Hindlimb proportions and locomotion of Emuarius gidju (Patterson & Rich, 1987) (Aves, Casuariidae)

6 Pages·1997·1.6 MB·

by  Walter E Boles

HINDLIMB PROPORTIONS AND LOCOMOTION OF EMUARIUSGIDJU
(PATTERSON & RICH, 1987) (AVES: CASUARIIDAE)
WALTER E. BOLES
Boles,W.E. 19970630: HindlimbproportionsandlocomotionofEmitariusgidjii(Patterson
& Rich, 1987) (Aves: Casuariidae). Memoirs ofthe Queensland Museum 41(2): 235-240.
Brisbane. ISSN 0079-8835.
Using proximal anddistal fragments,the lengthofthetarsometatarsusofEmuariusgidju is
estimatedandcomparedlothatofotherhindlimbelements.Fromtheseproportionsandother
hindlimb morphology, the inferred locomoiory mode off", gidju is compared with Recent
casuariids. Emuariusgidju appears to have been more cursorially adapted than Casuarius
anddwarfZ)romrt/Hi',suggestingatleastsomeopenhabitatintheRiversleighpalaeoenviron-
ment. Using the relationship between weight and least circumference of the femur and
tibiotarsus in Recent birds, the weight of E. gidju is suggested to have been 19-21kg.
Emuanus.Aves. hindlimb, locomotion.
Walter E. Boles, Australian Museum, 6 College Street, Sydney NSW 2000, Australia;
received4November1996.
Emus, Dromaius (Dromaiinae), form a promi- MATERIALS AND METHODS
nentelementofAustralia'savifauna.Theclosely
relatedcassowaries,Ccisiiarius(Casuariinae),are Osteological terminology follows Baumel &
more restricted in distribution. These twogroups Wilmer (1993). Measurements were made with
occupy verydifferent modern habitats, and loco- verniercalipers accurate lo0.05mmand rounded
motory adaptations correlated with these differ- lo the nearest 0.1mm. Institutional acronyms are
ent habitats are obvious in the relative pro- AM(AustralianMuseum),QM(QueenslandMu-
portions ofthe lower limb bones. Because ofthe seum) and SAM (South Australian Museum).
relationship between habitat and liinb propor- The type specimen of£. gidju (SAMP26779)
tions, fossil emusandcassowariesarepotentially is an associated distal tibiolarsal fragment, prox-
goodpalaeoenvironmentalindicators.Emushave imal tarsometatarsus including much most ofthe
a belter fossil record (Patterson & Rich, 1987). shaft, and complete set of pedal phalanges
than cassowaries (Vickers-Rich, 1991). Patter- (Patterson & Rich, 1987) and numerous speci-
son&Rich(1987)describedlowerlimbelements mens of Emuarius are known from Riversleigh
from the Mioceneofcentral Australianasasmall (Table 1).
emu, Dromaius gidju. Boles (1991) erected To compare changes in relative proportions of
Emuariusforthis speciesandconsidereditcloser the casuariid hindlimb, the following measures
toemusthancassowaries,but nearertheirdichot- were calculated for the bone lengths of the 3
omy than any otherdescribed laxon. extant species ofCasuarius and the 1 living and
The type material ofEmuarius gidju is part of 2 recently extinct dwarfspecies ofDrotnaius:
the Kutjamarpu Local Fauna, recovered fromthe TT^BoX-r4^?MR - -Tibiotarsus X 100
LeafLocality (UCMP V-6213) on the E shore of Femur
Lake Ngapakaldi in the E Lake Eyre Sub-Basin, TMXtMR = Tarsometatarsus x 100
South Australia. Much E. gidju material occurs Femur
NW
at Riversleigh Queensland. Archer el al. Tibiotarsus x 100
TBT/TMT
(1989, 1994) considered the Riversleigh pal- Tarsometatarsus
aeohabiial as rainforest, based on mammal re- Becausethepurposewasto findgeneral,rather
mains. than detailed, directions of change, measure-
This paper is to 1) estimate the lower limb mentsweretaken fromtheliterature(Table2)and
proportionsofE. gidju\2)comparethesetomod- rounded to the nearest mm. Means were used
ern emus andcassowaries and, by implication, to where available. The sample sizes were often
theirstyle oflocomotion; 3) make an initial esti- small, sometimescomprising single individuals.
