RAPID COMMUNICATION

THE EARLIEST TERTIARY THERIAN MAMMAL FROM SOUTH AMERICA

FRANCISCO J. GOIN1,*, ROSENDO PASCUAL1, MARCELO F. TEJEDOR2, JAVIER N. GELFO1,
MICHAEL O. WOODBURNE3, JUDD A. CASE4, MARCELO A. REGUERO1, MARIANO BOND1,

GUILLERMO M. LÓPEZ1, ALBERTO L. CIONE1, DANIEL UDRIZAR SAUTHIER1, LUCÍA BALARINO1,
ROBERTO A. SCASSO5, FRANCISCO A. MEDINA5, and MARÍA C. UBALDÓN6

1División Paleontología Vertebrados, Museo de La Plata, Paseo del Bosque s/n, 1900 La Plata, Argentina; 2LIEB, Facultad de
Ciencias Naturales, Sede Esquel, Universidad Nacional de la Patagonia “San Juan Bosco,” Sarmiento 849, 9200 Esquel,
Argentina; 3Department of Geology, Museum of Northern Arizona, Flagstaff, Arizona, U.S.A.; 4Department of Biology,

St. Mary’s College, Moraga, California 94575, U.S.A.; 5Departamento de Ciencias Geológicas, Facultad de Ciencias Exactas y
Naturales, Universidad de Buenos Aires, Pabellón 2, Ciudad Universitaria, 1428 Buenos Aires, Argentina; 6SEGEMAR/IGRM,

Pasaje Peñaloza 105, CC 90, 9200 Esquel, Chubut, Argentina

We describe the oldest South American Tertiary therian mam-
mal from Danian-equivalent strata of northwestern Chubut
province, Argentina (Fig. 1). The specimen is an isolated lower
molar most likely pertaining to a polydolopimorphian marsupial,
a group known from the Late Cretaceous of North America as
well as later Paleocene and Eocene deposits in South America.
The derived nature of this bunodont marsupial provides evi-
dence for an evolutionary radiation of marsupials in the earliest
Paleocene or Late Cretaceous of the Americas. Whereas the
presence of this radiation had been suspected previously, the
Lefipán specimen (Fig. 2) narrows the interval within which that
radiation may have occurred.

Up to now, the earliest glimpse of the diversity of therian
mammals of South America has been gained from the Tiupampa
local fauna of Bolivia and tentatively correlated to late Pa-
leocene sites in Argentina (Selandian; 58.7-61.7 Ma) (Somoza et
al., 1995; Marshall et al., 1997; radioisotopic ages after Luter-
bacher et al., 2004). In addition to relict gondwanan monotremes
and sudamericids, taxa from these sites reflect a growing diver-
sity of marsupial and placental immigrants from North America.
For marsupials, these include a few derived members of the basal
peradectian radicle (e.g., Roberthoffstetteria), along with ameri-
delphians represented by early didelphoids, with some (Didel-
phopsis, Eobrasilia, Gaylordia) showing probable affinities with
North American forms (Case et al., 2005). In addition, a rela-
tively broad array of ameridelphian taxa is allocated to groups
currently endemic to South America as far as their origin is
concerned: derorhynchid didelphids (Derorhynchus, Gaylordia,
Minisculodelphys), protodidelphids (Bobbschaefferia, Guggen-
heimia, Procaroloameghinia, Protodidelphis, Robertbutleria),
sternbergiids (Itaboraidelphys), ?sparassodontan carnivores
(Maylestes, Allqokirus, ?Nemolestes), and a variety of polydo-
lopimorphians, including epidolopine bonapartheriids (Epido-
lops). A third group consists of endemic microbiotheriid australi-
delphians (Khasia, Mirandatherium), with potential significance
for dispersal to, and radiation within, Australia. Most of these
taxa imply that an extensive evolutionary radiation took place in
South America prior to the medial Paleocene, and this usually is
taken as evidence in favor of a Late Cretaceous entry of the
necessary ancestral stocks (Muizon and Brito, 1993), for which
there is evidence of Late Cretaceous polydolopimorphians hav-
ing originated in North America (Case et al., 2005). The placen-
tal mammals are comprised mostly of condylarths and condy-

larth-like taxa considered to have had a likely early Paleocene
source in North America, with the possible exception of the
Pantodonta.

