Citation
Preliminary geochemical assessment of limestone resources and stone use at Maya sites in the Three Rivers Region, Belize

Material Information

Title:
Preliminary geochemical assessment of limestone resources and stone use at Maya sites in the Three Rivers Region, Belize
Publication Date:
Language:
English

Subjects

Subjects / Keywords:
Geology ( local )
Genre:
Article
serial ( sobekcm )

Notes

General Note:
Michael L. Brennan a, *, Eleanor M. King b, Leslie C. Shaw c, Stanley L. Walling d, Fred Valdez Jr. e aGraduate School of Oceanography, University of Rhode Island, Narragansett, RI 02882, USA bDepartment of Sociology and Anthropology, Howard University, Washington, D.C., USA cDepartment of Sociology and Anthropology, Bowdoin College, Brunswick, ME, USA dDepartment of Social Sciences, Community College of Philadelphia, USA eDepartment of Anthropology, University of Texas at Austin, USA The carbonate bedrock of northwestern Belize is poorly understood from the standpoint of both geochemistry and the use of stone in prehispanic Maya sites for buildings and monuments. The friable nature of the rock in this topographically rugged area makes it especially difficult to distinguish mon- uments from bedrock spall, as little carving, if any, survives, and identification rests on location and positioning. The research presented here analyzed 63 limestone samples collected from two sites in the Three Rivers Region of Belize. ICP-MS and ICP-AES were used to characterize the major, minor, and trace element chemistry of the limestone bedrock of the region and determine the amount of geochemical variability. Another important objective was attempting to trace the movement of monument stone and determining whether it was imported from outside of the sites. Bedrock, quarries, and possible monu- ments were all sampled for these purposes. Bedrock proved to be similar across wide areas. However, at Chawak But'o'ob, along the flank of the Rio Bravo, changes downslope in Mg concentration suggest a leaching of the bedrock by meteoric waters based on differences in porosity. At Maax Na, a hilltop site, in contrast, such leaching is not as apparent. Many monuments at both sites were found to be composed of stone similar in chemistry to the local bedrock, including several of the identifiable stelae. However, our analyses also revealed that a few monuments at Maax Na were made of material with a different chemical composition, apparently from stone imported to the site. These results suggest that the Maya deliberately selected certain types of limestone for certain purposes, and may even have traded in non- local rock. Overall, the methods used in this pilot study indicate there is real potential in more intensive, regional assessments of the materials used at archaeological sites, even in areas where the local stone does not have a distinctive geochemical signature.
Restriction:
Open Access - Permission by Author(s)
General Note:
See Extended description for more information.

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
K26-03265 ( USFLDC DOI )
k26.3265 ( USFLDC Handle )
17091 ( karstportal - original NodeID )

USFLDC Membership

Aggregations:
Karst Information Portal

Postcard Information

Format:
Serial

Downloads

This item has the following downloads:


Full Text

PAGE 1

Preliminarygeochemicalassessmentoflimestoneresourcesandstoneuseat MayasitesintheThreeRiversRegion,BelizeMichaelL.Brennana,*,EleanorM.Kingb,LeslieC.Shawc,StanleyL.Wallingd,FredValdezJr.eaGraduateSchoolofOceanography,UniversityofRhodeIsland,Narragansett,RI02882,USAbDepartmentofSociologyandAnthropology,HowardUniversity,Washington,D.C.,USAcDepartmentofSociologyandAnthropology,BowdoinCollege,Brunswick,ME,USAdDepartmentofSocialSciences,CommunityCollegeofPhiladelphia,USAeDepartmentofAnthropology,UniversityofTexasatAustin,USAarticleinfoArticlehistory: Received28September2012 Receivedinrevisedform 10April2013 Accepted11April2013 Keywords: Limestone Belize Maya MaaxNa GeochemistryabstractThecarbonatebedrockofnorthwesternBelizeispoorlyunderstoodfromthestandpointofboth geochemistryandtheuseofstoneinprehispanicMayasitesforbuildingsandmonuments.Thefriable natureoftherockinthistopographicallyruggedareamakesitespeciallydif culttodistinguishmonumentsfrombedrockspall,aslittlecarving,ifany,survives,andidenti cationrestsonlocationand positioning.Theresearchpresentedhereanalyzed63limestonesamplescollectedfromtwositesinthe ThreeRiversRegionofBelize.ICP-MSandICP-AESwereusedtocharacterizethemajor,minor,andtrace elementchemistryofthelimestonebedrockoftheregionanddeterminetheamountofgeochemical variability.Anotherimportantobjectivewasattemptingtotracethemovementofmonumentstoneand determiningwhetheritwasimportedfromoutsideofthesites.Bedrock,quarries,andpossiblemonumentswereallsampledforthesepurposes.Bedrockprovedtobesimilaracrosswideareas.However,at ChawakBut Â’ o Â’ ob,alongthe ankoftheRioBravo,changesdownslopeinMgconcentrationsuggesta leachingofthebedrockbymeteoricwatersbasedondifferencesinporosity.AtMaaxNa,ahilltopsite,in contrast,suchleachingisnotasapparent.Manymonumentsatbothsiteswerefoundtobecomposedof stonesimilarinchemistrytothelocalbedrock,includingseveraloftheidenti ablestelae.However,our analysesalsorevealedthatafewmonumentsatMaaxNaweremadeofmaterialwithadifferent chemicalcomposition,apparentlyfromstoneimportedtothesite.TheseresultssuggestthattheMaya deliberatelyselectedcertaintypesoflimestoneforcertainpurposes,andmayevenhavetradedinnonlocalrock.Overall,themethodsusedinthispilotstudyindicatethereisrealpotentialinmoreintensive, regionalassessmentsofthematerialsusedatarchaeologicalsites,eveninareaswherethelocalstone doesnothaveadistinctivegeochemicalsignature. 2013ElsevierLtd.Allrightsreserved.1.Introduction Investigationsofstoneusepatternsatarchaeologicalsitesrely oninterpretationsofamodernlandscapethat,inmanycases,has beenheavilymodi edbynaturalprocessessincethetimeofthe sites Â’ abandonment.Therefore,suchresearchrequiresanunderstandingofthelocalgeologyonbothintra-siteandregionalscales. Thisisespeciallyimportantinrainforest-coveredareaslikethe YucatanPeninsulawherealackofbedrockexposuresandrestricted accesshasinhibitedformalgeologicalstudies.Geologicalresearch inthisregion,however,canbene tfromlandscapemodi cations madebytheprehispanicMayaatthearchaeologicalsitesbeing investigated.Theextensiveareastheyquarriedontheperipheryof sitecentersprovidealargenumberofaccessiblebedrockexposures.ThisstudylooksattwoMayasitesindifferentkarstenvironmentsoftheThreeRiversRegionofnorthwesternBelize( Fig.1 ) tocharacterizetheintra-sitegeochemicalvariabilityinthelimestonebedrock,andalsotobegintoaddressdifferencesonaregional scale.TheprimaryobjectiveistoshedmorelightonhowtheMaya selectedandusedstoneresources.Thisisacrucialtopicinnorthern Belize,wherealackofcarvedmonumentshasimpededhistorical understandingoftheMayapresencethereandoftherelationship ofitsinhabitantstootherpartsoftheMayaworld.Thebedrockof thisareaissoftandeasilyweathered,makingitdif culttoidentify megalithsontherainforest oorasfallenmonumentsorrockthat Correspondingauthor.Tel.: 14018746186(Of ce). E-mailaddress: mlbrennan@mail.uri.edu (M.L.Brennan). Contentslistsavailableat SciVerseScienceDirectJournalofArchaeologicalSciencejournalhomepage:http://www. elsevier.com/locate/jas 0305-4403/$ e seefrontmatter 2013ElsevierLtd.Allrightsreserved. http://dx.doi.org/10.1016/j.jas.2013.04.005 JournalofArchaeologicalScience40(2013)3178 e 3192

PAGE 2

hasspalledoffthebedrockbynaturalprocesses.Thegeochemical investigationoflocalvariationsinthebedrockchemistrycanhelp determinethewaysinwhichtheMayamodi edthelandscapeand movedstonetocreatetheirconstructedenvironment.Italsoshows promiseasamethodtobeusedinothergeographicalareaswhere determiningculturalchoicesandpatternsofstoneusehasbeen similarlyproblematic. 1.1.GeologyofnorthernBelize Limestonesarebothapaleoenvironmentaldepositandamodernlandform,asthemorphologyandchemistryoftheserocks re ectbothformationconditionsandtheeffectsofpost-diagenesis weatheringprocesses.Carbonatesdissolveinacidicwaters,butdo notdissolveatequalrates,aprocessdependentonjointsand fracturesinthebedrock,aswellasporosity( Trudgill,1985 ). Geologicalinvestigationofthecarbonatebedrockistherefore importantforunderstandingbothpaleo-andmodernlandscapes. However,thestratigraphicsequenceofrockformationsinnorthern Belizehasnotbeenformalizedandrequiresfurtherstudy( King etal.,2004 ).ThelimestonesofthenortheasternPetnregionof GuatemalaandnorthwesternBelizeappeartobethesameformation,whichwasdescribedseparatelybytwoauthorsastheEl CayoandSantaAmeliagroups( Flores,1952 ; Vinson,1962 ).Ina reportsummarizingthegeologicalsurveysconductedforthe geologicalmapofBelize, Flores(1952) providedonlyinformaldescriptionsofthestratigraphyinnorthernBelize.Basedonthese descriptions,theThreeRiversRegioniscomposedoftheLower EocenelimestoneoftheElCayoGroup.Similarly,inareportonthe petroleumgeologyofthearea, Vinson(1962) de nedtheSanta AmeliaFormationasextendingfromthecentralPetnregionof northernGuatemalatowardthenortheastintothenorthernhalfof Belize.Thus,boththeformationsidenti edseemtoincorporatethe areaofnorthwesternBelize.Whilethiscon ictingnomenclature willnodoubtberesolvedinthefuture,fornowitservestounderscorethelackofdetailedgeologicalknowledgeofthisregion. Onabroaderscale,theYucatanpeninsulaisacarbonateplatformcomposedofshallowwatersedimentsoftheJurassicthrough Tertiaryperiods( Hartshornetal.,1984 ).Theselimestonesare formedfromback-reeflagoonalenvironments(e.g., Kingetal., 2004 ; Vinson,1962 ).AreportonthephysicalpropertiesofYucatanlimestonesdemonstratesalargeamountofheterogeneityin theirphysicalandmechanicalproperties,characteristicofthose formedinshallowtropicalseas( Espinosaetal.,1998 ).Thestepped escarpmenttopographyofthisareaonthenortheasternedgeofthe Petnplateauistheresultoftensionfaulting,whichhas Fig.1. MapoftheThreeRiversRegionofnorthwesternBelizeshowingtheboundaryoftheProgrammeforBelizeConservationandManagementAreaandmajorsite s.Adapted from Scarboroughetal.,2003 M.L.Brennanetal./JournalofArchaeologicalScience40(2013)3178 e 3192 3179

