Records of the Western Australian Museum Supplement No. 57: 255-264 (1999).

Additions to knowledge about ektopodontids (Marsupialia:Ektopodontidae): including a new species Ektopodon litolophus

Neville S. Pledge!, Michael Archer, Suzanne J. Hand2and Henk Godthelp2

1 South Australian Museum, North Terrace, Adelaide, SA 5000; email: npledge@senet.com.au2 School of Biological Science, University of New South Wales, Sydney, NSW 2052

Abstract - Information about the extinct phalangeroid family Ektopodontidaehas been increased following the discovery of new material from severallocalities. A new species, Ektopodon litolophus, described on the basis of an Mlfrom the Leaf Locality, Lake Ngapakaldi, Lake Eyre Basin, is characterized bythe extremely simple structure of the crests. Ektopodontids are recorded forthe first time from the northern half of the Australian continent throughdiscovery of a tooth fragment at Wayne's Wok Site, Riversleigh WorldHeritage area, northwestern Queensland. Comparisons of Ml of Olllnia andEktopodon species now allow evolutionary trends of simplification to bediscerned.

• Riversleigh

QUEENSLAND,

~-------------------------,

: SOlJTH AUSTRALIA :

: ~. L. Ngapakaldi• L. Eyre :: • L. PalankarlnnaI f ------- __,,--- .. , .......: LFromet1 : .... -

: L. Tarkarooloo '. L~ p'. Inpa

Ektopodontidlocalities

Figure 1 Locality map of central and eastern Australia,showing all localities from whichektopodontids have been recorded.

million years; Woodburne et al. 1985), followingpreliminary analyses by W.K. Harris of pollen fromthe Etadunna Formation at Mammalon Hill, LakePalankarinna. Subsequent work with

AgeAlthough Stirton (Stirton et al. 1961, 1968) initially

assessed the age of the Etadunna Formation and itsfaunas to be Oligocene, it became customary tointerpret these central Australian deposits asMiddle Miocene in age (approximately 12-15

INTRODUCTION

Ektopodon is a genus of extinct possum-likemarsupials established by Stirton et al. (1967) onisolated teeth found at the Early to Middle MioceneLeaf Locality (Kutjamarpu Local Fauna) at LakeNgapakaldi, northeastern South Australia (Figure 1).Further specimens from this locality were describedand interpreted by Woodburne and Clemens(1986b), together with new, slightly older Oligocenespecies in the plesiomorphic genus CJmnia (c.illuminata, C. sp. cf. C. illuminata and C. omega) fromsites in the Lake Eyre Basin (e.g. Tedford Locality,Lake Palankarinna, Ditjimanka Local Fauna) and theFrome Embayment (e.g. Tom O's Quarry, LakeTarkarooloo, Tarkarooloo Local Fauna). A secondLate Oligocene species Ektopodon stirtoni wasdescribed by Pledge (1986) from Mammalon Hill,Lake Palankarinna (Ngama Local Fauna), with areferred form from Tom O's Quarry. Rich (1986)described Darcius duggani from Hamilton, westernVictoria, a deposit radiometrically determined to bePliocene in age (ca. 4.46 Myr).

By 1986 the rich deposits of Riversleigh,northwestern Queensland, were beginning to yieldmany new species including an ektopodontid. ThisRiversleigh taxon as well as a new, second,ektopodontid from the Kutjamarpu Local Faunaare described in this paper.

256

foraminiferans from underlying lacustrinedolomites (Lindsay 1987) suggested that at leastsome of these deposits might be as old as LateOligocene. This assessment was based on theabundant presence of the foraminiferanBuliminoides sp. cf. B. chattonensis. The older age isalso supported by Truswell et al. (1985) on the basisof pollens from the Geera Clay (which appears tobe a lithological equivalent of basal NambaFormation, itself correlated with the EtadunnaFormation); by Norrish and Pickering (1983) whoreported a Late Oligocene Rb-Sr date for illite fromthe Etadunna Formation; and Woodburne et al.(1994) reporting the palaeomagnetic results ofBruce MacFadden.

