University of WollongongResearch Online

Faculty of Science, Medicine and Health - Papers Faculty of Science, Medicine and Health

2012

Usewear and phytoliths on bedrock grindingpatches, Pilbara, north-western AustraliaRichard FullagarUniversity of Wollongong, fullagar@uow.edu.au

Lynley A. WallisUniversity of Queensland

Research Online is the open access institutional repository for the University of Wollongong. For further information contact the UOW Library:research-pubs@uow.edu.au

Publication DetailsFullagar, R. & Wallis, L. A. (2012). Usewear and phytoliths on bedrock grinding patches, Pilbara, north-western Australia. Artefact:Pacific rim archaeology, 35 75-87.

Usewear and phytoliths on bedrock grinding patches, Pilbara, north-western Australia

AbstractBedrock grinding patches were recorded in the Fortescue Metals Group Ltd (FMG) Rail Corridor within theWooodstock/Abydos Aboriginal Heritage Area 130 km south of Port Hedland, Western Australia. WA StateMinisterial conditions required the salvage of representative samples, residue analysis and other detailedmicroscopic study to investigate the technology and function of these grinding patches. Following a pilotstudy and experimental work, we undertook microscopic study of 159 samples - including PVS (PolyVinylSiloxane™) peels and water extractions - from 81 grinding patches, collected at six sites. The worn stonesurfaces are microscopically similar to traces found on experimental and Aboriginal stone artefacts used forgrinding seeds, although the development of wear patterns is variable. The most common residues werephytoliths, which indicate that grinding patches were utilised for grinding grasses of the Panicoid andChloridoid sub-families, although the open nature of the sites means issues of taphonomy need to beconsidered. Spinifex phytoliths suggest seeds from this plant may also have been exploited. No traces ofpigment or ochre were found. We suggest that the grinding patches are linked with food processing associatedceremonial gatherings and rock art.

DisciplinesMedicine and Health Sciences | Social and Behavioral Sciences

Publication DetailsFullagar, R. & Wallis, L. A. (2012). Usewear and phytoliths on bedrock grinding patches, Pilbara, north-western Australia. Artefact: Pacific rim archaeology, 35 75-87.

This journal article is available at Research Online: http://ro.uow.edu.au/smhpapers/2681

1

Usewearandphytolithsonbedrockgrindingpatches,Pilbara,

north‐westernAustralia

RichardFullagar,UniversityofWollongongandScarpArchaeology

and

LynleyWallis,FlindersUniversityandWallisHeritageConsulting

Abstract

Bedrock grinding patches were recorded in the Fortescue Metals Group Ltd RailCorridorwithintheWooodstock/AbydosAboriginalHeritageArea130kmsouthofPortHedland, Western Australia. WA State Ministerial conditions required the salvage ofrepresentative samples, residue analysis and other detailed microscopic study toinvestigatethetechnologyandfunctionofgrindingpatchesinthearea.Followingapilotstudy and experimental work, we undertook microscopic study of 159 samples –including PVS (PolyVinyl Siloxane™) peels and water extractions – from 81 grindingpatches, collected at six recorded sites. The worn stone surfaces are similarmicroscopicallytotracesfoundonexperimentalandAboriginalstoneartefactsusedforgrinding seeds, although the development of wear patterns is variable. The mostcommonresidueswerephytolithswhichindicatethatgrindingpatcheswereutilisedforgrindinggrassesofthePanicoidandChloridoidsub‐families,althoughtheopennatureof the sites means issues of taphonomy need to be considered. Spinifex phytolithssuggest seeds from this plantmay also have been exploited. No traces of pigment orochrewerefound.Wesuggestthatthegrindingpatchesarelinkedwithfoodprocessingassociatedceremonialgatheringsandrockart.

Introduction

As part of a heritage mitigation strategy, Shaun Canning, ACHM (Australian CulturalHeritage Management) and Gavin Jackson (Gavin Jackson P/L) invited RF to salvagesome granite grinding patches in the Pilbara, Western Australia (WA). WA StateMinisterialconditionsrequiredthesalvageofrepresentativesamples,residueanalysisandotherdetailedmicroscopicstudytoinvestigatethetechnologyandfunctionofthegrinding patches prior to their destruction. The bedrock grinding patches had beenpreviously recorded in the FMG Rail Corridor within the Wooodstock/AbydosAboriginalHeritageArea130kmsouthofPortHedland,WA(Jackson2006).Theareais

60‐200subseqevaluatfieldwoon KarKarriyato samrecordiall worunwornresidueGavinJPVSanthis pamicrosgrindinethno‐h

StudyA

Thestusouthessenti(BoM2

Figure

0 km southquently invte possibilorkwithRFrriyarra coarracommple14griningandsamrn grindingn surfaces.es from thJacksonresndaqueousaper,we oucopic studngseeds,inhistoricale

Area

udyareaisof Port Heially a ston2012).The

1:Studyar

h of Portvited Alanlities for dF,AlanWaountry inunitymemndingpatchmplingmeg patches. An aqueohe abundansultedinthresiduesautline the rdy, suggestncludingpeevidence.