PAGE 3

dramaticallyalteredthecarbonatelandscape( Dunningetal., 2003 ).Weatheringandkarsti cationprocessescausedbythe exposureofunderlyingcarbonatestrataaddtothevariabilityin composition,density,andporosityofthebedrock( Trudgill,1985 ). Sinceitsdeposition,thelimestonehasbeenweatheredintofracturedformationsofridgesandhills.Aroundtheperimeterofthese uplandsareaseriesofswampylowlands( bajos ),whichwereused forresourceprocurementbytheMaya( Dunningetal.,2003 ; Kunen,2004 ; KunenandHughbanks,2003 ). 1.2.StoneuseandtheMaya Therelationshipbetweenpeopleandtheirnaturalenvironment hasdrawnincreasinginterestoverthepastfewdecadeswiththe developmentofenvironmentalandlandscapearchaeologyas complementaryapproachestoinvestigatingpastlandscapes(e.g., DavidandThomas,2008 ; Denham,2008 ; Dincauze,2000 ; Knapp andAshmore,1999 ). AstonandRowley(1974) describe “ landscape ” asapalimpsestonwhichpeopleleavetheirmark,building uponandoverwritingtheimprintofpreviousgenerations.An importantconsiderationforbothenvironmentalandlandscape studiesofabandonedsitesisthatthisprocessextendsintothe presentday.LowlandMayasites,inparticular,haveseenmultiple periodssincetheirabandonmentthathavelefttheirmarkand overprintedmuchoftheoriginalsites,includingrainforestreclamationfollowingtheMayaperiod,deforestationandroad-building intheeighteenthcenturyduetoinlandmahoganylogging( Bolland, 1977 ),andasecondperiodofsubsequentforestre-growth.These factors,alongwithkarstprocessesandweatheringovertime,have createdwhatweseetodayasthemodernlandscapeofthesesites. Themaskingoftheancientlandscapebecomescompoundedfor studiesofMayastoneuseatsitesinnorthernBelize,wherethe limestonebedrockissoftandhighlyweathered,and,thereforeso tooaremonumentsandblocksthatwereshapedfromit,blending alltooeasilyintothecontemporarynaturallandscape. InMayastudies,environmentalandlandscapearchaeologyare closelytied,becauseMayaarchitectureandengineeringtypically builtuponandaccentuated,butrarelycompletelyoverwrote,the naturallandscape( BradyandAshmore,1999 ).Understandinghow theMayavisualizedtheirworldrequiresabroadgraspofhowtheir constructedenvironmentre ectedtheirworldview,includingelementsoftheircivicplansandceremoniallandscapes,whichoften overlapped( Ashmore,1991 1992 2008 ; KnappandAshmore, 1999 ).Inmanycases,naturalfeaturesofthelandscapewerestrategicallyincorporatedintoasite ’ scivicplan,suchaswhenasite wasbuiltaroundacaveasan axismundi ( Ashmore,2008 ; Brady andAshmore,1999 ; Kingetal.,2012 ).TheMayaconstructed environmentwasalsoheavilyin uencedbyquarryingandwatermanagementengineering.Sitesthatusednaturalhills,suchasLa Milpa,requiredarti cialreservoirsbecausetheywerebuiltawayfromwatersources( Scarboroughetal.,1995 ).Quarriesaroundthe peripheryofthesehillscontributedtoconstructionandwereoften subsequentlyincorporatedintothewatermanagementsystem ( Scarborough,1998 ).Forexample,stonefromtheexcavationof reservoirsmightbeusedtoconstructadjacenttemplesandstructures.Atthesametimeotherareasofthenaturaltopographywere engineeredinto atsurfacesbycuttingawayor llingin,inorderto makeroomforstructuresandcontrolthe owofwater.Waterwas notjustapracticalnecessity,however.ThroughoutMesoamerica,it wasassociatedwiththeunderworldandwithalife-giving,sentient Earth accessibleincaves( Brady,1989 1996 1997 ; BradyandPrufer, 2005 ).Importantwatersourcescamefromspringsinwater mountainsatthemouthsofthesecaves( Scarborough,1998 :149). Thecreationofhilltopcentersasarti cial “ watermountains ” ( Scarboroughetal.,1995 ; Scarborough,1998 )thusexhibitsthe Maya ’ sabilitytoshapeandengineertheirnaturallandscapeas physicalrepresentationsoftheirworldview,whileattendingtothe practicalrequirementsofsiteconstruction. ThebuiltenvironmentoftheMayawascomposedpredominantlyoflimestone,withexceptionsinareasofthesouthern highlands,asatCopan,whereigneoustuffwaspredominantlyused ( Abrams,1994 ).Inadditiontoarchitecturalblocks,limestonewas alsoburnedtoproducelime,acomponentofplaster,whichwas usedasasurface nishonarchitectureandplazaswherethe landscapewas lledinandsmoothedover( Villaseor,2010 ). StudiesofMayastoneusehavefocusedonboththeenergeticsof laborrequiredforquarryingandconstruction(e.g., Abrams,1994 ; Folan,1982 ; WoodsandTitmus,1996 )andtheproductionofplaster (e.g., Mathews,2002 ; Villaseor,2010 ).Quarrieswerecommonly positionednearareasofconstruction,suggestingthatstonewould nothaveneededtobetransportedoverlargedistances( Folan, 1982 ).Forexample,atthesiteofKinalinGuatemala,extensive quarryingwasobservedalongthemarginsofthecentralprecinct ( Scarboroughetal.,1994 ).However,studiesofthemovementof stoneperseacrosstheMayalandscapehavegenerallybeenlimited toartifacts.Graniteandobsidianitemshavebeengeochemically tracedbacktospeci cigneoussourcesinthehighlands(e.g., Graham,1987 ; Riceetal.,1985 ),givingagoodideaofstoneuseand tradingpatterns.Limestone,however,likequartzite(e.g., Pitblado etal.,2008 ),ishardertosource,asitlacksacleargeochemical signature.StudiesthroughoutMesoamericahavesuccessfully sourcedthemovementoflimestoneonlyonasmallscale.Barba etal.(2009) usedX-rayFluorescence(XRF)tosourcelumpsof limestonefromplaster oorsatTeotihuacnincentralMexicoto localquarries. Nationetal.(2012) usedInductivelyCoupledPlasma MassSpectrometry(ICP-MS)toestablishtheprovenienceofspeleothemartifactsfromexcavationsintheSibunValleyRegionto speci ccavesafterchemicallycharacterizingthelimestoneof multiplecavesites.Similarly,thisstudyaimstocharacterizethe geochemistryoflimestonebedrockformationsfromtwoMayasites intheThreeRiversRegionofnorthernBelizetoassessthepotential formappingthemovementofquarriedstoneonbothsiteand regionalscales. ThesitesintheThreeRiversRegionofnorthwesternBelizehave beende nedasresource-specializedcommunities,atermthat emphasizesthedependencyrelationshipsbetweensitesofvarying sizesandfunctions( ScarboroughandValdez,2003 ).Thisregionis alsousedasaprimeexampleofaheterarchicalsocialstructure, whichshowshowthesitesinthisareadevelopedinanunranked systemasaresultofitsdispersedresources( Scarboroughetal., 2003 ). KunenandHughbanks(2003) showthatsiteslocatednear importantresourcesdevelopedsystemsofspecializedproduction throughtheprocurementanduseofthoseresources.TheThree RiversRegioncontainsalargenumberofsites,includinglarge centerssuchasLaMilpaandDosHombres,andmid-levelsites, suchasLasAbejasandBolsaVerde,whichmayhaveactedasintermediariesbetweenthelargersites( Tourtellotetal.,2003 ). Signi cantly,ofthesesitesonlyLaMilpahascarvedstelaewith readabletexts,asstoneinthisareaisgenerallytoosofttosupport carving.Interpretationoftherelationshipsamongsiteshasthereforehadtorelyonotherfactors,suchastheirrelativesizeand spacing.Importantfeaturesofthisregionaretheextensivebajo drainageareasthatcontrolledbothsitelocationsaswellasresources. Kunen(2004) de nedthreedistinctzonesforwhichthe Mayahaddifferentusesnearthebajos:residentialonhilltops, agriculturalonterracedslopes,andresource-extractiveinthe interior.Thissamedifferentiationcanbeextendedmorewidelyto incorporatepartsoftheThreeRiversRegionawayfromthebajos,as thepatternsfoundaresimilar.Mostcentersarelocatedonhilltops, afactorthatconstrainedbuilding.ThecoreofsitessuchasLaMilpaM.L.Brennanetal./JournalofArchaeologicalScience40(2013)3178 e 3192 3180

PAGE 4

andMaaxNa,forexample,waslimitedtotheplateausuponwhich theywerebuilt,sothespacehadtobeusedeffectively,planning architecturalconstructionandresourceprocurementaroundone anotherforbothstoneacquisitionandterrainshaping.Terraced hillsidesandlow-lyingbajosprovidedhighlyvariedlandscapesfor modi cationtoaidtheMayasettlementinaseasonallydynamic environmentandforaccesstoresourcessuchasclayandchert.This diversityoflandscapesandresourcesmakesthisregionan importantareaforstudyingtheintersectionoftheMaya Â’ snatural andbuiltenvironments( ScarboroughandValdez,2003 ). Thepresentresearchlooksatthebedrockgeochemistryofthe sitesofMaaxNa,alargecenteronahilltopneartheGuatemalan border,andChawakBut Â’ o Â’ ob,asmallerresidentialsiteneartheRio Bravo( Fig.1 ).Thesesitesprovidegoodaccesstobedrockexposures thatisnotcommonlyfoundinotherareasoftherainforest-covered Petnplateau.Inaddition,theycomefromdifferingenvironments withintheThreeRiversRegionandprovideagoodbasisforcomparison.Themodernconditionofthelandscape e upliftedcarbonatedeposits,weatheredrock,erodedquarryfaces,ruined buildings,andtherainforestcover e re ectssomeaspectsofthe ancient,overprintedlandscape.Geochemistryisrequired,however, tobetoabletodistinguishtheoriginalnaturalenvironmentfrom theonetheMayabuilt.Theobjectiveofthisresearchistoexamine thegeochemicaldatafrombedrocksamplesatthesetwositesto determinetheintra-siteandinter-sitevariability,thencompare thesetosamplestakenfrombuildingblocksandpossiblemonumentsateachsite.Theultimategoalistodeterminewhetherthe materialsusedincarvingandconstructioncomefromlocal quarriesatthetwositesorfromrockforeigntotheirunderlying bedrock.Byexaminingthemodernlandscapesofthesetwoareas throughthegeochemistryofEocenelimestoneformations,wecan begintoinferClassicMayastoneusepatternsandthemovementof quarriedmaterialacrossthelandscape. 1.2.1.MaaxNa MaaxNawasinitiallymappedin1995aspartofaProgramme forBelizeArchaeologyProjectsurveyandtheninvestigatedbythe MaaxNaArchaeologyProject( ShawandKing,1997 ; KingandShaw, 2003 ).Long-termgoalsoftheresearchincludelookingatthesite throughanintegrationofeconomic,political,andideological spheresinordertoexamineitsmultidimensionalrelationships withsitesintheregion( KingandShaw,2003 ).MaaxNa,anEarlyto TerminalClassiccenter(AD250 e 900),followsapatternofconstructionsimilartootherlargesitecentersintheThreeRivers Region,suchasLaMilpaandDosHombres( Fig.1 ),aswellasto centersintheneighboringPetnareaofGuatemala.Itexhibits someuniquefeatures,however.Mostnotablearealarge,centrally locatedreservoiranddamandtheexpansiveNorthPlaza( Fig.2 ), whichislargerthanthoseatallothercentersintheregionwiththe exceptionofLaMilpa Â’ s( Shaw,1999 ).Thisplazaalsoappearsempty Fig.2. MapofMaaxNasitecoreindicatinglocationofsamples.MapbyL.Shaw,samplelocationsbyM.Brennan. M.L.Brennanetal./JournalofArchaeologicalScience40(2013)3178 e 3192 3181

