Exploring the relationship between sampling efficiency and short range endemism for groundwater fauna in the Pilbara region, Western Australia

Citation
Exploring the relationship between sampling efficiency and short range endemism for groundwater fauna in the Pilbara region, Western Australia

Material Information

Title:
Exploring the relationship between sampling efficiency and short range endemism for groundwater fauna in the Pilbara region, Western Australia
Series Title:
Subterranean Ecology
Creator:
Barron, Harley J.
Cocking, James
Eberhard, Stefan
Halse, Stuart A.
Scanlon, Michael D.
Williams, Matthew R.
Publisher:
Subterranean Ecology, Scientific Environmental Services
Publication Date:
Language:
English

Subjects

Subjects / Keywords:
Biology ( local )
Cave Ecology ( local )
Genre:
Article
serial ( sobekcm )
Location:
Australia

Notes

General Note:
1. Identifying the existence of short or narrow range endemic species is an important issue when planning for conservation of groundwater fauna in the face of threats to groundwater quantity and quality. 2. Fourteen bores were sampled six times over 3 or 4 years to assess the reliability of nethauling sampling in broad-scale survey to collect the groundwater fauna present at a site and to identify short-range endemic (SRE) species. 3. Species accumulation curves suggested that one sample from a bore collected 23% and 46% of species occurring in low and high abundance, respectively, and two samples collected 38% and 65% of such species. False-negative rates provided a slightly higher estimate of the collection probability of species with low abundances. 4. The frequent failure to collect species present at a site means that some apparent shortrange endemism was probably an artefact of low sampling effort. Nevertheless, as is typical for subterranean fauna, a high proportion of the known species in the Pilbara region appeared to be SREs. About 55% had probable ranges 10 000 km2, the criterion proposed by Harvey (2002) for short-range endemism. 5. Consideration of species occurrence patterns, natural barriers and the scale of most disturbances suggest that 1000 km2 is a more satisfactory threshold for short-range endemism than 10 000 km2 but, as the threshold is reduced, more intensive sampling is required to determine whether a species qualifies as an SRE. 6. Extrapolation of the results of regional sampling suggested the Pilbara contains about 500â€"550 species of groundwater fauna, with the density of species being relatively uniform across the region. Attempts to use a T-S curve approach (sensu Ugland Gray, 2004) highlighted the lack of information about within-population dispersal -- Authors
Restriction:
Open Access - Permission by Publisher
General Note:
See Extended description for more information.

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
K26-01382 ( USFLDC DOI )
k26.1382 ( USFLDC Handle )
5991 ( karstportal - original NodeID )

USFLDC Membership

Aggregations:
Added automatically
Karst Information Portal

Postcard Information

Format:
serial

Downloads

This item has the following downloads:


Full Text
Description
1. Identifying the
existence of short or narrow range endemic species is an
important issue when planning for conservation of groundwater
fauna in the face of threats to groundwater quantity and
quality. 2. Fourteen bores were sampled six times over 3 or 4
years to assess the reliability of nethauling sampling in
broad-scale survey to collect the groundwater fauna present at
a site and to identify short-range endemic (SRE) species. 3.
Species accumulation curves suggested that one sample from a
bore collected 23% and 46% of species occurring in low and high
abundance, respectively, and two samples collected 38% and 65%
of such species. False-negative rates provided a slightly
higher estimate of the collection probability of species with
low abundances. 4. The frequent failure to collect species
present at a site means that some apparent shortrange endemism
was probably an artefact of low sampling effort. Nevertheless,
as is typical for subterranean fauna, a high proportion of the
known species in the Pilbara region appeared to be SREs. About
55% had probable ranges <10 000 km2, the criterion proposed
by Harvey (2002) for short-range endemism. 5. Consideration of
species occurrence patterns, natural barriers and the scale of
most disturbances suggest that 1000 km2 is a more satisfactory
threshold for short-range endemism than 10 000 km2 but, as the
threshold is reduced, more intensive sampling is required to
determine whether a species qualifies as an SRE. 6.
Extrapolation of the results of regional sampling suggested the
Pilbara contains about 500"550 species of groundwater fauna,
with the density of species being relatively uniform across the
region. Attempts to use a T-S curve approach (sensu Ugland
& Gray, 2004) highlighted the lack of information about
within-population dispersal --
Authors



PAGE 1

Exploringtherelationshipbetweensamplingefciency andshort-rangeendemismforgroundwaterfaunainthe Pilbararegion,WesternAustraliaSTEFANM.EBERHARD*,1,STUARTA.HALSE*,2,MATTHEWR.WILLIAMS†, MICHAELD.SCANLON*,2,JAMESCOCKING*,2ANDHARLEYJ.BARRON*,3* DepartmentofEnvironmentandConservation,ScienceDivision,Wanneroo,WA,Australia†DepartmentofEnvironmentandConservation,ScienceDivision,BentleyDeliveryCentre,WA,AustraliaSUMMARY 1.Identifyingtheexistenceofshortornarrowrangeendemicspeciesisanimportantissue whenplanningforconservationofgroundwaterfaunainthefaceofthreatsto groundwaterquantityandquality. 2.Fourteenboresweresampledsixtimesover3or4yearstoassessthereliabilityofnethaulingsamplinginbroad-scalesurveytocollectthegroundwaterfaunapresentatasite andtoidentifyshort-rangeendemic(SRE)species. 3.Speciesaccumulationcurvessuggestedthatonesamplefromaborecollected23 % and 46 % ofspeciesoccurringinlowandhighabundance,respectively,andtwosamples collected38 % and65 % ofsuchspecies.False-negativeratesprovidedaslightlyhigher estimateofthecollectionprobabilityofspecieswithlowabundances. 4.Thefrequentfailuretocollectspeciespresentatasitemeansthatsomeapparentshortrangeendemismwasprobablyanartefactoflowsamplingeffort.Nevertheless,asis typicalforsubterraneanfauna,ahighproportionoftheknownspeciesinthePilbara regionappearedtobeSREs.About55 % hadprobableranges<10000km2,thecriterion proposedbyHarvey(2002)forshort-rangeendemism. 5.Considerationofspeciesoccurrencepatterns,naturalbarriersandthescaleofmost disturbancessuggestthat1000km2isamoresatisfactorythresholdforshort-range endemismthan10000km2but,asthethresholdisreduced,moreintensivesamplingis requiredtodeterminewhetheraspeciesqualiesasanSRE. 6.ExtrapolationoftheresultsofregionalsamplingsuggestedthePilbaracontainsabout 500–550speciesofgroundwaterfauna,withthedensityofspeciesbeingrelativelyuniform acrosstheregion.AttemptstouseaT-Scurveapproach( sensu Ugland&Gray,2004) highlightedthelackofinformationaboutwithin-populationdispersalofthesespeciesand theareaofanaquiferthatiseffectivelysampledbyabore.Keywords :falsenegative,narrowrangeendemism,speciesaccumulation,stygofauna,surveydesignIntroductionThetaskofconservingbiodiversityinthefaceof increasingthreatfromhumanactivitiesisachallenge tobiologistsworldwide(Danielopol etal. ,2003;Mace etal. ,2005;Dudgeon etal. ,2006).Oneofthecentral tenetsofconservationisthatallspeciesshouldbe preventedfromextinctionandthereismuchlegislationtosupportthisaspirationatinternational, Correspondence:Dr.StuartA.Halse,BennelongiaPtyLtd., POBox384,Wembley,WA6913,Australia. E-mail:stuart.halse@bennelongia.com.au1Presentaddress:SubterraneanEcology,Scientic EnvironmentalServices,Greenwood,WA,Australia.2Presentaddress:BennelongiaPtyLtd,Wembley,WA,Australia.3Presentaddress:Barron,25DurackWay,Padbury,WA, Australia. FreshwaterBiology (2009) 54, 885–901doi:10.1111/j.1365-2427.2007.01863.x 2007TheAuthors,Journalcompilation 2007BlackwellPublishingLtd885