Of the immigrant groups, the presence in Danian-equivalent
deposits of a marsupial (with a possible age range of 61.7–65.5
Ma) was expected, but the derived nature of this taxon, if its
proposed polydolopimorphian affinity is borne out by further
material, further strengthens the implications in favor of a Late
Cretaceous entry of marsupials to South America. The presence
of polydolopimorphian taxa in Judithian as well as Lancian
(Maastrichtian) deposits in North America (Case et al., 2005)
indicates that the evolution of this group was underway by about
75 Ma, so that a member of this relatively derived marsupial
group in Danian-equivalent deposits of South America is com-
patible with the North American data.

SYSTEMATIC PALEONTOLOGY

MAMMALIA Linnaeus, 1758
Supercohort MARSUPIALIA Illiger, 1811

Cohort AMERIDELPHIA Szalay, 1982
Order ?POLYDOLOPIMORPHIA Ameghino 1897
COCATHERIUM LEFIPANUM, gen. et. sp. nov.

(Fig. 3)

Etymology—For “Coca” San Martín, of the Estancia San
Ramón, whose invaluable assistance to scientists of the Museo
de La Plata and particularly to us for many years, made our
work in the western Chubut River region possible (Fig. 1). Lefi-
panum from Lefipán, name of one of the most prominent native
(mapuche) inhabitants of the region.

Holotype—LIEB-PV 1001, an isolated right lower molar.
Locality, Horizon, and Age—The Grenier section (Fig. 2) of

the Lefipán Formation is located at grid coordinates 42° 46� S
and 69° 51� W, on Grenier Farm, about 20 km west of the town
of Paso del Sapo (Fig. 1), in the middle Chubut region, Chubut
Province, Argentina. The regional geology (Spalletti, 1996)
shows that the Lefipán Formation (and underlying Paso del Sapo
Formation) unconformably overlies older Cretaceous strata re-
ferable to the Chubut Group. In turn, the Lefipán Formation is
unconformably overlain by lower Tertiary (Eocene) strata of the
Laguna del Hunco Series (Wilf et al., 2003 and references
therein). The Lefipán Formation is a 200-m thick, marine silici-

Journal of Vertebrate Paleontology 26(2):505–510, June 2006
© 2006 by the Society of Vertebrate Paleontology

505

https://www.researchgate.net/publication/249549945_Estuarine_and_shallow-marine_sedimentation_in_the_Upper_Cretaceous-Lower_Tertiary_west-central_Patagonian_Basin_Argentina?el=1_x_8&enrichId=rgreq-53d4bc5d-e62d-414d-9095-791a1c386587&enrichSource=Y292ZXJQYWdlOzI0OTAyMzAwMjtBUzo5ODc3NjcxNDg0MjEzMEAxNDAwNTYxNjEyOTg3


clastic unit, mainly composed of massive and heterolithic mud-
stones with several intercalations of sandstone packets and co-
quina beds that have a well preserved molluscan fauna of Maas-
trichtian-Paleocene age (Spalletti, 1996). Sandstone beds are
massive, or display parallel or cross-bedding. Sometimes they
show well preserved bioturbation of the Skolithos-Cruziana like
ichnofacies and contain phosphatic concretions. The whole unit
was deposited in a shallow epeiric sea, in which the sandstones
were interpreted as bars and shoals that accumulated in a
shoreface environment. The lower part of the Lefipán Formation
(Spalletti, 1996) corresponds to strata assigned to Asociacion I
(Medina et al., 1990) of Late Maastrichtian age. The upper part
of the Lefipán Formation begins with a maximum flooding sur-
face considered to mark the beginning of the Paleocene (Spal-
letti, 1996), and is equivalent to Asociacion II in Medina et al.
(1990) paleontologically determined to be of Danian-equivalent
age. Although the part of the Lefipán Formation that yielded
Cocatherium is not discussed in either Spalletti (1996) or Medina
et al. (1990), Asociacion II is characterized by the restricted
presence of Meretrix chalcedonica, comparable to the beds that
yield Cocatherium.