PAGE 5

inrelationtoplazasatthoseothersites;structuresborderit,but withalessdensespacingthaniscommonelsewhere.Also,unlike sitessuchasDosHombresandLaMilpa,MaaxNa Â’ smaintempleis isolated.Positionedtothesouthofthesitecore,itisjuxtaposedto theNorthPlaza,inlinewiththeballcourt( KingandShaw,2003 ).To thewestofthetempleandNorthPlazaliesalarge,mostlyenclosed courtyardwithassociatedbuildingsknownastheWestPlaza.The NorthandWestPlazasarelinkedbyacausewaythatextendspast themaintempleanduponwhichStela2,theonlystandingstela remainingatthesite,waserected( Fig.2 ).MaaxNadatestothe EarlyClassic(AD250 e 600),withpossiblesmall-scaleoccupationin theLatePreclassic(400BC e AD250),andappearstohavebeen abandonedearlyintheLate-to-TerminalClassic(ca.AD700 e 800), beforetheotherlargesitesintheregion,afactorthatmayinpart accountforthepaucityofconstructionontheNorthPlaza( King andShaw,2003 ; Kingetal.,2012 ). Asnotedearlier,theMayaplacementofarchitectureandtheir constructedlandscapesweregreatlyin uencedbycosmology ( Ashmore,1991 1992 ).ThelayoutofMaaxNafollowscommon trendsinsiteplanningforthispartoftheMayaworld,includinga strongnorth e southaxislinkedbyaballcourtandtheuseof causewaystolinkgroupsofstructures( Ashmore,1991 1992 ; Houk, 1996 2003 ).However,thistemplateappearstohavebeenmodi ed toaccommodatethenaturaltopographyatthesiteandhighlight certainfeaturesofculturalsigni cance.MaaxNahasanumberof cavesandrockshelters,whicharelinkedtoasentient Earth andthe underworldinMayacosmology( Brady,1989 1996 1997 ; Bradyand Prufer,2005 ; Kingetal.,2012 ).MaaxNaappearstohavebeen intentionallypositionedinalignmentwiththelargestoftheseand awayfromthebulkofitsresidentialandagriculturalareas( King andShaw,2003 ; Kingetal.,2012 ),anotherfactoraccentuatingits distinctivelayout. SurveysatMaaxNahavedocumentedareasontheperipheryof thesitecorethatincludealargenumberofwhatareinterpretedto bequarrylocations.Twoknollsnearthereservoirappeartohave beenquarried,thoughsomeofthemodi cationsvisiblemaybe fromloggingroadconstruction,andotherquarrieshavebeen locatedtothewestandnorthoftheNorthPlazaandeastofthe maintemple.Otherpartsofthesitefartherontheperipheryappear tohavebeenmodi ed,too,likelyinpartduetoquarryingactivities. Small-scalewatermanagementfeaturesmayhavebeenassociated withsomequarries,askeepinglimestonemoistallowsforeasier cutting( Scarboroughetal.,1995 ).ThesefeaturesareexposedatMaaxNatoagreaterextentthanatmanyothersitesinthearea,in partbecausetheymaynothavebeenaswellmaskedbyarchitectureduetothesite Â’ srelativelyearlyabandonment.Forthisreason, MaaxNaistheidealsitewithintheThreeRiversRegionforthis study,asitslocationonahilltopandnumerousquarriesaroundits edgesallowforbroadsamplingacrosstheareaofthesitecore. 1.2.2.ChawakBut Â’ o Â’ ob ChawakBut Â’ o Â’ obisasmallresidential,agriculturalsitelocated ontheinclinedslopeoftheRioBravoEscarpmentabout7km southeastofMaaxNa,acrosstheRioBravofromthelargecenterof DosHombres( Fig.1 ).Thismid-to-lateLateClassic(AD600 e 800) siteconsistsofresidentialterraces,densecommonerhousing,anda seriesofcomplexwatermanagementfeatures( Walling,2005 2011 ; Wallingetal.,2006 ).Researchherehasuncoveredprimarily utilitarianartifactsandnoformalritualorelitestructures.The numberofhousesconstructedintheshortperiodoftimethesite wasoccupiedsuggeststhattherewasarapidpopulationgrowth, drivenbyanin uxofimmigrantstothearea( Walling,2005 ).A ballcourtatthesouthernextentofthesiteisoneofthefewceremonialconstructions.Itwasbuiltastwomoundsofsmallcobbles andappearstohavehadwaterrunningthroughit.Theballcourtis inlinewithoneofthetwocavespresentatthesite,whichcanbe seentothewestoftheballcourt. Theslopeofthesiteontheescarpmentplacesitinanactive karstregionwherewater owsthroughthebedrocktowardtheRio Bravo.TheMayaconstructedaseriesofterraces,collectionsurfaces,interconnectedwaterbasins,andacentralreservoirtodirect thewatertotheoccupationandagriculturalareasfurtherdownthe escarpment( Walling,2005 ).Thissiteservesasagoodcomparison toMaaxNaontwofronts.First,thesmallsizeofChawakBut Â’ o Â’ ob,in comparisontothelargecenterofMaaxNa,showcasesadifferent aspectofMayaculture.Second,thecarbonateenvironmentof ChawakBut Â’ o Â’ obdiffersdramaticallyfromtheplateauuponwhich MaaxNasits,andisdominatedbyridgesanddrainagegulliesfrom themovementofwateroverthebedrock.However,bothsiteslie withinthesamedrainagesystemoftheRioHondo,socomparisons arebetweendifferentkarstlandscapes,butoverasimilarhydrologicalregionandwithinthesamelimestoneformation. 2.Methods 2.1.Sampling 2.1.1.MaaxNa Sincefewstudieshaveattemptedthegeochemicalsourcingof limestonefromanarchaeologicalcontext,thegoalofthisresearch was rsttodetermineifdiscriminationbetweendifferentbedrock sampleswaspossible.Asurveyofthelimestoneresourcesand featuresatMaaxNawasundertakenduringthe2005and2006 eldseasons.Tobeginthisprocess,limestonesampleswere collectedthroughoutthesitecoreofMaaxNa( Fig.2 ),including multiplelocationswheresamplesweretakenbothfrommegaliths identi edaspossiblemonumentsandfromtheadjacentbedrock. Thesampleswerechiseledoffthestonewithahammerand collectedinlabeledbagswithbothadescriptionofthestoneand thezonecodefortheareaofthesite.Inthecaseofknownmonuments,weonlytookasampleofthestoneifpieceshadalready spalledorbrokenoffandcouldbecollectedwithoutfurther damagingthemonument.Atleastonesamplewastakenfromas manyoftheaccessiblebedrockoutcropsaspossiblewithinthe 1km2areaofthesitecore.Afewduplicatesampleswerediscarded inthe eldduetoweightlimitationsfortheirexport. SevensamplesweretakenfromandaroundMegalith1,whichis locatedsoutheastoftheWestPlazaadjacenttoalinearseriesof chultuns( Shaw,2002 ).NinesamplesweretakenfromtheunderlyingbedrockandpossiblemonumentsontheSouthAcropolis, includingthreesmallmegaliths(apossiblealtarca.50cmin diameterandtwopossiblestelae)thatmayhavebeenfacedwith plaster,andapossiblebrokenstelanearby.Sixsamplescamefrom megalithsandoutcropseastofthemaintemple,includingalarge sectionoflimestonethatmayhavebeenupliftedbyatree-fall,and alsofromStela1,oneofthefewpositivelyidenti edstelaeatthe site.IntheNorthPlaza,atotalofsevensampleswerecollected, includingonefromabeveledstoneremovedfromthefaceofthe ballcourtduringexcavationsin2003.Otherlocationssampledhere includeapossiblealtarintheplazanorthoftheballcourt,apossible monumenteastoftheNorthHill,andsamplesfrominsideand outsideSpiderCave,thelargestcaveatthesite,whichisbeneath theEastStructure( KingandShaw,2003 ; Kingetal.,2012 ).Six sampleswerecollectedfromaquarriedareanorthoftheNorth Plaza,includingtwopossiblemonuments.The nalareasampled duringthissurveyistheWestPlaza,wheresevensampleswere collected.FeaturesincludedAltar2inPlazaD,aquarriedarea northwestoftheWestAcropolis,twoquarriedknollsborderingthe reservoirentrance,andtheeasternmostofthreepossiblealtars fromsouthoftheSouthBuildingoftheWestPlaza.ThesethreeM.L.Brennanetal./JournalofArchaeologicalScience40(2013)3178 e 3192 3182

PAGE 6

small,roundstones,whicharegroupedtogether,aresimilarinsize andshapetothepossiblealtarsampledontheSouthAcropolis.In surfaceappearancetheyresembleboththatmegalithandthe possiblestelaewithwhichitisassociated. 2.1.2.ChawakBut Â’ o Â’ ob InJune2007,asimilarsurveyofthelimestonebedrockoutcrops andpossiblestonemonumentfeatureswasundertakenatChawak But Â’ o Â’ ob( Fig.3 ).Twenty-twosampleswerecollected,primarily frombedrockoutcropsalongtheslopedterrainofthesite.Samples ofculturalfeaturesincludedthreecobblesfromtheballcourt,a smallspalled-offchipfromafallenstelainGroupA,andpossible plasterfrombeneaththismegalith.Drainageandotherwater managementfeaturesatthesitewereincludedinthesampling,as wasaknollwestoftheballcourt,andoutcropsalongtheedgeofthe reservoir.AsatMaaxNa,sampleswerealsotakenfrominsideand outsidethecavejustwestofChawakBut Â’ o Â’ ob Â’ sballcourt. 2.2.Testanalyses Theinitialsetof37samplescollectedatMaaxNain2005was thinsectionedandexaminedunderapetrographicmicroscope.All ofthesamplesarecomposedentirelyofmicrocrystallinecalcite (micrite),withonesamplethathadsmallveinsofchertinthe micrite.Thelackofmicrofossilsinthelimestoneforidenti cation informedusthatgeochemicalanalysiswouldbenecessaryforthis study.Sixofthesesamples,amixofpossiblemonumentsand bedrock,wereselectedfortestanalyseswithascanningelectron microscope(SEM)andICP-MSbasedonthevariedcolorandhardnessoftherocksinhandsample.TheSEMresultsprovidedbulk chemistrydata,whichshowedthatMgwasanimportantelementin thesesamples,butthatadditionalminorandtraceelementswould benecessarytodifferentiatebetweenthem.ICP-MSisageochemical methodusedtoquantifytraceelementsindissolvedrockmaterial andiscapableofverylowdetectionoftraceelements.Ionsfroma dissolvedsampleareextractedfromaplasmamatrixintoapumped vacuumandfocusedwithalensintoamassspectrometer ( Rollinson,1993 ).ThetestusingICP-MSshowedthatSrandother minorelements(e.g.,therareearthelements,orREE),mayalsobe usefulindiscriminatingbetweensamples.Thismethodwasthereforeselectedforthecurrentstudyoverlaserablationandelectron microscopyanalyticalmethodsbecausethedissolutionofthe samplesallowsforanalysisofthebulktraceelementcomposition ratherthanaseriesofsmallpointsonthesample,thusenablinga betterdeterminationofaspecimen Â’ sbulkchemistry.Forthesame reason,atomicemissionspectroscopy(ICP-AES),whichmeasures bulkelementchemistry,wasusedforthisstudyinsteadoftheSEM. 2.3.Solutionpreparation ICP-MSandICP-AESanalysesrequiresamplestobedissolved entirelyintosolution.Inordertoensurethatallcomponentsof eachsample,includingpotentialnon-carbonateinclusions,were dissolved,thefollowingprocedureswerefollowed.First,arock hammerwasusedtobreakuptherocksamples,andpieceswere selectedthatdidnotcontainanyexterior,weatheredsurfaces.A mortarandpestlewerethenusedtogrindthestoneintoa ne powder,whichwasplacedintoa2mlcentrifugetubeforstorage. Themortarandpestlewerecleanedwithmethanolbetweeneach sample.Inacleanlab,0.05gofeachpowderedsamplewas weighedintoa23mlTe onSavillexbeaker,followedby2mlof nitricacid(HNO3)and2mlofultrapurewater,thenplacedonahot plateat < 100Covernight,followingmethodsforsediment dissolutionoutlinedin Kelleyetal.(2003) and Planketal.(2007) Sevenstandardscommonlyusedforcalibration(AGV-1,EN026100-3,JCP-1,NBS-88a,RGM-1,W-2,andeitherIOFMorJB-3)and aproceduralblankweredissolvedatthesametimeasthelimestonesamplesfollowingthesameprocedures.Oncecooled,each solutionwastransferredtoa2mlcentrifugetubeandcentrifuged at11,000rpmfor5min.Thesupernatantliquidwastransferredto 250mlNalgenedilutionbottles.Eachtubewasrinsed3 with ultrapurewater,andtheremainingsolidmaterialinthemwasthen returnedtotheTe onbeaker. Sincemicrocrystallinesilicawasobservedinsomeofthesamplesinthinsection,1.5mlofhydro uoricacid(HF)wasaddedto theundissolvedsolidalongwith3mlofHNO3andplacedonahot plateovernightat < 100C.Thebeakerswerethentappeddown andthecapswereremovedtoallowfortheHFtoevaporateoff. Fig.3. TopographicmapofChawakBut Â’ o Â’ obsitecoreindicatinglocationofsamples. MapbyC.Coughenour,samplelocationsbyM.Brennan. M.L.Brennanetal./JournalofArchaeologicalScience40(2013)3178 e 3192 3183