PAGE 2

nationalandlocalscales(Caughley&Gunn,1996; Krishnamurthy,2003).However,legislationrequires theexistenceofaspeciestobedocumented,andits conservationstatusassessed,beforeprotectionoccurs (e.g.IUCNRedListofThreatenedSpecies).The conventionalapproachtosuchassessmentinvolves samplinganumberoflocationstoproduceapresence–absencematrixofspeciesoccurrenceacross spatialunits.Inferencesarethendrawnaboutspecies distributions,abundancesandhabitatpreferences, withconsequentialdecisionsaboutconservationstatus(e.g.Mace,1995;Paran etal. ,2005;Dole-Olivier etal. ,2009).Theobviousimportanceofreliablesurvey datainreachingappropriateconservationdecisions hasledtoconsiderableinterestinquantifyingthe errorsassociatedwithspeciesdetection. Oneapproachtocalculatingdetectionerrorsuses speciesaccumulationcurvestomeasuretherelationshipbetweensamplingeffortandspeciesdetection (seeColwell&Coddington,1994;Colwell,Mao& Chang,2004).Morerecently,focushasshiftedto explicitcalculationoftheprobabilityoffailingto collectaspecieswheninfactitispresent(MacKenzie etal. ,2002;Tyre etal. ,2003).Suchfalse-negative(FN) recordsleadtounderestimatesofspecies’rangesand overestimatesofextinctionprobabilities. Severalstudieshavedocumentederrorsassociated withsamplinginterstitialandgroundwaterspecies (e.g.Rouch&Danielopol,1997;Mauclaire,Marmonier&Gibert,1998;Pipan&Culver,2005).Theyshow thatsamplingerrorpreventseasytranslationofthe resultsofmostsurveysintoconservationplanning (Castellarini etal. ,2007a,b)butprovidelittleinformationaboutthefactorsaffectingspeciesdetectability, norwhetherthereisanyrelationshipbetweendetectabilityandconservationstatus. Locallyrestrictedspeciestendtohavehighconservationstatusbecausetheyaremorevulnerableto extinction,followinghabitatdestructionorenvironmentalchange,thanarewidespreadspecies(Ponder &Colgan,2002).Themoreextremeexamplesof locallyrestrictedspeciesarereferredtoasnarrowrangeorshort-rangeendemics(SREs).Harvey(2002) denedSREsasthosespecieswithdistributions covering<10000km2.Subterraneanfaunasusually containhigherproportionsofSREsthannearby surfacecommunities(Gibert&Deharveng,2002)so thatissuesassociatedwithconservationofSREsare particularlyimportantbelowground. IdentifyingSREsisdifcultandmanyhighlyvisible plantspeciessuspectedtocomprisesmalllocalised populationsremainclassiedinformalconservation listsas‘datadecient’,ratherthanbeingassignedtoa categoryofdistribution,becausesurveyeffortis consideredinadequate(Coates&Atkins,2001).Paradoxically,despitecrypticoccurrenceandmuchless beingknownaboutinvertebratebiologyanddistributions,SREstatusisoftenreadilyinferredfor invertebratespeciesknownonlyfromasinglesite. Thechanceoferrorishigh,however,whensuch assessmentsarebasedononlyone,orfew,surveysof theregion.Manyofthespeciesrecordedatasingle sitemaybewidespreadbutoccupyingpoorlysampledhabitats,beatthelimitofabroaderdistribution contiguouswiththeareassurveyed,oroccuronly sporadically(seeHalse etal. ,2000;Pinder etal. ,2004). ThePilbararegioninnorth-westernAustraliacontainstherichestknowngroundwaterfaunainAustralia,withupto54speciesatindividualboresanda totalofabout350speciesrecorded(Eberhard,Halse& Humphreys,2005a;S.A.Halse etal. ,unpubl.data). Althoughthefaunaisstillbeingdocumented,itis apparentthatthePilbaracontainsgloballysignicant numbersofgroundwaterspecies(seeCulver&Sket, 2000;Gibert&Deharveng,2002).ThePilbaraalso containsthelargestconcentrationofmininginAustralia,withmuchofitoccurringinopenpitsthat extendbelowthewatertableandrequirede-watering (Johnson&Wright,2001).Thus,thereispotentialfor substantialconictbetweenminingandtheconservationrequirementsofgroundwaterfauna(Boulton, Humphreys&Eberhard,2003;Humphreys,Watts& Bradbury,2005). Earlysparse,somewhatclusteredsamplingof groundwaterfaunainthePilbaraidentiedmany SREs(e.g.Bradbury,2000)andstronglysuggested thatasystematic,broad-scalesurveywasneededto provideaframeworkformining,groundwateruse andfaunaconservation.Thus,a4-yearsurveyofthe Pilbarabeganin2002withtheaimsof(i)mapping regionalpatternsofdiversityofthegroundwater fauna;(ii)identifyingspeciesofconservationsignicance(mostlySREs)and(iii)relatingdiversityof groundwaterfaunatoenvironmentalparameterssuch asgeologyandwaterchemistry(Eberhard etal. 2005b). Thispaperreportstheresultsofintensivesampling, undertakenatselectedsiteswithinthePilbarasurvey,886 S.M.Eberhard etal. 2007TheAuthors,Journalcompilation 2007BlackwellPublishingLtd, FreshwaterBiology 54, 885–901

PAGE 3

toexplorethevalidityofbiodiversitypatterns obtainedfromlowerintensitysamplinginthe regionalsurveyasawhole.Specicobjectiveswere (i)todeterminetheprobabilityofspeciespresentata sitebeingretrievedinasinglesamplingevent;(ii)to determinewhetherthatprobabilitywasaffectedby speciesabundanceand(iii)toexaminewhetherlow samplingintensitymayhavecontributedtothehigh proportionofsitesingletons(andinferredSREs)in Pilbaragroundwater.MethodsThesemi-aridPilbararegionisageologicallycomplex andancientlandscape,consistingofvemajorcatchmentsandcoveringanareaofapproximately 178000km2(Fig.1).Groundwaterispredominantly fresh(totaldissolvedsolids<3000mgL) 1),occurring inunconsolidatedalluvium,chemicallydeposited sedimentswithhighsecondaryporosity(calcrete andpisoliticlimonite)andfracturedrocks.Groundwaterfaunaoccursinallthesedeepergroundwater environments,aswellasinshallowgroundwaterin springsandthehyporheos(Halse,Scanlon&Cocking, 2002;Eberhard etal. ,2005a). ThePilbaracontains>3700boresandwells.Atotal of424weresampledtwiceduringthePilbarasurvey, onceafterthewetseason(April–July)andonce towardstheendofthedryseason(August–October) (Eberhard etal. ,2005b).Boresandwellsweresampled bydroppingaweightedphreatobiologicalnettothe bottomofthewatercolumn,agitatingthenetto disturbbottomsediment,andthenretrievingthenet. Netsofvaryingdiameterwereused,accordingtosize oftheboreorwell.AMcCartneyvialwasttedto eachnet,withthebaseofthevialgroundoffand replacedwith50l mmeshscreentoimprovewater owthroughnetsastheywerehauledup. Eachnet-haulsamplingeventconsistedofdropping andretrievingnetssixtimes:therstthreehaulswere madewitha150l mmeshnetprincipallytocatch macrofaunaandthesecondthreehaulsweremade witha50l mmeshnettocapturemicrofauna.After sampling,netswerewashedinadecontaminant(5 % Fig.1 Pilbararegionshowingthevemajorcatchmentsandlocationsof14net-sampling(circles)and13combinationbores(triangles).(1)AshburtonRiverBasin,(2)DeGreyRiverBasin,(3)FortescueRiverBasin,(4)PortHedlandCoastBasin,(5)OnslowCoast Basin.Samplingefciencyandgroundwaterfaunadistributions 887 2007TheAuthors,Journalcompilation 2007BlackwellPublishingLtd, FreshwaterBiology 54, 885–901