partofbotedland, inny desertwvegetation

rea.[Sourc

Hedland iWatchmandating orgatchman,GDecember

mbers(Bruches from fthodology.(from theous samplent cracks iheapplicatamplesfromrecordingmting that therhapsthet

thKarriyarthe Pilba

with irreguisdominat

ce:FMG]

in Karriyan to underganic accreGavin Jacks2006 in

ceBungandfourAborig.Polyvinylsmootheste was alsoin the grantionofthismsixmoremethodolohe grindintinyseeds

rraandPalyara regionular and vetedbyacac

arra and Prtake geoloetions ononandBoothe compadAndrewGginal sitesSiloxane(t looking sremovednite surfacmethodoloesites(n=8gy and sumng patchesofspinifex

ykucountrof WA (F

ery low raiciaandspin

Palyku Aboogical invethe grindioneLaw(Aany of FMGordon).Ainorder toPVS)peelssurfaces) ausing a pie. Subsequogytoreco81grindingmmarise kwere mos,forwhich

ry,approximFigure 1).nfall, 150‐nifexscrub

original lanestigationsing patcheACHM) tooMG personnkeyobjectodevelopasweretakeand from apette to suuent fieldwoverafurthgpatchestokey findingst likely uhthereison

mately60‐The lands300mmp(Figure2)

2

nds. RFand toes. Thisokplacenel andtivewasaviableen fromadjacentuck anywork byher159otal).Ins of theused fornlyrare

200kmscape isper year.

3

Figure2:Spinifexandacaciaspeciesdominatethelandscapearoundthegranitedomesandgrindingpatches.[Photo:BooneLaw]

Watchman(2007:4)inhisgeologicalreportnotedthatthestudyareagenerallylies

‘between the Pilbara hinterland of iron‐rich sediments and the coast, and essentiallycompriseslowgranitehillsabout200minelevation.VariousArchaeangraniticrockscropoutintheareaandthesecomprisetheYuleBatholith(HickmanandLipple1978).’

Watchman (2007:4) described one typical sitewith unusual xenoliths (rocks that areenvelopedinalargerrockduringformation):

‘AtsiteKAR06‐02thisstructuraldomeofgranitoidsispartofamigmatitesuiteofrocksconsistingof fine tocoarse‐grainedequigranular toporphyriticbiotiteadamelliteandbiotite granodiorite. Leucosomes of thin aplite dykes and pegmatites have intrudedparalleltothefoliation.

At site KAR 06‐02 a narrow, even‐grained aplite has intruded the porphyritic biotiteadamellite,anditisonthisequigranularintrusiverockthatthegrindingpatch(#1)wasmade (Figure 3). The aplite dyke is approximately 55cm wide and contains pale,elongate xenoliths (near the colour scale in Figure 3). The dyke trends at 30° andalthoughgenerallyflatontophasaslightdipofapproximately10°tothenorthwest.’

Thegraniteincludesquartzthatvariesingrainsizeandisveryhard. Hundredsifnotthousands of grinding patches occur in this area and their function is thought to beeitherforpigmentpreparationorgrindingplantfoods(Jackson2006).

BedrockGrindingPatches

The granite commonly outcrops in the form of domes rising gently above the sandyplain.HundredsofgranitegrindingpatcheswererecordedbyACHMandGavinJacksoninanareawheretheFMGrailwasproposed(Jackson2006).Thesampledpatcheshavemacroscopicsmoothingvaryingfromabout30cmx15cminmaximumdimensionsup

4

toabout300cmx150cm.Grindingpatchmorphologymostlyfollowstheflattoslightlyconvextopographyofthegranitedomes.Somegrindingpatcheshavedepressionswornfrom use, forming a concave cross‐section. Some grinding patches are adjacent tostreambeds or natural shallowdepressions in the granite that have heldwater afterrain (Figure3 andFigure4), thoughothers are associatedwithno obvious source ofwater.

Thegranitegrindingpatchesinthestudyareaexhibitdifferentdegreesofgrindingandsome variation in the type of granite on which they occur. Some surfaces are veryweathered, andmodification by grinding shows up very clearly. Other granites haveexfoliatedandexposedrelativelyfreshsurfaces,someofwhichhavealsobeenusedforgrinding.

Figure3: Somegrindingpatches (on theedgeof thegranite slab,below thepinkandblackflaggingtape)haveshallowdepressionsandarelocatednearstreambeds.[Photo:BooneLaw]

Figure4:Somegrindingpatchesareonlowdomesofgranitebutstreambedsarelessthan a few hundred metres away. Left to right: Gavin Jackson, Bruce Bung, Andrew

5

Gordon,AlanWatchman,withRFcrouchingandpointingatthegrindingpatch.[Photo:Boonelaw]

Occasionally upper grinding stones or hand stones were found near the grindingpatches.Handstonesaretypicallymadeofaharddolerite(Figure5).

Figure 5: An upper grinding stone or hand stone (dolerite) found near a bedrockgrindingpatch.[Photo:BooneLaw]

Althoughweareawareoffewtool‐useexperimentsundertakenspecificallyongranite(butseebelow),manytool‐useexperimentshavebeenundertakenutilisingquartz,andit is the smoothing, polish, and striations on quartz‐rich rocks that provide a key forinterpretingtheusewearpatternsdescribedbelow(e.g.Fullagar1991,2006).

Karriyarra community members Andrew Gordon and Bruce Bung involved with thefieldwork project suggested that grinding patchesmight have been used for grindingochre for body paint. Consequently, the grinding patches are thought to be linked tobody painting and men’s ceremonies performed near adjacent Aboriginal rockengravings on dolerite outcrops, similar to the granophyre rock formations on theBurrup.

