. The Stratigraphy and Palaeontology of Teapot Creek, MacLaughUn River, NSW LeanneArmand'-^,W. D. L. Ride'and Graham Taylor^ 'DepartmentofGeology,AustralianNational University, Canberra,ACT0200; ^Current contactaddress: Institute ofAntarctic and Southern Ocean Studies, University ofTasmania, GPO Box 252-77, Hobart 7001, Tasmania,Australia ([email protected]); ^CooperativeResearch Centre forLandscape Evolution andMineralExploration, University ofCanberra,ACT2601 Armand, L., Ride, W.D.L. and Taylor, G. (2000). The stratigraphy and palaeontology of TeapotCreek,MacLaughlinRiver,NSW.ProceedingsoftheLineanSocietyofNewSouth Wales 122, 101-121. ThestratigraphyandtheQuaternarydepositionalhistoryofTeapotCreek,atributary oftheMacLaughlinRiverintheSouthernMonaro,southeasternNewSouthWales,aregiven anddiscussed.Otherfossil-mammalbearingdepositsintheMacLaughlinValleyarereported. ThevalleyofTeapotCreekcontainsasequenceofterraces.Thehighestandoldest oftheseischaracterisedbypoorlysortedconglomeratedepositstypicalofstream-deposited longitudinalbars.Palaeomagneticresultsobtainedfromthisterraceareinterpretabletothe Bruhnesnormalpolarityinterval(<0.78Ma).Asecond,lowerterraceissetintothehighest terrace.Depositionofthisterracebeganbefore5320±80yrBPandischaracterisedbyoverbank deposits.Thelowestandyoungestterrace,consistingmostlyoffineclaysandsilts,represents lateralaccretionsinsidemeanderloopsduringmodemfloodingevents. The youngest terrace contains the remains ofmodern introduced mammals. The intermediateterracehasyieldedfossilsofmodemMacropusandanunidentifiedmurid.The highest terrace contains the remains of fossil mammals found in Plio-Pleistocene fossil deposits elsewhere inEasternAustralia. Species identifiedfromit areMacropusaltus, M. ferragus, M. titan, Procoptodongoliah,P.pusio, Protemnodonanak, P. "roechus/brehus", Sthenurus atlas, S. occidentalis, S. newtonae, Troposodon minor, andPhascolonus gigas, butnomodemspeciesofMacropuskangaroosorwallabies. Pleistocenefossilmammalshavebeenlocatedinfourothersitesinthevalleyofthe MacLaughlinRiver.Afifth site.ChalkPool, is sedimentologicallydifferentandatalower levelrelativetothemodernriverthantheotherterraces.SpeciesidentifiedintheChalkPool depositareregardedasPlio/Pleistoceneforms(Macropuspan,Protemnodonchinchillaensis andP. "roechus/P. brehus"); ofthese,P. chinchillaensisisknownonlyfromthePliocene. Manuscriptreceived 1 July2000,acceptedforpublication 13 November2000. KEYWORDS:ChalkPool,Macropodidae,Marsupial,Monaro,Pleistocene,Pliocene,Teapot Creek,Vombatidae. INTRODUCTION Background The Monaro region is the highest part of the southern tablelands ofNew South Wales. Workonmammalfossils fromalluvialdeposits in theregionhaspreviouslybeen confinedtoastudyonthenorthernpartoftheregion (Rideetal. 1989; Davis 1996; Ride andDavis 1997), althoughinthe 1930'sfossilmammalremainshadbeencollectedfrom Proc.Linn.Soc.n.s.w., 122.2000 102 MACLAUGHLINRIVERFOSSILDEPOSITS a gravel pit at Jincumbilly in the southern Monaro (AustraHan Museum specimens and localpressreports).Morerecently,in 1987,studentsoftheUniversityofCanberra,engaged in geological mapping in the region, reported finding a large bone fragment (a diprotodontid)inBungarbyCreekonBrooklynStationandmacropodinefossilsinTeapot Creek on Sherwood and Boco Stations (Fig. 1). Further studies ofTeapot Creek were made principally by the first author (Dansie 1992). During these studies a further four sites were located, close by in the valley of the MacLaughlin River, of which Teapot Creek is atributary. In this paper, a study of the stratigraphy and palaeontology of Teapot Creek is presented. Two "'C dates arerecordedfortheintermediateterraceofTeapotCreek, anda palaeomagnetic result for the highest terrace. The other localities and their fossils are also listed briefly and theirrelationship to theTeapot Creekdeposits is discussed. General description Basalts of Paleocene to Oligocene age dominate the valley of the MacLaughlin River (K-Ar dates of 51±0.3 and 37.2±1.0 Ma, Taylor et al. 1990; Roach et al. 1994). Theseoverlie lacustrine sediments. The basalts provide theprincipal sourcematerial for theQuaternaryhillslope (colluvial)andalluvialdepositsofthevalley.Thebasalts, which are also the basement rocks at the Quaternary and Holocene fossil sites described here, are considered to have been derived locally from multiple eruption sites. Teapot Hill, closetothejunctionofTeapotCreekandthe MacLaughlinRiveris one oftheseeruption sites (Roach 1991; Roach et al. 1994) (Fig. 2). The lacustrine sediments which underlie the basalts are believed to represent the western edge ofa palaeolake (Lake Bungarby, Taylor et al. 1990). These sediments are thoughttobe the sourceofquartz sands andpebbles foundas ararercomponentin some ofthe post-basaltic deposits. Similarsub-basalticlacustrinedepositsoccurnearbyinthevalleyoftheMacLaughhn River. These contain a diverse fossil macroflora (Taylor et al. 1990; Hill 1991; Hill and Carpenter 1991; Christophel 1994) andonthebasisofcontainedfossilpollen,havebeen assignedtotheLygistepollenitesbalmeiZoneofthe late Palaeocene (Tayloretal. 1990). The sediments underlyingthebasaltsinTeapotCreekarethoughttobeextensions ofthe same deposits. WeinferthatduringthelateQuaternary (<75 ka)theareawas subjecttoperiglacial conditionsonseveraloccasions. IntheKosciuszkoregion,recentmoraineexposuredates haveindicatedanearlyglaciationphaseat-55-65 kaandalateglaciationperiodbewteen 15-35 ka with three progressively less extensive glacial advances occuring at 32,000 ± 2500, 19,100 ± 1600 and 16,800 ± 1400 years ago (Barrows et al. in press). Similar findings from glacial records are reported from the Tasmanian Highlands (Fitzsimmons andColhoun 1991; Colhoun2000). SedimentsintheACTreveal severalperiodsofslope instabilitythatarethoughttoindicateseparateperiodsofperiglacialclimatesinthesouth eastern highlands (Barrows 1995). In addition, the Monaroregion lies in therainshadow inducedbytheAustralianAlps(Tulipetal. 1982;Rideetal. 1989).Whiletherainshadow may have ameliorated the effect ofclimate shifts, the late Quaternary was undoubtedly cooler and drier than at present with attendant higher sediment mobility. Under these climatic influences erosion and alluviation have shaped the valleys. RESULTS Geomorphology ofTeapot Creek The TeapotCreek catchment covers 7 km' on Sherwood Station, approximately 12 km south-west of Nimmitabel in the southern Monaro (Fig. 1). Teapot Creek rises as channels in the steep, hilly terrain ofSherwins Range at an elevation of950 m, and over its length of4 km, falls to below 780 m to a wide, gently sloping valley at itsjunction with the southerly flowing MacLaughlin River, the latter being a tributary ofthe Snowy Proc.Linn.Srx;.n.s.w., 122. 2000 L.ARMAND,W.D.L.RIDEANDG.TAYLOR 103 Figure 1.TopographicmapindicatingfossilsitesandtheTeapotCreekstudyareadescribedinthetext.Sitesare identified as follows: 1= Bungarby Creek, Brooklyn Station, 2= Chalk Pool, 3=Toppings Creek, 4= Ben's Bin,5=RailwayCutting,6=TeapotCreekStudyArea,enclosedbybox.Sites2to6arelocatedonSherwood Station.Mapreferencescale:AustralianMapGrid(U.TM.) intervalsof 1000fromOriginofZone55. Proc.Linn.Soc.n.s.w., 122.2000 104 MACLAUGHLINRIVERFOSSILDEPOSITS Trib.p'- / Trib.E a. Trib.cVJY^>.,. b. Trib.F 16#-^17 abandoned ^.