PAGE 7

Followingthis,1.5mlofhydrogenperoxide(H2O2)and1.5mlof HNO3wereaddedtodissolveanyorganicmatterpresentinthe samples,whichwereagainheatedovernight.Calcium-richspikes werecalculatedforthefournon-carbonatestandardsinorderto matrix-matchtheigneousstandardstocalcite.Weaddedthe appropriateamountof10,000ppmCasolutiontoeachofthesefour samplestotheequivalentof56wt.%CaOintherock(equivalentto purecalcite).Halfoftheproceduralblankwasspiked,whilethe otherwasnot.Theremainingliquidinthebeakerswasaddedtothe dilutionbottles,thenthebeakerseachrinsedthreetimeswith MilliQH2O.Ultrapurewaterwasaddedtobringeachsolutionto 150gfora3000 dilution.Thebottlesweresonicatedfor5min and15mlofthesolutiontransferredtocentrifugetubesfor analysis. 2.4.ICP-MS ThesolutionswereanalyzedwithaThermoX-Series2quadrupoleICP-MSwithanESISC-2autosamplerattheUniversityof RhodeIsland Â’ sGraduateSchoolofOceanography.Eachsamplewas analyzedtwicetoconstrainanalyticalprecision.Theconcentrations of43elementsselectedforpotentialimportanceincarbonate mineralsweredeterminedforthesamples(Li,Be,K,Sc,Ti,V,Cr, Mn,Fe,Co,Ni,Cu,Zn,Ga,As,Rb,Sr,Y,Zr,Nb,Sb,Cs,Ba,La,Ce,Pr,Nd, Pm,Sm,Eu,Tb,Gd,Dy,Ho,Er,Tm,Yb,Lu,Hf,Ta,Pb,Th,andU). ResultsforBe,K,Ti,As,andSbarenotreportedbecausetheywere belowtheinstrument Â’ sdetectionlimits.EitherIOFMorEN026 standardswereanalyzedperiodicallythroughouteach10-sample runtocorrectfortheinstrument Â’ sdrifttoimproveaccuracy. 2.5.ICP-AES Finally,thesolutionswereanalyzedwithaJY2000UltraceICPAESwithaJYAS421autosamplerand2400g/mmholographic gratingatBrownUniversity.Theconcentrationsofsevenelements thataremajorelementsincarbonateswereanalyzedsimultaneously(Na2O,MgO,Al2O3,SiO2,CaO,MnO,andFe2O3).Fivesyntheticsolutionsofvaryingconcentrationsoftheseelementswere madeforcalibratingtheinstrument.ResultsforNa2OandSiO2are notreportedbecausetheywerebelowtheinstrument Â’ sdetection limitsformostofthesamplesandthedissolutionmethoddoesnot conserveSiinsolution.ThestandardsanalyzedwiththeICP-MS analysiswerealsoanalyzedwiththeICP-AESandtheresults fromthesesolutionswereusedtocalibratetheresultsfromthe limestonesamples.Calibrationcurvesarelinear( r 0.9993)and replicateanalysesare < 3%RSD. 3.Results 3.1.Rocksamples ThelimestoneoftheThreeRiversRegionofnorthwestern Belizeissoft,andmanyofthesamplescollectedexhibitedthe commonwhite,soft,micritictexture.Anumberofthebedrock samplesfromMaaxNa,however,werecomposedofadenser,tan limestone.Theselessporousrockscomprisethefewnaturalhills atthesite,includingtheNorthHill,twosmallknollsalongthe southeastedgeofthereservoir,thebedrockofthewestacropolis, andasmallhillsouthoftheWestPlaza.Afewsamplesfrom ChawakBut Â’ o Â’ obhadsmallblackinclusionsinthemicritematrix, whichwereidenti edwiththeSEMasmanganeseoxides,and correspondtoaharderwhiterock.Thistypeofrock,asopposedto themorecommon,softerlimestoneatChawakBut Â’ o Â’ ob,occurs onlyalongtwoofthethreeridgesthatrunacrossthesiteE e W.At bothsites,thesehardertypesoflimestonearethusfoundonthe naturallyelevatedfeaturesonthelandscape,whichislikelydueto thefactthatthelowerporosityoftheserocksallowedforless chemicalerosion. 3.2.Geochemicaldata Thegeochemicaldataprovideavery ne-scaledlookatthe carbonatechemistry,inbothmajorandtraceelements.Thenature ofthesedataissuchthatevensamplesfromthesamelocationmay nothavecompletelyidenticalresults.Therefore,astatisticalexaminationofsomeofthegeochemicaldataisnecessarytobeable todrawappropriatecomparisonsbetweensamplesetsandsites. Foreachsite,theresultsfromthebedrocksamplesalonewere examinedtodeterminethevarianceinbedrockchemistryacross thesite.Thecoef cientofvariation(CV),orpercentrelativestandarddeviation(%RSD),wascalculatedforeachofthe38reported elementswithinthebedrocksamplesfromeachsite.Thisvalueis thestandarddeviationofeachelementdividedbythemean.Thisis animportantcalculationbecausethevarianceofatraceelement, suchastheREEs,couldbehigh,butitsstandarddeviationwouldbe ordersofmagnitudelowerthanthatofaminorelement,suchasSr. TheCVinsteadre ectsthevariationperppmofeachelement, showingtherelativevarianceandillustratingwhichelementsvary themostinthebedrockofthesites. Figs.4 and 5 showgraphsofthe CVforMaaxNaandChawakBut Â’ o Â’ ob,respectively.Lowvalues re ectlowvariance.ElementssuchasCa,Mg,Sr,Cr,andNithat havelowCVvaluesarebetterelementstouseforcomparing possiblemonumentstobedrocksamples,relativetothosewitha Fig.4. Plotofthecoef cientofvariation(CV)forelementsfromMaaxNasamples. M.L.Brennanetal./JournalofArchaeologicalScience40(2013)3178 e 3192 3184

PAGE 8

highCV,suchasRb,Ba,orPb,becausethelowCV-elementsre ecta betteraveragestandardchemistryofthesites Â’ bedrock.The objectiveofthisworkisdeterminingtheexpectedvarianceover thespaceofeachsite Â’ sbedrock,whichisbestdonewiththeelementsthatvarytheleastforthepurposesofcomparingtopossible externalmaterial. 3.3.MaaxNa Itisexpectedthatthechemistryofthelimestonebedrockwill varyacrossthespaceofasiteduetotheepigeneticprocessesat worksinceitsdeposition.Forty-twolimestonesamplesfromMaax Nawereincludedinthisstudy(bedrocksamples n 18,possible monumentsamples n 24).As Fig.4 shows,elementsvary differentlyamongthebedrocksamplesanalyzed.Highlyvariable elements,suchasBaandPb,shouldnotbeusedforcomparison,as explainedabove,sotheirgeochemistrywillbediscussedseparately inSection 4.1 .Certainelementsimportantincarbonateprocesses, suchasCa,Mg,Sr,Cr,andNi,arethosethatexhibittheleastamount ofvariance.Speci cally,rangesinMgconcentrationsinlimestones indicatethedegreeofdolomitizationinthisformation.TheMaax Nabedrockrangesbetween3and22wt%MgO,allofwhichfall withintherangeofthedolomiticlimestoneclassi cation( Trudgill, 1985 ),withanaverageof14.5%MgO,correspondingto28%dolomite(MgCO3).OnesamplehasalowerconcentrationofMgdueto largeamountsofmicrocrystallinesilicaintherock,butitisnonethelessstilldolomitic.NoneoftheMaaxNabedrocksamples indicateapureCaCO3limestonecompositionwith < 5%MgCO3. Toassessthepossibilityofsourcingstonetowhereitwas quarried,asimilarityindexwasappliedfordatafromallsamples basedontheanalysisofsimilarityby Borchardtetal.(1972) .The similaritycoef cient( d(A,B))indicatesthesimilarityofeachsample comparedtoallothers.Avalueof1.0indicatesperfectlysimilar samples.However,thismethodwasdevelopedforcomparing volcanicashlayers,whichhavespeci cchemicalsignaturesthat canresultinasimilarityof d 1.0( Borchardtetal.,1972 ).Sucha similarityisnotexpectedforlimestones,whicharemorechemicallyvariable.All38elementsreportedinthisstudywere compared,aswellasthesetofsixelementswithlowCVvalues,to differentiatebetweensamplesmoreclearly.Certainsamplesfrom MaaxNathatcontainhighamountsofsilica(e.g.41and44)have lowerconcentrationsofCaOandMgOandhavelowsimilaritywith otherbedrocksamples( d(9,44) 0.484),butarestillsimilarrelative tooneanother,whichiswhyratiosofCaO/MgOwereusedfor comparativepurposes. Thesimilarityindexillustratessomeinterestingdistributional patternsamongthebedrocksamplesatMaaxNa.Twosamplesof thebedrockfromtheSouthAcropolisatMaaxNa(9and16; Fig.1 ) aresimilarat d(9,16) 0.706,whereassamples50,fromsouthofthe WestPlaza,and53,fromnorthoftheWestAcropolisaresimilarat d(50,53) 0.701,showinglittlespatialvariation.Twosamples(29 and30)weretakenasacontrolforweatheringfromtheinside oor ofSpiderCaveaswellastheoutsidefaceabovetheentranceofthe cave.Thesetwosamplesaresimilarat d(29,30) 0.726.However,the exteriorsampleismoresimilarat d(29,52) 0.846withasample (52)fromamegalithabovethecave.Thesamplefromtheinterior ofthecavecamefromwhatwasidenti edasatravertinesurface duringtheexcavationofthiscavein2008( Kingetal.,2012 ). Travertineisformedfromthemovementoffreshwaterovercarbonate( Flugel,2004 ).Theinteriorsampleisdepletedinallthe elementsrelativetotheexteriorsample,withtheexceptionof calcium,suggestingitisamorepurecarbonateformedduetothe leachingoftraceelementsfromtherockbythemovementof meteoricwater,asillustratedby Fig.6 3.4.ChawakBut Â’ o Â’ ob ThesamplescollectedatChawakBut Â’ o Â’ obshowmoreheterogeneitythanthosefromMaaxNa.Twenty-onelimestone sampleswereanalyzed(bedrocksamples n 15,possible monumentsamples n 6). Fig.5 plotstheCVfortheChawak But Â’ o Â’ obbedrocksamples,whichshowsasimilarpatterntothat fromMaaxNa,withCa,Mg,Sr,andNiamongtheleastvariable elements,andRboneofthehighest.TheMgconcentrationsat thissitearesomewhathigherthanatMaaxNa,rangingfrom5 to22wt%MgOwithanaverageof17.4wt.%,or33%dolomite, alsoallwithintherangeofdolomiticlimestone( Trudgill,1985 ). ChawakBut Â’ o Â’ obsitsonalowerescarpmentthanMaaxNa,near thebanksoftheRioBravo,andwhereMaaxNawasconstructed onaplateau,thissmallersiteissituatedinamoreactivekarst landscape,heavilyaffectedbythemovementofwaterdownslopethroughthebedrock.AsatMaaxNa,theharder,denser rocktypemakesupthehigherreliefatChawakBut Â’ o Â’ ob.However,unlikeatthathilltopcenter,Mgconcentrationsdecrease withelevation.ThehighestMgOwassample1atthe100m contouratthesitewith22.2wt.%.Downslopetothewestandin Fig.5. Plotofthecoef cientofvariation(CV)forelementsfromChawakBut Â’ o Â’ obsamples. M.L.Brennanetal./JournalofArchaeologicalScience40(2013)3178 e 3192 3185