The Wipajiri Formation (Stirton et al. 1967)occupies a channel cut into the underlyingEtadunna Formation, and contains partly lithifiedpebbles of the latter sediments. The containedmammalian fauna is similar enough to theNgapakaldi and other local faunas of the EtadunnaFormation for Woodburne et al. (1994) to suggestclosely similar ages for the two formations. Thiswas predicated on a perceived brief span of time(about 2 million years) for the fossiliferous part ofthe Etadunna Formation, which one of us (Pledge)considers too short because of the numerouslacunae and erosional surfaces that can bediscerned within the sequence. Pledge's estimatefor the age of the Wipajiri Formation is Early toMiddle Miocene.

The age of the Wayne's Wok Ektopodon fromRiversleigh is equally contentious. At present,because of a majority of shared species, weconclude that the Wayne's Wok Local Fauna isapproximately equivalent in age to Riversleigh'sUpper Site Local Fauna which Archer et al. (1989)have concluded is probably Late Oligocene to EarlyMiocene in age. In part, this estimate is based onintercontinental comparisons of the Riversleighhipposiderid bats (including Brachipposiderosnooraleebus from Microsite) with their Europeancongenors (Sige et al. 1982). The Nooraleeba LocalFauna thus appears to be most probablyAquitanian (Late Oligocene) to Burdigalian (EarlyMiocene) in age.

Approximate correlation of the Upper Site andWayne's Wok Local Faunas from Riversleigh withlocal faunas from northern South Australia (Figure2) is based on shared faunal elements includingclosely-related species of Namilamadeta (Upper Site,Wayne's Wok, Tarkarooloo [Namba Formation]and Ngama [Etadunna Formation] Local Faunas);Wakiewakie (Upper Site and Kutjamarpu LocalFaunas); and Neohelos (Wayne's Wok andKuijamarpu Local Faunas).

Collection methodsSpecimens from the Leaf Locality were found by

N.S. Pledge, M. Archer, S.J. Hand, H. Godthelp

excavation of the fossiliferous horizons or byscreen-washing the dried sediment through finescreen with mesh of 6 x 6 wires per cm2, andpicking the concentrate. The Riversleigh specimenwas obtained by acetic acid digestion of thelimestone matrix, after masses of fossiliferous rockhad been blasted from the outcrop.

TerminologyDental homology follows Luckett (1993). The

terminology for the crown morphology ofektopodontid teeth used here is that of Woodburneand Clemens (1986a) with modificationsrecommended by Tedford and Woodburne (1987).Museum abbreviations are as follows: SAM, SouthAustralian Museum; QM, Queensland Museum;NMV, Museum of Victoria; UCB, University ofCalifornia, Berkeley; UCMP, University ofCalifornia Museum of Paleontology localitynumber; SAM PL, South Australian Museumpalaeontological locality.

SYSTEMATIC PALAEONTOLOGY

Supercohort Marsupialia (Illiger, 1811) sens.Cuvier (1817)

Order Diprotodontia Owen, 1866

Superfamily Phalangeroidea (Thomas, 1888) sens.Aplin and Archer (1987)

Family Ektopodontidae Stirton, Tedford andWoodbume, 1967

Ektopodon litolophus sp. novoFigure 3

Material Examined

HolotypeSAM P30176, an isolated right Ml collected by M.

Archer, S. Hand and others in 1982.

Type Locality and AgeBasal coarse clastic unit, Leaf Locality (UCMP V­

6213, SAM PL 7014), Wipajiri Formation, easternstrand, Lake Ngapakaldi, Etadunna Station, SouthAustralia. Kutjamarpu Local Fauna ofapproximately Early Miocene age (see above).

DiagnosisEktopodon differs from the other ektopodontids,

Chunia and (presumably) Darcius, in having anteriormolars wider than long, more than 6 cusps on theprotolophs, and with the accessory crests notradiating from the cusps but arranged more-or-Iesslongitudinally.

Ektopodon litolophus differs from all other

A new species of ektopodontid 257

Myr EPOCH VICTORIA SOUTH AUSTRALIA QUEENSLAND

Pliocene

Olway Basin

-+__Pl_e_is_lo_c_en_e_--lH .::;. Portland ~

-(Nelson Bay Fm)-

~ * Hamilton ~

(Grange Burn Fm)s-I--------t--------j

10-

Miocene

Callabonna Sllb~basil1 Tirari Sllb~basll1 Riversieigh

IRackham's Roostl

_?--?-

-------11 1I I, System C ,, I, 1I~ J

is -

20 -

25 -

f---?--?-

I*Kutjamarpu I(Wipajiri Fm) - - - - - --I

I--.-?0' ~ i System B i'-- I*Wayne's Wok I

f---? __ ? _ ~."-;reasur: I' :