A stratigraphic section 65m thick (Fig. 2) has been developed
in the upper part of the Lefipán Formation at the Grenier Farm
site. The lower 12 m of the section are characterized by brown
mudstone with calcareous concretions and rare heterolithic bed-
ding, and rusty-brown massive sandstone beds. The mudstone
beds contain remains of Pterotrigonia windhausenia (Wilckens),
indicative of a Maastrichtian age. The brown mudstones are
overlain by about 5 m of massive sandstone that contains the
bivalves Meretrix chalcedonica and Venericardia, gastropods,
shark and ray teeth and claws of callianassid crabs.

The holotype specimen (LIEB-PV 1001) was found at the
Meretrix–Venericardia level, about 15 m stratigraphically above
the base of the measured section displayed in Figure 2. The lower
and middle stakes shown in Figure 2 denote the interval within
which the K-T boundary is considered to occur based on the
fossils recovered by one of us (MCU). The Meretrix level occurs
about 5 m stratigraphically above the K-T interval in the upper
member of the Lefipán Formation (Spaletti, 1996), who consid-
ered the entire upper member of the Lefipán Formation to rep-
resent the earliest Paleocene Series. Medina et al., (1990) assign
the relevant strata to Asociacion II, with fossils of Danian-
equivalent age. The stratigraphic section in Figure 2 was pre-
pared by some of us (MCU, RAS, FAM) and colleagues as a
refinement of the previous work. Although the new investigation

cannot yet precisely identify the stratigraphic location of the K-T
boundary in Figure 2, it is clear that Cocatherium and its asso-
ciated invertebrate fossils occur stratigraphically above this in-
terval, and are of Danian-equivalent age. In addition, Meretrix
chalcedonica is restricted to Association II, the oldest of the two
Paleocene biostratigraphic assemblages recognized in the
Lefipán Formation by Medina et al. (1990). Actinopterigian fish
and neoselachians found stratigraphically 2–3 m above and be-
low the Meretrix horizon are consistent with a Danian-equivalent
age for the deposit. Neoselachian teeth are comprised mostly of

FIGURE 1. Map of Chubut Province, Argentina and Mammal Locality
yielding Cocatherium. Chubut Province is black area in index map of
Argentina.

FIGURE 2. Stratigraphic section at Grenier Farm locality showing po-
sition of Cocatherium level with respect to temporal indicators discussed
in the text.

JOURNAL OF VERTEBRATE PALEONTOLOGY, VOL. 26, NO. 2, 2006506

https://www.researchgate.net/publication/249549945_Estuarine_and_shallow-marine_sedimentation_in_the_Upper_Cretaceous-Lower_Tertiary_west-central_Patagonian_Basin_Argentina?el=1_x_8&enrichId=rgreq-53d4bc5d-e62d-414d-9095-791a1c386587&enrichSource=Y292ZXJQYWdlOzI0OTAyMzAwMjtBUzo5ODc3NjcxNDg0MjEzMEAxNDAwNTYxNjEyOTg3
https://www.researchgate.net/publication/249549945_Estuarine_and_shallow-marine_sedimentation_in_the_Upper_Cretaceous-Lower_Tertiary_west-central_Patagonian_Basin_Argentina?el=1_x_8&enrichId=rgreq-53d4bc5d-e62d-414d-9095-791a1c386587&enrichSource=Y292ZXJQYWdlOzI0OTAyMzAwMjtBUzo5ODc3NjcxNDg0MjEzMEAxNDAwNTYxNjEyOTg3


Carcharias, not known prior to the Paleocene, and remains of
Hypolophodon, a taxon previously known only from Eocene
deposits (Kemp et al., 1990). The presence of this genus in this
part of the Lefipán Formation apparently represents a down-
ward range extension. It also is notable that neoselachian groups
otherwise typical of Cretaceous strata are absent in this part of
the Lefipán Formation (Hybodontidae, Acrodontidae, Anacora-
cidae, Polyacrodontidae, Sclerorhynchidae, Ptychodontidae,
Rhombodontidae, Scapanorhynchus, Igdabatis), consistent with
its correlation to the Danian.