mateofthe weightof£". gidju\and4)interpretthe Predicted body weight ofEmuarius gidju was
possible palaeohabitat at Riversleigh where E. calculated(Campbell &Marcus, 1991) fromAM
QMF
gidju occurs. F78585 (near complete femur) and 16827
236 MEMOIRSOFTHEQUEENSLAND MUSEUM
(distal libiolarsal fragment). Tape was wrapped TABLE I. Specimens that have been referred to
around the bones at their least circumferences, Emuarius gidju. *=speciniens described by Boles
marked at the point ofoverlap, straightened and (1991).
measured with calipers. These results were used SITE REG. NO. ELEMENT
in theequation logio(weighl)=a*logio(circumfer- LakeNgapakaldi Holoiype:associated
ence)+b. where the values of a (slope) and b SAM distaltibiolarsus,
(inierecept) were those determined by Campbell P26779 proximal
& Marcus (1991) for all birds for the respective tarsometarsus.
completepes
elements(femur: a=0.41 l,b=-0.065; libiotarsus:
AMF78585* Femur
a=2.424,b=0.076).
Riversleigh 1
RESULTS System A orB QMF16827'*= Tibiotarsus
WhileHunter
S\stem B QMF29720 Vertebrae
CamelSpuiuni
Increased cursoriality in these birds is associ- QMF29721 Vertebrae
ated with an increase in the lengths of the tibio-
QMFI6828* Femur
larsus and tarsometalarsus relative to that ofthe
QMFI6829* Femur
femur (Howell, 1944). Relative proportions of
the hindlimb contributions of the long bones in AMF78586* Tibiotarsus
Casuariits, Dromaius and Emuahus (Fig, 2) QMF29722 Tibiotarsus
show that thai ofthe libiotarsus remains more or QMF29723 Tcu-sometatarsus
less consistent in all taxa; that of the femur de- QMF29724 Tarsometalarsus
creases with increased cursoriality, whereas that OMF29725 Tarsometalarsus
ofthetarsometalarsusincreases.Therelationship
QMF29726 Tarsometalarsus
between these changes is consistent for Recent
species: TBT/FMR = 0.45 [TMT/FMR] +113; QMF29727 Tarsometalarsus
r=0.84. Upper QMF16830* Rostrum
These changes appear independent of overall QMFt683l* Scapulocoracoid
size when compared between genera, but within QMF29728 Vertebrae
TgeBneTr/aTtMheTsmaanldlersmmaelmlbeerrsTMhaTv/eFtMhRe.greTatheer TSo>wsteerms7B, Dirks QMF29729 Tiirsomeiaiarsus
TBT/FMRin Casuariusremainsconstant. Itsug- SystemC QMF16832* Femur
Gag
gests from these figures that the smaller species AMF78587* Tarsometalarsus
in each genus are the leasi cursorial members. Using atarsometatarsal lengthof340mmanda
The Kutjamarpu femur (AMF78585; Boles, femoral length of 198mm, gives aTMT/FMR o{
1991)oi'E. gidju is 194mm long, which, because 172, greater than that of any Recent casuariid
mm
of abrasion, is a few less than its original except Dromaius novaehollandiae. Using these
length, of about 198mm and very likely not valueswiththeregressionequation forthefamily
200mm. Complete libiolarsi are unknown for E. (Fig. 4) gives a predicted TBT/FMR of 190,
gidju (Table 1). Nevertheless, it is possible to which corresponds to a tibiotarsal length of
predict the size and relative proportions of this 376mm. These 3 values give a combined length
bone from other hind limb elements. Tar-
of916mm, virtually the same as the hindlimb of
sometatarsi are represented by the holotypical
Casuarius casuarius, although the proportional
proximal end and shaft, and several distal frag-
ments. An estimate ofthe tarsometatarsal length contribution ofeach bone to this length is differ-
wasmadebyusingtheproximalendandshaftand ent (Fig. 2).