PAGE 9

thedrainagearroyointhesouthsamplesrangedfrom5.2to 11.4wt.%.Themoreporous,softertypeofstonefoundoffthe ridgesatthesiteismoreeasilyeroded,whichleachesMgout. Fig.7 showsthepositiverelationshipofMgOconcentrationwith elevationatChawakBut Â’ o Â’ ob. SimilaritytestswerealsorunforallthesamplesatChawak But Â’ o Â’ ob.Asetofcontrolsampleswastakenfromtheinteriorand exteriorofthecaveneartheballcourtatChawakBut Â’ o Â’ ob,which aresimilarat d(15,16) 0.811.Inthiscave,theinteriorsamplewas takenfromtheceilingofthecave,asthereisnoevidenceofa travertinesurfacehere.Contrastsbetweenbedrocksampleswere strongerwithelevationthandistance.Sample1,fromGroupEat thenorthwesternpartofthesiteatthe100mcontour,andsample 17,collectedfromahillatthesouthwestofthesite,bothofwhich havesimilarlyhighMgOconcentrations,aresimilarat d(1,17) 0.832.Thesetwosampleswerefromoppositeendsofthe site( Fig.3 ),butatsimilarelevations.Incontrast,sample6fromthe easternendofthesiteanddownslopecorrelateswithsample1at d(1,6) 0.646.Elevation,porosity,andsubsequentelementleaching apparentlygovernthegeochemistryofthelimestoneatthissite. Thedifferencesbetweenthetwotypesoflimestoneobservedin handsampleareinporosityanddensity,aswellasMnO concentrations,duetotheMn-oxideinclusions.Thethreesamples withthehighestMnOconcentrations e 4,10,and14 e arethose thathadthemostvisibleblackoxideinclusions.Thesemineral inclusionsarevisibletothenakedeyeinhandsample,andare thereforeusefulforidentifyingthistypeinthe eld. 4.Discussion 4.1.Limestonegeochemistry Forthisresearch,itisimportanttounderstandcarbonate geochemistryinordertointerpretthevariationsindatafrom samplescollectedacrosslargeareasofthetwosites.Thebehavior ofminorandtraceelements,aseitherpartofthecalcitecrystal latticeorasinclusionsinthelimestoneitself,areimportantfor distinguishingbetweendifferentlimestonesamplesassuchelementsre ectdifferentformationandepigeneticconditions.Not onlyischaracterizingthespatialgeochemicalvariationsatthese sitesimportantforunderstandingtheevolutionofthesurface bedrock,butalsofordistinguishingbetweenrocknaturaltothe localbedrockandrockmovedintoplacebytheMaya.Magnesium andotherdivalentmetals,suchasSr,areabletoreplaceCainthe calcitelattice.Theconcentrationsoftheseelementsinlimestone dependonmanyfactors,includingthetemperature,salinity,and chemistryoftheseawaterduringformation( Flugel,2004 ).Postdiageneticprocesses,suchasthemovementofwateroverthe carbonatelandscape,alsogreatlyaffectthechemistryofmodern limestoneformations.MeteoricwaterhasaverylowMgO/CaO ratio,andthereforeleachesMgfromcalcite,whichoccurstoa greaterextentinmoreporousrock( Flugel,2004 ).Thisexplainsthe higherlevelsofMginthedenserrocksoftheridgesatChawak But Â’ o Â’ ob,andthedepletionofMginthesofterrocksin-between. Thelatticestructureofthecalcitecrystalishighlyrobustand canabsorbavarietyofdivalentmetalcations,inadditiontoMgand Sr,withoutbecomingdisordered( FordandWilliams,1989 ). Reeder etal.(1999) showedthatdivalentCo,Zn,Pb,andBacommonly substituteintheCasiteofthecalcitestructurewithvaryingdegrees oflocaldistortion,eitherdilationaroundBaandPborcontraction aroundthesmallerions,CoandZn.Thismodeofsharingcationsof Fig.7. PlotofMgOwt.%vselevationforsamplesatChawakBut Â’ o Â’ ob. Fig.6. Traceandminorelementplotofsamples30and52frominsideandaboveSpiderCave. M.L.Brennanetal./JournalofArchaeologicalScience40(2013)3178 e 3192 3186

PAGE 10

varyingsizesisfavoredbythecalcitelatticebecauseitscornersharingstructurecantakeinionslargerorsmallerthancalcium ( Reeder,1996 ; Reederetal.,1999 ).Othermetalscommonlyadsorb ontothecalcitecrystalsurfaceratherthanbeingincorporatedinto thecrystallattice.Studieshaveinvestigatedthebehaviorsofelements Â’ interactionwithcalciteincludingBa,Sr,Cd,Mn,Cr,Zn,Co, andNi( Garcia-SanchezandAlvarez-Ayuso,2002 ; Lakshtanovand Stipp,2007 ; Zacharaetal.,1991 ).Ahydratinglayerofwatercontainingtracemetals,adsorbedontothecalcitesurface,initially interactswithpartiallyunsatis edchargesoftheCaandCO3ions, whichresultsinahydrolyzedlayeronthemineralsurfacethatcan becomeincorporatedintothestructureuponrecrystallization ( LakshtanovandStipp,2007 ).However,theelementsthatsubstitutedirectlyintotheCasiteinthelatticearethemostcommon minorcomponents.TheresultsfromMaaxNaandChawakBut Â’ o Â’ ob showlowvarianceforMgandSr,buthighCVvaluesforBaandPb. ThesevariableelementshavemuchlargerionicradiithanCa, requiringthecalcitelatticetodilatearoundthesiteofincorporationinordertoallowforcoprecipitation( Reederetal.,1999 ),which explainsthehighvariabilitybetweensamplesfromacrossthearea ofthesites Â’ bedrock,astheconditionsforacceptingaBaorPbion intothemineralaremoredemandingthanthoseforMgandSr,or smallerradiuselementsthatcan llthatspace,suchasCoandZn. Therareearthelements(REE)areindicatorsofthechemical evolutionofwaters owingthroughcarbonates( Vanimanand Chipera,1996 ).TheTertiarycarbonatesedimentsontheYucatan peninsulabecameemergentbythelateOligoceneandhavebeen exposedtosubaerialprocessessincethen( Kingetal.,2003 ).The identi cationofallthesamplescollectedfromthesesitesas microcrystallinecalciteshowsthatthemovementofmeteoric wateroverandthroughthecarbonatelandscapesincediagenesis hasrecrystallizedtheinitialrockstructureintomicrite,leavingno microfossilspreserved.Whiletherearemanyepigeneticfactors involvedinsuchprocesses,recrystallizationcausestheleaching andmigrationoftraceandminorelements( Wolfetal.,1967 ).REE concentrationsarebelievedtobealteredbydolomitizationaswell asotherepigeneticweatheringprocesses,buttheexactrelationshipisunclearandunderstudy(e.g., Northdurftetal.,2004 ). Jietal. (2004) showthatweatheringofdolomitescausesuneventransport ofREE,whichisdependentontheclimaticweatheringconditions andthestabilityoftheREE-carryingrock.CeriumanomaliesinREE pro lesre ecttheoxidationofCeanditssubsequentincorporation intomanganesenodules,depletingCeinseawater( Piper,1974 ). However,thesamplesinthisstudydonotdepictaconsistent depletioninCe,asshownintheplotofthemonumentsamples fromMaaxNain Fig.8 ,furtherdiscussedbelow.InthecaseofMaax NaandChawakBut Â’ o Â’ ob,theporosityofthebedrock,whichwe haveshownvariesatbothsites,greatlyaffectsthedissolutionand weatheringrates. Strontiumisalsoamobileelementduringcarbonateweathering anddissolution(e.g., VanimanandChipera,1996 ).ApositivecorrelationbetweenSrand P REE,asshownin Fig.9 fortheMaaxNa bedrocksamples,re ectstheepigeneticin uenceofmeteoric watersmovingoverthebedrock( TligandM Â’ Rabet,1985 ).The differencesshowninthetraceandminorelementcompositionof thebedrockatthesesitesisdemonstratedintheBa,Rb,andPb concentrations,andREEpro lesillustratethevariablenatureof bedrockcomponents.Thisiswhyanunderstandingofthebedrock Fig.8. REEplotofMaaxNamonumentsamples.Colorsinwebversion.Redlines(MN18,31,36)representthreestela;orangelines(MN24,25)aretheballcourtan dNorthPlaza altar;greenlines(MN10 e 15,43)arethesevensmallmonuments;possiblemonumentsamplesareinlightblue(MN35,37,38,52);anddarkbluelines(MN2,4 e 6,8)show samplescollectedfromMegalith1. M.L.Brennanetal./JournalofArchaeologicalScience40(2013)3178 e 3192 3187

PAGE 11

geochemistryisessentialtotheidenti cationofasetofelements appropriatefordiscriminatingimportedfromlocalrockateachof thesitesbeinginvestigated.Forexample, Nationetal.(2012) used RbandNiandsomeREEtosourcespeleothemsintheSibunValley, buttheseelementswouldnotapplytoMaaxNabecauseRbandthe REEaremorevariablethereduetounevenweathering.Thecombinationofelementsthatbestrepresentsthesites Â’ bedrockderives fromthosewithlowCVs(see Figs.4 and 5 )withinthecontextofthe geologicalframeworkdiscussedabove.Theternarydiagramin Fig.10 showsthebedrocksamplesplottedinblue.CaO/MgOindicatesthedolomitepercentageaswellastheleachingofMginthe softerrocks.TheSr/LaratiolinksthemobilityofSrandtheREEs duringchemicalweathering.Finally,Cr/Ni,thetwotransition metalswiththelowestCVvalues,representtheelementsadsorbed andincorporatedintothecalcitecrystal.Theternaryplotclearly illustratesaclusteroftheMaaxNabedrocksamples,whichcanbe usedasastartingpointforcomparisontosamplesfrommegaliths andpossiblemonuments. 4.2.Sourcingmonuments Sincelimestonevariessowidelyduetoformationconditions andpost-diagenesisprocessesaffectingitschemistry,absolute sourcingmethodslikethoseusedforobsidiancannotbeappliedto limestoneartifacts.Identifyingaspeci cgeochemicalsignaturefor MaaxNaandChawakBut Â’ o Â’ obisthereforenotanendgoalofthis pilotstudy.Rather,weproposeamultifacetedapproachincorporatingbothgeochemistryandarchaeologytomakeinferences aboutstoneuseatthesesites.Usingthesetofelementsidenti ed aboveasmostimportantinthecarbonategeologyofthebedrock samples,thisresearchcomparesthemtothesamplescollected frommegalithsandpossiblemonumentsateachsite.Theprovenienceofeachsamplewithinthesiteandanyotherarchaeological informationaboutitarethenassessedtohelpplacethe geochemicaldataincontext. 4.2.1.MaaxNa Aseriesofmonuments,megaliths,andotherstonefeaturesat MaaxNaweresampledtoprovideacomparativedatasettothatof thebedrockinordertodetermineifgeochemistrycouldhelp identifymonumentsand,potentially,sourcethestonetolocal quarries.Itwasexpectedthatmanyofthemegalithswerespall fromthebedrockandwouldhaveverysimilarchemistry,whichin andofitselfisagoodtestforthismethodology.Suchprovedtobe thecase.Forexample,samples16and17camefromamegalithand theadjacentbedrockdirectlyunderneath,northoftheSouth Acropolis.Anexcavationconductedherein2004foundlittleartifactualevidencesupportingthisasamonumentfeature.Thesetwo samplesaresimilarat d(16,17) 0.913,suggestingthatthemegalith isindeeddirectspallfromtheunderlyingbedrockand insitu inits originallocation.Anumberofotherpossiblemonumentsamples (e.g.37,38,and52)arealsosimilartonearbybedrocksamplesat d 0.8,asareseveral(e.g.4 e 8)collectedfromMegalith1,aseries ofmegalithsgivingtheappearanceofabrokenmonumentsouthof thesoutheastcorneroftheWestPlaza.Conversely,apossiblestela attheeasternbaseoftheNorthHillandabedrocksamplefromthe hillaresimilarat d(18,36) 0.389,supportingtheidenti cationof thestelaasamonumentthatwasmovedintoplacefromelsewhere atthesiteorfromoutsideMaaxNa. Fig.10 plotsallthesamplesfromMaaxNaandChawakBut Â’ o Â’ ob ontheternarydiagramagainsttheratiosofCaO/MgO,Cr/Ni,andSr/ Ladiscussedabove.Thebedrocksamplesareplottedinblue,the possiblemonumentsinred,andthesetof vesmallmegalithsin theSouthAcropolisandWestPlaza(10 e 15,and43),whichwillbe discussedbelow,areplottedingreen.Thereisageneralclustering ofbedrocksampleswithlowCaO/MgOandhighSr/Lavaluesdueto thedolomiticlimestonecompositionofthebedrockatMaaxNa. Mostofthemonumentsplotwithintherangeofthebedrock samples,showingthatmanyofthemegalithssampledareeither notmonumentsorweremadefromthelocalbedrock.Thereare, however,somesamplesthatfallwelloutsidetherangesestablished fortheMaaxNabedrock.Threenoticeableoutliersamongthe monumentsamples(reddots)arethreestela.Stela1totheeastof themaintemple,andthemonumenteastoftheNorthHill,insight oftheentrancetoSpiderCave,plotnearestthepointofCr/Ni,while Stela3northoftheNorthPlazaplotstothefarrightonthegraph. Theremainingmonumentsamplesplotwithinthegeneralrangeof thebedrock,althoughthecutballcourtblockandthepossiblealtar intheNorthPlaza(24and25)plotsomewhatabovetheMaaxNa range. Theclearestandmostintriguingoutliersinthedatasetin Fig.10 areplottedseparatelyingreenandfalltowardtheCaO/MgOcorner. ThesewerethesmallstonesfromgroupingsofthreeintwolocationsatMaaxNa:theSouthAcropolis(samples11 e 15)andsouthof Sr/LaCaO/MgO Cr/Ni Fig.10. TernarydiagramplottingCr/Ni,Sr/La,andCaO/MgO.Blueopencircles Maax Nabedrock,reddots MaaxNamonuments,greenopensquares MaaxNasmall monuments,andyellow lledsquares ChawakBut Â’ o Â’ obsamples.(Forinterpretation ofthereferencestocolorinthis gurelegend,thereaderisreferredtothewebversion ofthisarticle.) Fig.9. PlotofSrvs P REEforMaaxNasamples. M.L.Brennanetal./JournalofArchaeologicalScience40(2013)3178 e 3192 3188

