Citation
Journal of cave and karst studies

Material Information

Title:
Journal of cave and karst studies
Series Title:
Journal of Cave & Karst Studies
Alternate Title:
Continues NSS bulletin (OCLC: 2087737)
Creator:
National Speleological Society
Publisher:
National Speleological Society
Publication Date:
Language:
English

Subjects

Subjects / Keywords:
Geology ( local )
Genre:
serial ( sobekcm )
Location:
United States

Notes

Restriction:
Open Access - Permission by Publisher
Original Version:
Vol. 70, no. 1 (2008)

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:
K26-02047 ( USFLDC DOI )
k26.2047 ( USFLDC Handle )
6753 ( karstportal - original NodeID )
0146-9517 ( ISSN )

USFLDC Membership

Aggregations:
Karst Information Portal

Postcard Information

Format:
Serial

Downloads

This item is only available as the following downloads:


Full Text

PAGE 1

NEWFINDINGSATANDRAHOMANACAVE, SOUTHEASTERNMADAGASCAR D.A.B URNEY 1,* ,N.V ASEY 2 ,L.R.G ODFREY 3 ,R AMILISONINA 4 ,W.L.J UNGERS 5 ,M.R AMAROLAHY 6 AND L.R AHARIVONY 6 Abstract: Aremoteeolianitecaveandsinkholecomplexonthesoutheastcoastof MadagascarhasplayedamajorroleinthehistoryofpaleontologyinMadaga scar. AndrahomanaCavehasyieldedarichfossilrecordoftheextinctmegafauna Expeditionsin2000and2003producedawealthofnewmaterialandprovided the firstsystematicinformationconcerningthegenesis,stratigraphy,and taphonomyofthe site.Recoveredbonesofoneofthemostpoorlyunderstoodextinctlargele murs, Hadropithecusstenognathus ,includemanyskeletalelementspreviouslyunknown. Radiocarbondatesshowthatthesitehassampledthisdisappearedfaunain themidto-lateHolocene,butthatbone-bearinglayersarestratigraphicallymi xed,probably owingtotheeffectsofreworkingofthesedimentsbyextrememarineevents .Thediverse biotarecoveredcontainselementsofbotheasternrainforestandsouthwe sternarid bushland,reflectingthecave’spositioninthezoneoftransitionbetwee nwetanddry biomes.Bonesoftwounusualsmallmammalsaddtothepreviouslylongfauna llistfor thesite:1)thefirstfossilevidencefor Macrotarsomyspetteri ,alarge-bodiedendemic nesomyidrodentpreviouslyknownonlyfromasinglemodernspecimen;and2 )thetype specimenandadditionalmaterialofanewlydescribedextinctshrew-tenr ec( Microgale macpheei ).Evidenceforprehistoricandcolonial-erahumansincludesartifacts, hearth deposits,andremainsofhumandomesticatesandotherintroducedspecies .Although previouslyprotectedbyitsextremeisolation,theuniquesiteisvulnera bletoexploitation. Anincipienttouristindustryislikelytobringmorepeopletothecave,an dthereis currentlynoformofprotectionaffordedtothesite. I NTRODUCTION Madagascarisnotedforhavingmanylargeand spectacularcaves,andseveralofthesehaveyieldedtroves offossilsoftheextinctgiantlemurs,hugeelephantbirds, pygmyhippos,andotherelementsoftheisland’slost Holocenesubfossilbiota(e.g.,Simonsetal.,1995;Burney etal.,1997,2004).Oneoftherichestofthese,AndrahomanaCaveonMadagascar’ssoutheastcoast,aremote collapsed-cavefeaturethathasbeenvisitedbyonlya handfulofoutsiders,isbothvisuallyspectacularand scientificallyimportant(seeAppendix1forachronologicalaccountofexploration). Thesiteisnormallyreachedbyclimbingdowngneissic cliffstotheeast,thenwadingatlowtideoverexposedcoral reefs.Duringthe2003expedition,analternativeroutewas developedthatwasmuchfasterandlessdependentonthe tides,consistingofdirectapproachtothenorthern,inland rimofthesinkhole,followedbyamoderaterappelor cable-ladderdescentdowntheverticalface.Anapproach alongtherockyshorelinefromthewestwasalsousedafew times,butprovedtobetoodangerousbecausethedeep rillsinthegneissicbedrockformdeadlyborewaveswhen theseaswellisverylarge. Thestonymatrixfromwhichthecaveissculptedisa roughly100mthickpileofeolianite,withatleastthree separateunitsofcalcareoussandstoneslaiddownbyduneformingwinds.Currenttheoryholdsthatthesedepositsare normallyformedattheendofPleistoceneglacialcyclesas sealevelrisesrapidlyfollowingdeglaciation,grindingup coralreefsandothermarinedeposits(seeBurneyetal., 2001).Theseeolianiteunitslieexposedinasheersouthfacingclifffacecutbytheroughseasthatbringhugeswells ashorefromAntarctica.Agapingentrancenearthebaseof thiscliffleadsintoadoublesinkhole,whosefigure-eightshapedfootprintispartiallyroofedbyasweepingmassive triple-archshapedbymultiplecollapses.Beyondthis complexnaturalbridge,withitslargecentralpillar,isa smallsunkenrainforestthrivinginthemicrohabitat providedinsidethecavesystem.Thismoistrefugeis surroundedabovebyamuchdrierlandscapeofspiny semiaridbushland. *Correspondingauthor 1 NationalTropicalBotanicalGarden,3530PapalinaRoad,Kalaheo,HI9674 1, USA,dburney@ntbg.org 2 DepartmentofAnthropology,PortlandStateUniversity,Portland,OR972 07, USA 3 DepartmentofAnthropology,UniversityofMassachusetts-Amherst,Amhe rst, MA01003,USA 4 Muse ed’Artetd’Arche ologie,17RueduDocteurVillette,Antananarivo101, Madagascar 5 DepartmentofAnatomicalSciences,StonyBrookUniversity,StonyBrook, NY 11794,USA 6 De partmentdePale ontologieetd’AnthropologieBiologique,Faculte desSciences, B.P.906,Universite d’Antananarivo,Antananarivo101,Madagascar D.A.Burney,N.Vasey,L.R.Godfrey,Ramilisonina,W.L.Jungers,M.Ramar olahy,andL.Raharivony–NewFindingsat AndrahomanaCave,SoutheasternMadagascar. JournalofCaveandKarstStudies, v.70,no.1,p.13–24. JournalofCaveandKarstStudies, April2008 N 13

PAGE 2

Aroundtheedgesofthecollapsedareas,theremaining portionsofthehighvaultedceilingareperforatedby dozensofskylightsandthewallsareadornedwithgiant speleothemsandunusualruiniformfeaturesshapedlike towers,fins,andbuttresses.Asinterestingasthese erosionalanddepositionalfeaturesare,thegreatattraction forthesparseassemblageofpaleontologistswhohavebeen cominghereatlongintervalssinceFranzSikoraÂ’s1899 visitisthesoftsandysedimentsthatmantlethefloor, becausethesecontainwell-preservedbonesofextinct megafauna,aswellasadiversecollectionofbonesof smallervertebratesandshellsoflargeornatelandsnails andotherinvertebrates.Remarkably,despitetheproven potentialofthesiteandthenumberofpaleontologistswho havevisited,previousexcavationshavebeenlargely undocumentedanduncontrolled. Wereportheretheresultsofabriefsurveyin2000and amoreextensiveexcavationprogramin2003thatincludes preliminarystratigraphicdescriptions,faunallists,radiometricdating,andpaleoecologicalinferences.Thegoalsof thisworkincludednotonlyrecoveringmorecomplete materialforraretaxadescribedpreviouslyfromthesite, butalsothecreationofaprovisionalchronostratigraphic contextwithwhichtointerpretthismaterial.Thesenew detailsallowthefirstinferencesregardingthepaleoecology oftheregion,particularlythepossibleroleofnaturaland human-inducedchangesinthehistoryofthesite. S ITE D ESCRIPTION Asidefromtherichnessandpreservationofitsfaunal remains,thesiteisalsoimportantbecauseitistheonly documentedfossilvertebratesitefromtheentiresoutheasternportionoftheisland(S25 u 11 55 E46 u 37 59 ; Fig.1A).Situatedinanareaofrapidtransitionfrom semiaridbushlandtothewestandhumidrainforesttothe east,thesitepotentiallysamplesbothtypesofenvironments. Theeolianitebodyhereiscomplexandextensive, composedofthreethicklayersofcalcareoussandstonethat varyconsiderablyintextureandcolorandinmanyplaces exhibitsharpzonesofcontact.Theentiredepositlies unconformablyondensegraniticandgneissicrocksthat areexposedintheseacliffshereandoutcropinlandasa seriesofweatheredhills,ridges,andinselbergs. Figure1.A.LocationofAndrahomanaCave,onthesoutheast cornerofMadagascarnearTolagnaro(formerlyknownasFt. Dauphin).EntrancecoordinatesareS25 u 11 55 E46 u 37 59 r Figure1.B.SurveymapofthefloorofAndrahomanaCave. Placenamessuppliedbytheauthors.CapitallettersB,F, andIindicatelocationsofexcavationsfromthe2000and 2003expeditions.Thecaveroofiscollapsed,exceptinthe areasofDaliÂ’sPlayhouse,theTripleBridge,and AeolianPipes. N EWFINDINGSAT A NDRAHOMANA C AVE ,S OUTHEASTERN M ADAGASCAR 14 N JournalofCaveandKarstStudies, April2008

PAGE 3

Theoverlyingthree-parteolianiteinthisareaprobably correspondsto,beginningatthebottomlayer,Tatsimian, Karimbolian,andAepyornian(Besairie,1972).Although thismaterialhasnotbeendated,itismostlikelymid-to late-Pleistoceneinage;eachunitprobablyreflectsthe depositionalresultsofapost-glacialsea-levelrisefollowing amajorice-agecycle.Theseunitsarelikelytohavebeen depositedatleast100,000yearsapart,reflectingthemajor componentofglacial-interglacialcycles.Thusthecave’s materialwouldhavebeenatleast200,000yearsinthe making,althoughitisnotknownwhetherthesedeposits arefromthemorerecent,orearliercycles. Inanycase,thesedunesandswereconsolidated throughaprocessofrecrystallizationinwhichoverlying acidicwatersfromthesurfacepercolatedthrough, dissolvinghighlysolublearagoniticcalciumcarbonate andreplacingit,ondrying,withharderandlesssoluble calcite.Theadditionofsilicafromoverlyingsoilsand atmosphericdustwouldhavehelpedfurtherhardenthis naturalcement.Thisrecrystallizationphasewasfollowed muchlaterbydissolution,asgroundwaterflowingbetweentheporouseolianitesandtherelativelyimpervious basementrocksetchedawaythesandstonefrombelow, producingacavern.Furtherdissolution,probablyaidedby subsequentmarineincursion,hasledtothecollapseofthis caveinthecentralchamber.Considerabletime,perhaps 100,000yearsormore,passedbeforethecavecollapsed,as thelargeinactivespeleothemsofupto2metersindiameter aroundtheedgeoftheopeningwouldhaveformedinthe humidconditionsofamoreenclosedcave.Growthratesof similar-sizedspeleothemselsewhereinMadagascarandin similarclimatesinAfricahavebeenmeasuredbyU-series techniques(Brooketal.,1990;Burneyetal.,1994,1997). Thesespeleothemsarelikelytobeatleastasoldifthey formedatsimilarrates. Sincethiscollapse,Andrahomanahasservedasa naturaltrapforterrestrialsedimentsandbioticremains. Evidenceinthelowerpartsofthecaveshowsthateither highsea-standsorextrememarineeventshavebroughtthe seabackintothecaveonoccasion,addingmarine materialstothesedepositsaswell. Workfromthe2000and2003expeditionsincluded mappingwithcompass,clinometer,andhip-chain,surface survey,sedimentdescriptionfrombucket-augercoresand testpits,andcontrolledexcavation.Contrarytotheclaim ofAlluaud(1900)thatSikorahadexcavatedtheentire place( toutfouille )theyearbefore,coupledwiththedigging effortsofsubsequentvisitors(seeAppendix1),aconsiderableamountofpreviouslyunexcavatedsedimentswas located.Earlierexcavationswereeasilyrecognizedbythe derangednatureofthesedimentlayersandlackoflarge bonesintheoldspoilpiles.Fine-screeningofthismaterial rejectedbyearlierinvestigators,however,didprovidea richyieldofsmallerbones,teeth,andartifacts,and providedinsightintothecoarsermethodsofearlier investigators. ControlledexcavationswerecarriedoutatlocationBin 2000,andlocationsFandIin2003(Fig.1B).Eachshowed asimilarpattern:anupperunit,generally10–20cmthick, ofbrownhumicsiltysand,containingprimarilybonesof thepresentfauna,andamuchthickerlowerunit,ofca. 1mto 3mthickness,ofcoarser,lighter-coloredsands, includingalargercomponentofremainsofextinctfauna. Artifactsofpost-Europeancolonizationprovenance,includingglassandspentcartridges,occurnearthetopofthe upperunit,withnodefiniteevidenceofhumansdeeperin theunit.DatesonbonesfromthisunitarelateHolocene, butarenotinstratigraphicorder(Table1). Thelowerunitlackshumanartifactsandcontains bonesdatingtothemid-Holocenethatarenotinconsistent stratigraphicorder.ThissuggeststhatlateHolocene sedimentsherehavebeendisturbedandperhapslargely erodedbyextrememarineevents,perhapsoneormore tsunamiorstormoverwashes.Recentevidenceforthe BurckleImpactCrater,betweensouthernMadagascarand Antarctica(Masseetal.,2006)suggeststhatatleastone mega-tsunamioffargreatermagnitudethananyrecorded historicallywasgeneratedinthisareaintheHolocene,as supportedbytheoccurrenceofchevrondunedepositsin thecoastalregionsofextremesouthernMadagascar. F AUNA Table2providesafulllistoftaxafoundinthesubfossil depositsatAndrahomanain2003andadditionaloccurrencesinpriorcollections.Certainspeciesarerepresented inbothcollections,butafewarerepresentedinsmall numbersonlyinearliercollections,andmanyotherspecies (especiallynon-mammals)arereportedhereforthefirst time. Taxatypicallyconfinedtoeitherdryorhumidhabitats arefoundinthecollection,asmightbeexpectedgiventhat Andrahomanaissituatedinanareaofabrupttransition fromsemiaridspinybushlandtothewestandhumid rainforesttotheeast,adividecreatedinlargepartbythe AnosyenneMountains.Extanttaxatypicalofdrier habitatsinclude Numidameleagris amongthebirds, Geogaleaurita and Echinopstelfairi amongtheafrosoricidans, Triaenopsfurculus amongthebats, Lemurcatta amongtheprimates,and Macrotarsomysbastardi among therodents.Thesetaxaarefoundinwesterndrydeciduous forestsandsouthernspinybushareasofMadagascar (CarletonandGoodman,2003;EgerandMitchell,2003; Goodman,2003). Lemurcatta regularlyfrequentsthe premisesofthecavetodayanditislikelytheothertaxado aswell.Inadditiontotheseextanttaxa,severalextincttaxa recoveredfromAndrahomanaalsolikelyoccupieddrier habitats,includingtheprimates Megaladapisedwardsi and Hadropithecusstenognathus ,whicharedominantelements ofthesubfossilfaunaoftheextremesouthandsouthwest, butnotelsewhere(seeGodfreyandJungers,2002,fora review),andtherodent Hypogeomysaustralis ,althoughthe D.A.B URNEY ,N.V ASEY ,L.R.G ODFREY ,R AMILISONINA ,W.L.J UNGERS ,M.R AMAROLAHY AND L.R AHARIVONY JournalofCaveandKarstStudies, April2008 N 15

PAGE 4

naturalhistoryanddistributionofthelatterspeciesisstill poorlyknown(GoodmanandRakotondravony,1996). Extanttaxaconfinedtomorehumidhabitatsinclude Microgaleprincipula amongtheafrosoricida, Avahilaniger amongtheprimates,and Nesomysrufus amongtherodents (Jenkins,2003;Ryan,2003;Thalmann,2003).These speciesarefoundonlyinearliercollections;nonewere recoveredfromourexcavationsin2003despitedeliberate effortnottobiasourcollectioninfavoroflarger-bodied taxa. Avahilaniger and Nesomysrufus nolongerliveinthe regionandtheiroccurrenceinthesubfossildepositsof Andrahomanahasnotbeenpreviouslypublished. Avahi laniger isrepresentedbyasingleskullapparentlyfoundby SikoraandcurrentlyhousedintheBritishMuseum (NaturalHistory)(BMNHZD.1960.2103),andanumber ofskullscollectedbySikoraandhousedintheBMNH wereattributedto Nesomysrufus .Thesubfossiloccurrence ofanotherlocallyextinctspecies, Macrotarsomyspetteri furthersuggestsatleastslightlymorehumidclimatic regimesinthisregionofMadagascarinrecentmillennia (Goodmanetal.,2006). Macrotarsomyspetteri wasonly recentlydescribedandnamed,knownfromasingle specimencollected450kmnorthwestofAndrahomanain theMikeaforestnorthofToliara(Goodmanand Soarimalala,2005). AmongthevertebratesrecoveredatAndrahomana,the largesttaxaareextinct,asarethelargestspeciesinseveral smaller-bodiedorders.Elephantbirds( Aepyornis Mullerornis ),giantlemurs( Hadropithecusstenognathus Archaeolemurmajori Megaladapisedwardsi ,and Pachylemur insignis ),thecarnivore, Cryptoproctaspelea ,thepygmy hippo, Hippopotamuslemerlei ,thegiantjumpingrat, Hypogeomysaustralis ,andthegiantlandtortoise, Geochelonegrandidieri ,aregone,knownonlyfromfossils. Ostensiblysmallertaxafaredbetter.Nevertheless,analyses underwayontheminifaunaofAndrahomanademonstrate thatsmallertaxawerenotimmunetoanthropogenicand climaticchangesoccurringinrecentmillennia.Inarecent publication,wedescribeanewsubfossilshrew-tenrecfrom Andrahomana, Microgalemacpheei ,todate,theonly knownafrosoricidanthathasgoneextinctinrecent millennia(Goodmanetal.,2007).Furthermore,biostratigraphicanalysisdemonstratesthedeclineofendemic rodentsintandemwithincreasesintherepresentationof theintroducedrodents Rattus sp.and Musmusculus (Vasey andBurney,unpublisheddata). Domesticdogs( Canislupusfamiliaris ),cattle( Bos indicus ),andthepossiblynon-nativesoricomorphid, Suncusmadagascariensis ,arealsopresentamongthebones recoveredin2003,amplyindicatingthepost-human settlementcontextofthemorerecentlydepositedsediments inthecave.Wenote,asdidWalker(1967),thepresenceof burnedbonesonthesurface.However,weareofthe opinionthatthesearefullymodernremains,probablyleft bylocalfishermen,whooccasionallyenterthecavetocook ameal.Ontheotherhand,askullof Archaeolemur from Table1.Accelerationmassspectrometer 14 CpurifiedcollagendatesfromAndrahomanaCave. Material/Method AgeB.P. 6 1 s Calibrated2 s Range a SampleI.D.Provenanceor AccessionNo. Lab. Number d 13 CComments/Other Hadropithecusstenognathus bone 6724 6 54B.P.7660–7490BJ-HS-1AA-45963 9.1Humeralfragment T-16044A Hypogeomysaustralis bone4440 6 60B.P.5840–4420NotGivenBeta-73370…InGoodmanandRakotondravony(2006) Hypogeomysaustralis tooth1536 6 35A.D.428–618AHA-I-L1-CD1,2NZA-18996 20.0… B.P.1522–1332AHA-03-1R-28421/1 Macrotarsomyspetteri bone2480 6 40B.C.790–410AHA-I-L1-CD3Beta-212739…Demi-mandible;toosmallfor 13 C B.P.2740–2360 Macrotarsomyspetteri bone1760 6 40A.D.150–390AHA-I-L6-CD1,2Beta-212738…Demi-mandible;toosmallfo r 13 C B.P.1800–1560 Megaladapis sp.bone4566 6 25B.P.5436–5059AHA-03-7NZA-18997 20.1Infantfemur R-28421/7 a CalibrationsfromStuiveretal.(1998). N EWFINDINGSAT A NDRAHOMANA C AVE ,S OUTHEASTERN M ADAGASCAR 16 N JournalofCaveandKarstStudies, April2008

PAGE 5

Andrahomanaoncethoughttobearsignsofafatal crushingblowfromanaxe,orsimilartool(Walker, 1967),hasrecentlybeenradiocarbondatedtothemidHolocene(3975 6 53 14 CyrBP;Perezetal.,2005),making itentirelyunlikelythathumansdirectlycausedthedeathof thisanimal.Presumablymostofthelargertaxaenteredthe cavebyaccidentallyfallingthroughoneofmanyskylights, whichtodayarefrequentlycamouflagedatthesurfaceby vegetation.Manyofthesmalleranimalswerelikely depositedbyraptorsfromroostsitesintheformof discardedpreyremainsorregurgitatedpellets. Anoverviewoftheprimatesrecoveredbearsmentioningbecausethisisthemosttaxonomicallyrichorder representedatAndrahomanaandbecausethemost significantfindofthe2003excavationsdrawsfromthis order:theremainsof Hadropithecusstenognathus .The faunallistincludeselevenspeciesofprimates,consistingof sixextinctspecies,andfiveextantspecies( Lemurcatta Propithecusverreauxi Microcebus sp., Cheirogaleusmedius and Avahilaniger ),oneofwhich( Avahilaniger )nolonger livesonthissideoftheAnosyenneMountainsdivideas discussedabove.Onlytwospecimensof Cheirogaleus are presentintherecentcollectiondespitethereported coexistenceinthegeneralregionofTolagnarotodayof C.medius C.crossleyi ,and C.major (Hapkeetal.,2005). Presently, Lemurcatta regularlyfrequentsthecave,and Propithecusverreauxi and Microcebus sp.canbeobserved innearbyforests.Theabundant Microcebus subfossilsin thecavedepositsappeartobelongtoasinglespecies,most likely M.griseorufus ,anotherdenizenofdryforest. Oftheextinctprimatespecies,four( Hadropithecus stenognathus Archaeolemurmajori Megaladapisedwardsi and Pachylemurinsignis )arewellrepresentedinsubfossil depositsfromAndrahomana.Two( Megaladapismadagascariensis and Archaeolemur sp.,cf. edwardsi )arepoorly represented. Megaladapismadagascariensis isrepresented byasinglespecimen(distalradius)intheSikoracollection intheViennaNaturhistorischesMuseum(VNM 1934.V.56).Mostofthe Archaeolemur fromAndrahomana fitcomfortablywithinthesizerangeof A.majori ,butthere areafewspecimensinboththeBMNHandVNM collections(representingbothpostcranialandcranial fragments)thatarewelloutsidethisrangeofvariation, andappeartorepresent A.edwardsi (Godfreyetal.,1999; LRG,unpubl.observations). Numerouselementsofasinglesubadult Hadropithecus stenognathus wererecoveredfromlocationI(Godfreyet al.,2006a).Inadditiontoisolatedteethandcranial fragments,theseelementsincludethefirstassociatedforeandhind-limbbones,andanumberofpreviouslyunknown elements:thefirstscaphoid,hamate,metacarpals,scapular fragment,wholesacrum,vertebrae(includingthefirst severalcaudalvertebrae),andribs(Godfreyetal.,2006a; Lemelinetal.,2006,2008).AlanWalkerpointedoutthe possibilitythatthecranialfragmentsbelongedtoaskull thatSikorahadfoundin1899atAndrahomanaandthat LorenzvonLiburnau(1902)described.TimRyanhasnow usedCTscanningofthenewlyfoundorbitsandSikoraÂ’s craniuminViennatoconfirmthisassociation(Ryanetal., 2008). Thesediscoverieshaveledtoareassessmentofprior hind-limbattributionsfor Hadropithecus ,andare-evaluationofitslocomotorbehavior.Thenewlydiscovered skeletalelementsconfirmthehind-limballocationsfor Hadropithecus describedbyGodfreyetal.(1997)and refuteearlierattributionsbyLamberton(1938).They provethat Hadropithecus ,like Archaeolemur ,hadalong tail,relativelyshortlimbs,andarobustbodybuild.Of particularinterestarethenewcarpalsandmetacarpalsthat provideevidenceforpronogradequadrupedalismand terrestriality(Godfreyetal.,2006a;Lemelinetal.,2006, 2008).Anisolateduppermolarof Hadropithecus from AHA-IwassectionedbyGarySchwartz,andfromthis specimen,dataonenamelthickness,enamelprismcharacteristics,andthechronologyofdentaldevelopmentin Hadropithecus werederived(Godfreyetal.2005,2006a,b). Researchonthedentalmicrostructureofthatmolarhas demonstratedthatmolarcrownformationtimewas prolongedin Hadropithecus andismorelikethatofextant hominoids(gorillas,chimpanzees,orangutans)thanextant lemursorevenotherextinctgiantlemurs,whetherlargeror smallerinbodysize.Thecuspalenamelof H.stenognathus isnotasthickorheavilydecussatedasin Archaeolemur ; thesedifferences,alongwithdifferencesintheirstable isotopesignatures,underscorevariationintrophicadaptationswithintheArchaeolemuridae(Godfreyetal.,2005; Godfreyetal.,2008). Terrestrialgastropodsarewell-representedinthe Andrahomanafossilrecord.Someofthetypespresentas fossilsarestilllivinginthearea,includinglargespeciesin thegenera Helicophanta Tropidophora and Clavator WithinthePleistoceneeolianites,shellsofa Clavator specieswerefairlycommon.Thegeneralimpressionfrom theHolocenestratigraphyisthattheprehumansnailfauna hasgenerallysurvivedtothepresentinthearea,but detailedanalysis,includingidentificationtospecies,should beundertaken.TheAndrahomanavicinityisrichinboth fossilandlivingsnails.Emberton(1997)hasidentified80 extantlandsnailspeciesinextremesoutheasternMadagascar. D ISCUSSION B IOGEOGRAPHYAND P ALEOECOLOGY LocatedonthesoutheasterncoastofMadagascar, AndrahomanaisjustwestoftheAnosyenneMountains, andthusindryhabitat.Today,thehighplateauofcentral Madagascarbehavesasaformidablebarrierseparatingthe wetforestsoftheeastandthedryforestsofthewest,and therearestrikingdifferencesbetweentherespectivefaunas aswellasflorasoftheseregions.Theboundarybetween wetanddrybiotaisverysharpinplaces;theeasternand D.A.B URNEY ,N.V ASEY ,L.R.G ODFREY ,R AMILISONINA ,W.L.J UNGERS ,M.R AMAROLAHY AND L.R AHARIVONY JournalofCaveandKarstStudies, April2008 N 17

PAGE 6

Table2.ListofsubfossiltaxarecoveredfromdepositsatAndrahomanabyB urney etal. (thispublication)andby previouscollectors. ClassOrderFamilyBurney etal. PreviousCollectors AvesAepyornithiformesAepyornithidae Aepyornis sp. a,b,c Mullerornis AnseriformesAnatidae Centrornis sp. a ProcellariiformesProcellariidae Puffinus sp. FalconiformesFalconidae Falconewtoni GalliformesNumidae Numidameleagris GruiformesTurnicidae Turnixnigricollis Rallidae Gallinulachloropus ColumbiformesColumbidae Streptopeliapicturata PsittaciformesPsittacidae Coracopsisvasa CuculiformesCuculidae Couacursor Couacristata StrigiformesTytonidae Tytoalba Strigidae Otusrutilis ApodiformesApodidae Apus sp. CoraciiformesUpupidae Upupamarginata PasseriformesAlaudidae Mirafrahova Sylviidae Nesillas cf. lantzii Thamnornischloropetoides Monarchidaecf. Tersiphonemutata Zosteropidae Zosteropsmaderaspatana Vangidae Vangacurvirostris Leptopterusviridis Cyanolaniusmadagascarinus Corvidae Corvusalbus Ploceidae Ploceussakalava Foudiamadagascariensis MammaliaAfrosoricidaTenrecidae Geogaleaurita ‘‘ Cryptogaleaustralis ’’ c Microgalebrevicaudata Microgalelongicaudata Microgalenasoloi Microgalepusilla Microgalemacpheei ‘‘ Paramicrogaledecaryi ’’ c ( Microgale principula ) Echinopstelfairi Setifersetosus Tenrececaudatus ‘‘ Centetesecaudatus ’’ b,c SoricomorphaSoricidae Suncusmadagascariensis ChiropteraPteropodidae‘‘ Pteropusedwardsi ’’ a,b ( Pteropus rufus ) Eidolondupreanum Rousettusmadagascariensis ‘‘ Pelophilousmadagascariensis ’’ c Hipposideridae Hipposideroscommersoni Triaenopsfurculus Vespertilionidae Miniopterusgleni Molossidae Mormopterusjugularis Mopsleucostigma N EWFINDINGSAT A NDRAHOMANA C AVE ,S OUTHEASTERN M ADAGASCAR 18 N JournalofCaveandKarstStudies, April2008

PAGE 7

westernslopesofAndohahelainsoutheasternMadagascar provideagoodexample(Goodman,2000).Yetinother places,thehighplateauservesasaregionofconsiderable exchange(seeforexample,Goodmanetal.,2007).The sharpecotonalboundariesthatexistontheeasternand westernslopesofmountainssuchasAndohahelamaybe recenteffectsofrapidorogeny(seedeWit,2003).Itisalso clearthat,inthepast,thegeographicboundariesof typicallywet-anddry-adaptedspecieswerequitefluid; fingersofeasternforestoncespreadacrosstheisland. ClassOrderFamilyBurney etal. PreviousCollectors PrimatesCheirogaleidae Microcebus sp.‘‘ Cheirogaleusmyoxinus ’’ c Cheirogaleusmedius ‘‘ Cheirogaleussamati ’’ a Lemuridae LemurcattaLemurcatta c,d Pachylemurinsignis ‘‘ Lemurinsignis ’’ a,b,c Indriidae PropithecusverreauxiPropithecusverreauxi a,b,c,d Avahilaniger d Archaeolemuridae ArchaeolemurmajoriArchaeolemurmajori a,c,d Archaeolemuredwardsi d HadropithecusstenognathusHadropithecusstenognathus d Megaladapidae MegaladapisedwardsiMegaladapisedwardsi a,b,c,d,e Megaladapismadagascariensis d CarnivoraViverridae‘‘ Viverrafosa ’’(var.nov. alluaudi largesize) a,c Eupleridae Cryptoproctaferox (var.nov. spelea largesize) b,c Canidae CanislupusfamiliarisCanisfamiliaris ArtiodactylaBovidae Bosindicus ‘‘ Bosmadagascariensis ’’ a,b,c Hippopotamidae HippopotamuslemerleiHippopotamuslemerlei a,d RodentiaNesomyidae Eliurusmyoxinus Eliurus sp.(largerspecies) HypogeomysaustralisHypogeomys nov.sp. c Macrotarsomysbastardi MacrotarsomyspetteriNesomysrufus d Muridae MusmusculusMusmusculus b Rattus sp.‘‘ Musdecumanus ’’ a,b,c ReptiliaTestudinesPelomedusidae Pelomedusasubrufa Testudinidae Dipsochelys sp.‘‘ Geochelone ( Testudo ) grandidieri ’’ a,b Geocheloneradiata ‘‘ Geochelone ( Testudo ) radiata ’’ b SquamataChamaeleonidae Furcifer cf. verrucosus Gekkonidae Paroedura sp. Gerrhosauridae Zonosaurus cf. trilineatus SerpentesBoidae Acrantophis sp. Colubridae Leioheterodon sp. CrocodyliaCrocodylidae Crocodylusniloticus Crocodilusrobustus a,b,c AmphibiaAnuraMantellidae Laliostomalabrosa Ranidae Ptychadenamascareniensis Microhylidae Scaphiophryne sp. Note:Furtherinformationconcerningpreviouscollectionscanbefoundi nAppendix1.TheAlluaud,GaubertandGrandidierCollectionswereallpub lishedinGrandidier (1902).Earliernomenclature(i.e.,synonyms)ofvarioustaxafromprevi ouscollectionsisretainedinthepreviouscollectionscolumn,listedin quotesandparentheses. a AlluaudCollection b GaubertCollection c GrandidierCollection d SikoraCollection e WalkerCollection Table2.Continued. D.A.B URNEY ,N.V ASEY ,L.R.G ODFREY ,R AMILISONINA ,W.L.J UNGERS ,M.R AMAROLAHY AND L.R AHARIVONY JournalofCaveandKarstStudies, April2008 N 19

PAGE 8

Isolatedforestswitheasternelementsoccurinthewest, e.g.,theregionofZombitse-Sakaraha,eastofMorondava, andthemistoasisoftheAnalavelonaMassif,nearToliara (DuPuyetal.,1994;LangrandandGoodman,1997; Carletonetal.,2001,Ganzhorn,2006),andtypicaldrylovingspeciesoccurinsomeforestsoftheEast(reviewedin Ganzhorn,2006).Anumberoflemurspeciesandother faunaaretoday,orwereintherecentpast,spreadacross theeast-westdivide.Forexample,itisnowknownthat Hapalemursimus ,restrictedtodaytohumidforestsinthe southeast(Irwinetal.,2005),oncelivedintheextreme north(AnkaranaandMt.desFranc ais),centralMadagascar(Ampasambazimba),thenorthwest(Anjohibe),and theextremewest(Bemaraha)(Godfreyetal.,2004). Paleoecologicaldataconfirmthatthesouthwestwas recentlymoisterthanitistoday(Burney,1993),andthe Holocenedistributionsofrodentsandbirdsprovethe existenceofacorridorsouthof 20 u latitudethrough whichanimalsadaptedtomoistenvironmentswereableto spreadwestward(GoodmanandRakotondravony,1996; GoodmanandRakotozafy,1997). Thereareseveralpossibleexplanationsfortheobserved mixtureofwet-anddry-adaptedfaunaatAndrahomana. Oneisthatthesetaxawerenotactuallysynchronousbutare amixedassemblageofspeciesthatlivedintheregionat differenttimes;theirpresenceatAndrahomanawouldthen reflecttemporalfluctuationintheclimateandvegetationof thearea.Anotheristhatthetoleranceofcertainspeciesto variationinhabitatwasgreaterinthepastthanitistoday.A thirdpossibilityisthatthehabitatsthemselveswerericher, supportingalargernumberofsympatricspecieswith differentresourcerequirements.Directandindirectanthropogenicfactors(hunting,deforestation,introductionof exoticspecies)mayhavecontributedtochangesand diminutionsinthegeographicrangesofendemicspecies. Regardlessofwhetherpasttransitionsbetweenwetanddry biomesweregradualorabruptinsouthernMadagascar (eithertemporallyorspatially),themixtureofwet-anddryadaptedspeciesatAndrahomanacertainlyreflectsthe positionofthecavewithinthezoneoftransition. C AVE G ENESISAND C HRONOLOGY AlthoughtheeolianiteunitsatAndrahomanaare almostcertainlytoooldforradiocarbondatingandeven lessconventionalmethodssuchaswholerockamino-acid racemization(seeHeartyetal.,2000),somerough estimatescanbemadeforAndrahomanaÂ’schronologyof formation.Forexample,thecavehasbeenformedby solutionsincethethirdeolianiteunitwasdeposited(the topmostAepyornian).Thatsurelymeansthatthecavewas formedsincethelastinterglacialdepositionalevent,a minimumofapproximately130kyrBP(isotopestage5e). Speleothems 1mindiameterarefoundinside,and thesewerealmostcertainlydepositedpriortocavecollapse (theairistoodryinthecavetodaytopermitsignificant dripstoneformation).Similarformationsinanothercavein Madagascarrequiredmorethan40kyrtoreachasimilar size(Burneyetal.1997).Combiningtheseobservations, thefollowingscenarioisplausible,butnotprovenbythe datapresented:1)thelimestonewaslaiddownoverthelast quartermillionyearsormore;2)thecavewasexcavatedby ground-waterdissolutionoverthelastca.100,000years;3) afterperhaps40,000yearsofspeleothemgrowth,theceiling ofthecavebegantocollapse,dryingouttheinsideofthecave andpermittingrapidsedimentationfromterrestrialsources; 4)duringtheHolocene,sealevelreachedthebottomofthe lowermostcaveentrance;5)extrememarineevents,since mid-Holocenetimes,occasionallybreachedtheinteriorof thecave,contributingtostructuralcollapseanddepositing marinesandsandothermaterialinsideuptothehighest partsofthefloor, 8mabovepresentsealevel. Iftheselinkedscenariosarecorrect,mid-to-late Pleistocenefossilsmaybeincorporatedintotheeolianite, theuppermostsedimentsareamixtureofHolocene terrestrialandmarinematerials,anddeepdown,below anyexcavationlevelsreachedsofar,andperhapsinfissure fillshigherup,therecouldbemid-to-latePleistocenecave breccias,perhapscontainingbonesandshellsfromtimes poorlysampledinfossilsitesinMadagascar.Thus Andrahomanahassomepotentialforfillingthelater portionofthelargeCenozoicblindspotthathas characterizedpaleontologyinMadagascar(seeTable1in Krauseetal.,2006).Forresolvingtheinterestingquestions regardingthetimingofhumanarrivalandthefateofthe extinctsubfossilfauna,however,thecaveisnotideal. Becauseoftheapparentdisturbance,removal,and reworkingofHolocenesediments,onlylimitedstratigraphicresolutionhasbeenobtainedintheexcavatedsediments. Itisabundantlyclearfromthedetailedhistoryofsite explorationanddevelopmentinAppendix1that,despite itsextremelyremotelocation,thesitehasattractedagreat dealofattentionfrompaleontologists.Ironically,however, noneoftheseaccomplishedcollectorsprovidedany stratigraphicdetailorinformationconcerningtheageor provenanceofthemanyfossilscollected.Thisisunfortunate,asmuchoftheaccessibledepositswereremovedor disturbedbeforetheinitiationofthepresentstudy. Confirmingassociationsandfindingpreviouslyundescribedelementsof Hadropithecusstenognathus would alonejustifycontinuedefforts.Thediscoveryofassociated forelimbandhind-limbboneshasindeedinformedand improvedreconstructionsofthebehaviorof Hadropithecus (seereviewsbyGodfreyetal.,1997,2006a).Farfrom completingthestudyofthisinterestingcaveandexhaustingitspotential,ourrecentworkatAndrahomana,which hasyieldedasitemap,newinsightsonthestratigraphyand geochronologyofthesite,evidenceformega-tsunamisor otherextrememarineevents,andthecollationofhistorical recordsforthesite,merelyunderscorestheneedformore research.Twohighprioritiesforfutureresearchwouldbe larger-scaleexcavationandmethodicalsearchingfor Pleistoceneagebrecciasandfissurefills. N EWFINDINGSAT A NDRAHOMANA C AVE ,S OUTHEASTERN M ADAGASCAR 20 N JournalofCaveandKarstStudies, April2008