1*Tarkaroo.loo II*- Ngama I:-;s~e~ ~ - :. --'-_.....J Ngapakaldi 1_ - - - - - -'

OligoceneEricmas

* Pinpa

(Namba Fm)

*Ditjimanka

Minkina

(Etadunna Fm)

("Carl CreekLimestone")

30 - -?--?----it--?--?--t--?--?-

BMcH

Figure 2 Stratigraphic distribution of ektopodontids. Occurrences indicated by':+::, '7' indicates formation boundaryage uncertainty. Modified in part from Woodbume et al. (1994). Local fauna names are shown in bold.'Minkina' is the new name for the 'Wynyardiid', pre-Di~imankalocal fauna.

Ektopodon species in the following features: largersize; Ml relatively and absolutely wider than thattooth in all species except E. serratus and 40% largeroverall than that species; parastyloph more regular,with cusps of more uniform size; more cusps (10)on main lophs; uniform size and withsimple pre- without obvious ribs,struts Or cusps);

EtymologySpecific name from the Greek Wos (plain or

simple) combined with lOp/70S (a crest, mane orridge), referring to the simple morphology of thecrests on the molar cusps, relative to those of otherspecies of Ektopodon.

DescriptionThe holotype is a first upper molar, which has been

258 N.S. Pledge, M. Archer, S.}. Hand, H. Godthelp

Figure 3 Ektopodon litolophus sp. nov., right Ml, SAM P30176, Leaf Locality, Lake Ngapakaldi, Ku~arnarpu LocalFauna. A, stereopair of occlusal surface, buccal side at top, anterior to right; B, lingual; C, anterior; 0, buccal;E, posterior views. Scale bar = 2 mm.

somewhat abraded, is relatively and absolutelylarger than that of any other known species (seeTable 1) and has a smoother, more triangular outline(Figure 3). The truncated buccal face is also smootherand more regular than in other species.

The crest of the metaloph is straight andtransverse while that of the protoloph curvesbackwards buccally and would intersect, just

beyond its buccal end, a line extrapolated from thecurving parastyloph. Protoloph and metaloph areabout equal in length, but slightly offset.

The parastyloph is about half the width of theprotoloph, with four cusps. The first cusp has ashort postcrista joining a lingual cingulum thatcrosses the end of the transverse valley to theanterior base of the protocone. It is separated from

A new species of ektopodontid

the slightly larger and bladed second cusp by ananterolingual crevice and a posterior crevice. Thecrest of the second cusp extends a.nterolingually tothe anteriormost point of the tooth and there is ashort postcrista. The third cusp is similar to thesecond but minute and squeezed between thesecond and fourth. The fourth cusp has a precristafrom its lingual side, a very short postcrista and ashort curving posterobuccal crest. The transversevalley is short, and none of the crests except thelingual cingulum crosses it.

The protoloph is almost straight, but its apicalcrest curves posterobuccally. There are 10 cusps.The protocone is by far tl1e largest and is markedby a deep anterolingual groove and a slightlylarger posterolingual groove. The pre- andpostcristae thus formed are inclined apically to thebuccal side of the tooth and are simple with nobifurcations, ribs or struts. Cusp 2 is simple withsingle, plate-like unornamented pre- andpostcristae. The precrista curves lingually at thebase of the anterior transverse valley. Thesucceeding 3rd to 8th cusps are similar wifu theirsubparallel and imbricated pre- and postcristae.The postcrista of cusp 3, however, differs in beingbifurcated at its base. Cusp 9 is slightly smaller. Itsprecrista merges into fue buccal face of the toothand fue postcrista is shorter because the apex ismore posteriorly positioned. Cusp 10 shows a veryshort precrista and postcrista, and a buccal crestthat curves backwards. None of the postcristaecrosses fue main transverse valley.