Diagnosis—Bunodont marsupial of relatively large size; trigo-
nid with distinct paraconid; protoconid and metaconid trans-
versely opposite each other; cristid obliqua attaches to trigonid
posterior to protoconid and well labial to protocristid notch;
hypoconulid distinct, terminal, and closely appressed with, but
distinctly separate from, much taller entoconid; entoconid with
distinctly curved preentocristid that meets the postmetacristid to
block the talonid basin; talonid shorter than trigonid, with shal-
low talonid basin; lingual surface of hypoconid flat to labially
slightly concave, rather than basally rounded. Labial surface of
the tooth slopes basally and is deeper than the nearly vertical
lingual surface. This diagnosis is provisional pending discovery of
additional materials of this or other South American Paleocene
polydolopimorphian marsupials.

Measurements—Total length, 2.90 mm; trigonid width, 2.20
mm; talonid width, 2.27 mm.

Description—LIEB-PV 1001 is a fully tribosphenic bunodont
well-worn right lower molar. Although somewhat shortened due
to the angulate paracristid, the trigonid is distinctly longer than
the talonid. The talonid is slightly wider than the trigonid. The
anterior and posterior roots are anteroposteriorly compressed. A
narrow anterior cingulum is present at the anterior base of the
tooth, and apparently was somewhat wider at the anterolabial
surface of the protoconid than at the anterior base of the par-
conid. No other cingula are present.

The paraconid is set somewhat labially relative to the meta-
conid but, judging from the well-developed wear pattern, still
was a distinct cuspid. The paraconid portion of the paracristid
extended directly transversely to meet the anterolingually ori-
ented protocristid portion at a distinct, although oblique, angle,
at the anterior midpoint of the tooth. The paraconid is set closely
adjacent to the anterolingual base of the metaconid, with the two
cuspids being separated by a distinct, narrow, crevice. Thus the
trigonid basin likely was very narrow.

The protoconid and metaconid are set transversely opposite to
each other, and are about equally large. The labial surface of the
protoconid slopes strongly toward the base of the tooth, whereas
the lingual border of the metaconid is more nearly vertical. It is
likely that the protoconid and metaconid were connected by a
transverse protocristid. The enamel surface preserves a weak,
short, postmetacristid. Based on the current wear pattern, it ap-
pears that the metaconid and protoconid were the tallest cuspids
of the tooth, with the hypoconid, entoconid, paraconid and hy-
poconulid being progressively lower.

The talonid is shorter than typical of the plesiomorphic therian
condition. The hypoconid is about as strongly worn as the trigo-
nid cuspids. As for the protoconid, the labial surface of the hy-
poconid slopes basally, in contrast to the more nearly vertical
lingual surface of the entoconid. The cristid obliqua apparently
was anterolingually oriented. It now reaches the rear of the tri-
gonid at the posterior side of the protoconid; at this stage of
wear, it is continuously connected to the worn trigonid surface.
Previous to this stage of wear it is likely that the cristid obliqua
abutted the rear of the trigonid below the occlusal edge of the
trigonid wall. The short posthypocristid extends posterolingually
to the hypoconulid, which apparently also was oriented oblique-
ly. A remnant of a valley that separates the posterior surface of
the hypoconid from that of the hypoconilid suggests that the

posthypocristid sloped basally toward a comparably sloping pre-
hypoconulid cristid. The entoconid still preserves much of its
original structure, and is a relatively tall cuspid that is laterally
compressed. It also is inclined labially and slightly anteriorly. In
contrast to the other cuspids which have nearly flat occlusal wear
surfaces, that of the entoconid clearly slants labially. The ento-
conid is separated from the adjacent base of the hypoconulid by
a sharp, distinct, crevice, which broadens slightly anteriorly to
separate the entoconid from the lingual slope of the hypoconid.
The talonid basin thus is shaped as an asymmetrical valley, with
a steeper lingual than labial surface, rather than as a basin. The
lingual surface of the hypoconid is not expanded and does not
occlude the talonid basin. A distinct preentocristid extends an-
teriorly from the entoconid and weakly meets the opposing post-
metacristid so that the talonid basin is closed lingually. The hy-
poconulid is very close to the entoconid, although distinctly sepa-
rated from it.

An interdental wear facet is present on both the anterior and
posterior surfaces of the tooth so that it is most likely an inter-
mediate molar, neither the first nor the last. In anterior and
posterior views (Fig. 3d, e), the asymmetrical slope of the tooth
is apparent with much more sloping, and taller, labial rather than
lingual surfaces.