PAGE 4

solutionofDecon90;BactoLaboratories,Sydney Australia)andrinsedwithdistilledwatertoreduce thepossibilityoffaunalcontaminationbetweensamplingsites.NetsamplingThedatasetonwhichmostofthepresentedanalyses arebasedcamefrom14boresthatweresampledby nethaulingonsixoccasions(threewetseasonand threedryseasoncollections)overa3-or4-yearperiod (2002–05)(Table1).Thebores,whichwerechosen afterinitialsamplinghadoccurredatmanysites acrossthePilbara,representedarangeofhydrogeologicalconditionsandassociatedgroundwaterfauna. Theycoveredallvemajorcatchmentsandboth coastalandinlandsettings(Fig.1).CombinationsamplingTheeffectivenessofnet-haulsamplingwasfurther investigatedat13boresbycombinationsampling, whichconsistedofanet-haulsamplingeventfollowedimmediatelybypumpingthreetimesthebore volumeofgroundwaterthrougha50 l mnet.Assoon astheborerelled,anothersetofnethaulswastaken. Whencalculatingsamplingeffort,combinationsamplingwasconsideredtoconsistofthreesampling events(onepumpsampleandtwonethauls). Elevenofthecombinationsamplingboreswere locatedinalluvialaquifersnearthecoast;theother twowereinland(Fig.1andTable1).Combination samplingwasundertakentwiceatanintervalof 2yearsatsixoftheboresandonceattheremaining sevenbores.Alltheboresweresampledwithnets onlyonatleasttwootheroccasionsaspartofthe standardPilbarasurveyprotocol.SampleprocessingandidenticationEachtimethenetwaspulledtothesurface,contents oftheMcCartneyvialweretransferredtoa120mL polycarbonatecontainer.Oncompletionofthesixnet hauls,waterwasdrainedoffandthesamplewas preservedin100 % analyticalgradeethanol.Samples ofanimalscollectedinpumpwaterweresimilarly preserved. Priortosortingspecimensunderadissecting microscope,sampleswereseparatedintothreesize fractionsinthelaboratorybysievingthrough250,90 and53 l mmetalEndecottsieves(EndecottLtd, London,U.K.).Allanimalswereidentiedtothe lowesttaxonomicalrankpossibleusingpublishedand informalkeys,andthenumbersofindividualsof eachtaxonwererecorded.Identicationfrequently requireddissectionandexaminationunderacompoundmicroscope.AllostracodswereidentiedbyI. KaranovicorJ.ReevesandT.Karanovicidentied copepodscollectedin2002and2003.AnalysesSpeciesabundancewascategorisedintwowaysfor thepurposeofcalculatingspeciesaccumulationrates. First,foreachspeciesthemeannumberofindividuals retrievedfromallsamplesinwhichthespecies occurredwascalculatedandspeciesinthelowest50 percentilesofabundanceweredesignated‘rare’and others‘abundant’.Inasecondanalysis,eachspecies wasassignedtoanabundancecategoryforeachbore, basedonthespecies’averageabundanceonlyinthe samplesfromthatboreinwhichitoccurred(rare £ 3 animalsandabundant>3animals).Aspecieswas sometimesclassiedasrareatoneboreandabundant atanother. Accumulationcurvesateachboreforrarespecies, abundantspeciesandallspeciesweregenerated usingColwell’s(2005)ESTIMATES ESTIMATESsoftware(version 7.5.1).Thetotalnumberofspeciesateachborewas estimatedusingtheChao2estimator(orICEifthe coefcientofvariationforincidencewas>0.5andthe ICEestimatewashigher)becauseofthepatchynature ofspeciesrecoverythroughtime(Colwell&Coddington,1994;Foggo etal. ,2003).Resultsfromallbores werecombinedtoexaminethegeneralpatternof accumulationofrareandabundantspecies,although itshouldbeemphasisedthatratesinindividualbores wereheterogeneousbecauseofvariationingeology anddifferencesinthebiologyandbehaviourofthe particularspeciespresent. False-negativeratesforabundantspecies,rare speciesandallspecies(basedonabundanceinall net-sampledbores)werecalculatedforeachborefrom thesixsurveysas FN 1 no= 6 S 1 where noissumofoccurrencesofallspeciesatthe boreand S isnumberofspeciesrecorded.Ifallspecies888 S.M.Eberhard etal. 2007TheAuthors,Journalcompilation 2007BlackwellPublishingLtd, FreshwaterBiology 54, 885–901

PAGE 5

recordedataborewerecollectedduringmost samplingevents, nowillapproach6 S andtheFNrate willbelow.Animportantassumptionisthatall speciespresentattheborewerecollectedsometime duringthesequenceofsamplingevents.Aswith estimatesofspeciesaccumulationrates,theFNrates ofallintensivelysampledboreswereaveragedto estimateoverallFNratesforabundant,rareandall species.UsingFNrates,theproportionofspeciesata borecollectedbyseveralsampleswascalculatedas f 1 FNk 2 where k isnumberofsamplestaken. Speciesaccumulationratesduringtherstthree samplingevents(i.e.twonethaulsandonepump sample)atcombinationbores,whichtookplaceovera fewhours,werecomparedwithratesforthreeseasonallyseparatedsamplingeventsatrepeatborestoassess theextentofspeciesturnoverbetweenseasons.If turnoveroccurred,ahigherproportionofspecies wouldbeexpectedintherstsamplefromcombinationthanrepeatbores,assuminganyincreased efciencyofpumpsamplingwaslessthantheextent ofseasonalturnover.Inafurthertestofwhether groundwatercommunitiesexhibitedseasonalchange, variationsinspeciesrichnessandabundanceat net-sampledboreswereexaminedusingrepeatedmeasuresANOVA ANOVA.Abundancedatawerelogarithmicallytransformed[log( x +1)]toensureapproximate normalityandhomoscedasticityofresiduals;species richnessdatadidnotrequiretransformation. ValidityofthehypothesisthatmanyapparentSREs detectedduringthePilbara-widesamplingprogram weresamplingartefacts,becauseinfacttheyhave muchlargerrangesthansuggestedbysampling results,wasexaminedbycomparingspatialoccurrencesofsiteswithspeciesrecordedatonlytwoor threesites(sitedoubletonsandtripletons,respectively) withthespatialdistributionofsamplingsites.Data from397boresandwellssampledtwicewereusedin thisanalysis;thenet-sampledandcombinationsamplingboreswereexcluded.Thedistancebetween recordsofanSREspeciesshouldbeclosertothe minimumdistancebetweenbores,reectinglocalised distribution,thanoverallborespacing(i.e.theaverage distancebetweeneachboreandeveryotherbore). Thetotalnumberofspeciesofgroundwaterfauna inthePilbarawasestimatedusingtheICEestimator withinESTIMATES ESTIMATES(seeabove)andthemorerecently derivedT-Scurvetechnique(Ugland,Gray&Ellingsen,2003;Ugland&Gray,2004).Themaximum numberofsitesthatcouldbeprocessedinthe softwareavailabletocalculateT-Scurveswas240, soregionalsurveyboreswherenospecieswas recordedwereomittedandadditionalsiteswere randomlydroppeduntilonly240boresorwellsand 239speciesremainedinthedataset.Extrapolations withESTIMATES ESTIMATESweremadeusingboththisreduced datasetandtheregionaldatasetof397boresand wells.ForcalculationofT-Scurves,siteswerestratiedaccordingtocatchment.Results NetsamplingNinety-threespeciesbelongingto10highertaxonomicgroupswerecollected:Crustacea(69species), Oligochaeta(10),Nematoda(3),Arachnida(3),Rotifera(2),Gastropoda(2),Aphanoneura(1),Polychaeta (1),Hirudinea(1)andTurbellaria(1).Sixordersof Crustaceawerecollected:Ostracoda(32species), Copepoda(20),Amphipoda(7),Isopoda(5),Syncarida(4)andThermosbanacea(1).Cumulativespecies richnessatindividualboresrangedfrom0to36 (123;meanSE).Thetotalnumberofanimals collectedperborerangedfrom0to1402(7624)and thenumberofanimalsperspecieswas3310.No animalwascollectedfromborePSS056.Thethree bores(PSS003,PSS016andPSS058)withrelatively highspeciesrichness(>20species)alsohadhigh numbersofanimals(350–475)butthesitewithmost animals(PSS032)hadonly12species. Morespecieswerecollectedinsmallthanlarge numbersandtheabundancedistributionwasoverdispersed(Fig.2).Twenty-fourspecieswererepresentedbyonlyoneanimalinthesample(s)inwhich theyoccurredand69specieswererepresentedby £ 5 animalspersample.Forty-sevenpercentofallspecies werefoundinonlyonesample(herereferredtoas samplesingletons),24 % wererecordedinonlytwo samples(sampledoubletons)andonly29 % were recordedmorethantwice(Fig.3a).Samplesingletons wereusuallyrepresentedbyfeweranimalsina samplethansampledoubletonsandothermore frequentlyoccurringspecies(Fig.3b). Unsurprisingly,thespeciesthatoccurredinhigh numbersweremostlycollectedearlierinthesamplingSamplingefciencyandgroundwaterfaunadistributions 889 2007TheAuthors,Journalcompilation 2007BlackwellPublishingLtd, FreshwaterBiology 54, 885–901

PAGE 6

Table1 Boresandwellssampled,aquifercharacteristicsincludingboredepth(metresbelowgroundlevel),standingwaterlevel(SWL),numberofsamplinge vents(netsampling orcombinationsampling),geologyandselectedphysicochemicalparametersat ) 1mSWL(meanofsamplingevents) Hydrographic basin(catchment areainkm2)AquifernameGeology Sampling type No. events Bore code Species richness Bore depth (mbgl) SWL (mbgl) Temperature ( C) pH range Salinity (mgL) 1) DO (mgL) 1) 1Ashburton River(78893) DuckCreekColluvium: unconsolidated sandandgravel Repeat6PSS172167631.46.5–6.96782.0 TureeCreekAlluvium:uvial sand,silt andgravel 6PSS0560493931.56.6–7.54944.2 6PSS0582710630.36.9–7.26603.3 2DeGrey River(56717) PearCreekAlluvium:silt, sandandgravel withcalcrete 6PSS140954532.06.4–7.41000.2 WestStrelley River Alluvium:silt, sandandgravel 6PSS0321251531.95.7–7.25043.1 3Fortescue River(49232) EthelCreekAlluvium:silt, sandandgravel withcalcrete 6PSS0032223327.96.6–8.08441.5 WeeliWolli Creek Alluvium: unconsolidatedsilt, sand,graveland cobblesoverlying fractured-rock (Brockmaniron formation) 6PSS006722328.06.7–7.22851.6 6PSS009234426.16.9–7.32951.8 Warp2Alluvium: unconsolidated silt,sandand gravel 6PSS0445841528.16.9–7.43484.4 4PortHedland Coast(35172) BallaBalla River Calcrete6PSS0272461030.86.4–7.04651.2 TabbaTabba Creek Alluvium:silt, sandandgravel 6PSS0251216631.96.4–7.76271.5 5Onslow Coast(15689) CaneRiverAlluvium:clay,sand, siltandgravel partlycalcreted 6PSS0864301031.27.1–8.02573.8 Yarraloola Well Unconsolidated uviatiledeposits 6PSS08817541631.06.0–6.32770.6 RobeRiverAlluvium,calcrete andlimestone 6PSS0163613631.16.7–7.14804.1890 S.M.Eberhard etal. 2007TheAuthors,Journalcompilation 2007BlackwellPublishingLtd, FreshwaterBiology 54, 885–901

