Content uploaded by Lilian Paglarelli Bergqvist
Author content
All content in this area was uploaded by Lilian Paglarelli Bergqvist on Feb 09, 2018
Content may be subject to copyright.
323
GEODIVERSITAS • 2004 • 26 (2) © Publications Scientifiques du Muséum national d’Histoire naturelle, Paris. www.geodiversitas.com
The Xenarthra (Mammalia) of São José
de Itaboraí Basin (upper Paleocene, Itaboraian),
Rio de Janeiro, Brazil
Lílian Paglarelli BERGQVIST
Departamento de Geologia/IGEO/CCMN/UFRJ, Cidade Universitária,
Rio de Janeiro/RJ, 21949-940 (Brazil)
bergqvist@ufrj.br
Érika Aparecida Leite ABRANTES
Departamento de Geologia/IGEO/CCMN/UFRJ, Cidade Universitária,
Rio de Janeiro/RJ, 21949-940 (Brazil)
Leonardo dos Santos AVILLA
Departamento de Geologia/IGEO/CCMN/UFRJ, Cidade Universitária,
Rio de Janeiro/RJ, 21949-940 (Brazil)
and Setor de Herpetologia, Museu Nacional/UFRJ,
Quinta da Boa Vista, Rio de Janeiro/RJ, 20940-040 (Brazil)
Bergqvist L. P., Abrantes É. A. L. & Avilla L. d. S. 2004. — The Xenarthra (Mammalia) of São
José de Itaboraí Basin (upper Paleocene, Itaboraian), Rio de Janeiro, Brazil. Geodiversitas
26 (2) : 323-337.
ABSTRACT
Here we present new information on the oldest Xenarthra remains. We
conducted a comparative morphological analysis of the osteoderms and post-
cranial bones from the Itaboraian (upper Paleocene) of Brazil. Several osteo-
derms and isolated humeri, astragali, and an ulna, belonging to at least two
species, compose the assemblage. The bone osteoderms were assigned to
Riostegotherium yanei Oliveira & Bergqvist, 1998, for which a revised diagno-
sis is presented. The appendicular bones share features with some “edentate”
taxa. Many of these characters may be ambiguous, however, and comparison
with early Tertiary Palaeanodonta reveals several detailed, derived resem-
blances in limb anatomy. This suggests that in appendicular morphology, one
of the Itaboraí Xenarthra may be the sister-taxon or part of the ancestral stock
of Palaeanodonta.
KEY WORDS
Mammalia,
Xenarthra,
Cingulata,
Riostegotherium,
Astegotheriini,
Palaeanodonta,
armadillo,
osteoderm,
appendicular skeleton.
INTRODUCTION
The discovery of osteoderms in Itaboraí was first
reported 50 years ago (Paula-Couto 1949), but
they were only studied in the mid-1970s by
Scillato-Yané (1976), who described two isolated
osteoderms, assigning them to Prostegotherium
aff. P. astrifer Ameghino, 1902 a dasypodid from
the Casamayoran of Argentina. These osteo-
derms, together with new material, were later
studied by Oliveira & Bergqvist (1998), who
considered them distinct from P. astrifer as well
as from any early Tertiary Patagonian Astego-
theriini Vizcaíno, 1994. They proposed a new
taxon for this material, Riostegotherium yanei, the
oldest known cingulate.
Cifelli (1983) described two different tarsals of
Itaboraí cingulates, assigning one to a dasypodid
and the other to a glyptodontid(?); the latest
assignment was questioned by Bergqvist & Oliveira
(1998). New xenarthran remains (humeri and
ulna), allocated by Bergqvist & Oliveira (1995)
to cingulates, were only briefly described by the
authors.
In this paper we gather all available information
on Itaboraí Xenarthra, providing a broad descrip-
tion and illustrations of the forelimb bones men-
tioned above, and review the material already
described. As the bones and osteoderms were
collected separately, and there is no confident in-
formation from where they were recovered, no di-
rect association could be made among the bones
themselves and between them and the osteo-
derms. Each bone is thus discussed separately.
The São José de Itaboraí Basin, located in the
County of Itaboraí, Rio de Janeiro State, is the
oldest (Itaboraian, upper Paleocene) and smallest
basin of the Continental Rift of Southern Brazil.
The sediments that filled this basin are grouped
in two sequences: the lower sequence (S1) is basic-
ally composed of interbedded clastic and chemi-
cal carbonates that are originated from debris
flows in a tectonic lake with hydrothermal fonts.
Fossil mollusks are abundant in this sequence.
The upper sequence (S2) comprises fissure fill
deposits with marls and collapse breccias where
abundant fossil vertebrates (specially mammals)
were recovered (Medeiros & Bergqvist 1999a).
An updated list of the fossils recovered from
Itaboraí Basin is provided in Medeiros &
Bergqvist (1999b).
ABBREVIATIONS
DNPM LE Departamento Nacional da
Produção Mineral, Lote de
Entrada (“entrance lot” – a
temporary number sometimes
given to a group of fossils
before they are cataloged);
Bergqvist L. P., Abrantes É. A. L. & Avilla L. d. S.
324 GEODIVERSITAS • 2004 • 26 (2)
RÉSUMÉ
Les Xenarthra (Mammalia) du bassin de São José de Itaboraí (Paléocène
supérieur, âge itaboraiense), Rio de Janeiro, Brésil.
Nous présentons ici de nouvelles données anatomiques sur les plus anciens
restes de Xenarthra. L’essentiel de ce travail porte sur des analyses morpholo-
giques comparées entre les plaques osseuses et les os postcrâniens trouvés à
Itaboraí (Paléocène supérieur, Brésil). L’ensemble du matériel se compose de
plusieurs plaques osseuses, d’humérus, d’astragales isolés et d’un ulna qui
appartiennent à au moins deux espèces. Les plaques osseuses sont attribuées à
Riostegotherium yanei Oliveira & Bergqvist, 1998, pour lequel une diagnose
révisée est proposée. Les os des ceintures partagent des caractères communs à
ceux de quelques taxons d’édentés. Pourtant beaucoup de ces caractères sont
ambigus et la comparaison avec les Palaeanodonta du début du Tertiaire sou-
ligne des ressemblances dans l’anatomie du membre. Ainsi le Xenarthra
d’Itaboraí pourrait être le groupe-frère de tout ou partie des Palaeanodonta.