^^^ • T^Teapot channels "MCreek i^^ 14i Jl3 • 11 10^ 9| 8 ) Trib.B \ y^ - wj(/^^ '%^ V abandoned #4 channels Teapot / rZ ,.- Trib.H Creek >3 ''^<%P!=-a=>^- \^ ^ Trib.A V KEY \ _^_^ I Terrace 1 Basalt '^c"? Terrace2 Alluvialfan ^ ^0 Alluvium/ Terrace 3 Colluvium Figure 2a: Teapot Creek fossil localities described in text. 2b: Corresponding generalised cross sections of TeapotCreek (from 2aj illustrating alluvial terraces. Illustrationsin2barenottoscale. River. The modem TeapotCreek flows down the north-eastern sideofits valley whereit incises asequenceofgently sloping alluvial plains. Lateral fansmerge withtheseandare incisedbysmall tributaries(Figs 1 and2).ThedepthofincisionbyTeapotCreekdecreases down stream from approximately 15 m to 2 m. Alluvial fans merge laterally with the alluvial plain along the lower valley, where current entrenchment of the stream shows some evidence ofmeandering prior to its present incision. Proc.Linn.Soc.n.s.w., 122. 2000 L.ARMAND,W.D.L. RIDEANDG.TAYLOR 105 Thefossil-bearingsitesarelocatedinbanksexposedbytheentrenchmentofTeapot Creek.Themoderncreekis confinedto a singlemajorchannel (approximately 1 mwide and about50 cmdeep). Flooding in the creekraises the waterlevel upto 2 m. Currently, TeapotCreekisasinuousstreamwhichhaslowpoint-barsandthalwegdepositsofbasalt- derived gravel, sands and mud, and deep, steep entrenched banks. Basalt crops out in manyplacesalongthecreek,evenintheupperreaches,whereentrenchmentdowntothis level has produced well worn surfaces and minorpools. Sequence and sedimentology oftheAlluvialTerraces Stratigraphic information forTeapot Creek valley was obtained from exposures in escarpments, cut stream banks, and four trenches excavated with a back-hoe. From the information gathered, terrace sequence was determined and an informal stratigraphy provided (Dansie 1992). There is no formal stratigraphy associated with the deposits of the Southern Monaro and the description of stratigraphy is hindered by the lack of superposition.Severalfaciesareobservedineachofthesedimentaryunitswithinexposures andtheirdescriptionsfollow lithofaciesdefinedbyMiall (1977, 1978), andRust (1978). Three alluvial benches are identified along the length of Teapot Creek. These morphological terraces areinset withinone anotheralongtracts ofthecreek(see section W-XinFig.2b).Relativeelevationsandlithostratigraphicsimilarityoftheexposedterraces determined correlation of the terraces along the creek. Figure 2b shows a generalised valley-fillsequenceforTeapotCreekatthreepointsofitslength.Eachterraceiscomposed of a single alluvial morphostratigraphic unit often overlain by colluvial and alluvial deposits, which has been derived from the valley sides and during flood events. The contact between the terrace deposits and the overlying colluvial/alluvial veneer varies fromdiffuse,incaseswhereintensepedogenesishasoccurred,tosharphiatus.Theterraces alloverlieabasaltbasement, exceptatone site (Fig. 2a, site5) wheretheterraceoverlies amixedquartz sand andbasalt-derivedclayey facies, consideredtohavebeenreworked from the Palaeocene lacustrine deposits and surrounding basalt material. Terrace 3 is the highest alluvial deposit in the area. It is predominantly a fluvial deposit, containingcrudelybeddedconglomeratic units withweaklyimbricatedgravels. EachexposureofTeixace3 (Sites3 to7,9to 12, and 14to 17)hasauniqueinternalfacies sequence, but at all sites examined there is a comparable mix of facies consisting of: matrixsupported,poorlybedded,medium-tocoarse-grainedwellsortedchannelgravels; andpebbly