PAGE 7

sequenceataborethanthoseoccurringinlow numbers,reectingastrongrelationshipbetween abundanceanddetectability(Fig.4).BasedonacrossboreabundancecategoriesandChao2estimatesofthe truenumbersofspeciesateachbore,therst samplingeventcollectedonly236 % ofrarespecies presentatasite,467 % ofabundantspeciesand 335 % ofallspecies,whilesixsamplingevents collected7922 % ,9216 % and8216,respectively(Fig.4a).Theeffectofabundancewaseven morepronouncedwhenwithin-boreabundancecategorieswereused(Fig.4b). AnalysisofFNratesprovidedasimilarpicture, althoughtheyoverestimatedtheefciencywithwhich rarespecieswererecovered.Theprobabilityof collectingararespeciesinasinglesamplewas 363 % ,forabundantspecies503 % andforall species393 % .FNratessuggestedsixsamples wouldprobablycollect95 % ofallspecies.The discrepanciesbetweenFNandspeciesaccumulation estimatesweremainlytheresultofFNcalculations overestimatingtherateofaccumulationofrarespecies.CombinationsamplingResultsfromcombinationboressuggestedtherewas nosignicantseasonalturnoverinspeciescompositionatasite.Therstnet-haulingeventatthesebores collectedasmallerproportion(<16 % )ofallspecies collectedbythenet-pump-netsamplesthandidthe rstofthreenet-haulingeventsindifferentseasonsat net-sampledsites(438 % versus513 % ).The probablereasonforthersteventatcombination boresyieldingalower,ratherthansimilar,proportion ofspeciestothatobtainedatnet-samplingsiteswas thatpumpsamplingwasmoreefcientandinated thetotalspecieslistatcombinationbores.Pumping collectedonaverage6.81.5speciescomparedwith 5.41.2fromanet-haulingeventatthesamebores, althoughthedifferencewasnotsignicant( P 0.2, paired t -test, n 18)(butseeHancock&Boulton, 2009). InclusionofborePSS0016inbothnetandcombinationsamplingmeantthat11samplingevents occurredatthissite,whichallowedpredictionsbased onspeciesaccumulationcurvesandFNratestobe tested.Cumulativespeciesrichnessappearedto stabiliseafter10sampleevents(Fig.5),whichwas Table1 ( Continued ) Hydrographic basin(catchment areainkm2)AquifernameGeology Sampling type No. events Bore code Species richness Bore depth (mbgl) SWL (mbgl) Temperature ( C) pH range Salinity (mgL) 1) DO (mgL) 1) 5OnslowCoastRobeRiverAlluvium,calcrete andlimestone Purge2PSS0165013631.16.7–7.14804.1 2PSS0152523831.86.3–7.27424.9 2PSS0171216631.06.7–7.28404.3 1PSS072828731.87.1–7.45404.0 1PSS0751020630.76.8–7.18253.8 3Fortescue River Lower Fortescue River Alluvium,calcrete, limestoneand conglomerate Purge2PSS012249729.58.1–8.91740.4 2PSS0131425731.16.4–6.84462.2 1PSS076470930.77.7–8.82780.4 1PSS0771120930.76.8–7.15354.1 2PSS0781525830.66.8–6.911085.2 1PSS4471316930.76.7–6.93834.4 Coondiner Creek Alluviumand colluvium:sand andclay 1PSS5032571130.36.7–7.05903.3 Marillana Creek Pisoliticlimonite (vuggyporosity) 1PSS5041162–29.66.4–6.54103.4Samplingefciencyandgroundwaterfaunadistributions 891 2007TheAuthors,Journalcompilation 2007BlackwellPublishingLtd, FreshwaterBiology 54, 885–901

PAGE 8

0510152025303540451 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93Mean no. individuals per sample Fig.2 Distributionofanimalabundancesamongspeciesfromnet-sampledbores.Meanabundancewascalculatedforeachspecies onlyfromsamplesinwhichthespecieswaspresent.SeeAppendixforspeciesnamesaccordingtonumberson y -axis.892 S.M.Eberhard etal. 2007TheAuthors,Journalcompilation 2007BlackwellPublishingLtd, FreshwaterBiology 54, 885–901

PAGE 9

ingeneralagreementwithpredictionsofthespecies accumulationcurves(Fig.4)otherthanthatthe actualnumberofspeciescollectedwas15 % higher thanthepredictiongeneratedbyChao2aftersix samplingevents.AbundanceandrichnesspatternsAnalysisofpatternsoftotalanimalabundanceatthe net-samplingboresshowedstrongdifferencesbetween sitesbutnosignicantdifferencesbetweentimesof yearsampled(Table2).Thesamelackofresponseto seasonwasapparentinspeciesrichness(resultsnot shown).DistributionalpatternsinthePilbaraSamplingthroughoutthePilbaraidentiedabout350 species.Theproportionsofspeciesrepresentedby samplesingletonsanddoubletonswere37 % and 20 % ,respectively,andtheproportionsofspecies recordedatonlyone,two,threeormoreboreswere 44 % ,20 % ,10 % and26 % For29ofthe50speciesrepresentedbysitedoubletons,bothcollectinglocationswerewithinthesame catchment(Table3).However,eventhese29species hadaveragedistancesbetweenoccurrencesthatwere >50 % oftheaveragedistanceofanyboretoanyother withinthesamecatchment,whichsuggeststhatmany ofthespecieshadcatchment-widerangesthatdonot twithlocaliseddistributionsexpectedofSREs.A similarpatternwasobtainedforsitetripletons, althoughonlysevenofthe25speciesrecordedat onlythreeboresappearedtoberestrictedtosingle catchments(Table3). Despitetherelativelyweakevidenceoflocalised occurrenceamongsitedoubletonsandtripletons,28of 0 20 40 60 80 SingletonsDoubletonsTripletons & others% total species Net sampling Combination 0 20 40 60 80 100 SingletonsDoubletonsTripletons & othersMean no. individuals per sample(a) (b) Fig.3 Frequencyofspeciesoccurrenceanditsrelationshipwith animalabundance.(a)Meanproportion(SE)ofsingletons, doubletonsandotherspeciesatnet-samplingandcombination bores.(b)Meananimalabundance(SE)forsamplesingletons, doubletonsandotherspeciesatnet-samplingandcombination bores. 0 2 4 6 8 10 12 14 16Cumulative no. speciesRare Abundant All0 2 4 6 8 10 12 14 16 123456 Number of samplesCumulative no. species(a) (b) Fig.4 Speciesaccumulationcurvesforrarespecies,abundant speciesandallspecies,basedonaverageddatafromallnetsampledbores.(a)Rare averagespeciesabundanceacrossall boreswasinlowest50percentilesofabundance.(b)Rare speciesabundancewithinborebeingsampledwas £ 3animals persampleinwhichthespeciesoccurs.Samplingefciencyandgroundwaterfaunadistributions 893 2007TheAuthors,Journalcompilation 2007BlackwellPublishingLtd, FreshwaterBiology 54, 885–901