MOTS CLÉS
Mammalia,
Xenarthra,
Cingulata,
Riostegotherium,
Astegotheriini,
Palaeanodonta,
tatou,
ostéoderme,
squelette appendiculaire.
MCN-PV Museu de Ciências Naturais,
Fundação Zoobotânica do Rio
Grande do Sul, Porto Alegre,
RS;
MCT-M (ex-DGM) Museu de Ciências da Terra,
Departamento Nacional de
Produção Mineral, Rio de
Janeiro, RJ;
MLP Museo de La Plata, La Plata;
UFRJ-DG-M Departamento de Geologia,
Universidade Federal do Rio de
Janeiro, RJ.
SYSTEMATICS
XENARTHRA Cope, 1889
CINGULATA Illiger, 1811
Family DASYPODIDAE Gray, 1821
Tribe ASTEGOTHERIINI Vizcaíno, 1994
Riostegotherium yanei
Oliveira & Bergqvist, 1998
REFERRED MATERIAL. — Buckler osteoderms (MCN-
PV 1774 [holotype], 1775); movable osteoderms
(MCN-PV 1776, 1778, 1779, MCT 2081-M, MLP
75-XII-26-1, MLP 75-XII-26-2, UFRJ-DG 317-M);
caudal osteoderm (MCN-PV 1777).
DIAGNOSIS (emended from Oliveira & Bergqvist
1998). — Differs from all known early Tertiary aste-
gotheriines in having more than 10 pits in the groove
around the main figure on buckler osteoderms (25 pits
on the holotype, MCN-PV 1774; and 16 pits on
MCN-PV 1775).
DESCRIPTION
The former description of the osteoderms
(Oliveira & Bergqvist 1998) is partly repeated
here, with additions and comments. No articulat-
ed osteoderms were preserved. The buckler osteo-
derms are larger than the movable ones,
sub-rectangular in shape and bear more than
10 pits in the groove around the main figure
(Fig. 1). In comparison with buckler osteoderms
of other astegotheriines, Riostegotherium yanei has
the largest number of foramina for the tribe. The
lateral and medial borders are slightly concave
and smooth. The anterior and posterior borders
are irregular; the posterior has a U shaped
concavity when viewed internally. The anterior
border has a small, weakly defined articulation
zone. The external surface is very punctated and
bears fine irregularities; in combination with the
presence of small depressions, these irregularities
give the surface a slightly wrinkled appearance,
as in Prostegotherium Ameghino, 1902 and
Stegosimpsonia Vizcaíno, 1994. The main figure has
an inverted U-shape, with a subcircular anterior
The Xenarthra (Mammalia) of Itaboraí, Brazil
325
GEODIVERSITAS • 2004 • 26 (2)
AB
a
b
c
FIG. 1. — Riostegotherium yanei Oliveira & Bergqvist, 1998, dorsal view of buckler osteoderms; A, holotype (MCN-PV 1774);
B, MCN-PV 1775; a, main figure; b, peripheral figures; c, pits. Modified from Oliveira & Bergqvist 1998. Scale bar: 1 cm.
outline, and covers almost the whole osteoderm
surface, a very characteristic feature of some
genera of the tribe (Prostegotherium, Astego-
therium Ameghino, 1902, Stegosimpsonia and
Nanoastegotherium Carlini, Vizcaíno & Scillato-
Yané, 1997). It is limited by a shallow groove
with 16 to 25 pits. Although the anterior outline
of the main figure varies slightly, it is never trian-
gular as in some Astegotheriini. Its U-shape is
very similar to the osteoderms of Prostegotherium
and Nanoastegotherium. Two to four small
peripheral figures are present, limited by shallow
radial grooves with at least two pits, as in
Prostegotherium. A well developed central keel is
present on the external surface of the main figure,
noted by Vizcaíno (1994) as one of the synapo-
morphies of Astegotheriini. The internal surface
of the osteoderms is smooth and slightly concave.
No foramina of the piliferous system are observed
in the posterior border.
The typical movable osteoderms vary in shape,
ranging from sub-quadrangular to sub-rectangu-
lar (Fig. 2), which is characteristic of Dasypo-
didae. The anterior articular surface is poorly
developed, and the external surface resembles
that of the buckler osteoderms. The main figure
has a subcircular anterior outline, but with a
reduced number of pits (seven to 12) which is
similar to Prostegotherium and Stegosimpsonia.
The external surface also bears a well developed
central keel. No foramina are present on the pos-
terior border. The internal side of the posterior
border is moderately inclined toward the edge of
the osteoderm.
The shape of the osteoderm MCN-PV 1777
resembles caudal osteoderms of the extant dasy-
podid Dasypus. It differs from typical movable
osteoderms in: articular surface more developed
laterally, smoother external surface, pits more
widely spaced in the groove limiting the main
figure, and a sharp posterior border.
XENARTHRA incertae sedis
REFERRED MATERIAL. — Humeri (MCN-PV 1780,
1781, MCT 2396-M, 2397-M); ulna (MCN-PV
3606); astragali (MCN-PV 1340, 1380, MCT 2394-
M, 2395-M).
DESCRIPTION
Humeri (MCN-PV 1780, 1781, MCT 2396-M,
2397-M)
The humeri can be separated into two mor-
photypes, slightly distinct in shape but clearly
distinct in size.
The larger humeri (MCT 2396-M, 2397-M –
HUM-morph 1) are powerfully built (Figs 3; 4).
The head has an elliptical outline and is more
posteriorly directed than dasypodids, in a way
similar to Manis pentadactyla Linnaeus, 1758.
The articular area for the scapular acromion, a
very characteristic feature of fossorial armadillos,
is shallow. As in other xenarthrans, the shaft is
flat and wider than deep in its proximal half. The
greater tuberosity is higher than the lesser (but
Bergqvist L. P., Abrantes É. A. L. & Avilla L. d. S.
326 GEODIVERSITAS • 2004 • 26 (2)
FIG. 2. — Riostegotherium yanei Oliveira & Bergqvist, 1998,
dorsal view of a movable osteoderm (UFRJ-DG 317-M). Scale
bar: 1 cm.
not projected above the head), but less transverse-
ly expanded. There are well marked impressions
for the infraspinatus and subscapularis muscles on
the greater and lesser tuberosities, respectively.