a distal fragment. The proximal tarsometatarsal This figure must be used with caution. The
fragment is 276mm long; the longest edge of its fragmentson which it is basedrepresentdifferent
shaft is straight and shows noevidenceofflaring individuals from different localities. As such,
outward to trochlea metatarsi II. The distal piece thereareseveral sourcesofpotentiallysignificant
is 64mm long; the small portion ofshaft remain- variation between components. Geographical
ing is just proximal to the flaring of trochlea variationmaynothavebeenofmajorimportance;
metatarsi11.Becausethereislittle,ifany,overlap livingD. novaehollandiaeexhibits littledifferen-
betweenthesetwopieces, minimumlengthofthe tiation across its range. There is much greater
tarsometalarsus is 340mm (Fig. 3). size variation across this species' chronological
HINDLIMB PROPORTIONS OFEMUARIUSGIDJU 131
thananyknown speciesofemu
(illustrated in Patterson &
Rich, 1987). In emus, phalanx
ungualis of digit III is longer
than thai ofdigit II, whereas in
cassowaries it is reversed, wilh
phalanx ungualis ofdigit II ex-
tended intoalongspikeseveral
times the length ofphalanx un-
gualis o( digit III. Emuarius
doesnot havephalanx ungualis
of digit II developed into a
spike, but il is still longer than
thai ofdigit III.
Cassowaries have digits II
and IV relatively long com-
pared wilh digit III (ratios oi
11:111 and IV:III, respectively,
without phalanges unguales
0.82, 0.76). Both are substan-
FIG. 1. Leaf Locality femur oi' Emuarius gidju compared wilh femora ol tially reduced in relative length
Recent casuariids, in cranial view. From left, Casuarius Cositariits, C. in emus (ratios as above -0.55,
bennettii, Emuariusgidju, Dromaiusnovaehollandiae andD. aler. 0.63). Both digits II and IV of
Emuarius are comparatively
range, with mainland fossils usually smallerthan longer than those in emus (digit IV lo a lesser
modern birds. Patterson & Rich(1987)suggested degree), butnot lo the extent seen in cassowaries
that it 'may have been at any one lime in the (0.57, 0.68). Patterson & Rich (1987) suggested
Pleistocene both larger or smaller than at that the reduction ofdigit IV and particularly of
present'. Considerable intraspecifie variation in digit II in E. gidju, compared wilh the highly
living D. novaehollandiae is probably related to cursorial D. novaehollandiae, appears lo parallel
age and sex. as well as individual differences similar reductions in other groups of lerresirial
(Marchant & Higgins. 1990). For example, birds and mammals.
among 22 modern specimens, Patterson & Rich
In comparison wilhthe pedal phalangesofcas-
(1987) found arange in tarsometatarsal length of sowaries, those ofemus are dorsoplaniarly com-
332-422mm (mean 383mm: s.d. 18.0). pressed. Emuarius is somewhat intermediate,
Miller( 1963)describedDromaiusocypus from with its phalanges substantially more
the middle lo late Pliocene Palankarinna Local dorsoplaniarly compressed than those of casso-
Fauna from Lake Palankarinna, northern South waries, but less compressed than (but more sim-
Australia. This species was intermediate in si/.e ilar lo) the condition in emus.
between Enuiarius gidju and living Dromaius Campbell & Marcus (1991) staled 1he least
novaehollandiae. The tarsometatarsus of D. oc-
shaft circumference ofeither [femur or libiotar-
ypus ismarkedly shorterrelative lo its width than sus] can be a reliable indicatorofthe weight ofa
is that ofD. novaehollandiae, but less so than in fossil bird\ From their equation, and measure-
Casuarius. Vickers-Rich (1991) inlerpreled this mentsoff. gidjubones, predictedweightsofthis
as suggesting a less cursorial lifestyle for /). species were 21kg based on the femur and 19kg
oPactytpeurssotnh&anRifcohr D.98n7ovdaiedhnooltlagnidvieaceo.mpBaercaatuisvee basedon the tibiotarsus. The leastcircumference
( 1 ) oi'the tibiotarsus is ahiiosl always at or distal lo
figures ft>r relative widths, it is difficult to quan- the midpointofthe bone; in mostbirds il is in the
tViitsautailvleyly, EcmoumaprairuesEa.ppgeiadrjsu twoitbhe ptrhoepsoertsipoencailelsy. dtiasrtsaall tshpiercdi(mCeanmpobieEl.lg&idjMuarisccuosmp1l9e9t1e),. Nalothliobuigoh-
thinner lor its length than is D. ocypus, but not lo the length ofthat measured is about athird ofthe
the degree ofD. novaehollandiae.
predicted length for this bone. Il is possible that
Patterson & Rich (1987) pointed out that, al- the least circumference occurs on the missing
though, the foot of E. gidju is more like thai oi' section of libotarsus, and the value given here
emus than cassowaries,it ismorecassowary-like would have to be adjusted. The predicted
238 MEMOIRS OFTHEQUEENSLAND MUSEUM
TABLE 2. Measurements (mm) and measures of relative proportions of watercourses.Itwasnotuntilat
hindlimb bones ofRecent emus and cassowaries and ofEmuariusgidju. leastthemiddleMiocene(c. 15
Forcalculationofpredicted measurements ofE. gidjuand ofmeasures of mya)thatdryingoftheclimate
relativeproportions, see text. beganandopenhabitatsstarted
TBT/TMT TBT/ to appear on a large scale.