ExperimentalArchaeology

Twotool‐useexperimentswereundertakenbyRFinthefield.SpinifexandAcaciaseedswere harvested and were ground to flour in a period of about 30 minutes, using adoleritehandstoneandagraniteslab(Figure6).ThemainpurposeofthisexperimentwastoenableRFtoobservethenatureofgrindingusewearongranite.Itislikelythatthemainformofabrasivewearisaconsequenceofgrindinghardstonestogether(asitisforsandstoneseedgrindingimplements).However,thepresenceofsiliceousparticles(phytoliths) and other plant tissue also contributes to polish formation, as has beendocumentedforawiderangeofprocessedmaterialsinnumeroustool‐useexperiments(e.g.Hamon2008).

6

Grindingexperimentswerealsoundertakenwithredochre(hematite)which leavesadistinctive, bright, extremely fine‐grained residue that was extremely difficult toremove.Inthisinstancethegrindingprocessrapidlyceasedtobeeffectivebecausetheochre firmly filled all the depressions and cracks, creating a hard smooth film andcloggingthehardsharpgrainsoftheoriginalroughgranitesurface.Becausethestonesweretoolargetofitonamicroscopestage,polyvinylsiloxane(PVS)peels(seebelow)weretakentofacilitatethestudyofthegrindingsurfacesathighmagnificationusingalightmicroscope.

Polishdevelopmentunderthesegrindingconditionsprobablycannotextendbeyondastagewhen grains are plucked from the surface (see Figure 7). These plucked grainsprovide free abrasives that will constrain the extent of polish forming as the entiresurface is gradually worn down (Kamminga 1979). More compacted, homogenousstones(suchasquartzite)cansustainmuchmoredevelopedpatchesofpolish,butthesurfaceofthegranitefromthisstudyareaisconstantlyandrapidly

Figure 6: Experimental grinding tools used to grind Spinifex and Acacia seeds. Left:lowergrindingstone(granite).Right:Handstone(dolerite).Scalebarisincentimetres.[Photo:RF]

Polishdevelopmentunderthesegrindingconditionsprobablycannotextendbeyondastagewhengrainsarepluckedfromthesurface(seeFigure7).

FigureLeft: sobeenpgrindinScale:D

beingwgrindin

Method

Weusepatchescleanedresidue1998),adjacenreplicatunderadisposawhereagitatewaterremoveexaminvisiblemicros

Inaddiremovesurfaceweren

7:UseweaomebroadluckedfromngstoneshDiagonalis

wornwithngtoolscom

dology:sam

edtwotechs. First, thd by wipines.Then,awasusedtnt area witedwithexametallograblenylonthe PVS sadwiththeand any red with thnation reveintheserecopeatx20

tiontotheed by Alanes microscootvisibleo

aronexperd striationsmthesurfahowingfineabout1mm

use, in themmonlyfou

mplinggrin

hniquestoge areas tong the sudentalimptotakeahith no grinxcellentclaraphicrefledroppertoampleswenylontiparesidues ithe dropperealed that aesiduesam00.

peelsandn Watchmeopically; thonthegrind

rimental loand abras

face.Scaleoestriations,m.[Photo:R

esamewayundinothe

dingpatch

gathersam be samplrfaces vigopressioncoighresolutnding (Figaritysuchectedlightmodeliverpere taken oandlefttohad ‘suckand transawide ranpleswhen

residuesamen so thathis was codingpatche

owergranitsive smootonleft is in,smoothinRF]

y(butperherpartsofa

hesforusew

mplesforstued (thoseorously wompound,ptionimpresgure 8). Usthatnaturmicroscopepurifiedwaon the grinsoakintoked up’ frosferred tonge of partviewedmi

mples,tinyt I could eonsidered nesatlowm

tegrindingthing. Notenmillimetrgandpolis

haps lessraridAustra

wearandre

udyingthedeemed to

with ethanopolyvinylsissionofthesing this mal fractureeatx200matertothepnding patchcracksintom the crsealed nylticles includicroscopica

yflakesfromexamine thnecessary bmagnificatio

g stone (sh the pitswres.Right:sh(lighter

apidlythanalia.

esidues

functionofo be the ‘mol to remoiloxane(PVegrindingmethod qufeaturesw

magnificatioporousroch (Figure 9thegraniteacks andon tubes (ding phytoallyundera

mthegrindhe polishedbecause usoninthefie

hown inFigwhere grainPVSpeelocoloureds

n)sandsto

fbedrockgmost wornove looseVS)(seeMasurfaceanuartz grainwereeasilyon.Nextwecksurface9). Thewaforaminurock surfa(Figure 10olithswereatransmitt

dingsurfacd and unpse‐wear steld.