The tubercle for insertion of the teres major and
latissimus dorsi muscles is more prominent than
in any known Xenarthra, although less than in
Metacheiromys Wortman, 1903, and is placed at
the level of the deltoid tuberosity, as in this
genus. These two muscles are frequently well
developed and fused in burrowers, in which they
have an important digging function (Reed 1951
apud MacPhee 1994; MacPhee 1994). The bici-
pital groove differs from that of any known “eden-
tate” in being wide, shallow and well defined in
the proximal third of the shaft. A similar condi-
tion is present in most pilosans and some derived
cingulates, but the groove is not as defined in the
shaft as in the Itaboraí humeri. The deltopectoral
crest, as in Eurotamandua Storch, 1981, Palaea-
nodon Matthew, 1918 and Metacheiromys is
broad, shelf-like and extends more than halfway
down the shaft, unlike the condition in most
dasypodids. In Priodontes and Cabassous
McMurtrie, 1831, the deltopectoral crest is quite
longer but not as robust as in the Itaboraí
humeri. However, they form a derived mono-
phyletic group (Priodontina sensu Abrantes
2002), and this feature may represent a new
acquisition compared to the pattern observed in
other cingulates. The shelf is proximodistally
concave, but transversely concave only on the dis-
tal half; its outline is very close to that of
Palaeanodon ignavus Matthew, 1918 (Rose 1999:
figs 2, 3). Unlike Dasypodidae, the medial (pec-
toral) border of the crest overhangs the shaft.
This is a distinctive feature of palaeanodonts,
even though the pectoral tuberosity is more
marked. It is distal to the deltoid tuberosity,
which is also well marked, but more distally
located on the lateral border than in Euro-
tamandua and palaeanodonts. As in palaean-
odonts, manids and Eurotamandua, but unlike
xenarthrans, the deltopectoral shelf is anterome-
dially oriented.
The distal extremity is incomplete in both speci-
mens, but its preserved portion suggests that the
distal end was very broad transversely, more than
in most Xenarthra currently known, with the
possible exception of Priodontes. The entepi-
condyle is prominent, a typical feature of “eden-
tates” (Rose 1999). A small portion of the
supinator crest is preserved in specimen MCT
2397-M, showing that it was very prominent,
probably like Eurotamandua and palaeanodonts.
Its shape recalls Palaeonodon ignavus (Rose 1999)
in having a straight border parallel to the shaft.
However, its proximal end forms an angle of
almost 90° with the proximodistal axis of the
shaft, and is straighter than in P. ignavus.
Tamandua Gray, 1825 also presents a straight
supinator crest, but it is considered here a paral-
lelism, as this genus is a derived Myrmeco-
phagidae, which is a derived group within
Xenarthra (Gaudin & Branham 1998).
The other humeri (MCN-PV 1780, 1781 –
HUM-morph 2) are shorter than HUM-
morph 1 (Fig. 5). They are very similar to mor-
photype 1, except in having the tuberosities more
equally developed and transversely expanded, and
the tuberosity of the teres major muscle less
prominent. Compared to morphotype 1, the del-
toid tuberosity is more proximal, less projected,
more rounded and rougher.
Ulna (MCN-PV 3606)
The ulna, like the smaller humerus, is gracile
(Fig. 6). Only the proximal portion is preserved.
The olecranon process, like that of Dasypus
Linnaeus, 1758, is straight and long, with a broad
and medially inflected epiphysis, in the same way
as in some Dasypodidae. The enlargement and
inflection of the olecranon is associated with
great enlargement of the triceps muscles, and per-
haps also the digital flexor (K. Rose pers. comm.),
which powerfully extend the forearm during the
power stroke of digging (Puttick & Jarvis 1977
apud Rose & Emry 1983). The medial surface of
the olecranon is concave and the lateral is flat,
but from the trochlear notch to the distal pre-
served portion, it bears a deep and narrow con-
cavity. The trochlear notch is shallow and very
similar in orientation to that of Proeutatus
Ameghino, 1891. The anconeal process is little
The Xenarthra (Mammalia) of Itaboraí, Brazil
327
GEODIVERSITAS • 2004 • 26 (2)
Bergqvist L. P., Abrantes É. A. L. & Avilla L. d. S.
328 GEODIVERSITAS • 2004 • 26 (2)
A
B
C
A'
B'
C'
Gt Hh
Lt
TmLdm
Ent
Dc
Dt
Dps
Ef
Pt
Pc
Bg
Bg
Sa
FIG. 3. — Xenarthra incertae sedis, left humerus (MCT 2396-M); A, caudal view; B, cranial view; C, proximal view; A’, B’ and C’,
interpretive drawings of the specimen. Abbreviations: Bg, bicipital groove; Dc, deltoid crest; Dps, deltopectoral shelf; Dt, deltoid
tuberosity; Ef, entepicondylar foramen; Ent, entepicondyle; Gt, greater tuberosity; Hh, humeral head; Lt, lesser tuberosity; Pc, pectoral
crest; Pt, pectoral tuberosity; Sa, scapular acromion articulation surface; TmLdm, teres major-latissimus dorsi muscles. Scale bar:
1cm.
The Xenarthra (Mammalia) of Itaboraí, Brazil
329
GEODIVERSITAS • 2004 • 26 (2)
AA'
BB'
Sc
FIG. 4. — Xenarthra incertae sedis, right humerus (MCT 2397-M); A, cranial view; B, caudal view; A’ and B’, interpretive drawings of
the specimen. Abbreviation: Sc, supinator crest. Scale bar: 1 cm.
projected. The radial facet is comparable to
Tolypeutes, but with its lateral portion in a more
distal position than in this taxon. As in other
“edentates”, the ulnar shaft is deeper than wide.
Astragali (MCN-PV 1340, 1380, MCT 2394-M,
2395-M)
There are two different patterns and sizes of
astragali (Fig. 7). The larger (MCN-PV 1340,
MCT 2395-M – AST-morph 1) is morphologi-
cally similar to Utaetus Ameghino, 1902. The
smaller (MCN-PV 1380, MCT 2394-M – AST-
morph 2) in some ways (especially in the shape of
trochlea) resembles Peltephilus Ameghino, 1887.