PAGE 9

Untilrecently,thegreatisolationofthesitehas providedsomeprotectionfromvandalismandlootingof thenaturaltreasuresofthesite.Theornatespeleothems, abundantfossilbonesandshells,andmineraldepositsofthe caveareaarepotentiallyvulnerabletothegrowingillegal tradeinthesematerials(Krauseetal.,2006).Local authorities,quitenaturally,areanxioustoseeaccesstothe siteimprovedinanticipationofecotourismrevenues.Several localtouroperatorsnowoffervisitstothesiteandmoreare plannedaccordingtolocalauthorities.Itisimportant, however,thatanyfuturedevelopmentplansforthearea includeprovisionofsomeformofprotectionforthesite. A CKNOWLEDGMENTS ThisresearchwassupportedbyNSFBCS-0129185(to DAB,LRG,andWLJ),BCS-0237388(toLRG),anda CollegeofLiberalArtsandSciencesResearchStipendfrom PortlandStateUniversity(toNV)andconductedunder collaborativeagreementswiththeDe partementdePale ontologieetd’AnthropologieBiologique,Universite d’Antananarivo,andtheAcade mieMalgache.PaulS.Martinand AlanC.Walkerencouragedustoinvestigatethesiteand providedunpublishednotesontheir1966expedition.Steve Goodman,ChrisRaxworthy,andAlexanderWolfassisted inidentifyingspecimens,andMichaelCarletonandSteve Goodmanprovidedhelpfulreviewsofthemanuscript. R EFERENCES Alluaud,C.,1900,Correspondance(LetteraddressedtoMr.Guillaume GrandidierinFortDauphin,August30,1900):BulletinduMuse um Nationald’HistoireNaturelleParis,v.6,p.327–330. Battistini,R.,1964,E tudege omorphologiquedel’extre ˆmesudde Madagascar:LaboratoiredeGe ographie,Antananarivo;E ditions Cujas,Paris. Besairie,H.,1972,Ge ologiedeMadagascarI.Lesterrainsse dimentaires: AnnalesGe ologiquesdeMadagascar,v.35,p.1–463. Brook,G.A.,Burney,D.A.,andCowart,J.B.,1990,Desertpaleoenvironmentaldatafromcavespeleothemswithexamplesfromthe Chihuahuan,Somali-Chalbi,andKalaharideserts:Palaeogeography, Palaeoclimatology,Palaeoecology,v.76,p.311–329. Burney,D.A.,1993,LateHoloceneenvironmentalchangeinarid southwesternMadagascar:QuaternaryResearch,v.40,p.98–106. Burney,D.A.,Brook,G.A.,andCowart,J.B.,1994,AHolocenepollen recordfortheKalahariDesertofBotswanafromaU-seriesdated speleothem:TheHolocene,v.4,no.3,p.225–232. Burney,D.A.,James,H.F.,Grady,F.V.,Rafamantanantsoa,J.-G., Ramilisonina,Wright,H.T.,andCowart,J.B.,1997,Environmental change,extinction,andhumanactivity:evidencefromcavesinNW Madagascar:JournalofBiogeography,v.24,p.755–767. Burney,D.A.,James,H.F.,Burney,L.P.,Olson,S.L.,Kikuchi,W., Wagner,W.L.,Burney,M.,McCloskey,D.,Kikuchi,D.,Grady, F.V.,Gage,R.,andNishek,R.,2001,Fossilevidenceforadiverse biotafromKaua’ianditstransformationsincehumanarrival: EcologicalMonographs,v.71,no.4,p.615–641. Burney,D.A.,Burney,L.P.,Godfrey,L.R.,Jungers,W.L.,Goodman,S.M., Wright,H.T.,andJull,A.J.T.,2004,Achronologyforlateprehistoric Madagascar:JournalofHumanEvolution,v.47,p.25–63. Carleton,M.D.,andGoodman,S.M.,2003, Macrotarsomys ,Big-footed mice, in Goodman,S.M.,andBenstead,J.P.,eds.,Thenaturalhistory ofMadagascar,ChicagoandLondon,UniversityofChicagoPress, p.1386–1388. Carleton,M.D.,Goodman,S.M.,andRakotondravony,D.,2001,Anew speciesoftufted-tailedrat,genus Eliurus (Muridae:Nesomyinae),from westernMadagascar,withnotesonthedistributionof E.myoxinus in ProceedingsoftheBiologicalSocietyofWashington,v.114,p.977–987. Decary,R.,1927,fortheyear1926,UnemissionscientifiquedanslesudestdeMadagascar:Bulletindel’Acade mieMalgache,v.9(nouvelle se rie),p.79–86. Decary,R.,1928,fortheyear1927,Contributiona `labotaniqueetla ge ologiedelare gionFort-Dauphin—Andrahomana:Bulletinde l’Acade mieMalgache,v.10(nouvellese rie),p.13–18. Decary,R.,andAndre Kiener,1971,Inventairesche matiquedescavite s deMadagascar:AnnalesdeSpe le ologie,v.26,p.31–46. deWit,M.J.,2003,Madagascar:Headsit’sacontinent,tailsit’sanislan d: AnnualReviewofEarthandPlanetarySciences,v.31,p.213–248. Dorr,L.J.,1997,PlantcollectorsinMadagascarandtheComoroIslands, RoyalBotanicGardens,Kew. DuPuy,B.,Abraham,J.P.,andCooke,A.J.,1994,Lesplantes, in Goodman,S.M.,andLangrand,O.,eds.,InventairebiologiqueFore ˆt deZombitse.Recherchespourlede veloppement,Se rieSciences biologiques,No.Spe cial,Antananarivo,Centred’Informationetde DocumentationScientifiqueetTechnique,p.15–29. Eger,J.L.,andMitchell,L.,2003,Chiroptera,Bats, in Goodman,S.M., andBenstead,J.P.,eds.,ThenaturalhistoryofMadagascar,Chicago andLondon,UniversityofChicagoPress,p.1287–1302. Emberton,K.C.,1997,Diversitiesanddistributionsof80landsnail speciesinsoutheastern-mostMadagascanrainforests,withareport thatlowlandsarericherthanhighlandsinendemicandrarespecies: BiodiversityandConservation,v.6,p.1137–1154. Ganzhorn,J.U.,2006,Lemurbiogeography, in Lehman,S.M.,and Fleagle,J.G.,eds.,Primatebiogeography,progressandprospects, NewYork,Springer,p.229–254. Geay,F.,1908,Rapportd’explorationsauxre gionsNord-Est,Sud-SudOuest,Sud,etSud-Sud-EstdeMadagascar,1904–1907:Privately printedbytheauthor,74AvenuedesGobelins,Paris. Godfrey,L.R.,1977,Structureandfunctionin Archaeolemur and Hadropithecus (subfossilMalagasylemurs):Thepostcranialevidence [Ph.D.thesis],Cambridge,HarvardUniversity. Godfrey,L.R.,andJungers,W.L.,2002,Quaternaryfossillemurs, in Hartwig,W.C.,ed.,Theprimatefossilrecord,NewYork,Cambridge UniversityPress,p.97–121. Godfrey,L.R.,Jungers,W.L.,Wunderlich,R.E.,andRichmond,B.G., 1997,Reappraisalofthepostcraniumof Hadropithecus (Primates, Indroidea):AmericanJournalofPhysicalAnthropology,v.103, p.529–556. Godfrey,L.R.,Jungers,W.L.,Simons,E.L.,Chatrath,P.S.,andRakotosamimanana,B.,1999,Pastandpresentdistributionsoflemursin Madagascar, in Rakotosamimanana,B.,Rasamimanana,H.,Ganzhorn,J.U.,andGoodman,S.M.,eds.,NewDirectionsinLemurStudies, NewYork,KluwerAcademic/PlenumPublishers,p.19–53. Godfrey,L.R.,Simons,E.L.,Jungers,W.L.,DeBlieux,D.D.,andChatrath P.S.,2004,Newdiscoveryofsubfossil Hapalemursimus ,thegreater bamboolemur,inwesternMadagascar:LemurNews,v.9,p.9–11. Godfrey,L.R.,Semprebon,G.M.,Schwartz,G.T.,Burney,D.A.,Jungers, W.L.,Flanagan,E.K.,Cuozzo,F.P.,andKing,S.J.,2005,New insightsintooldlemurs:thetrophicadaptationsoftheArchaeolemuridae:InternationalJournalofPrimatology,v.26,p.825–854. Godfrey,L.R.,Jungers,W.L.,Burney,D.A.,Vasey,N.,Ramilisonina, Wheeler,W.,Lemelin,P.,Shapiro,L.J.,Schwartz,G.T.,King,S.J., Ramarolahy,M.F.,Raharivony,L.L.,andRandria,G.F.N.,2006a, Newdiscoveriesofskeletalelementsof Hadropithecusstenognathus fromAndrahomanaCave,southeasternMadagascar:Journalof HumanEvolution,v.51,p.395–410. Godfrey,L.R.,Schwartz,G.T.,Samonds,K.E.,Jungers,W.L.,and Catlett,K.K.,2006b,Thesecretsoflemurteeth:Evolutionary Anthropology,v.15,p.142–154. Godfrey,L.R.,Jungers,W.L.,Schwartz,G.T.,andIrwin,M.T.,2008, Ghostsandorphans:Madagascar’svanishingecosystems, in Fleagle, J.G.,andGilbert,C.C.,eds.,ElwynSimons,ASearchforOrigins, NewYork,Springer,p.361–395. Goodman,S.M.,2000,DescriptionoftheReserveNaturelleIntegrale d’Andohahela,Madagascar,andthe1995biologicalinventoryofthe reserve:FieldianaZoology,newseries,v.94,p.1–7. Goodman,S.M.,2003,Checklisttotheextantlandmammalsof Madagascar, in Goodman,S.M.,andBenstead,J.P.,eds.,Thenatural D.A.B URNEY ,N.V ASEY ,L.R.G ODFREY ,R AMILISONINA ,W.L.J UNGERS ,M.R AMAROLAHY AND L.R AHARIVONY JournalofCaveandKarstStudies, April2008 N 21

PAGE 10

historyofMadagascar,ChicagoandLondon,UniversityofChicago Press,p.1187–1190. Goodman,S.M.,andRakotondravony,D.,1996,TheHolocene distributionof Hypogeomys (Rodentia:Muridae:Nesomyinae)on Madagascar, in Lourenc o,W.R.,ed.,Bioge ographiedeMadagascar, Paris,ORSTOMe ditions,p.283–293. Goodman,S.M.,andRakotozafy,L.M.A.,1997,Subfossilbirdsfrom coastalsitesinwesternandsouthwesternMadagascar:Apaleoenvironmentalreconstruction, in Goodman,S.M.,andPatterson,B.D., eds.,NaturalchangeandhumanimpactinMadagascar,Washington D.C.,SmithsonianInstitutionPress,p.257–279. Goodman,S.M.,andSoaramalala,V.,2005,Anewspeciesof Macrotarsomys (Rodentia:Muridae:Nesomyinae)fromsouthwestern Madagascar, in ProceedingsoftheBiologicalSocietyofWashington, v.118,p.450–464. Goodman,S.M.,Vasey,N.,andBurney,D.A.,2006,Thesubfossil occurrenceandpaleoecologicalimplicationsof Macrotarsomyspetteri (Rodentia:Nesomyidae)inextremesoutheasternMadagascar:ComptesRendusPalevol,v.5,p.953–962. Goodman,S.M.,Raselimanana,A.P.,andWilme ,L.(eds),2007, Inventairesdelafauneetdelafloreducouloirforestierd’AnjozorobeAngavo.RecherchespourleDeveloppement,Se rieSciencesBiologiqueno.24.Centred’InformationetdeDocumentationScientifiqueet Technique,Antananarivo,Madagascar. Goodman,S.M.,Vasey,N.,andBurney,D.A.,2007,Descriptionofanew speciesofsubfossilshrew-tenrec(Afrosoricida:Tenrecidae: Microgale )fromcavedepositsinsoutheasternMadagascar, in Proceedings oftheBiologicalSocietyofWashington,v.320,p.367–376. Grandidier,A.,1885,Histoiredelage ographie,Paris,Imprimerie Nationale. Grandidier,G.,1902,Observationssurlesle muriensdisparusde Madagascar:CollectionsAlluaud,Gaubert,Grandidier:Bulletindu Muse umd’HistoireNaturelle,v.8,p.497–505,587–592. Grandidier,G.,1903,Descriptiondel’ Hypogeomysaustralis ,unenouvelle espe `cederongeursubfossiledeMadagascar:BulletinduMuse um d’HistoireNaturelle,Paris,v.9,p.13–15,withfigures. Grandidier,G.,1928,Descriptiondedeuxnouveauxmammife `res insectivoresdeMadagascar:BulletinduMuse umNationald’Histoire Naturelle,Paris,v.34,p.63–70. Hapke,A.,Fietz,J.,Nash,S.D.,Rakotondravony,D.,Rakotosamimanana,B.,Ramanamanjato,J.-B.,Randria,G.F.N.,andZischler,H., 2005,Biogeographyofdwarflemurs,Geneticevidenceforunexpected patternsinsoutheasternMadagascar:InternationalJournalof Primatology,v.26,p.873–901. Hearty,P.J.,Kaufman,D.S.,Olson,S.L.,andJames,H.F.,2000, Stratigraphyandwhole-rockaminoacidgeochronologyofkey HoloceneandlastinterglacialcarbonatedepositsintheHawaiian Islands:PacificScience,v.54,p.423–442. HeimdeBalsac,H.,1972,Insectivores, in Battistini,R.,andRichardVindard,G.,eds.,BiogeographyandecologyofMadagascar,The Hague,Dr.W.Junk,p.629–660. Irwin,M.T.,Johnson,S.E.,andWright,P.C.,2005,Thestateoflemur conservationinsouth-easternMadagascar:populationandhabitat assessmentsfordiurnalandcathemerallemursusingsurveys,satellite imageryandGIS:Oryx,v.39,p.204–218. Jenkins,P.D.,2003, Microgale ,Shrewtenrecs, in Goodman,S.M.,and Benstead,J.P.,eds.,ThenaturalhistoryofMadagascar,Chicagoand London,UniversityofChicagoPress,p.1273–1278. Jungers,W.L.,1976,Osteologicalformandfunction:Theappendicularsk eleton of Megaladapis ,asubfossilprosimianfromMadagascar(Primates, Lemuroidea),[Ph.D.thesis],AnnArbor,UniversityofMichigan. Krause,D.W.,O’Connor,P.M.,Rasoamiaramanana,A.H.,Buckley, G.A.,Burney,D.A.,Carrano,M.T.,Chatrath,P.S.,Flynn,J.J., Forster,C.A.,Godfrey,L.R.,Jungers,W.L.,Rogers,R.R.,Samonds, K.E.,Simons,E.L.,andWyss,A.R.,2006,Theimportanceofkeeping theisland’svertebratefaunainthepublicdomain:Madagascar Conservation&Development,v.1,p.43–47. Lamberton,C.,1938,fortheyear1937.Contributiona `laconnaissancede lafaunesubfossiledeMadagascar,NoteIII,LesHadropithe `ques: Bulletindel’Acade mieMalgache(nouvellese rie),v.20,p.127–170. Langrand,O.,andGoodman,S.M.,1997,Bre `vedescriptionbiologiquede laregiondesfore ˆtsdeVohibasiaetd’Isoky-Vohimena, in Langrand, O.,andGoodman,S.M.,eds.,InventairebiologiqueFore ˆtde Vohibasiaetd’Isoky-Vohimena,Recherchespourlede veloppement, Se rieSciencesbiologiques,No.12,Antananarivo,Centred’InformationetdeDocumentationScientifiiqueetTechnique,p.11–28. Lemelin,P.,Hamrick,M.W.,Godfrey,L.R.,Jungers,W.L.,andBurney, D.A.,2006,Newhandbonesof Hadropithecusstenognathus : ImplicationsforthepaleobiologyoftheArchaeolemuridae:American JournalofPhysicalAnthropology,v.Suppl.42,p.120. Lemelin,P.,Hamrick,M.W.,Richmond,B.G.,Godfrey,L.R.,Jungers, W.L.,andBurney,D.A.,2008,Newhandbonesof Hadropithecus stenognathus :ImplicationsforthepaleobiologyoftheArchaeolemuridae:JournalofHumanEvolution,v.54,p.405–413. LorenzvonLiburnau,L.,1899,HerrCustosDr.Ludwigv.Lorenzberichtet u ¨bereinenfossilenAnthropoidenvonMadagascar:Anzeigerder KaiserlichenAkademiederWissenschafteninWien,v.36,p.255–257. LorenzvonLiburnau,L.,1901,U ¨ bereinigeResteausgestorbener PrimatenvonMadagaskar:DenkschriftderKaiserlichenAkademie derWissenschafteninWien,v.70,p.1–15. LorenzvonLiburnau,L.,1902,U ¨ ber Hadropithecusstenognathus Lz. NebstbemerkungenzueinigenanderenaustestorbenenPrimatenvon Madagascar:DenkschriftderKaiserlichenAkademiederWissenschafteninWien,v.72,p.243–254. LorenzvonLiburnau,L.R.,1905, Megaladapisedwardsi G.Grandidier: DenkschriftderKaiserlichenAkademiederWissenschafteninWien, v.77,p.451–490. MacPhee,R.D.E.,1987,TheshrewtenrecsofMadagascar:systematic revisionandHolocenedistributionof Microgale (Tenrecidae, Insectivora),NewYork:AmericanMuseumofNaturalHistory (AmericanMuseumNovitates,No.2889). Major,C.I.F.,1899,OnsubfossilmammalsfromMadagascar, in ProceedingsoftheZoologicalSocietyofLondon,v.64,p.988–989. Masse,W.,Bryant,E.,Gusiakov,V.,Abbott,D.,Rambolamana,G., Raza,H.,Courty,M.,Breger,D.,Gerard-Little,P.,andBurckle,L., 2006,HoloceneIndianOceancosmicimpacts:Themega-tsunami chevronevidencefromMadagascar[abs.]:Eos(Transactionofthe AmericanGeophysicalUnion),v.87,p.43B–1244. Perez,V.R.,Godfrey,L.R.,Nowak-Kemp,M.,Burney,D.A.,Ratsimbazafy,J.,andVasey,N.,2005,Evidenceofearlybutcheryofgiantlemurs inMadagascar:JournalofHumanEvolution,v.49,p.722–742. Ryan,J.M.,2003, Nesomys ,Redforestrat, voalavomena in Goodman, S.M.,andBenstead,J.P.,eds.,TheNaturalHistoryofMadagascar, ChicagoandLondon,UniversityofChicagoPress,p.1388–1389. Ryan,T.M.,Burney,D.A.,Godfrey,L.R.,Go ¨hlich,U.,Jungers,W.L., Vasey,N.,Ramilisonina,Walker,A.,andWeber,G.,2008,A reconstructionoftheViennaskullof Hadropithecusstenognathus : AmericanJournalofPhysicalAnthropology,v.suppl.46,p.184. Sikora,F.,1897,Septansa `Madagascar:Bulletindege ographied’AixMarseille,v.21,p.163–174,277–285. Simons,E.L.,Burney,D.A.,Chatrath,P.S.,Godfrey,L.R.,Jungers, W.L.,andRakotosamimanana,B.,1995,AMS 14 Cdatesforextinct lemursfromcavesintheAnkaranaMassif,northernMadagascar: QuaternaryResearch,v.43,p.249–254. Stuiver,M.,Reimer,P.J.,Bard,E.,Beck,J.W.,Burr,G.S.,Hughen,K.A., Kromer,B.,McCormac,F.G.,Plicht,J.v.d.,andSpurk,M.,1998, INTCAL98 14 C,Radiocarbonv.40,p.1041–1083. Tattersall,I.,1973,CranialanatomyoftheArchaeolemurinae(Lemuroidea,Primates):AnthropologicalPapersoftheAmericanMuseumof NaturalHistory,v.52,p.1–110. Tattersall,I.,1982,TheprimatesofMadagascar,NewYork,Columbia UniversityPress. Thalmann,U.,2003, Avahi ,Woollylemurs, Avahy Fotsy-fe Ampongy Tsarafangitra Dadintsifaky in Goodman,S.M.,andBenstead,J.P., eds.,TheNaturalHistoryofMadagascar,ChicagoandLondon, UniversityofChicagoPress,p.1340–1342. Thevenin,A.,1907,Notesurdesfossilesrapporte sdeMadagascarparM. Geay:BulletinduMuse umd’HistoireNaturelle,v.13,p.85–88. Walker,A.C.,1967,Locomotoradaptationinrecentandfossil Madagascarlemurs,[Ph.D.thesis],London,UniversityofLondon. Walker,A.,2002,LookingforlemursontheGreatRedIsland, in Bowen, T.,ed.,BackcountryPilot:FlyingAdventuresofIkeRussell,Tucson, UniversityofArizonaPress,p.99–104. Zapfe,H.,1971,CatalogusFossiliumAustriae,EinsystematischesVerze ichnisalleraufo ¨sterreichischemGebietfestgestelltenFossilien,Heft.XV: IndexPalaeontologicorumAustriae,Vienna,Springer-Verlag. N EWFINDINGSAT A NDRAHOMANA C AVE ,S OUTHEASTERN M ADAGASCAR 22 N JournalofCaveandKarstStudies, April2008

PAGE 11

A PPENDIX 1:H ISTORYOFTHEEXPLORATIONOF A NDRAHOMANA 1840AmapofthesoutherncoastofMadagascarpublishedbyLegue veldeLacombe(laterreproducedinA.Grandidier, 1885)showsavillageattheBayofAndrahomanalabeled‘‘Andrahoum.’’ 1855Marguin(citedbyGrandidier,1885)publishesamapofthesiteofthec averns(‘‘CapAndavaka’’–inMalagasy, the‘‘placeofthecaverns’’)andadetailedmapofthe‘‘Baied’Andrahoman a.’’NeitherofMarguin’smapsis reproducedinGrandidier(1885). 1885AlfredGrandidier(1885)brieflydescribestheexistenceofcaverns onthesoutherncoastofMadagascarinhis ‘‘Histoiredelage ographie.’’In1892thisbookwasrevised,reprintedandissuedasthefir stvolumeofhis monumental‘‘HistoirePhysique,NaturelleetPolitiquedeMadagascar.’ ’Sometimescreditedwithhaving discoveredthecaves(e.g.,seeDecary,1927),AlfredGrandidieractuall ylearnedofthemfromtheMalagasyand throughtheexperienceofotherwesternerswhohadtraveledtoMadagascar insomecaseswellbefore Grandidier’sfirsttripthere(in1865).Grandidier(1885)statesthatCa pAndavakawasknownto16 th century EuropeanexplorersofMadagascar. 1899Austrianpipe-carverandnaturalist,Franz(orFranc ois)Sikora,andateamofsometwentyMalagasyworkers, excavateAndrahomanaCaveforaperiodof18days.Specimensofextinctlem ursandassociatedfaunacollected duringthisexpeditionweresenttotheNaturhistorischesMuseuminVienn aandtotheBritishMuseumof NaturalHistoryinLondon.Sikoraapparentlyleftnowrittenrecordsdocu mentingtheexactlocationofhisdig withinthecave,nordidhemakealogofstratigraphicassociations.Hismo stdetailedtreatmentofhis explorations,‘‘SeptAnsa `Madagascar,’’waspublishedtwoyearspriortohisexpeditiontoAndraho mana (Sikora,1897).However,intermsofthesheervolumeofspecimensbelongi ngtosubfossillemurs,Sikora’s osteologicalcollectionfromAndrahomanaremainsunsurpassedbythoseo fsubsequentexplorers.Among Sikora’sdiscoverieswerethefirstcompleteskullandhemimandible,plu sassociatedpostcranialbones,of Archaeolemurmajori (senttotheNaturalHistoryMuseum,London)aswellasanearlycompletesk ullandother skeletalelementsof Hadropithecusstenognathus (senttotheViennaNaturhistorischesMuseum).Sikora’s expeditiontoAndrahomanawasbrieflydocumentedbyAlluaud(1900).Thes pecimensthatSikoracollected therewerestudiedinitiallybyC.I.ForsythMajor(1899)andbyLorenzvon Liburnau(1899,1901,1902,1905), andlaterbyWalker(1967),Tattersall(1973,1982),Jungers(1976),andG odfrey(1977).Sikoraalsocollected over40specimensofanextinctspeciesof Hypogeomys (seebelow),nowattheBritishMuseum(NaturalHistory). 1900FrenchentomologistCharlesAlluaudconductsthefirstofaseriesof expeditionstoAndrahomanaandother localitiesinsouthernMadagascar.Onlyhisfirstexpedition(August21– August27,1900)iswelldocumentedin theliterature(Alluaud1900,aletterdatedAugust30,1900andaddressed toGuillaumeGrandidier).Onthis expedition,AlluaudwasaccompaniedbyaFrenchsoldier,LieutenantGaub ert.Thetwotraveledtogetherfrom thePostedeManambarotoRanopiso,andthentotherat-infested,recently constructed,Posted’Andrahomana, locatedonahilloverthesmallpeninsulaborderingtheBaied’Andrahoman a.Alluaud’s(1900)letterto Grandidierdescribesthedifficultythetwoexperiencedingainingacces stothecaveaswellasvariousother frustrations.Hecomplainsthatthecavehadbeencompletelyexcavated( toutfouille )byhispredecessor,Franz Sikora.Itisnoteworthy,however,thatAlluaudandGaubertspentonlythr eedays(August23–25)actually excavatinginornearthecave,andtheyhadhiredMalagasyhelpersononlyt woofthem.Theteamfound relativelyfewsubfossilspecimens(althoughthereweresomegianttorto iseandextinctlemurbonesinthelot). PartofAlluaud’scontributionwastheconstructionofapathforeasierac cesstothecave,andabotanicalsurvey. ThebountyofthisfirstexpeditionwassenttotheMuse umNationald’HistoireNaturelleinParisin1900. 1901–02Onseparateoccasions,CharlesAlluaud,LieutenantGaubertandG uillaumeGrandidier(thesonofAlfred)visitor revisitAndrahomana,tocollectmorespecimensofextinctlemurs.Thesee xpeditionswereapparentlymore successfulthanAlluaudandGaubert’sfirstone,buttheyarepoorlydescr ibedintheliterature.SeparateAlluaud, Gaubert,andG.GrandidiercollectionsweresenttoParisinthespringof1 902andweredescribedbyG. Grandidier(1902).No Hadropithecusstenognathus waslistedashavingbeencollectedbyanyoftheseexplorers. However,some Hadropithecus bonesthatmayhavecomefromAndrahomanaareinthecollectionsofthePari s museum(Godfreyetal.,2006a).Prizespecimensofthesemissionswerethe types(MNHNMAD1646and MNHNMAD1647)ofanewspeciesofextinctrodent, Hypogeomysaustralis ,foundbyAlluaudandinitially describedbyGrandidier(1903)(GoodmanandRakotondravony,1996). 1902FranzSikoradiesattheageof39,fewerthanthreeyearsafterhehadle dhispioneeringexpeditionto Andrahomana(Zapfe,1971). D.A.B URNEY ,N.V ASEY ,L.R.G ODFREY ,R AMILISONINA ,W.L.J UNGERS ,M.R AMAROLAHY AND L.R AHARIVONY JournalofCaveandKarstStudies, April2008 N 23

PAGE 12

1906FrenchpharmacistandnaturalistMartinFranc oisGeaycollectsbonesoflemurs,rodents,andbatsatthecaverns ofAndrahomana.GeayhadstudiednaturalhistoryinFranceunderthetutel ageofAlphonseMilne-Edwards.He spentseveralyearsinMadagascar(1904–1907)collectingplants,wood,f ossils,mollusks,fish,birds,reptiles,and rocksfortheParisnaturalhistorymuseum(Thevenin,1907;Dorr,1997).D uringthattime,hesent70crates containingabout14,000specimenstoParis.Inhisprivatelypublishedbo ok,Geay(1908,p.99–100)briefly describeshistriptoAndrahomana,statingonlythathecollectedlemur,r odent,andbatbonesthere,and describingtheterrestrialdepositsintowhichthecavewascut.Geayalso offersthefollowingexplanationforthe nameofthebay,Andrahomana(literally,‘‘an’’ atorwhere,‘‘rano’’ water,‘‘homana’’ eaten:‘‘wherethe waterisdevoured’’).AttheendofAndrahomanaBayisameanderingbrookth atwindsaroundtall,gloomy,and entirelydenudedrocksborderingtheocean.Thewaterofthebrookdisappe arsinto(i.e.,iseatenorconsumedby) thesandsofthebeach. 1906?Gaston-JulesDecorse(aFrenchmilitarydoctorwhocollectedplant s,insectsandfossilsforthenaturalhistory museuminParis)explorestheregionofCapAndavaka(Decary,1927,forthe year1926).Decorsewasvery interestedinterrestrialmollusksandhecollectedspecimensinthecalc areousdunesnearAndrahomana. Thevenin(1907)describesseveralexplorers,includingGrandidierandD ecorse,collecting Aepyornis eggshellin theregion. 1926FrenchmilitaryofficerandnaturalistRaymondDecaryleadsascient ificexploratoryexpeditiontosouthern Madagascar(includingAndrahomana)fortheMuseumNationald’HistoireN aturelle(Paris).Thismissionbegan inJune,1926.Decary(1927)foundlittlehereinthewayofgiantsubfossil s,buthedidfindnumerousbonesof micromammalsandothersmallvertebrates.Healsopublishedasurveyofth eplantsandgeologyoftheregion (Decary,1928). 1927DecarysendssomeofthespecimensthathecollectedatAndrahomanato Grandidierforstudy.Overthefollowing 12years,bothGrandidierandDecarydispatchspecimensfromAndrahomana totheMuseumofComparative Zoology,HarvardUniversity.Theseincludesome Lemurcatta Microcebus sp.(probably M.griseorufus ),and someinsectivores(nowrecognizedas Microgaleprincipula and Geogaleaurita ). 1928GuillaumeGrandidier(1928)namesseveralnewspeciesofinsectivor esonthebasisofspecimensfrom AndrahomanaCave.Hegavethename Paramicrogaledecaryi tospecimenswitharelativelyshortandwideskull, and Cryptogaleaustralis tootherswithalongandnarrowskull. Paramicrogaledecaryi waslatersynonymized with Microgale byHeimdeBalsac(1972)andthenwith M.principula byMacPhee(1987). Cryptogaleaustralis waslatersynonymizedwith Geogaleaurita byHeimdeBalsac(1972). 1964Rene Battistini(1964)publisheshisdissertationmonographonthegeomorph ologyofsouthernMadagascar,with notesonthegeologyofAndrahomana. 1966BushpilotIkeRussellandhiswifeJeanaccompanyAlanWalkerandPaul MartinonaflightfrommainlandAfrica toAntananarivo(Walker,2002;PaulMartinunpublishednotes),andthent oanumberofsubfossilsites, includingAndrahomana.ToreachAndrahomana,Russelllandedhissmallai rcraftonthedrybedofLake Erombo,severalmilesfromthecave.Fromthere,WalkerandMartinhikedto thecavewhiletheRussells remainedwiththeirplane.Onceinthecave,WalkerandMartinfoundapiece ofthescapulaandafingerboneof Megaladapisedwardsi ,aswellaspiecesofasmashedcarapaceandplastronofagianttortoisetha thadfalleninto thecavefromabove.Theycouldonlyvisitthecaveforashortperiodbefore havingtoreturnfirstbyfootto EromboandthenbyplanetoFortDauphin. 1971RaymondDecaryandAndre Kiener(1971)publishaninventoryofcavesofMadagascar. 1983RossMacPhee,ElwynSimons,PrithijitChatrath,NeilWells,andMart ineVuillaume-Randriamanantenavisitthe caveAugust18forafewhoursinconnectionwithaDukeUniversityPrimateC enterexpeditiontosouthern Madagascar.Althoughimpressedwiththebeautyofthearea,thegroupfoun donly‘‘veryrecentbonesof Microcebus Propithecus Rattus ,birds,andturtles.Afewhand/footbonesoflargesizewerediscovered…t hey couldbethoseofagiantlemur’’(R.MacPhee,pers.comm.,2007). 2000DavidBurney,Malagasystudents,andlocalguidesmakeareconnaissa nceofthecaveAugust24–27,conductinga smallcontrolledexcavationatAHA-Bandmakingsurfacecollectionsands edimentborings. 2003July16toAugust8,BurneymakesexpeditiontoAndrahomanaaccompani edbythecoauthorsofthispaperand localguides.Groupundertakesmapping,andexcavatessitesAHA-FandAHA -I. 2006LaurieGodfreyandcolleaguesrevisepostcranialattributionsfor Hadropithecusstenognathus onthebasisof associatedskeletalmaterialsfoundatAHA-Iin2003(Godfreyetal.,2006 a).CollaborationswithSteve Goodmanleadtonamingasubfossilspeciesofshrew-tenrec( Microgalemacpheei ;Goodman,Vasey,andBurney, 2007)andpublishingthefirstknownfossiloccurrenceoftherecentlydes cribednesomyidrodent Macrotarsomys petteri (Goodmanetal.,2006). Appendix1.Continued. N EWFINDINGSAT A NDRAHOMANA C AVE ,S OUTHEASTERN M ADAGASCAR 24 N JournalofCaveandKarstStudies, April2008