The metaloph is almost straight wifu a slightlysinuous posterior face, and 10 cusps. Themetaconule is not as large as the protocone but isrounder. The apices of fuese two cusps are level.There is a deep anterolingual groove and a finerposterolingual one which apically tends to be moretransverse. Cusps 2-9 are similar in form andbecome closer and slightly smaller buccally. Thereis a distinct flexure of fue precristae (as in fuose ofthe protoloph) near the base of the transversevalley and of the postcristae at the posterior edge.In contrast to those of the protoloph, the cristaeshow a tendency to bifurcate and very weak ribsare developed. Cusp 10 may be interpreted as afusion of two cusps because it has two divergingprecristae and two diverging postcristae with anintermediate crest. The two buccal cristae arethickened basally. A remnant postcingulum ispreserved buccally between the postcristae ofcusps 9 and 10 but interdental appression hasobliterated it lingually.

The tooth shows considerable, although notextensive, wear and most cristae have wear facets.The transverse valleys are deep and very narrowat the bottom but the lophs themselves are broader,their crests subtending an angle of about 90°, partlyas a result of the degree of wear.

259

The roots of the holotype specimen are betterpreserved than those described by Stirton et al. (1967)for fue holotype of E. serratus, but fue configurationis basically the same. There is a bladed, slightlycrescentically cross-sectioned root under theparastyloph, parallel to its buccal face. This fusesbasally with a smaller laterally-compressed rootunder fue buccal end of fue protoloph. A broad, thin,transverse root supports the bulk of fue metaloph.The best-developed root supports both protoconeand metaconule. This root is a fusion of twostructures, indicated on fue lingual face by a deepgroove and on its buccal face by a shallower groove.It leans anterolingually and tapers rapidly but onlyfue last 1-2 mm of its estimated 7-8 mm length aremissing. The tip of fue root is below fue protocone.This root configuration differs slightly from fuatinferred for Clmnia illuminata, on fue basis of maxillafragment QM F10641, where fue bone lamina buccalto fue roots has been partly broken away. There fueposterior transverse root is fuicker and fue anteriorbuccal root seems to be transverse and thicker fuanfue small parastyloph root to which it appears to befused.

RemarksThe holotype and only known specimen of

Ektopodon litolophus was found in fue same coarsebasal unit of fue Wipajiri Formation and fue samequarry (fue Leaf Locality) as the holotype and allother specimens of E. serratus. Although thisconjunction suggests that fue two species may havebeen sympatric, there is no way of being certainfuat the single toofu of E. litolophus had not beentransported from some distance away. Its veryrarity might be taken as evidence fuat it was notnormally encountered in the same environment asE. serratus. However, the significant size difference(Figure 4) would probably have been sufficient tohave enabled these species to be members of asingle ecological guild in the same way thatdifferent-sized but comparably-shaped congenericdasyurids are known to coexist (e.g. Antechinusswainsonii and A. stuartii).

A strong molar gradient is apparent in the maxillaof Clnmia illuminata (QM F10641), and a lessergradient may be seen in fue maxilla SAM P35309,and inferred from the dentaries SAM P19509 andP29577, of Ektopodon stirtoni. It is probable fuen thatE. serratus would show even less molar gradientand, in fact, the isolated upper molars referred to E.serratus do not differ greatly in size (Woodburneand Clemens 1986b). In view of this, and thedifferences in morphology (cusp number andcomplexity), it is unlikely that Leaf Localityspecimens previously referred to E. serratus includeexamples of E. litolophus posterior molars.

For the same reasons SAM P30176 is almostcertainly not an extreme variant of E. serratus. The

260

size difference alone is beyond that shown by thecomparable teeth of natural species. Nor is thetooth likely to be an abnormal tooth of E. serratus.Its regular wear facets indicate that it occludedwith a lower counterpart of similar size and shape,a situation so far unknown in marsupials exhibitingabnormalities on such a scale (Archer 1975).

In representing a fourth species of Ektopodon, thetooth shows that some phylogenetic trends can bediscerned within both the family and the genus,viz. a general increase in size, increase in numberof cusps and decrease in complexity of structure.Also, it shows that the genus was still diversifyingduring the Miocene.

Ektopodon sp. cf. E. serratusFigure 5

Material ExaminedQM F36365, a fragment of a right upper molar,

comprising a quarter of the protoloph and most ofthe metaloph of M2 or M3. Collected by M. Archer,H. Godthelp, S.]. Hand and others in May 1986.