Species FMR TBT TMT Source FMR /FMR TMTII There is no evidence ofgrass-
lands before the endofthe late
Dromaius Patterson& Miocene to Pliocene (Martin,
203 401 383 198 189 105
novaehoUandiae Rich. 1987 this volume).
Dromaius 164 293 234 Morgan& 179 143 125 The graviportal locomotion
baudinianus Sutton. 1928 of Casuarius is associated
Dromaiusater 178 331 274 Morgan& 186 154 121 with movement through the
Sutton. 1928 dense vegetation of the Aus-
C.asuanus 236 388 306 Richetal.. 1988 164 130 127 tralo-Papuan rainforests. In-
unappendiculatus creased cursoriality seems
Casuartus 232 384 300 Richetal., 1988 166 129 128 correlated with the appearance
casuarius
ofmore open habitat, in which
Casuariusbennetd 221 365 229 Richetal., 1988 165 104 159 sustained running could take
Emuariusgidju 198 376 340 Thispaper 190 172 111 place. (While able to run if
required, cassowaries have
weights from the two bones are close enough to limited opportunities in such habitat to work up
give an acceptable first estimate forE. gidju. and sustain areasonabledegreeofspeedbecause
sufficiently open areas are restricted.) Morpho-
DISCUSSION
logical correlates with cursoriality include pro-
portional lengthening of the tibiotarsus and
Throughout the early Tertiary, much of Aus- tarsometatarsus, and a reduction in the relative
tralia was covered in closed forest, and the cli- size ofdigit II. Conditions between the extremes
mate was considerably more humid than at of the states found in Casuarius and Dromaius
present (Frakes el al., 1987). The dominant veg- are suggestive oflevels ofcursorial ability inter-
etation type over much of the continent was mediate between theirs, although at what point
rainforest. Nothofagus-dom'iwdtQd rainforest
along the scale cannot be determined. This, in
covered the Lachlan River valley during the late
EocenetolateOligocene-earlyMiocene(Martin, turn, suggests the possibility that the amount of
1987). Even where closed forests were not pres- open habitat might be also somewhere between
ent, gallery rainforest probably occurred along that available to cassowaries and emus.
TAR.'^OMETATARSUS TIBIOTARSUS FEMUR
^^^^^M
Casuarius
bennetti III
Casuarius
casuarius
Emuarius
gidju
Dromaius
ater
Dromaius
novaehoUandiae
FIG.2.Comparativeproportionsofbonesinthehindlimbsoi'Emuariusgidju,twospeciesofCasuariusandtwo
species of Dromaius. Note that while the tibiotarsal proportion remains relatively constant, the proportion
comprisingthe femurdecreasesas thatofthetarsometatarsus increases.
HINDLIMB PROPORTIONS OFEMUARIUSGIDJU 239
locomotion (Prange el al 1979), more closely
resembled Casuarius.
There are alternative explanations that accom-
modate both the cursorial hindlimb proportions
of£. gidju and the absence ofopen spaces. One
possibility is that the ground cover of the rain
forests was sufficiently open for cursorial ani-
mals to move rapidly without obstruction. For
example, modern Nothofagus forests are fre-
quentlyopenbelowthecanopy,withoutthedense
understory ofsome other rain forest types (pers.
obs.).
Neville Pledge (pers. comm.) suggested that a
situation may have existed similar to that which
occurred on Kangaroo Island when the dwarf
emu Dromaiusbaudinianuswas extant. Much of
the island's vegetation was very thick and would
have prevented rapid passage of a large animal
such as the emu. Large mammals, however,
forced runways through the vegetation, permit-
ting them to move with comparable, albeit re-
stricted, ease. The emus apparently took
advantage of these runways for their own prog-
ress. Likewise, the large mammals known from
FIG. 3. 'Combined' length ofLeafLocality proximal Riversleigh may have opened similar pathways
tarsomctatarsus and Riversleigh distal tarsometatar- throughthethick undergrowth, whichcould have
sus compared with tarsometatarsi of Recent been used by £". gidju. Nonetheless, D.
casuariids, in cranial view. From left, Dromaius baudinianus was also noted for being very fast
navaeliollancUae,D. ater.Emuariusgidju,Casuarius and virtually uncapturable in open areas.
cosuariusand C. bennettH.
