.S.A.(OF). (TIF)FigureS4Acomparisonofoverallnematodeassemblagesbetweentemperate(OlympicForest,OF)and tropical(LaSelva,LS)rainforestsatthefamilylevelof taxonomicresolution. Familiesweregroupedbytheirtrophic guildsandsortedwithineachguildbytheirproportionate representation(highesttolowestwithinLS).BF=bacterial feeders,FF=fungalfeeders,RA=rootassociates,PP=plant parasites,OM=omnivores,PR=predators,AP=animal parasites,AL=algivores. (TIF)AcknowledgmentsWethankThomasNylen(GeologyDept.atPortlandStateUniversity)for helpincollectingsamplesandNicoleDeCrappeo(USGSForestand RangelandEcosystemScienceCenter,Corvallis,OR)andRussellIngham (Dept.ofBotanyandPlantPathologyatOregonStateUniversity, Corvallis,OR)forsharingtheirfacilitiesforprocessingofthesamples collectedattheOlympicNationalForest,WA.WealsothankAlejandro Esquivel(UniversidadNacionalatHeredia,CostaRica)forproviding logisticalsupportandsharingoffacilitiesinCostaRica.Wearealso gratefultoDeedraMcClearnforprovidingfacilitiesattheLaSelva BiologicalStation.WethankOrganizationforTropicalStudies(OTS)for accesstotheLaSelvaBiologicalStationforsampling,MaterialTransfer Agreementc-1-0620-0955fortransferofbiologicalmaterialsfromMar a delosAngelesMoraLo pez,holderofcollectingpermit001-2005-OTCONAGEBIO.NopermitswerenecessaryforsamplingattheOlympic NationalForest.AuthorContributionsConceivedanddesignedtheexperiments:DLPRMGDTOP.Performed theexperiments:DLPRMGDTOP.Analyzedthedata:DLP.Contributedreagents/materials/analysistools:RMGDWKT.Wrotethepaper: DLP.References1.BaasBeckingLGM(1934)Geobiologieofinleidingtotdemilieukunde.VPVan Stockkum&Zoon,TheHague,TheNetherlands. 2.FinlayBJ(2002)Globaldispersaloffree-livingmicrobialeukaryotespecies. Science296:1061–1063. 3.FenchelT,FinlayBJ(2004)Theubiquityofsmallspecies:patternsoflocaland globaldiversity.BioScience54:777–784. 4.FinlayBJ,CorlissJO,EstebanG,FenchelT(1996)Biodiversityatthemicrobial level:Thenumberoffree-livingciliatesinthebiosphere.QuartRevBiol71: 221–237. 5.FenchelT,EstebanGF,FinlayBJ(1997)Localvs.globaldiversityof microorganisms:Crypticdiversityofciliatedprotozoa.Oikos80:220–225. 6.MedingerR,NolteV,PandeyRV,JostS,Otternva ¨lderB,etal.(2010)Diversity inahiddenworld:potentialandlimitationofnextgenerationsequencingfor surveysofmoleculardiversityofeukaryoticmicroorganisms.MolEcol19:32– 40. 7.NolteV,PandeyRV,JostS,MedingerR,Ottenwa ¨lderB,etal.(2010) Contrastingseasonalnicheseparationbetweenrareandabundanttaxaconceals theextentofprotestdiversity.MolEcol19:2908–2915. 8.RobesonMS,KingAJ,FreemanKR,BirkyJrCW,MartinAP,etal.(2011)Soil rotifercommunitiesareextremelydiversegloballybutspatiallyautocorrelated locally.PNAS108:4406–4410. 9.WuT,AyresE,BardgettRD,WallDH,GareyJR(2011)Molecularstudyof worldwidedistributionanddiversityofsoilanimals.PNAS108:17720–17725. 10.BlaxterM,MannJ,ChapmanT,ThomasF,WhittonC,etal.