They were first described by Cifelli (1983),
whose description is partially transcribed here
with additions provided by two new specimens
not seen by this author.
The body of AST-morph 1 is low, transversely
broad and anteroposteriorly shortened, as in sev-
eral cingulates. The trochlea is moderately deep,
with the medial crest shorter and less defined
than the lateral one, as seen in Utaetus. A rugose
fossa, possibly a remnant of the superior astra-
galar foramen, is present posterolaterally on the
trochlea. The lateral wall of the astragalar body is
vertical, and a small fibular shelf protrudes from
the anteroinferior angle of the lateral wall, a very
common feature of dasypodids, but less so in
other “edentates”. The medial wall bears a dis-
crete protuberance for the medial collateral liga-
ment. Both ectal and sustentacular facets are
posterolaterally-anteromedially aligned. The
alignment of ectal and sustentacular facets on the
same axis of rotation was proposed by Szalay &
Schrenk (1998) as a synapomorphy of Xenarthra
and Palaeanodonta. The sustentacular facet
extends from the poorly defined groove for the
digital flexor tendons, which is more distinct
from the trochlea in the specimen MCN-PV
1340. Distally, the sustentacular facet touches the
navicular facet by a medial prolongation. The
ectal facet has the typical triangular shape and
concavity of cingulates. It is separated from the
sustentacular facet by a developed sulcus tali.
As in other cingulates, but differently from
palaeanodonts and pholidotans, the neck is rela-
tively wide, short, shallow and markedly oblique
to the anteroposterior axis of the trochlea. A well
defined crest occupies the dorsodistal extremity
of the neck, which, as in living dasypodids,
would have given rise to the astragalonavicular
ligament. The head is transversely narrow, as in
other dasypodids, and extends posteriorly me-
dially, almost reaching the groove for the digital
flexor tendons, as in Utaetus, for example.
The AST-morph 2, like morphotype 1, lacks the
superior astragalar foramen. The trochlea is
deeper than in the AST-morph 1 and strongly
constricted, as in Peltephilus, Propraopus Ame-
ghino, 1881 and most Glyptodontidae, as well as
Myrmecophaga Linnaeus, 1758. Its lateral border
is condyloid and does not extend anteriorly
beyond the middle of the astragalus. Both medial
and lateral walls of the body are vertical, and the
protuberance for the medial collateral ligament is
poorly developed in specimen MCT 2394-M.
Compared to AST-morph 1, the fibular shelf is
more prominent. The ectal and sustentacular
facets are placed the same way as in other astra-
gali. The latter facet is rounded and does not
contact either the navicular facet or the trochlea
(the digital flexor groove is not distinct from the
posteroinferior margin of the trochlea). The ectal
facet is narrower than morphotype 1 and more
concave.
DISCUSSION
The monophyly of Edentata has been subject of
discussion since the 18th century. However, from
the 20th century on, for most authors Edentata
comprises Xenarthra, Pholidota, and Palaeano-
donta (e.g., Novacek & Wyss 1986; Patterson et
al. 1992; Gaudin & Wible 1999). Morphological
studies have suggested alternative relationships
within Edentata: Palaeanodonta closer to
Pholidota (e.g., Emry 1970; Shoshani et al.
1997) and Palaeanodonta closer to Xenarthra
(e.g., Simpson 1931; Patterson et al. 1992;
Gaudin & Wible 1999). The clade Xenarthra-
Pholidota is an old idea, resurrected by Novacek
& Wyss (1986), contrary to Simpson’s (1931,
Bergqvist L. P., Abrantes É. A. L. & Avilla L. d. S.
330 GEODIVERSITAS • 2004 • 26 (2)
1945) conclusions that it lacks any definite pale-
ontological support.
Recent molecular cladistic analyses (Waddell
et al. 1999; Delsuc et al. 2001) also failed to sup-
port a Xenarthra-Pholidota clade, but proposed
different phylogenetic relationships for
Pholidota: closer to Carnivora, and included in a
higher-taxon called Ferae (DeJong 1982;
Waddell et al. 1999), and sister-group of Cetar-
tiodactyla (Cetacea + Artiodactyla; Delsuc et al.
2001). A detailed taxonomic history of Edentata
can be found in Szalay & Schrenk (1998).
The record of mammalian osteoderms in Itaboraí
Basin is a clear indication of the presence of
Xenarthra-Cingulata, as these bones are exclusive
to the group (the osteoderms of some Tardigrada
are isolated bony nodules without articular sur-
faces). Cingulata is considered the most primitive
Xenarthra (e.g., Patterson et al. 1992), with the
Astegotheriini as its oldest member (Vizcaíno
1994; Oliveira & Bergqvist 1998). Until the first
recovered cingulates from Itaboraí Basin
(Itaboraian), the oldest record of astegotheriines
was from the Casamayoran of Argentina, approx-
imately 5 My younger (Scillato-Yané 1976).
The forelimb bones indicate that their owners
were proficient diggers. The prominent tubercle
for the teres major and latissimus dorsi muscles
and their insertion marks on the humeral shaft
suggest that these muscles were well developed
and fused (Reed 1951; MacPhee 1994), which
corroborates the first argument. Moreover, these
humeral features and the well developed olecra-
non of the ulna reinforce the proposition of well
adapted burrowers in the Itaboraí fossil record.
The allocation of the forelimb bones to the
Xenarthra would be easy, as they present an
expected xenarthran general morphology, and no
other “edentates” than Xenarthra are found in
South America. These bones show closer similari-
ties with Priodontini (sensu McKenna & Bell
1997), but this group is a very derived dasypodid
lineage (Engelmann 1985; Abrantes 2002) that
appeared recently in the fossil record (middle
Pleistocene; McKenna & Bell 1997). So, the sim-
ilarities observed between the Itaboraí Xenarthra
and Priodontini humeri are assumed here as par-
allelism, as the probability of homoplasy increases
with time since divergence from a common
ancestor (Simmons & Geisler 1998).
On the other hand, several features of the
Itaboraí forelimb bones are derived attributes
shared by other “edentates”, mainly Palaeano-
donta. This similarity hinges on the medially
turned elongation of the pectoral crest (Szalay &
Schrenk 1998) and the very prominent supinator
crest (Rose & Emry 1993; Rose 1999). This lat-
ter feature was indicated as diagnostic for
Metacheiromys (Palaeanodonta) by Simpson
(1931: 342) and it was used to relate Eurota-
mandua and Palaeanodonta (Rose 1999). So we
pose the classic question: do these similarities
indicate convergences or phylogenetical signals?