PAGE 12

theWestPlaza(sample43; Fig.2 ),aswellasasinglesimilarstone fromoutsidethegroupattheSouthAcropolis(sample10).As noted,thefunctionofthesestonesisasyetundetermined,butthey weresampledineachlocationbecausethe “ altars ” weresimilarly shaped,andallsevenhadasmootherweatheredsurfacethanmost megaliths( Fig.11 ).Theirweatheringpatternisexplainedbythe factthattheyarecomposedofadense,tan, ne-grainedlimestone withaconchoidalfractureratherthanthecommonsoft,white limestoneatMaaxNa.Thechemistryofthesestonefeaturesdiffers drasticallyfromtheMaaxNabedrock,primarilyintheSrandMgO concentrations.Magnesiumis < 0.5wt.%,classifyingtheseaspure limestoneasopposedtothedolomiticlimestonefoundelsewhere. Strontiumconcentrationsarehigherthanthoseinmostoftheother samples,generally,becausedolomitescontainlessSrthancalcite ( Verheydenetal.,2000 ).TheMgO/CaOandSr/CaOratiosforallthe MaaxNasamplesareplottedin Fig.12 (after Aharonetal.,2006 ). Thedolomiticlimestoneofthebedrockandtheothermonuments atMaaxNashowsapositiverelationship,andthesamplesplot togetheralongoneline.Thesevenlimestonesamplesplottogether onanearlyverticalline,indicatingthepurerlimestonecomposition.TheconsistentdolomiticnatureofthebedrockatbothMaax NaandChawakBut ’ o ’ obindicatesthatthesesmallmonumentsmay havebeenquarriedatsomedistancefromthesesites.Thesmall sizeofthesestonesmakesitfeasiblethattheyweretransported fromoutsidethisarea. DifferencesintheREEpro lesforthemonumentsamplesfrom MaaxNaalsosupportsomeofthemhavingbeenimportedfrom quarriesoutsidethesitecore. Fig.8 plotstheREEpro lesforthe MaaxNapossiblemonumentandmegalithsamples,normalizedto anaveragebedrockcompositionforMaaxNa,whichwascalculated fromthebedrocksamplesanalyzed.Fourpossiblemonument samplesareplottedinlightblue,comprisingmegalithswhose identi cationasnaturalorculturalwasunclearduringthesurvey. DarkbluelinesplotsamplescollectedfromMegalith1,theidenti cationofwhichasamonumentwasalsoinquestion.Allofthese samplesplotwithinthegeneralclusterofsamplesfortheMaaxNa bedrockin Fig.10 andhaverelatively atREEpro leswhen normalizedtotheMaaxNabedrockaverage.Alternately,the monumentsthatplotoutsidetheclusterofthebedrocksamples showmorevaryingREEpro les.Thethreestelaesampledduring thissurveyareplottedin Fig.8 inred(inthewebversion),the ballcourtsampleandNorthPlazaaltarinorange,andtheseven smallmonumentsingreen.ThesesamplesshowmuchmorepronouncednegativeCeanomalies.Thesmallmonumentsingreen alsoshowaconsistentnegativeEuanomalynotpresentinthe othersamples,whennormalizedtotheMaaxNabedrock,further con rmingadifferencebetweentheserocksamplesandthose fromthesite ’ sbedrock. 4.2.2.ChawakBut ’ o ’ ob ThesamplingatChawakBut ’ o’ obwasfocusedonthe geochemicalcharacterizationofthesiteinanactivekarstenvironmentasacomparisontoMaaxNa,whichwasconstructedatthe topofaplateau.Fewermonumentswereincludedinthesampling atChawakBut ’ o ’ ob.Threecobblesfromtheballcourt,astelaand thepossibleplastersurfacebeneathit,andabridgebaseinastream weretheonlyculturalfeaturessampled.Thesamplesfromthissite areplottedinyellowagainstthosefromMaaxNaontheternary diagramin Fig.10 ,andtheyplotprimarilyinthesameareaasthe MaaxNabedrock,withsomevariationinthesamerange.This showsthat,despitea7kmdistanceandaverticalchangeoverthree escarpments,thegeneralbedrockchemistryisrelativelyconsistent betweenthesesites. Afewexceptionscanbenoted,butthey,too,stillfallwithinthe bedrockrangeandcanbeexplainedbylocalformationandpostdiageneticprocesses.FoursamplesplotneartheCr/Nipointof thediagram,awayfromtherest.Thesearetwosamples(3and4) fromdrainagechannels,orarroyos,oneatthesouthernendofthe siteandonenearGroupG( Fig.3 ),aswellastwocobblesfromthe ballcourt.Thesecobbleslikelycamefromthesoutherndrainage, whichisnearwheretheballcourtwasconstructed.Thissetof samplesmaythusre ectthesomewhatdifferentchemistrypresentinalowerstratumofthebedrockthantherestoftheoutcrops thatweresampled.Anothersample,number12,plotstotherightof theothers.Thiswascollectedfromapossiblebridgebaseinasmall streamattheeasternedgeofthesite.Thesamplehasextremely highvaluesofSr,Ba,andREE,andverylowMgOcomparedtothe othersamples.Thisfeatureisatthebottomoftheslopeonwhich ChawakBut ’ o ’ obwasbuilt,andthewater owingthroughthe bedrockpoolshereand owsthroughthisstreamandoverthe stonefeature.Thehighconcentrationsarelikelyduetothe movementofleachedelementsfromthebedrockabove.Theratios ofSr/LaandCr/Niaresimilartotheotherbedrock,andthissample wouldplotintheclusterin Fig.10 withtheexceptionofahigher CaO/MgOratioduetothelowMgconcentration.Meteoricwaters haveverylowCa/Mg,makingMgleachingmorepronounced ( Flugel,2004 ). Insummary,despitetheminordifferencesdiscussed,allthe culturalfeaturessampledatChawakBut ’ o ’ obappeartobemadeof stonelocaltotheareaofthesite.Sample7AfromthestelaatGroup Fig.12. BivariateplotofMgO/CaOvsSr/CaforMaaxNasamples.Blue diamonds bedrock,redsquares monuments,andgreentriangles smallmonuments.(Forinterpretationofthereferencestocolorinthis gurelegend,thereaderis referredtothewebversionofthisarticle.) Fig.11. Photographofsmallunknownmonument.PhotobyM.Brennan. M.L.Brennanetal./JournalofArchaeologicalScience40(2013)3178 e 3192 3189

PAGE 13

Amatchesthebedrock,butitiscomposedofahard,tanstone, similarinappearance,butnotchemistry,tothatofthesmall limestonemonumentsatMaaxNa.Itseemslikely,therefore,that thestoneforthisstelawascarefullyselected,althoughfroma nearbysource. 4.3.Mayastoneuse TheresultsofthispilotstudyindicatethattheMayawerevery knowledgeableabouttherockbeneaththeirfeetand,insome cases,mayhavegonetoextremeefforttoacquirestonefrom elsewhereforspeci cpurposes.ThethreestelaesampledatMaax NaareclearlycutfromstonethatisnotlocaltotheareaofMaax Na Â’ ssitecore.GiventheresultsfromtheChawakBut Â’ o Â’ obarroyo samples,thechemistryofthestelaemaybelocal,butfroman unknownquarrysomewherenearthesite.Thesevensmall,pure limestonemonumentsfromtwoareasofMaaxNa,however,are unlikelytohavecomefromnearthesite,asnootherlow-Mg limestonewassampledinthisstudy.ThetwounsampledmonumentsfromsouthoftheWestPlazalikelysharethischemistry, basedontheirappearanceaswellastheirgroupingwiththe monumentthatwassampled.TheseresultsshowthattheMayaat MaaxNacarefullyselectedstonefortheerectionofcertain monuments,butwereabletousethelocalbedrockforotherstone features.Thesamecarefulselectionofstoneisalsoapparentat ChawakBut Â’ o Â’ obbasedonthephysicalappearanceoftherockthat comprisesthestelainGroupA.AllofthesampledculturalfeaturesatChawakBut Â’ o Â’ ob,however,wereoflocalbedrock.Thissite wasconstructedlateintheLateClassicperiodwithoutthe monumentalfocusofsitecenterssuchasMaaxNa,solesscostand effortmayhavebeenputintoacquiringmaterialsfromoutsidethe site. Despitethis,ChawakBut Â’ o Â’ obstillunderwentextensivelandscapemodi cationsthatincludedterracingandwaterdrainage andcatchmentfeatures( Walling,2005 ).Thesefeaturesareprimarilysmall-scaleandengineeredtotakeadvantageoftheslope oftheescarpmenttochannelwaterasandwhereneeded.Maax Na,ontheotherhand,isahilltopcenterthatunderwentsigni cantquarryingtoshapethelandscapeoftheplateau,likeLaMilpa andKinal.Thereservoirwascarvedout,butcertaintopographical featureswereleftinplaceatthesite,althoughshaped,suchasthe twoknollsformingthesoutheastedgeofthereservoirandthe NorthHill.ThesmallhillcontainingSpiderCaveisalsonatural, andthelayoutofthesiteandtheincorporationoftheNorthHill intotheplanoftheNorthPlazaarehypothesizedtobelinkedto thepresenceofcaves( KingandShaw,2003 ; Kingetal.,2012 ).The speci cmovementofstonefromquarriesatMaaxNatothe nal locationofarchitectureandmonumentswasnotpossibleto determinegiventhesmallnumberofsamplesinthisstudy,and, especially,thefewthatcamefromquarries.However,sucha researchquestionwouldbepossiblewithmoreintensivesampling,usingthisstudyasabaseline.NowthatageneralunderstandingofthevariabilityofthebedrockatMaaxNaandChawak But Â’ o Â’ obhasbeenacquired,furthercharacterizationmaybe warranted.Moreintensivesampling,speci callyalongthe perimeterofMaaxNa,couldhelppinpointthequarriesfrom whichcertainstonewasexcavated.Additionally,moreregional researchmaybeabletoelucidatethesourceofthepurelimestone thatthesmallmonumentswerequarriedfrom,whichmaybe somedistancefromMaaxNa. 5.Conclusion Ashasbeenshownforquartzite( Pitbladoetal.,2008 ),variabilitywasexpectedinthelimestonesamplesfromthesesites.However,theanalysisdiscussedaboveshowsawaytointerpret suchdatasothatgeologicallyimportantcomparisonscanbemade todeterminethepossiblesourcesofstoneatthesesites.While therewasarangeinthegeochemicalresultsfromsamplesacross theareasofbothsites,themajorityofthemclustertogether. Furtherinterpretationbasedonthecontextofsampledmonumentshelpunderstandcertainvariations.Whilethesampling strategywastoosparsetosourcemonumentstonearbyquarries successfully,severalstonefeatureswereidenti edasoriginating fromquarriesoutsidethesampledcoreofMaaxNa:threestelae andthesevensmallpossiblealtars.Thesourcesofthesestelae,and ofthestelafromChawakBut Â’ o Â’ ob,mightbenearbyquarrieswith betterqualitylimestone,butthechemistryoftheotherseven monumentssuggeststhatthatmaterialmayhavecomefromquite somedistancebasedonthelow-Mg,nearlypurelimestone composition.Thebedrocksamplesfrombothsitesinvestigated provedtobeofsimilarcompositionsfortheelementsplottedin Fig.10 ,despitethefactthatChawakBut Â’ o Â’ obisbelowtheescarpmentand7kmsoutheastofMaaxNa.Furtherregionalinvestigationscouldhelpdiscernfromhowfarawayothertypesof limestonewouldhaveneededtobeimportedbyidentifying quarries. Thisisthe rststudytolookatthegeochemicalcompositionof thesurfacebedrockinthisareaofnorthernBelize,inpartthanksto theareastheMayacutawayandquarried,leavingeasilyaccessible bedrockoutcropsforsampling.WhilethebedrockoftheThree RiversRegionisanextensionofalargeEocenelimestoneformation spanningmuchofnorthernGuatemalaandBelize,itisimportantto characterizeitmorelocallytodeterminemicrovariationswithin theformationthatcanillustratebothpost-diageneticeffectsand themovementofstonebytheancientMaya.Thebedrockofthese twositesshowsheavyepigeneticeffectsfromthemovementof wateroverthelandscapethathaschangedboththechemistryand mineralogyofthelimestone.Thedolomiticlimestonebedrockhas beenrecrystallizedintomicrite,andrangesinporosityfromavery soft,powderywhiterocktoahard,densetantypeofrock.While thedolomitecompositionofthebedrockvariesatbothsites,the variationatChawakBut Â’ o Â’ obindicatestheremovalofMgfromthe bedrockduetoleachingbymeteoricwatersmovingthroughthe bedrock.BecauseofMaaxNa Â’ slocationatthetopoftheplateau, thatwasnotgenerallythecase.However,thissortofleachingof tracemetalsbythemovementofmeteoricwaterisalsoseeninthe travertinesurfaceonthe oorofSpiderCave,whichwe geochemicallycon rmedhadwater owingthroughit.More importantly,thestudyshowsthatthebedrockhasbeenaffected unevenlybasedonthevariabilityinporosityanddensityofthe limestone. ThiscarbonatelandscapewaswhattheMayausedtoconstruct theirbuiltenvironmentandshapeitaccordingtotheirworldview. Theirclearunderstandingofthelocalgeologyisevidentinthe typesofstoneusedforcertainstelaeandmonuments,andthe transportofsomeofthemtoMaaxNafromoutsidethesitecore. Ourmostimportantresultistheidenti cationofstonenotderived fromthesamecarbonatematerialasthebedrockatMaaxNa,from whichmonumentsweremadeandwhosesourcemaybesome distanceaway.ThisdeliberatechoicespeakstotheMaya Â’ sunderstandingofstonequalityandtheefforttheyexpendedtoacquire certaintypesofstone.Graham( 1987 :754)describeshowfarthe MayawenttoprocurelithicresourcesintheLowlands,evennoting thattheMayadidnotalwaysacquirerockatthesource.Whilethe stonematerialsshementionsincludeonlyobsidian,granite,and chert,ourresultssuggestthatlimestonemayhavebeenanother stonetypethatwastraded.Thesevenpossiblealtarsandstelaeat MaaxNaaresmallenoughtohavebeenpartofsuchastonetrade, whichmightalsohaveincludedotherlimestoneartifactsregularlyM.L.Brennanetal./JournalofArchaeologicalScience40(2013)3178 e 3192 3190