PAGE 10

the50speciesrepresentedbysitedoubletonsappeared tomeetHarvey’s(2002)criterionforSREsofrange <10000km2ifcircularrangeswereassumed.Eighteen ofthesitedoubletonsappearedtohaveranges <1000km2.Ofthe25sitetripletons,11andthreehad rangesof<10000km2and<1000km2respectively.RegionalspeciesrichnessUsingareducedmatrixof240boresandwellsthat yielded239species,theICEestimatorsuggestedthat about400speciesofgroundwaterfaunaoccurinthe Pilbara(Fig.6a).Thisestimateomittedseveralgroups ofanimalsthatwereexcludedfromanalysesbecause theyarepoorlyresolvedtaxonomicallyinthePilbara andpresentidenticationdifculties.Extrapolating ratioofcollectedanduncollectedtaxa(1.67)tothefull numberofspeciesknownfromthe240bores(320) providesamorerealisticestimatethatabout500–550 speciesoccurinthePilbara.Asimilarestimatewas obtainedusingthefullregionaldataset. Thecumulativenumbersofspeciesatterminal pointsofthecatchmentplotsusedtocalculateaT-S curve(Fig.6b)showedaverystronglinearrelationshipwiththesquarerootofsamplesize(ratherthan 0 10 20 30 40 50 60Pre-pump net haul Post-pump net haul PumpNet haulNet haulNet haulPre-pump net haul PumpPost-pump net haul Net haulNet haul 14/11/024/04/0311/05/0421/10/0411/11/0414/05/056/08/05 Cumulative no. species Other species Sample singletons Fig.5 IncreaseincumulativenumberofspeciesrecordedatborePSS016,anddecreaseinnumberofspeciesrecordedonlyonceatthe site(referredtoinlegendassamplesingletons),assamplingeffortincreased. Table2 Comparisonoftotalgroundwaterfaunaabundance acrosssamplingseason(asmonth)usingmixed-modelANOVA ANOVAwithsamplingmonthbeingaxedandsitearandomeffect. Abundancevariedsignicantlyacrosssitesbutnotmonths Sourced.f. TypeIII sumof squares Mean squares FP Site13115388.715.3<0.001 Month668.511.41.970.08 Error643715.8 Table3 NearestneighbourandaveragedistancetoanyotherborewithineachofthevecatchmentsofthePilbaracomparedwith distancesbetweenoccurrencesofsitedoubletonandtripletonspecies.Weightedaverageswereusedtosummarisewithin-catchment data Catchment Boresandwells Averageinter-boredistances (km) Doubletons Inter-occurrencedistances (km) Tripletons Inter-occurrencedistances (km) n NearestAverage n AverageMax. n AverageMax. 1Ashburton1327.5151114414616767 2DeGreyRiver1127.01232222214747 3FortescueRiver1096.5174714436312828 4PortHedlandCoast766.410846011824790 5OnslowCoast405.5602549422238 Withincatchments4696.8135297226374057 Acrosscatchments2422122750218259400894 S.M.Eberhard etal. 2007TheAuthors,Journalcompilation 2007BlackwellPublishingLtd, FreshwaterBiology 54, 885–901

PAGE 11

thelogarithmofsamplesizeasfoundbyUgland etal. 2003),yieldingtheequation S 17 : 1 p a 26 R2 1 : 00 3 where a representsareaasamultipleofthesampling unit.T-Scurve-derivedestimatesoftheregional numberofspecieswereverysensitivetoestimated sizeofthesamplingunit(i.e.howmuchofthe surroundingaquiferwascapturedwhensampling bores)andarenotpresented.However,assumingthe ICEestimatewasofthecorrectmagnitude,T-Scurve calculationshadtheinterestingimplicationthatthe faunawithinaradiusofabout5kmoftheborewas sampled. Therewaslittledifferencebetweenmajorcatchmentsinpatternsofoccurrenceofthegroundwater fauna.Theproportionofboresandwellsthatyielded nospeciesfromtwosamplingeventsvariedbetween 20 % inPortHedlandCoastaland28 % inthe Ashburton(seeFig.1forlocations)andtherateof accumulationofspeciesshowedalmostnovariation amongcatchments(Fig.6b).DiscussionResultsoftheintensivesamplinginthePilbaraare similartothoseofothersubterraneanfaunastudiesin thatadditionalspeciescontinuedtobecollectedas samplingeffortincreased(e.g.Culver etal. ,2004; Hancock&Boulton,2009).Speciesaccumulation curvesshowedthat,inthePilbara,onenet-haul samplingeventcollectedonly33 % ofthespecies presentataboreandsixsamplingeventscollected only82 % ofspecies.IneasternAustralia,10nethauls atabore(i.e.1.6Pilbarasamplingevents)yielded 31 % ofthespeciescollectedbyfourcombinationsof nethaulingandpumping(Hancock&Boulton,2009, recalculatedfromTable3).Thesendingshaveprofoundimplicationsforthedesignofstudiesthatare intendedtoprovideacompletelistofthespeciesinan area,asisthecaseforenvironmentalassessment (EPA,2003).Currently,itisunusualforbores inAustraliatobesampledmorethantwicein biodiversitysurveysorenvironmentalassessment programmesbecauseoftimeconstraintsandthecost ofeldwork,particularlyinremoteregionssuchas thePilbara.SpeciesabundanceanddetectabilityTherelationshipbetweenspeciesabundanceand detectabilityformsthebasisforseveralfrequently usedestimatorsofspeciesrichness,suchasChao1 (Foggo etal. ,2003).Mostanimalcommunitiescontain afewabundantspeciesandmanyspeciesthatoccur inlownumbers(e.g.Fig.2)andtheabundantspecies aremorelikelytobecollectedinarandomsampleofa fewanimalsthantherarespecies.Assamplingeffort increases,thenumberofspeciescollectedincreases accordingtothegeneralformulationthataspecies willbecollectedifitsproportioninthecommunity multipliedbythetotalnumberofanimalsinthe sampleis>1(Courtemanch,1996). (b) 0 50 100 150 200 250 300 Number of samplesCumulative no. species DeG Ash For PHC Robe (a) 0 50 100 150 200 250 300 350 400 450 050100150200 050100150200Cumulative no. species Sobs ICE Fig.6 SpeciesaccumulationpatternsinthePilbararegional survey.(a)EstimatedtotalnumberofspeciesinthePilbara accordingtonumberofsamplescollected,usingtheICE estimator,andcumulativenumberofspeciesobserved( Sobs). (b)Speciesaccumulationcurvesbasedondatafromone,and thentwocatchmentsetc(seeUgland etal. ,2003).Legend indicatestheadditionalcatchmentineachcurve.Notethat thattheDeGreycurve,incorporatingallcatchments,isthe Sobscurvefrom(a).Samplingefciencyandgroundwaterfaunadistributions 895 2007TheAuthors,Journalcompilation 2007BlackwellPublishingLtd, FreshwaterBiology 54, 885–901

PAGE 12

Weareunawareofpreviousstudiesoftherelationshipbetweenabundanceanddetectabilityforsubterraneananimals,despitethesamplingdecienciesand thesmallnumbersofanimalscollectedinstudiesof subterraneanenvironments(e.g.Eberhard etal. 2005b;Hahn&Matzke,2005;Schneider etal. ,2005), makingthemmoresusceptibletoeffectsondetectabilitythanstudiesofsurface-waterfauna.Usingonly twoabundancecategories,thisstudyshowedthat speciesrepresentedbyhighnumbersofanimalswere twoorthreetimesmorelikelytobecollectedinone samplingeventthanspeciesoccurringatlowabundance(Fig.4).Inreality,thedifferenceincollection probabilitiesofthemost,andleast,abundantspecieswillbemuchgreaterandmuchofwhathas previouslybeentreatedasstochasticvariationin speciesrecoverycanprobablybeexplainedinterms ofspeciesabundances.Recognisingthiswillimprove ourunderstandingofthestructureofgroundwater faunacommunitiesandimproveinterpretationof samplingresults.FNratesESTIMATES ESTIMATEShasoftenbeenusedtoprovideestimates ofthetruenumberofspeciesatasite,orinaregion, fromwhichsamplingefciencycanthenbeinferred. Suchestimatesareunreliableatlowsamplingefforts (Foggo etal. ,2003)and,therefore,weexploreduseof FNratesasanalternativemethodofexamining samplingefciency.TheFNratecalculatedwasbased ontheassumptionthatallspeciespresentatabore weredetectedinatleastoneofthesixsampling events.However,thisassumptionisunlikelytobe correctforsubterraneanfauna,andotherorganisms occurringinlowabundance,asshownbythisstudy andHancock&Boulton(2009). AmoreaccurateFNforallspeciescanbecalculated byinsertingtheobservedFNrate(0.392),derived fromeqn1,intoaformulaprovidedbyTyre etal. (2003)togeneratecollectionprobabilitiesthatinclude failuretocollectaspecies L y j ^ p ; ^ q p m y ^ qy 1 ^ q m yy > 0 4 where1 ) q istheFNrate; p istheprobabilitythat thespeciesutilisedthesitethroughoutsampling (assumed 1); m isthenumberofsamplingevents (6)and y isthenumberoftimesthespecieswas observed.Fortheall-speciesdataset,thelikelihoodof aspeciesnotbeingrecordedinsixsampleswas0.05, therevisedFNratewasapproximately0.65,and revisedprobabilityofaspeciesbeingcollectedina singlesamplewas0.35.Thisissimilartotheestimate basedonspeciesaccumulationcurves(0.33).However,adiscrepancyremainedbetweenestimatesofthe detectionprobabilityofrarespeciesbasedonthe revisedFNrateandChao2(0.30versus0.23).RegionalheterogeneitySpeciesrichnessoftenexhibitsheterogenouspatterns acrossregions(Ugland etal. ,2003;Culver etal. ,2004) andtakingheterogeneityintoaccountwhenestimatingregionalspeciesrichnessislikelytoimprove accuracy.Unfortunately,theT-ScurvemethodproposedbyUgland etal. (2003)provedtobevery sensitivetoassumptionsabouttheareaofaquiferthat wascapturedbysamplingabore.Thereisno experimentalinformationaboutthedistanceover whichgroundwaterfaunawillmoveintoboresbutTScurvecalculationssuggestedingressoccursfromthe surrounding50–80km2ofaquifer.Thisisamuchlarger areathaninferredinotherstudies(e.g.Malard etal. 1997;Hahn&Matzke,2005).Whileindependent conrmationisneededthatthemobilitysuggestedby T-Scurvesisnotamathematicalartefact,suchmobility hasconsiderableramicationsformanagementof groundwaterfauna.Inparticular,re-colonisationof de-wateredsitesfromsurroundingareasmayoccur quicklyifanimalscanmoveseveralkilometres.ColonisationofboresTheextenttowhichthecompositionofgroundwater faunainboresaccuratelyreectsfaunalcomposition insurroundinggroundwaterisanimportant,unresolvedissue.Resultstodateareinconsistent,with somestudiessuggestingboresprovidebiasedsamplesandothersimplyingarepresentativeselectionof aquiferspeciesisobtained(seeHahn&Matzke,2005). Biasmaybecausedbythedifferentialattractionof variousspeciestoboresbecauseoftheirfeedingand habitatpreferences,orbythephysicalexclusionof largerspeciesfrombores. Construction,slottingandscreeningmethodsfor boresvaryconsiderably,andareunrecordedformost896 S.M.Eberhard etal. 2007TheAuthors,Journalcompilation 2007BlackwellPublishingLtd, FreshwaterBiology 54, 885–901