7

gure6).ns haveof lowertreaks).

neseed

grindingn’) weresurfaceandikosdofthens wereyvisibleeusedaclosetoterwasute.Theace was). Latere clearlytedlight

eswerepolishedriations

8

Figure8:RFdeliveringPVS(withmixingdispenser)tothegrindingpatchsurfaceKAR06‐21GP16.[Photo:BooneLaw]

Figure9:RFtakingaresiduesamplewiththedisposablenylondropper.[Photo:BooneLaw]

Figure10:PVS (greenpeels)andresidue (sealed tube) samplesonandoffagrindingpatch.Theredstainisformedbyveryfinesedimentatthemarginofdepressionsthatfillwithrain.ThegreenPVSpeelswereremovedandsealedinplasticbags.LargerPVS

peels ([Photo:

Results

Usingastriatiofrom gprovidepeelsothisme

FigureWidth1mm.[

FigurePVSSamThephabout2

(the size of:BooneLaw

s:Use‐wear

astereoscoonswere nrinding paedanegationfreshlyfethod(Figu

11:Striatioof field ab[Photo:RF]

12:Impresmple11tahotowasta2mm.[Pho

f standardw]

robservati

picmicrosnoted to beatches (Figuiveimpresfracturedqure12).

onsvisibleout 1cm.R

ssionoffreakenawayfakenatx10oto:RF]

microscop

ons

copeinthee very comure 11). Adsionbutwuartzgrain

onrocksaRight:Deta

eshfracturfromthem00magnific

pe slides)w

elaboratorymmon on thdditionallywithsufficiensindicate

amplefromail of usew

resandstremostwornacation,and

weremade

y,withlowhe small flay, the PVS pentdetailtotheexcelle

mgrindingpwear on roc

esslinesonareaatgrinthemaxim

e for some

wangledpoaked rockpeels fromoseethesentresolut

patchKAR0ck sample.

naquartzndingpatchmumdimen

grinding p

ointsourcesamples re

m grindingamefeaturtionachiev

06211A(Diagonal i

grain(circhKAR062nsion(diag

9

patches.

oflight,emovedpatchesres.PVSablevia

4).Left:is about

cled)on1GP38.gonal)is

Athighfrom gLowpowithhaflatwit

ThePVareflecofpolisothers.

FigureGP16.bottomRight:KRF]

hermagnifigrinding paoints on thardgrainsothfewerstr

VSimpressictedlightmshandstria

13:PVSpeNotefines

m left. LighKAR0601

icationundatches showhepeel (i.eofquartzthriations(Fi

ionsfromemicroscopeationswer

eelsfromsmstriations(shter patcheGP1Samp

derametalw abundan. highpoinhathavemgure13an

eachgrindie.Twokindrecommon

moothpolisslightlydares with smple2.Scale

llographicnt fine strints on themoresmootndFigure1

ngpatchandsofsmootnonsomeg

shedsurfacrkerscratcmall depreseforbothim

(reflectediations indgrinding stthinganda4).

ndoff‐sitethingorpogrindingpa

ceofgrindiches)andassions (damages:dia

light)micrdicating abrtone surfacappeartoh

controlswolishwerenatchesand

ingpatcheslignmentsrker patchgonalisab

roscope,PVrasive smoce) are asshavebeenp

werestudienoted.Aligrareorab

s.Left:KARrunningtohes) are pbout1mm.

10

VSpeelsoothing.sociatedpolished

dundergnmentsbsenton

R06‐21opright‐olished.[Photo:

11

Figure 14: PVS peels from grinding patches. Left: PVS peel from KAR 06 19 GP 1,showingexceptionallywelldevelopedpolishonhighpoints(lighterareas)withdarkerpatches indicatingdepressionsonthestone itself.Right:PVSpeel fromPALVII0602GP35showingmuchmoreabrasiveconditionswithstriationsoriented inat least twodirections.Scaleforbothimages:diagonalisabout1mm.[Photo:RF]

Nevertheless some surfaceswere highly smoothed and flat; and had sustained brightpolishwhich is presumably a consequenceof the stone types and thematerialwhichwasbeingground.

Results:Phytolithsinresidues

Atotalof97residuesamplesfrom95grindingpatchesat10siteswerestudiedbyLW.Thesesampleswereanalysedbyexaminationof themountedresidueonaglassslideusinga transmitted lightmicroscope.This revealed that theaqueous residuesamplestypicallyhadabundantresiduespresent,includingphytolithsandstarchgrains(Figure15). Phytolithswererecordedasbelongingtooneof23 ‘distinctive’(ierecognisable)categories:sinuousedgedrod,straightedgedrod,spinyedgedrod,blockyquadrilateral,haircellovoid,trichome,elongatedhaircell,unilobe,bilobewithalimitedwaist,bilobewith a defined waist, spheroid [which typically indicate the presence of spinifexgrasses], angular quadrilateral, Eriachne [bilobe on a base], Aristida [bilobe with adistinctive long shaft], polylobate, sphere, quadrilateral plate, sinuous edged plate,bulliform, Acacia [irregular amorphous form with surface ornamentation], stomata,Cyperaceae[platewithdistinctivepatternednodes]and‘others’.Additionally,piecesofamorphous plate and amorphous fragments of thicker, irregularly shaped, non‐repetitivemorphologieswerealsonoted;theselattercategorieswereexcludedfromthetotal phytolith countsusedboth to compare the relativephytolith abundance in eachsampleandtoconstructthephytolithdiagrams.