PAGE 14

foundinexcavations,suchasmanosandmetates.Ourresults indicatethatbasicgeochemicaldatacouldbecollectedinthe eld thatwouldhelplocatepossiblesourcesforthisstoneaswellas nd othermonumentsmadefromit.WhileICP-MSandICP-AESanalysesofsamplesareoptimalforinitialcharacterization,handheld XRFandothersuch elddevicescanprovideimmediatebulk geochemicalresults,suchasinthisareatheMgconcentration,or percentdolomite.Suchimplementationcouldaccelerateand expandthisformofresearchintheThreeRiversRegionand beyond.Finally,andperhapsmostsigni cantly,theresultsofthis pilotstudyindicatethatitispossibletodiscriminateamongstone sources,eveninareaswherethetypeofstonedoesnotlenditselfto clearvisualand/orgeochemicaldistinctions.Geochemistrybyitself cannotprovidealltheanswers,particularlywheremonuments weremadefromlocalbedrock.Byusingacombinationof geochemicalandarchaeologicalapproaches,however,wemay begintodecipherstoneselection,movement,andusageacrosspast landscapes. Acknowledgments Beforeacknowledgingthemanyinstitutionsandpeoplewho madethisresearchpossible,wewouldliketorememberourcoauthorLeslieShaw,whodiedunexpectedlywhilethisarticlewas underreview.Herknowledgeandthoughtfulinputguidedthe researchoverthesevenyearssinceitsinceptionandledtothe nal productyouseetoday.Shewillbegreatlymissed.Wewishtothank theInstituteofArchaeologywhograntedpermissiontosampleand exportthematerialsfromBelize,whichwereexportedbyDr.Fred ValdezoftheUniversityofTexasatAustin,DirectoroftheProgrammeforBelizeArchaeologicalProject.Fundingfortheproject duringyearsinwhichsampleswerecollectedwasprovidedbythe NationalGeographicSociety,HowardUniversity,andBowdoin College;fundsforsampleprocessingwereprovidedbytheGSO AlumniAssociationandtheURIGraduateSchool Â’ sEnhancementof GraduateResearchAwards.Inaddition,wewishtothankRobert Ballard,MatthewBrennan,MaryjoBrounce,IlyaBuynevich,Steve Carey,AllanCobb,ChanceCoughenour,PeterDavis,JamesDelgado, RobynDodge,IashaDoumanoff,KatieKelley,DavidKing,Jason Krumholz,MarionLytle,DavidMurray,ReneOlson,SandraWitten, andtheGSOICP-MSlab. ReferencesAbrams,E.M.,1994.HowtheMayaBuiltTheirWorld:EnergeticsandAncientArchitecture.UniversityofTexasPress,Austin. Aharon,P.,Rasbury,M.,Murgulet,V.,2006.CavesofNiueIsland,SouthPaci c: speleothemsandwatergeochemistry.In:Harmon,R.S.,Wicks,C.M.(Eds.), PerspectivesonKarstGeomorphology,Hydrology,andGeochemistry:aTribute toDerekC.FordandWilliamB.White,pp.283 e 295. Ashmore,W.,1991.Site-planningprinciplesandconceptsofdirectionalityamong theancientMaya.LatinAmericanAntiquity2(3),199 e 226. Ashmore,W.,1992.DecipheringMayaarchitecturalplans.In:Danien,E.C., Sharer,R.J.(Eds.),NewTheoriesontheAncientMaya.UniversityMuseum Monograph77.UniversityMuseum,UniversityofPennsylvania,Philadelphia, pp.173 e 184. Ashmore,W.,2008.Visionsofthecosmos:ceremoniallandscapesandcivicplans. In:David,B.,Thomas,J.(Eds.),HandbookofLandscapeArchaeology.LeftCoast Press,Inc.,WalnutCreek,CA,pp.167 e 175. Aston,M.,Rowley,T.,1974.LandscapeArchaeology.DavidandCharles,London. Barba,L.,Blancas,J.,Manzanilla,L.R.,Ortiz,A.,Barca,D.,Crisci,G.M.,Miriello,D., Pecci,A.,2009.ProvenanceofthelimestoneusedinTeotihuacn(Mexico):a methodologicalapproach.Archaeometry51,525 e 545. Bolland,O.N.,1977.TheFormationofaColonialSociety:Belize,fromConquestto CrownColony.TheJohnsHopkinsUniversityPress,Baltimore. Borchardt,G.A.,Aruscavage,P.J.,MillardJr.,H.T.,1972.CorrelationoftheBishopAsh, aPleistocenemarkerbed,usinginstrumentneuronactivationanalysis.Journal ofSedimentaryPetrology42,301 e 306. Brady,J.E.,1989.AnInvestigationofMayaRitualCaveUsewithSpecialReferenceto NajTunich,Petn,Guatemala.UnpublishedPh.D.dissertation.Archaeology Program,UniversityofCalifornia,LosAngeles. Brady,J.E.,1996.SourcesfortheStudyofMesoamericanRitualCaveUse.In:Studies inMesoamericanCaveUse,PublicationNo.1.GeorgeWashingtonUniversity, Washington,D.C. Brady,J.E.,1997.Settlementcon gurationandcosmology:theroleofCavesatDos Pilas.AmericanAnthropologist99(3),602 e 618. Brady,J.,Ashmore,W.,1999.Mountains,caves,water:ideationallandscapesofthe AncientMaya.In:Ashmore,W.,Knapp,A.B.(Eds.),ArchaeologiesofLandscape: ContemporaryPerspectives.BlackwellPublishers,Oxford,pp.124 e 145. Brady,J.E.,Prufer,K.M.,2005.Mayacavearchaeology:anewlookatreligionand cosmology.In:Prufer,K.M.,Brady,J.E.(Eds.),StoneHousesandEarthLords: MayaReligionintheCaveContext.UniversityPressofColorado,Boulder, pp.365 e 379. David,B.,Thomas,J.,2008.Landscapearchaeology:introduction.In:David,B., Thomas,J.(Eds.),HandbookofLandscapeArchaeology.LeftCoastPress,Inc., WalnutCreek,CA,pp.27 e 43. Denham,T.,2008.Environmentalarchaeology:interpretingpractices-in-thelandscapethroughgeoarchaeology.In:David,B.,Thomas,J.(Eds.),Handbook ofLandscapeArchaeology.LeftCoastPress,Inc.,WalnutCreek,CA,pp.468 e 481. Dincauze,D.,2000.EnvironmentalArchaeology:PrinciplesandPractice.Cambridge UniversityPress,Cambridge. Dunning,N.,Jones,J.G.,Beach,T.,Luzzadder-Beach,S.,2003.Physiography,habitats, andlandscapesoftheThreeRiversRegion.In:Scarborough,V.L.,ValdezJr.,F., Dunning,N.P.(Eds.),Heterarchy,PoliticalEconomy,andtheAncientMaya. UniversityofArizonaPress,Tucson,pp.14 e 24. Espinosa,L.,Ceron,M.,Sulub,Y.A.,1998.LimestonerocksoftheYucatanPeninsula. Descriptionofthelithologyandphysicalpropertiesbasedontheresultsof exploration,investigation,andlaboratorytests.InternationalJournalofRock MechanicsandMiningSciences35,410 e 411. Flores,G.,1952.SummaryReportofPreliminaryGeologicalStudiesoftheAreaNof 17NLatitude,BritishHonduras.BahamasExplorationCompany,Freeport, Bahamas. Flugel,E.,2004.MicrofaciesofCarbonateRocks:Analysis,Interpretationand Application.Springer,NewYork. Folan,W.J.,1982.MiningandquarryingtechniquesoftheLowlandMaya.Anthropology6,149 e 174. Ford,D.,Williams,P.,1989.KarstGeomorphologyandHydrology.UnwinHyman, London. Garcia-Sanchez,A.,Alvarez-Ayuso,E.,2002.SorptionofZn,CdandCroncalcite. Applicationtopuri cationofindustrialwastewater.MineralsEngineering15, 539 e 547. Graham,E.,1987.ResourcediversityinBelizeanditsimplicationsformodelsof lowlandtrade.AmericanAntiquity52(4),753 e 767. Hartshorn,G.,Nicolait,L.,Hartshorn,L.,Bevier,G.,Brightman,R.,Cal,J.,Cawich,A., Davidson,W.,DuBois,R.,Dyer,C.,Gibson,J.,Hawley,W.,Leonard,J.,Nicolait,R., Weyer,D.,White,H.,Wright,C.,1984.BelizeCountryEnvironmentalPro le,a FieldStudy.USAIDContract.Trejos.Hnos.Sucs.S.A.,SanJose. Houk,BrettA.,1996.TheArchaeologyofSitePlanning:AnExamplefromtheMaya SiteofDosHombres,Belize.UnpublishedPh.D.dissertation.Departmentof Anthropology,UniversityofTexasatAustin. Houk,BrettA.,2003.Thetiesthatbind:siteplanningintheThreeRiversRegion. In:Scarborough,V.L.,ValdezJr.,F.,Dunning,N.P.(Eds.),Heterarchy,Political Economy,andtheAncientMaya.UniversityofArizonaPress,Tucson, pp.52 e 63. Ji,H.,Wang,S.,Ouyang,Z.,Zhang,S.,Sun,C.,Liu,X.,Zhou,D.,2004.Geochemistryof redresiduaunderlyingdolomitesinkarstterrainsofYunnan-GuizhouPlateau II.Themobilityofrareearthelementsduringweathering.ChemicalGeology 203,29 e 50. Kelley,K.A.,Plank,T.,Ludden,J.,Staudigel,H.,2003.Compositionofalteredoceanic crustatODPsites801and1149.GeochemistryGeophysicsGeosystems4(6), 8910. KingJr.,D.T.,Petruny,L.W.,Pope,K.O.,2003.Shallow-marinefaciesoftheOrange WalkGroup,Miocene-Pliocene,northernBelize(CentralAmerica).GCAGS/ GCSSEPMTransactions53,384 e 397. KingJr.,D.T.,Pope,K.O.,Petruny,L.W.,2004.StratigraphyofBelize,northofthe 17thparallel.GulfCoastAssociationofGeologicalSocietiesTransactions54, 289 e 303. King,E.,Shaw,L.,2003.Aheterarchicalapproachtositevariability:theMaaxNa ArchaeologyProject.In:Scarborough,V.L.,ValdezJr.,F.,Dunning,N.P.(Eds.), Heterarchy,PoliticalEconomy,andtheAncientMaya.UniversityofArizona Press,Tucson,pp.64 e 76. King,E.M.,Brady,J.E.,Shaw,L.C.,Cobb,A.B.,Kieffer,C.L.,Brennan,M.L.,Harris,C.L., 2012.Smallcavesandsacredgeography:acasestudyfromthePrehispanic MayasiteofMaaxNa,Belize.LatinAmericanAntiquity23(4),611 e 628. Knapp,A.B.,Ashmore,W.,1999.Archaeologicallandscapes:constructed,conceptualized,ideational.In:Ashmore,W.,Knapp,A.B.(Eds.),Archaeologiesof Landscape:ContemporaryPerspectives.BlackwellPublishers,Oxford,pp.1 e 30. Kunen,J.L.,2004.AncientMayaLifeintheFarWestBajo:SocialandEnvironmental ChangeintheWetlandsofBelize.TheUniversityofArizonaPress,Tucson. Kunen,J.L.,Hughbanks,P.J.,2003.Bajocommunitiesasresourcespecialists.In: Scarborough,V.L.,ValdezJr.,F.,Dunning,N.P.(Eds.),Heterarchy,Political Economy,andtheAncientMaya.UniversityofArizonaPress,Tucson, pp.92 e 108. Lakshtanov,L.Z.,Stipp,S.L.S.,2007.Experimentalstudyofnickel(II)interactionwith calcite:adsorptionandcoprecipitation.GeochimicaetCosmochimicaActa71, 3686 e 3697.M.L.Brennanetal./JournalofArchaeologicalScience40(2013)3178 e 3192 3191