(2005)Defining operationaltaxonomicunitsusingDNAbarcodedata.PhilosTransRSocLond BiolSci360:1935–1943. 11.BassettY(2001)Invertebratesinthecanopyoftropicalrainforests:Howmuch dowereallyknow?PlantEcol153:87–107. 12.WardleDA(2002)CommunitiesandEcosystems:LinkingtheAbove-ground andBelow-groundComponents.PrincetonUniversityPress,Princeton,NJ. 13.BoagB,YeatesGW(1998)Soilnematodebiodiversityinterrestrialecosystems. BiodivConserv7:617–630. 14.Amaral-ZettlerL,ArtigasLF,BarossJ,BharathiL,BoetiusA,etal.(2010)A globalcensusofmarinemicrobes.In:McIntyreA,editor.LifeintheWorld’s Oceans:Diversity,DistributionandAbundance.Oxford:BlackwellPublishing Ltd.223–45. 15.PowersTO,NeherDA,MullinP,EsquivelA,Giblin-DavisRM,etal.(2009) Tropicalnematodediversity:Verticalstratificationofnematodecommunitiesin aCostaRicanhumidlowlandtropicalrainforest.MolEcol18:985–996. 16.PorazinskaDL,Giblin-DavisRM,EsquivelA,PowersTO,SungW,etal.(2010) Ecometageneticsconfirmshightropicalrainforestnematodediversity. MolEcol 19:5521–5530. 17.BikHM,SungW,DeLeyP,BaldwinJG,SharmaJA,etal.(2011)Metagenetic communityanalysisofmicrobialeukaryotesilluminatesbiogeographicpatterns indeep-seaandshallowwatersediments.MolEcol21:1048–1059. 18.BeijerinckM(1918)DeInfusiesendeontdekkingderBackterien.Jaarboekvan deKoninklijkeAkademiev.Wetenschappen,Muller,Amsterdam,TheNetherlands. 19.KiontkeK,Fe lixMA,AilionM,RockmanMV,BraendleC,etal.(2011)A phylogenyandmolecularbarcodesfor Caenorhabditis ,withnumerousnewspecies fromrottingfruits.BMCEvolBiol11:339. 20.FontanetoD,FicetolaGF,AmbrosiniR,RicciC(2006)Patternsofdiversityin microscopicanimals:aretheycomparabletothoseinprotistsorinlarger animals?GlobalEcologyandBiogeography15:153–162. 21.FulthorpeRR,RoeschLFW,RivaA,TriplettEW(2008)Distantlysampledsoils carryfewspeciesincommon.TheISME2:901–910. 22.TaylorJW,TurnerE,TownsendJP,DettmanJR,JacobsonD(2006)Eukaryotic microbes,speciesrecognitionandgeographiclimitsofspecies:examplesfrom thekingdomFungi.PhilTransRSocB361:1947–1963. 23.NeherDA(2010)Ecologyofplantandfree-livingnematodesinnaturaland agriculturalsoil.AnnRevPhytopath48:371–94. 24.SollinsP,SanchoFM,MataRCh,SanfordRLJr.(1994)Soilsandsoilprocess research.In:McDadeLA,BawaKS,HespenheideHA,HartshornGS,editors. LaSelva:ecologyandnaturalhistoryofaneotropicalrainforest.Chicago:The UniversityofChicagoPress.34–53. 25.CreerS,FonsecaVG,PorazinskaDL,Giblin-DavisRM,SungW,etal.(2010) Ultrasequencingofthemeiofaunalbiosphere:practice,pitfallsandpromises. MolEcol19:4–20. 26.YeatesGW,BongersT,deGoedeRGM,FreckmanDW,GeorgievaSS(1993) Feedinghabitsinsoilnematodefamiliesandgenera–anoutlineforsoil ecologists.JNematol25:315–331. 27.DeLeyP(2006)Aquicktourofnematodediversityandthebackboneof nematodephylogeny, WormBook ,ed.The C.elegans ResearchCommunity, WormBook,doi/10.1895/wormbook.1.41.1. 28.PorazinskaDL,Giblin-DavisRM,FarmerieW,KanzakiN,MorrisK,etal. (2009)Evaluatinghighthroughputsequencingasamethodformetagenomic analysisofnematodediversity.MolEcolRes9:1439–1450. 