Locality and AgeWayne's Wok Site is in an unnamed freshwater

limestone, "System B Sediments" (Archer et al.1989, 1991), Hal's Hill, Riversleigh Station,northwestern Queensland. The Wayne's Wok LocalFauna, a member of Riversleigh System B, ispresently interpreted to be between Late Oligoceneand Early Miocene in age on the basis ofbiocorrelation (Archer et al. 1989, 1991), and sharesmany elements with the Upper Site Local Fauna(Godthelp Hill, Riversleigh), which in turn has taxathat are congeneric and (in some cases) conspecificwith forms in the Kutjumarpu and TarkaroolooLocal Faunas of northern South Australia (seeIntroduction).

N.S. Pledge, M. Archer, S.J. Hand, H. Godthelp

DescriptionAlthough not greatly worn, this specimen has

been badly corroded and broken and lacks thebuccal edge of the crown. The transverse ratherthan oblique nature of the lophs indicates that it isan upper molar. A full cusp count cannot be made.Most of the protoloph, including the protocone, ismissing but it would have borne at least six cusps.The metaloph bears six cusps with evidence of aseventh on the buccal end. The metaconule isrounded with an anterolingual groove and abifurcated premetaconulecrista. The precrista ofcusp 2 also bifurcates lingually. Cusps 3 and 4 aresimilar with single cristae. Their precristae eachhave three small lingual ribs towards the anteriorend. The cristae of cusps 5 and 6 are single andsimple with no obvious ribs or divisions. Cusps arelinked apically with single deep-set fine struts. Thebuccal face is damaged or missing.

RemarksQM F36365 is the first ektopodontid described

from the northern half of Australia and the onlyektopodontid specimen yet recovered from theotherwise richly fossiliferous Riversleigh deposits.Its referral to E. serratus is on the basis of similarcrestal spacing and development. The relativerarity of the species suggests that ektopodontidswere uncommon in the Riversleigh rainforests inOligo-Miocene times. The damaged nature of thetooth might be seen as support for the hypothesisthat this specimen was transported to theRiversleigh area, perhaps by a bird of prey.However, to date several other Riversleigh animals(e.g. yalkaparidontids, Archer et al. 1988) were firstdiscovered as small fragments and only much laterappeared in abundance when other sitesrepresenting slightly different habitats ordepositional environments were explored. Thus, it

10mmI I I I I I I 11 I I

Figure 4 Comparison drawings of known ektopodontid M1S in occlusal view. A, Chunia illuminata; B, Ektopodonstirtoni; C, E. serratus; D, E. litolophus. Scale =10 mm.

A new species of ektopodontid

Figure 5 Ektopodon sp. cf. E. serratus, upper molarfragment QM F36365, Riversleigh, Wayne'sWok Local Fauna. Scale bar = 2 mm.

is probably too early to conclude anything aboutthe significance of the apparent rarity ofektopodontids in the Riversleigh deposits. Further,investigation of the Wayne's Wok Site material isin its early stages with many kilograms oflimestone blocks still remaining to be prepared.

The relative simplicity and regular spacing ofcusps on this tooth fragment most closely matchthe condition seen in E. serratus, but too little of thetooth is preserved to exclude the possibility that itrepresents a distinct taxon.

In view of many other overlaps in taxa betweenthe Kutjamarpu Local Fauna and variousRiversleigh local faunas (e.g. Neohelos tirarensis,Wakiewakie lawsoni, Wakaleo sp. cf. W. oldfieldi,Emuarius gidju, and the lungfish Mioceratodusanemosyrus) (Archer et al. 1989, Kemp 1997), it isnot improbable that this Riversleigh ektopodontidwill be confirmed as representing anotherKu~amarpu taxon.

DISCUSSION

The family Ektopodontidae is now known tocontain three to five species of Ektopodon (E. serratus,

261

E. stirtoni, E. litolop/IllS, E. sp. cf. E. stirtoni, and tl1eRiversleigh taxon), two or three species of Chunia(c. illuminata, C. sp. cf. C. illwninata, C. omega), andOarcius duggani. No ektopodontid species occurs inmore than a single local fauna and in all but twolocal faunas (Tarkarooloo and Ku~amarpu) there isonly a single ektopodontid species. Further, mostare not common in their respective local faunas andthree (E. litolophus, C. omega and the Riversleightaxon) are known only from single teeth. In thefaunas where more than one species occurs, thesympatric forms are distinct in terms of size andmorphology, a situation that may indicateecological partitioning of species into feedingguilds. Food preferences for these possums areunclear but possibilities include seeds or grains,nuts and insects (Pledge 1982, 1986, 1991).