PAGE 1

EDITORIAL JournalofCaveandKarstStudies Listinginthe JournalofCitationReport :What DoesitMean? M ALCOLM S.F IELD U.S.EnvironmentalProtectionAgency,NationalCenterforEnvironmenta lAssessment(8623P),1200PennsylvaniaAve.,N.W.,Washington,D.C.204 60, USA,tel:(703)347-8601,fax:(703)347-8692,email:field.malcolm@epa .gov T HOMSON ISIL ISTING Severalyearsagothe JournalofCaveandKarstStudies achievedalistinginthedistinguishedThomsonInstitute forScientificInformation(ISI)whichincludedScience CitationIndexExpanded,ISIAlertingServices,and CurrentContentsPhysical,Chemical,andEarthSciences. AsignificantaspectofanISIlistingisameasureofthe influencealistedjournalhasonscientificresearch.Thisis wheretheThomsonISI JournalCitationReports (JCR) becomesimportant. J OURNAL C ITATION R EPORTS TheJCRcurrentlycoversmorethan7,500ofthe worldÂ’smosthighlycited,peer-reviewedjournalsin approximately200disciplines;theScienceEditioncovers over5,900leadinginternationalsciencejournalsfromthe ThomsonScienceCitationIndex(SCI)database.Itisthe recognizedauthorityforevaluatingjournals(Thomson, 2008). AfterajournalisacceptedforlistingbyThomsonISI andhasbeeninthesystemforaminimumofthreeyears, thejournalbecomeseligibleforanadditionallistinginthe JCR,whichisamultidisciplinaryjournalevaluationtool andistheonlyevaluationresourcethatprovidesstatistical informationbasedoncitationdata.Bycompilingcited references,ameasureofresearchinfluenceandimpactat thejournallevelisobtainedandillustratestherelationships betweencitingandcitedjournals(Thomson,2005a). TheimportanceofajournalÂ’sinfluenceonscientific researchisevidentbytheimportanceuniversityadministratorsandfederalresearchinstitutionshaveplacedonSCI data,whicharenowbeingusedasameasureoffaculty productivityandfacultyproductivityquality.Itisno longerenoughtopublishinrefereedjournals,nowthey mustbeSCIjournalsaswell.ToavoidU.S.karstwork goingtoforeignkarstjournalsthatarelistedintheSCI,it iscriticalthatthe Journal continuetobepublished regularlyandvoluminouslywithhigh-qualitypeer-reviewedarticles. C ITATION S TATISTICS TheJCRliststhetotalcites,whichrepresentsthe numberoftimesthatthe Journal hasbeencitedbyall journalsincludedintheproductdatabasewithinthe currentproductyearandthetotalnumberofarticles publishedinajournalinthecurrentproductyear.Italso providesfourbasicstatisticstomeasureaparticular journalÂ’sinfluence.Thesestatisticsareimpactfactor, immediacyindex,citedhalf-life,andcitinghalf-lifeand areexplainedbelow(Thomson,2005a,b). I MPACT F ACTOR Theimpactfactorisameasureofthefrequencythatan articleinthe Journal iscitedinaparticularyear.Itis calculatedbydividingthenumberofcurrentcitationsto itemspublishedinthetwopreviousyearsbythetotal numbersofarticlesandreviewspublishedinthetwo previousyears. I MMEDIACY I NDEX Theimmediacyindexisameasureofhowquicklythe averagearticleinthe Journal iscited.Ittellsauserhow oftenarticlesinajournalarecitedwithinthesameyear.It iscalculatedbydividingthenumberofcitationstoarticles publishedinagivenyearbythenumberofarticles publishedinthatyear. C ITED H ALF -L IFEAND C ITING H ALF -L IFE Thecitedhalf-liferepresentsthemedianageof Journal articlescitedinthecurrentJCRyear.Itisusefulfor evaluatingtheagerangeofthearticlesfromthejournal andmaybeusedformakingarchivingandretention decisions.Thecitinghalf-liferepresentsthemedianageof thearticlescitedbythe Journal inthecurrentJCRyear.It isusefulforevaluatingtheageofthemajorityofarticles referencedbyajournal. J OURNALOF C AVEAND K ARST S TUDIES S TATISTICS Sohowdoesthe JournalofCaveandKarstStudies measureup?Figure1showstheJCRstatisticsfor2005and 2006,theonlytwoyearsavailableasofthiswriting.The Journal impactfactorincreasedfrom0.357in2005to0.576 in2006suggestingthattheimpactthatarticleshaveon scientificresearchincreasedin2006.Unfortunately,the Journal immediacyindexdecreasedfrom0.235in2005to 0.000in2006suggestingarticlespublishedin2006werenot beingcitedveryquickly.However,amorerealistic explanationmaybethesmallnumberofarticlesactually publishedinthatyear.Thisstatisticissomewhatmisleadingbecauseitisdependentonthenumberofarticles publishedin2006,whichtheJCRaccuratelylistsasnine E DITORIAL JournalofCaveandKarstStudies, April2008 N 1

PAGE 2

becausetheDecember2006issuecameoutin2007.Had theDecember2006issuebeenincludedinthecounting,a totalof16articleswouldhavebeenlisted.Inaddition, becausethe Journal hasonlybeenincludedintheThomson ISIdatabaseforashorttime,categoriessuchasthe immediacyindexaregoingtolagforawhile. The Journal citedhalf-lifeincreasedfromnodatain 2005to6.3yrin2006whichsuggeststhattherewere nopublishedarticlesinthe Journal citedin2005but somecitedworksin2006.The Journal citinghalf-life decreasedfrom 10.0yrinto2005to9.2yrin2006 suggestingthatthenumberofarticlescitedinthe Journal decreasedfrom2005to2006.Bycomparison,theonly otherkarstjournalintheThomsonISIdatabasehadno listingfor2006. WhentheJCRstatisticsforthe JournalofCaveand KarstStudies arecomparedwithotherjournals(e.g., Science )itdoesnÂ’tappeartorateverywell.However,other journalshavebeenincludedinJCRforamuchlongertime thanthe JournalofCaveandKarstStudies andarenot nearlysospecialized.Theextremelyparochialnatureofthe JournalofCaveandKarstStudies limitsthenumberof papersthatwillgetsubmittedforpublicationwhichthen limitsthenumberofpapersthatgetcited. Overall,the Journal isonanupswingasevidencedby theimproved2006 Journal impactfactoroverthe2005 Journal impactfactor 1 .Asthe Journal continuestoexpand becausemoreandmorepapersarebeingsubmitted,itis veryprobablethatthe Journal statisticslistedintheJCR willalsocontinuetoincrease.Forexample,for2007itis expectedthat 20articlestobelistedintheJCRbecauseit includestheDecember2006issue,butnottheDecember 2007issueduetothelatenessassociatedwiththesetwo issues 2 .While30 publishedarticlesperyearinthe Journal havenotbeencommoninthepast,itisverylikelyto becomecommoninthenottoodistantfuture.This increaseinpublishedarticlesinthe Journal willalmost definitelyleadtohigherratingsforthe Journal R EFERENCES Thomson,2005a,JournalCitationReportsontheWebv.4.0:Thomson ISI,URLhttp://scientific.thomson.com/media/scpdf/jcr4_sem_0305. pdf,[accessedJanuary10,2008]. Thomson,2005b,JournalCitationReportsTutorial(v.4.0):Thomson ISI,URLhttp://scientific.thomson.com/tutorials/jcr4/jcr4tut6.ht ml, [accessedJanuary14,2008]. Thomson,2008,JournalCitationReportsontheWebv.2.0:Thomson ISI,URLhttp://scientific.thomson.com/products/jcr/,[accessedJan uary10,2008]. Figure1. JournalCitationReports statisticsforthe JournalofCaveandKarstStudies for2005and2006. 1 TheJCRreportreleasedjustpriortopressofthisissuereportsanimpactf actorfor the JournalofCaveandKarstStudies 1.000. 2 Notethatweexpecttobereturningtoourpublishingscheduleshortly. E DITORIAL 2 N JournalofCaveandKarstStudies, April2008



PAGE 1

VARIABLECALCITEDEPOSITIONRATESASPROXYFOR PALEO-PRECIPITATIONDETERMINATIONASDERIVED FROMSPELEOTHEMSINCENTRALFLORIDA,U.S.A. P HILIP E.V AN B EYNEN *,1 ,L IMARIS S OTO 2 AND J ASON P OLK 1 Abstract: Depositionratesderivedfromspeleothemshavebeenshowntobeauseful paleoclimaticproxy.Paststudieshaveshownthatthemostcommonclimati cparameter measuredbyvariabledepositionratesisprecipitation,whereincreased precipitationleads toincreasedcalcitedeposition.Thiswasthepremiseofourstudy,wheret hreeFloridian stalagmites’depositionratesweremeasuredandcomparedtopaleohydrol ogicindicators takenfromthesampleorfromotherregionalrecords.Depositionrateswer emeasuredby determiningthevolumeofcalciteprecipitatedbetweenTIMSU-seriesdat es(mm 3 yr 1 ), therebyaccountingformorphologicalchangesonthestalagmiteoveritsd epositional history.Mostpriorresearchreliedonasimplelinearinterpolationbetw eenknownagesto calculaterate(mmyr 1 ).Resultsshowthreedistinctperiodsofincreaseddepositionfor ourstalagmitescenteredon2.0,1.25and0.5kaBP.AcomparisonwithMg/Ca andSr/Ca ratiosandcalcitedepositiontentativelyshowselevatedelementalrati osduringthethree aforementionedperiods.Elevatedtraceelementratioshavebeenshownto becorrelated withincreasedresidencetimeofpercolationwatersintheoverlyingbedr ockabovecaves andconsequentlydecreasedrainfall.Tocorroboratethisfinding,paleo -precipitation recordsfromLittleSaltSpring,FloridaandLakeMiragoane,Haiti,weree xaminedfor coevalaridperiodswithourstalagmites.Bothrecordsdopossesssimilar dryperiodsand provideaddedsupportthattheregionexperiencedperiodsofabruptaridi tyoverthelast twomillennia.ThecombinedeffectofachangeinthemeanpositionoftheIn tertropical ConvergenceZoneandtheeasterlywindsassociatedwiththeNorthAtlanti cHighappear tobethemajorcausesforthesetimesofaridity. I NTRODUCTION Thedepositionratesofspeleothemscanprovide informationaboutpaleoclimaticvariations,includingprecipitation,temperature,andsoilactivity,aboveacave (KaufmannandDreybrodt,2004;White,2004).Therate ofspeleothemdepositionisdeterminedbysoilcarbon dioxideconcentration,driprate,andtemperature,allof whichareclimaterelatedandcanaffecttherateandshape ofdeposition(Dreybrodt,1999).Hiatusesinaspeleothem canindicateperiodsofclimatechangethatincludeglacial periods,droughts,andchangesinthehydrologicpatternof waterflow.Fasterdepositionratescanbeindicativeof awarmer,wetterclimateinthearea,whereasslower depositionratescanresultfromcooler,drierconditions abovethecave(Hennigetal.,1983;Musgroveetal.,2001; vanBeynenetal.,2004). Thedepositionrateofspeleothemsisafunctionof supersaturation( P CO 2 )ofpercolationwatersanddriprate (rainfallamount)(Bakeretal.,1998).Thedepositionrate ofstalagmitescanthenbeusedasapaleo-environmental proxyforsurfaceprecipitation(Bakeretal.,1993;Genty andQuinif,1996;Holmgrenetal.,1999;Qinetal.,1999). Thedeterminationofspeleothemdepositionrateshas evolvedtoahighlevelofaccuracyduetoradiogenicdating methods,suchasU-seriesTIMSmassspectrometry(Liet al.,1989;Doraleetal.,1992;Gascoyne,1992;Musgroveet al.,2001).ThroughtheuseofaccurateU/Thdates combinedwithspeleothemlengths,depositionratescan becalculatedwithhighprecision,alsoprovidingevidence ofhiatusesanddepositionratevariability(Dreybrodt, 1999). Thepurposeofthisstudyistodevelopandimprovethe understandingofthepaleoclimateofFlorida,with emphasisonwhetherspeleothemscanrecordhydrologic responsestochangingatmosphericcirculationpatterns thataffecttheFloridaPeninsula.Theapproachtakenin thisstudywastomeasureshiftsinspeleothemdeposition rateanddeterminethepaleoclimaticsignificanceofthose shifts.Hence,wetestedthefollowinghypothesis:Increased depositionratesderivedfromspeleothemsareindicativeof wetterconditionsabovethecave. Totestsuchahypothesisrequiresanaccurate chronologyofthesechangesincalciteprecipitationrate. Uranium-seriesdisequilibriadatingisthemethodusedhere toachievethisgoal.Manystudieshavedemonstratedthat thespeleothemsprovidereliablechronologiesusingtheU*Correspondingauthor. 1 DepartmentofEnvironmentalScienceandPolicy,UniversityofSouthFlor ida, 4202EFowlerAve,NES107,Tampa,FL33620,vanbeyne@cas.usf.edu 2 DepartmentofGeology,UniversityofSouthFlorida,4202EFowlerAve,Tam pa, FL33620 P.E.VanBeynen,L.Soto,andJ.Polk–Variablecalcitedepositionratesas proxyforpaleo-precipitationdeterminationasderivedfrom speleothemsinCentralFlorida,U.S.A. JournalofCaveandKarstStudies, v.70,no.1,p.25–34. JournalofCaveandKarstStudies, April2008 N 25

PAGE 2

seriesmethod(Harmonetal.,1978;Doraleetal.,1992; Gascoyne,1992;FrumkinandStein,2004;Lachneitetal., 2004;Polyaketal.,2004).Tofullyaddressthehypothesis, thevolumeofcalciteprecipitatedbetweenthesedatesfor multiplestalagmiteswilldeterminethevariabledeposition rateforeachspeleothem. S TUDY A REA Thestudyareaconsistsoftwocaves,BrownRatCave (BRC),inHernandoCounty,andBriarsCaveinMarion County,bothinFlorida(Fig.1).Theyarelocatedinthe BrooksvilleRidgesectionoftheOcalaArchthatcontains sinkholes,drykarstvalleys,andinterfluvialhills(Reeder andBrinkmann,1998).Thegeologyintheregionis dominatedbyaseriesofcarbonatesthatincludethe EoceneOcala(locationofbothcaves)andOligocene SuwanneeLimestones,whichareunconformablyoverlain bytheHawthornGroupcharacterizedbycarbonates interspersedwithsiliclasticsandphosphoriteredeposition formingadmixtureswithdolomite,quartzsand,andMgrichclays(Scott,1997).Approximatelytwometersof Pleistocene-agedquartzsandsoverlietheseunitsinmost partsoftheregion,althoughtheydrainrapidly. Withalengthof 1km,BRCinBrooksvilleRidgeis oneofthelongestdrycavesinFlorida,andhasasingle, man-madeentrance.Consequently,relativehumiditylevels inthecavewouldhavebeenat 100%untiltheentrance wascreated.ThecavedevelopedwithintheOcala Limestone,andthevegetationabovethecaveischaracterizedashardwoodhammockspopulatedbyoak,hickory andmaples(Armstrongetal.,2003).Theaverage temperatureoftheBrooksvilleRidgeareais21.3 u C,and theaveragetotalprecipitationis1356mmperyear (SoutheastRegionalClimateCenter).Thestalagmite BRC03-02wascollectedfromthiscave,whichdoesnot floodduringhurricanes(asinHurricanesJeanneand Francesof2004)orthunderstorms. BriarsCaveislocatedonthesouthernoutskirtsof Ocala.Thecaveunderliesalowhillbetweentwosinkholes. BriarsCavetrendsNE-SWandconsistsofadryupper levelandapartiallyfloodedlowerpassage(Floreaetal., 2003).Thecaveisapproximately1kmlong,makingitone ofthelongestcavesinthestateofFlorida,withonlyone smalltightentrance.Thevegetationinthevicinityofthe caveischaracterizedbyhardwoodspecies,includingoak. TheaveragetemperatureofMarionCountyis22 u C,and theaverageannualprecipitationis1330mm(Southeast RegionalClimateCenter).Twostalagmitesamples, BRIARS04-01andBRIARS04-02,werecollectedfrom thesameareaofthecave.AHendytest(Hendy,1971), undertakenforthelatterofthetwospeleothemsto determinewhethertheisotopesweredepositedinequilibriumwiththeambientwaters,showedrelativehumidity levelsinthecavewouldhaveremainedcloseto 100%. M ETHODS S AMPLE C OLLECTION Uponcollection,thethreestalagmiteswerecutverticallyalongtheirdepositionalaxes,andpolishedtoclarify thepositioningofthelamina.Calcitesamplesof250mg werecollectedforU-seriesdatingfromallspeleothems usingadental-bitequippedDremel H tool.Sampleswere drilledatapproximately20–30mmintervalsalongthe growthaxisofthestalagmite(Fig.2).Exactlyequal Figure1.Locationofstudysites,BriarsCaveandBrown RatCave,inrelationtooneanother. Figure2.Photographsofthethreespeleothemsusedin thisstudy. V ARIABLECALCITEDEPOSITIONRATESASPROXYFORPALEO-PRECIPITATIONDETER MINATIONASDERIVEDFROMSPELEOTHEMSIN C ENTRAL F LORIDA U.S.A 26 N JournalofCaveandKarstStudies, April2008

PAGE 3

incrementscouldnotbeattainedininstanceswherethe potentialsamplingareapossessedasmallvoid,or adramaticcolorchangeorpossibledirtycalcite.Wetried tosamplefromclear,homogenouscalcite,whichcouldnot beachievedatapredetermineddistancefromthe stalagmitebase. USERIES D ATING SuccessfuldatingofspeleothemsbyU/Thmethods dependsonsatisfyingthreecriteria:1)thespecimenmust containameasurableamountofuranium( 0.02ppm)and thorium;2)thespecimenmustcontainnegligibleinitial 230 Th.Thiscanbemeasuredthroughthepresenceofthe isotope 232 Thbecauseitwillbepresentinanythoriumbearingmineralsthatmayoccurintheinsolubledetritus, butisnotpartofthedecaychain;and3)thespecimenmust nothaveundergonedissolutionandre-crystallizationthat wouldaltertheinitialU/Thisotopiccomposition(Latham andSchwarcz,1992;White,2004). ThesampleswereanalyzedattheRadiogenicLaboratoryattheUniversityofNewMexico.UandThare measuredonaMicromassSector54thermalionization massspectrometerwithahigh-abundancesensitivityfilter (Lachnietetal.,2004).Allisotopesofinterestwere measuredonanion-countingDalymultiplierwithabundancesensitivityintherangeof20ppbatonemassunitin themassrangeofUandTh,requiringverylittle backgroundcorrection,evenforsampleswithlarge 232 Th content.Uisotopicstandards,suchasNBL-112,were measuredalongwithsamples.Typicalanalyticaluncertaintiesareintherangeof0.2%forUisotopecomposition, similarorsomewhatlowerprecisionforTh,dependingon theageandsizeofthesamplemeasured.Adetailed accountofthetechniquecanbefoundinPolyakand Asmerom(2001). D EPOSITION R ATES Whencalculatingdepositionrates,itisimportantto considerthatthespeleothemmorphologyisnotaconstant linearrelation(Franke,1965;Curl,1973;Gams,1981; Baldini,2001).Becausetheshapeofthestalagmitecanvary significantlyoveritsdepositionalhistory,wecalculatedthe depositionratesasavolumeandnotsimplylineardistance ofcalcitedepositionbetweenknownages.Thisapproachof measuringspeleothemdepositionrateissimilaronetaken byBaldini(2001).Halfofthestalagmitewasphotocopiedto haveaone-dimensionalpictureofthesample.Deposition linesweredrawontothecopytomarkthedifferentvisible layersbetweenthespeleothem.Becausethestalagmiteswere alreadydated,eachknowndatewasmarkedonthe photocopy.Withthedatesandtheexactlocationofthem onthespeleothem,aquantitativeapproachwasapplied.We usedtheformulaofthefrustumofacone V 1 3 p h ( r 2 rR R 2 ) 1 tocalculatethevolume(mm 3 )ofcalcitedepositedbetween datedhorizons.Figure3showshowthecross-sectionofthe frustumofaconeisfittedtoaportionofBRIARS04-02 betweenknowndates.Onoccasion,toreplicatethe depositionbetweentwodates,severaldifferentsized frustumshadtobeused.Tocalculatethedepositionrate (calcitedeposition),thevolumeofcalcitewasdividedbythe timeofdeposition(mm 3 yr 1 ). R ESULTS USERIES D ATING TheentiresuiteofTIMSU-seriesdatesforallthree speleothemsusedinthisstudyispresentedinTable1.All agesarereportedasyearsbeforepresent.Thespeleothems meetallthreecriteriamentionedabove.Theuraniumand thoriumlevelsaresufficientfordating.AllTh/Uagesare incorrectstratigraphicorder(Fig.4).Thisresultsuggests thatallthespeleothemsremainedclosedsystems(i.e.,no dissolutionorrecrystallization)fortheirentiredepositional histories;hence,nouraniummigrationoccurredwithinany speleothem.The 230 Th/ 232 Thratios(Table1)areindicative ofverylittledetritalThbeingpresentinanyofthesamples. Cleancalcite 230 Th/ 232 Thratiosaregenerallygreaterthan 50,andthevaluesforBRCandBCsamplesarewellabove thatthresholdvalue,somevaluesapproaching1,136. Depositionaltrendsforeachspeleothemshowalinear relationshipforthemajorityofthegrowthperiod(Fig.4). Onlyduringthefirst30mmdideachspeleothemdeviate Figure3.Diagramshowinghowthefrustumofaconeis fittedtothegrowthlayersforparttheupperportion ofBRIARS04-02. P.E.V AN B EYNEN ,L.S OTO AND J.P OLK JournalofCaveandKarstStudies, April2008 N 27

PAGE 4

fromthislinearity.Suchdeviationrepresentsaslower depositionrate,whichmaybecausedbytheavailable calcitebeingspreadoveralargerarea.Forthecarrottype stalagmitesusedinthisstudy,suchaninitialslower depositionrateiscommon. D EPOSITION R ATES ThedepositionratescalculatedusingtheU-seriesdates arepresentedinTable2.Eightdatesinstratigraphicorder wereusedtodevelopdepositionratesforsampleBRC0302(Fig.5a).Thisstalagmitegrewfromaround2.6to 0.35kaBP.Theaveragedepositionrate(calcitedeposition)forthissamplewas375mm 3 yr 1 .Intheearlystage ofaccumulation,thedepositionratewas63mm 3 yr 1 averylowcalciteaccumulationforthespeleothem.A significantchangeoccurredfrom1.8to1.3kaBPwhenthe rateincreasedwithadepositionrateof337mm 3 yr 1 .A continuedincreaseindepositionrateoccurredfrom1.3to 1.2kaBP,whencalciteaccumulatedat326mm 3 yr 1 (Fig.5a).Ataround0.5kaBP,thedepositionrateincreasedconsiderably,beingcloseto1164mm 3 yr 1 ,the highestratemeasuredinanyspeleotheminthisstudy. SpeleothemscollectedfromBriarsCavewerealso analyzedfordepositionrates.ForBRIARS04-01seven datesinstratigraphicorderestablishthatthespeleothem grewfrom3.6to1.2kaBP(Table2).Theaveragerateof depositionforthespeleothemwas224mm 3 yr 1 (Fig.5a). Thelowestratewasfrom 3.6to2.5kaBP,attaining avalueof40mm 3 yr 1 .Thestalagmitedepositionfrom 2.1to1.9kaBPincreasedconsiderably,averaging 385mm 3 yr 1 ,thehighestrateforthisspeleothem, althoughitwasfollowedbyanotherrapidperiodofcalcite depositionfrom1.9to1.8kaBP,witharateof 366mm 3 yr 1 .Adecreaseindepositionrateoccurred after1.8to1.2kaBPwherethevaluesdiminishedto nearlyhalfofthepreviousones. StalagmiteBRIARS04-02hastendatesinstratigraphic ordershowingthatithasgrowncontinuouslyfrom 4.5kaBPuntilthepresent(Table2).However,the youngestdateof32years 6 198hassuchahigherror comparedtotheactualdate(Fig.5a)thatitwillnotbe includedinanyfurtheranalysis.Theaveragerateof depositionforthissamplewas542mm 3 yr 1 ,thehighest ofallthespeleothems.Aslowdepositionrateof 51mm 3 yr 1 wasrecordedatthebaseofthestalagmite from4.1to2.2kaBP.However,withonlytwodates delineatingthis2,000yearperiod,itisnotpossibleto identifyanyperiodsofmorerapidcalcitedepositionthat mayhaveoccurredduringspeleothemgrowth.BRIARS0402containsafewpossiblehiatusesduringthisperiod, whichwouldaccountforthelowdepositionrates.Three distincthorizonsarepresentwithinthiszonethatarenot foundinanyotherintervalsforthisortheother speleothems.Onlywithmorefrequentdatingcouldthe timingofthesepotentialhiatusesberesolved.Atapproximately2.2to1.8kaBP,asignificantincreaseindepositionratewasrecorded;thisisconsistentwithan increaseddepositionraterecordedatthesametimefor BRIARS04-01.Theratewasthehighestfrom1.35to 1.3kaBP,attainingavalueof972mm 3 yr 1 .Another majorchangewasrecordedfrom0.6to0.5kaBP,with adepositionrateof915mm 3 yr 1 :thisincreasein depositionratewasalsorecordedinthespeleothem BRC03-02. Asaforementioned,mostspeleothemstudiesexamining depositionratescalculatetherateinmmyr 1 ,asinthe numberofmmcalcitedepositedbetweeneachdate.To showhowthistechniquemaygiveanunrealisticimpressionofhowquicklyaspeleothemwasdeposited,Figure5b allowsacomparisonbetweenthevolumetric(mm 3 yr 1 ) andlinear(mmyr 1 )methods.Itisreadilyapparentthat thelinearmethodtendstoexaggeratedepositionrateas seeninBRC03-02at 0.5kaBPandforBRIARS04-02at 1.3kaBP.Thevolumetrictechniqueaccentuatesepisodesofaugmenteddepositionbutalsoperiodsthatare notapparentinthelinearmethod.Forexample,BRC03-02 hasanoticeableincreaseat 1.3kaBPthatcannotbe seeninFigure5a. D ISCUSSION Ourinitialhypothesisstatedthatincreaseddeposition rateswereindicativeofwetterconditionsabovethecave. Thishypothesisisderivedfrompreviouspaleoclimate studiesusingspeleothemsthatshowedtherateofdepositioniscontrolledbytheamountofprecipitationfalling abovethecave,suchthatanincreaseinprecipitationleads tomorespeleothemgrowth(Hennigetal.,1983;Lauritzen andLundberg,1999).However,notallstudiessupportthis conclusion,withsomeshowingthatdrierconditionscan leadorhaveledtoanincreaseindepositionrate (Dennistonetal.,1999;Fairchildetal.,2000).Whileour hypothesisproposesthefirstscenario,wetesteditby correlatingthedepositionratesoutlinedabovewithtrace Figure4.Agevs.depthplotfortheFloridianspeleothems. Solidlinesshowthelineartrendforeachspeleothem. V ARIABLECALCITEDEPOSITIONRATESASPROXYFORPALEO-PRECIPITATIONDETER MINATIONASDERIVEDFROMSPELEOTHEMSIN C ENTRAL F LORIDA U.S.A 28 N JournalofCaveandKarstStudies, April2008

PAGE 5

Table1.U-seriesdatesforFloridianspeleothems.Allagesassuminga 230 Th/ 232 Thinitialratioof10 6 2ppm. Sample Distance frombase (mm) U(ppm) Error 3 10 3 Th(ppm) 3 10 3 Error 3 10 5230 Th/ 232 Th 234 U/ 238 U Error 3 10 3 230 Th/ 238 U Error 3 10 3 Uncorrected Age(yrB.P.) Corrected Age(yrB.P.) Briars04-011301.031 6 0.68.4 6 0.1233.8 6 3.91.008 6 0.50.012 6 0.11,263 6 381,210 6 40 1101.249 6 1.02.5 6 0.21136.1 6 102.21.010 6 0.50.014 6 0.11,495 6 501,482 6 50 951.201 6 0.75.5 6 0.2601.5 6 21.41.012 6 0.30.017 6 0.11,824 6 531,794 6 53 771.353 6 1.09.1 6 0.2446.1 6 8.31.013 6 0.40.018 6 0.11,987 6 491,943 6 54 581.030 6 0.63.4 6 0.2989.2 6 45.11.012 6 0.30.020 6 0.12,185 6 652,163 6 65 270.953 6 0.76.0 6 0.2622.3 6 18.31.017 6 0.70.024 6 0.12,587 6 362,546 6 37 01.193 6 0.643.9 6 0.5158.7 6 2.01.018 6 0.20.036 6 0.13,891 6 1053,650 6 115 Briars04-022350.474 6 0.215.3 6 0.311.5 6 0.21.004 6 0.40.002 6 0.2246 6 16732 6 198 2150.399 6 0.216.1 6 0.227.9 6 0.31.004 6 0.40.007 6 0.1748 6 65480 6 149 1950.399 6 0.213.6 6 0.236.1 6 0.61.008 6 0.50.008 6 0.1817 6 120591 6 164 1700.381 6 0.29.5 6 0.160.3 6 0.91.002 6 0.40.009 6 0.11,009 6 51842 6 98 1500.403 6 0.210.2 6 0.281.2 6 1.91.015 6 0.40.013 6 0.41,359 6 4411,192 6 448 1350.418 6 0.39.9 6 0.290.6 6 1.71.002 6 0.60.013 6 0.11,441 6 1221,283 6 145 1030.349 6 0.210.6 6 0.275.9 6 1.11.004 6 0.40.014 6 0.11,546 6 781,344 6 128 780.285 6 0.24.2 6 0.2193.4 6 7.11.002 6 0.60.017 6 0.11,925 6 1041,826 6 115 430.346 6 0.26.8 6 0.2178.6 6 5.11.004 6 0.40.021 6 0.12,345 6 1002,215 6 120 00.607 6 0.336.1 6 0.3112.6 6 1.11.006 6 0.40.041 6 0.34,553 6 3134,138 6 369 BRC03-021410.897 6 0.32.9 6 0.2184.4 6 34.11.068 6 0.30.004 6 0.1367 6 32347 6 33 1320.972 6 0.43.9 6 0.3202.4 6 17.01.075 6 0.30.005 6 0.1500 6 36475 6 38 1101.102 6 0.54.6 6 0.3211.8 6 12.51.069 6 0.60.005 6 0.1550 6 29524 6 32 901.111 6 0.58.3 6 0.3182.9 6 54.01.071 6 0.60.008 6 0.1851 6 66805 6 70 500.928 6 0.217.0 6 0.3118.1 6 2.11.073 6 0.70.013 6 0.11,350 6 851,236 6 102 310.825 6 0.16.3 6 0.2299.6 6 9.61.076 6 0.40.014 6 0.11,424 6 561,377 6 61 170.839 6 0.620.7 6 0.3130.2 6 1.91.075 6 0.40.020 6 0.12,007 6 881,855 6 116 01.033 6 0.55.1 6 0.1849.9 6 19.01.080 6 0.30.026 6 0.12,601 6 452,571 6 47 P.E.V AN B EYNEN ,L.S OTO AND J.P OLK JournalofCaveandKarstStudies, April2008 N 29

PAGE 6

elementsextractedfromoneofthespeleothemsandwith otherpaleo-precipitationproxydatafromtheregion. Thefirstapproachistotrytorejectthehypothesisby comparingtraceelementswithdepositionratesforthe samespeleothem.Fairchildetal.(2000)demonstrated alinkbetweenMg/CaandSr/Caratiosinthedripwatersin twocavesandtheresidencetimeofwaterinthebedrock abovethecave.Longerresidencetimeswillleadto enhanceddolomitedissolutionofthebedrock,thereby augmentingMgandSrlevels.TheHawthornGroup, whichliesaboveBriarsCave,hasMgrichclaysandSr bearingsiliciclasticsanddolomiteinterspersedwithinthe carbonates(Scott,1997).Decreasedrainfallabovethecave promoteslongerresidencetimesofpercolatingwaters. Takingthesefindingsandapplyingthemtoourstudy,Mg/ CaandSr/Caratiosanddepositionrateswerecompared forBRIARS04-02(Fig.6)asthisiscurrentlytheonly speleothemexaminedinthisstudywithatraceelement record.Whilewedonothavethesamehighresolutionof thetraceelementsforthedepositionrates,itisapparent thatperiodsofincreasedcalcitedepositioncoincidewith higherMg/CaandSr/Caratios,especiallyduringperiods centeredon2.0and1.25kaBP.Thisresultthensuggests thatthedepositionrateisnegativelyrelatedtorainfall, causingustorejectourinitialhypothesis. Thereasonforanincreaseindepositionrateduring drierperiodsisprobablyduetotheincreaseinthecalcite saturationstateoftheseepagewaters.Aslowerflowof waterthroughthebedrockwouldallowmoretimefor dissolutionofthebedrock.Withinthecave,adecreasein dripratewouldleadtoathinnerfilmofwateronthe speleothemandmoredegassing.Bothsituationswould increasethedepositionrateforthespeleothem.Thisagrees withFairchildetal.(2000),whofoundthatincreased calciteprecipitationinthecaveoccurredduringthewinter whentherechargetothecavewasreducedduetothe frozenground.Baldini(2001)providesadifferentexplanationwherebyadecreaseinthedriprateduringdrier periodsallowsmoreCO 2 degassingonthestalactiteabove thestalagmite,therebyincreasingdepositionratesonthe stalactiteanddecreasingdepositiononthepairedstalagmite.Determiningwhichoftheseexplanationsappliesto thecentralFloridianstalagmitesrequirescomparisonof depositionratewithregionalpaleo-precipitationrecords. LittleSaltSpring(LSS)islocatedwithin100kmofour studysitesandAlvarezZarikianetal.(2005)interpreted variationsinthestableoxygenisotoperatio( d 18 Ovalues) of Cytheridellailosvayi (anostracodspecies)asmeasuresof changingprecipitation.Morenegative d 18 Ovaluescorrespondedtodrierperiodsandconversely,morepositive valuesindicatedwetterconditions.Figure7showsacomparisonbetweentheir d 18 Orecordandourdepositionrates fromBRIARS04-02.Itisreadilyapparentthatduring morenegativephasesinthe d 18 O(dryperiods),deposition Table2.DepositionratesforFloridianspeleothems.*Dateof32years 6 198hasbeendeletedfromthedatasetdueto unacceptableerrorrelativetoage. TimeIntervals(years)TimeofDeposition(years)VolumeofCalcite(mm 3 )CalciteDeposition(mm 3 /yr) BRC03-02(2571-1855)7164482963 BRC03-02(1855-1377)478160898337 BRC03-02(1377-1236)14145900326 BRC03-02(1236-805)43167057156 BRC03-02(805-524)28165220232 BRC03-02(524-475)49570221164 BRC03-02(475-347)12865313510 BRC03-02(347-0)34773788213 BRIARS04-01(3650-2546)11044364940 BRIARS04-01(2546-2163)38376308199 BRIARS04-01(2163-1929)22084738385 BRIARS04-01(1929-1794)14954585366 BRIARS04-01(1794-1482)31254333174 BRIARS04-01(1482-1210)27249388182 BRIARS04-02(4138-2215)19239805951 BRIARS04-02(2215-1826)389309237795 BRIARS04-02(1826-1344)482106266220 BRIARS04-02(1344-1283)6159322972 BRIARS04-02(1283-1192)9179194870 BRIARS04-02(1192-842)35093840268 BRIARS04-02(842-591)251115830461 BRIARS04-02(591-480)111101601915 BRIARS04-02(480-0)*48052969110 V ARIABLECALCITEDEPOSITIONRATESASPROXYFORPALEO-PRECIPITATIONDETER MINATIONASDERIVEDFROMSPELEOTHEMSIN C ENTRAL F LORIDA U.S.A 30 N JournalofCaveandKarstStudies, April2008