29.IsbellF,CalcagnoV,HectorA,ConnollyJ,HarpoleWS,etal.(2011)High plantdiversityisneededtomaintainecosystemservices.Nature477:199–203. 30.SudhausW,KiontkeK(2007)Comparisonofthecrypticnematodespecies Caenorhabditisbrenneri sp.n.and C.remanei (Nematoda:Rhabditidae)withthestem speciespatternofthe CaenorhabditisElegans group.Zootaxa1456:45–62. 31.GuilN,Sa nchez-MorenoS,MachordomA(2009)Localbiodiversitypatternsin micrometazoans:Aretardigradeseverywhere?SystemBiodiver7:259–268. 32.DeryckeS,DeLeyP,TandinganDeLeyI,HolovachovO,RigauxA,etal. (2010)LinkingDNAsequencestomorphology:crypticdiversityandpopulationSpatialPatternsofNematodeDiversity PLOSONE|www.plosone.org10September2012|Volume7|Issue9|e44641

PAGE 11

geneticstructureinthemarinenematode Thoracostomatrachygaster (Nematoda, Leptosomatidae).ZoologicaScripta39:276–289. 33.DeryckeS,VanaverbekeJ,RigauxA,BackeljaouT,MoensT(2010)Exploring theuseofthecytochromeoxidasecsubunit1(COI)forDNAbarcodingoffreelivingmarinenematodes.PLoSONE5:10e13716.doi:10.1371/journal.pone. 0013716. 34.JenkinsWR(1964)Arapidcentrifugal-flotationtechniqueforseparating nematodesfromsoil.PlantDisReport48:692. 35.ViglierchioDR,SchmittRV(1983)Onthemethodologyofnematodeextraction fromfieldsamples:Baermannfunnelmodifications.JNematol15:438–444. 36.PorazinskaDL,SungW,Giblin-DavisRM,ThomasWK(2010)Reproducibility ofreadnumbersinhigh-throughputsequencinganalysisofnematode communitycompositionandstructure.MolEcolRes10:666–676. 37.BlaxterML,DeLeyP,GareyJR,LiuLX,ScheldemanP,etal.(1998)A molecularevolutionaryframeworkforthephylumNematoda.Nature392:71– 75. 38.FonsecaVG,CarvalhoGR,SungW,JohnsonHF,PowerDM,etal.(2010) NatureComms1,doi:10.1038/ncomms1095. 39.ChouH,HolmesM(2001)DNAsequencequalitytrimmingandvector removal.Bioinformatics12:1092–1104. 40.ZhangZ,SchwartzS,WagnerL,MillerW(2000)Agreedyalgorithmfor aligningDNAsequences.JComputBiol7:203–214. 41.PorazinskaDL,Giblin-DavisRM,SungW,ThomasWK(2010)Linking operationalclusteredtaxonomicunits(OCTU)fromparallelultrasequencing (PUS)tonematodespecies.Zootaxa2427:55–63. 42.AltschulSF,MaddenTL,SchafferAA,ZhangJ,ZhangZ,etal.(1997)Gapped BLASTandPSI-BLAST:anewgenerationofproteindatabasesearch programs.NuclAcidsRes25:3389–3402. 43.PorazinskaDL,Giblin-DavisRM,SungW,ThomasWK(2012)Thenatureand frequencyofchimerasineukaryoticmetageneticsamples.JNematol44:In press. 44.ColwellRK(2009)EstimateS:Statisticalestimationofspeciesrichnessand sharedspeciesfromsamples.Version8.2.User’sGuideandapplication publishedat:http://purl.oclc.org/estimates. 45.ChaoA,HwangW-H,ChenY-C,KuoC-Y(2000)Estimatingthenumberof sharedspeciesintwocommunities.StatisticaSinica10:227–246. 46.MagurranAE(2004)Measuringbiologicaldiversity.Oxford:Blackwell Publishing.SpatialPatternsofNematodeDiversity PLOSONE|www.plosone.org11September2012|Volume7|Issue9|e44641