Discussion

Based on its Danian-equivalent age, Cocatherium is the oldest
Paleocene record of a marsupial or any therian mammal in the
Southern Hemisphere. As preserved, its morphology and worn
occlusal surfaces show that Cocatherium had a full complement
of therian cuspids and cusps, including a protocone in the upper
dentition.

LIEB-PV 1001 contains all the features characteristic of a the-
rian mammal, including the requisite cusps, their appropriate
locations in the trigonid and talonid, and pertinent wear facets. A
numbered series of surfaces has been documented (Crompton,
1971) for shearing occlusion between the upper and lower molars
which define the tribosphenic molar morphology of therian
mammals. Even though LIEB-PV 1001 is very worn, it still re-
tains evidence of three distinct shearing surfaces. A large occlu-
sal wear facet is present on the anterolabial surface of the pro-
toconid along the paracristid prior to its lingual bend, which

FIGURE 3. Drawings of the lower molar LIEB-PV 1001, Cocatherium
lefipanum. Labial (a), lingual (b), occlusal (c), anterior-occlusal (d), and
posterior-occlusal (e) views. Scale: 1 mm.

GOIN ET AL—THE EARLIEST TERTIARY THERIAN MAMMAL FROM SOUTH AMERICA 507



corresponds to shearing surface N° 2 (Crompton, 1971; and all
subsequent numbered surfaces), that occludes with the post-
metacrista on the upper molar. A second, but smaller, occlusal
facet on the anterolabial surface of the hypoconid is referred to
shearing surface N° 3, which would occlude with the upper mo-
lar’s postparacrista. The posterolabial surface of the hypoconid
also possesses a wear facet; it corresponds to shear facet N° 4,
and occludes with the premetacrista of the upper molar. In short,

the cusps and observable shearing surfaces correspond with
those of a therian mammal. We interpret that the peculiar, ex-
tended, and almost continuous wear facets of the masticatory
surface resulted from the prolonged medial movement of the
lower jaw that followed its upward movement during occlusion.
The almost continuous surface that extended throughout the tri-
gonid—to connect the paracristid and the metacristid—appears
to have been produced by the joint action of the post-metaconule

FIGURE 4. Cladogram of selected marsupial and other therian taxa. Strict Consensus of 25 equally parsimonious trees. TL � 1175; CI � 0.402;
RI � 0.829; HI � 0.598. Cocatherium is shown within polydolopimorophians. Boldface numbers refer to nodes. Unless otherwise indicated, the
classification follows Case et al. (2005, hereafter C’05, which includes Marshall et al., 1990). Node 1 comprises the outgroup. Node 2 includes
Metatheria (Kokopellia; Cifelli, 1993). Nodes 3-5 comprise Peradectia. Node 6 is after C’05, plus Turgidodon. Nodes 7, 8, and 9 comprise Marsupialia,
Incertae Sedis, after C’05, with Caroloameghiniidae (minus Glasbius) grouped with Polydolopimorphia. Node 10 is Alphadelphia; with Perrodelphys
homoplasiously added. Node 11 is an unnamed plesion including Pediomyidae and Dakotadens. Node 12 is an unnamed plesion containing Ameri-
delphia and Australidelphia. Node 13 is Australidelphia (C’05) plus Chulpasia, likely a homoplasy. Nodes 14 and 15 are Ameridelphia (C’05) plus
Caenolestoidea and Borhyaenoidea. Node 16 is Didelphimorphia (C’05), with Herpetotheriidae indicated. The other taxa (Pucadelphys – Bobb-
schaefferia) are variously allocated in C’05 to Pucadelphidae, “Didelphidae,” Eobrasiliinae, Derorhynchidae, and Protodidelphidae. The Majority
Rule insert (below) shows Cocatherium nested within Polydolopimorphia.

JOURNAL OF VERTEBRATE PALEONTOLOGY, VOL. 26, NO. 2, 2006508

https://www.researchgate.net/publication/275345017_The_origin_of_tribosphenic_molar?el=1_x_8&enrichId=rgreq-53d4bc5d-e62d-414d-9095-791a1c386587&enrichSource=Y292ZXJQYWdlOzI0OTAyMzAwMjtBUzo5ODc3NjcxNDg0MjEzMEAxNDAwNTYxNjEyOTg3


crista and the post-metacrista. The wear surface that joins the
hypoconid and hypoconulid must have been produced by the
joint action of the post-paracrista and the pre-metacrista of the
upper molars. In turn, the wear surface affecting the entoconid
must have been produced by the action of the protocone.