PAGE 7

ratesforBRIARS04-02increasedandduringmorepositive d 18 Ointervals,calciteprecipitationdecreased.LSSattained maximumaridityforthelateHolocenebetween2.6– 2.0kaBP(AlvarezZarikianetal.,2005),corresponding withthefirstmajordryperiodintheBRIARS04-02 speleothemrecord.Inall,thethreephasesofaridityinLSS duringthelast3.0kacoincidewithincreasedcalcite deposition.Thisprovidesmoreevidenceforrefutingour initialhypothesisandsuggeststhatforthisparticular study,decreasedprecipitationleadstoanincreasein speleothemdepositionrate.Itshouldbenotedthatour resolutionofdatingfrom2.5to4.2kaBPprecludesany detailedcomparisonoftrends,althoughtheotherlow depositionratesdoappeartocoincidewithmorepositive d 18 OvaluesinLSS. Anotherregionalcomparisonofpaleo-precipitationcan bemadewithLakeMiragoane,Haiti(Hodelletal.,1991). d 18 Ovaluesfromostracodesfoundinthesedimentsreflect changesinprecipitation/evaporation(P/E)ratios.More negative d 18 OvaluesshowaP/Eincreaseindicatinghigher lakelevels.Figure5showsthatLakeMiragoanerecorded thesamearidperiodcenteredon2.0kaasthespeleothem andLSSrecords.Thewetperiodsbefore2.0kaandlaterat 1.0kaasmeasuredintheotherrecordsarealsofoundin Haiti.Fromthesetwopaleo-records,itappearsthatthe Fairchildetal.(2000)explanationholdsforourFloridian speleothems.However,wecannotirrefutablydiscount apotentialcontributionoftheBaldini(2001)degassing mechanism. P ALEOCLIMATIC I NTERPRETATION Withthenumberofdateswehaveforeachspeleothem, thebestresolutionattainedis 100years.Consequently wecannotdiscussdecadal-scaleclimateinfluence,only Figure5.a)Variabledepositionrates(mm 3 yr 1 )foreachstalagmite.Notethedifferenttimescalesonthey-axes.b) Extensionrates(mmyr 1 )foreachspeleothemusingthetraditionallinearmethod. P.E.V AN B EYNEN ,L.S OTO AND J.P OLK JournalofCaveandKarstStudies, April2008 N 31

PAGE 8

thoseinthecentennialtimeframe.Thisexcludesthe AtlanticMultidecadalOscillation(AMO-Enfieldetal., 2001).DuringthecoldphaseoftheAMO,CentralFlorida experiencesdrierconditions,withaperiodof60–80years. However,thisisofshorterdurationthancanberesolved byourrecords.Anotherfactorthatproducesdroughtin theregionisthewesternextensionoftheNorthAtlantic High(NAH),whichdirectssubsidingairintoFlorida leadingtoreducedrainfall.However,atothertimesthe NAHispositionedfurthertotheeast,andcanalsodirect warmmoistCaribbeanairintothepeninsula,increasing rainfall. Ithasbeensuggestedthatincreasedtransportof CaribbeansurfacewatersandmoistureintotheGulfof Mexicoassociatedwiththenorthwardmigrationofthe averagepositionoftheITCZwouldinfluenceprecipitation intheregion(Hodelletal.,1991;Pooreetal.,2003).The morenortherlypositionoftheITCZwouldenhance easterlywindsbringingprecipitationtoHaiti(Hodellet al.,1991)andtheGulfofMexico(Pooreetal.,2003). Consequently,asouthwardshiftintheITCZmean positionwouldproduceperiodsofaridity.Theseshifts wereproposedtohaveoccurredoveramillennialtimescale.Thesimilaritiesofthedepositionratesfromthe FloridianspeleothemsandtheLSS-Haitipaleoprecipitationinterpretations(Fig.7)appeartorecordthisregional atmospheric-oceaninfluencesuggestedbyPooreetal. (2003).Theeasterlywindsdiscussedbytheseauthorsare generatedbytheNAHandconsequentlyprovidesome insightintohowthishighpressuresystemhasaffected Florida’sclimateduringthelateHolocene. C ONCLUSIONS Bycalculatingthevolumeofcalcitedepositedbetween U-seriesdates,anddividingthisbytheyearsbetweenthose dates,amorerealisticestimateofchangingdepositionrates canbeachieved,asopposedtothemethodofmerely measuringthemillimetersofcalcitedepositedeachyear alongthegrowthaxisbetweenknownages.Thislatter methoddoesnotrecognizechangesintheshapeofthe speleothemovertime. Theresultsfromourstudysuggestthatforthe speleothemsexamined,depositionrateiscontrolledby rainfallabovethecave,butnotintheacceptedmanneras suggestedbyHennigetal.(1983),Gascoyne(1992)or LauritzenandLundberg(1999).Theysuggestedwetter Figure7.Regionalpaleoprecipitationcomparisonfora) d 18 OvalueLakeMiragoanerecord,Haiti(5pointrunning mean);b) d 18 OvaluerecordfromLittleSaltSpring,Florida; c)DepositionratesfromBRIARS04-02.Shadedareas denoteperiodsofaridity.Notethat d 18 Ovaluesareplotted onaninvertedaxis. Figure6.Comparisonbetweenthetraceelementsand depositionratesforBRIARS04-02.a)Fivepointrunning meanappliedtotheSr/Caratios,b)Fivepointrunningmean appliedtotheMg/Caratios. V ARIABLECALCITEDEPOSITIONRATESASPROXYFORPALEO-PRECIPITATIONDETER MINATIONASDERIVEDFROMSPELEOTHEMSIN C ENTRAL F LORIDA U.S.A 32 N JournalofCaveandKarstStudies, April2008

PAGE 9

periodsledtogreatercalciteprecipitation.Ourstudy appearstocontradictthatfinding;periodsofaridity increasetherateofspeleothemgrowth.Sucharesult supportstheresearchdonebyFairchildetal.(2000).They foundMg/CaandSr/Caratiosincavecalcitearenegatively relatedtorainfallandourincreaseddepositionrates correspondwithperiodsofhigherMg/CaandSr/Caratios. Apaleo-precipitationrecordfromtheregion(Alvarez Zarikianetal.,2005)agreeswithourperiodsofaridityin CentralFlorida.Theseperiodsofaridityaremostlycaused byasouthwardshiftinthemeanpositionoftheITCZand aweakeningoftheNAHeasterlywindsthatdirectwarm waterandmoistairintotheNorthernGulfofMexico. FutureworkwillinvolveacalibrationstudyofBriars Cave,measuringtheMg/CaandSr/Caratiosincave dripwatersandrelatingittorainfalldataabovethecaveto testifourexplanationofthetraceelementscorrespondsto thatofFairchildetal.(2000).Secondly,asmorefunds becomeavailable,wehopetoimprovetheresolutionofour depositionratereconstructionwithmoreU/Thdates. A CKNOWLEDGEMENTS WethankAmyFrappier,PeterHarriesandPeterRowe fortheirthoughtfulreviewofthemanuscript.Ethan GoddardanalyzedthetraceelementsandVictorPolyak andYemaneAsmeromhelpedintheproductionofthe TIMSdates.ThisarticlewaspartialfundedbyUniversity ofSouthFloridaNewResearcherGrant. R EFERENCES AlvarezZarikian,C.A.,Swart,P.K.,Gifford,J.A.,andBlackwelder,P.L ., 2005,HolocenepaleohydrologyofLittleSaltSpring,Florida,based onostracodassemblagesandstableisotopes:Palaeogeography, Palaeoecology,Palaeoclimatology,v.225,p.134–156. Armstrong,B.,Chan,D.,Collazos,A.,andMallams,J.L.,2003,Doline andaquifercharacteristicswithinHernado,Pasco,andNorthern HillsboroughCounties:KarstStudiesofWestCentralFlorida. SouthwestFloridaWaterManagementDistrict,Brooksville,p.39–51. Baker,A.,Genty,D.,Dreybrodt,W.,Barnes,W.L.,Mockler,N.J.,and Grapes,J.,1998,Testingtheoreticallypredictedstalagmitedepositio n ratewithrecentannuallylaminatedsamples:Implicationsforpast stalagmitedeposition:GeochimicaetCosmochimicaActa,v.62, p.393–404. Baker,A.,Smart,P.L.,andFord,D.C.,1993,NorthwestEuropean palaeoclimateasindicatedbygrowthfrequencyvariationsof secondarycalcitedeposits:Palaeogeography,Palaeoclimatology, Palaeoecology,v.100,no.3,p.291–301. Baldini,J.U.L.,2001,Morphologicanddimensionallinkagebetween recentlydepositedspeleothemsanddripwaterfromBrownsFolly Mine,Wiltshire,England:JournalofCaveandKarstStudies,v.63, no.3,p.83–90. Curl,R.L.,1973,Minimumdiameterstalagmites:BulletinoftheNational SpeleologicalSociety,v.36,p.1–9. Dorale,J.A.,Gonza lez,L.A.,Reagan,M.K.,Pickett,D.A.,Murrell, M.T.,andBaker,R.G.,1992,Ahigh-resolutionrecordofHolocene climatechangeinspeleothemcalcitefromColdwaterCave,northeast Iowa:Science,v.258,p.1626–1630. Denniston,R.F.,Gonza lez,L.A.,Asmerom,Y.,Baker,R.G.,Reagan, M.K.,andBettis,E.A.,1999,Evidenceforincreasedcoolseason moistureduringthemiddleHolocene:Geology,v.27,no.9,p.815–818. Dreybrodt,W.,1999,Chemicalkinetics,speleothemgrowthandclimate: Boreas,v.28,no.3,p.347–356. Enfield,D.B.,Mestas-Nun ez,A.M.,andTrimble,P.J.,2001,TheAtlantic multidecadaloscillationanditsrelationtorainfallandriverflowsin thecontinentalU.S:GeophysicalResearchLetters,v.28,p.277–280. Fairchild,I.J.,Borsato,A.,Tooth,A.F.,Frisia,S.,Hawkesworth,C.J. Huang,Y.,McDermott,F.,andSpiro,B.,2000,Controlsontrace element(Sr-Mg)compositionsofcarbonatecavewaters:implications forspeleothemclimaticrecords:ChemicalGeology,v.166, p.255–269. Florea,L.,Hashimoto,T.,Kelley,K.,Miller,D.,andMrykalo,R.,2003, KarstGeomorphologyandRelationtothePhreaticSurface,Briars Cave,MarionCounty,Florida:KarstStudiesofWestCentral Florida.SouthwestFloridaWaterManagementDistrict,Brooksville, p.9–19. Franke,H.W.,1965,Thetheorybehindstalagmiteshapes:Studiesin Speleology,v.1,p.86–95. Frumkin,A.,andStein,M.,2004,TheSahara—EastMediterraneandust andclimateconnectionrevealedbystrontiumanduraniumisotopesin aJerusalemspeleothem:EarthandPlanetaryScienceLetters,v.217, p.451–464. Gams,I.,1981,Contributionstomorphometricsofstalagmites, in Beck, B.F.,ed.,ProceedingsoftheInternationalCongressofSpeleology, v.1,p.276–278. Gascoyne,M.,1992,Paleoclimaticdeterminationfromcavecalcite deposits:QuaternaryScienceReviews,v.11,p.609–632. Genty,D.,andQuinif,Y.,1996,Annuallylaminatedsequencesinthe internalstructureofsomeBelgianstalagmites;importancefor paleoclimatology:JournalofSedimentaryResearch,v.66,no.1, p.275–288. Harmon,R.S.,Schwarcz,H.P.,andFord,D.C.,1978,Stableisotope geochemistryofspeleothemsandcavewatersfromFlintRidgeMammothCaveSystem,Kentucky:Implicationsforterrestrial climatechangeduringtheperiod230,000to100,000yearsB.P.: JournalofGeology,v.86,p.373–384. Hendy,C.H.,1971,TheisotopicgeochemistryofspeleothemsI.The calculationoftheeffectsofdifferentmodesofformationonthe isotopiccompositionofspeleothemsandtheirapplicabilityas paleoclimateindicators:GeochimicaetCosmochimicaActa,v.35, p.801–824. Hennig,G.J.,Gruen,R.,andBrunnacker,K.,1983,Speleothems, travertines,andpaleoclimates:QuaternaryResearch,v.20,no.1, p.1–29. Hodell,D.A.,Curtis,J.H.,Jones,G.A.,Higuera-Gundy,A.,Brenner,M., Binford,M.W.,andDorsey,K.T.,1991,ReconstructionofCaribbean climatechangeoverpast10,500years:Nature,v.352,p.790–793. Holmgren,K.,Karlen,W.,Lauritzen,S.E.,Lee-Thorp,J.A.,Partridge, T.C.,Piketh,S.,Repinski,P.,Stevenson,C.,Svanered,O.,andTyson, P.D.,1999,A3000-yearhigh-resolutionstalagmite-basedrecordof palaeoclimatefornortheasternSouthAfrica:TheHolocene,v.9, no.3,p.295–309. Kaufmann,G.,andDreybrodt,W.,2004,Stalagmitedepositionand paleo-climate:aninverseapproach:EarthandPlanetaryScience Letters,v.224,p.529–545. Lachniet,M.S.,Burns,S.J.,Piperno,D.R.,Asmerom,Y.,Polyak,V.J., Moy,C.M.,andChristenson,K.,2004,A1500-yearElNin o/Southern oscillationandrainfallhistoryfortheIsthmusofPanamafrom speleothemcalcite:JournalofGeophysicalResearch,v.109, D20117p. Latham,A.G.,andSchwarcz,H.P.,1992,Carbonateandsulphate precipitates, in Ivanovich,M.,andHarmon,R.S.,eds.,Uraniumseriesdisequilibrium;applicationstoEarth,marine,andenvironmentalsciences:Oxford,ClarendonPress,p.423–459. Lauritzen,S.E.,andLundberg,J.,1999,Speleothemsandclimate:aspeci al issueoftheHolocene:Holocene,v.9,p.643–647. Li,W.-X.,Lundberg,J.,Dickin,A.P.,Ford,D.C.,Schwarcz,H.P., McNutt,R.,andWilliams,D.,1989,High-precisionmass-spectrometricuranium-seriesdatingofcavedepositsandimplicationsfor palaeoclimaticstudies:Nature,v.339,p.524–536. Musgrove,M.L.,Banner,J.L.,Mack,L.E.,Combs,D.,James,H.,and Edwards,R.L.,2001,GeochronologyoflatePleistocenetoHolocene speleothemsfromcentralTexas:Implicationsforregionalpaleoclimate:GSABulletin,v.113,no.12,p.1532–1543. Polyak,V.J.,Rasmussen,J.B.,andAsmerom,Y.,2004,Prolonged droughtperiodinthesouthwesternUnitedStatesthroughthe YoungerDryas:Geology,v.32,no.1,p.5–8. P.E.V AN B EYNEN ,L.S OTO AND J.P OLK JournalofCaveandKarstStudies, April2008 N 33

PAGE 10

Polyak,V.J.,andAsmerom,Y.,2001,LateHoloceneclimateandcultural changesinthesouthwesternUnitedStates:Science,v.165,p.971–981. Poore,R.Z.,Dowsett,H.J.,Verardo,S.,andQuinn,T.M.,2003, Millennial-tocentury-scalevariabilityinGulfofMexicoHolocene climaterecords:Paleoceanography,v.18,no.2,1048,doi:10.1029/ 2002PA000868. QinXiaoguang,TanMing,LiuTungsheng,WangXianfeng,LiTieying,and LuJinpo,1999,Spectralanalysisofa1000-yearstalagmitelamina thicknessrecordfromShihuaCavern,Beijing,China,anditsclimatic significance:TheHolocene,v.9,p.689–694. Reeder,P.,andBrinkmann,R.,1998,Paleoenvironmentalreconstruction onanOligoceneagedislandremnantinFlorida,USA:Caveand KarstScience,v.25,no.1,p.7–13. Scott,T.M.,1997,MiocenetoHolocenehistoryofFlorida, in Randazzo, A.F.,andJones,D.S.,eds.,TheGeologyofFlorida,TheUniversityof FloridaPress,Gainesville,p.57–68. SouthEastRegionalClimateCenter,http://cirrus.dnr.state.sc.us/cg i-bin/ sercc/cliMAIN.pl?fl1046 VanBeynen,P.E.,Henry,P.S.,andDerek,C.F.,2004,Holoceneclimatic variationrecordedinaspeleothemfromMcFail’sCave,NewYork: JournalofCaveandKarstStudies,v.66,no.1,p.20–27. White,W.,2004,Paleoclimaterecordsfromspeleotheminlimestonecaves, in Sasowsky,I.D.,andMylroie,J.,eds.,Studiesof CaveSediments-PhysicalandChemicalRecordsofPaleoclimate,KluwerAcademic/PlenumPublishers,NewYork,p.135– 175. V ARIABLECALCITEDEPOSITIONRATESASPROXYFORPALEO-PRECIPITATIONDETER MINATIONASDERIVEDFROMSPELEOTHEMSIN C ENTRAL F LORIDA U.S.A 34 N JournalofCaveandKarstStudies, April2008



PAGE 1

CASTILEEVAPORITEKARSTPOTENTIALMAPOFTHE GYPSUMPLAIN,EDDYCOUNTY,NEWMEXICOAND CULBERSONCOUNTY,TEXAS:AGIS METHODOLOGICALCOMPARISON K EVIN W.S TAFFORD 1,2 ,L AURA R OSALES -L AGARDE 1,2 AND P ENELOPE J.B OSTON 1,2 Abstract: CastileFormationgypsumcropsoutover 1,800km 2 inthewestern DelawareBasinwhereitformsthemajorityoftheGypsumPlain.Karstdevel opmentis wellrecognizedintheGypsumPlain(i.e.,filledandopensinkholeswitha ssociated caves);however,thespatialoccurrencehasbeenpoorlyknown.Inorderto evaluatethe extentanddistributionofkarstdevelopmentwithintheCastileportiono ftheGypsum Plain,combinedfieldandGeographicInformationSystem(GIS)studieswe reconducted, whichenableafirstapproximationofregionalspeleogenesisanddelinea tekarst-related naturalresourcesformanagement.Fieldstudiesincludedphysicalmappi ngof50,1-km 2 sites,includingidentificationofkarstfeatures(sinkholes,caves,an dsprings)and geomorphicmapping.GIS-basedstudiesinvolvedanalysesofkarstfeatur esbasedon publicdata,includingDigitalElevationModel(DEM),DigitalRasterGra phic,(DRG) andDigitalOrthophotoQuad(DOQ)formats.GISanalysesconsistentlyund erestimate theactualextentanddensityofkarstdevelopment,basedonkarstfeature sidentified duringfieldstudies.However,DOQanalysescoupledwithfieldstudiesap pearsto produceaccuratemodelsofkarstdevelopment.Asaresult,akarstpotenti almapofthe Castileoutcropregionwasdevelopedwhichrevealsthatkarstdevelopmen twithinthe CastileFormationishighlyclustered.Approximately40%oftheregionef fectively exhibitsnokarstdevelopment( 1feature/km 2 ).Twosmallregions( 3km 2 each) displayintensekarstdevelopment( 40features/km 2 )locatedwithinthenorthernextent oftheGypsumPlain,whilemanyregionsofsignificantkarstdevelopment( 15features/ km 2 )aredistributedmorewidely.Theclustereddistributionofkarstdevelo pment suggeststhatspeleogenesiswithintheCastileFormationisdominatedby hypogenic, transverseprocesses. I NTRODUCTION ThegypsumfaciesoftheCastileFormationcropsout overanareaof 1800km 2 inEddyCounty,NewMexico andCulbersonCounty,Texasonthewesternedgeofthe DelawareBasin(Fig.1).Theregionhastraditionallybeen referredtoastheGypsumPlain(Hill,1996),whichcovers anareaof 2800km 2 andiscomposedofoutcropsofthe CastileandRustlerFormations(Fig.2).Theregionis locatedinthesemi-aridsouthwestonthenorthernedgeof theChihuahuanDesert,whereannualprecipitation averages26.7cmwiththegreatestrainfalloccurringas monsoonalstormsinlatesummer(July–September) (Sares,1984).Annualtemperatureaverages17.3 u Cwith anaverageannualminimumandmaximumof9.2 u Cand 25.2 u C,respectively. ThroughoutCastileoutcrops,surficialkarrenoccurs extensivelyinregionsofexposedbedrock,includingwelldevelopedrillenkarren,spitzkarren,kamenitzasandtumuli.Sinkholedevelopmentiswidespread,includingboth closedandopensinkholesrangingfromnear-circular featurestolaterallyextensive,incisedarroyo-likefeatures. Cavedevelopmentrangeswidely,fromsmallepigenic rechargefeaturestolarge,complexpolygeneticfeatures (Stafford,2006).Theregionhoststhesecondlongest documentedgypsumcaveinNorthAmerica,ParksRanch Cave,EddyCounty,N.M.,withasurveyedlengthof 6596m(Stafford,2006).Inaddition,manyothersignificantgypsumcaveshavebeendocumentedbytheTexas SpeleologicalSurvey(TSS)(e.g.,ReddellandFieseler, 1977)andGYPsumKArstProject(GYPKAP)(Eaton, 1987;Belski,1992;Lee,1996).However,nosystematic investigationhasbeenconductedwithintheregionwith respecttokarstdevelopment.Priortothisstudy,246karst features,primarilycaves,weredocumentedwithinthe Castileoutcropregion.TheBLM(BureauofLand Management)documented45ofthetotalreportedkarst features(JonJasper,2006,pers.com.);whiletheTSS 1 DeptofEarthandEnvironmentalScience,NewMexicoInstituteofMiningan d Technology,Socorro,NM87801,USA.kwstafford@juno.com,lagarde@nmt. edu andpboston@nmt.edu 2 NationalCaveandKarstResearchInstitute,Carlsbad,NM,88220,USA K.W.Stafford,L.Rosales-Lagarde,andP.J.Boston–Castileevaporiteka rstpotentialmapoftheGypsumPlain,EddyCounty,New MexicoandCulbersonCounty,Texas:AGISmethodologicalcomparison. JournalofCaveandKarstStudies, v.70,no.1,p.35–46. JournalofCaveandKarstStudies, April2008 N 35

PAGE 2

documented201ofthetotalreportedkarstfeatures(Jim Kennedy,2006,pers.com.). Therapidsolutionkineticsandhighsolubilityof gypsumpromotesextensivekarstdevelopment.Gypsum solubility(2.53gL 1 )isapproximatelythreeordersof magnitudegreaterthanlimestone(1.5mgL 1 )inpure waterandtwoordersofmagnitudelessthanhalite(360g L 1 )(Klimchouk,1996).Thehighsolubilityandnearlinearsolutionkineticsofevaporitesencourageintense surfacedissolutionthatoftenformslargesinkholes,incised arroyosandcavesthatarelaterallylimitedwithrapid decreasesinpassageapertureawayfrominflowsthrough epigenicspeleogenesis(Klimchouk,2000a).Additionally, thehighsolubilitiesofevaporitesfavorthedevelopmentof hypogenictransversespeleogenesisdrivenbymixedconvection(forcedandfree)(Klimchouk,2000b).Forced convectionisestablishedbyregionalhydraulicgradientsin confinedsettings,whilefreeconvectionisgeneratedwhere steepdensitygradientsestablishasfresh-watersare continuouslysuppliedtothedissolutionfronts(theupper levels)throughthesimultaneoussinkingofsaturatedfluids bydensitydifferences(AndersonandKirkland,1980). Thereforeepigenicandhypogenickarsticfeatureslikely bothexistinthestudyarea,oftensuperimposedoneach other. Theworkwereportherefocusesondelineatingthe extentanddistributionofkarstdevelopmentwithinthe outcropregionoftheCastileFormation,inorderto predictregionsofintenseversusminimalkarstdevelopment,whichcanbeusedforkarstresourcemanagementas wellasafirstapproximationforunderstandingregional speleogenesis.AdualapproachinvolvingfieldandGeographicInformationSystem(GIS)analyseswereutilizedin ordertodefinekarstvariabilitywithinthestudyarea, includingfieldmappingof50,1-km 2 regionsandGIS analyses,usingESRIArcGIS9.2software,ofpublicdata (i.e.,DigitalElevationModel[DEM];DigitalRaster Graphic[DRG];andDigitalOrthophotoQuad[DOQ]) fortheentireregion.Thecombinedresultswereusedto developakarstpotentialmapoftheCastileFormation outcropregion,whilesimultaneouslyevaluatingdifferent GIS-basedtechniquesforkarstanalyses. G EOLOGIC S ETTING TheCastileFormationwasdepositedduringthelate Permian(earlyOchoan),subsequenttodepositionofthe GuadalupianCapitanReef,whichiswell-knownforthe cavesithostsintheGuadalupeMountains(e.g.,Hoseand Pisarowicz,2000).Castileevaporitesrepresentdeep-water depositswithinastratified,brine-filledbasin(i.e.,DelawareBasin)(KendallandHarwood,1989),boundedbelow byclasticsoftheBellCanyonFormation,onthemargins byCapitanReefcarbonates,andabovebyadditional evaporiticrocksoftheSaladoandRustlerFormations (Fig.2)(Kelley,1971).Castileevaporitescropoutalong theirwesterndissolutionfrontintheGypsumPlain (Fig.1),diptotheeastwheretheyreachamaximum thicknessof480minthesubsurface(Hill,1996),andare characterizedasmassivetolaminatedsulfates(gypsum/ anhydrite)interbeddedwithhalite(Dietrichetal.,1995). Increasedthicknessintheeasthasbeenattributedto dissolutionofintrastratalhalitetothewestandincreased depositiontotheeastintheOchoaTroughduringthe Permian(Andersonetal.,1972). TheCastileFormation,includingoutcropsinthe GypsumPlain,hasexperiencedminimaltectonicdeformationalthoughlocatedontheeasternedgeofmajortectonic events.TriassicandLaramidetectonismproducedregional tiltingtothenortheast,broadflexuresandfracturingwith minimaloffsetwithinsoutheasternNewMexicoandwest Texas.ThefarwesternedgeoftheDelawareBasinhas beendown-droppedalongthefareasternmarginofBasin Figure1.LocationmapshowinglocationofGypsumPlain includingoutcropareasoftheCastileFormation(solid white)andtheRustlerFormation(solidblack)withinthe DelawareBasin(darkgray),EddyCounty,NMand CulbersonCounty,Texas.LocationoftheDelawareBasin inrelationtoTexasandNewMexicoisillustratedinbottom leftcorner,withtheenlargedregionoutlinedbythesmall blackrectangle(adaptedfromKelley,1971,Dietrichetal., 1995andHill,1996). C ASTILEEVAPORITEKARSTPOTENTIALMAPOFTHE G YPSUM P LAIN ,E DDY C OUNTY ,N EW M EXICOAND C ULBERSON C OUNTY ,T EXAS :AGIS METHODOLOGICALCOMPARISON 36 N JournalofCaveandKarstStudies, April2008

PAGE 3

andRangeblockfaulting;however,withintheremaining DelawareBasin,theeffectsarelimitedtonear-vertical joints(Horak,1985).Asaresultoftectonism,Castile evaporitescurrentlydip3to5degreestothenortheastwith abundantconjugatejointsetsorientedat N75 u Eand N15 u W.Associatedwithjointsetsalongthewestern dissolutionfront,solutionsubsidencevalleyshavedevelopedfromsubsurfacedissolutionofhalitebeds(Hentz andHenry,1989). Inadditiontotectonicdeformation,somesulfaterocks havebeenexposedtosignificantdiagenesis.Original laminated(varved)gypsumoftenexhibitsmassiveand nodularfabricsthatarelikelytheresultofplastic deformationassociatedwithanhydrite/gypsummineral conversion(MachelandBurton,1991).Calcitizedevaporitesarecommon(oftenreferredtoascastilesorcalcitized masses),generallyformingclustersorlineartrendsof biogeniclimestoneassociatedwithbacterialsulfatereduction(KirklandandEvans,1976).Seleniteislocally abundant,forminglinearfeaturesandfracturefillings (likelyassociatedwithmineralconversion),aswellas lenticularmasses(probablyassociatedwithcalcitization processes).DiageneticfabricalterationwithinCastile evaporitesprobablyhasexertedsignificantinfluenceon establishingpreferentialflowpathsforkarstdevelopment withintheGypsumPlain. F IELD S TUDIES Fieldmappingwasconductedat50,1-km 2 siteswithin theCastileoutcroparea(Fig.3A).Fieldsiteswere randomlyselectedusingESRIArcGIS9.2softwarein ordertoobtainanaccuraterepresentationofkarst developmentwithintheCastileoutcropregionand minimizeanyhumanbiasesthatmightbeintroducedinto siteselection.Tenfieldmappingsiteswereshiftedupto twokilometersawayfromGIS-definedlocations,inorder toavoidanthropogenicfeatures(i.e.,roads,houses, quarries),whiletwositeswereshifteduptofourkilometers toavoidregionswherelandaccesswasnotavailable. Eachfieldsitewasdefinedasaonekilometersquare region.Transectsurveyswereconductedon100-meterline spacing,suchthatten,onekilometerlongtransectswere traversedineachofthe50fieldsites.Smallerline-spacing (40m)fortransectsurveyswascomparedwith100-meter linespacingthroughindependentsurveysbytwoofthe authorsatfivefieldsites,whichidentifiedlessthan10% additionalkarstfeatures(i.e.,sinkholesandcaves). Becauseoftheresultsofsub-samplingandthelocation ofthestudyregionwithinthesemi-aridsouthwest,where vegetationissparseanddoesnotcommonlyobscurekarst features,100-meterspacedtraversesurveyswerefoundto besufficienttodocumentmorethan90%ofsurficialkarst features.Duringfieldmapping,identifiedfeaturelocations wererecordedwithahand-heldGPS(GlobalPositioning System)andindividualfeatureswerecharacterizedbased onsize(length,width,depth),geomorphicexpression (closedsink,opensink[i.e.,cave],spring)andgeologic occurrence(laminated,massiveandnodulargypsum; gypsite;calcitizedevaporite). Fieldmappingidentified389individualkarstfeatures, including236opensinkholeswithfreedrains(i.e.,cavesor smallersolutionalconduitsthatconnectdirectlytosinkholes),147filledsinkholes,fourcaveswithnoassociated sinkhole,andtwosprings.However,ofthe236open sinkholes,only39containedcavesthatwerelargeenough tobehumanlyenterable.Ofthe50fieldsites,12contained nokarstfeaturesand14sitescontainedmorethan10 features(Fig.4).Onlytwositescontainedmorethan30 features,onewith31andonewith48. Featureswerefoundinawiderangeofgypsumfabrics (Fig.5).Cavesarelargelydevelopedinlaminated( 43% offeatures)(Fig.5A)andmassivefabrics( 26%of features)(Fig.5B);however,numeroussmallsurficial cavesformingypsite( 28%offeatures)(Fig.5D).Caves wereoccasionallyfoundinselenite( 2%offeatures) (Fig.5C)andcalcitizedmasses( 2%offeatures) (Fig.5E).Filledsinkholesweregenerallyfoundingypsite oralluvium;however,thislikelyonlyrepresentssurficial mantlingoverdeeperfeaturesinmostcases. Sinkholeareaandvolumerangedwidelywithinthe surveyedsites.Theaverageopensinkholeareawas1.99 3 10 3 m 2 (0.3to4.12 3 10 4 m 2 )withanaveragevolumeof Figure2.DiagrammaticrepresentationoflatePermian (GuadalupianandOchoan)depositsassociatedwiththe GuadalupeMountains(left)andDelawareBasin(right). NotethattheCastileFormationfillsinthebasinandmarks thebeginningoftheOchoan(dashedwhiteline)(adapted fromHill,1996). K.W.Stafford,L.Rosales-Lagarde,andP.J.Boston JournalofCaveandKarstStudies, April2008 N 37

PAGE 4

1.73 3 10 3 m 3 (8.0 3 10 2 to4.71 3 10 4 m 3 ).Theaverage areaofclosedsinkholeswas1.01 3 10 3 m 2 (3.0 3 10 2 to 2.36 3 10 4 m 2 )withanaveragevolume3.70 3 10 2 m 3 (5.0 3 10 3 to6.54 3 10 3 m 3 ).Sinkholeareawascalculatedby treatingfeaturesassimpleellipsesbasedonthemaximum widthandlengthmeasuredinthefield,whilesinkhole volumewascalculatedbytreatingfeaturesasconical ellipsesbasedonellipticalareaandsinkholedepth. Therefore,approximatedsinkholeareasandvolumes probablyoverestimatetruevalues. GISA NALYSES InthelastdecadeGIShasbeenrecognizedasapowerful toolforgeographicanalysesandhasbecomeausefultool forcaveandkarststudies(e.g.,Szukalskietal.,2002). Publicdataisavailableinmultipleformatsthrough governmentagencies,suchasUnitedStatesGeological Survey(USGS),NewMexicoBureauofGeologyand MineralResources(NMBGMR),andTexasNatural ResourceInventoryService(TNRIS),whichenablesGIS analysesoflargekarstregionsatzerocost. GISanalysesofkarstterrainshavebeenusedinvarious studiestodelineatekarstdevelopment.Floreaetal.(2002) combinedknownpointlocationsforkarstfeatureswith digitizedsinkholesfromDRGstodevelopkarstpotential mapsinKentucky,whileDenizman(2003)conducted similarstudiesinFlorida.Tayloretal.(2005)demonstratedtheuseofDEMsfordelineatingsinkholesinKentucky. Hungetal.(2002)usedanintegratedapproachinvolving analysesofmultispectralimagery,aerialphotography,and DEMstoevaluaterelationshipsbetweenlineamentsand cavedevelopment. BecausemostpreviouskarststudiesusingGIShave focusedononeortwotechniques,multiplepublicdata formats(DEM,DRGandDOQ)werecomparedand evaluatedinthisstudy,notonlytocharacterizetheextent ofkarstdevelopmentbuttoalsotesttheintercomparability ofdifferentmethodologies.Physicalmappingofkarst featuresinthefield,describedintheprevioussection,was furthercomparedwithGIStechniquestofullyevaluatethe accuracyofGIS-basedapproaches.Whilefieldmapping identifiedthetrueoccurrenceofkarstfeatureswithin specificregions,theGISanalysesonlyrepresentapproxFigure3.A)Castileoutcropregion(gray)showinglocationofthe50rando mlyselected1km 2 siteswherefieldmappingwas conducted;B)Castileoutcropregion(gray)showingsinks(closeddepres sions)determinedbyDEManalysis(boxedarea includes 75%ofthecloseddepressionsidentifiedthroughDEManalysis). C ASTILEEVAPORITEKARSTPOTENTIALMAPOFTHE G YPSUM P LAIN ,E DDY C OUNTY ,N EW M EXICOAND C ULBERSON C OUNTY ,T EXAS :AGIS METHODOLOGICALCOMPARISON 38 N JournalofCaveandKarstStudies, April2008