A small Miocene dromaiid from Alcoota, NT,
Casuariushas been considered more primitive is known from 2 phalanges and 3 unassociated
than Dromaiuson the basis ofdistribution, habi- trochleae (Patterson & Rich, 1987). These are of
tat preferencesand hindlimb. Schodde& Calaby comparable size with £". gidju. There are slight
a(s19e7l2e)maenndtsScohfodthdeeT(u19m8b2u)ncaitnedatvhiefacuansas,owwahriicehs dPaitftfeerrseonnce&s iRnicphha(l1a9n8g7e)alretmaoinrepdhotlheosgey,speacnid-
represents the earliest lineages of extant Aus- mens as Dromaiussp. indct. until more complete
tralo-Papuan birds. The emus were placed by material is available. The Alcoota palaeohabitat
Schodde (1982) in the autochthonous Eyrean has been interpreted as a lake, bordered by sedge
offautnhae, Awuhsitcrhaleivaonlvhaebditiant.resBpoolnesse(t1o9t9h1e)ocpoennsiindg- oesrtg(rMaussrlraanyd,&gMreagdiirnigatno,w1o9o92d)l.aTnhdisanidsagduilflfyerfeonrt-
ered Emuarius to be closer to the dichotomy of environment than that interpreted for the older
cassowaries and emus than any other known Riversleigh deposits. Even if the Alcoota
taxon. buttoocursorially adaptedandDromaius- dromaiid proves to be E. gidju, it would have
liketohavebeen thecommonancestor. Itappears limited relevance to reconstructing the
tomarkin theCasuariidaeastageinthetransition Riversleigh habitats because of the age differ-
from a graviportal to a more cursorial locoino- encesofthedeposits. It would, however, suggest
lion. Although some characters of the hindlimb that E. gidju was preadapted for the more open
Alcootaenvironment.
of£. gidju are more similar to either Casuarius
orDromaius, many areintermediatebetween liv- Models of this species' locomotory mode de-
ingmembersofthesegenera(Boles, 1991). Boles pend on extrapolations from living, non-conge-
(1991) drew attention to the fact that the lower neric relatives. These are speculative, and must
limb bones were more similar to those of be treated as such, while palaeoenvironmental
Dromaius,whereasthefemur,whoseproportions reconstructions based on them require an even
are more dependent on the bird's mass than its greaterdegree ofcaution.
MEMOIRS OFTHEQUEENSLAND MUSEUM
24flt
BOLES,W-B- 1991.lie\\skmoWromoiusgidjuPatter-
son & Rich. 1987. with a reassessment of its
genericposition. Natural History MuseumofLos
Angeles CotmiyScience Series 36: I95-20S.
CAMPBELL K.E.'jR. & MARCUS, L. K>7! Tlie
relationshipofhindlinibhonedimensionslobody
weigh! in birds. Natural History Museum ofb;>.s
AngelesCounty Science Series36:395-412-
FRAKES. LA.. McGOWRAN. B. & BOWLER. J.M.
1987. Evolution ofAustralian environments. Pp.
1-16. In Dyne.G.R. ^ Walton. DAV.(edsl.Fauna
ofAustralia vol. lA.General articles. (Australian
Government PublishingService: Canberra).
HOWELL. A.B. 1944.Speedinanimals.Theirspecial-
ization for running and leaping. (Universiiy of
FEMOROMETATAFSALtNOE?t Chicago Press: Chicago).