Species of Ektopodon range in age from LateOligocene to ?Middle Miocene. The oldest (E. sp.d. E. stirtoni) occurs in the Tarkarooloo LocalFauna of probably latest Oligocene age and the twoyoungest (E. serratus and E. litolophus) in theKutjamarpu Local Fauna of possible Early toMiddle Miocene age. The age of the Riversleighektopodontid has been interpreted by Archer et al.(1989) to be most probably Early Miocene. Speciesof Chunia range in age from Late Oligocene(Ditjimanka Local Fauna) to latest Oligocene(Tarkarooloo Local Fauna) in age. Oarcius dugganiis only known from the Early Pliocene HamiltonLocal Fauna, although an undescribedektopodontid ascribed tentatively to Oarcius (T.H.Rich, pers. comm.) is reported in passing by Archeret al. (1997) from the Early Pleistocene Nelson BayLocal Fauna.

The Ektopodon serratus-like ektopodontid from theWayne's Wok Local Fauna supports previoussuggestions that Riversleigh's System B local faunasshare taxa with the Kutjamarpu Local Fauna(Archer et al. 1989).

Table 1 Measurements (in millimetres) of ektopodontid upper molar teeth (M').

Specimen number Tooth Length Anterior Posteriorwidth width

Ektopodon litolophus sp. novoSAM P30176 RM' 10.6 11.7 12.1Ektopodon serratusSAM P13847 LM' 7.1 7.8 8.4Ektopodon stirtoniSAM P22504 RM' 8.7 8.9 9.2SAM P35309 LM' 8.8 8.3 9.0Ektopodon sp. cf. E. stirtoniNMV P4875Q-l LM' 8.5 80SAM P19962 RM' 8.5Cllllnia illuminataSAM P29081 LM' 7.3 6.3 7.0

262

....----------- ChuniaiIIuminata

r-------- Ektopodonstirtoni

....--__ Ektopodonserratus

L----16

L..-__ Ektopodonlitolophus

DarciusL...- duggani

Figure 6 PAUP-derived cladistic tree forektopodontids, based on morphology of Mls.At node 6, synapomorphic characters are: 1,3,15; at node 7: 4,7,14; and at node 8: 2, 6, 9,11.

Woodburne and Clemens (1986c) interpreted thephylogenetic relationships of ektopodontids thenknown. Analysis of the four known ektopodontidspecies M1s in terms of cusp and crest numbers

N.S. Pledge, M. Archer, S.J. Hand, H. Godthelp

(and extrapolating for Darcius from its lower molarsby comparison with E. stirtoni) (Table 2), and thenature of the crests and overall proportions of theseteeth (Table 3), using PAUP 3.1 (Swofford 1993),gave a single simple tree (Figure 6). Because of theunique morphology of the M l

, no meaningfuloutgroup could be chosen, so the tree is unrooted.

We suggest here that the new species Ektopodonlitolophus is best regarded as a sister taxon of E.serratus and that both may have been derived froman E. stirtoni-like ancestor. This conclusion followsfrom the observation that E. litolophus and E.serratus share the evidently synapomorphiccondition of transversely widened lophs which alsocontain a relatively larger number of cusps.Considering that both of the more derived speciesare contemporaneous, neither is likely to be theother's ancestor. Because of the poor preservationof the Riversleigh tooth fragment, it is not possibleto be confident about its affinities, although itappears to be more similar to E. serratus in style ofcrista development and spacing.

ACKNOWLEDGEMENTS

We thank the owners of Etadunna Station andthe Riversleigh World Heritage property forpermission to work on their land and for theirforbearance of our activities. Numerous studentsand volunteers helped in the field, both in SouthAustralia and at Riversleigh.

The Riversleigh Project has had vital supportfrom: the Australian Research Grants Scheme (toM. Archer); The University of New South Wales;the National Estate Grants Scheme (Queensland)(toM. Archer and A. Bartholomai); the RiversleighSociety Inc.; the Queensland National Parks andWildlife Service; Century Zinc Pty Ltd; PasmincoPty Ltd; ICI Australia Pty Ltd; the AustralianGeographic Society; the Queensland Museum; the

Table 2 Comparison of cusp and crest numbers on Ml of different ektopodontids. Abbreviations: a = approximatecount because of complexity of tooth morphology; e = estimated number by comparison with lower molar.