For Hennig (1966), in cases that it is impossible
The Xenarthra (Mammalia) of Itaboraí, Brazil
331
GEODIVERSITAS • 2004 • 26 (2)
AB
C
Gt
Dc
Dt
Bg
Lt
TmLdm
Bg
Sa
FIG. 5. — Xenarthra incertae sedis, right humerus (MCN-PV
1781); A, cranial view; B, caudal view; Cproximal view.
Abbreviations: Bg, bicipital groove; Dc, deltoid crest; Dt, deltoid
tuberosity; Gt, greater tuberosity; Lt, lesser tuberosity;
Sa, scapular acromion articulation surface; TmLdm, teres
major-latissimus dorsi muscles. Scale bar: 1 cm.
to decide whether the common character is a
synapomorphy, parallelism, homology or even a
convergence, kinship must be assumed a priori
(auxiliary principle).
A same problem was raised by Rose (1999) when
he analyzed forelimb bones of Eurotamandua. In
spite of the biogeographical problem, this “eden-
tate” from the lower Eocene of Germany has
always been viewed as related to the Xenarthra
(Rose & Emry 1993; Storch 1981; Storch &
Habersetzer 1991). However, after Shoshani et
al. (1997), who argued that Eurotamandua is a
pholidotan and not a xenarthran, several studies
were conducted in order to place Eurotamandua
within Eutheria (Gaudin & Branham 1998;
Szalay & Schrenk 1998; Rose 1999). In fact, only
Shoshani et al. (1997) related Eurotamandua to
Pholidota. What appears to be a consensus since
Rose (1999) is that Eurotamandua shows many
more detailed similarities to primitive Palaea-
nodonta than to South American anteaters or any
other xenarthrans. Moreover, T. Gaudin &
K. Rose (pers. comm.) suggested that Eurota-
mandua belongs or is related to the Palaeano-
donta, based on the homology of the features of
forelimb elements (including humeri and ulna).
Except for the shallow articular area for the
scapular acromion, the Itaboraí humeri do not
share other “exclusive” similarity to Xenarthra.
Additionally, the ulnar morphology appears to be
much more informative on a functional perspec-
tive (see Vizcaíno et al. 1999), and as noted by
MacPhee (1994), all fossorial mammals have sim-
ilar ulnae (with well developed olecranon). Thus,
these humeri from the Paleocene of Itaboraí share
several similarities with primitive palaeanodonts
than to South American anteaters or any other
xenarthrans.
It should be emphasized that the resemblances
observed here between the humeri from the
Paleocene of Itaboraí and those of palaeanodonts
are suggestive of a closer affinity between
xenarthrans and paleonodonts. This hypothesis
meets the conclusion of several authors that had
recently suggested that palaeanodonts are more
closely related to xenarthrans than to pholidotans
(Storch & Habersetzer 1991; Patterson et al.
1992; Szalay & Schrenk 1994, 1998; Gaudin &
Branham 1998; Gaudin & Wible 1999).
Additionally, the humeri described here from the
Itaboraian of Brazil are, together with the osteo-
derms of Riostegotherium, one of the most ancient
records of South American “edentates”.
The value of the astragalus in mammalian taxon-
omy has long been recognized (at least since
Matthew [1909]). Simpson (1931) noticed the
difficulty of designating a general xenarthran type
of astragalus, for in the ground sloths and tree-
sloths this bone has extreme divergent specializa-
tions. Szalay & Schrenk (1998), reviewing
Eurotamandua and Edentata, presented the fol-
lowing features as tentative apomorphies of the
tarsus of the higher clade Xenarthra: transversely
broad astragalar body with greater arc laterally,
and calcaneal facets aligned virtually on the same
axis of rotation. Both characters are present in the
Itaboraí astragali, supporting their placement is
this clade, but their assignment to one of the
xenarthran higher taxa was made on a tentative
basis.
Bergqvist L. P., Abrantes É. A. L. & Avilla L. d. S.
332 GEODIVERSITAS • 2004 • 26 (2)
A B
Op
Ap
Tn
Rf
FIG. 6. — Xenarthra incertae sedis, left ulna (MCN-PV 3606);
A, cranial view; B, caudal view. Abbreviations: Ap, anconeal
process; Op, olecranon process; Rf, radial facet; Tn, trochlear
notch. Scale bar: 1 cm.
The Xenarthra (Mammalia) of Itaboraí, Brazil
333
GEODIVERSITAS • 2004 • 26 (2)
C
D
A
B
C'
D'
A'
B'
Fs
Mc
Mcl
N
Nc
Fs
Rf
Dfg
Sf
Nf
Ef
FIG. 7. — Xenarthra incertae sedis, rigth and left astragali; A, B, MCT 2395-M; A, dorsal view; B, plantar view; A’ and B’, interpretive
drawings of the specimen; C, D, MCT 2394-M; C, dorsal view; D, plantar view; C’ and D’, interpretive drawings of the specimen.
Abbreviations: Dfg, digital flexor tendon groove; Ef, ectal facet; Fs, fibular shelf; Mc, medial crest; Mcl, protuberance for the medial
collateral ligament; N, neck; Nc; neck crest; Nf, navicular facet; Rf, rugose fossa; Sf, sustentacular facet. Scale bar: 1 cm.
The two astragali previously studied by Cifelli
(1983) (DNPM LE 449A and 449B – now MCT
2394-M and 2395-M, respectively) were assigned
to the Cingulata based on the absence of derived
pilosan characters rather than on the presence of
distinctively cingulate features. The author
assigned the smaller astragalus to the family
Glyptodontidae, despite of the absence of two
distinctive and derived features of the glyptodont
astragalus – a flattened sustentacular facet and a
shorter neck. His assumption was supported by
the presence of an anterior constriction of the
trochlea and the morphology of its medial crest-
condyloid and shorter than half of the astragalar
length. Bergqvist & Oliveira (1998) pointed out
that the crests of the tibial trochlea of
Hoplophorus, which is the most condyloid among
glyptodonts, are less condyloid than the speci-
mens MCN-PV 1340 and MCT 2394-M (AST-
morph 2), and the other features, although
present in the glyptodontid Propaleohoplophorus
Ameghino, 1891, are also seen in the dasypodid
Propraopus and Cabassous. The anterior constric-
tion of the tibial trochlea is also shared by
Hapalops Ameghino, 1887 and Myrmecophaga
(though not well developed), showing that this
feature is widespread among the Xenarthra, and
cannot be used as distinctive for Glyptodontidae.