PAGE 5

imationsbasedonthegeomorphicexpressionofkarst features(Fig.6,7). Digitalelevationmodels(DEM)wereanalyzedto definecloseddepressions(i.e.,sinks)withintheCastile outcropregion.Closeddepressionswereidentifiedby creatinganewDEMwithfilledsinksthroughGIS processing,whichwascomparedwiththeoriginalDEM todeterminethedifferencebetweendatasets(Fig.3B,6B) (Tayloretal.,2005).Theresultingdataincluded554 individualsinkholeswithanaverageareaof2.57 3 10 4 m 2 (6.0 3 10 2 to8.70 3 10 5 m 2 );however,approximately80% oftheidentifiedfeaturesoccurredwithina26km(16mile) widestripimmediatelysouthoftheNewMexico–Texas stateline.Lessthan5%ofthefeaturesoccurrednorthof thestripofabundantcloseddepressions,whilethe remainderwasdistributedsouthofthestrip(Fig.6B). AlthoughallpublicdatausedforDEManalyseshad10meterpostings,theresultingsinkholemapsuggeststhat thereissignificantvariabilityinthesourcematerialusedto createtheseDEMs.Theregionofsinkholeabundance appearstorepresentwelltheactualcloseddepressions withinthestudyarea,whileregionsoutsidethisarea appeartosignificantlyunderestimatefeatureabundance. Digitalrastergraphics(DRG)of1:24,000USGS topographicmapswereanalyzedforthestudyareaand allcloseddepressionsweredigitizedasindicatorsof individualkarstfeatures(Fig.6C,7B);however,itislikely thatmultiplekarstfeaturesexistwithinsomelarge,closed depressions.FromDRGs,552individualcloseddepressionswereidentified(Fig.7B),withanaverageareaof1.54 3 10 4 m 2 (53m 2 to1.74 3 10 6 m 2 ),basedonGISspatial analyses.Becausetopographicmapsofthisregionare basedon20foot(6.1m)contourintervals,numeroussmall sinkholes,includingmostofthefeaturesdocumented duringfieldmapping,arenotrepresented.However,most ofthekarstfeaturesdocumentedbytheBLMandTSSare representedassinkholesonDRGsbecausetopographic mapshavetraditionallybeenusedforlocatingand identifyingkarstfeatures. Digitalorthophotoquads(DOQ)withinthestudy regionhavearesolutionofonemeter.DOQanalyseswere conductedbyvisuallypickingprobablekarstfeatures (Fig.6D)ataresolutionof1:4,000.Featureswere identifiedbasedongeomorphicexpressionthroughcomparisonwithknowncaveandkarstfeatures,either documentedbytheBLMinNewMexicoandtheTSSin Texasorfeaturesdocumentedduringfieldmapping.Based oncomparisonwithknownfeatures,3,237individual featureswereidentifiedwithintheCastileoutcropregion (Fig.7C). Spatialanalysesoffeaturedensitieswereperformedin ordertodelineatekarstdevelopmentwithinthestudyarea. Threesetsofdatawereprocessedseparatelytoevaluate karstdensity,including:1)knowncavesdocumentedby theTSSandBLM(Fig.7D);2)DRGdefinedsinks (Fig.7E);and3)featuresidentifiedthroughDOQanalyses (Fig.7F).Densityanalysesoffeaturesidentifiedfrom DEMdatawasnotconductedbecauseoftheapparenthigh degreeofvariabilityinqualityofthesepublicdatasets.All densityanalysesindicateintensekarstdevelopmentwithin thenorthwesternportionofthestudyareaandageneral decreaseinfeatureabundancetowardstheeast. D ISCUSSION Studiesconductedtodeterminetheextentanddistributionofkarstdevelopmentvarywidely(Veni,2002),but GIS-basedstudieshaveenabledsignificantadvancesin geographicanalyseswithinthelastdecade.Analysesof knownkarstdistributionsandfeaturesidentifiedthrough GISwithintheCastileoutcropregionallshowsimilar trendsforareasofsignificantkarstdevelopment.However, thedegreeofresolutionofvariouspublicdatausedinGIS analysesproducessubstantialdifferencesinevaluationof karstdevelopmentthroughouttheentireregion(Fig.8), suggestingthatfieldstudiesshouldalwaysbecoupledwith anyGIS-basedstudies.SinkholesidentifiedthroughDEM analyseswerenotusedtodevelopkarstdensitymaps becauseoftheapparentvariabilitywithintheoriginaldata usedtodeveloptheDEMs.However,theDEMvariability illustratesanimportantpoint,inthatpublicdatamustbe interpretedwithcaution. Analysisofpreviouslydocumentedcaveandkarst featureswithintheCastileoutcropregionindicatesmall clustersofcaves,focusedinthenorthwesternregionofthe studyarea,largelyalongthedissolutionalmarginofthe CastileFormation;however,onlyminorregionsofkarst developmentareobservedscatteredthroughouttherestof Figure4.Comparativeplotshowingkarstfeaturesidentifiedduringfieldmappingcomparedwithfeaturesidentified throughDOQanalysesforthe50,1km 2 fieldsites.Field mappingandDOQanalysesareonlyshownbecausemost DEMandDRGanalysesshowednofeaturesintheregions wherefieldmappingwasconducted.NotethatDOQanalysis identified 35%offeaturesthatwerelocatedduring fieldmapping. K.W.Stafford,L.Rosales-Lagarde,andP.J.Boston JournalofCaveandKarstStudies, April2008 N 39

PAGE 6

thestudyarea(Fig.7D).Basedonpreviouslydocumented features,approximately95%ofthestudyareaeffectively exhibitsnokarstdevelopment( 1feature/km 2 ).Studies basedondocumentedkarstfeaturesinherentlycreate biasedresultsthatmaynotaccuratelydepictthecomplete distributionofkarstdevelopment.Biasesareintroducedby variableaccesstoportionsofakarstregion,suchas regionswherelandowneraccessisnotavailableorregions thatareremotewithpoorroadaccess. AnalysisofcloseddepressionsdepictedonDRGs (Fig.7E)showssimilarpatternsofkarstdevelopmentas documentedkarstdistributions(Fig.7D),butdonotshow anyregionswithdensitiesgreaterthan10features/km 2 DRGanalysesshowsgreaterdistributionsofkarstfeatures thandocumentedcaveanalyses,expandingthepredicted boundariesofkarstdevelopment;however,themajorityof thestudyarea( 90%)stillappearstohaveminimalkarst development( 1feature/km 2 ).Aswithanalysesof documentedcaves,DRGsappeartounderestimatethe actualextentofkarstdevelopmentbecausethecontour intervalofDRGspreventsdistinguishablerepresentation ofsmallcloseddepressionsandnarrow,incisedkarst arroyos. AnalysisofkarstfeaturesidentifiedonDOQsindicates asignificantlygreaterdegreeofkarstdevelopmentdensity anddistribution(Fig.7F)asopposedtootherGIS-based analyses.Regionsofminimalkarstdevelopmentwere reducedtoapproximately50%andseveralregionswith karstfeaturedensitiesgreaterthan15features/km 2 were identified.Intensekarstdevelopmentstillappearsconcentratedwithinthenorthwesternportionofthestudyarea; however,regionsofextensivekarstdevelopmentare Figure5.CavedevelopmentintheCastileFormationoccurswithinawidera ngeoflithologicfabrics.A)PlummetCave: laminatedgypsum;B)ParksRanchCave:massivegypsum;C)BlackWidowHole :selenite;D)PokeyCave:gypsite,ande) DeadBunnyHole:biogeniclimestone(calcitizedevaporite). C ASTILEEVAPORITEKARSTPOTENTIALMAPOFTHE G YPSUM P LAIN ,E DDY C OUNTY ,N EW M EXICOAND C ULBERSON C OUNTY ,T EXAS :AGIS METHODOLOGICALCOMPARISON 40 N JournalofCaveandKarstStudies, April2008

PAGE 7

identifiedthroughouttheentirewesternhalfoftheCastile outcroparea,aswellasseveralsmallerregionsclosertothe easternmarginofthestudyarea.AlthoughDOQanalysis showsmoreextensivekarstdevelopment,itisinherently biasedbecausefeatureswerevisuallypickedbasedon comparisonwiththegeomorphicexpressionofknown features.Comparisonofkarstfeaturesphysicallydocumentedduringfieldstudieswithfeaturesidentifiedthrough DOQanalysis,withintheboundariesoffieldsitesmapped, indicatesthatDOQanalysisconsistentlyunderestimates Figure6.Variabilityinkarstidentificationthroughvariousmethodolo gieswithinarepresentative1km 2 fieldsite(each squareregionmeasures1kmby1km).A)filledblackcirclesrepresenteigh tkarstfeaturesdocumentedthroughphysical mappingoffieldsite;B)originalDEMoffieldsitefromwhichnokarstfeat ures(closeddepressions)wereidentifiedduring GISanalysis(notedarkershadinginupperleftisthehighestelevations) ;C)DRGoffieldsiteshowingnocloseddepressions, butablind-terminated,ephemeralstreamsuggestsinkpoint(arrow);and D)DOQoffieldsiteshowinggeomorphicvariability andthelocationofthreefeatures(blacktriangles)whichcouldberesolv edthroughDOQanalysis. K.W.Stafford,L.Rosales-Lagarde,andP.J.Boston JournalofCaveandKarstStudies, April2008 N 41

PAGE 8

Figure7.ComparisonofdatausedfordensityanalyseswithintheCastileo utcropregion(grey).A)pointdataforindividual karstfeaturespreviouslydocumentedbytheTSSandBLM;B)closeddepress ionsdigitizedfromDRGs;C)pointdatafor individualkarstfeaturesidentifiedthroughDOQanalysis;D)karstfeat uredensitymapbasedonpreviouslydocumentedkarst featuresinFig.6A;E)karstfeaturedensitymapbasedondistributionofi ndividualcloseddepressionsdigitizedfromDRGs showninFig.6B;andF)karstdensitymapbasedonfeaturesidentifiedthro ughDOQanalysisshowninFig.6C.Colorshading inkarstdensitymapsrepresentthenumberofkarstfeatures/km 2 ,where:gray 1feature/km 2 ;blue = 1 5features/km 2 ; green = 5 10features/km 2 ;yellow = 10 15features/km 2 ;andred 15features/km 2 C ASTILEEVAPORITEKARSTPOTENTIALMAPOFTHE G YPSUM P LAIN ,E DDY C OUNTY ,N EW M EXICOAND C ULBERSON C OUNTY ,T EXAS :AGIS METHODOLOGICALCOMPARISON 42 N JournalofCaveandKarstStudies, April2008

PAGE 9

thetotalnumberofkarstfeaturespresent(Fig.4).This underestimationislikelytheresultofthe1-meter resolutionoftheDOQpublicdata. DOQanalysisappearstobestrepresentkarstdevelopmentwithintheoutcropregionoftheCastile Formation;however,allGIS-basedanalysesappearto underrepresenttheextentofactualkarstdevelopmentas comparedtophysicalkarstsurveysconductedinthefield (Fig.8).DOQanalysesgenerallyidentify36%ofthe featuresdocumentedduringfieldstudies(Figs.4,6,7,and 8),whileotherGISanalysescommonlyidentifylessthan 5%ofthefeaturesdocumentedduringfieldstudies. Therefore,DOQdensityanalysiswasweightedbyafactor of2.77usingSpatialAnalyst,inordertoadjustthe densitiescalculatedthroughGISascomparedtodensities documentedduringfieldstudies.Asaresult,akarst potentialmapwasdevelopedfortheentireoutcropregion oftheCastileFormation(Fig.9),whichindicatesthatless than40%oftheoutcropregioncontainseffectivelyno karstdevelopment( 1feature/km 2 ),whiletwosmall regions( 3km 2 each)withinNewMexicoexhibitintense karstdevelopment( 40features/km 2 ).Comparativetests oflinespacingusedintransect-basedfieldmapping, suggeststhattheactualdensityofkarstfeaturesmaybe atleast10%greater(Fig.8).Thekarstpotentialmaplikely representskarstdevelopmentrelativelyaccuratelywithin thestudyarea,butacompletephysicalsurveyoftheentire 1,800km 2 regionwouldprobablyshowdiscrepancies. C ONCLUSIONS Thedevelopmentofakarstpotentialmapforthe CastileFormationshowsthatkarstdevelopmentis distinctlyclusteredwithintheGypsumPlain(Fig.9). Visualinterpretationoftheclusteringdistributionofkarst featureswithintheCastileoutcropregionwasconfirmed throughGIS-basednearestneighboranalyses(Fordand Williams,1989),whichyieldedanearestneighborindexof 0.439.Anearestneighborindexof1isclassifiedasrandom whilevaluesgreaterthan1approacharegular,evenly spacedpatternwhilevalueslessthan1approachgreater clustering(FordandWilliams,1989).Anearestneighbor indexof0.439indicatessignificantclustering.Large regionsexhibitminimalsurficialkarstexpressions,primarilyalongthesouthernandeasternedgesoftheCastile outcroparea. Thedensestregionsoccurinthenorthwesternportionof theoutcroparea,andcommonlycontainmorethan20 features/km 2 (Fig.9),withmorethan40features/km 2 locally.Thenorthernofthetwodensestregionscontains thesecondlongestknowngypsumcaveinNorthAmerica, ParksRanchCave,andislargelyincludedwithinaBLM criticalresourceareathatdoesnotallowsurfaceoccupancy, thusprotectingtheextensivekarstdevelopmentwithinthis area.However,theseconddensekarstregionshouldbe evaluatedthroughmoreintensefieldstudiestodetermineif itshouldalsobeprotectedasacriticalresourcearea. GIS-basedanalyseshavebecomeanimportanttoolfor karststudies.DOQanalysis,coupledwithfieldstudies,has beenshowntobethemosteffectivemethodfordelineating theactualextentandintensityofkarstdevelopmentwithin theCastileoutcroparea,becauseofthesparsevegetation associatedwiththesemi-aridsouthwesternUnitedStates. However,thismaynotbethemosteffectivetechniquein otherregionswherevegetationisdenser.Although commonlyusedinmanykarstregions,DRGanalysis withinthestudyareaprovedtopoorlyrepresenttheactual extentofkarstdevelopmentwithintheregionbecauseof thelowresolutionofcontourintervals,includingsignificantlyunderestimatingtheactualabundanceofkarst featureswithinthetwodensestregions.DEManalysis provedtobeoflittleusewithinthestudyarea,because apparentvariabilityinoriginaldatafromwhichtheDEMs wereconstructeddoesnotconsistentlyrepresentthesame resolution. AlthoughDEMandDRGanalysesprovedineffective inthestudyarea,itislikelythatthesemethodologiescould Figure8.Comparativegraphsoftheresultsfromvarious methodologiesusedtoevaluatekarstdevelopmentwithinthe Castileoutcropregion.NotethatTotalEstimaterefersto the10%additionalkarstfeaturesexpectedbasedonfield testsofsmallertransectsurveylinespacing.A)Cumulative methodologyresultsfromthe50,1-km 2 sitesthatwere physicallymappedduringfieldstudies.B)Cumulativekarst featuresfortheentireCastileoutcropregionbasedon differentmethodologies,whereDOQCorrectedrepresents theweightingDOQ-definedfeaturesbyafactorof2.77 basedontheratiooftruefeaturesdocumentedduringfield mappingwiththoseidentifiedthroughDOQanalysis. K.W.Stafford,L.Rosales-Lagarde,andP.J.Boston JournalofCaveandKarstStudies, April2008 N 43

PAGE 10

Figure9.KarstpotentialmapoftheCastileFormationoutcropregiondefi nedinthisstudy.Notethetwodenseareasofkarst developmentwithinthenorthernportionofthestudyareawithdensitiesg reaterthan40features/km 2 C ASTILEEVAPORITEKARSTPOTENTIALMAPOFTHE G YPSUM P LAIN ,E DDY C OUNTY ,N EW M EXICOAND C ULBERSON C OUNTY ,T EXAS :AGIS METHODOLOGICALCOMPARISON 44 N JournalofCaveandKarstStudies, April2008

PAGE 11

beeffectivefordelineatingkarstdevelopmentinother regionswherehigherresolutionDEMorDRGdatais available.Ultimately,thescaleofkarstfeatureswithin regionsbeingevaluatedwithGISmethodologiesmustbe comparedwiththeresolutionofavailableGISdata,in ordertodeterminetheeffectivenessofGIS-basedstudies. Therefore,cautionmustbetakenwhenconductingGISbasedkarstanalyses,whichshouldalwaysbecoupledwith fieldstudiesforverification,notonlyindenselykarsted areas,butalsoinregionsthatappeartohaveminimalkarst development. Thedistinctclusteringpatternofkarstprovidessome insightintothenatureofspeleogenesiswithintheregion (Figs.7,9).Klimchouk(2003)andFrumkinandFischhendler(2005)suggestthathypogenickarsttendstoform indenseclustersseparatedbyregionsofminimalkarst developmentbecauseheterogeneitieswithinsolublestrata promotetransversespeleogenesisinregionswhererising fluidsbecomefocusedalongfavorableflowpaths.In contrast,epigenickarstisgenerallyexpressedasmore widelydistributedfeatureswheredescendingmeteoric watersattempttoutilizeallavailableirregularitiesnear thesurfaceandconvergewithdepth.Becauseconvergence occurswithdepthinepigenickarst,surficialexpressions tendtobelessclusteredinepigenicdominatedkarstas opposedtohypogenickarstwhereconvergenceoccursnear thesurface. CurrentstudiesofkarstdevelopmentwithintheCastile Formationbytheauthorshavefoundsignificantmorphologicalevidencewithinindividualcavesthatsupportsthe interpretationofspeleogenesisdominatedbyhypogene processes.Theseincludethediagnosticsuiteofhypogenic features(e.g.risers,channelsandcupolas)reportedby Klimchouk(2007),aswellasthewidespreadoccurrenceof blanketbreccias(Andersonetal.,1978),brecciapipes (AndersonandKirkland,1980),evaporitecalcitization (KirklandandEvans,1976)andnativesulfurdeposits (HentzandHenry,1989)previouslyreportedwithinthe region.Currentresearchisfocusingoninterpretingthe speleogeneticevolutionoftheCastileFormation,including thediageneticalterationofcalciumsulfaterocksandthe developmentofcavernousporosity.However,thisis beyondthescopeofthismanuscriptandwillbereported separatelyinthenear-future.WhileGIS-basedanalyses provideinsightintothespeleogeneticprocessesofthe region,detailedfieldstudiesofspecificfeatureswillbe requiredinordertointerpretthespeleogeneticevolutionof theregion. WhilethekarstpotentialmapoftheCastileFormation outcropregionalonecanonlyprovidelimitedinsightinto regionalspeleogenesis,itcanprovideaneffectivetoolfor landmanagementwithinEddyCounty,NewMexicoand CulbersonCounty,Texas.Delineationofkarstintense regionscanbeusedinlandmanagementplanningforroad constructionandoilfieldwellandpipelineplacements,in ordertonotonlyavoidregionsofpotentialgeohazards associatedwithcollapse,butalsotoprotectregionsof significantground-waterrecharge.WhetherCastilekarstis primarilytheresultofhypogenicorepigenicspeleogenesis, mostexposedfeaturescurrentlyactasground-water rechargefeatures,thusthedelineationofdensekarstregions iscrucialforthesustainedmanagementofsparsewater resourceswithinthisportionofthesemi-aridsouthwest. Ultimately,karstpotentialmapscanbeusedtodelineate sensitiveregionsforkarstresourcemanagement. A CKNOWLEDGEMENTS TheauthorsareindebtedtoJackBlake,Draper Brantley,StanleyJobe,LaneSumnerandClayTaylor fortheirgenerousaccesstoprivateranchesinTexas throughoutthisstudy.TimHuntprovidedusefulinformationandassistancewithUniversityLandinTexas. JonJasperandJimGoodbarprovidedessentialinformationaboutknowngypsumkarstdevelopmentwithinNew Mexico.JimKennedyprovidedessentialinformation aboutknowngypsumkarstdevelopmentwithinTexas. Theauthorsarethankfulfortheusefulcommentsprovided byananonymousreviewerandAmosFrumkinwhich helpedtoimprovethismanuscript.Thisresearchwas partiallyfundedthroughgrantsfromtheNewMexico GeologicalSocietyandtheNewMexicoTechGraduate StudentAssociationandsupportfromtheNationalCave andKarstResearchInstitute(NCKRI). R EFERENCES Anderson,R.Y.,Dean,W.E.,Kirkland,D.W.,andSnider,H.I.,1972, PermianCastilevarvedevaporitesequence,WestTexasandNew Mexico:GeologicalSocietyofAmericaBulletin,v.83,p.59–85. Anderson,R.Y.,Kietzke,K.K.,andRhodes,D.J.,1978,Developmentof dissolutionbreccias,northernDelawareBasinandadjacentareas: NewMexicoBureauofMinesandMineralResourcesBulletin159, p.47–52. Anderson,R.Y.,andKirkland,D.W.,1980,Dissolutionofsaltdeposits bybrinedensityflow:Geology,v.8,p.66–69. Belski,D.,ed.,1992,GYPKAPReportVolume 2:SouthwesternRegion oftheNationalSpeleologicalSociety,Albuquerque,N.M.,57p. Denizman,C.,2003,Morphometricandspatialdistributionparametersof karsticdepressions,lowerSuwanneeRiverbasin,Florida:Journalof CaveandKarstStudies,v.65,no.1,p.29–35. Dietrich,J.W.,Owen,D.E.,Shelby,C.A.,andBarnes,V.E.,1995, GeologicatlasofTexas:VanHorn-ElPasoSheet:UniversityofTexas BureauofEconomicGeology,Austin,Texas, Eaton,J.,ed.,1987,GYPKAP1987AnnualReport:SouthwesternRegion oftheNationalSpeleologicalSociety,Alamogordo,N.M.,35p. Florea,L.J.,Paylor,R.L.,Simpson,L.,andGulley,J.,2002,KarstGIS advancesinKentucky:JournalofCaveandKarstStudies,v.64, no.1,p.58–62. Ford,D.C.,andWilliams,P.W.,1989,KarstGeomorphologyand Hydrology:London,UnwinHymam,601p. Frumkin,A.,andFischhendler,I.,2005,Morphometryanddistribution ofisolatedcavesasaguideforphreaticandconfinedpaleohydrologicalconditions:Geomorphology,v.67,p.457–471. Hentz,T.F.,andHenry,C.D.,1989,Evaporite-hostednativesulfurin Trans-PecosTexas:Relationtolate-phaseBasinandRangedeformation:Geology,v.17,p.400–403. Hill,C.A.,1996,GeologyoftheDelawareBasin,Guadalupe,Apacheand GlassMountains:NewMexicoandWestTexas:PermianBasin Section:Midland,Texas,SEPM,480p. K.W.Stafford,L.Rosales-Lagarde,andP.J.Boston JournalofCaveandKarstStudies, April2008 N 45

PAGE 12

Horak,R.L.,1985,Trans-PecostectonismanditsaffectsonthePermian Basin, in Dickerson,P.W.,andMuelberger,W.R.,eds.,Structureand TectonicsofTrans-PecosTexas:Midland,Texas,WestTexasGeologicalSociety,p.81–87. Hose,L.D.,andPisarowicz,J.A.,eds.,2000,TheCavesoftheGuadalupe Mountains:JournalofCaveandKarstStudies,v.62,no.2,157p. Hung,L.Q.,Dinh,N.Q.,Batelaan,O.,Tam,V.T.,andLagrou,D.,2002, RemotesensingandGIS-basedanalysisofcavedevelopmentinthe SuoimuoiCatchment(SonLa–NWVietnam):JournalofCaveand KarstStudies,v.64,no.1,p.23–33. Kelley,V.C.,1971,GeologyofthePecosCountry,SoutheasternNew Mexico:NewMexicoBureauofMinesandMineralResources,78p. Kendall,A.C.,andHarwood,G.M.,1989,Shallow-watergypsuminthe CastileFormation–significanceandimplications, in Harris,P.M., andGrover,G.A.,eds.,SubsurfaceandOutcropExaminationofthe CapitanShelfMargin,NorthernDelawareBasin:SEPM,Core WorkshopNo.13,SanAntonio,Texas,p.451–457. Kirkland,D.W.,andEvans,R.,1976,Originoflimestonebuttes,Gypsum Plain,CulbersonCounty,Texas:AmericanAssociationofPetroleum GeologistsBulletin,v.60,p.2005–2018. Klimchouk,A.,1996,Dissolutionandconversionofgypsumand anhydrite:InternationalJournalofSpeleology,v.25,no.3–4, p.263–274. Klimchouk,A.,2000a,Speleogenesisingypsum, in Klimchouk,A.,Ford, D.C.,Palmer,A.N.,andDreybrodt,W.,eds.,Speleogenesis: EvolutionofKarstAquifers:Huntsville,NationalSpeleological Society,Inc.,p.261–273. Klimchouk,A.,2000b,Speleogenesisunderdeep-seatedandconfined conditions, in Klimchouk,A.,Ford,D.C.,Palmer,A.N.,and Dreybrodt,W.,eds.,Speleogenesis:EvolutionofKarstAquifers: Huntsville,NationalSpeleologicalSociety,Inc.,p.244–260. Klimchouk,A.,2003,Conceptualizationofspeleogenesisinmulti-story artesiansystems:amodeloftransversespeleogenesis:Speleogenesis andEvolutionofKarstAquifers,v.1,no.2,p.1–18. Klimchouk,A.,2007,HypogeneSpeleogenesis:Hydrogeologicaland MorphometricPerspective:Carlsbad,NationalCaveandKarst ResearchInstitute,SpecialPaperNo.1,106p. Lee,J.,ed.,1996,GYPKAPReportVolume3:SouthwesternRegionof theNationalSpeleologicalSociety,69p. Machel,H.G.,andBurton,E.A.,1991,Burial-diageneticsabkha-like gypsumandanhydritenodules:JournalofSedimentaryPetrology, v.61,no.3,p.394–405. Reddell,J.R.,andFieseler,R.G.,1977,TheCavesofFarWestTexas: Austin,TexasSpeleologicalSurvey,103p. Sares,S.W.,1984,Hydrologicandgeomorphicdevelopmentofalowrelief evaporitekarstdrainagebasin,southeasternNewMexico[M.S. Thesis]:Albuquerque,UniversityofNewMexico,123p. Stafford,K.W.,2006,GypsumkarstoftheChosaDrawarea, in Land,L., Lueth,V.W.,Raatz,W.,Boston,P.,andLove,D.,eds.,Cavesand KarstofSoutheasternNewMexico:Socorro,NewMexicoGeological SocietyFifty-seventhAnnualFieldConference,NewMexicoGeologicalSociety,p.82–83. Szukalski,B.W.,Hose,L.D.,andPisarowicz,J.A.,eds.,2002,Caveand karstGIS:JournalofCaveandKarstStudies,v.64,no.1,93p. Taylor,C.J.,Nelson,H.L.,Hileman,G.,andKaiser,W.P.,2005, Hydrogeologic-frameworkmappingofshallow,conduit-dominated karst—componentsofaregionalGIS-basedapproach, in U.S. GeologicalSurveyKarstInterestGroupProceedings,RapidCity, SouthDakota,U.S.GeologicalSurveyScientificInvestigations Report2005–5160,p.103–113. Veni,G.,2002,RevisingthekarstmapoftheUnitedStates:Journalof CaveandKarstStudies,v.64,no.1,p.45–50. C ASTILEEVAPORITEKARSTPOTENTIALMAPOFTHE G YPSUM P LAIN ,E DDY C OUNTY ,N EW M EXICOAND C ULBERSON C OUNTY ,T EXAS :AGIS METHODOLOGICALCOMPARISON 46 N JournalofCaveandKarstStudies, April2008



PAGE 1

CHARACTERIZATIONOFCAVEAEROPHYTICALGAL COMMUNITIESANDEFFECTSOFIRRADIANCELEVELS ONPRODUCTIONOFPIGMENTS J ANEZ M ULEC* 1 ,G ORAZD K OSI 2 AND D ANIJEL V RHOVS EK 3 Abstract: Aerophyticalgaegrowonvarioussubstrataunderfavourableecological conditions.Intheilluminatedpartsofcaves,whererelativehumidityre aches100%,they colonizesediments,rockysurfaces,andartificialmaterials.Anaeroph yticalgal communityfromthecaveentranceiscomposedalmostexclusivelyofcyanob acteria, incontrasttolampenflorawheregreenalgaebecomemoredominant.Inthel aterstage ofspeciessuccessioninthelampenfloracommunity,cyanobacteriaaremo reabundant andthuscommunitystructurebecomesmoresimilartothecommunityfromth ecave entrance.Absenceofcorrelationbetweenphotonfluxdensityandchlorop hyll a concentrationindicatesthatsubstratumcharacteristicsatthemicrole velnotably influencealgalgrowth.Chl a concentrationpersurfaceunitinthecaseoftheepilithic algaefromthecaveentranceislower(max.1.71 gcm 2 )comparedtothatforthe lampenfloraalgae(max.2.44 gcm 2 ).Atcavetemperatures,thelightsaturationpoint isquicklyreached.At9.0 u Candfrequentlowphotonfluxdensitiesinacaveentrance andaroundlampsinshowcaves,biosynthesisofaccessoryphotosynthetic pigmentsfor twotypicalcaveaerophyticorganisms,cyanobacterium Chroococcusminutus andgreen alga Chlorella sp.,isconsiderablyelevated. I NTRODUCTION Cavesareoneoftheextremeenvironmentsgenerally characterizedbylownutrientinput(Pedersen,2000).Low nutrientinputisalimitingfactorformanygroupsof organisms,althoughsomespecies,likealgae,findthis environmentstillsuitableforcolonizationandgrowth.In caves,algaecanbefoundinbodiesofwater(Kuehnetal., 1992;Sanchezetal.,2002)andaerophytichabitats (Golubic ,1967;Dobat,1970).Incaves,manysurfaces serveforalgalcolonizationincluding:sediments,rocky surfacesandartificialmaterial.Werecentlypublishedthe compositionofalgalcommunitiesfromanothertwo interestingcaveaerophytichabitats,fromstromatolitic stalagmitesandfromstalactites,wheregrowthisenhanced bycarbonatedepositionpromotedbycyanobacteria towardssunlight(Mulecetal.,2007).Developmentof aerophyticvegetationisinfluencedbylight,temperature, highrelativehumidity(reaching100%)and/orseeping water,andsubstratumcharacteristics(Golubic ,1967; Martinc ic etal.,1981;ChangandChang-Schneider, 1991).Aerophyticalgaeareeasilyobservedinthecave entranceilluminatedbydirectorindirectsunlightand,in showcavesequippedwithartificialillumination,asapart ofalampenfloracommunityaroundlamps(Mulec,2005). Severalapproacheshavebeentestedtocontrolgrowthof thisalienlampenfloravegetation(Olson,2006).Illuminatedspotsinagenerallynutrient-poorcaveenvironmentare quicklycolonizedbyaerophyticalgae.Thelargeamountof energyandconsequentbiomassintroducedintothecave ecosystemindirectlyinfluencecavefauna,aswellas affectingthesurvivalandtransportoforganismsentering thecave,eitheractivelyorpassively.Highernutrientinput enablesnewcomerstobemorecompetitiveinthecave environmentthanspecializedtroglomorphicorganisms. Consequently,obligatecave-dwellingorganismsarethreatenedandmaybecomeextirpatedorextinctwithout restorationofpreviousnaturalconditions(Pipan,2005). Cavesaregenerallynotconsideredtobeisolated habitats;however,thereisanexampleofthespatially isolatedMovilecavewheretheexistenceofcomplexanimal communitiesisbasedonlyonbacterialchemolithotrophy (Sarbuetal.,1996;KinkleandKane,2000).Threekey modesoftransportofviablealgalpropagulesintothekarst undergroundcanbedistinguished:aircurrents,waterflow, andintroductionbyanimalsandhumans(Dobat,1970). Theexistenceoflampenfloradeepinshowcavesprovesthe efficienttransportofpropagulesfromsourcesabovecaves, whichismoreorlessconstant.Speciescompositionof aerophyticalgalcommunitiesfromcaveentrancesdiffers comparedtolampenflora.Inilluminatedcaveentrances cyanobacteriaprevail(Palik,1964a;Golubic ,1967;Buczko andRajczy,1989;Vinogradovaetal.,1995,1998;Asencio andAboal,2000).Vinogradovaetal.(1998)established thatlightisakeyfactorthatinfluenceszonationof cyanobacteriainthecaveentrance. 1 KarstResearchInstitute,ScientificResearchCentreoftheSlovenianAc ademyof SciencesandArts,Titovtrg2,SI-6230Postojna,S LOVENIA, janez.mulec@guest. arnes.si 2 NationalInstituteofBiology,Vec napot111,SI-1000Ljubljana,S LOVENIA gorazd.kosi@nib.si 3 LIMNOS,Podlimbarskega31,SI-1000Ljubljana,S LOVENIA ,dani@limnos.si J.Mulec,G.Kosi,andD.Vrhovs ekCharacterizationofcaveaerophyticalgalcommunitiesandeffectso firradiancelevelson productionofpigments. JournalofCaveandKarstStudies, v.70,no.1,p.312. JournalofCaveandKarstStudies, April2008 N 3

PAGE 2

Incaves,newalgalspecieswereidentified(Jones,1964; Palik,1964b;VanLandingham,1966a,b;Sant'Annaetal., 1991;Hernande z-Marine andCanals,1994).Beyond taxonomy,ecophysiologicalstudiesoncavealgaearerare, althoughcavesrepresentanalmostidealnaturallaboratoryforalgologicalstudieswithpracticallyconstantecologicalparameters.Thepurposeofthisstudywastoascertain howdifferentirradiancelevelsinfluencequantityand qualitychangeinthealgalcommunity,howlowirradiancesaffectstheratioofphotosyntheticpigmentsinalgae, andtocomparefloristicanalysisfromthesamecaves conducted22yearsagobyMartinc ic etal.(1981). S TUDY A REA Sixcaves(Postojnskajama,Kostanjevis kajama,Pekel priZalogu,Pivkajama,S kocjanskejame,Z upanovajama) andtwomines(Idrija,Mez ica)werestudiedinthekarst regionofSlovenia(Table1).S kocjanskejameis,duetoits importancefromthespeleologicalandecologicalpointof view,listedintheUNESCOWorldheritagelistandas importantundergroundwetlands(Ramsarconvention). Samplingwasperformedinthesummerof2003.Where lampenfloraweresampledinshowcaves,lampsare periodicallyturnedonduetotouristvisitsormaintenance ofthetouristinfrastructure.OnlyintheMez icaleadand zincminearelampson24hoursduetotheconstant monitoringofundergroundwaterflow.Samplingsitesfor lampenflorawereselectedrandomly.SamplingofaerophyticalgaeinthecaveentranceofS kocjanskejamewas performedinoneoftheentrancesnamedSchmidlova dvorana,whichisacavespaceofhugedimensions,22m wide,25mhighand100mdeep.Thislargeshadyareain thecavemouthisdirectlyilluminatedbysunlightinthe morning.Duetotheorientationandpositionofthecave entrance,intheearlyspringandlateautumn,aneven widerareaisdirectlyilluminated.Growthexperiments werecarriedoutinthepartofPostojnskajamathatisnot opentothepublic. M ATERIALSAND M ETHODS Samplesforfloristicanalysisweretakenfromeight cavesandmines.Priortotakingspecimensphotonflux densityatthesiteswasmeasuredusingaLICORLI 1000 DataLogger(USA).Inthecaveentranceseveralmeasurementsofphotonirradianceweremade;themostrepresentativeoneswereusedforstatisticalanalysis.Weinoculated Jaworskimedium(Warrenetal.,1997)atthesampling sitesusingsterilescrapesofthealgalmat.Inthecaseof lampenflora,ifconfluentgrowthwasobservedarounda lamp,uptosevensamplesweretakenaroundthesame lampatdifferentdistancestoobservedifferencesinthe communitycomposition.Mixedandpurealgalcultures wereisolatedinJaworskiliquidandonsolid1%Jaworski agarmedia.Jaworskimediumisfrequentlyusedin algologyasitsupportsgrowthavarietyofgroupsof algae.Cyanobacteriawereselectivelyisolatedwhenthe mediumwassupplementedwith100 gml 1 ofDCMU (diuron,N-3,4-dichlorophenyl-N -dimetilurea).Cultivationconditionswere:20 u C,8:16light/darkperiodwitha photonfluxdensityof100 molm 2 s 1 forseveral weeks.Cultureswereregularlyscreenedusingamagnifying glassandlightmicroscope(NikonEclipseTE300).Floristic dataobtainedfromculturematerialweresupplemented withmicroscopicdataofthesamefieldmaterialfixedwith 4%formalinsolution.Diatomsampleswereprocessedand identifiedasdescribedbyClescerietal.(1998).Sputtered goldspecimenswerescreenedusingaSEMmicroscope (JSM-840,JEOL,USA).Severalkeysandarticleswere usedtoidentifyalgalspecies:Abdelahad(1985,1989), AsencioandAboal(2000),Coute (1982),EttlandGa rtner (1995),Garbackietal.(1999),Geitler(1932),Golubic (1967),Hindak(1996),Hoffmann(1986),Koma rekand Table1.Locationandcharacterizationofthecavesandminesinthisstudy Cave/mineMunicipality Altitude (m) Length (m) Depth (m) Orientationofcave entranceLithology PostojnskajamaPostojna52919555115NWCretaceouslimestones Kostanjevis kajamaKrs ko170172647SCretaceouslimestonesand dolomites PekelpriZaloguZ alec314115940NETriassiclimestonesanddolomites PivkajamacavePostojna54079477ECretaceouslimestones Leadandzincmine Mez ica Ravnena Koros kem 5003500300SCarnianlimestones,Triassic dolomites,shales Mercurymine Idrija Idrija330100022SPermocarbonianshalesand dolomites S kocjanskejameSez ana4255800250NWCretaceousandPaleogene limestones Z upanovajamaGrosuplje46868270NWJurassiclimestones C HARACTERIZATIONOFCAVEAEROPHYTICALGALCOMMUNITIESANDEFFECTSOFIRRA DIANCELEVELSONPRODUCTIONOFPIGMENTS 4 N JournalofCaveandKarstStudies, April2008