muds and very fine sands ofoverbankdeposits. Theconglomerate facies are % predominantlybasaltderivedbutincluderare(< 1 ofconglomeratecomposition)quartz andjasperpebblesindicatinganOrdovicianprovenanceforsomematerial.Granitepebbles found at Site 7 indicate that material possibly derived from the Silurian aged Berridale Batholith, was available at the time ofdeposition. However, source areas for such older rocks are unknown in the present catchment. It is possible they are recycled from pre- andintra-basaltic alluvial deposits, which are common throughoutthe Monaro. Soils on Terrace3 varyfrom5 cmto370cminthickness,andhaveanaveragethicknessof60cm They comprise a dark, yellowish brown clay with discrete calcrete nodule layers. In instanceswherethesoilshavebeeneroded,thesecalcretebandsareexposedonthesurface. Multiple channel deposits are common within Terrace 3, indicating that they representalongperiodofdeposition.Terrace 3 sitesincludethebulkofthe stratigraphic sections investigated (Fig. 2). Site 13 is thoughtto representTerrace 3, but it occurs out m ofstratigraphic context on top ofa 15 high escarpment, and in addition its sediment characters remain undetermined due to its inaccessibility. AtthetimeTerrace3wasdeposited.TeapotCreekwasabraidedstreamcomparable totheScotttype(proximalbraidedstream)ofMiall(1977, 1978, 1982),whichisindicative ofscour-based,massiveconglomeratebedsoverlainbysandyfaciesthatarethenscoured by superimposingbars.Thecyclicity intheTerrace 3 deposits is interpretedasrecording rapidaggradationfromfloodingevents.Thebraidedstreamisestimatedtohavebeen40- m 50 wideintheupperreachesofthecreek(Sites 15 and 11,Fig. 2a),whichhaddecreased Proc.Linn.Soc.n.s.w., 122.2000 106 MACLAUGHLINRIVERFOSSILDEPOSITS m inbreadth to around 15-20 at Site 7 (Fig. 2a). Terrace 2 is distinguished from Terrace 3 by its lower elevation and by black- coloured fine-grained sediments and conglomeratic channel deposits. The colour is the resultofahighorganiccontent.Atsites 1 and2 (Fig. 2a),Terrace2showsfiningupward sequencesoffinelylaminatedclays,siltsandsands,andlowangledcross-beddedcoarse- to fine-grained sands. Both facies indicate a stream that developed a significant flood plain with waning flow. The overbank deposits increase in thickness up-sequence, and contain charcoal deposits, and evidence of bioturbation. Conglomeratic deposits are generally ofwell-rounded and sorted basalt clasts. The area ofdeposition ofTerrace 2 decreases upstream in Teapot Creek as the valley decreases in width and depth and the entrenchment of the creek increases. The terrace is also found in tributaries ofTeapot Creekbut similarlydisappears upstream. Soildevelopmentonthe surfaceis upto 37 cm deep and is composed ofbasalt-derivedblackclays. Terrace 1 is the lowest terrace and occurs on the inside bends ofmeander loops, where it forms lateral accretion deposits of fine clays and silts. It is very young by comparison with Terrace 2 and contains remains of modern species introduced by Europeans. Terrace 1 occurs upstreamofSite 8 (Terrace 2) and along mosttributaries in the study area. Itis coveredinplacesbyrecently depositedflooddebris. Chutechannels often cross some ofthe Terrace 1 benches down stream in the open valley (abandoned channels in Fig. 2b.) Minimal black clayey soil development occurs at the surface of theseterraces,havingamaximumdepthof10cm,whichisfrequentlydisturbedbymodem flooding events. Such development is termed stratic: a low alluvial bench frequently flooded, ofraw, pedologicallyunmodifiedalluiviumthatisencompassedbyalluvialsoil classification (Walker and Coventry 