PAGE 13

boresinthePilbara,butboreattributesappear unlikelytohavehadmucheffectonthenumberof speciescollectedfromboreswherenetsampling occurred.Down-holevideorecordingsshowedisopodsofthegenus Pygolabis ,thelargestgroundwater invertebratesinthePilbaraandmeasuringover1cm (Keable&Wilson,2006),movingfreelyinandoutof boresthroughslotting(Fig.7c).Amphipodswerealso observedswimmingthroughslotting.Amphipods and Pygolabis specieswerebothobservedentering boresthroughsmallssuresatthebase(Fig.7a,b)and belowthecasing(Fig.7d)ofbores.Furthermore, Pygolabis oramphipodsofthegenera Nedsia Pilbarus or Chydaekata wererecordedfromahighproportionof bores.Theaboveevidencesuggeststherewerefew impedimentstocolonisationbylargerspeciesand thatanybiasinspeciescompositionwithinbores, relativetothesurroundingaquifer,wasmorelikely becauseoffactorsassociatedwiththeenrichmentof bores(Hahn&Matzke,2005).SingletonsandSREsOneofthedeningcharacteristicsofsubterranean faunagloballyisthehighlevelofregionalandshortrangeendemism.InEurope,subterraneanfauna usuallycomprises>50 % SREs,withsomeregions having>90 % (Gibert&Deharveng,2002),andhigh proportionsofSREshavebeenreportedfromgroundwateroftheAustralianaridzone(Cooper etal. ,2002; Harvey,2002).SurveyeffortinthePilbaraisinsufcienttomakedenitivestatementsabouttheproportionofgroundwaterfaunathatareSREsbut, nonetheless,itseemslikelythatveryfewspecies collectedatmorethanthreesitesintheregional surveyareSREs(range<10000km2).Fewerthan30 % ofspeciesrepresentedbysitetripletonsand<50 % of sitedoubletonsarelikelytobeSREs. Assessingtheproportionofspeciesrecordedassite singletonsthatarelikelytobeSREsisdifcult.About 85 % ofthemwerealsosamplesingletons,whichwas similartothe88 % expectediftheyhadanaverage collectionprobabilityof0.23(usedincalculations below).Averagedensityofboresandwellsinthe Pilbarasurveywasabout23/10000km2andspecies withacollectionprobabilityof0.23shouldhavebeen collectedfrommorethanoneboreiftheirrangeswere ‚ 10000km2(probabilityofbeingcollectedtwoorthree timeswas>0.99and0.94,respectively).Thus,even allowingforboredistributionbeingsomewhatirregularwithinacatchment,mostofthespeciesrecordedas sitesingletonsmusthaveranges<10000km2and qualifyasSREsaccordingtoHarvey’s(2002)criterion, unlesstheyarerestrictedtogroundwaterhabitatsthat (a) (b) (c)(d) Fig.7 Groundwaterspeciesentering boresinthePilbarafromdown-holevideo footage.(a&b)Amphipod(arrowedina) enteringthroughsmallvoidinbaseof borePSS190at35mdepthintheFortescueBasin.(c) Pygolabis isopodentering throughslotofborePSS172at7mdepth intheAshburtonBasin.(d) Pygolabis isopodsenteringundercasingofborePSS179 at41mdepthintheAshburtonBasin. Scalebar 1cm.Samplingefciencyandgroundwaterfaunadistributions 897 2007TheAuthors,Journalcompilation 2007BlackwellPublishingLtd, FreshwaterBiology 54, 885–901

PAGE 14

werarelysampled.Ifourassumptionsaboutcollection probabilityanddistributionofsingletonsinrelation withborespacingsarecorrect,thenabout55 % of knownPilbaragroundwaterspeciesareSREs.Atrue pictureofendemisminthePilbara,however,mustalso takeintoaccountthespecieswefailedtocollect. Despitesomesmallgapsinoursamplingcoverage, mostofthesespecieswouldhavebeenmissedbecause theirrangeswere £ 10000km2,whichimpliesthat about70 % ofallPilbaragroundwaterspeciesareSREs. TheproportionofSREsis,however,dependenton thethresholdcriterionused.Harvey’s(2002)decision tousearangeof10000km2todeneSREswas arbitraryandwesuggestitisratherlarge.Itfailsto distinguishgroundwaterspecieswithsub-regional distributionsthataresecurefromrange-related naturalandanthropogenicthreatsfromthosespecies withsufcientlylocaliseddistributionsthatmostof theirpopulationmaybeatriskfromanactivitysuch asde-wateringorfromapollutionevent.Noproject involvingbelowwatertableminingandgroundwater abstractioninthePilbarahascausedsignicant groundwaterdrawdownbeyonda10km2radiusof pumping(Johnson&Wright,2001),whichequatesto animpactareaofabout350km2,andinmostcases theimpactareahasbeen<100km2.Therefore,we suggestthatamoreappropriatecriterionforPilbara groundwaterSREsisarangeof<1000km2.Mine de-wateringandgroundwaterextractionareunlikely tothreatenspecieswithdistributionsattheupperend ofthisrangebutathresholdof1000km2representsa precautionaryapproachandisofascalethatmatches naturalbarriers.ManyPilbaragroundwaterspecies appeartoberestrictedtosectionsofhydrographic basins,ortributarieswithinthem,andtohaveranges oftheorderof1000km2(seeFinston etal. ,2006; Reeves,DeDeckker&Halse,2007). Inconclusion,highregionalrichnessofgroundwaterfaunaisusuallyattributedtoageofthe landscape,habitatfragmentation,SREandexistence ofsuitablehabitat(e.g.Christman&Culver,2001; Humphreys,2001;Gibert&Deharveng,2002).The Pilbaraisanoldlandscapewithgeologicalheterogeneity(McPhail&Stone,2004)and>40 % ofknown Australiangroundwaterspecieshavebeenrecorded fromthisregion(Humphreys etal. ,2005).Although thereismuchstilltobelearnedaboutthegroundwaterfaunaofthePilbara,therecentlycompleted regionalsurveyislikelytohaveidentieditsmajor characteristicsandsub-regionalhotspots.Results fromthenet-sampledboressuggestthat,aswith Sloveniancaves(Culver etal. ,2004),tworoundsof samplingisusuallysufcienttoidentifyPilbarasites thatarerichingroundwaterspecies.Continued samplingis,however,requiredtodocumentthefull richnessofthesites.Thiswasdemonstratedatbore PSS016where11samplingeventsyieldedthreetimes morespeciesthancollectedintwoevents(Fig.5).AcknowledgmentsJessicaReevesandIvanaKaranovicidentiedall ostracodsinthisstudy;TomKaranovicidentied manyofthecopepods.Somespeciesidentications wereconrmedbyAdrianPinder(oligochaetes)and BuzWilson(isopods).Adviceonothergroupswas providedbyJohnBradburyandTerrieFinston (amphipods),MarkHarvey(mites)andPeter Serov(bathynellids).WethankDaveRobertsonfor calculatingdistancesbetweenbores,LesleyGibson forhelpfuladvice,andJanineGibertandtwo anonymousrefereesforveryconstructivecriticism ofthemanuscript.ReferencesBoultonA.J.,HumphreysW.F.&EberhardS.M.(2003) ImperilledsubsurfacewatersinAustralia:biodiversity,threateningprocessesandconservation. Aquatic EcosystemHealthandManagement 6 ,41–54. BradburyJ.H.(2000)WesternAustralianstygobiont amphipods(Crustacea:Paramelitidae)fromtheMt NewmanandMillstreamregions. RecordsoftheWestern AustralianMuseumSupplement 60 ,1–102. CastellariniF.,Dole-OliverM.-J.,MalardF.&GibertJ. (2007a)Modellingthedistributionsofstygobiontsin theJuraMountains(easternFrance).Implicationsfor theprotectionofgroundwaters. DiversityandDistributions 13 ,213–224. CastellariniF.,Dole-OliverM.-J.,MalardF.&GibertJ. (2007b)Usingenvironmentalheterogeneitytoassess stygobioticspeciesrichnessintheFrenchJuraregion withaconservationperspective. Fundamentaland AppliedLimnology 169 ,69–78. CaughleyG.&GunnA.(1996) ConservationBiologyin TheoryandPractice .BlackwellScience,Oxford,pp.459. ChristmanM.C.&CulverD.C.(2001)Therelationship betweencavebiodiversityandavailablehabitat. JournalofBiogeography 28 ,367–380.898 S.M.Eberhard etal. 2007TheAuthors,Journalcompilation 2007BlackwellPublishingLtd, FreshwaterBiology 54, 885–901