As a therian, Cocatherium also may be considered as a mem-
ber of the Boreosphenida (Luo et al., 2001), and even though
therian-like australosphenidans are known in late Jurassic strata
of Chubut Province, South America (Rauhut et al, 2002), there
is no evidence that this group persisted into the Cenozoic of that
continent. In addition to its own apomorphic bunodonty in this
context, Cocatherium lacks any australosphenidan apomorphies,
such as a ‘wrapping mesial cingulid.’ Although the specimen is
well worn, enough can be seen of its morphology to suggest
polydolopimorphian affinities for Cocatherium. Chief among
these features is the strong association of hypoconulid and en-
toconid in conjunction with the large size and elevation of the
entoconid, which typically is a marsupial characteristic (Cifelli
and Muizon, 1997). Among South American marsupials, only the
polydolopimorphians also are as bunodont and relatively as large
as shown in Figure 3 (Goin and Candela, 2004). The relative
development of the other cuspids, and the longer (and sloping)
labial versus the shorter (and more nearly vertical) lingual sur-
face of the tooth combine to suggest a polydolopimorphian af-
finity. In Epidolops and polydolopines, in particular, there is a
tendency to reduce the hypoconulid and increase the size of the
entoconid (Goin and Candela, 2004). In the Prepidololpidae and
Bonapartheriidae the entoconid is large and strongly developed,
but doesn’t show the anterior and posterior crests of Co-
catherium. In contrast to some polydololpimorphians (e.g.,
Bonapartheriids), the paraconid still is relatively strong, and still
lingual of the midline of the tooth. A phylogenetic analysis of
therian molar characters (Fig. 4) based on PAUP 4.0b (Swofford,
2001), using a combination of ordered, unordered, and unique
character trees (Appendixes I, II, Supplementary Data; www
.vertpaleo-org/jvp/JVPcontents.html). Whereas aspects of the
cladogram reflect the homoplasious nature of molar characters,
Cocatherium clearly is contained within polydolopimorphian
marsupials at Node 8 (also Majority Rule insert, Fig. 4), and
separated from Alphadelphia and Ameridelphia at Node 12.
Other examples of homoplasy include the diversity of associa-
tions among Didelphimorphia at Node 16, and within the Poly-
dolopimorpia in Node 8, but the overall affinities of Cocatherium
are consistent with its limited known morphology. With respect
to therians, Fig. 5 is a phylogenetic analysis of South American
condylarth and ‘ungulate’ taxa based on the character states re-
fined in Gelfo (2004) and shown in Appendix III (Supplemen-
tary Data, www.vertpaleo.org/jvp/JVPcontents.html). In this
analysis, Cocatherium shows virtually plesiomorphic states com-
parable to the hypothetical ancestor out-group and exhibits no
compelling affinity with any American condylarth or ‘ungulate.’

Although a few of the more plesiomorphic ‘ungulates,’ con-
dylarths and condylarth-like taxa reviewed in Cifelli and Muizon
(1997) reveal individual and selected similarities to Cocatherium,
the trigonid of Cocatherium is derived in the appressed paraco-
nid and metaconid, the trigonid being virtually closed lingually,
and the protoconid and metaconid lie opposite each other trans-
versely. Cocatherium probably had a planar posterior surface of
the trigonid, a plesiomorphic character, and a weak precingulid
also is plesiomorphic. An affinity of Cocatherium with condy-
larths and other South Amerian ungulates is virtually ruled out
by the paracristid in Molinodus and other South American con-
dylarths (and didolodontids and primitive litopterns) having a
distinctively arcuate pattern where it actually extends anteriorly
from the protoconid before curving lingually (and almost recurv-
ing) to reach the paraconid. It is one of the strongest synapo-
morphies to unite Kollpaniinae + Didolodontidae + litopterns

(Muizon and Cifelli, 2000), and its absence in Cocatherium vir-
tually excludes this genus from an ‘ungulate’ association.