1976: 307). Terrace chronology Terraces 1 and 2 both occur close to the stream bed. Terrace 1 in the downstream reachandTerrace 2 fartherupstream. Terrace 3 occurs well aboveTerraces 1 and2 (Fig. 2b).Progressiveloweringofthealluviallandscape,producingtheseyoungerinsetterraces, relates to the reduction of stream discharge due to late Quaternary change in climate. Charcoal and sediment from one stratigraphic 'bed' (Fig 2a, site 2) were obtained and senttoBetaAnalyticInc. (Florida,U.S.A.)fordating.Twocarbondatesataapproximate depth of 550 cm and 560 cm, provided ages of 4940±140 BP (TPC/B/1- Beta-50156, charcoal) and 5320±80 BP (TPC/B/2-Beta-50157, sediment) respectively. ThermoluminescencedateswerenotobtainablefromthefluvialdepositsofTeapotCreek because the deposits lack sufficient quartz material and no molluscan remains were encountered, thus eliminating the possibility ofdelta'^0 analysis. Terrace 3 is predominantly composed ofchannel deposits and occurs high in the landscape relative to Terrace 2. Palaeomagnetic samples were taken from fine-grained facies within fossil-bearing Terrace 3 conglomerate beds at Sites 4 and 15 (Fig. 2a). These samples were then step-wise alternate field demagnetised and measured on a MolspinspinnermagnetometeratthejointAGSO/ANUPalaeomagneticLaboratory,Black Mountain, Canberra. Palaeomagnetic results indicate a normal polarity for fine-grained facies from these sites, which although not conclusive, is interpretable to the Brunhes normal polarity interval (<0.78 Ma) (B. Pillans pers. comm. 1996). Teapot Creek fossil mammals: identification and physical condition From the Teapot Creek study area (Fig. 1, no. 6) 244 fragmentary specimens of fossil mammals were collected between 1988 and 1995. Ofthis total only 62 specimens have been identified to family, and only 42 to species. Of the latter, all but Macropus titan are sparsely represented (in some cases only by a few parts ofbroken molars). The probable causes of the fragmentation include transport by water and post- depositional weathering by episodic clay expansion and contraction, and also calcrete impregnation. There does not seem to be any evidence ofbite marks on the bone. There Prcx;.Linn. Soc.n.s.w., 122. 2000 . L.ARMAND,W.D.L. RIDEANDG.TAYLOR 107 is little evidence ofbone mineralisation, butpartial 'opalisation' is noted in some ofthe teethrecovered. Species identifications ofthe fragmentary material are presentedhere on the basis thateachspecimenisnotmorphologicallyseparablefromanequivalentpartofasecurely identified specimen of the species named (preferably the holotype or a topotypical specimen). To enable the identifications to be evaluated by subsequent workers, each species listed is accompanied by the number and a briefdescription ofthe specimen(s) on which the species identification principally depends (the voucher specimens), the collection number of the museum specimen with which it was compared, and a bibliographic reference to adescription ofthe diagnostic characters. Voucher specimens forall taxa identified are deposited in theAustralian Museum, Sydney. All taxa listed above were identified by the senior author and Ride; the latterthen visitedmuseumcollections andmadedirectcomparisons with identified specimens (and whereverpossiblewithtypesandtopotypicalspecimens). ComparisonswithSthenurinae were confirmed forus by Dr G. Prideaux. Noformalfaunisticnameis given in this workto the species assemblagesreported fromthethreeterraces(thespeciesassemblageofTerrace3 depositshaspreviouslybeen referred to by Dansie (1992) as theTeapotLocal