PAGE 15

CoatesD.J.&AtkinsK.A.(2001)Prioritysettingandthe conservationofWesternAustralia’sdiverseandhighly endemicora. BiologicalConservation 97 ,251–263. ColwellR.K.(2005)EstimateS:statisticalestimationof speciesrichnessandsharedspeciesfromsamples, version7.User’sguideandapplication.(http://viceroy. eeb.uconn.edu/estimates). ColwellR.K.&CoddingtonJ.A.(1994)Estimating terrestrialbiodiversitythroughextrapolation. PhilosophicalTransactionsoftheRoyalSocietyofLondon,Series B 345 ,101–118. ColwellR.K.,MaoC.X.&ChangJ.(2004)Interpolating, extrapolating,andcomparingincidence-basedspecies accumulationcurves. Ecology 85 ,2717–2727. CooperS.,HinzeS.,LeysR.,WattsC.&HumphreysW. (2002)Islandsunderthedesert:molecularsystematics andevolutionaryoriginsofstygobiticwaterbeetles (Coleoptera:Dytiscidae)fromcentralWesternAustralia. InvertebrateSystematics 16 ,589–598. CourtemanchD.L.(1996)Commentaryonthesubsamplingproceduresusedforrapidbioassessments. JournaloftheNorthAmericanBenthologicalSociety 15 381–385. CulverD.C.&SketB.(2000)Hotspotsofsubterranean biodiversityincavesandwells. JournalofCaveand KarstStudies 6 ,11–17. CulverD.C.,ChristmanM.C.,SketB.&TronteljP.(2004) Samplingadequacyinanextremeenvironment:speciesrichnesspatternsinSloveniancaves. Biodiversity andConservation 13 ,1209–1229. DanielopolD.L.,GrieblerC.,GunatilakaA.&NotenboomJ.(2003)Presentstateandfutureprospectsfor groundwaterecosystems. EnvironmentalConservation 30 ,104–130. Dole-OlivierM.J.,CastellariniF.,CoineauN.,Galassi D.M.P.,MoriN.,ValdecasasA.&GibertJ.(2009) Towardsanoptimalsamplingstrategyto assessgroundwaterbiodiversity:comparison acrosssixEuropeanregions. FreshwaterBiology 54 777–796. DudgeonD.,ArthingtonA.H.,GessnerM.O. etal. (2006) Freshwaterbiodiversity:importance,threats,status andconservationchallenges. BiologicalReviews 81 163–182. EberhardS.M.,HalseS.A.&HumphreysW.F.(2005a) StygofaunainthePilbararegion,north-westWestern Australia:areview. JournaloftheRoyalSocietyof WesternAustralia 88 ,167–176. EberhardS.M.,HalseS.A.,ScanlonM.D.,CockingJ.S.& BarronH.J.(2005b)Assessmentandconservationof aquaticlifeinthesubsurfaceofthePilbararegion, WesternAustralia.In:ProceedingsofanInternational SymposiumonWorldSubterraneanBiodiversity(Ed. J.Gibert),pp.61–68.Villeurbanne,France,8–10 December2004.UniversityofLyon,France. EPA(2003) GuidancefortheAssessmentofEnvironmental Factors:ConsiderationofSubterraneanFaunainGroundwaterandCavesDuringEnvironmentalImpactAssessment inWesternAustralia .GuidanceStatement54.Western AustralianEnvironmentalProtectionAuthority,Perth, Australia. FinstonT.L.,JohnsonM.S.,HumphreysW.F.,Eberhard S.M.&HalseS.A.(2006)Crypticspeciationintwo widespreadsubterraneanamphipodgenerareects historicaldrainagepatternsinanancientlandscape. MolecularEcology 16 ,355–365. FoggoA.,AttrillM.J.,FrostM.T.&RowdenA.A.(2003) Estimatingmarinespeciesrichness:anevaluationof sixextrapolativetechniques. MarineEcologyProgressSeries 248 ,15–26. GibertJ.&DeharvengL.(2002)Subterraneanecosystems:atruncatedfunctionalbiodiversity. BioScience 52 ,473–481. HahnH.J.&MatzkeD.(2005)Acomparisonofstygofaunacommunitiesinsideandoutsidegroundwater bores. Limnologica 35 ,31–44. HalseS.A.,ScanlonM.D.&CockingJ.S.(2002)Do springsprovideawindowtothegroundwaterfaunaof theAustralianaridzone?In: BalancingtheGroundwater Budget:ProceedingsofanInternationalGroundwater Conference,Darwin2002 (Ed.D.Yinfoo),pp.1–12. InternationalAssociationofHydrogeologists,Darwin, Australia. HalseS.A.,ShielR.J.,StoreyA.W.,EdwardD.H.D., LansburyI.,CaleD.J.&HarveyM.S.(2000)Aquatic invertebratesandwaterbirdsofwetlandsandriversof thesouthernCarnarvonBasin,WesternAustralia. RecordsoftheWesternAustralianMuseumSupplement 61 ,217–265. HancockP.J.&BoultonA.J.(2009)Samplinggroundwaterfauna:efciencyofrapidassessmentmethods testedinboresineasternAustralia. FreshwaterBiology 54 ,902–917. HarveyM.(2002)Short-rangeendemismamong theAustralianfauna:someexamplesfromnonmarineenvironments. InvertebrateSystematics 16 555–570. HumphreysW.F.(2001)Groundwatercalcreteaquifers intheAustralianaridzone:thecontexttoanunfolding plethoraofstygalbiodiversity. RecordsoftheWestern AustralianMuseumSupplement 64 ,63–83. HumphreysW.F.,WattsC.H.S.&BradburyJ.H.(2005) Emergingknowledgeofdiversity,distributionand originsofsomeAustralianstygofauna.In:Proceedings ofanInternationalSymposiumonWorldSubterranean Biodiversity(Ed.J.Gibert),pp.57–60.Villeurbanne,Samplingefciencyandgroundwaterfaunadistributions 899 2007TheAuthors,Journalcompilation 2007BlackwellPublishingLtd, FreshwaterBiology 54, 885–901