Polydolopimorphians have been recently reported (Case et al.,
2005) from the Late Cretaceous of North America, suggesting
that the evolution of this otherwise South American group was
already underway. In this context it is not surprising to find a
relatively derived member of that group in Danian-equivalent
deposits of South America. The North American presence of
polydolopimorphians also raises questions about the timing, as
well as place of evolution, of other nominally South American
groups with potential links to North American clades. The new
polydolopimorphian reported upon here not only narrows the
focus on the timing of the implied evolutionary diversification of
the group to the Late Cretaceous (as opposed to early Pa-
leocene), but also implies that a significant record of this radia-
tion is to be expected in South America, hopefully to stimulate
further search for its expected manifestation.

LITERATURE CITED

Ameghino, F. 1897. Mammifères crétacés de l’Argentine. Deuxiéme con-
tribution à la conaissance de la fauna mammalogique des couches à
Pyrotherium. Boletin Institut Geographie Argentino xvii:1–117.

Case, J., Goin, F. and M. Woodburne, M. 2005. “South American” mar-
supials in the Late Cretaceous of North America and the Origin of
Marsupial Cohorts. Journal of Mammalian Evolution 12, 3/4:
461–494.

Cifelli, R.L. 1993. Early Cretaceous mammal from North America and
the evolution of marsupial dental characters. Proceedings National
Academy of Sciences 90:9413–9416.

Cifelli, R.L., and Muizon, C. de. 1997. Dentition and jaw of Kokopellia
juddi, a primitive marsupial or near-marsupial from the medial Cre-
taceous of Utah.. Journal of Mammalian Evolution 4:242–258.

FIGURE 5. Cladogram of selected condylarth and condylarth-like
taxa. Consensus of 30 equally parsimonious trees, 1000 random input
orders. TL � 59; CI � 0.644, RI � 0.817; HI � 0.356. Cocatherium is
shown outside of all condylarth taxa.

GOIN ET AL—THE EARLIEST TERTIARY THERIAN MAMMAL FROM SOUTH AMERICA 509

https://www.researchgate.net/publication/11469557_A_Jurassic_mammal_from_South_America?el=1_x_8&enrichId=rgreq-53d4bc5d-e62d-414d-9095-791a1c386587&enrichSource=Y292ZXJQYWdlOzI0OTAyMzAwMjtBUzo5ODc3NjcxNDg0MjEzMEAxNDAwNTYxNjEyOTg3
https://www.researchgate.net/publication/258998033_The_condylarths_archaic_Ungulata_Mammalia_from_the_early_Palaeocene_of_Tiupampa_Bolivia_Implications_on_the_origin_of_the_South_American_ungulates?el=1_x_8&enrichId=rgreq-53d4bc5d-e62d-414d-9095-791a1c386587&enrichSource=Y292ZXJQYWdlOzI0OTAyMzAwMjtBUzo5ODc3NjcxNDg0MjEzMEAxNDAwNTYxNjEyOTg3
https://www.researchgate.net/publication/11994421_Dual_origin_of_tribosphenic_mammals?el=1_x_8&enrichId=rgreq-53d4bc5d-e62d-414d-9095-791a1c386587&enrichSource=Y292ZXJQYWdlOzI0OTAyMzAwMjtBUzo5ODc3NjcxNDg0MjEzMEAxNDAwNTYxNjEyOTg3


Crompton, A.W. 1971. The origin of the tribosphenic molar; pp 65–87 in
D.M. Kermack & K.A. Kermack (eds.) Early Mammals, Linnaean
Socitety Zoological Journal 50, Suppl. 1, London.

Gelfo, J.N. 2004. A new South American mioclaenid (Mammalia: Un-
gulatomorpha) from the Tertiary of Patagonia, Argentina.
Ameghiniana 43:475–484.

Goin, F.J., and A. M. Candela. 2004. New Paleogene marsupials from the
Amazonian basin, Southeastern Perú. Pp. 15–60 In: K. E. Campbell
Jr. (Ed.), The Paleogene Mammalian Fauna of Santa Rosa, Ama-
zonian Perú. Natural History Museum of Los Angeles County, Sci-
ence Series 40, Los Angeles.

Illiger, C. 1811. Prodromus systematis mammalium et avium additis ter-
minis zoographicis utriusque classis. Berlin: C. Salfeld.