Fauna). TaxonomicattributionstogeneraarethoseadoptedinRideetal. (1989).Macwpus titan andM. altus are both morphologically distinctfromM. giganteus andM. robustus and are conventionally treated as full species. The use ofthe names Procoptodon and P. pusio are maintained as cases under consideration by the International Commission on Zoological Nomenclature, in accordance withArticle 82.1 ofthe International Code of Zoological Nomenclature (Davis andRide 2000). Fossil occurrences in theTeapot Creek Terraces. Fossil mammals recovered from each of the three Terraces are listed in Table 1 Terrace 1 contained only the remains ofmodern introduced species. Terrace 2 deposits (Site 2)haveyieldedthreefossils only. Ofthese, only partofapelvishasbeenidentified to agenus (Macwpus). The presence ofaspecies ofthe modernkangaroo (M. giganteus orM. robustus) inthe deposit, which has aprobable minimum age ofbetween4660 and 5480 BP (see above), is consistent with the presence ofmodem species ofMacwpus in similarly dated deposits in the northern Monaro (Ride et al. 1989; Davis 1996). AllotherfossilsintheTeapotCreekvalleywererecoveredfromTerrace3 deposits. Many fossils werecollectedat sites fromwhich theyhadweatheredout, butothers were excavated directly fromwithin the deposits; those excavated were foundpredominantly withinconglomeraticfacies and, less frequently, inoverbankandsandypoint-barfacies. The overbank and sandy point bar facies yielded the least damaged fossils, yet such facieswererareinTerrace3 deposits (beingconfinedto Site4) and,eveninthosefacies, the bone structure was crackedby clay expansion and desiccation. As aresult, itis very unlikelythatfossilmaterialwouldhavebeenreworkedfromTerrace3toTerrace2deposits without disintegrating. No fossils in that condition were found inTerrace 2. Specimenscollectedfromthe surfaceinthevalleyofTeapotCreekby students and others mapping in the area are all attributed by us to Terrace 3 deposits. Information given by the donors satisfied us that all came from upstream ofthe Site 2 (Terrace 2). Identifications ofthese arerecordedafterthe siterecords underthe heading "Terrace 3 - unlocalized". Otherfossil sites in the valley ofthe MacLaughlin River Five otheralluvial sites in the valley ofthe MacLaughlin River, southern Monaro, haveyieldedfossilmammals(Fig. 1: Ben'sBin,RailwayCutting,ToppingsCreek,Chalk Pool, and Bungarby Creek-the firstfourare within Sherwood Station andthe lastis on Brooklyn Station). At Ben's Bin, Railway Cutting and Toppings Creek sites the fossils are eroding fromcutbanks.Thefossilsareveryfragmentedanddisassociated, andallidentifications Proc.Linn.Soc.n.s.w., 122.2000 108 MACLAUGHLINRIVERFOSSILDEPOSITS Table 1.SpecieslistsofTeapotCreekSitesbyTerraceandsiteaffiliation.Thenumberofspecimens, whetheridentifiableornot,fromeachsiteisprovidedin column2 andincludesbundledcollectionsofbone shardsrecoveredfromthesites.Acompleteindexofthespecimensandmaterial collectedisfoundin Dansie(1992). Speciestaxonomy (authorship anddates)ofthespecieslistedinTaMe 1 isformallytreated intheAppendix. Sites No. specimens Specimens identified tospecies recovered level Terrace 1 Between Sites 1 and8 not recorded Vulpesvulpes Bostaurus Ovisaries Terrace2 Sitel Site2 3 Macropus (eithergiganteus orrobustus) Terrace3 Site3 1 Site4 to6 28 Macropus titan Protemnodonanak Protemnodon "roechus/brehus" Procoptodon goliah Sthenurusoccidentalis Troposodonminor Site? 21 Phascolonusgigas Protemnodonanak Procoptodonpusio Macropus titan Macropusferragus Site9 5 Macropus titan Site 10 3 Site 11 3 Macropus titan Site 12 10 Sthenurusnewtonae Site 14 12 Macropus titan Site 15 