Figure 15: Left: Residue extraction from KAR 06 10 GP1 Sample 3, with phytoliths(somewithserrateedges)Scale:diagonalisabout1mm.Right:Starchgrain(diameter18microns – too large to be from grass) fromKAR06 01GP1. Cross polarised light.[Photo:RF]

AnalysiabundaresultophytoliSpinifeproducsub‐fam

FigureAcacia,bottomare20

Althougofgrindsupporutilisedphytoliwas simfamilie

isconfirmeance of phyoftaphonoith types oxgrasses(cedbyavamilies.Sixo

16: Somesecondro

m row: Spinmicrons,B

ghitisnotdingowingrt forthepd for the gith analysismilar in eas.

edtheubiqytoliths prmicfactorsobserved i(Figure16)arietyofgrofthesites

of the phyowdown:Pnifex (alsolackbarsa

tpossibletg to theoppropositiongrinding os suggestsach case, b

quitouspreesent in easratherthancluded th),andassemasses,partarediscuss

ytolith typePanicoideaein the Chlare5micro

tobeentirepennature that inallof plant mathat the p

being typica

esenceofpach sampleandifferenhose frommblageswticularlymsedhere(F

es observedegrasses,tloridoideaens.[Photo:

elycertainof thesitelcasestheaterials, asplantmaterally grasse

phytolithsie varied concesinorigAcacia, an

weredominembersofFigure17).

d in this stthirdrowde subfamilyLW]

thatallofessampled,sampledgs opposedrial proceses of the P

nallsamplonsiderablyginalusepand PanicoidatedbymotheChlorid

tudy. SEMdown:chloy). Scale:W

theresidue,phytolithgrindingpato ochresed on theanicoid an

lesexaminy,most likeatterns.Diad, ChloridoorphologicadoidandP

images. ToridoideaegWhite strip

eswerethanalysispatchfeaturepreparatio

e grindingnd Chlorido

12

ned.Theely as aagnosticoid andaltypesPanicoid

op row:grasses,ed bars

eresultprovideseswereon. Thepatchesoid sub‐

13

Figuresamplephytoli

Of partSpinifeofspinthesiteSpinife(which

Table1

Site

FMGKA

FMGKA

FMGKA

FMGKA

FMGKA

FMGPA

Figuresample

17:Graphe (and acrithscounte

ticular intexongrindifexarepres.Thissugx and Acaisveryhig

1:Abundan

ARIII06‐1

ARIII06‐1

ARIII06‐4

ARIII06‐5

ARIII06‐5

ALVII06‐0

18 showsesfromthe

showingthoss the siedacrosste

erest are tingpatcheresentatalggests(butcia may ngh)butrath

nceofphyto

Numgrinpatsite

8 26

9 4

0 2

0 14

5 20

02 13

s a graphicKARIII06

hecomparix differententransects

the low inswithinthllsitesbuttdoesnotot simplyhertocultu

olithsatsit

mberndingches at eace

cal represe6‐19site.

isonbetwet sites), bs.[Wallis2

ncidence ofhesamesittheirpresunequivocbe relateduralselectio

tes.

of

ch

PatcheAcaciapresen

1

1

0

2

1

0

entation of

enphytolitased on to008]

f Acacia antecomplexsencevariecallydemond to their aonattimes

es withant

PS(p

2

4

2

1

1

6

f the distin

thabundanotal numb

nd the var(Table1).satgrindinstrate)thaabundanceinthepast

PatcheswiSpinifex(spheroidspresent

26

4

2

13

19

6

nctive phy

nceineachbers of dis

riable presSpheroidsingpatchesatthepres in the lant.

ith

s)

Chlorito Parelativabund

EvenGP46

High

High

Tendineven

Evenvariatpatche

Variatpatche

ytoliths in

14

residuestinctive

ence ofstypicalswithinsenceofndscape

idoidanicoidvedance

except

ng

bution ines

tion ines

residue

Figure

AlthougconsistChloridphytolihomogChloriddegree(water,possibl

Figure

Unfortuincorpomateriaaction.presennotpronot metherefoParr 2

18:Phytoli

gh only foutency in thdoidgrasseiths in reenous,withdoidgrasseto which, wind andletodemon

19:Phytoli

unately theorated intoal,asopposWhilethecewas theocessburnean it is aore more li2006). Nev

ithdiagram

ur sampleshe presences,withversidue samhthevarioes,withonlresidue sad fire) is dnstrateAbo

ithdiagram

ere is no eo a grindinsedtothospresenceoeresultofwtgrasses),always poikely the rvertheless,

mofresidue

swere exace of spinrylittleinpmples fromusmorpholyverylimamples midifficult tooriginalpre

mofresidue

easyway tong patch resethathaveofburnt(stawindrathethe fact thssible to aresult of dethe exam

esamplesf

amined fromnifex grasseputfromPam the FMGologicaltypmitedinputight be affdetermineeferencefo

esamplesf

o distinguiesidue as aesettledonained)phyer thanhumhatburninassume uneliberate pmination of

fromFMGK

mKAR IIIes (ie sphanicoidgraG KAR IIIpesindicatifromspinifected by le with thisrparticula

fromFMGK

ish betweea result ofnthesurfaytolithsinamanagentgdoesnotnstained pplant procef ‘off‐site’

KARIII06‐

06‐19, theheroid typeasses.IncoI 06‐50 sngamixtuifexgrassesocal taphodata set argrasstaxa

KARIII06‐

n phytolithf human prcefortuitouaresiduems(sincepetalwayscaphytoliths aessing activcontrol s

‐19.[Wallis

ey show ree phytolithontrast,dissite are alureofPanics(Figure1onomic conand it is tha.