PAGE 16

France,8–10December2004.UniversityofLyon, France. JohnsonS.L.&WrightA.H.(2001) CentralPilbara GroundwaterStudy.HydrogeologicalRecordSeriesHG8 DepartmentofWater,Perth,Australia. KeableS.J.&WilsonG.D.F.(2006)Newspeciesof Pygolabis Wilson2003(Isopoda,Tainisopidae,Crustacea)fromWesternAustralia. Zootaxa 1116 ,1–27. KrishnamurthyK.V.(2003) TextbookofBiodiversity .SciencePublishers,Eneld.pp.260. MaceG.M.(1995)Classicationofthreatenedspeciesand itsroleinconservationplanning.In: ExtinctionRates (EdsJ.H.Lawton&R.M.May),pp.197–213.Oxford UniversityPress,Oxford. MaceG.,MasundireH.,BaillieJ. etal. (2005)Biodiversity.In: EcosystemsandHumanWell-being:CurrentState andTrends,Vol.1.FindingsoftheConditionandTrends WorkingGroupoftheMilleniumEcosystemAssessment (EdsR.Hassan,R.Scholes&N.Ash),pp.77–122. IslandPress,Washington,D.C. MacKenzieD.I.,NicholsJ.D.,LachmanG.B.,DroegeS., RoyleJ.A.&LangtimmC.A.(2002)Estimatingsite occupancyrateswhendetectionprobabilitiesareless thanone. Ecology 83 ,2248–2255. MalardF.,ReygrobelletJ.-L.,LaurentR.&MathieuJ. (1997)Developmentsinsamplingthefaunaofdeep water-tableaquifers. Archivfu rHydrobiologie 138 ,401– 432. MauclaireL.,MarmonierP.&GibertJ.(1998)Sampling waterandsedimentininterstitialhabitats:acomparisonofcoringandpumpingtechniques. Archivfu r Hydrobiologie 142 ,111–123. McPhailM.K.&StoneM.S.(2004)AgeandpalaeoenvironmentalconstraintsonthegenesisoftheYandi channelirondeposits,MarillanaFormation,Pilbara, northwesternAustralia. AustralianJournalofEarth Sciences 51 ,497–520. ParanF.,MalardF.,MathieuJ.,LafontM.,GalassiD.M.P. &MarmonierP.(2005)Distributionofgroundwater invertebratesalonganenvironmentalgradientina shallowwater-tableaquifer.In:Proceedingsofan InternationalSymposiumonWorldSubterraneanBiodiversity(Ed.J.Gibert),pp.99–106.Villeurbanne, France,8–10December2004.UniversityofLyon, France. PinderA.M.,HalseS.A.,McRaeJ.M.&ShielR.J.(2004) Aquaticinvertebrateassemblagesofwetlandsand riversintheWheatbeltregionofWesternAustralia. RecordsoftheWesternAustralianMuseumSupplement 67 ,7–37. PipanT.&CulverD.C.(2005)Estimatingbiodiversityin theepikarsticzoneofaWestVirginiacave. Journalof CaveandKarstStudies 67 ,103–109. PonderW.&ColganD.J.(2002)Whatmakesanarrowrangetaxon?InsightsfromAustralianfreshwater snails. InvertebrateSystematics 16 ,571–582. ReevesJ.,DeDeckkerP.&HalseS.A.(2007)GroundwaterostracodsfromthearidPilbararegionof northwesternAustralia:distributionandwaterchemistry. Hydrobiologia 585 ,99–118. RouchR.&DanielopolD.L.(1997)Speciesrichnessof microcrustaceainsubterraneanfreshwaterhabitats. Comparativeanalysisandapproximateevaluation. InternationaleRevuedergesamtenHydrobiologie 82 121–145.SchneiderK.,CulverD.C.,HobbsH.H.&FongD.(2005) Theblindmisleadingtheblind:falsenegativesand estimationsofsubterraneanbiodiversity.In:ProceedingsofanInternationalSymposiumonWorldSubterraneanBiodiversity(Ed.J.Gibert),p.138.Villeurbanne, France,8–10December2004.UniversityofLyon, France. TyreA.J.,TehumbergB.,FieldS.A.,NiejalkeD.,ParrisK. &PossinghamH.P.(2003)Improvingprecisionand reducingbiasinbiologicalsurveysbyestimatingfalse negativeerrorratesinpresence–absencedata. EcologicalApplications 13 ,1790–1801. UglandK.I.&GrayJ.S.(2004)Estimationofspecies richness:analysisofthemethodsdevelopedbyChao andKarakassis. MarineEcologyProgressSeries 284 ,1–8. UglandK.I.,GrayJ.S.&EllingsenK.E.(2003)The species-accumulationcurveandestimationofspecies richness. JournalofAnimalEcology 72 ,888–897. (Manuscriptaccepted28July2007)Appendix1Taxonomiclistofthe93taxacollected duringrepeatsamplingof14boresinthePilbara(see Fig.2).Manyoftheidenticationsareonlyto morphospecieslevel.Speciesnumbersrelatetothose inFig.2Species numbers Turbellaria Turbellariasp.70 Nematoda Nematodasp.324 Nematodasp.430 Nematodasp.1123 Rotifera Dissotrocha sp.72 Gastropoda Planorbidaesp.42 Hydrobiidaesp.22900 S.M.Eberhard etal. 2007TheAuthors,Journalcompilation 2007BlackwellPublishingLtd, FreshwaterBiology 54, 885–901

PAGE 17

Appendix1 ( Continued )Species numbers Oligochaeta Hirudineasp.68 Aeolosoma sp.359 Phreodrilussp.WA3226 Phreodrilidaesp.DVC45 Phreodrilidaesp.SVC64 Ainudrilus sp.WA2754 Tubicidaesp.183 Tubicidaesp.229 Tubicidaesp.WA2827 Deronivea Aiyer19 Pristina sp.WA369 Enchytraeidaesp.225 Enchytraeidaesp.186 Polychaeta Nereidaesp.31 Arachnida Guineaxonopsis sp.S112 Arrenurus n.sp.252 Peza sp.41 Oribatidasp.153 Ostracoda Gomphodellahirsuta Karanovic82 Limnocytherestationis Vavra16 Limnocythere sp.115 Candonopsispilbarae Karanovic88 Deminutiocandonaaporia Karanovic48 Deminutiocandona cf. atope Karanovic66 Deminutiocandonaaenigma Karanovic44 Deminutiocandonastomachosa Karanovic62 Humphreyscandonawoutersi Karanovic& Marmonier 80 Humphreyscandona sp.275 Humphreyscandonaventosa Karanovic67 Notacandonaboultoni Karanovic&Marmonier17 Origocandonaposteriorecta Karanovic18 Pilbaracandonacolonia Karanovic&Marmonier32 Pilbaracandonaeberhardi Karanovic&Marmonier33 Pilbaracandonakosmos Karanovic85 Pilbaracandonatemporaria Karanovic39 Pilbaracandonarosa Karanovic43 Areacandonamulgae Karanovic(11) Areacandonalepte Karanovic61 Areacandonacylindrata Karanovic58 Areacandonatriangulum Karanovic55 Areacandonaiuno Karanovic91 Areacandona cf.sp.137 Areacandona sp.738 Areacandonaastrepte Karanovic50 Areacandonaatomus Karanovic10 Leicacandonacarinata Karanovic51 Leicacandonajimi Karanovic14 Kencandonaverrucosa Karanovic13 Candonidaen.gen.77 Appendix1 ( Continued )Species numbers Syncarida Bathynella sp.21 Notobathynella sp.47 Chilibathynella sp.40 Atopobathynella sp.A20 Thermosbaenacidae Halosbaenatulki Poore&Humphreys79 Copepoda Mesocyclopsbrooksi DeLaurentiis etal .5 Inermipes sp.24 Diacyclopseinslei Karanovic3 Diacyclopshumphreysi s.str. X unispinosus Karanovic 78 Diacyclopshumphreysihumphreysi Karanovic81 Diacyclopscocking Karanovic76 Diacyclopsscanloni Karanovic90 Diacyclopssobeprolatus Karanovic60 Halicyclops(Rochacyclops)rochi Karanovic87 Orbuscyclopswestaustraliensis Karanovic6 Elaphoidellahumphreysi Karanovic84 Schizoperaroberiveri Karanovic46 Abnitocrella sp.356 Archinitocrellanewmanensis Karanovic92 Parapseudoleptomesochratureei Karanovic65 Stygonitocrellabispinosa Karanovic36 Stygonitocrellatrispinosa Karanovic63 Stygonitocrellaunispinosa Karanovic74 Parastenocarisjane Karanovic35 Pseudectinosomagalassiae Karanovic7 Amphipoda Chydaekata sp.73 Paramelitidaen.gen.2 Paramelitidaesp.928 Paramelitidaesp.289 Nedsia sp.93 Melitidaesp.149 Bogidiellidaesp.1 Isopoda Speocirolana n.sp.157 Pygolabiseberhardi Keable&Wilson71 Pygolabishumphreysi Wilson34 Pygolabisparaburdoo Keable&Wilson9 Microcerberidaesp.8Samplingefciencyandgroundwaterfaunadistributions 901 2007TheAuthors,Journalcompilation 2007BlackwellPublishingLtd, FreshwaterBiology 54, 885–901


printinsert_linkshareget_appmore_horiz

Download Options

close
Choose Size
Choose file type
Cite this item close

APA

Cras ut cursus ante, a fringilla nunc. Mauris lorem nunc, cursus sit amet enim ac, vehicula vestibulum mi. Mauris viverra nisl vel enim faucibus porta. Praesent sit amet ornare diam, non finibus nulla.

MLA

Cras efficitur magna et sapien varius, luctus ullamcorper dolor convallis. Orci varius natoque penatibus et magnis dis parturient montes, nascetur ridiculus mus. Fusce sit amet justo ut erat laoreet congue sed a ante.

CHICAGO

Phasellus ornare in augue eu imperdiet. Donec malesuada sapien ante, at vehicula orci tempor molestie. Proin vitae urna elit. Pellentesque vitae nisi et diam euismod malesuada aliquet non erat.

WIKIPEDIA

Nunc fringilla dolor ut dictum placerat. Proin ac neque rutrum, consectetur ligula id, laoreet ligula. Nulla lorem massa, consectetur vitae consequat in, lobortis at dolor. Nunc sed leo odio.