44 
Records of the Australian Museum (2009) Vol. 61 
buccal ridge small, descends posteriorly from stylar cusp D, 
terminating at posterobuccal corner of tooth; molar enamel 
heavily crenulated. 
Remarks. QMF52288 is regarded as an M 2 due to its 
trapezoidal occlusal outline that tapers posteriorly and by 
its possession of a relatively small neometaconule, identical 
to that of M 2 in other species of Phascolarctos. QMF52288 
is significantly larger than corresponding teeth of Koobor, 
Madakoala, Perikoala, Nimiokoala and Litokoala, and Ph. 
cinereus. However, the tooth is morphologically similar to 
other species of Phascolarctos. In comparison to Ph. stirtoni, 
the posterior cingula is relatively smaller, molar enamel is 
less crenulated (although that feature may be slightly variable 
judging from variation expressed in large samples of modern 
Ph. cinereus), and the tooth is smaller overall (Fig. 3B). 
However, judging by the range of morphometrical variation 
exhibited in modern Ph. cinereus (Fig. 3B), QMF52288 
could easily fall within the lower size range of Ph. stirtoni. 
QMF52288 is somewhat similar in morphology to extant 
Ph. cinereus. However, it is larger and falls outside the 
morphometrical range of variation of modern populations 
(Fig. 3B). Pledge (1987) suggested that Ph. maris (a species 
questionable distinct from Ph. stirtoni ; see Black, 1999) is 
intermediate in size between the smaller Ph. cinereus and 
larger Ph. stirtoni. Thus, QMF52288 potentially represents 
Ph. maris. However, it is not possible to compare QMF52288 
to Ph. maris (nor to the significantly larger Cundokoala 
(?Ph.) yorkensis) as corresponding teeth are unknown in 
those species. 
Phascolarctos stirtoni Bartholomai, 1968 
Holotype. QMF5707, right maxillary fragment with P 3 , 
M 1-2 , Cement Mills, Gore, southeastern Queensland (Late 
Pleistocene; Price et al., 2009). 
Diagnosis. See Bartholomai (1968). 
Phascolarctos ?stirtoni 
Figs 4B, C 
Referred material. QMF52289, isolated RM 12or3 fragment, 
QML7, Chinchilla, southeast Queensland, Australia (Middle 
Pliocene); QMF52290 isolated RM 2 fragment, QML1420 
Marmor Quarry, central eastern Queensland, Australia 
(Middle Pleistocene; Table 1). 
Description. RM 12 013 , description based on QMF52289: 
Metacone only major cusp preserved, very large, worn; 
postmetacrista moderately developed, descends apex of 
metacone posterobuccally to small stylar cusp E; buccal crest 
small but distinct, directed anteriorly from stylar cusp E, 
slightly ascending buccal margin of tooth; posterior cingulum 
large, worn; molar enamel crenulated at posterolingual base 
of metacone. 
RM 2 , description based on QMF52290: Protocone 
only major cusp preserved; preprotocrista well developed, 
descends anterobuccally to lingual portion of anterior 
cingulum; postprotocrista descends posterobuccally to base 
of protocone; rib descends protocone between pre- and 
postprotocristae, well developed; anterior fossette well 
developed at anterior base of protocone; lingual cingulum 
well developed, extending along lingual margin at base of 
protocone, ascending and terminating at lingual corner of 
tooth; molar enamel crenulated on all sides of protocone. 
Remarks. The wear pattern on the posterior margin of the 
metacone molar fragment (QMF52289) is consistent with 
wear from abrasion with a succeeding tooth whilst still in 
the maxilla. This suggests that the specimen represents either 
an M 1,2 or M 3 , rather than an M 4 (the most posterior tooth in 
phascolarctids). On the basis of molar morphology, the teeth 
are referable to Phascolarctos due to: (a) their well-developed 
molar crenulations; ( b) minor degree of development of stylar 
cusps and associated stylar shelf (QMF52289); (c) well- 
developed lingual cingulum (QMF52290); and (d) being 
higher-crowned than all other phascolarctids (excepting 
Cundokoala (?Ph.) yorkensis). Morphometrically, both teeth 
are larger than corresponding teeth of Ph. cinereus, but are 
smaller than corresponding teeth of C. ( ?Ph.) yorkensis. 
The protocone fragment (QMF52290) is morphologically 
similar to the corresponding M 2 of Ph. stirtoni, particularly 
in the development of the anterior fossette at the base of the 
protocone, and the anterolingual extension of the lingual 
cingulum. The metacone fragment (QMF52289) lacks a well- 
developed stylar shelf as exhibited in C. ( ?Ph.) yorkensis, 
and in that respect, closely resembles the condition exhibited 
in Ph. stirtoni. Corresponding teeth of Ph. maris and C. ( ?Ph.) 
yorkensis are either not known or are poorly represented, thus, 
preventing further comparison to the material described here. 
However, the molar fragments described here are similar 
in size to Ph. stirtoni, a species that is intermediate in size 
between Ph. maris and C. ( ?Ph .) yorkensis (Pledge, 1987, 
1992). Thus, QMF52289 and QMF52290 are unlikely to 
be referable to those poorly known taxa. The fragmentary 
nature of the material precludes additional comparison to 
Ph. stirtoni. 
Discussion 
New koala material described here significantly extends the 
temporal and spatial distribution of Plio-Pleistocene koalas 
in Australia. Previously, Pliocene records of koalas were 
restricted to southern Australian deposits (i.e., Corra Lynn 
Cave and Waikerie). Thus, the new records from the Chinchilla 
Local Fauna (phascolarctid indet. and Phascolarctos ?stirtoni) 
represent a significant north-eastern geographic extension for 
Pliocene koalas, and provide the first unambiguous evidence 
for the presence of koalas in the assemblage. Similarly, new 
records of Phascolarctos from the Marmor and Mt. Etna 
deposits extend the northern geographic distribution of Pleisto¬ 
cene koalas from central eastern Australia (e.g., Price, 2008a) 
to northeastern central Queensland. Thus, the new koala 
records are important because they show that Phascolarctos 
has been closely associated with eastern Australian habitats 
since the Middle Pliocene. 
Palaeoecology. The modem Koala, Phascolarctos cinereus, 
is mostly restricted to open forests and woodlands of eastern 
Australia (Moore & Foley, 2000), but has been recorded, 
albeit as a rare component, in rainforest communities 
(Williams et al., 1996). Koalas are folivores and feed almost 
exclusively on the leaves of certain species of Eucalyptus 
(Moore & Foley, 2000). Thus, because fossil species of 
Phascolarctos such as Ph. stirtoni, have morphologically 