Kemp, D., Kemp, and Ward, D.J. 1990. An illustrated guide to the British
Middle Eocene vertebrates. David Ward, London, p. 59.

Linnaeus, C. 1758. Systema naturae per regna tria naturae, secundum
classes, ordines, genera, species, cum characteribus, differentiis, syn-
onymis, locis. Vol. 1: Regnum animale. Editio decima, reformatea.
Stockholm: Laurentii Salvii.

Luo, Z.-X., Cifelli, R.L., and Kielan-Jaworowska, Z. 2001. Dual origin of
tribosphenic mammals. Nature 409:53–57.

Luterbacher, H.P., Ali, J.R., Brinkhuis, H., Gradstein, F.M., Hooker, J.J.,
Monechi, S., Ogg, J.G., Powell, J., Rohl, U. Sanfilippo, A., and
Schmitz, B. 2004. The Paleogene Period, in A Geologic Time Scale,
F. Gradstein, J. Ogg, and A. Smith (eds.), Cambridge University
Press. Cambridge, pp. 384–408.

Marshall, L.G., Case, J.A., and Woodburne, M.O. 1990. Phylogenetic
relationships of the families of marsupials, in H. H. Genoways (ed.),
Current Mammalogy 2. Plenum Press. New York, pp. 433–505.

Marshall, L.G., Sempere, T., and Butler, R.F. 1997. Chronostratigraphy
of the mammal-bearing Paleocene of South America. Journal South
American Earth Sciences 10(1):49–70.

Medina, F.A., Camacho, H.H., and Malagnino, E.C. 1990. Bioestrati-
grafía del Cretácio superior-Paleocene marino de la Formación

Lefípan, Barranca de los Perros, Río Chubut, Argentina. V. Cong.
Arg. Paleont. y. Bioestrat. Actas I:137–142.

Muizon, C. de., and Brito, I. 1993. Le basin calcaire de Sao Jose de
Itaborai (Rio de Janeiro, Bresil): les relaciones faunique avec le site
de Tiupampa (Cochabamba, Bolivie). Annales de Palaeontologie
79:233–269.

Muizon, C. de., and Cifelli, R.L. 2000. The “condylarths” (archaic Un-
gulata, Mammalia) from the early Palaeocene of Tiupampa (Bo-
livia): implictions on the origin of the South American ungulates.
Geodiversitas 22:47–150.

Pascual, R., Oritz Jaureguizar, E., and Prado, J.L. 1996. Land Mammals:
Paradigm for Cenozic South American Geobiotic Evolution; pp.
265–319 in G. Arratia (ed), Contribultions of South America to
Vertebrate Paleontology, Münchner Geowissenschaftliche Abhand-
lungen A. 30, München.

Rauhut, O.W., Martin, T., Ortiz-Jaureguizar, E., and Puerta, P. 2002. A
Jurassic mammal from South America. Nature 416:165–168.

Somoza, R., Cladera, G., and Archangelsky, S. 1995. Una nueva Tafo-
flora Paleocena de Chubut Patagonia. Su Edad y Ambiente de de-
positación. VI Congreso Argentino de Paleontología y Bioestrati-
grafía, Trelew. Actas. 265–269.

Spalletti, L.A. 1996. Estuarine and shallow-marine sedimentation in the
Upper Cretaceous—Lower Tertiary west-central Patagonian Basin
(Argentina). In Geology of Siliciclastic Shelf Seas, M. de Batist and
P. Jacobs (eds.). Geology Society Special Paper 117. London.

Swofford, D.L. 2001. PAUP*. Phylogenetic Analysis Using Parsimony
(*and Other Methods), v. 4. Sinauer, MA.

Szalay, F.S. 1982. Phylogenetic relationships of the marsupials. In Pylog-
enie et paleobiogeographie, Livre jubilaire en l’honneur de Robert
Hoffstetter (E. Buffetaut, ed.). Geobios Mem. Spec. 6:177–190.

Wilf, P., N. R. Cú neo, K. R. Johnson, J. F. Hicks, S. L. Wing, and J. D.
Obradovich. 2003. High plant diversity in Eocene South America:
evidence from Patagonia. Science 300:122–125.

Submitted 2 February 2006; accepted 20 March 2006.

JOURNAL OF VERTEBRATE PALEONTOLOGY, VOL. 26, NO. 2, 2006510