‐50.[Wallis

hs that havrocessing ouslythrougmightindicaeoplegeneause staininare unburvities (thouamples (ie

15

s2008]

elativelyhs) andstinctivell quitecoidand19).Thenditionshus not

s2008]

ve beenof plantghwindatetheirrallydongdoesrnt, andugh seee those

16

collected from the rock surface adjacent but not on grinding patches) might allowelucidationofthisissue.

Discussion

Taphonomic issues notwithstanding, the phytolith evidence provides support for theproposition that inallcases thesampledgrindingpatch featureswereutilised for thegrindingofplantmaterials.Starchgrainswereveryrareandnotlikelytosurviveopenairconditions.Noochrestainsorparticleswereobserved,noranyothersignificantuse‐residues.

Thephytolithanalysisshowsthattheplantmaterialprocessedonthegrindingpatcheswas similar in each case, being typically grasses of the Panicoid and Chloridoid sub‐families. Moreover, usewear on the grinding patches is consistent with experimentalwearpatternsfromgrindingseeds.

ExtensivebedrockgrindingpatchesareknowninGulfcountryofwesternQueensland(Gorecki and Grant 1994). A grass seed grinding function for these Pilbara grindingpatches (rather than ochre grinding) is not surprising particularly given the polishedsurfaces noted by previous researchers, and the extensive distribution and long timedepthofgrindingtechnologyinAustralia(Fullagaretal.2008;Goreckietal.1997).Noris it surprising to find phytoliths in abundance (see Hart andWallis 2003). What issurprising is the evidence of spinifex phytoliths and a rare archaeological example ofpossibleexploitationofspinifexseedsforfood.

There are about 14 species of Triodia and Plechtrachne genera (commonly calledspinifex) in central Australia and it is notoriously difficult to determine species (Latz1995: 288). There are at least three species in the Pilbara: Triodia wiseana (hardspinifex),Triodiapungens(softspinifex)andTriodiaepactia(softspinifex).AboriginalpeopleuseSpinifex foranextraordinaryrangeof things: food(seedsand internodes),medicine(i.e. resin impregnatedanthillsandsmoke), shelter,bedding(!),glue (resin)for tools andornaments,waterproofing of rafts, huntinghides, torches and firesticks,nets(hardspinifexclumps)andfibres(Gott1992;JuluwarluAboriginalCorporationnd;Latz 1995). Pitman andWallis (2012) list 11 main categories including reference tospinifex processing using stone to grind clumps of leaves (for fibre and string) andseeds (for food). The presence of spinifex phytoliths on grinding patches raises aninterestingquestionabouttheusebyAboriginalpeopleofspinifexseedsinthePilbaraandelsewhereinAustralia.BethGott’sdatabaseofAboriginaluseofplants(Gott1992)documentedfewspecificreferencestogrindingspinifexseedsinsemi‐aridmarginsofthe continent: in northern Flinders Ranges (Cleland and Johnston 1939) andnorthwesternNSW(Turner1905).

17

IncentralAustralia,Latz(1995:291)notedthatthe‘Warlpiriarereputedtoeatspinifexinternodesintimesofhardshipbutconsiderthematasteless,inferior,stopgapfood....Aboriginalpeopleinthisareadenyanyuseofspinifexseedsforfood’.

InnorthwesternAustralia,intheHamersleyPlateauofthePilbara,JuwularluAboriginalCorporation (nd:118) documented edible seeds including Acacia inaequilatera, colei,pyrifolia,tumida,Panicumdecompositum,Calandrinapolyandra,PortulacaoleraceaandEucalyptus camaldulensis; but no spinifex. However, Brehaut and Vitenbergs (2001:146)documentthatedibleseedsreportedbytheGurumaelders includedhillspinifex(Triodiasp.) (Gurumaname:paru).[ThislistofedibleplantswascompiledmostlybyPeter Stevens and Nelson Hughes with identification assistance from Stephen vanLeeuwen and Michael Hughes fromWestern Australia’s Department of ConservationandLandManagement.]

ThereisalsoindirectevidencethatSpinifexseedsweregroundforfoodinthePilbara.AlthoughVeth(1993)andWalsh(1987)notednospinifexseedsintheMarturesourceschedule, there is evidence from the Burrup. Bird and Hallam (2006:5) referred toTurner(1981)andreportedthat ‘AstudyofthegrindingpatchesrecordedduringtheDampier Archaeological Project in the early 1980s showed that they were mostcommonincampingareasclosetospinifexgrasslands,suggestingthattheyweremainlyusedforgrindingspinifexseedsintoflour.Manygrindingpatcheshadclearlybeenusedoverlongperiodsoftime,fromtheamountofwearandthefactthattheirsurfaceshadoften been rejuvenated and re‐roughened through pecking or incising lines.’ TheJaburara Heritage Trail brochure also says Aboriginal people harvested spinifex forseeds.ItislikelythatspinifexseedswereusedforfoodinthePilbaraparticularlywhennothingelsewasavailableintherequiredquantityatparticulartimesoftheyear.