PAGE 3

Anagnostidis(2000,2005),KrammerandLange-Bertalot (1986,1988,1991),Lemmermannetal.(1915),Sulek (1969). Cyanobacterium Chroococcusminutus andgreenalga Chlorella sp.,isolatedinpureculturefromthelampenflora community,whichfrequentlyinhabitaerophytichabitats includingthecaveentrance,wereusedingrowthexperimentsinthatpartofPostojnskajamathatisnotopento thepublic(9.0 u C,RH95%).CulturesinliquidJaworski mediumwerecultivatedintriplicateusingaphotonflux gradientof100,50,20,10,5,2.5and0 molm 2 s 1 with thesamelightingperiod(8:16light/darkperiod)and regularmixed.Theinoculumwas10 4 cellsperml.After 25daysofincubation,cellswerecounted,harvested,and theconcentrationofphotosyntheticpigmentsestablished: Chl a ,Chl b andcarotenoidsfor Chlorella sp.afterWetzel andLikens(1995)andphycocyaninafterLeeetal.,(1994) andChl a afterVollenweideretal.(1974)for C.minutus SitesforascertainmentofChl a levelsofepilithiccave algaewerecarefullyselected.Knownflatrockysurfaces withminorsubstratumirregularitieswithconfluentovergrowthofalgaewerescrapedoffwithanalcoholflame sterilizedpocketknifeandcollectedinatesttube.The concentrationofChl a wasestablishedusingtheprocedure describedbyVollenweideretal.(1974).Ateachsitephoton fluxdensitywasmeasured. R ESULTS Intheaerophyticalgalcommunityfromthecave entranceofS kocjanskejamecyanobacteriaprevailed (69%ofidentifiedtaxa)whileChlorophyta(19%)and Chrysophyta(12%)representedtheminorpartofthe community(Table2).Sampleswerealsotakenatthesame sitestodetermineChl a levelsofepilithicalgae,which rangedfrom0.14to1.71 gcm 2 (Table2).Thereisno correlation(r 0.04,p 0.05)betweenirradianceandChl a aswellasbetweenthenumberofalgaltaxaandChl a concentration(r 0.19,p 0.05).Takingintoaccount onlythecyanobacterialcomponent,weestablishedthat withincreasingirradiancethenumberofcoccoidcyanobacterialowers(r 0.42,p 0.05). Lampenfloracanbeobservedintheimmediatevicinity ofartificiallighting.InSloveniancaveszonationof vegetationisnotobserved.Weidentified60algaltaxain thelampenfloracommunityfromeightshowcaves (Table3).Cyanobacteriawerethemostabundant(47%) followedbyChlorophyta(30%)andChrysophyta(23%). InPekelpriZalogucave,samplesweretakenaround oneselectedlampwithanilluminationgradientand confluentphototrophicgrowthtodetermineChl a levels ofepilithicalgae.Concentrationsrangedfrom0.57to 2.45 gcm 2 (datanotshown).Asinthecaseof aerophtyticalgaefromSchmidlovadvorana,lampenflora showednocorrelationbetweenirradianceandChl a levels (r 0.19p 0.05). Irradiancesusedinourgrowthexperimentweresimilar tothemeasurementsofirradianceexperiencedincaves aroundlampsandintheshadypartsofcaveentrances. FromFigure1,itisevidentthatincyanobacterium C. minutus theconcentrationofChl a increaseduptoa photonfluxdensityof50 molm 2 s 1 andat 100 molm 2 s 1 aslightdeclineisobserved.Concentrationofphycocyaninatlowphotonfluxdensities(e.g.,2.5, 5and10 molm 2 s 1 )werehighercomparedto20and 50 molm 2 s 1 .Kirk(1983)experiencedasimilareffect thatatlowirradiancetheratioofbiliproteinstoChl a increases.Atthehighestphotonfluxdensityof 100 molm 2 s 1 phycocyaninconcentrationelevateda littlemore.Withgreenalga Chlorella sp.atlowphotonflux densities(e.g.,2.5,5and10 molm 2 s 1 )theconcentrationofaccessoryphotosyntheticpigmentsisalso elevated(Fig.2).Bygraduallyincreasingfrom20to 100 molm 2 s 1 ,themolarratioofaccessorypigments (i.e.,Chl b andcarotenoids)wasloweredinfavourofChl a .At100 molm 2 s 1 theratioofChl a vs.Chl b was 3:1,whichistypicalforgreenalgae(Kirk,1983). D ISCUSSION Algaefrequentlygrowintheilluminatedpartsofcaves. Incaves,twodistinctaerophyticmicrohabitatscolonized byepilithicalgaecanbedistinguished:(1)caveentrances illuminatedbysunlightand(2)areasaroundlampsinshow cavesandmines.Cyanobacteriaprevailinthealgal communityfromcaveentrances.Theycancolonizeinto thedeepestpartsofthecaveentrancewherebiodiversityof phototrophicorganismsisthelowest(Vinogradovaetal., 1998).Manyoftheidentifiedalgaefrompoorlyilluminatedcaveenvironmentscannotbeconsideredtypicalcave species(Hoffmann,2002),althoughthepresenceofsome ofthem(e.g., Pediastrumboryanum ,Table2)indicatesnot onlyefficienttransportbutalsotheexistenceofasuitable nicheinwaterdropletsfornon-aerophyticalgae. Asreportedinotherpapers,thelampenforacommunity showslowerbiodiversitycomparedtothealgaefromcave entrances.In1981Martinc ic etal.publishedafloristic analysisoflampenflorafromsixSlovenianshowcaves: C rnajama,jamaPekelpriZalogu,Pivkajama,Postojnska jama,S kocjanskejameandTaborskajama(nowcalled Z upanovajama)wheretheyidentifiedatotalof44algal species,withthehighestportionofcyanobacteria.Comparingtheseresultsandtheresultsofthepresentstudy20 yearslater,wedidnotobserveanymajordifferenceinthe compositionofthelampenfloracommunity.Thegreenalga Trentenpohliaaurea appearswiththehighestfrequencyin Slovenianshowcaves. Thekeyquestionherewouldbe:Whatisspecies successionlikeinthecaseoflampenflora?Although cyanobacteriaarethemostadaptablephototrophsunder stressedconditions,inmicrohabitatswithlessenvironmentalstress,likeilluminatedspotsaroundlamps,theyare J.M ULEC ,G.K OSI AND D.V RHOVS EK JournalofCaveandKarstStudies, April2008 N 5

PAGE 4

Table2.CompositionofalgalcommunitywithregardtoconcentrationsofC hl a andphotosyntheticactiveradiationatsamplingsitesintheSchmidlovad vorana caveentrancefromS kocjanskejame. Chl a ( gcm 2 ) Species\Photosyntheticactiveradiation( molm 2 s 1 ) Samplingsite 0.140.200.490.560.940.610.371.711.340.540.510.930.370.320.570.1 40.950.811.870.930.210.50 0.300.270.280.240.210.310.110.900.100.800.400.700.200.060.570.4 50.510.480.110.750.411.96 PROKARYOTA CYANOPHYTA Aphanocapsa sp.Na geli ....... .. .... .... Aphanocapsamuscicola (Meneghini)Wille .. .. Aphanothececastagnei (Bre bisson)Rabenhorst ... ........ ......... Chondrocystisdermochroa (Na geli)Koma reket Anagnostidis ..... ........ .. .. Chroococcus sp.Na geli ................... Chroococcushelveticus Na geli ................ .... Chroococcuslithophilus Ercegovic ................. ... Chroococcusminutus (Ku tzing)Na geli ................. .... Chroococcusmontanus Hansgirg ................. .... Chroococcusvarius A.BrauninRabenhorst ... .................. Chroococcusturgidus (Ku tzing)Na geli ................. .... Cyanotheceaeruginosa (Na geli)Koma rek ... ... . .......... Geitleriacalcarea Friedmann .. .. . ... ... Gloeocapsa sp.Ku tzing ..... ........ .... .. Gloeocapsaaeruginosa Ku tzing ..... ........ ... Gloeocapsaatrata Ku tzing ..... ...... ... . Gloeocapsakuetzigiana Na geli .. .............. .... Gloeocapsarupestris Ku tzing ................. .. Gloeocapsopsis sp.GeitlerexKoma rek ................. Gloeocapsopsispleurocapsoides (Nova c ek)Koma reket Anagnostidis ..... .... .......... Gloeothecepalea (Ku tzing)Rabenhorst ................. .. Gloeothecerupestris (Lyngbye)Bornet,inWittrocket Nordstedt .............. .. .... Leptolyngbyafoveolarum (RabenhorstexGomont) AnagnostidisetKoma rek ................... Leptolyngbyafragilis (Gomont)AnagnostidisetKoma rek .... . . ... Leptolyngbyagracillima (ZopfexHansgirg)Anagnostidiset Koma rek .............. ....... Leptolyngbyaperelegans (Lemmermann)Anagnostidiset Koma rek ................... Leptolyngbyaschmidlei (Limanowska)Anagnostidiset Koma rek .............. ....... Leptolyngbyatenuis (Gomont)AnagnostidisetKoma rek .................... Lyngbya sp.C.AgardhexGomont .. Lyngbyaattenuata Fritsch ... .................. Nostocminutum DesmazexBorn ................... .. Oscillatoria sp.VaucherexGomont .......... ... ... Oscillatoriasubbrevis Schmidle ... .... ............. Phormidiuminundatum Ku tzingexGomont ... ..... ............ C HARACTERIZATIONOFCAVEAEROPHYTICALGALCOMMUNITIESANDEFFECTSOFIRRA DIANCELEVELSONPRODUCTIONOFPIGMENTS 6 N JournalofCaveandKarstStudies, April2008

PAGE 5

Chl a ( gcm 2 ) Species\Photosyntheticactiveradiation( molm 2 s 1 ) Samplingsite 0.140.200.490.560.940.610.371.711.340.540.510.930.370.320.570.1 40.950.811.870.930.210.50 0.300.270.280.240.210.310.110.900.100.800.400.700.200.060.570.4 50.510.480.110.750.411.96 Planktolyngbyalimnetica (Lemmermann)Koma rkova Legnerova etCronberg ......... ...... Plectonema sp.ThuretexGomont .............. ... Pseudophormidiumtenue (ThuretexGomont)Anagnostidis etKoma rek ..................... Scytonema sp.Agardh ................. ... Scytonemahofmanni Agardh ................ ..... Synechococcuselongatus Na geli ................ ..... Synechocystis sp.Sauvageau . ..... ... EUKARYOTA CHRYSOPHYTA Eunotiaargus (Ehrenberg)Ku tzing ................. .... Cymbellaehrenbergii Ku tzing ............... ...... Navicula sp.Bory ............... .... Naviculacontenta var. biceps (Arnott,GrunowinVan Heurck)Cleve .............. .. Naviculagallica var. perpusilla (Grun)Lange-Bertalot .............. ....... Naviculamutica Ku tzing .............. ...... Naviculastroemii Hustedt ................ ..... CHLOROPHYTA Chlorella sp.Beijerinck ... ..... .. . Chlorosarcina sp.Gerneck .............. ....... Gloeocystispolydermatica (Ku tzing)Hinda k .................. ... Klebsormidiumflaccidum Silva,MattoxetBlackwell .................... Muriella sp.J.B.Petersen ................ ..... Pediastrumboryanum (Turpin)Meneghini .. ...... .... . Pleurococcus sp.Meneghini ..................... Scenedesmus sp.Meyen ..................... Scenedesmusbijugatus (Turpin)Meneghini .............. ....... Stichococcusbacillaris Na geli ..... ... . ... Trentepohliaaurea Martius .............. ... Table2.Continued. J.M ULEC ,G.K OSI AND D.V RHOVS EK JournalofCaveandKarstStudies, April2008 N 7

PAGE 6

Table3.Speciescompositionofthelampenfloraalgaefrom8showcavesand mines. Species Cave/Mine IdKoMePePiPoS kZ u PROKARYOTA Cyanophyta Aphanocapsabiformis A.BrowninRabenhorst .. ... Aphanocapsafusco-lutea Hansgirg ...... Aphanocapsamuscicola (Meneghini)Wille Aphanocapsaparietina Na geli ... .... Chondrocystisdermochroa (Na geli)Koma reketAnagnostidis .... Chroococcuslithophilus Ercegovic ...... Chroococcusminutus (Ku tzing)Na geli ... Chroococcusschizodermaticus W.West ...... Chroococcusvarius A.BrauninRabehorst .. ..... Chroococcuswestii (W.West)Boye-Petersen ... .. Gloeocapsa sp.Ku tzing .... Gloeocapsaatrata Ku tzing ..... Gloeocapsabituminosa (Bory)Ku tzing ...... Gloeocapsapunctata Na geli ..... .. Gloeocapsarupestris Ku tzing ...... Leptolyngbyafoveolarum (RabenhorstexGomont)AnagnostidisetKoma rek ...... Leptolyngbyafragilis (Gomont)AnagnostidisetKoma rek ... .. Leptolyngbyaperelegans (Lemmermann)AnagnostidisetKoma rek ..... .. Leptolyngbyascytonemicola (Gardner)AnagnostidisetKoma rek ... .... Lyngbya sp.C.AgardhexGomont . Oscillatoria sp .VaucherexGomont... .... Planktolyngbyabipunctata (Lemmermann)AnagnostidisetKoma rek ..... .. Planktolyngbyalimnetica (Lemmermann)Koma rkova -Legnerova etCronberg .... ... Plectonema cf. puteale Hansgirg ... . Pseudoanabaenacatenata Lauterborn ...... Pseudocapsa sp.Ercegovic ...... Scytonemahofmanni Agardh ... .... Synechocystis sp.Sauvageau EUKARYOTA Chrysophyta Chlorocloster sp.Pascher . Cymbellaehrenbergii Ku tzing ..... .. Elipsoidon sp.Pascher ...... Elipsoidonoocystoides Pascher ....... Fragilariapinnata Ehrenberg .. ... Heterococcus sp.Chodat ...... C HARACTERIZATIONOFCAVEAEROPHYTICALGALCOMMUNITIESANDEFFECTSOFIRRA DIANCELEVELSONPRODUCTIONOFPIGMENTS 8 N JournalofCaveandKarstStudies, April2008

PAGE 7

Species Cave/Mine IdKoMePePiPoS kZ u Heterococcusfurcatus Pitschmann ...... Monodus sp.Chodat .. ... Navicula sp.Bory .... Naviculacontenta var. biceps (Arnott,GrunowinVanHeurck)Cleve .. Naviculagallica var. perpusilla (Grun)Lange-Bertalot .. Naviculamutica Ku tzing . Nitzschia sp.Hassall .. ..... Pinnulariaborealis Ehrenberg ...... Chlorophyta Apatococcus cf. lobatus (Chodat)J.B.Petersen ....... Chlorella sp.Beijerinck Chlorotetraedron sp.MacEntee ....... Gloeocystispolydermatica (Ku tz.)Hinda k ... .... Klebsormidiumflaccidum Silva,MattoxetBlackwell ..... Microthamnion cf. strictissimum Rabenhorst .. ..... Muriella sp.J.BPetersen .. .... Myrmecia sp.Printz ....... Pediastrumboryanum (Turpin)Meneghini .. .. .. Pseudochlorella sp.Lund ...... Pseudoclonium cf. basiliense Vischer .... Scenedesmusbijugatus (Turpin)Meneghini .... ... Scenedesmusobliquus (Turpin)Ku tzing .... ... Scotiellopsis sp.Vinatzer .... Stichococcusbacillaris Na geli Stichococcusexiguous Gerneck .. ..... Stichococcusundulatus Vinatzer ....... Trentepohliaaurea Martius Note: IdmercurymineIdrija,KoKostanjevis kajama,MeleadandzincmineMez ica,PePekelpriZalogu,PiPivkajama,PoPostojnskajama,S kS kocjanskejame,Z uZ upanovajama Table3.Continued. J.M ULEC ,G.K OSI AND D.V RHOVS EK JournalofCaveandKarstStudies, April2008 N 9

PAGE 8

overgrownbyfast-growingeukaryoticalgae.Thisobservationissupportedbya10-timeshigherincreaseincell countof Chlorella sp.after25daysofcultivationunder caveconditionscomparedto C.minutus (datanotshown), whichissimilartothefindingsofGaylardeandGaylarde (2000)whoconcludedthatthefirstcolonizersonthewalls ofbuildingsareeukaryoticalgae,butcyanobacteria becomepredominantlaterinthespeciessuccession.This laterstagedoesnotdevelopifthebiofilmisfrequently removedfromthebuildings(GaylardeandGaylarde, 2000).Thebestsimilarexamplecanbeseeninthecaseof thelampenfloracommunityfromS kocjanskejamewith oneofthehighestpercentageofcyanobacteria(57%)we found,whichcouldbeduetothefactthatlampenflora havenotbeenremovedsincetheestablishmentofelectric illuminationin1959(Mulec,2005).Neverthelesslampenfloraalgaeareusuallyubiquitous,fastreproducing,and adaptablesoilalgae(Rajczy,1989).Althoughsome authorsreportlampenfloracommunitieshavingmore eukaryoticalgaethancyanobacteria(Faimonetal., 2003),itcouldbethattheysampledlampenflorainthe earlystageofspeciessuccession. Weidentified59algaltaxafromtheSchmidlova dvoranacaveentranceofS kocjanskejame.Ahigher numberofidentifiedtaxabelongingtoOscillatoriales comparedtoNostocalesconfirmedthatthiscaveenvironmentisexposedonlytolowphotonfluxdensity(Table2). NamelycyanobacteriafromtheorderOscillatorialesare betteradaptedtoconstantlowirradiancelevelsthan Nostocales(Albertanoetal.,2000).Ifwetakeintoaccount thatinthecaveentrancerelativeairhumidityisconstant andonlythephotonirradiancelowersthedeeperwego intothecave,acorrelationbetweenirradianceandChl a concentrationisexpected.Ourresultsshowedthereisno correlationthatindicatesasubstratumand/orpresenceof seepingwaterwithavailablenutrientsonsitesinkarst cavesplayanimportantroleinthecolonizationand growthofaerophyticalgae.Nocorrelationbetween irradianceandbiomass(Chl a )wasobservedasinthe caseoflampenflora.However,thereisadifferenceinChl a concentrationpersurfaceunitbetweenthesetwomicrohabitats.Lampenfloraalgaedemonstratedhighervalues (max.2.44 gcm 2 )comparedtothealgaefromcave entrance(max.1.71 gcm 2 ).Thisdifferencecanbe explainedduetothedifferentlightregimeinboth microhabitats(i.e.,changinglightqualityandirradiance levelsduringthedayinthecaveentrance),differentperiods ofillumination,different insitu moisturelevels,and differentspeciescomposition.Generallyspeaking,areas aroundlampshavemorestableconditionssincetheyare deeperintotheconstantzoneofthecave.However,these valuesareafewmagnitudeslowerwhencomparedwiththe concentrationsfromnon-caveenvironmentslikethe Niagaracliffwhereitwas7.3 gcm 2 (Matthes-Searset al.,1997). Resultspresentedhereindicatethatthelightisnotthe onlykeyfactorforhighorlowalgalbiodiversity.Our resultsfurthershowthat,ifwetakeintoaccountonlythe cyanobacterialpartofthecommunity,coccoidforms toleratelowirradiancemoreeasily;andthus,they representthemajorpartofthecommunity.Aerophytic algaemustdevelopanintimateinteractionwithsubstratum onwhichtheyareattached.Whichcomponentofthe limestonesubstratumdirectsorlimitsalgalcolonization? Incarbonaterocks,severaldifferenttraceelementscanbe foundinsignificantconcentrations:Al,Ba,Cd,Co,Cu, Fe,K,Mg,Mn,Na,Si,Sr,Ti,UandZn(Morseand MacKenzie,1990).Someoftheseelementscouldalsobe lethalforalgaeorcertaingroupsofalgae.Incaves,algae oftencolonizeflowstonesurfaces.Variousmineralscanbe tracedinflowstoneinwhichCo,Cr,Cu,Fe,Mg,Mn,Ni andZnarethemostfrequentelementsdepositedtogether withtheflowstone(HillandForti,1997).Tomasellietal. (2000)determinedthatsomealgaehaveapreferencefora specificsubstratumtocolonize.Nevertheless,oneshould takeintoaccountthatincavesvarioussubstancesin seepinganddrippingwatercannotablyinfluencealgal growth. Figure1.Chl a andphycocyaninconcentrationsatdifferent photonfluxdensitiesafter25daysofcultivationof Chroococcusminutus inPostojnskajama. Figure2.ConcentrationsoftheChl a ,Chl b andcarotenoidsatdifferentphotonfluxdensitiesafter25daysof cultivationof Chlorella sp.inPostojnskajama. C HARACTERIZATIONOFCAVEAEROPHYTICALGALCOMMUNITIESANDEFFECTSOFIRRA DIANCELEVELSONPRODUCTIONOFPIGMENTS 10 N JournalofCaveandKarstStudies, April2008

PAGE 9

Physiologicaladaptationsofalgaeincaveconditions arenotwellstudied.Whenlightislimitingforalgalgrowth theChl a contentincreasesanditdecreaseswhenlightis notlimiting(Meeks,1974).Forcyanobacterium C.minutus atcavetemperature(9.0 u C),thelightsaturationlies between50and100 molm 2 s 1 .Asimilartrendis observedforgreenalga Chlorella sp.atphotonflux densitiesslightlyhigherthan100 molm 2 s 1 .Based onourexperiencesfromcaves,themajorityofalgaegrow generallyatphoton-fluxdensitymuchlowerthanthe saturationvalueandevenlowerthanthecompensation point.Diatomsandcyanobacteriahaveanaverage compensationpointbetween5and6andgreenalgaeat 21 molm 2 s 1 (Hill,1996).Algaeincavesmustinclude intheirsurvivalstrategyotherwaystoobtainenergy,like heterotrophy.Giordanoetal.(2000)suggeststhatcells livingincavesatlowphotonirradiancecouldhaveabetter yieldofavailablephotons.Tocaptureasmanyavailable photonsaspossibleatlowirradiance,cellssynthesise accessoryphotosyntheticpigments.At9.0 u Candbelow 10 molm 2 s 1 ,thebiosynthesisofaccessorypigments intwotypicalaerophyticorganisms( C.minutus and Chlorella sp.)waselevated.With Chlorella sp.atvalues higherthan10 molm 2 s 1 ,theratioofcarotenoidsto Chl ab approached1:3,whichisanexpectedratioforgreen algae.At100 molm 2 s 1 ,atypicalratioChl a :Chl b 3:1 forgreenalgaewasobserved(Kirk,1983).Different photonirradiancealsoinfluencestheratioofChl a vs. phycocyaninin C.minutus .Atthehighestirradianceused intheexperiment,theconcentrationofChl a decreased, butphycocyaninslightlyincreased.ZilinskasBraunand ZilinskasBraun(1974)explainedsuchaphenomenonwith loweredefficiencyofenergytransferfromphycocyaninto Chl a .Inthecavehabitat,veryadaptablealgaecanprosper iftheycanuselowphotonirradiancedosages. A CKNOWLEDGEMENTS Thisworkwaspartlysupportedbythefinancial contributionoftheSlovenianResearchAgencyZ6-7072, roleandsignificanceofmicroorganismsinkarstprocesses. R EFERENCES Abdelahad,N.,1985,Osservazionisualcunecianoficeecavernicolerare e interesanti:GiornaleBotanicoItaliano,v.119,no.2,p.4546. Abdelahad,N.,1989,Onfour Myxosarcina -likespecies(Cyanophyta) livingintheInfernigliocave(Italy):Archivfu rHydrobiologie: Supplementband,v.82,Algologicalstudies,no.9,p.313. Albertano,P.,Bruno,L.,D'Ottavi,D.,Moscone,D.,andPalleschi,G., 2000,EffectofphotosynthesisonpHvariationincyanobacterial biofilmsfromRomancatacombs:JournalofAppliedPhycology, no.12,p.279384. Asencio,A.D.,andAboal,M.,2000,AlgaefromLaSerretacave(Murcia, SESpain)andtheirenvironmentalconditions:Archivfu rHydrobiologie:Supplementband,v.131,Algologicalstudies,no.96, p.5978. Buczko ,K.,andRajczy,M.,1989,Contributionstothefloraofthe HungariancavesII.FloraofthethreecavesnearBeremend,Hungary: StudiaBotanicaHungarica(Antea:FragmentaBotanica),v.21, p.1326. Chang,T.P.,andChang-Schneider,H.,1991,Algeninviersu ddeutschen Ho hlen:BerichtederBayerischenBotanischhenGesellschaft,v.62, p.221229. Clesceri,L.S.,Greenberg,A.E.,andEaton,A.D.,1998,Standardmethods fortheexaminationofwaterandwastewater,20 th edition,Washington,AmericanPublicHealthAssociation,p.1027. Coute ,A.,1982,Ultrastructured'unecyanophyce eae riennecalcifie e cavernicole: Geitleriacalcarea Friedmann:Hydrobiologia,v.97, p.255274. Dobat,K.,1970,Conside rationssurlave ge tationcryptogamiquedes grottesduJuraSouabe(sud-ouestdel'Allemagne):Annalesde Spe le ologie,v.25,no.4,p.872907. Ettl,H.,andGa rtner,G.,1995,SyllabusderBoden-,Luft-und Flechtenalgen,Stuttgart,GustavFischerVerlag,p.721. Faimon,J.,S telcl,J.,Kubes ova ,S.,andZima k,J.,2003,Environmentally acceptableeffectofhydrogenperoxideoncavelampflora,calcite speleothemsandlimestones:EnvironmentalPollution,v.122, p.417422. Garbacki,N.,Ector,L.,Kostikov,I.,andHoffmann,L.,1999, Contributiona `l'e tudedelafloredesgrottesdeBelgique:Belgian JournalofBotany,v.132,no.1,p.4376. Gaylarde,P.M.,andGaylarde,C.C.,2000,Algaeandcyanobacteriaon paintedbuildingsinLatinAmerica:InternationalBiodeterioration andBiodegradation,v.46,p.9397. Geitler,L.,1932,Cyanophyceae, in Rabenhorst,L.,ed.,KryptogamenFlora,Akademische,Leipzig,Verlagsgesellschaft,p.1196. Giordano,M.,Mobili,F.,Pezzoni,V.,Hein,M.K.,andDavis,J.S.,2000, PhotosynthesisinthecavesofFrasassi(Italy):Phycologia,v.39, no.5,p.384389. Golubic ,S.,1967,AlgenvegetationderFelsen:Eineo kologische AlgenstudieimdinarischenKarstgebiet, in Elster,H.J.,andOhle, W.,eds.,DieBinnengewa sser,Stuttgart,BandXXIII,E.Schweizerbart'scheVerlagsbuchhandlung,p.183. Hernande z-Marine ,M.,andCanals,T.,1994, Herpyzonemapulverulentum (Mastigocladaceae),anewcavernicolousatmophyticandlimeincrustedcyanophyte:Archivfu rHydrobiologie:Supplementband, v.105,Algologicalstudies,no.75,p.123136. Hill,C.,andForti,P.,1997,Cavemineralsoftheworld,2 nd edition, Huntsville,NationalSpeleologicalSociety,p.463. Hill,W.R.,1996,Effectsoflight, in Stevenson,R.J.,Bothwell,M.L.,and Lowe,R.L.,eds.,Algalecology:freshwaterbenthicecosystems,San Diego,AcademicPress,p.121144. Hindak,F.,1996.Kl'u c naurc ovanienerozkona reny chvla knity ch zeleny chrias(Ulotrichineae,Ulotrichales,Chlorophyceae),Slovenska botanicka spoloc nost'priSAV,Bratislava,p.77. Hoffmann,L.,1986,Cyanophyce esae riennesetsubae riennesduGrandDuche deLuxembourg:BulletinduJardinBotaniqueNationalde Belgique,v.56,p.77127. Hoffmann,L.,2002,Cavesandotherlow-lightenvironments:aerophitic photoautotrophicmicroorganisms, in Bitton,G.,ed.,Encyclopediaof EnvironmentalMicrobiology,NewYork,JohnWiley&Sons, p.835843. Jones,H.J.,1964,AlgologicalinvestigationsinMammothcave,Kentucky : InternationalJournalofSpeleology,v.1,no.4,p.491516. Kinkle,B.K.,andKane,T.C.,2000,Chemolithoautotrophicmicroorganismsandtheirpotentialroleinsubsurfaceenvironments, in Wilkens,H.,Culver,D.C.,andHumphreys,W.,eds.,Ecosystemsof theWorld30.Subterraneanecosystems,Amsterdam,Elsevier, p.309318. Kirk,J.T.O.,1983,Lightandphotosynthesisinaquaticecosystems, Cambridge,CambridgeUniversityPress,p.401. Koma rek,J.,andAnagnostidis,K.,2000,Cyanoprokaryota1.Teil: Chroococcales, in Ettl,H.,Ga rtner,G.,Heying,H.,andMollenhauer, D.,eds.,Su b wasserfloravonMitteleuropa19/1,Heidelberg,Spectrum AkademischerVerlag,p.548. Koma rek,J.,andAnagnostidis,K.,2005,Cyanoprokaryota2.Teil: Oscillatoriales, in Bu del,B.,Ga rtner,G.,Krienitz,L.,andSchagerl,M., eds.,Su sswasserfloravonMitteleuropa19/2,Heidelberg,Elsevier,p.759. Krammer,K.,andLange-Bertalot,H.,1986,Bacillariophyceae1.Teil: Naviculaceae, in Ettl,H.,Gerloff,J.,Heying,H.,andMollenhauer, D.,eds.,Su b wasserfloravonMitteleuropa2/1,Stuttgart,Gustav FischerVerlag,p.876. J.M ULEC ,G.K OSI AND D.V RHOVS EK JournalofCaveandKarstStudies, April2008 N 11

PAGE 10

Krammer,K.,andLange-Bertalot,H.,1988,Bacillariophyceae2.Teil: Bacillariaceae,Epithemiaceae,Surirellaceae, in Ettl,H.,Gerloff,J., Heying,H.,andMollenhauer,D.,eds.,Su b wasserfloravonMitteleuropa2/2,Stuttgart,GustavFischerVerlag,p.596. Krammer,K.,andLange-Bertalot,H.,1991,Bacillariophyceae3.Teil: Centrales,Fragilariaceae,Eunotiaceae, in Ettl,H.,Gerloff,J.,Heying, H.,andMollenhauer,D.,eds.,Su b wasserfloravonMitteleuropa2/3, Stuttgart,GustavFischerVerlag,p.576. Kuehn,K.A.,Oneil,R.M.,andKoehn,R.D.,1992,Viablephotosynthetic microalgalisolatesfromaphoticenvironmentsoftheEdwardsaquifer (centralTexas):Stygologia,v.7,no.3,p.129142. Lee,T.,Tsuzuki,M.,Takeuchi,T.,Yokoyama,K.,andKarube,I.,1994, Invivo fluorimetricmethodfordetectionofcyanobacterialwaterblooms:JournalofAppliedPhycology,v.6,p.489495. Lemmermann,E.,Brunnthaler,J.,andPascher,A.,1915,Heft5: Chlorophyceae2, in Pascher,A.,ed.,DieSu b wasserfloraDeutschlands,O sterreichsundderSchweiz,GustavFischer,Jena,p.250. Martinc ic ,A.,Vrhovs ek,D.,andBatic ,F.,1981,Floravjamahzumetno osvetlitvijo:Biolos kiVestnik,v.29,no.2,p.2756. Matthes-Sears,U.,Gerrath,J.A.,andLarson,D.W.,1997,Abundance, biomass,andproductivityofendolithicandepilithiclowerplantson thetemperatezoneofNiagaraEscarpment,Canada:International JournalofPlantSciences,v.158,p.451460. Meeks,J.C.,1974,Chlorophylls, in Stewart,W.D.P.,ed.,Algalphysiology andbiochemistry,Botanicalmonographs,Volume10,Oxford, Blackwellscientificpublications,p.161175. Morse,J.W.,andMacKenzie,F.T.,1990,Geochemistyofsedimentary carbonates.Developmentsinsedimentology48:Amsterdam,Oxford, Elsevier,p.707. Mulec,J.,2005,AlgaeinthekarstcavesofSlovenia[Ph.D.thesis], Ljubljana,UniversityofLjubljana,p.149. Mulec,J.,Kosi,G.,andVrhovs ek,D.,2007,Algaepromotegrowthof stalagmitesandstalactitesinkarstcaves(S kocjanskejame,Slovenia): CarbonatesandEvaporites,v.22,no.3,p.610. Olson,R.,2006,Controloflampfloraindevelopedcaves, in HildrethWerker,V.,andWerker,J.C.,eds.,Caveconservationand restoration,Huntsville,NationalSpeleologicalSociety,p.343348. Palik,P.,1964a,U berdieAlgenweltderHo hleninUngarn:International JournalofSpeleology,v.1,no.12,p.3544. Palik,P.,1964b,EineneueAulakochlorisArtausderTropfsteinho hlevon Abaliget:InternationalJournalofSpeleology,v.1,no.12,p.2528. Pedersen,K.,2000,Explorationofdeepintraterrestrialmicrobiallife : currentperspective.MiniReview:FEMSMicrobiologyLetters,v.185, p.916. Pipan,T.,2005,Epikarst-apromisinghabitat:copepodfauna,its diversityandecology:acasestudyfromSlovenia(Europe),Ljubljana, ZRCPublishing,p.101. Rajczy,M.,1989.ThefloraofHungariancaves,Karszte sBarlang, Specialissue,p.6972. Sanchez,M.,Alcocer,J.,Escobar,E.,andLugo,A.,2002,Phytoplankton ofcenotesandanchialinecavesalongadistancegradientfromthe northeasterncoastofQuintanaRoo,YucatanPeninsula:Hydrobiologia,v.467,no.13,p.7989. Sant'Ana,C.L.,Branco,L.H.Z.,andSilva,S.M.F.,1991,Anewspeciesof Gloeothece (Cyanophyceae,Microcystaceae)fromSa oPaoloState, Brazil:Archivfu rHydrobiologie:Supplementband,v.89,Algological studies,no.89,p.15. Sarbu,S.M.,Kane,T.C.,andKinkle,B.K.,1996,Achemoautotrophically basedcaveecosystem:Science,v.272,no.5270,p.19531955. Sulek,J.,1969,TaxonomischeU bersichtderGattung Pediastrum Meyen, in Fott,B.,andKoma rek,J.,eds.,StudiesinPhycology,Stuttgart,E. Schweizerbart'scheVerlagsbuchhandlung,p.197261. Tomaselli,L.,Lamenti,G.,Bosco,M.,andTiano,P.,2000,Biodiversity ofphotosyntheticmicro-organismsdwellingonstonemonuments: InternationalBiodeteriorationandBiodegradation,v.46,p.251258. VanLandingham,S.,1966a,Threenewspeciesof Cymbella from Mammothcave,Kentucky:InternationalJournalofSpeleology, v.2,no.12,p.133136. VanLandingham,S.,1966b,Anewspeciesof Gomphonema (Bacillariophyta)fromMammothcave,Kentucky:InternationalJournalof Speleology,v.2,no.4,p.405406. Vinogradova,O.N.,Kovalenko,O.V.,Wasser,S.,Nevo,E.,Tsarenko, P.M.,Stupina,V.V.,andKondratiuk,E.S.,1995,AlgaeoftheMount CarmelNationalPark(Israel):Algologia,v.5,no.2,p.178192. Vinogradova,O.N.,Kovalenko,O.V.,Wasser,S.P.,Nevo,E.,and Weinstein-Evron,M.,1998,Speciesdiversitygradienttodarkness stressinbluegreenalgae/cyanobacteria:amicroscaletestina prehistoriccave,MountCarmel,Israel:IsraelJournalofPlant Sciences,v.46,p.229238. Vollenweider,R.A.,Talling,J.F.,andWestlake,D.F.,1974,Amanualon methodsformeasuringprimaryproductioninaquaticenvironments, IBPHandbookNo.12.Internationalbiologicalprogramme,2 nd edition,Oxford,Blackwellscientificpublications,p.225. Warren,A.,Day,J.G.,andBrown,S.,1997,Cultivationofalgaeand protozoa, in Hurst,C.J.,Knudsen,G.R.,McInerney,M.J.,Stetzenbach,L.D.,andWalter,M.V.,eds.,Manualofenvironmental microbiology,Washington,AmericanSocietyforMicrobiology, p.6171. Wetzel,R.G.,andLikens,G.E.,1995,Limnologicalanalyses,2 nd edition, NewYork,Springer-Verlag,p.139165. ZilinskasBraun,G.,andZilinskasBraun,B.,1974,Lightabsorption, emissionandphotosynthesis, in Stewart,W.D.P.,ed.,Algalphysiologyandbiochemistry,Botanicalmonographs,Volume10,Oxford, Blackwellscientificpublications,p.346390. C HARACTERIZATIONOFCAVEAEROPHYTICALGALCOMMUNITIESANDEFFECTSOFIRRA DIANCELEVELSONPRODUCTIONOFPIGMENTS 12 N JournalofCaveandKarstStudies, April2008