8 Macropus titan Macropusferragus Site 16 13 Macropus titan Site 17 2 Macropus titan MacropusaItus Terrace3 - unlocalised 35 Protemnodon "roechus/brehus" Macropus titan Macropusferragus Sthenurusatlas Proc. Linn. Soc.n.s.w., 122.2000 L.ARMAND,W.D.L. RIDEANDG.TAYLOR 109 are based on isolated teeth or fragments of teeth and a few more complete dentitions (suchas several fromToppings Creek). On thebasis ofthefossils the ages ofthesethree sites are not different from Terrace 3 sites in Teapot Creek and the deposits are either remnantpoint-bar sequences orconglomeratic thalweg deposits. An exposure at Chalk Pool, on the MacLaughlin River, close below its junction with Teapot Creek (Fig. 1), is significantly different from other sites in the region. The site occurs at lowerelevations (between 750 and 760 m) than any otherfossil site in the MacLaughlinRivervalley.Herethefossilsarealsocontainedinpointbarorconglomerate facies. Materialfromthepointbarisencasedina2-4cmthickcalcareous shelloforange coloured,medium-sizedsandgrains.Jawsthatcontainportionsofmolarrowsandpartial longboneshavebeenrecovered. Fossils ofatleastthree speciesareidentifiedandall are known fromthe Plio-Pleistocene elsewhere (one only from the Pliocene). Two fossil post-cranial fragments have been recovered from a site in Bungarby Creek, onBrooklyn Station- also alargelyconglomeratic deposit. Neitherspecimen has been identified to species, although one is certainly a large diprotodontid. Although attemptsatlocatingotherfossil-bearingdepositsinthevicinityonBrooklynStationhave beenmade,nootherfossilsorlikely siteshavebeenlocated.Thespeciesfromthese sites are listed inTable 2. DISCUSSION Interpretation offossils All but one species of fossil mammal {Phascolonus gigas, the giant wombat), recovered so farfrom Terrace 3 deposits ofTeapotCreek are Macropodidae (kangaroos andwallabies,andshort-facedkangaroos). Elevenspeciesoffourmacropodidgeneraare represented. The most common is Macropus titan; the other species comprise less than 35% ofindividuals identified (i.e., Macropusaltus, M.ferragus, Procoptodon goliah, P. pusio, Protemnodon anak, P. "roechus/brehus", Sthenurus andersoni, S. atlas, S. occidentalis, S. newtonae, Troposodon minor). From Murray's estimates ofweights of fossil mammals (Murray 1991:1155 - 56, tables 16, 17) most, when adult, would have beenbetween50and 100kginweight. BecauseallotherPleistocenefossildepositsinthe EasternHighlands(includingthenorthernMonaro-seeRideetal. 1989)containnumbers of smaller species, the composition of the fossil sample indicates that the processes responsibleforits accumulation in theTerrace 3 deposits werehighly selective, eitherat death orpost-mortem (orboth); itis notpossible, atthisjuncture, to determine which of thetwo (orboth) occurred. The condition of the specimens indicates that they are derived from earlier accumulations. No specimens are articulated and all are fractured, probably the resultof theirhavingbeentransported. Butnothingindicatesthatthey are significantlyolderthan the sediments, orthat more than a short period oftime is represented by them. We have been unsuccessful in obtaining any dates from the fossils themselves. We attempted to have collagen dated from fossil bone collected from Terrace 3 (Site 7), but the dating laboratoryfoundthat"thecollagenfraction(basalboneprotein),whichisverysusceptible to decay when exposed to nature's weathering mechanisms, has decayed away entirely .