Ml Chunia Ektopodon Ektopodon Ektopodon Darciusilluminata stirtoni serratus litolophus duggani

Parastylophprecristae 1 3 4 4 ?cusps 3 3 4 4 ?postcristae 6 6 5 6 ?

Protolophprecristae 11 11 13 11 e5cusps 6 8 8 10 e4postcristae 13a 14 16 13 e8

Hypolophprecristae 14 15 15 13 e7cusps 7 8 8 10 e5postcristae 12 13 14 13 e7

A new species of ektopodontid 263

Table 3 Matrix of character states in Ml of different ektopodontids. Plesiomorphic character states indicated by '0'. 1,three or fewer parastyloph cusps; 2, four or less protoloph cusps; 3, six or less protoloph cusps; 4, seven orless protoloph cusps; 5, eight or less protoloph cusps; 6, five or less hypoloph cusps; 7, seven or lesshypoloph cusps; 8, eight or less hypoloph cusps; 9, generally more than two precristae per cusp; 10,generally two or more precristae per cusp; 11, crests radiating from cusps; 12, crests bifurcating, parallel; 13,generally more than one pre- or postcrista; 14, length equals or exceeds overall width; 15, l:w ratio morethan 0.9; 16, l:w ratio more than 0.85.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

Darcius ? 0 0 0 0 0 0 0 1 0 1 0 0 ? 0 0Chunia 0 1 0 0 0 1 0 0 0 0 0 0 0 0 0 0E. stirtoni 0 1 1 1 0 1 1 0 1 0 1 0 0 1 0 0E. serratus 1 1 1 1 0 1 1 0 1 0 1 0 0 1 1 0E. litoloplllls 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

Australian Museum; the Royal Zoological Societyof New South Wales; the Linnean Society of NewSouth Wales; private supporters; and colleaguesand co-workers who have helped to collect, processand interpret Riversleigh's resources.

Ben McHenry drew Figures 1 and 2.

REFERENCES

Aplin, K.P. and Archer, M. (1987). Recent advances inmarsupial systematics with a new syncreticclassification. In Archer, M. (ed.), Possums andopossums: studies in evolution: xv-lxxii, Surrey Beatty& Sons, Chipping Norton, N.s.W.

Archer, M. (1975). Abnormal dental development andits significance in dasyurids and other marsupials.Memoirs of the Queensland Museum 17: 251-265.

Archer, M., Black, K. and Nettle, K. (1997). Giant ringtailpossums (Marsupialia, Pseudocheiridae) and giantkoalas (Phascolarctidae) from the Late Cainozoic ofAustralia. Proceedings of the Linnean Society of NewSouth Wales 117: 3-15.

Archer, M., Godthelp, H., Hand, S.J. and Megirian, D.(1989). Fossil mammals of Riversleigh, northwesternQueensland: preliminary overview ofbiostratigraphy, correlation and environmentalchange. The Australian Zoologist 25: 29-65.

Archer, M., Hand, S.J. and Godthelp, H. (1988). A neworder of Tertiary zalambdodont marsupials. Science239: 1528-1531.

Archer, M., Hand, S.J. and Godthelp, H. (1991).Riversleigh. The story of animals in ancient rainforests ofinland Australia: 1-264, Reed Books, Sydney, NSW.

Kemp, A. (1997). A revision of Australian Mesozoic andCenozoic lungfish of the Family Neoceratodontidae(Osteichthyes: Dipnoi), with a description of fournew species. Journal of Paleontology 71: 713-733.

Lindsay, J.M. (1987). Age and habitat of a monospecificforaminiferal fauna from near-type EtadunnaFormation, Lake Palankarinna, Lake Eyre Basin.South Australian Department of Mines Report Book 87/93: 1-10. (Unpublished).

Luckett, W.P. (1993). An ontogenetic assessment of dentalhomologies in therian mammals. In Szalay, F.5.,Novacek, M.J. and McKenna, M.C. (eds), Mammal

phylogeny: Mes020ic differentiation, multituberculates,monotremes, early therians, and marsupials: 1: 182-204,Springer-Verlag, New York, N.Y., U.s.A.