PAGE 1

HYPOGENICSPELEOGENESISWITHINSEVENRIVERS EVAPORITES:COFFEECAVE,EDDYCOUNTY, NEWMEXICO K EVIN W.S TAFFORD 1,2 ,L EWIS L AND 2,3 AND A LEXANDER K LIMCHOUK 3,4 Abstract: CoffeeCave,locatedinthelowerPecosregionofsoutheasternNewMexico, illustratesprocessesofhypogenicspeleogenesisinthemiddlePermianS evenRivers Formation.CoffeeCaveisarectilineargypsummazecavewithatleastfour stratigraphically-distincthorizonsofdevelopment.Morphologicalfe aturesthroughout thecaveprovideunequivocalevidenceofhypogenicascendingspeleogene sisinaconfined aquifersystemdrivenbymixed(forcedandfree)convection.Morphologic featuresin individualcavelevelsincludeacompletesuitethatdefinesoriginalris ingflowpaths, rangingfrominletsforhypogenicfluids(feeders)throughtransitional forms(risingwall channels)toceilinghalf-tubeflowfeaturesandfluidoutlets(cupolasa ndexposed overlyingbeds).Passagemorphologydoesnotsupportoriginsbasedonepi genic processesandlateraldevelopment,althoughthepresenceoffine-graine dsedimentsinthe cavesuggestsminimaloverprintingbybackflooding.Feederdistributio nsshowalateral shiftinascendingfluids,withdecreasingdissolutionaldevelopmentin upperlevels.Itis likelythatadditionalhypogenickarstphenomenaarepresentinthevicin ityofCoffee Cavebecauseregionalhydrologicconditionsareoptimumforconfinedspe leogenesis, withartesiandischargestillactiveintheregion. I NTRODUCTION CoffeeCaveislocatedontheeastsideofthePecos Rivervalley,approximately20kmnorthofCarlsbad,New MexicoatthebaseoftheMcMillanEscarpment.Thecave isformedintheevaporitefaciesbeltoftheSevenRivers Formation(Fig.1).EvaporitekarstdevelopmentisextensivethroughoutthelowerPecosregion,notonlyinthe SevenRiversFormation,butalsoinotherPermian evaporitefaciesoftheArtesiaGroup(includingtheSeven Rivers),andtheYeso,SanAndres,Castile,Saladoand RustlerFormations.Numerousfilledsinkholesandcaves havebeendocumentedwithinarangethatextendsseveral kmeastandwestofthePecosRiver,andfromTexastoas farnorthasSantaRosaineast-centralNewMexico.Most solutionalopeningsarerelativelysmallandnothumanly enterable,butmanyfeaturesareextensivewithcomplex morphologies,suggestiveofmultiplephasesofspeleogenesis.Cavepatternsandabundantdiagnosticmorphologic featuresatmeso-scalewithinindividualcavesappeartobe theresultofhypogenic,largelyconfined,speleogenesis, whilecavesedimentsandminorentrenchmentinsome cavessuggestalaterphaseofunconfineddevelopment. Althoughhypogenicfeaturesareseeninmanycaveswithin theregion,thispaperwillfocusonexamplesfromCoffee Caveinrelationtothecurrentunderstandingofregional hydrologyandspeleogenesisintheSevenRiversFormation. EvaporitekarstdevelopmentwithintheSevenRivers Formation,aswithmostevaporitekarstphenomenainthe UnitedStates,hasnotbeenthoroughlyinvestigated.Most cavedevelopmentwithintheSevenRiversFormationhas onlybeendocumentedinanecdotalreports(Eaton,1987; Belski,1992;Lee,1996),althoughevaporitekarsthasbeen recognizedinassociationwithregionalaquifers(HendricksonandJones,1952)anddamleakagealongthePecos River(Cox,1967).Theoccurrenceoflargegypsum sinkholesatBottomlessLakesStateParknearRoswell dramaticallyillustratestheoccurrenceofartesianspeleogenesiswithintheSevenRiversFormation(Quinlanetal., 1987;Land,2003;2006).Althoughpoorlydocumented withintheUnitedStates,hypogenicevaporitekarsthas beenextensivelystudiedintheWesternUkraine,where largemazecaveshavedevelopedinconfinedconditions andwerelaterbreachedbysurfacedenudationandfluvial entrenchment(e.g.,Klimchouk,1996a;2000a).Other examplesofhypogenicgypsumkarstareknownfrom Germany,Russia,SpainandUnitedKingdom(Klimchouk etal.,1996). G EOLOGIC S ETTING TheSevenRiversFormation,alongwiththeotherfour membersoftheArtesiaGroup,representsthebackreef faciesequivalentoftheCapitanReef,whichdefinedthe shelfmarginoftheDelawareBasinduringmiddlePermian (Guadalupian)time(Figs.2and3).TheSevenRivers 1 DeptofEarthandEnvironmentalScience,NewMexicoInstituteofMiningan d Technology,Socorro,NM87801,USA(kwstafford@juno.com) 2 NationalCaveandKarstResearchInstitute,Carlsbad,NM,88220,USA 3 NewMexicoBureauofGeologyandMineralResources,Socorro,NM87801,USA 4 UkrainianInstituteofSpeleologyandKarstology,Simferopol,95007,Uk raine K.W.Stafford,L.Land,andA.Klimchouk–Hypogenicspeleogenesiswithin SevenRiversEvaporites:CoffeeCave,EddyCounty, NewMexico. JournalofCaveandKarstStudies, v.70,no.1,p.47–61. JournalofCaveandKarstStudies, April2008 N 47

PAGE 2

FormationconsistspredominantlyofdolomiteinitsnearbackreefsettingintheGuadalupeMountains,butbecomes increasinglyevaporiticfurthertothenorthonthe NorthwesternShelf(Fig.2),changingfaciesintointerbeddedgypsumandredmudstone(Scholleetal.,2004). GuadalupianrockswerelaterburiedbyextensivedepositionoflatePermian(Ochoan)evaporitesthatfilledthe basinandsurroundingshelfareas(Fig.3)(Bachman, 1984).BytheendofthePermian,marinesedimentation hadeffectivelyceased(Dickenson,1981).Duringtheearly Triassic,theentireareawasupliftedabovesealevelandthe LaramideOrogenyproducedregionaldeformationlimited touplift(1–2km),tiltingtotheeastandbroadanticlinal flexures(Horak,1985).Bythemid-Tertiary,Laramide compressionhadceasedandshiftedtoBasinandRange extension(ChapinandCather,1994).Asaresultof tectonism,regionaldipofGuadalupianstratainthispart ofsoutheasternNewMexicois 1to2 u totheeastand southeast(Fig.4),withbroadflexuresandabundanthigh anglefracturesandjointsexhibitingminimaloffset.Since thelatePermian,southeasternNewMexicohasbeen dominatedbyfluvialerosion,associatedsedimentation, andkarsticdissolution(Kelley,1971). TheevaporitefaciesoftheSevenRiversFormationis upto150mthickinthestudyarea,withanhydrite (CaSO 4 )andbeddedsalt(NaCl)inthesubsurfaceand gypsum(CaSO 4 2H 2 O)nearthesurfaceasaresultof sulfatehydration(Kelley,1971).Dolomite(CaMg (CaCO 3 ) 2 )interbedsarecommonthroughouttheevaporite facies,forminglaterally-continuouslayersthatthicken towardsthereefandthinawayfromit.Theentiregypsum sequenceiscappedbydolomiteoftheAzoteaTongue Member.SevenRiverssulfatesaregenerallywhitetogrey, nodulartomicrocrystallineanhydrite/gypsum,forming individualbedsrangingfromcentimeterstometersin thickness(Hill,1996). H YDROLOGIC S ETTING CoffeeCaveisformedatthebaseoftheMcMillan Escarpment,whichlocallydefinestheeasternmarginofthe PecosRiverValley(Fig.4).Thecaveislocatedonthe easternshoreofoldLakeMcMillan,anartificialimpoundmentthatformerlystoredwaterfortheCarlsbad IrrigationDistrict(CID).TheoriginalMcMillanDamwas constructedin1893,andthereservoiralmostimmediately beganexperiencingleakageproblemsthroughsinkholes formedinthelakebed.Waterflowedthroughkarstic conduitsintheunderlyingSevenRiversgypsumand returnedtothePecosRiverbydischargefromsprings downstreamfromthelake.Attemptstoisolatetheworst Figure1.RegionalmapdelineatingtheRoswellArtesian Basin,outcropregionoftheSevenRiversevaporitefacies, andlocationofCoffeeCave.SREF = SevenRivers EvaporiteFacies.sre = SevenRiversEmbayment.BLSP = BottomlessLakesStatePark(adaptedfromKelley,1971 andLand,2003). Figure2.Paleogeographicreconstructionofsoutheastern NewMexicoduringthemiddlePermian,showingthe depositionalrelationshipbetweentheDelawareBasinand theNorthwesternShelfwhereSevenRiversevaporitefacies weredeposited(adaptedfromScholleetal.,2004). H YPOGENICSPELEOGENESISWITHIN S EVEN R IVERS E VAPORITES :C OFFEE C AVE ,E DDY C OUNTY ,N EW M EXICO 48 N JournalofCaveandKarstStudies, April2008

PAGE 3

areasofsinkholeformationbyconstructionofadikealong theeasternlakeshorewereonlypartiallysuccessful(Cox, 1967).McMillanDamwasbreachedin1991andthewater allowedtoflowintothenewlyconstructedBrantley Reservoir,whichislocatedinthedolomiticfaciesbeltof theSevenRiversFormation. TheLakeMcMillanarealiesnearthesouthernendof theRoswellBasin,akarsticartesianaquifersystem occupyingseveralhundredsquarekilometersinthelower PecosregionofsoutheasternNewMexico(Fig.1).Ground waterintheRoswellBasinisstoredinmultiplehighly porousandtransmissivezoneswithinGuadalupiancarbonatesoftheSanAndreslimestoneandtheoverlying GrayburgandQueenFormationsoftheArtesiaGroup (Fig.4).Secondaryporosityisdevelopedinvuggyand cavernouslimestonesandintraformationalsolution-collapsebreccias,theresultofsubsurfacedissolutionof evaporites.RechargetotheaquiferoccursonthePecos Slope,abroadareaeastoftheSacramentoMountains wheretheSanAndreslimestonecropsout.Redbedsand gypsumoftheSevenRiversFormationserveasaleaky upperconfiningunitfortheartesianaquifer(Fig.4) (Welder,1983). Water-bearingzoneswithintheartesianaquifersystem risestratigraphicallyfromnorthtosouth,occurringnear themiddleoftheSanAndresFormationinthenorthern RoswellArtesianBasin,andincarbonaterocksofthe GrayburgFormationinthesouthernpartoftheBasinnear LakeMcMillan(Fig.4).Thesouthernboundaryofthe ArtesianBasinisnotwell-defined,butisusuallylocated, somewhatarbitrarily,alongtheSevenRiversHills southwestofLakeMcMillan. TheSevenRiversconfiningunitisoverlainbyashallow water-tableaquifercomposedlargelyofTertiaryand Quaternaryalluvialsediment.Thismaterialwasdeposited onthePecosRiverfloodplainasitmigratedeastwarddue toupliftoftherisingSacramentoMountainstothewest.A substantialpercentageofrechargetotheshallowaquiferis derivedfromupwardflowthroughleakyconfiningbeds fromtheunderlyingartesianaquifer(Welder,1983).Very locally,inthevicinityofLakeMcMillan,theSevenRivers Formationmakesupalargepartoftheshallowaquifer, probablyassolutionconduitsintheSevenRiversgypsum. Sincetheinceptionofirrigatedagricultureinthelower PecosValleyintheearly20 th Century,mostofthe dischargefromtheartesianaquiferhasbeenfrom irrigationwells.However,substantialnaturaldischarge stilloccursalongthePecosRiver,flowingupwardthrough fracturesandsolutionchannelsintheoverlyingSeven Riversgypsum.Thisnaturaldischargehasformed acomplexofkarstsprings,sinkholelakes,andextensive wetlandslocatedalongthewestsideofthePecosRiver, eastofthecityofRoswell(Land,2005).Alongtheeastern marginofthePecosRivervalleysoutheastofRoswell, dischargefromtheartesianaquiferhascausedsubsurface dissolutionofgypsumandupwardpropagationofcollapse chimneys,forminglargegypsumcenotesatBottomless LakesStatePark(Land,2003;2006). Intheearly20 th Century,manywellsintheArtesianBasin flowedtothesurface,withyieldsashighas21,500Lmin 1 Figure3.StratigraphicchartofPermianfaciesinsoutheasternNewMexic owithcomparisonofstratigraphicunitswithinthe PecosValley-NorthwesternShelfandthenorthernDelawareBasin-Guadal upeMountains.CoffeeCaveisdevelopedinthe SevenRiversFormation(highlightedingray).W = Wolfcampian;LNDRN = Leonardian(adaptedfromZeigler,2006). K.W.Stafford,L.Land,andA.Klimchouk JournalofCaveandKarstStudies, April2008 N 49

PAGE 4

(Welder,1983).Althoughdecadesofintensivepumpinghave causedsubstantialdeclinesinhydraulichead,manywellsstill displaystrongartesianflow(LandandNewton,2007).As recentlyasthe1940s,wellsinthevicinityofLakeMcMillan flowedtothesurface,withwaterlevelsreportedupto12m abovegroundlevel(U.S.GeologicalSurvey,2007).HydrologicconditionswithinthesouthernArtesianBasinthus continuetoprovidestrongpotentialforhypogenicspeleogenesis. H YPOGENIC S PELEOGENESIS Karstdevelopmentisgenerallydescribedintermsof geomorphologyorhydrology,wheredissolutioniseither hypogenicorepigenic(hypergenic).Epigenicspeleogenesis, whichiswell-documentedinkarstliterature,involves surficialfeaturesthatactasentrancepointsfordescending watersthatmayeitherrechargelocalgroundwaterorform integratedcavenetworksthatfunctionassubsurface bypassfeaturesforoverlandflow(e.g.,FordandWilliams, 1989;White,1988).Epigenickarstiswellstudiedbecauseit naturallyformsnumeroussurfacemanifestationsthatare easilyrecognizedandhumanlyaccessible.Incontrast, hypogenicspeleogenesisisoftenoverlookedbecauseit formswithoutadirectsurfaceconnection,usuallyin confinedorsemi-confinedsettings.Hypogenickarstis oftenonlyexposedbysurfacedenudation,andisoften overprintedbyepigenicprocesses(Palmer,1991;Klimchouk,1996a;2000b).However,hypogeniccavesare occasionallyinterceptedduringmininganddrillingoperations,wherefeaturesshowlittleornooverprinting (Kempe,1996;Klimchouk,2000b;2003). Hypogenicspeleogenesishasbeenreferredtobroadly (e.g.,deep-seated,confined,semi-confined,artesian,transverse)andisoftenattributedtospecificfluidproperties (e.g.,sulfuricacid,hydrothermal),butintermsof hydrogeology,alltypesofhypogenickarstaresimilar. Hypogenickarstphenomenahavebeendescribedin evaporiticandcarbonaterocks.Inevaporites,bothmaze caves(e.g.,Klimchouk,1996b;2000a)andisolatedvoids (e.g.,Kempe,1996)arewell-documentedinEuropeand havebeenattributedtoconfinedspeleogenesis.Inmature carbonates,hypogenickarsthasbeenassociatedwithdeepseatedprocessesinvolvingacidic(e.g.,Palmer,1991;Lowe etal.,2000)andhydrothermal(e.g.,Dublyansky,2000) fluids.However,sulfuricacidandhydrothermalspeleogenesisaresimplyspecialsubsetswherefluidchemistryand temperature,respectively,increasethesolubilityofhost rock. FollowingthesuggestionofFord(2006),weadoptthe hydrogeological,ratherthanthegeochemical,notionof hypogenicspeleogenesis:‘‘…theformationofcavesby waterthatrechargesthesolubleformationfromunderlying strata,drivenbyhydrostaticpressureorothersourcesof energy,independentofrechargefromtheoverlyingor immediatelyadjacentsurface.’’HypogenickarstcandeFigure4.West-Easthydrostratigraphicsectionacrosssouthernendofth eRoswellArtesianBasin,showingrelationshipof CoffeeCavetotheunderlyingartesianaquifer.Theaquiferisrechargedo nthePecosSlopetothewest,wheretheSanAndres andArtesiaGroupcarbonatesareexposedinoutcrop.Groundwaterflowsdo wngradienttowardthePecosRiverandupward throughSevenRiversevaporites,whichserveasaleakyconfiningunitfor theaquifer.TQa = TertiaryandQuaternary alluvium,mostlyfloodplaindepositsoftheancestralandmodernPecosRi verandlacustrinesedimentsinthebedofoldLake McMillan.Noteverticalexaggerationis 29 3 H YPOGENICSPELEOGENESISWITHIN S EVEN R IVERS E VAPORITES :C OFFEE C AVE ,E DDY C OUNTY ,N EW M EXICO 50 N JournalofCaveandKarstStudies, April2008

PAGE 5

velopinanyenvironmentwherefluidsentersolublehost rockfrombelow,beingundersaturatedwithrespecttothe hostrockoracquiringaggressivenessduetomixingwith shallowerflowsystems(Klimchouk,2000b;2003).When thehydrogeologicframeworkisestablishedforhypogenic transversespeleogenesis,dissolutionmaydevelop3-D patternswithstratiformcomponents,dependingonthe specificlithologicandstructuralhostrockproperties. Pressurizedfluidswillattempttomigrateupwardtoward regionsoflowerpressure,oftenvalleysorothertopographiclows(To th,1999),wheretheexactflowpath dependsonthepermeabilityoflocalrockunitsandcrossformationalfractures.Flowmaybehorizontalthrough relativelyhighpermeabilityunits,oftensandsorcarbonates,orverticalthroughoriginallylowpermeability media,oftensolubleunits(Klimchouk,2000b;2003). Hypogenicprocessesmaybetheresultofeitherforced orfreeconvectionofgroundwater,oracombinationof both.Forcedconvectionisdrivenbydifferencesin hydraulicheadwithinanaquifersystem.Freeconvection isdrivenbyvariabilitywithinfluidproperties,whichsets updensitydifferenceswithinthefluid.Lighterfluidsrise anddenserfluidssink,usuallybecauseofdifferencesin salinityortemperatureoftheconvectingfluids(e.g., Kohout,1967;Kohoutetal.,1988).Andersonand Kirkland(1980)physicallymodeledthisprocessand showedthatbrinedensityconvectioncouldresultin significantdissolutionofsolublerocks,whereundersaturatedandsaturatedfluidssimultaneouslyriseandsink, respectively,withinaconfinedsystem.Iflimitedconnectivityoccursbetweensourcefluidsandsolublerock, simultaneousflowcanoccurthroughthesameporethroat, butinregionsofgreaterconnectivityseparateflowpaths forascendinganddescendingfluidsmaydevelop.AndersonandKirkland(1980)consideredbrinedensityconvectionastheprimarymechanismthroughwhichlarge verticalbrecciapipesdevelopedintheDelawareBasin, wherefluidsoriginatinginacarbonateaquiferand undersaturatedwithrespecttohalite(NaCl)rosethrough overlyinghalitebeds.Asrisingfluidsdissolvedhalite,they becamedenserandsubsequentlysankbacktothe carbonateaquifer,thusundersaturatedwaterswerecontinuouslyrejuvenatedatthedissolutionfront.Kempe (1996)showedthatspeleogenesisdrivenbyfreeconvection cancreatelargecavitiesatthebaseofthickevaporitic formationswithlowfracturedensity.Klimchouk(2000a; 2000b)invokedmixedconvectionforearlystagesof confinedtransversespeleogenesisinfracturedgypsumbeds intheWesternUkraine,withfreeconvectioneffects becomingincreasinglyimportantduringsubsequentstages, whenheaddistributionwithinanaquifersystemwas homogenizedduetoincreasedhydraulicconnectivityof aquiferousunits. Althoughmazecaveshaverecentlybeenshowntoform inhypogenicconfinedsettings(Klimchouk,2003),traditionalmodelsinvolveepigenicspeleogenesis.Theclassic mechanismofepigenicmazecavedevelopmentwas suggestedbyPalmer(1975;2000),wherewaterinfiltrates fromabovethroughporous,insolublerock.Asthefluids descendtheyareevenlydistributedthroughfractured solublerock,suchthatindividualpassagesenlargeat similarratesandconvergetoformmazepatterns.Other epigenicmodelsformazecavedevelopmentinvoke epigenicfluidsthataredeliveredlaterallytofractured rock,primarilyintheformofbackflooding,withan associatedshiftfromvadosetophreaticdissolution. Whetheramazecaveistheresultofhypogenicorepigenic speleogenesis,lithologicvariabilityandfracturingdominatemazecavedevelopment. Klimchouk(2000a;2003)andFrumkinandFischhendler(2005)havedescribedspecificmorphologicalfeatures thatareindicativeofhypogenicspeleogenesis.The morphologicsuiteconsistsoffeeders,masterpassages andoutlets,whichoccurwithinindividuallevelsof hypogenicsystems(Fig.5)(Klimchouk,2003).Feeders, orrisers,arethelowestelevationcomponentwithinthe suiteandarecharacterizedasverticalornear-vertical conduitsthroughwhichundersaturatedfluidsrisefrom loweraquifers.Feedersmayformasisolatedfeaturesor featureclusters.Masterpassagesarethecommonly exploredportionsofhypogeniccaves,whichareoften extensiveandformthelargestpassagesbecauseofthe existenceoflaterallywell-connectedandextensivefracture networksencasedwithincertainlithologichorizons.Outlets(i.e.,cupolasanddomes)occuratthehighestelevations withinasinglelevelofahypogeniccaveandformthe dischargefeaturesfortransverseflowtohigherelevations andlowerpressures.Isolatedrisersandoutletscan convergethroughcontinueddissolutionsuchthatrift-like featuresmaydevelopthatconnectlevelsinamulti-level, hypogenicsystem.Hypogeniccavesforminsluggishflow conditionsandshownoevidenceoffast-flowingfluids,but insteadexhibitsmoothwallswithirregularsolution Figure5.Diagrammaticrepresentationofmorphological featuresuiteindicativeofhypogenicspeleogenesis.Transmissivezonesaredolomiteandsolublebedsaregypsumin CoffeeCave. K.W.Stafford,L.Land,andA.Klimchouk JournalofCaveandKarstStudies, April2008 N 51

PAGE 6

Figure6.GeomorphicmapofCoffeeCave.A)PlanviewmapofCoffeeCaveshow ingmazepassagemorphology,hypogenic feederdistributionanddelineationbetweenlowerfloodedcaveleveland upperdrylevels.Theboundarybetweenthegrayand whiteregionsalongthenorthernedgeofCoffeeCaverepresentstheedgeof theMcMillanEscarpment.;B)Crosssectionsof CoffeeCavewith2 3 verticalexaggerationshowingrelationshipofpassagelevelsanddolomi teinterbeds(solidgraylines). NotethatthedashedlinesinFigure6Adelineatelocationswherepassagew idthwasmeasuredperpendiculartotheMcMillan EscarpmentforFigure11. H YPOGENICSPELEOGENESISWITHIN S EVEN R IVERS E VAPORITES :C OFFEE C AVE ,E DDY C OUNTY ,N EW M EXICO 52 N JournalofCaveandKarstStudies, April2008

PAGE 7

pocketsandresidualpendants(FrumkinandFischhendler, 2005),andvariousmorphologicimprintsofrisingfree convectioncirculation(Klimchouk,2000b;2003). C OFFEE C AVE Recently,aresurveyofCoffeeCavewasconductedin ordertobetterdocumentthecaveÂ’sdistributionandextent, lithologicvariability,andoccurrenceofmorphological featuresindicativeofhypogenictransversespeleogenesis. Atpresent,thelowermostlevel,whichcomprisesatleast halfoftheknowncave,basedontheoriginalsurvey (Belski,1972),isflooded(Fig.6A). Throughoutmostofthecave,fine-grainedsediments andangularcollapseblockscommonlycoverthefloor, obscuringmuchofthedissolutionalfloormorphologyand displayingabimodaldistributionofallogenicsedimentand autogenicbreakdown.Passagesproximaltothescarpedge havepartiallycollapsedduetoscarpretreat,creatinglarge earthfissuresonthelandsurfaceandacomplexentrance network(Fig.7).Theoccurrenceofabundantsurface fissuresbeyondtheknownextentofCoffeeCavestrongly suggeststhatnumerousmazecavesexistalongthesame erosionalscarp,butmostarelargelyblockedbybreakdown.SeveralkmtotheeastofCoffeeCave,clustersof cavesoccurwithintheBurtonFlatsareathatexhibit similarmorphologicfeaturesindicativeofhypogenic transversespeleogenesis. Mostpassageswithinthecaveareroughlyrectangular incross-sectionwiththindolomitebedsformingthe ceiling,floorsandintermittentledges(Fig.6B,8).The caveisathree-dimensionalmazewithmostpassages orientednorthwestandnortheast,whichprobablyrepresentsaconjugatefractureset(Fig.6A).Mostpassages interceptatsharpangles,whilemanyindividualpassages terminateinblindalcovesornarrowfractures,often recognizedasfeeders.Basedonpreviousmapping,the lowermostlevelappearstobetheregionofmostintense lateraldevelopment(Fig.6A).Successivelyhigherlevelsof thecavecontainprogressivelysmallerpassages,creatingat leastfourdistinctcavelevels,althoughthetwohighest levelsaregenerallytoosmalltobehumanlyaccessed.In severalregions,individualpassagestransectallfourlevels formingmajortrunkpassages,whichareintermittently separatedintodistinctlevels(Fig.6B),whilemanyregions containupperlevelpassagesthattransecttwoorthree levelsinlimitedareas.Itshouldbeemphasizedthatthe designationoflevelswithinCoffeeCaverefersonlyto distinctcavehorizonsthatarelithologicallyseparatedand laterallyextensiveandwhichwereformedconcurrently underaconstant,stablehydrologicregime.Thetermlevel doesnotimplyhydrologicallydistinctsolutionaleventsor levelsrelatedtochangesinhydrologicconditions. Mostoftheindividualcavepassagesexhibitcomplex surficialsculpturingwithinindividualgypsumbedsand betweendifferentlithologies;however,nodiscernable patternscommontoepigeniccaveswithorientedscallops ofsimilarshapeandsizewereobserved.Residualpendants arepresentthroughoutthecave.Additionalmorphological featuresindicativeofrisingtransverseflowcommonly occuronfloors,wallsandceilings,including:1)feeders,2) risingwallchannels,3)cupolas,and4)ceilinghalf-tubes. Generally,complexsuitesoffeaturesoccurinacontinuous seriesorincloseproximity. Feedersfunctionasfluidinletlocations,eitherjoining differentlevelsofacaveorterminatingblindlyinbedrock, generallytransmissivedolomitelayers.ThroughoutCoffee Cave,feederscommonlyoccuraspointsource(Fig. 9A,B,D),denseclusters(Fig.9F)orlinearfissure-like features(Fig.9E).Pointsourcefeedersareindividual Figure7.McMillanescarpmentshowinglargeearthfissures(A)andcomple xentrancenetwork(B). K.W.Stafford,L.Land,andA.Klimchouk JournalofCaveandKarstStudies, April2008 N 53

PAGE 8

featuresthatgenerallyformalongdolomite-gypsum contactsinwalls(Fig.9B)oratthemarginsoffloors (Fig.9A,C).Theyarecrudelyconicalfeaturesexhibiting anincreaseinaperturewidthtowardcavepassagesthat generallyincludeminordomingproximaltotheopen passage(Fig.9A,B,C).Feederclusterscommonlyexistas adenseoccurrenceofsmall-tomedium-sizedfeeders withinalimitedarea,butindividualfeederswithinclusters aremorphologicallysimilartopointfeeders.Linearfeeders developalongtheaxisofpassages,formingnarrowfissures alongfractures,andareoftenassociatedwithpassagesthat aretriangularincrosssectionandbroadenupwards (Fig.10E).Linearfeedersarelaterallyextensive,as comparedtopointfeeders. Cupolasarewell-documentedinthespeleogenetic literatureasdomalceilingfeatures(Osborne,2004). However,forthepurposeofthispaper,theyareviewed inabroadersensethatencompassestraditionalcupolasas wellasdomesandsimilaroutletfeaturesthatmayeitherbe closedoropen,concaveceilingfeatures.Cupolasare generallyellipticalinplanview(Fig.10A),butmayrange fromnear-circulartolenticular.Cupolaheightrangesfrom centimeterstometerswithinnerwallsthatmayvaryfrom gentlyslopingtonearvertical.Closedcupolas,whichfall withinthetraditionaldescriptionsofcupolasanddomes, areconcavefeatureswheretheentireinnersurfaceis gypsumbedrock,withtheuppersurfacecommonlyformed alongthecontactoftransmissivedolomiteinterbeds (Fig.10C).Incontrast,opencupolashaveopeningsin uppersurfaces,eitherinthecenteroroffsettooneside. Cupolasarerecognizableinthosepassageswherethe ceilingisstillwithingypsum(Fig.10A,D).Inmany passages,thenextupperdolomitebediscontinuously exposedattheceiling,butitisapparentthatthese exposureshaveformedbymergingofcloselyspaced cupolas(Fig.10B,C). Half-tubes(risingwallchannels,ceilingchannels)are elongateconcavestructuresthatoccuronceilingsandwalls andvaryfromshallowindentationstodeep,incised channels,ranginginwidthfromcentimeterstometers withcorrespondingdepths.Generally,half-tubesexhibit smoothroundedinteriorsurfacesandabrupt,well-defined marginswithadjacentwallsorceilings(Fig.9D,10C). Featuresonwallsaregenerallyverticallyoriented,butmay shiftlaterallyfrombottomtotop(Fig.9D).Ceilingfeatures areusuallydevelopedontheundersideofdolomitelayers andcommonlydisplayirregularmarginsresultingfromthe coalescingofserialcupolas(Fig.10B,C).Whenceilinghalftubesdonotformbeneathdolomitebeds,theygentlyslope Figure8.CompositelithologicsectionthroughCoffeeCave inrelationtothefouridentifiedcavelevels(designatedI–IV ondiagram).LevelsII,IIIandIVrepresenttheupperlevels thatwereresurveyedforthisstudy.LevelIiscurrently flooded,hencelithologyisunknown(presumablygypsum r withdolomiteinterbeds).Passagecross-sectionsareaverage passagerepresentationsofdistinctlevelsanddonotinclude connectionsbetweenlevels. H YPOGENICSPELEOGENESISWITHIN S EVEN R IVERS E VAPORITES :C OFFEE C AVE ,E DDY C OUNTY ,N EW M EXICO 54 N JournalofCaveandKarstStudies, April2008

PAGE 9

Figure9.FeederfeaturesinCoffeeCave.Blackscalebarsinfiguresareal lapproximately0.5mandcameraangleisnearhorizontalinallfeederfeaturesphotos.A)pointsourcefeedershowingp rominentdomingmorphologyproximaltomaster passage;B)typicalfeedershowingdevelopmentatthetopofadolomiticin terbed;C)completehypogenicmorphologicsuite showingriser(whitearrow),wallchannel(dashedwhitelines),ceilingc hannel(solidyellowlines)andoutlet(yellowarrow);D) welldevelopedwallriserwithassociatedwallchannel(dashedyellowlin es);E)linearriserdevelopedalongaxisofmaster passage;F)denseclusterofsmallfeedersabovedolomiteinterbedwithmi norvadoseoverprintingbelowdolomiteinterbed. K.W.Stafford,L.Land,andA.Klimchouk JournalofCaveandKarstStudies, April2008 N 55

PAGE 10

Figure10.OutletfeaturesinCoffeeCave.Blackscalebarsinfiguresarea llapproximately0.5m.A)seriesoftypicalceiling cupolas(cameraangleis 60 u upfromhorizontal,lookingtowardstheceiling);B)seriesofcupolastha tareintheprocessof coalescing(cameraangleis 70 u upfromhorizontal,lookingtowardtheceiling);C)ceilingchannelforme dbycomplete coalescingofserialcupolas(cameraangleis 30 u upfromhorizontal,lookingtowardstheceiling);D)completehypogenic morphologicsuiteshowingriser(whitearrow),wallchannel(yellowdash edlines),andceilingcupola(yellowarrow)(camera angleisroughlyhorizontal);E)rift-likepassageshowinglinearfeeder (yellowarrow),triangularpassageandupperdolomite bedthathaspartiallycollapsedduetolossofbuoyantsupport(whitearro w)(cameraangleisroughlyhorizontal).Figure10B isfromFuchslabyrinthCave,Baden-Wu ¨rtenberg,Germany(developedinverticallyheterogeneousbedsofTrias siclimestone showinglithologicvariabilitybetweentwodistinctbeds),insteadofCo ffeeCave,inordertobetterillustratetheintermediate stageofcupolacoalescence. H YPOGENICSPELEOGENESISWITHIN S EVEN R IVERS E VAPORITES :C OFFEE C AVE ,E DDY C OUNTY ,N EW M EXICO 56 N JournalofCaveandKarstStudies, April2008