MARCHANT. S. & HIGGINS. PJ. (coonJinalor^K
FIG.4. RelationshipofTMT/FMR and TBT/FMR o\' 1990. HandbookofAustralian, New Zealandand
RDevcoemnatUisspfecclioesscdocfirCcalcxst»t.arSipuesci(eospneunmbceirrceldesu)salonld- AnenyU)i.rciicbirds, 1.(Oxford UniversityPress:Syd-
iows: I. Casnanu:s benne/it; 2. C. cusitanus, 3. C. MARTIN.H.A. 1987.Cainozoichistory ofIhcvegeta-
Dii.nfuipopyeaneihJoiJciIialnaitiuisa:e,4,RFeJ.grheasasdimonu^lniines;is5T,BDT./Ia'ie"rM.K6=, StioountahnWdaclleism.atPeroofcetehdeiLnagcshollanthReiLvienrnReeagni.onS,ocNieetwy
0-45ITMT/FMR|-(- 1 13;r=0,84.Thefinevertical line ofNewSouthWales 19: 213-257.
represenis the inierseciiun ^^ith this line of the
TiliMeTte/xtF.MPRa^d(i1c7i2c)dfTorBETm/iuFtMriRnsisi^1i9Jj0t.tascalculatedin MILLTEAeuRrs.ttiraaArl.yiHao.nfMS1ou9u6st3eh.uAmFuosst1sr4i;all4iJar.a3iRn4ee2c0ob.ridrd.ssooffthtchSeoulatthe
ACKNOWLBDGEMENTS MORGAN. A.M. & SUTTON.1 1928. A crilical de-
scriptionofsomerecentlydiscoveredK^nesofthe
In addiiion lo acknowledgcmenis in Boles extinct Kangaroo Island Emu {Dromaius
t(h1e99K1a)nIgtahmanokIsNleavnidllheahiPilaeid,gLeynfonredHisocuasnsdioMnauo-n MURRdiAtY't.nenPi.aSniusM).EGEImRuIA28N:,1D-1.91992.Continuityand
riincfeorOmraitieig)an ofonr fEi.gugriedsj,u.aTnhdeARninvaerGsilleliegshpimeatfeo-r ccoonnlunntusntiiiinesmiTdrdolme ailnide NlaotnehMeimocTeenreriivocrrytcbrTuhlec
rial wascollected via an ARC Programme Grant PATTBEeRaSglOeN9,: 9C1,5&-21P8..V. RICH. l9S7.The iossii his-
lo M. Archer; a grajii from the Department of
tory ofihc emus. Drotnaius (A\es: Daimaiinae).
Arts, Sport, the Environineni. Tourism and Ter- Records o\' the South Australian Museum 21:
ritories to M. Archer, S. Hand and H. Godthclp; 85-117.
saupgproarnttffrroommthtehUeniNvaetrisointaylofENsteawteSoPuntihgWraulmemse: PRANSGcEal.inHgDof.sAkeNleDtEalRmSaOssN.toJb.oEd&y mRaAsIsIiNn.bIitr.ds19a7n9d.
GruntsSchemetoM. ArcherandA. Bartholomai; mammals, American Naturalist 113: 103-122.
and grants in aid to the Riversleigh Research SCUODDE, R i9S2, Origin, adaplalionandevolution
Projecl from Wang Australia Piy Ltd. ICl Aus- ofbirdsin arid Au.slralia. Pp. 191-224- In BrnkcT.
tralia and the Australian Geographic Society. W.R.&Grcenslade,PJ.M.(eds).Evolutionofthe
flora .aid faunaofaridAuslralia. (Peacock Puhli-
REFERENCES ealions: Frewville).
SCHODDE, R. & CALABY,J.H. 1972.The biogeog-
ARCHER. M . HAND. S.J. & GODTHELP. H. 1994. raphy ofthe Auslralo-Papuan bird and mammal
Riversleigh. 2nded (Reed: Svdnevl- faunasinrelationtoTorresStrait. Pp. 257-3(X). In
ARCHER. M.. HAND. S.. C}ObTHELP. H. & D Walker. D. (cl\X Bridge and harrier: the natural
MEGIRIAN. 1989- Fossil mammals of and cultural heritage ofTones Strait. tAustraiian
Riversleigh, norlhsvestem Queensland: prelimi- Nalion;il University Press; Canberra).
naryover%icw ol'hioslraligraphv.corretatinn and VICKERS-RICH. P- 199! Hie MesozoicandTertiary
envimnmcnial change. Ausualian Zoologist 25: history ofbirds on the Australian plalc Pp. 721-
29-65. SOS. In Vickers-Rich. PA^, Monaghan. J.M..
BAUMEL. J.J. & L.M. WfTMER. 1093. Osieologia. Baird. R.F. & Rich, T.H, (eds). Vertebrate pal-
Puhlicaiions of the Nuttnll Oniitholncical Cluh aeontology of Australasia. (Pioneer Design Stu-
23:45-132. dio: Melbourne).