PeterVeth(Vethetal.1993andindiscussion)hassuggestedthatthesemi‐aridmarginsaroundthearidcoremayhavebeenhardertoliveinatcertaintimesbecausetheyarelesspredictable (whenandwhere rainfallhappens); and largeaggregationsofpeopleclose to the grinding patches are suggested by the vast body of engravings,which isindeedthecaseintheWoodstock/Abydosstudyarea.Followingthisargument,bedrockgrindingpatchesonthishugescaleareasignatureofsubstantial foodproduction(i.e.seed cakes or damper) to feed large gatherings, made possible by the peculiarconfigurationofsocialnetworks,seasonalresourceavailability(waterandgrassseeds),androckformations(bothgranitedomesforgrindingpatchesandgranophyrebouldersforrockart.).Basedonethnographyandoraltradition,itislikelythattheselargeareasof intensive seed processing reflect women’s activities, to feed families involved inceremoniesassociatedwiththerockart.Domesticcampsitesarelikelytobeassociatedwiththeseseedgrindingactivities.

Conclusions

18

Other than noting the presence of grinding patches andmeasuring their dimensions,researchintoAboriginalgrindingpatchesistypicallyminimalinAustralia.Inthisstudywe present a methodology for recovering PVS peels of grinding stone usewear andaqueous solutions of phytolith residues in order to allow the further study of suchfeatures to assist in the determination of their possible function(s). Seed grindingexperimentswereundertakenwithgraniteanddoleritetocomplementothertool‐useexperiments.Usewearstudyof95grindingpatchesat10sitelocationsindicatesthatallwereprobablyusedforgrindingseeds,althoughtheincidenceofstriationsandpolishdevelopment is variable andmay relate to different task conditions, or, alternativelyothertaskssuchastheprocessingofspinifexleavesandstemsforfibres.

Phytolithanalysisofresiduesrecoveredfromthesurfacesofgrindingpatchesindicatesthedominanceofgrassphytoliths.Despiteunresolvedtaphonomicconditions inopensite settings with likely contamination from rain and wind, the grass phytolithassemblagesupports theusewearstudyandsuggests theuseofSpinifex (TriodiaandPlechtrachnegenera)inthePilbaraforseeds.Newmicro‐stainingtechniquesdevelopedby Birgitta Stephenson (2011) have high potential for reliable identification othergrindingstonetracesandfurtherevaluatingthefunctionofthesesites

A high frequency and extensive distribution of grinding stones (including bedrockgrinding patches) are probably linked intimately with food production for largegatheringsofAboriginalpeople.

Acknowledgements

Initial fieldwork for this study in December 2006 included Karriyarra communitymembers Andrew Gordon and Bruce Bung, Alan Watchman, Boone Law, RichardFullagarandGavinJackson(whoundertooksubsequentfieldwork).

Foradviceandpermissionstopublishthisinformation,weareparticularlygratefulto:

TheKarriyarrapeople(inparticularDonnyWilson,Elder)whohasbeensupportiveofouraimtolearnhowpeoplewereliving‘insuchhardcountry’andtoletotherpeopleknow. He discussed the presented paper and carefully edited some of the text andimages.

ThePalykupeople(inparticularDavidMilroy,ChairPalykuNativeTitleWorkingParty).Davidaskedmetoincludethistextwiththepaper:‘OurRockart,grindingpatchesandcultural material are not only a record of Palyku past, but also an essential part ofPalyku'spresentdayandfuture.WethankRichardfortakingthetimetotalktousandwehopethismaterialwillhighlighttheneedforfurtherprotectingPalykuHeritageandCulture.’

19

WealsothankAlexaMorcombe,FMGforpermissionstopublishthedetailincluded.Wealso thankvariouspeoplewhoprovided logistical supportand/or technicaladviceonthis paper, including: ShaunCanning (ACHM); Phil Czerwinski (ACHM);NealeDraper(ACHM);Gavin Jackson (GJPtyLtd);BooneLaw(ACHM)whosephotographsarealsodulyacknowledged;BlairMcGlew(FMG);PeterVeth(ANU),andAlanWatchman.

We are both indebted to Beth Gott whose friendship and research have beeninspirational. Her botanically focussed publications and meticulous databases haveinspired us here to focus our discussion on one plant taxon, spinifex. We are alsograteful to organisers and fellow participants at: A Symposium inHonour of Dr BethGott, Ethnobotanist, Friday 12th September 2008, 9:30am – 5:30pm at the KoorieHeritageTrust,MelbourneVictoria.

References

Bird, C. and S. J. Hallam. 2006. A review of archaeology and rock art in the DampierArchipelago.UnpublishedreportpreparedfortheNationalTrustofAustralia,Perth.

Brehaut,L.andA.Vitenbergs(eds).2001.TheGurumaStory.Guruma‐yharntuwangka.ToldbytheGurumaEldersGroup,ledbyPeterStevens.JukurrpaBooks,IADPress,AliceSprings.

Cleland, J.B.andT.Harvey Johnstone.1939.Aboriginalnamesandusesofplants intheNorthernFlindersRanges.TransactionsoftheRoyalSocietyofSouthAustralia63.2:172‐179.

Fullagar, R.L.K. 1991. The role of silica in polish formation. Journal of ArchaeologicalScience18:1‐25.

Fullagar,R.2006.Residuesandusewear.InJ.BalmeandA.Paterson(eds),Archaeologyin Practice: a Student Guide to Archaeological Analysis, pp. 207‐234. BlackwellPublishing,Malden.