PAGE 11

uptoanadjoininghalf-tubethatcontactsdolomite.All observedwallhalf-tubesjoinwithfeedersatlowerelevations (Fig.9C,D)andconvergewithceilinghalf-tubesathigher elevations(Fig.10D).Ceilinghalf-tubesjoinwithcupolas, whichwhenopen,formthelowersurfaceofriserstothenext highercavelevel;however,whencupolasareclosed,they appearasdeeperconcavestructureswithinacontinuous ceilinghalf-tube(Fig.10A,B).Bothwallandceilinghalftubescommonlyconvergefromsmallertolargerfeatures formingacomplexnetwork. Cupolas,feedersandhalf-tubesformacompositesuite ofmorphologicalfeaturesthroughoutCoffeeCave. Therefore,duringtheresurveyofCoffeeCave,feeders weremappedasaproxyforthedistributionofthis featuresuite(Fig.6A).Smallfeeders(lessthan10cm wide)areabundantthroughoutthecaveandcouldnotbe representedatthescaleofmapping;therefore,mapping waslimitedtomedium(0.1to1.0mdiameter)andlarge ( 1.0mdiameter)features.Duringmappingoffeeders, 25featureswereidentifiedthatconnectthelowerflooded portionofthecavetotheupperlevels,including12 mediumand13largefeeders.Intheupperlevelsofthe cave,107featureswereidentified,alllessthan1mwide. Itisprobablethatmanymorefeedersexist,buttheyare eitherobscuredbybreakdownandsedimentsorare locatedinpassagesthatweretoosmalltobesurveyed. Feedersarewelldistributed,butthereappearstobe anorthwardshiftinfeederabundancefromthelower leveltotheupperlevels. ObservationswithinCoffeeCaveduringthefirst quarterof2007indicatethatwaterlevelsinthecavehave risenbyatleast2minlessthan3months.Thisrisein waterlevelsmaybetheresultofanincreaseinhydraulic headintheartesianaquifer,oritmayreflectariseinwater levelsinthesurficialaquifer.Bothfactorsmaybeinplayin thevicinityofthecave,sincethewatertableaquiferis rechargedtoalargeextentbyupwardleakagefromthe underlyingartesianaquifer.HydraulicheadintheArtesian Basinhasbeenrisingsincethelate1970s(Landand Newton,2007),andfloodingofthelowerlevelsofCoffee Cavemaythusrepresent,inpart,increasedartesianflow fromtheunderlyingGrayburgFormation(Fig.4). D ISCUSSION CoffeeCaveisaclassicrectilinearmazecavewithat leastfourstratigraphiclevelsofdevelopmentandabundant morphologicfeaturessuggestiveofhypogenictransverse originswiththepronouncedroleoffreeconvection dissolution.Hydrologically,CoffeeCaveislocatedatthe southernendoftheRoswellArtesianBasin(Fig.1),where artesiandischargefromwellsandspringsiswell-documented(e.g.,Cox,1967;Welder,1983).Althoughwater extractionforagriculturalusehassignificantlylowered ground-waterlevelsintheregionoverthepreviouscentury, artesiandischargeisstillactive.Submergedspringsinthe gypsumcenotesatBottomlessLakesStateParkcontinue todischargesignificantvolumesofartesianwaters(Land, 2003;2006),whilefree-flowingwellsintheCoffeeCave areahavebeenreportedasrecentlyasthe1940s(U.S. GeologicalSurvey,2007).Inadditiontothehydrologic regimeinwhichCoffeeCaveislocated,interbedded dolomiteandgypsum(Fig.6B,8),coupledwithextensive tectonicfracturingintheLakeMcMillanarea,makethe SevenRiversFormationidealforthedevelopmentof multi-levelhypogenicmazecaves.However,priortothis study,hypogenicoriginsforgypsumcaveswerenot reportedinthisregion. Itisinstructivetoemphasizethedifferenceinthedegree ofkarstificationandslopegeomorphologybetweenthe gypsiferousSevenRiversoutcropswithinthePecosRiver ValleyandoutcropsintheSevenRiversEmbayment (Fig.1)oftheGuadalupeMountains,westofthevalley.In theSevenRiversEmbayment,theSevenRiversoutcrops areextensivebutlargelyintact,withstableslopesshowing minimalsignsofkarstification,whereaswithinthePecos RiverValleyanditsvicinity,theslopesaredramatically disturbedbygravitationalprocesseswithnumerouscollapsefeatures,apparentlyinducedbythehighdegreeof karstdevelopmentinthesubsurface.Thesemorphologies indicateintensekarstificationwithintheSevenRivers sequencebeneaththemigratingandincisingvalley,inducedbyrisingartesianflow(Fig.4),inagreementwith thegeneralconceptofspeleogenesisinconfinedsettings (Klimchouk,2000b;2003). CoffeeCavehastraditionallybeencharacterizedas amazecaveformedbybackfloodingalongthePecosRiver associatedwiththeconstructionofLakeMcMillaninthe late19 th Century.Inthismodel,floodingproduced dissolutionalongfracturesproximaltothescarp.Reports ofleakagefromLakeMcMillanthroughkarstconduits withintheSevenRiversFormation(Cox,1967)wereused asevidencefortheoriginofCoffeeCavethroughepigenic, floodingprocesses(alternatively,thisleakagecanbe perfectlyexplainedbythepresenceofhypogenicconduit systems).SedimentsandorganicdetrituswithinCoffee Cavewereusedasfurtherevidencetosupportanorigin throughbackflooding.Therefore,anynewmodelforthe proposedspeleogenesisofCoffeeCavemustconsider previous,althoughunpublished,interpretationsofcave origin. Cavesexhibitingmazepatternshavebeenshownto forminbothepigenicandhypogenicsettings;however,the completemorphologicalfeaturesuite,indicativeofrising flowandtheroleoffreeconvection,observedinCoffee Cave(i.e.,feeders,half-tubes,andoutletcupolas)hasonly beenreportedfromhypogeniccaves(Klimchouk,2003; FrumkinandFischhendler,2005).IfCoffeeCavehad formedbybackflooding,assuggestedbythepresenceof allogenicsediments,thenthecaveshouldexhibitan averagedecreaseinpassageaperturewidthawayfrom McMillanEscarpment,becausehighgypsumsolubility K.W.Stafford,L.Land,andA.Klimchouk JournalofCaveandKarstStudies, April2008 N 57

PAGE 12

promotesincreaseddissolutionproximaltothesourcein epigenicsettings(Klimchouk,2000c).However,mapping inCoffeeCavehasrevealedthatthereisnosystematic increaseinpassagewidthtowardthescarpface.Analyses ofpassagewidthinrelationtodistancefromthe escarpmentindicatethatpassagewidthactuallyincreases withdistancefromthescarp(Fig.11).Becauseofpresence oforganicdetritusinthecave,itislikelythatwaters derivedfromLakeMcMillanhavemodifiedCoffeeCave tosomeextent,primarilythroughsedimentinfilling,but lakewatersdonotappeartohavesignificantlymodified thecavethroughdissolution.Mostoftheoverprintingof CoffeeCaveÂ’shypogenicoriginistheresultofscarpfailure andretreat,whichhasproducedthecomplexcollapse entranceareaalongthescarpface,includingmorethan thirtydocumentedentrances(Fig.6A). ThemorphologicfeaturesuiteobservedinCoffeeCave, aswellasthepresenceofwallpendantsandtheabsenceof discernablescalloppatterns,stronglysupportsamodelof hypogenicspeleogenesisdrivenbyrisingcross-formational flowasasignificantpartoftheevolutionofkarstfeatures intheLakeMcMillanarea.Itissuggestedherethatthe cavewasformedunderconfinedconditions,whenthefloor ofthePecosValleywasatconsiderablyhigherelevations thanatpresent.Thedistributionoffeedersandtheoverall morphologyofCoffeeCaveareremarkablysimilartowelldocumentedhypogeniccavesinMiocenegypsumdeposits oftheWesternUkraine,suchasOzernaCave(Fig.12) (Klimchouk,1991).CoffeeCaveandthepartofOzerna Caveonthefigurebothshowalateralshiftinpassage developmentatadjacentlevels,whereclustersofconduits atthelowerlevelservedasafeedingsubsystemandrising fluidshavemigratedlaterallythroughtheupperlevel. Lateralmigrationresultsfromdiscordanceindistribution ofwater-bearingzonesintheunderlying(feeding)aquifer andpreferentialpathsfordischargethroughtheleaky confiningunits(argillaceousgypsumabovetheupper dolomitebedinCoffeeCave).Inthesluggishflow conditionsoftheconfinedsystem,densitydifferences readilydevelopedwhenfluidsrosefromthedolomitebeds throughgypsum,sothatfreeconvectioncellsoperated extensively.Thisisevidencedbycharacteristicmorphologicimprintsofrisingbuoyantcurrents(ear-likeorificesof feeders,risingwallchannels,ceilingchannelsandcupolas) andcharacteristicnarrowingoffeederstodepthdueto shieldingbysinkingconvectionlimbs.Thesimilarity betweenCoffeeCaveandwell-documentedhypogenic transversecaveselsewherefurthersupportstheroleof hypogeneprocessesintheoriginofCoffeeCave. AconceptualmodelforthespeleogenesisofCoffeeCave hasbeendeveloped(Fig.13).Priortokarstdevelopment, evaporitefaciesintheSevenRiversFormationnearLake McMillanprovidedaleakysealforconfinedartesianfluids intheRoswellBasinaquifer(Fig.13A).Groundwater initiallyfloweddowngradientthroughporouscarbonatesof theGrayburgandSanAndresFormationsandrosetoward thePecosRivervalley.Beneaththevalley,karstinitiation Figure11.Plotshowingrelationshipbetweenpassagewidth anddistancefromMcMillanEscarpment.Notethatthe dashedlinesinFigure6Adelineatelocationswherepassage widthsweremeasuredperpendiculartotheMcMillan Escarpment. Figure12.MapfragmentfromOzernaCave,Western Ukraine,showingcavemorphologyandfeederdistribution (Klimchouk,1991).NotethesimilaritieswithCoffeeCave depictedinFigure6A. H YPOGENICSPELEOGENESISWITHIN S EVEN R IVERS E VAPORITES :C OFFEE C AVE ,E DDY C OUNTY ,N EW M EXICO 58 N JournalofCaveandKarstStudies, April2008

PAGE 13

begantodevelopverticalflowpathsalongfracturesinthe evaporitefacies(Fig.13B).Highlyfractureddolomitelayers withintheSevenRiversFormationservedaslaterally transmissiveunitstodistributeundersaturated,aggressive waterstoavailablefissuresintheverticallyadjacentgypsum beds.Locally,risingconduitscouldhavedevelopedeven withoutforcedflowthroughguidingfracturesinthe gypsum,solelybybuoyancy-drivendissolution.Controlled byarrangementoffractureswithinagypsum-dolomite intercalatedsequence,themulti-levelmazecavedeveloped (Fig.13C,D).Eventually,confinementwasbreachedbythe entrenchingPecosRiverValley,andprimaryconduitswere establishedforartesiandischarge,mostlikelyalongalimited numberofflowpaths(Fig.13E,F).Baselevelswithinthe regionloweredandthecavewasexposedtoepigenic processes.Duringthisfinalphase,minorvadoseoverprintingandcollapseofthindolomitebedsoccurreddueto lossofbuoyantsupport(Fig.10E).Recently,inthe20 th Century,floodingofthecavethroughtheconstructionof LakeMcMillanintroducedallogenicsediments.Today, continuedscarpretreathascreatedacomplexentrance network. C ONCLUSIONS CoffeeCaveprovidesdirectevidenceforhypogenic transversespeleogenesisdrivenbycross-formationalflow anddensityconvectionwithintheSevenRiversFormation andmorebroadlyintheDelawareBasinregionof southeasternNewMexico.Thecomplex,three-dimensionalmazepatternofthecaveissuggestiveofhypogenic originwherenon-competitiveconfinedflowresultedin uniformdissolutionalongplanesofbrittledeformation. Thecompletesuiteofobservedmorphologicalfeatures withinthecaveprovidesunequivocalevidenceofhypogenicspeleogenesisbyrisingmixed-convectionflow,where smoothwallsandconcavemorphologiesdelineateprevious freeconvectioncells.Thepresenceofdolomiteinterbeds andregionalfracturingmakestheSevenRiversFormation idealfordevelopmentofhypogeniccaves;however,natural heterogeneitiesinmostcarbonateandsulfaterocksare sufficientforhypogenicspeleogenesisintheproper hydrologicregime.Theoccurrenceofabundantsurface fissuresbeyondtheknownextentofCoffeeCavesuggests thatnumeroushypogenic,mazecavesexistalongthesame erosionalscarp,butmostarelargelyblockedbybreakdown.Moreover,clusteredhighlykarstifiedfieldsexist beyondthescarp,althoughstillwithinthebroadlimitsof thePecosValleyandwithintheevaporitefaciesofthe SevenRiversandRustlerFormations.Thesekarstified fields(e.g.,BurtonFlats,NashDraw)containnumerous caveslackinggeneticrelationshipswiththesurfaceandare likelyhypogenicsystemsthatarecurrentlybeingpartially denuded.Itisalsofeasibletoassumethattheleakagefrom LakeMcMillanthroughkarstconduitswithintheSeven RiversFormationwasrelatednottoepigenickarst developmentbutthepresenceofpre-existinghypogenic conduitsystemsbeneaththevalley. Basedoncurrentandongoingstudiesbytheauthorsof karstdevelopmentwithintheNewMexico-WestTexas region,thesignificanceofhypogenicspeleogenesisappears tobepoorlyrecognized.Karstdevelopmentinother regionalgypsumformations(i.e.,Castile,Yeso,Rustler, andSanAndres)includesnumerouscavesthatarethree dimensionalmazesand/orcontaincompletesuitesof morphologicalfeaturesindicativeoftheroleofdensity drivendissolution.However,thesefeaturesarenotlimited togypsumformationsbuthavealsobeenobservedin carbonatekarstwithintheregion,including,butnot limitedto,thecavesoftheGuadalupeMountains(e.g., CarlsbadCavern,LechuguillaCave,McKittrickHill caves).AlthoughlimestonecavesoftheGuadalupe Mountainshavebeenattributedtosulfuricacid(H 2 SO 4 ) speleogenesis,theyarehypogeniccaves,notonlybythe sourceofacidity,butalsohydrologically.Thesecavesare hypogenictransversefeaturescontainingmorphologic suitesindicativeofrisingflowandfreeconvectioneffects. Therefore,theroleofhypogenicspeleogenesisislikelyto beextensivethroughoutthesouthwestUnitedStates,butis currentlynotrecognizedeitherbecauseofextensive epigenicoverprinting,misinterpretation,orbecausefluid chemistry(e.g.,sulfuricacid)isproposedastheprimary criterionfordistinguishinghypogenicspeleogenesis. Theidentificationofhypogenicspeleogenesiswithin southeasternNewMexico,beyondthecavesofthe GuadalupeMountainsandbrecciapipeswithinthe DelawareBasin,suggeststhatmorestudiesneedtobe conductedwithintheregion,includingre-evaluationofthe originofmanyindividualcavesandkarstregions.The implicationsofanimprovedunderstandingofhypogenic speleogenesiswithintheregionwillhavesignificant impactsondelineatingareasofpotentialengineering geohazards,investigationofpetroleumresources,mineral resourcesandground-waterbehaviorassociatedwith karst.Ultimately,recognitionoftheimportanceofmixed convectionprocessesrelatedtohypogenicdissolutionwill enablethedevelopmentofimprovedmodelsforthe speleogeneticevolutionandbasindiagenesisoftheentire DelawareBasinregion. A CKNOWLEDGEMENTS Theauthorswishtoexpresstheirappreciationtothe numerousindividualsinvolvedthroughoutthisproject. TheresurveyofCoffeeCavewasconductedbytheauthors withhelpfromStanAllison,PaulBurger,JonJasper, LucasMiddletonandPatSeiser.DaveBelski,JimGoodbarandRayNanceprovidedessentialinformationrelated toregionalcavedevelopmentandexistingcavesurveys withinthearea.ReviewersDavidLevyandPaulBurgerare thankedfortheirusefulcommentswhichhelpedto improvethemanuscript. K.W.Stafford,L.Land,andA.Klimchouk JournalofCaveandKarstStudies, April2008 N 59

PAGE 14

Figure13.ConceptualmodelforthespeleogeneticevolutionofCoffeeCav einrelationtotheeastwardmigrationofthePecos Rivervalley,associatedsurfacedenudationandevolvingground-waterf lowpaths.A)regionalconceptualizationofgroundwatercirculation(solidblacklines)inmid-to-lateTertiarytime;B)lo calconceptualizationofhydrologyinCoffeeCaveregion, inrelationtoFigure13A,showingforcedflow(solidblacklines)alongve rticalfracturesintheSevenRiversFormation;C) regionalconceptualizationofground-watercirculationinmid-to-late QuaternarywhenCoffeeCavewasprimarilyforming;D) localconceptualizationofCoffeeCaveinrelationtoFigure13C,showing forcedconvection(solidarrows)andfreeconvection (dashedarrows)circulationinvolvedinspeleogenesis;E)currenthydro logicregimeinthelowerPecosRiverValley;andF) conceptualizationofthecurrenthydrologicregimeofCoffeeCaveafters urficialbreaching,relatingtoFigure13E.Note: FiguresB,DandFareschematicillustrationsthatarenotdrawntoscale,a nddonotrepresenttheactuallevelsofcave development,duetotheresolutionofthefigure.PSA = PermianSanAndresFm;PG = PermianGrayburgFm;PSR = Permian SevenRiversFm;PY = PermianYatesFm;apra = ancestralPecosRiveralluvium;TQa = TertiaryandQuaternaryalluvium. H YPOGENICSPELEOGENESISWITHIN S EVEN R IVERS E VAPORITES :C OFFEE C AVE ,E DDY C OUNTY ,N EW M EXICO 60 N JournalofCaveandKarstStudies, April2008

PAGE 15

R EFERENCES Anderson,R.Y.,andKirkland,D.W.,1980,Dissolutionofsaltdeposits bybrinedensityflow:Geology,v.8,p.66–69. Bachman,G.O.,1984,RegionalgeologyofOchoanevaporites,northern partofDelawareBasin:NewMexicoBureauofMinesandMineral Resources,Circular184,22p. Belski,D.,1972,CoffeeCave:unpublishedmap. Belski,D.,ed.,1992,GYPKAPReportVolume 2:Albuquerque,N.M., SouthwesternRegionoftheNationalSpeleologicalSociety,57p. Chapin,C.E.,andCather,S.M.,1994,Tectonicsettingoftheaxialbasins ofthenorthernandcentralRioGrandeRift, in Keller,G.R.,and Cather,S.M.,eds.,Structure,stratigraphy,andtectonicsetting: GeologicalSocietyofAmericaSpecialPaper291,p.5–25. Cox,E.R.,1967,GeologyandhydrologybetweenLakeMcMillanand CarlsbadSprings,EddyCo.,NewMexico:U.S.GeologicalSurvey Water-SupplyPaper1828,48p. Dickenson,W.R.,1981,Platetectonicevolutionofthesouthern Cordillera:ArizonaGeologicalSocietyDigest,v.14,p.113–135. Dublyansky,Y.V.,2000,Hydrothermalspeleogenesis–itssettingsand peculiarfeatures, in Klimchouk,A.,Ford,D.C.,Palmer,A.N.,and Dreybrodt,W.,eds.,Speleogenesis:EvolutionofKarstAquifers: Huntsville,Ala.,NationalSpeleologicalSociety,Inc.,p.292–297. Eaton,J.,ed.,1987,GYPKAP1987AnnualReport:Alamogordo,N.M., SouthwesternRegionoftheNationalSpeleologicalSociety,35p. Ford,D.C.,2006,Karstgeomorphology,cavesandcavedeposits:areview ofNorthAmericanContributionsduringthepasthalfcentury, in Harmon,R.S.,andWicks,C.W.,eds.,PerspectivesonKarst Geomorphology,HydrologyandGeochemistry:Boulder,Colo., GeologicalSocietyofAmericaSpecialPaper505,p.1–14. Ford,D.C.,andWilliams,P.W.,1989,KarstGeomorphologyand Hydrology:London,UnwinHymam,601p. Frumkin,A.,andFischhendler,I.,2005,Morphometryanddistribution ofisolatedcavesasaguideforphreaticandconfinedpaleohydrologicalconditions:Geomorphology,v.67,p.457–471. Hendrickson,G.E.,andJones,R.S.,1952,GeologyandGround-Water ResourcesofEddyCounty,NewMexico:Socorro,N.M.,New MexicoBureauofMinesandMineralResources,189p. Hill,C.A.,1996,GeologyoftheDelawareBasin,Guadalupe,Apacheand GlassMountains:NewMexicoandWestTexas:Midland,Texas, PermianBasinSection–SEPM,480p. Horak,R.L.,1985,Trans-PecostectonismanditseffectsonthePermian Basin, in Dickerson,P.W.,andMuelberger,eds.,Structureand TectonicsofTrans-PecosTexas:Midland,Texas,WestTexasGeologicalSociety,p.81–87. Kelley,V.C.,1971,GeologyofthePecosCountry,SoutheasternNew Mexico:Socorro,N.M.,NewMexicoBureauofMinesandMineral Resources,78p. Kempe,S.,1996,GypsumkarstinGermany:InternationalJournalof Speleology,v.25,no.3–4,p.49–60. Klimchouk,A.B.,1991,LargemazecavesingypsumintheWestern Ukraine:speleogenesisunderartesianconditions:NationalSpeleologicalSocietyBulletin,v.53,p.71–82. Klimchouk,A.,1996a,Thetypologyofgypsumkarstaccordingtoits geologicalandgeomorphologicalevolution:InternationalJournalof Speleology,v.25,no.3–4,p.49–60. Klimchouk,A.,1996b,GypsumkarstofthewesternUkraine:InternationalJournalofSpeleology,v.25,no.3–4,p.263–278. Klimchouk,A.,2000a,Speleogenesisofthegreatgypsummazecavesin thewesternUkraine, in Klimchouk,A.,Ford,D.C.,Palmer,A.N., andDreybrodt,W.,eds.,Speleogenesis:EvolutionofKarstAquifers: Huntsville,Ala.,NationalSpeleologicalSociety,Inc.,p.431–442. Klimchouk,A.,2000b,Speleogenesisunderdeep-seatedandconfined conditions, in Klimchouk,A.,Ford,D.C.,Palmer,A.N.,and Dreybrodt,W.,eds.,Speleogenesis:EvolutionofKarstAquifers: Huntsville,Ala.,NationalSpeleologicalSociety,Inc.,p.244–260. Klimchouk,A.,2000c,Speleogenesisingypsum, in Klimchouk,A.,Ford, D.C.,Palmer,A.N.,andDreybrodt,W.,eds.,Speleogenesis: EvolutionofKarstAquifers:Huntsville,Ala.,NationalSpeleological Society,Inc.,p.261–273. Klimchouk,A.,2003,Conceptualizationofspeleogenesisinmulti-story artesiansystems:amodeloftransversespeleogenesis:Speleogenesis andEvolutionofKarstAquifers,v.1,no.2,p.1–18. Klimchouk,A.,Lowe,D.,Cooper,A.,andSauro,U.,eds.,1996,Gypsum KarstoftheWorld:InternationalJournalofSpeleology,v.25,no.3– 4,307p. Kohout,F.A.,1967,Ground-waterflowandthegeothermalregimeofthe FloridianPlateau:TransactionsoftheGulfCoastAssociationof GeologicalSocieties,v.17,p.339–354. Kohout,F.A.,Meisler,H.,Meyer,F.,Johnston,R.,Leve,G.,andWait, R.,1988,HydrogeologyoftheAtlanticContinentalMargin, in Sheridan,R.,andGrow,J.,eds.,TheGeologyofNorthAmerica;the AtlanticContinentalMargin,U.S.:Boulder,Colo.,Geological SocietyofAmerica,v.I-2,p.463–480. Land,L.,2003,Evaporitekarstandregionalgroundwatercirculationin thelowerPecosValley, in Johnson,K.S.,andNeal,J.T.,eds., EvaporiteKarstandEngineeringandEnvironmentalProblemsinthe UnitedStates:OklahomaGeologicalSurveyCircular109,p.227–232. Land,L.,2005,Evaluationofgroundwaterresidencetimeinakarstic aquiferusingenvironmentaltracers:RoswellArtesianBasin,New Mexico, in ProceedingsoftheTenthMultidisciplinaryConferenceon SinkholesandtheEngineeringandEnvironmentalImpactsofKarst, SanAntonio,Texas,2005:ASCEGeotechnicalSpecialPublication no.144,p.432–440. Land,L.,2006,HydrogeologyofBottomlessLakesStatePark, in Land, L.,Lueth,V.,Raatz,B.,Boston,P.,andLove,D.,eds.,Cavesand KarstofSoutheasternNewMexico:NewMexicoGeologicalSociety, Guidebook57,p.95–96. Land,L.,andNewton,B.T.,2007,Seasonalandlong-termvariationsin hydraulicheadinakarsticaquifer:RoswellArtesianBasin,New Mexico:JournaloftheAmericanWaterResourcesAssociation(in press). Lee,J.,ed.,1996,GYPKAPReportVolume3:Alamogordo,N.M., SouthwesternRegionoftheNationalSpeleologicalSociety,69p. Lowe,D.J.,Bottrell,S.H.,andGunn,J.,2000,Somecasestudiesof speleogenesisbysulfuricacid, in Klimchouk,A.,Ford,D.C.,Palmer, A.N.,andDreybrodt,W.,eds.,Speleogenesis:EvolutionofKarst Aquifers:Huntsville,Ala.,NationalSpeleologicalSociety,Inc., p.304–308. Osborne,R.A.L.,2004,Thetroubleswithcupolas,ActaCarsologica, v.33,no.2,p.9–36. Palmer,A.N.,1975,Theoriginofmazecaves:NationalSpeleological SocietyBulletin,v.37,no.3,p.56–76. Palmer,A.N.,1991,Originandmorphologyoflimestonecaves:GeologicalSocietyofAmericaBulletin,v.103,p.1–21. Palmer,A.N.,2000,Mazeoriginbydiffuserechargethroughoverlying formations, in Klimchouk,A.,Ford,D.C.,Palmer,A.N.,and Dreybrodt,W.,eds.,Speleogenesis:EvolutionofKarstAquifers: Huntsville,Ala.,NationalSpeleologicalSociety,Inc.,p.387–390. Quinlan,J.F.,Smith,R.A.,andJohnson,K.S.,1987,Gypsumkarstand saltkarstoftheUnitedStatesofAmerica, in Attisymposio interntionalsulcarsimonelleevaporiti,LeGrotted’Italia,v.4, no.XIII,p.73–92. Scholle,P.A.,Goldstein,R.H.,andUlmer-Scholle,D.S.,2004,Classic UpperPaleozoicReefsandBiohermsofWestTexasandNewMexico: Socorro,N.M.,NewMexicoInstituteofMiningandTechnology, 166p. To th,J.,1999,Groundwaterasageologicagentandoverviewofthe causes,processes,andmanifestations:HydrogeologyJournal,v.7, p.1–14. U.S.GeologicalSurvey,2007,NationalWaterInformationSystem (NWIS):http://nwis.waterdata.usgs.gov. White,W.B.,1988,GeomorphologyandHydrologyofKarstTerrains: OxfordUniv.Press,NewYork,NY,464p. Welder,G.E.,1983,GeohydrologicframeworkoftheRoswellGroundWaterBasin,ChavesandEddyCounties,NewMexico:NewMexico StateEngineerTechnicalReport42,28p. Zeigler,K.E.,2006,Stratigraphicchart, in Land,L.,Lueth,V.,Raatz,B., Boston,P.,andLove,D.,eds.,CavesandKarstofSoutheasternNew Mexico:NewMexicoGeologicalSociety,Guidebook57,p.inside backcover. K.W.Stafford,L.Land,andA.Klimchouk JournalofCaveandKarstStudies, April2008 N 61



PAGE 1

ERRATAIthasbeenbroughttotheJournalsattentionthatonpage113inVolume69Issue1thecontributionofFigure8byJimPapadakiswasinadvertentlyomittedfromtheAcknowledgementssectionofthemanuscriptbyWilliamHallidaytitledPseudokarstinthe21stCentury.BOOKREVIEW72NJournalofCaveandKarstStudies,April2008



PAGE 1

Journal of Cave and Karst StudiesVolume 70 Number 1 April 2008 CONTENTSEditorial 1Journal of Cave and Karst Studies Listing in the Journal of Citation Report: What Does it Mean? Malcolm S. FieldArticle 3Characterization of Cave Aerophytic Algal Communities and Effects of Irradiance Levels on Production of Pigments Janez Mulec, Gorazd Kosi, and Danijel VrhovekArticle 13New Findings at Andrahomana Cave, Southeastern Madagascar D.A. Burney, N. Vasey, L.R. Godfrey, Ramilisonina, W.L. Jungers, M. Ramarolahy, and L. RaharivonyArticle 25Variable Calcite Deposition Rates as Proxy for Paleo-Precipitation Determination as Derived from Speleothems in Central Florida, U.S.A. Philip E. Van Beynen, Limaris Soto, and Jason PolkArticle 35Castile Evaporite Karst Potential Map of the Gypsum Plain, Eddy County, New Mexico and Culberson County, Texas: A GIS Methodological Comparison Kevin W. Stafford, Laura Rosales-Lagarde, and Penelope J. BostonArticle 47Hypogenic Speleogenesis Within Seven Rivers Evaporites: Coffee Cave, Eddy County, New Mexico Kevin W. Stafford, Lewis Land, and Alexander KlimchoukArticle 62Ground-Water Storage Calculation in Karst Aquifers with Alluvium or No-Flow Boundaries Ezatollah RaeisiErrata 72 J O UR NA L O F C A V E AN D K A R ST ST UDI ESApril 2008 V olume 70, Number 1 I SSN 1090-6924 A Publication of the National Speleological Society Madagascar paleo-rainfall IN THIS ISSUE:

PAGE 2

The Journal of Cave and Karst Studies is a multidisciplinary journal devoted to cave and karst research. The Journal is seeking original, unpublished manuscripts concerning the scientic study of caves or other karst features. Authors do not need to be members of the National Speleological Society, but preference is given to manuscripts of importance to North American speleology. LANGUAGES: The Journal of Cave and Karst Studies uses American-style English as its standard language and spelling style, with the exception of allowing a second abstract in another language when room allows. In the case of proper names, the Jour nal tries to accommodate other spellings and punctuation styles. In cases where the Editor-in-Chief nds it appropriate to use nonEnglish words outside of proper names (generally where no equivalent English word exists), the Journal italicizes them. However, the common abbreviations i.e., e.g., et al., and etc. should appear in roman text. Authors are encouraged to write for our combined professional and amateur readerships. CONTENT: Each paper will contain a title with the authors names and addresses, an abstract, and the text of the paper, including a summary or conclusions section. Acknowledgments and references follow the text. ABSTRACTS: An abstract stating the essential points and results must accompany all articles. An abstract is a summary, not a promise of what topics are covered in the paper. STYLE: The Journal consults The Chicago Manual of Style on most general style issues. REFERENCES: In the text, references to previously published work should be followed by the relevant authors name and date (and page number, when appropriate) in parentheses. All cited references are alphabetical at the end of the manuscript with senior authors last name rst, followed by date of publication, title, publisher, volume, and page numbers. Geological Society of America for mat should be used (see http://www.geosociety.org/pubs/geoguid5. htm). Please do not abbreviate periodical titles. Web references are acceptable when deemed appropriate. The references should follow the style of: Author (or publisher), year, Webpage title: Publisher (if a specic author is available), full URL (e.g., http://www. usgs.gov/citguide.html) and date when the web site was accessed in brackets; for example [accessed July 16, 2002]. If there are specic authors given, use their name and list the responsible organization as publisher. Because of the ephemeral nature of websites, please provide the specic date. Citations within the text should read: (Author, Year). SUBMISSION: Effective July 2007, all manuscripts are to be submitted via AllenTrack, a web-based system for online submission. The web address is http://jcks.allentrack2.net. Instructions are provided at that address. At your rst visit, you will be prompted to establish a login and password, after which you will enter information about your manuscript (e.g., authors and addresses, manuscript title, abstract, etc.). You will then enter your manuscript, tables, and gure les separately or all together as part of the manuscript. Manuscript les can be uploaded as DOC, WPD, RTF, TXT, or LaTeX. A DOC template with additional manuscript specications may be downloaded. (Note: LaTeX les should not use any unusual style les; a LaTeX template and BiBTeX le for the Journal may be downloaded or obtained from the Editor-inChief.) Table les can be uploaded as DOC, WPD, RTF, TXT, or LaTeX les, and gure les can be uploaded as TIFF, EPS, AI, or CDR les. Alternatively, authors may submit manuscripts as PDF or HTML les, but if the manuscript is accepted for publication, the manuscript will need to be submitted as one of the accepted le types listed above. Manuscripts must be typed, double spaced, and single-sided. Manuscripts should be no longer than 10,000 words plus tables and gures, but exceptions are permitted on a case-bycase basis. Authors of accepted papers exceeding this limit may have to pay a current page charge for the extra pages unless decided otherwise by the Editor-in-Chief. Extensive supporting data will be placed on the Journals website with a paper copy placed in the NSS archives and library. The data that are used within a paper must be made available. Authors may be required to provide supporting data in a fundamental format, such as ASCII for text data or comma-delimited ASCII for tabular data. DISCUSSIONS: Critical discussions of papers previously published in the Journal are welcome. Authors will be given an opportunity to reply. Discussions and replies must be limited to a maximum of 1000 words and discussions will be subject to review before publication. Discussions must be within 6 months after the original article appears. MEASUREMENTS: All measurements will be in Systeme Internationale (metric) except when quoting historical references. Other units will be allowed where necessary if placed in parentheses and following the SI units. FIGURES: Figures and lettering must be neat and legible. Figure captions should be on a separate sheet of paper and not within the gure. Figures should be numbered in sequence and referred to in the text by inserting (Fig. x). Most gures will be reduced, hence the lettering should be large. Photographs must be sharp and high contrast. Color will generally only be printed at authors expense. TABLES: See http://www.caves.org/pub/journal/PDF/Tables. pdf to get guidelines for table layout. COPYRIGHT AND AUTHORS RESPONSIBILITIES: It is the authors responsibility to clear any copyright or acknowledgement matters concerning text, tables, or gures used. Authors should also ensure adequate attention to sensitive or legal issues such as land owner and land manager concerns or policies. PROCESS: All submitted manuscripts are sent out to at least two experts in the eld. Reviewed manuscripts are then returned to the author for consideration of the referees remarks and revision, where appropriate. Revised manuscripts are returned to the appropriate Associate Editor who then recommends acceptance or rejection. The Editor-in-Chief makes nal decisions regarding publication. Upon acceptance, the senior author will be sent one set of PDF proofs for review. Examine the current issue for more information about the format used. ELECTRONIC FILES: The Journal is printed at high resolution. Illustrations must be a minimum of 300 dpi for acceptance.The Journal of Cave and Karst Studies (ISSN 1090-6924, CPM Number #40065056) is a multi-disciplinary, refereed journal published three times a year by the National Speleological Society, 2813 Cave Avenue, Huntsville, Alabama 35810-4431 USA; Phone (256) 852-1300; Fax (256) 851-9241, email: nss@caves.org; World Wide Web: http://www.caves.org/pub/journal/. The annual subscription fee is $23 US, $44 US for 2 years, and $65 US for 3 years. Check the Journal website for international rates. Back issues and cumulative indices are available from the NSS ofce. POSTMASTER: send address changes to the Journal of Cave and Karst Studies, 2813 Cave Avenue, Huntsville, Alabama 35810-4431 USA. The Journal of Cave and Karst Studies is covered by the following ISI Thomson Services Science Citation Index Expanded, ISI Alerting Services, and Current Contents/ Physical, Chemical, and Earth Sciences. Copyright 2008 by the National Speleological Society, Inc. Front cover: Feeder feature in Coffee Cave, New Mexico, see article by Stafford et.al. beginning on page 47.Published By The National Speleological SocietyEditor-in-Chief Malcolm S. FieldNational Center of Environmental Assessment (8623P) Ofce of Research and Development U.S. Environmental Protection Agency 1200 Pennsylvania Avenue NW Washington, DC 20460-0001 703-347-8601 Voice 703-347-8692 Fax eld.malcolm@epa.govProduction EditorScott A. EngelCH2M HILL 304 Laurel Street, Suite 2A Baton Rouge, LA 70801-1815 225-381-8454 scott.engel@ch2m.comJournal Proofreader441 S. Kearney St Denver, CO 80224 303-355-5283 dgdavis@nyx.netJOURNAL ADVISOR Y BOARD Dave Culver Annette Summers Engel John Mylroie Megan Porter Stephen Worthington BOARD OF EDITORSAnthropology University of Kentucky211 Lafferty Hall Lexington, KY 40506-0024 Conservation-Life Sciences Lewis & Associates, LLC. Cave, Karst & Groundwater Biological Consulting Department of Geology and Environmental Science Paul BurgerCave Resources Ofce Stephen R. Mosberg, M.D Microbiology Department of Biology State University of New York Plattsburgh, NY 12901 Paleontology Park Museum Management Program National Park Service 1201 Oakridge Dr. Suite 150 Fort Collins, CO 80525 Social Sciences Joseph C. DouglasHistory Department Volunteer State Community College Department of Earth Sciences State University of New York Oneonta, NY 13820-4015