PAGE 15

Mathews,R.,2002.Geology,Environment,andLimeProductionVariationinthe MayaLowlands.UnpublishedMastersthesis.UniversityofTexas,SanAntonio. Nation,Humberto,Peterson,PollyA.,Brady,JamesE.,Neff,Hector, McAnany,PatriciaA.,2012.UsinginductivelycoupledplasmamassspectrometrytosourcespeleothemsfromarchaeologicalcontextsintheSibunValley RegionofBelize,CentralAmerica.In:Brady,JamesE.(Ed.),HeartofEarth: StudiesinMayaRitualCaveUse.Bulletin,vol.23.AssociationforMexicanCave Studies,Austin,pp.137 e 142. Northdurft,L.D.,Webb,G.E.,Kamber,B.S.,2004.Rareearthelementgeochemistryof LateDevonianreefalcarbonates,CanningBasin,WesternAustralia:con rmationofaseawaterREEproxyinancientlimestones.GeochimicaetCosmochimicaActa68(2),263 e 283. Piper,D.Z.,1974.Rareearthelementsinferromanganesenodulesandothermarine phases.GeochimicaetCosmochimicaActa38,1007 e 1022. Pitblado,B.L.,Dehler,C.,Neff,H.,Nelson,S.T.,2008.Pilotstudyexperiments sourcingquartzite,GunnisonBasin,Colorado.Geoarchaeology23(6), 742 e 778. Plank,T.,Kelley,K.A.,Murray,R.W.,Stern,L.Q.,2007.Chemicalcompositionof sedimentssubductingattheIzu-Bonintrench.GeochemistryGeophysicsGeosystems8,Q04I16. http://dx.doi.org/10.1029/2006GC001444 Reeder,R.J.,1996.Interactionofdivalentcobalt,zinc,cadmium,andbariumwiththe calcitesurfaceduringlayergrowth.GeochimicaetCosmochimicaActa60, 1543 e 1552. Reeder,R.J.,Lamble,G.M.,Northrup,P.A.,1999.XAFSstudyofthecoordinationand localrelaxationaroundCo2 ,Zn2 ,Pb2 ,andBa2 traceelementsincalcite. AmericanMineralogist84,1049 e 1060. Rice,P.M.,Michel,H.V.,Asaro,F.,Stross,F.,1985.Provenienceanalysisofobsidians fromtheCentralPetnLakesRegion,Guatemala.LatinAmericanAntiquity50 (3),591 e 604. Rollinson,H.,1993.UsingGeochemicalData:Evaluation,Presentation,Interpretation.PearsonEducationLimited,Harlow,England. Scarborough,V.L.,1998.Ecologyandritual:watermanagementandtheMaya.Latin AmericanAntiquity9,135 e 159. Scarborough,V.L.,Becher,M.E.,Baker,J.L.,Harris,G.,ValdezJr.,F.,1995.Waterand landattheancientMayacommunityofLaMilpa.LatinAmericanAntiquity6, 98 e 119. Scarborough,V.L.,ValdezJr.,F.,2003.Theengineeredenvironmentandpolitical economyoftheThreeRiversRegion.In:Scarborough,V.L.,ValdezJr.,F., Dunning,N.P.(Eds.),Heterarchy,PoliticalEconomy,andtheAncientMaya. UniversityofArizonaPress,Tucson,pp.3 e 13. Scarborough,V.L.,ValdezJr.,F.,Dunning,N.P.,2003.Heterarchy,PoliticalEconomy, andtheAncientMaya.UniversityofArizonaPress,Tucson. Scarborough,V.L.,Connolly,R.P.,Ross,S.P.,1994.ThePre-HispanicMayareservoir systematKinal,Petn,Guatemala.AncientMesoamerica5,97 e 106. Shaw,L.C.,1999.Constructedlandscapeasideology:archaeologyandmappingat MaaxNaintheThreeRiversRegionofBelize.In:PaperPresentedatthe64th AnnualMeetingoftheSocietyforAmericanArchaeology,Chicago. Shaw,L.C.,2002.TheMa Â’ axNaArchaeologyProject:DocumentationofStelae,Altars,andCaveEntrancesintheWestCeremonialGroup.FAMSIReport. http:// www.famsi.org/reports/00100/ Shaw,L.C.,King,E.M.,1997.Researchinhighplaces:thehilltopcenterofMaaxNa, Belize.In:PaperPresentedatthe62ndAnnualMeetingoftheSocietyfor AmericanArchaeology,Nashville,Tenn. Tlig,S.,M Â’ Rabet,A.,1985.Acomparativestudyoftherareearthelement(REE) distributionswithintheLowerCretaceousdolomitesandlimestonesofCentral Tunisia.Sedimentology32,897 e 907. Tourtellot,G.,Belli,F.E.,Rose,J.J.,Hammond,N.,2003.LateClassicMayaheterarchy, hierarchy,andlandscapeatLaMilpa,Belize.In:Scarborough,V.L.,ValdezJr.,F., Dunning,N.P.(Eds.),Heterarchy,PoliticalEconomy,andtheAncientMaya. UniversityofArizonaPress,Tucson,pp.37 e 51. Trudgill,S.,1985.LimestoneGeomorphology.LongmanInc.,NewYork. Vaniman,D.T.,Chipera,S.J.,1996.Paleotransportoflanthanidesandstrontium recordedincalcitecompositionsfromtuffsatYuccaMountain,Nevada,USA. GeochimicaetCosmochimicaActa60,4417 e 4433. Verheyden,S.,Keppens,E.,Fairchild,I.J.,McDermott,F.,Weis,D.,2000.Chemical Geology169,131 e 144. Villaseor,I.,2010.BuildingMaterialsoftheAncientMaya:aStudyofArchaeologicalPlasters.LambertAcademicPublishing,Saarbrucken,Germany. Vinson,G.L.,1962.UpperCretaceousandTertiarystratigraphyofGuatemala. BulletinoftheAmericanAssociationofPetroleumGeologists46,425 e 456. Walling,S.,2005.ArchaeologicalinvestigationofPrehispanicMayaresidential terraces,commonerhousingandhydrologyatChawakBut Â’ o Â’ ob,Belize.Antiquity79.ProjectGallery.Availableat: http://antiquity.ac.uk/Projgall/walling/ Walling,S.L.,2011.OverviewofrecentballcourtcomplexinvestigationsatChawak But Â’ o Â’ ob,Belize.In:Houk,Brett,ValdezJr.,Fred(Eds.),ResearchReportsfrom theProgrammeforBelizeArchaeologicalProject.OccasionalPapersNumber12, vol.5.MesoamericanArchaeologicalResearchLaboratory,UniversityofTexas, Austin,pp.251 e 261. Walling,S.L.,Davis,P.,Hanna,J.,Prasarn,N.,Matthews,L.,Taylor,C.,2006.Residentialterracing,watermanagement,suburbanceremonialismandmatrix analysisatChawakBut Â’ o Â’ ob:reportofthe2005RioBravoArchaeologicalSurvey.In:ValdezJr.,Fred(Ed.),ProgrammeforBelizeArchaeologyProjectReports. OccasionalPapers,Number6,vol.3.MesoamericanArchaeologicalLaboratory, UniversityofTexas,Austin,pp.41e 87. Wolf,K.H.,Chilingar,G.V.,Beales,F.W.,1967.Elementalcompositionofcarbonate skeletons,mineralsandsediments.In:Chilingar,G.V.,Bissell,H.J., Fairbridge,R.W.(Eds.),CarbonateRocks:PhysicalandChemicalAspects. Elsevier,Amsterdam,pp.23 e 149. Woods,J.C.,Titmus,G.L.,1996.Stoneonstone:perspectivesonMayacivilization fromlithicstudies.In:Robinson,M.G.,Macri,M.J.,McHargue,J.(Eds.),Eighth PalenqueRoundTable,1993.ThePre-ColumbianArtResearchInstitute,San Francisco. Zachara,J.M.,Cowan,C.E.,Resch,C.T.,1991.Sorptionofdivalentmetalsoncalcite. GeochimicaetCosmochimicaActa55,1549 e 1562.M.L.Brennanetal./JournalofArchaeologicalScience40(2013)3178 e 3192 3192


Description
Michael L. Brennan
a, *, Eleanor M. King
b, Leslie C. Shaw
c, Stanley L. Walling
d, Fred Valdez Jr.
e
aGraduate School of Oceanography, University of
Rhode Island, Narragansett, RI 02882, USA
bDepartment of Sociology and Anthropology, Howard
University, Washington, D.C., USA
cDepartment of Sociology and Anthropology, Bowdoin
College, Brunswick, ME, USA
dDepartment of Social Sciences, Community College
of Philadelphia, USA
eDepartment of Anthropology, University of Texas
at Austin, USA The carbonate bedrock of northwestern Belize
is poorly understood from the standpoint of both geochemistry
and the use of stone in prehispanic Maya sites for buildings
and monuments. The friable nature of the rock in this
topographically rugged area makes it especially difficult to
distinguish mon- uments from bedrock spall, as little
carving, if any, survives, and identification rests on
location and positioning. The research presented here
analyzed 63 limestone samples collected from two sites in the
Three Rivers Region of Belize. ICP-MS and ICP-AES were used
to characterize the major, minor, and trace element chemistry
of the limestone bedrock of the region and determine the
amount of geochemical variability. Another important
objective was attempting to trace the movement of monument
stone and determining whether it was imported from outside of
the sites. Bedrock, quarries, and possible monu- ments were
all sampled for these purposes. Bedrock proved to be similar
across wide areas. However, at Chawak But'o'ob, along the
flank of the Rio Bravo, changes downslope in Mg concentration
suggest a leaching of the bedrock by meteoric waters based on
differences in porosity. At Maax Na, a hilltop site, in
contrast, such leaching is not as apparent. Many monuments at
both sites were found to be composed of stone similar in
chemistry to the local bedrock, including several of the
identifiable stelae. However, our analyses also revealed that
a few monuments at Maax Na were made of material with a
different chemical composition, apparently from stone
imported to the site. These results suggest that the Maya
deliberately selected certain types of limestone for certain
purposes, and may even have traded in non- local rock.
Overall, the methods used in this pilot study indicate there
is real potential in more intensive, regional assessments of
the materials used at archaeological sites, even in areas
where the local stone does not have a distinctive geochemical
signature.