The absence of a diagnostic glyptodont astragalar
feature in AST-morph 2, and of glyptodont
osteoderms at Itaboraí, weakens their assignment
to the family Glyptodontidae. The presence of
dasypodid osteoderms in the basin and astragalar
features shared by members of the families
Dasypodidae and Peltephilidae (Bergqvist &
Oliveira 1998) is suggestive of a Dasypodoidea
affinity, but no conclusive placement can be
made until the derived features of the astragalus
of Dasypodoidea are defined to the exclusion of
Glyptodontoidea and vice-versa. A phylogenetic
analysis within Cingulata is urged.
The larger astragali possess most of the characters
listed by Simpson (1931) as the primitive type of
xenarthran astragalus: 1) body relatively broad
and short; 2) groove of trochlea broad, deep,
oblique; its lowest part relatively medial; 3) later-
al crest higher, sharper and longer; 4) malleolar
facets nearly vertical; 5) neck distinct, constricted,
but short; and 6) articular surface of head convex
from side to side and extending up almost to the
body of the astragalus medially. The only excep-
tion is the presence of a superior astragalar fora-
men, in a vestigial stage, as in Priodontes maximus
F. Cuvier, 1827.
Simpson (1931) and Cifelli (1983) mentioned
that dasypodid ankle morphology shows the
primitive pattern of Cingulata. The body of
AST-morph 2 is narrower and longer than in
Dasypodidae, the trochlea is less oblique and the
neck longer and more slender, resembling the
pattern seen in Palaeanodon, which Simpson
(1931) considered more primitive and closer to
the most generalized insectivore type. In some
ways, it also resembles the pattern seen in
Protoungulatum, which would be predicted for a
primitive eutherian morphotype (Cifelli 1983).
As mentioned above, the astragali also present
features broadly distributed among xenarthrans
and probably primitive for the group. Even if the
primitive pattern cannot be confidently recog-
nized without a phylogenetic analysis, the mor-
phology of the smaller astragali seems to be the
primitive pattern for Xenarthra, despite of the
absence of a vestige of the astragalar foramen.
Although not discussed here, the rugose fossa
present in the trochlea of AST-morph 1 is lateral-
ly placed, and might be related to ligaments,
rather than the superior opening of the astragalar
channel. Ameghino (1905, 1906) observed that
in Priodontes maximus the superior astragalar
foramen is medially placed, in a plantar tendi-
nous groove, as is also seen in Itaboraí ungulates
(Cifelli 1983; Bergqvist 1996). The lateral fossae
present in several armadillos may be related to
ligaments, not to the foramen. If this is true, then
both morphotypes may be considered derived
with respect to the superior astragalar foramen.
Cifelli (1983) noted the near absence of osteo-
derms in the Itaboraí Basin compared to the
number of postcranial remains, to the good
preservation of the bones and to the abundance
of herpetofauna. Although more osteoderms have
been recovered since then, they are still very few
compared to the common ratio of osteoderms/
Bergqvist L. P., Abrantes É. A. L. & Avilla L. d. S.
334 GEODIVERSITAS • 2004 • 26 (2)
bones in other sites (e.g., Bergqvist 1989). The
senior author of this paper has thoroughly exam-
ined all vertebrate fossil specimens collected in
Itaboraí, numbered or non-numbered, and only
the osteoderms listed here were found. We agree
with Cifelli (1983) that it is unlikely that this
small number is an artifact of collecting, and that
is likely that these primitive cingulates possessed
fewer osteoderms than later cingulates.
AST-morph 2 is the size of the astragalus of
Tolypeutes Illiger, 1811 whose osteoderms are
about the size of the osteoderms recovered in the
Itaboraí Basin. The astragalus of Stegotherium
Ameghino, 1887 as described by Scott (1903),
resembles the AST-morph 2. The size and resem-
blance suggest that AST-morph 2 might belong
to Riostegotherium. However, no confident associ-
ation can be supposed until a detailed evaluation
of the correlation between size of the osteoderms
and astragalus among armadillos is done or an
associated skeleton is discovered.
Rose & Emry (1993) proposed that Xenarthra
was a relatively ancient eutherian group, long iso-
lated in South America and with no obvious close
relationship to any other eutherian. However, the
evidence presented here by the limb bones sug-
gests the possibility that the Itaboraí Basin cingu-
lates might represent part of an ancestral stock
of a monophyletic group uniting Xenarthra
and Palaeanodonta (including Eurotamandua).
Despite of the lack of geological evidence, the
idea of a landbridge connection during the Late
Cretaceous/early Paleocene between Gondwana
and Laurasia cannot be discounted, as biogeo-
graphical patterns presented by some terrestrial
vertebrates (see Schoch & Lucas 1985; Muizon
& Marshall 1992; Salgado & Calvo 1997) and
plants (see Wang 1978; Hay et al. 1999) support
the presence of such a connection.
CONCLUSION
The validity of Riostegotherium as the oldest
cingulate is confirmed.
The morphology of forelimb bones and their
muscle scars indicate that well adapted burrowers
were part of the mammalian biota of the Itabo-
raian SALMA of Brazil.
The mosaic of derived features presented by the
appendicular bones prevents us from assigning
them to any particular known Xenarthra, princi-
pally to Astegotheriini. However, their shared
similarities to palaeanodonts add support to the
proposal that Palaeanodonta may have been
ancestral to, or is the sister-taxon of Xenarthra.
Acknowledgements
We are indebted to Ana Maria Ribeiro (Fundação
Zoobotânica do Rio Grande do Sul) for the
photographs of the specimens deposited in this
collection, and to Marcelo Adorna Fernandes
(Programa de Pós-Graduação em Geologia/
UFRJ) and Luis Antônio Alves Costa (Setor de
Entomologia/Museu Nacional) for the drawings.