Fullagar,R.2007.UsewearandresiduesongrindingpatchesintheWoodstock‐Abydosarea, Pilbara, W.A. Unpublished report prepared for Australian Cultural HeritageManagementPtyLtd.(contractedbyFortescueMetalsGroupPtyLtd.),Hindmarsh.

Fullagar, R. 2008. Usewear on Aboriginal grinding patch samples from the FortescueGroupRailConstructionCorridorProject,Pilbara,W.A.UnpublishedreportpreparedforFortescueMetalsGroupPtyLtd,EastPerth.

Fullagar,R., J.FieldandL.Kealhofer2008.Grindingstonesandseedsofchange:starchandphytolithsasevidenceofplantfoodprocessing.InY.M.RowanandJ.R.Ebling(eds),New approaches to old stones: recent studies of ground stone artifacts, pp. 159‐172.EquinoxPress,London.

20

Jackson,G.2006.ExtractoftheReportonAboriginalsitessubjecttosalvagewithintheKariyarra andPalyuku sections of the proposedCloudbreak to PortHedlandRailwayConstructionCorridorintheWoodstock/AbydosProtectedArea,EastPilbaraRegions,WesternAustralia.

Gorecki, P. and M. Grant. 1994. Grinding patches from the Croydon Region, Gulf ofCarpentaria. InM.Sullivan,S.BrockwellandA.Webb(eds),ArchaeologyintheNorth:Proceedingsofthe1993AustralianArchaeologicalAssociationConference,pp.227‐38.NARU,Darwin.

Gorecki, P.,M. Grant, S. O’Connor andP. Veth. 1997.ThemorphologyandantiquityofAustraliangrindingimplements.ArchaeologyinOceania32.2:141‐150.

Gott, Beth, 1992.EthnobotanyofSouthEastAustralia:NSWUSEandSAUSEDatabases.AIATSIS,Canberra.

Hart,D.M.andL.A.Wallis(eds).2003.PhytolithandStarchResearch intheAustralian‐Pacific‐Asian Regions: the State‐of‐the‐Art. Pandanus Press, Australian NationalUniversity,Canberra.

HickmanandLipple.1978.MarbleBar.GeologicalSurveyofWesternAustralia1:250,000GeologicalSeries‐ExplanatoryNotes.SheetSF50‐8.Perth.

Kamminga,J.1979.Thenatureofuse‐polishandabrasivesmoothingonstonetools.InB.Hayden (ed), LithicUse‐wearAnalysis, pp. 143‐157. Academic Press,NewYork.Korff,Jens. Jaburara–Heritage Trail Brochure. Retrieved fromhttp://www.creativespirits.info/ozwest/karratha/jaburara.html.

Mandikos,M.N.1998.Polyvinylsiloxaneimpressionmaterials:Anupdateonclinicaluse.AustralianDentalJournal43.6:428‐34.

Mathews,A.G.1964.ThepreparationofspinifexgumbyAustralianAborigines.WesternAustralianNaturalist9.

Stepehnson, B. 2011. In the groove: an integrated functional analysis of arid zonemillstones fromQueensland.UnpublishedBA(Hons)Thesis,UniversityofQueensland,StLucia.

Turner, F. 1905. Botany of the north‐western New South Wales. Proceedings of theLinneanSocietyofNewSouthWales30:32–91.

Turner, J. 1981. Murujuga: a spatial analysis of the engraved rock of Withnell Bay.Unpublished B.A. (Hons) Thesis, Department of Anthropology, University ofWesternAustralia,Perth.

Parr, J.F. 2006.Effect of fire on phytolith coloration.Geoarchaeology:An InternationalJournal21.2:171‐185.

Pitman,H.T.andL.A.Wallis2012Thepointofspinifex:areviewofAboriginalusesofspinifexgrassesinAustralia.EthnobotanicalResearchandApplications10:109‐131.

21

Roth, W.E. 1905. North Queensland Ethnography Bulletins 1‐18. Records of theAustralianMuseumpublishedbetween1898‐1910,AustralianMuseum,Sydney.

Latz, P. 1995.Bushfiresand bushtucker:Aboriginalplantuse inCentralAustralia. IADPress,AliceSprings.

Juwularlu Aboriginal Corporation. 2003.Wangalili: Yindibarndi andNgarluma Plants.JuwularluAboriginalCorporation,Roebourne.

Veth, P., E. Bradshaw, T. Gara, N. Hall, P. Haydock and P. Kendrick. 1993. BurrupPeninsula Aboriginal heritage project. Unpublished report to the Department ofConservationandLandManagement,Perth.

Veth,P.M.1993. Islands in the Interior: theDynamicsofPrehistoricAdaptationswithintheAridZoneofAustralia.InternationalMonographsinPrehistory,AnnArbor.

Wallis, L. 2008. Residue Samples from the FortescueMetals Group Rail ConstructionProject ‐Phytolithanalysis results.A reportprepared forFortescueMetalsGroupPtyLtd,EastPerth.

Walsh,F.J.1987.PatternsofplantresourceusebyMartujarraAborigines.UnpublishedM.Sc.(Prelim.)Thesis,UniversityofWesternAustralia,Perth.

Watchman,A.2007.Observationsonrockvarnishassociatedwithagrindingpatch inthe Woodstock‐Abydos area, Pilbara, W.A. A report prepared for Australian CulturalHeritageManagementPtyLtdforFortescueMetalsGroupPtyLtd,Hindmarsh.