Norrish, K. and Pickering, J.G. (1983). Clay minerals. InSoils: an Australian viewpoint: 281-308, CSIRO/Academic, Melbourne, Victoria.

Pledge, N.s. (1982). Enigmatic Ektopodon: a case historyof palaeontological interpretation. In Rich, P.V. andThompson, E.M. (eds), The fossil vertebrate record ofAustralasia: 479-488, Monash University OffsetPrinting Unit, Clayton, Victoria.

Pledge, N.S. (1986). A new species of Ektopodon(Marsupialia; Phalangeroidea) from the Miocene ofSouth Australia. University of California Publications inGeological Sciences 131: 43-67.

Pledge, N.5. (1991). Reconstructing the natural historyof extinct animals: Ektopodon as a case history. InVickers-Rich, P., Monaghan, J.M., Baird, R.F. andRich, T.H. (eds), Vertebrate palaeontology ofAustralasia: 247-266, Pioneer Design Studio,Lilydale, Victoria.

Rich, T.H.V. (1986). Darcius duggani, a newektopodontid (Marsupialia; Phalangeroidea) fromthe early Pliocene Hamilton Local Fauna, Australia.University of California Publications in GeologicalSciences 131: 68-74.

Sige, B., Hand, S.J. and Archer, M. (1982). An AustralianMiocene Brachipposideros (Mammalia, Chiroptera)related to Miocene representatives from France.Palaeovertebrata 12: 149-172.

Stirton, R.A., Tedford, R.H. and Miller, A.H. (1961).Cenozoic stratigraphy and vertebrate paleontology ofthe Tirari Desert, South Australia. Records of the SouthAustralian Museum 14: 19-61.

Stirton, RA., Tedford, RH. and Woodburne, M.O. (1967).A new Tertiary formation and fauna from the TirariDesert, South Australia. Records of the South AustralianMuseum 15: 427-462.

Stirton, RA., Tedford, RH. and Woodburne, M.O. (1968).Australian Tertiary deposits containing terrestrialmammals. University of California Publications inGeological Sciences 77: 1-30.

Swofford, D.L. (1993). PAUP. Phylogenetic Analysis usingParsimony. Version 3.1. Laboratory of MolecularSystematics, Smithsonian Institution, Washington,D.e., U.SA.

264

Tedford, R.H. and Woodburne, M.O. (1987). TheIlariidae, a new family of vombatiform marsupialsfrom Miocene strata of South Australia and anevaluation of the homology of molar cusps in theDiprotodontia. In Archer, M. (ed.), Possums andopossums: studies in evolution: 401-418, Surrey Beatty& Sons, Chipping Norton, NSW.

Truswell, E.M., Sluiter, I.R. and Harris, W.K. (1985).Palynology of the Oligocene - Miocene sequence inthe Oakvale-1 corehole, westem Murray Basin, SouthAustralia. Bureau of Mineral Resources Journal ofAustralian Geology & Geophysics 9: 267-95.

Woodburne, M.O. and Clemens, W.A. eds (1986a).Revision of the Ektopodontidae (Mammalia;Marsupialia; Phalangeroidea) of the AustralianNeogene. University of California Publications inGeological Sciences 131: 1-114.

Woodbume, M.O. and Clemens, W.A. (1986b). A newgenus of Ektopodontidae and additional comments

N.S. Pledge, M. Archer, S.}. Hand, H. Godthelp

on Ektopodon serratus. University of CaliforniaPublications in Geological Sciences 131: 10-42.

Woodbume, M.O. and Clemens, W.A. (1986c). Phyleticanalysis and conclusions. University of CaliforniaPublications in Geological Sciences 131: 94-102.

Woodbume, M.O., MacFadden, B.j., Case, j.A., Springer,MS, Pledge, N.5., Power, j.D., Woodbume, j.M., andSpringer, K.B. (1994). Land mammal biostratigraphyand magnetostratigraphy of the Etadunna Formation(Late Oligocene) of South Australia. Journal ofVertebrate Paleontology 13: 483-515.

Woodbume, M.O., Tedford, R.H., Archer, M., Tumbull,W.D., Plane, M.D. and Lundelius, E.L. jr (1985).Biochronology of the continental mammal record ofAustralia and New Guinea. Special Publication, SouthAustralian Department ofMines and Energy 5: 347-363.

Manuscript received 12 March 1998; accepted 20 May 1998.