Luiz Antônio Sampaio Ferro made the figure edi-
tions. We are especially grateful to Dr. Kenneth
Rose (Johns Hopkins University), Dr. Sergio
Fabian Vizcaíno (Museo de La Plata), Dr. Gerardo
De Iuliis and Dr. Gregory McDonald for review-
ing an early version of the manuscript and for
helpful suggestions. This research was partially
supported by FAPERJ, CAPES and UFRJ. This
is a contribution to Instituto Virtual de Paleon-
tologia/FAPERJ.
REFERENCES
ABRANTES E. A. L. 2002. — Revisão filogenética dos
Dasypodidae (Mammalia: Cingulata). Dissertação
de Mestrado, Programa de Pós-Graduação em
Geologia/Universidade Federal do Rio de Janeiro,
Brazil, 73 p.
AMEGHINO F. 1905. — La perforación astragaliana em
los mamíferos no es un carácter originariamente
primitivo, in TORCELLI A. J. (ed.), Obras completas
y correspondencia científica de Fiorentino
Ameghino. Volume XV. Anales del Museo Nacional
de Buenos Aires 3 (IV): 349-460.
AMEGHINO F. 1906. — La perforación astragaliana em
Priodontes, Canis (Chrysocyon) y Typotherium, in
TORCELLI A. J. (ed.), Obras completas y correspon-
dencia científica de Fiorentino Ameghino. Volume
XV. Anales del Museo Nacional de Buenos Aires
3 (VI): 1-19.
The Xenarthra (Mammalia) of Itaboraí, Brazil
335
GEODIVERSITAS • 2004 • 26 (2)
BERGQVIST L. P. 1989. — Mamíferos pleistocênicos do
estado da Paraíba, Brasil, depositados no Museu
Nacional, Rio de Janeiro. Dissertação de Mestrado,
Programa de Pós-Graduação em Zoologia/Museu
Nacional/Universidade Fedeeral do Rio de Janeiro,
Brazil, 174 p.
BERGQVIST L. P. 1996. — Reassociação do pós-crânio às
espécies de ungulados da bacia de S. J. de Itaboraí
(Paleoceno), Estado do Rio de Janeiro, e filogenia dos
“Condylarthra” e ungulados sul-americanos com base
no pós-crânio. Tese de doutorado, Programa de Pós-
Graduação em Geociências/Universidade Federal
do Rio Grande do Sul, Brazil, 407 p.
BERGQVIST L. P. & OLIVEIRA E. V. 1995. — Novo
material pós-craniano de Cingulata (Mammalia-
Xenarthra) do Paleoceno médio (Itaboraiense) do
Brasil. Atas do XIV Congresso Brasileiro de Paleon-
tologia, Uberaba: 16, 17.
BERGQVIST L. P. & OLIVEIRA E. V. 1998. —
Comments on the xenarthran astragali from the
Itaboraí Basin (middle Paleocene) of Rio de Janeiro,
Brazil. Acta Geológica Lilloana 18 (1): 153, 154.
CIFELLI R. L. 1983. — Eutherian tarsals from the Late
Paleocene of Brazil. American Museum Novitates
2761: 1-31.
DEJONG W. 1982. — Eye lens proteins and vertebrate
phylogeny, in GOODMAN M. (ed.), Macromolecular
Sequences in Systematics and Evolutionary Biology.
Plenum Press, New York: 75-114.
DELSUC F., CATZEFLIS F. M., STANHOPE M. J. &
DOUZERY E. J. P. 2001. — The evolution of armadil-
los, anteaters and sloths depicted by nuclear and mito-
chondrial phylogenies: implications for the status for
the enigmatic fossils Eurotamandua. Proceedings of the
Royal Society of London B 268: 1605-1615.
EMRY R. J. 1970. — A North American Oligocene
pangolin and other additions to the Pholidota.
Bulletin of the American Museum of Natural History
142 (6): 457-510.
ENGELMANN G. F. 1985. — The phylogeny of the
Xenarthra, in MONTGOMERY G. G. (ed.), The
Evolution and Ecology of Armadillos, Sloths and
Vermilinguas. Smithsonian Institution Press,
Washington DC, 451 p.
GAUDIN T. J. & BRANHAM D. G. 1998. — The phy-
logeny of the Myrmecophagidae (Mammalia,
Xenarthra, Vermilingua) and the relationship of
Eurotamadua to the Vermilingua. Journal of
Mammalian Evolution 5 (3): 237-265.
GAUDIN T. J. & WIBLE J. R. 1999. — The entotym-
panic of pangolins and the phylogeny of the
Pholidota (Mammalia). Journal of Mammalian
Evolution 6 (1): 39-65.
HAY W. W., DECONTO R. M., WOLD N., WILSON K.
M., VOLGT S., SCHULTZ M., WOLD A. R., DULLO
W., RONOV A. B., BALUKHOVSKY A. N. & SODING
E. 1999. — Alternative global Cretaceous paleo-
geography. Geological Society of America Special
Paper 332: 1-47.
HENNIG W. 1966. — Phylogenetic Systematics.
University of Illinois Press, Urbana, 263 p.
MCKENNA M. C. & BELL S. K. 1997. — Xenarthra, in
MCKENNA M. C. & BELL S. K. (eds), Classification
of Mammals Above the Species Level. Columbia
University Press, New York: 82-102.
MACPHEE R. D. E. 1994. — Morphology, adapta-
tions, and relationships of Plesiorycteropus, and a
diagnosis of a new order of mammals. Bulletin of
the American Museum of Natural History 220:
1-214.
MATTHEW W. D. 1909. — The Carnivora and
Insectivora of the Bridger Basin, Middle Eocene.
Memoirs of the American Museum of Natural History
9: 291-567.
MEDEIROS R. A. & BERGQVIST L. P. 1999a. —
O Paleogeno das bacias continentais do Brasil. Anales
del Servicio Geologico Minero Argentino 33: 79-87.
MEDEIROS R. A. & BERGQVIST L. P. 1999b. —
Paleocene of the São José de Itaboraí Basin, Rio de
Janeiro, Brazil: lithostratigraphy and biostratigra-
phy. Acta Geologica Leopoldensia 22 (48): 3-22.