(oralmost entirely)" (in litt. BetaAnalytic Inc., 2April 1993). No species inTerrace 3 deposits is representedby sufficient individuals toprovide areliable statementofthe distributionofmortality inthe population. However, themost numerous species, Macropus titan (26 identified individuals), allows some tentative interpretation of, at least, whether it represents a random sample of a stable breeding population. In the probably closely-related species, M. giganteus, the age ofindividuals may be established from molar progression (molar index; Kirkpatrick 1964, 1965a). In thesampleofM. titan,nospecimensaresufficientlycompletetoenablemolarprogression stages to be measured directly by Kirkpatrick's method, but the crown wear ofmolars enables stages of dental eruption and dental progression to be inferred assuming that Proc.Linn. Soc.n.s.w., 122. 2000 no MACLAUGHLINRIVERFOSSILDEPOSITS 1-100 « (0 -80 (0 ">O Macropus giganteus -60 percentage breeding 9- 1-40 -^ 0) 0) E 6- 03 20 0) 3- (0 —————— — — 0-' II r^- -i I I I I I r^-T I 1.39 2.42 3.01 3.44 3.77 4.04 4.27 4.47 4.64 Kirkpatrick molar index Figure 3. Histogram ofmolar indexes ofindividuals ofMacropus titan in theTeapot Creek, Terrace 3 deposit, shownagainstagraphindicatingthepercentageofbreedingfemalesatthesamemolarindexes in the closely related species, M. giganteus (data ofM giganteus from Kirkpatrick, 1965b; 50% of wearfollowsthepatternobservedinM. giganteus(detailsofthemethod,andjustification forit, are tobepublishedby RideandA.C. Davis in aworkonthe Quaternary vertebrate faunaofPilotCreek, nearCooma, N.S.W.). Estimates canbe made ofmolarindex from 17 ofthe 24 specimens. ThepatterninferredfromtheMacropustitansamplefromTerrace3 (Fig.3)indicates thatmostindividualsdiedafterM2haderupted, andbeforethedentinewasfullyexposed across both lophs (-ids) ofM4. In the case ofmodern M. giganteus, individuals ofthat age are activelybreeding adults (Kirkpatrick 1965b - i.e., individuals with amolarindex between 1.39and4.6).ThisconclusioniscomparablewiththoserecordedbyGillespieet al. (1978:1046); and Flannery and Gott (1984:413) for samples obtained from the PleistoceneLancefieldSwampandSpringCreekdeposits.Althoughtheseauthorsconclude that a more aged part of the adult population is represented in their samples compared with thatfromPilotCreek. Weareunabletocommentonthedifference,butsuchnarrow andunimodaldistributionsofdeathsrevealedbyallthreepopulationswouldbeunexpected iftheywere stablemammalpopulations withlongbreedinglivesundergoingprogressive mortality. In such cases, the remains of juvenile and aged animals are expected to predominate in samples (especiallyjuveniles - for the interpretation ofage distribution from similar samples, see Ride and Tyndale-Biscoe 1962; Hughes 1965; Flannery and Gott 1984; Turnbull and Martill 1988). Since the sample reveals neither a distribution pattern suggesting progressive mortality in a stable population, norone resulting from a catastrophic occurrence (such as aflash flood), eitherthe age structure ofthe population was disti:rbed irregularly as to annual increments, or some selective agent resulted in juvenileandagedclasses(particularlythejuvenile)tobeunder-representedinthesample. The Teapot Creek accumulation also displays a strong bias in favour of the representation of larger mammal species as does the Lancefield deposit (see Van Huet 1999:338-9). At Lancefield, the remains of small mammal species are present but they are comparatively rare (there, only 24 of317 individuals are of species smaller than M. titan which isrepresented by263 individuals; othermegafaunaintheLancefielddeposit, comparable with or larger than M. titan, are Diprotodon, Zygomaturus, Procoptodon, Sthenurus and Protemnndon). Proc.Lln.\.Soc.N.S.W.. 122. 2000