MUIZON C. DE & MARSHALL L. G. 1992. —
Alcidedorbignya inopinata (Mammalia: Pantodonta)
from the early Paleocene of Bolivia. Phylogenetic
and paleobiogeographic implications. Journal of
Paleontology 66 (3): 499-520.
NOVACEK M. J. & WYSS A. R. 1986. — High level
relationships of the recent eutherian orders: mor-
phological evidence. Cladistics 2: 257-287.
OLIVEIRA E. V. & BERGQVIST L. P. 1998. — A new
Paleocene armadillo (Mammalia, Dasypodoidea)
from the Itaboraí Basin, Brazil. Asociación
Paleontologica Argentina, Publicacion Especial 5:
35-40.
PAULA-COUTO C. 1949. — Novas observações sobre
paleontologia e geologia do depósito calcário de São
José de Itaboraí. DGM/DNPM. Notas Preliminares
e Estudos 49: 1-13.
PATTERSON B., SEGALL W., TURNBULL W. D. &
GAUDIN T. J. 1992. — The ear region in Xenarthra
(= Edentata: Mammalia). Part. II. Pilosa (sloths,
anteaters), palaeanodonts, and a miscellany.
Fieldiana (Geology) 1438: 1-79.
PUTTICK G. M. & JARVIS J. U. M. 1977. — The
functional anatomy of the neck and forelimbs of
the cape golden mole, Chrysochloris asiatica
(Lipotypla: Chrysochloriidae). Zoologica Africana
12: 445-458.
REED C. A. 1951. — Locomotion and appendicular
anatomy in three soricoid insectivores. American
Midland Naturalist 45 (3): 513-671.
ROSE K. D. 1999. — Eurotamandua and Palaeano-
donta: convergent or related? Paläontologische
Zeitschrift 73 (3/4): 395-401.
ROSE K. D. & EMRY R. J. 1983. — Extraordinary fos-
sorial adaptations in the Oligocene palaeanodonts
Epoicotherium and Xenocranium (Mammalia).
Journal of Morphology 175: 33-56.
Bergqvist L. P., Abrantes É. A. L. & Avilla L. d. S.
336 GEODIVERSITAS • 2004 • 26 (2)
ROSE K. D. & EMRY R. J. 1993. — Relationship of
Xenarthra, Pholidota, and Fossil “Edentates”: the
morphological evidence, in SZALAY F. S., NOVACEK
M. J. & MCKENNA M. C. (eds), Mammal
Phylogeny: Placentals. Springer-Verlag, New York:
81-102.
SALGADO L. & CALVO J. O. 1997. — Evolution of
titanosaurid sauropods. II: the cranial evidence.
Ameghiniana 34: 33-48.
SCHOCH R. & LUCAS S. 1985. — The phylogeny and
classification of the Dinocerata (Mammalia,
Eutheria), in LUCAS S. G. & MATEER N. J. (eds),
Studies of chinese fossil vertebrates. Bulletin of the
Geological Institutions of the University of Uppsala
(new series) 11: 31-58.
SCILLATO-YANÉ G. J. 1976. — Sobre un Dasypodidae
(Mammalia, Xenarthra) de edad Riochiquense
(Paleoceno superior) de Itaboraí (Brasil). Anais da
Academia de Ciências 48 (3): 527-530.
SCOTT W. B. 1903. — Mammalia of the Santa Cruz
Beds. Part I. Edentata, I. Dasypoda. The University
of Princeton, Princeton, 106 p.
SHOSHANI J., MCKENNA M. C., ROSE K. D. &
EMRY R. J. 1997. — Eurotamandua is a pholidotan
not a xenarthran. Journal of Vertebrate Paleon-
tology 17: 76A.
SIMMONS N. B. & GEISLER J. H. 1998. — Phylo-
genetic relationships of Icaronycteris, Archaeo-
nycteris, Hassianycteris, and Palaechiropteryx to
extant bat lineages, with comments of the evolution
of echolocation and foraging strategies in
Microchiroptera. Bulletin of the American Museum
of Natural History 235: 1-182.
SIMPSON G. G. 1931. — Metacheiromys and the
Edentata. Bulletin of the American Museum of
Natural History 59: 295-381.
SIMPSON G. G. 1945. — The principles of classifica-
tion and a classification of mammals. Bulletin of the
American Museum of Natural History 85: 1-350.
STORCH G. 1981. — Eurotamandua joresi, ein
Myrmecophagide aus dem Eozan der “Grube
Messel” bei Darmstadt (Mammalia, Xenarthra).
Senckenbergiana lethaea 61: 247-289.
STORCH G. & HABERSETZER J. 1991. — Rück-
verlagerte Choanen und akzessorische Bulla-
tympanica bei rezenten Vermilingua und
Eurotamandua aus dem Eozan von Messel
(Mammalia: Xenarthra). Zeitschrift für Säuge-
tierkunde 56: 257-271.
SZALAY F. S. & SCHRENK F. 1994. — Middle Eocene
Eurotamandua and the early differentiation of the
Edentata. Journal of Vertebrate Paleontology 14:
48A.
SZALAY F. S. & SCHRENK F. 1998. — The middle
Eocene Eurotamandua and a Darwinian phyloge-
netic analysis of “Edentates”. Darmstadter Beitrage
zur Naturgeschichte 7: 97-186.
VIZCAÍNO S. F. 1994. — Sistematica y anatomia de los
Astegotheriini Ameghino, 1906 (Nuevo rango)
(Xenarthra, Dasypodidae, Dasypodinae).
Ameghiniana 31 (1): 3-13.
VIZCAÍNO S. F., FARIÑA R. A. & MAZZETTA G. V.
1999. — Ulnar dimensions and fossoriality in
armadillos. Acta Theriologica 44 (3): 309-320.
WADDELL P. J., OKADA N. & HASEGAWA M. 1999. —
Towards resolving the interordinal relationships of
placental mammals. Systematic Biology 48 (1): 1-5.
WANG Z. 1978. — Cretaceous Charophytes from the
Yangze-Han river basin with a note on the classifi-
cation of Porocharaceae and Characeae. Memoirs of
the Nanjing Institute of Geology and Paleontology 9:
61-88.
Submitted on 16 January 2003;
accepted on 21 January 2004.
The Xenarthra (Mammalia) of Itaboraí, Brazil
337
GEODIVERSITAS • 2004 • 26 (2)
